Skip to main content

Full text of "The University of Kansas science bulletin"

See other formats 


we 


> 
cm 
\4 
’ 
* 
. 
id 
* 
7 


EE 


S 


an 


~ 
a 


ee. ve - 


ee 


me) 


ik 


ae aa a : 
ver. Ware: ar a 
ny 4 ne is 7 i M 7 > = 
ay) 7 a : ~ | ; 
ao a) 7 i ; > Ve i . = 
ust vey 
wah 7.4 a 
’ i By. "I vy 7 , ; i 
7 ta po oe 
eE ¢ -“ ‘ nd i 
7 or . ie { : &, 7 4p 4 p oa 
ul t 7 i 4 a3 aT 
’ i " 7 7 7 : 
- 7 ‘ a lL ran 1 
a) ay 7 "7 ' 7 
j a - J 7 = ¢, ; “a 
- - a _ 7 ; _ 
ie . : Peo — , : a 
“U 7 7 S an Sy an, ’ TT 
le - ae woya _ 
rn uy : ' ; un 
af 7 _ 
* 7 ae a” > ; 
- - § af 7 i . 
_ : ca i | 6 wm 
oa ; 5 ey 2 
i 7 . ; i 
he ae - Mae wy rs 
: ° we : 
‘ = \ 
- 7 “ 
7 ; 
7 
- 
iy 
; 7] 
v . arnt 
; i ow ay 
is 
ane) 
1 
: j - ’ 
a ; \* @ ' ‘ 
; +) 
; . 7 i" i 
- - - : 
; ; : - - : 
a ; 
; fl { ye 
A r Oi 
- ; o) 
1 : a . 
ay : 
- if . 
; > FA : 
. = 7 : 7 7 a - 
i] : H a 
: ¥ : ! 
j ; i i 7 7 : : 
f 
: 7 - : i os 
a - ; 2 ‘ 7 a 
7 ! _ . a) oe 
_ 1" 7 . : ? _ 
: . : i 
- ss 7 ' “a ; ) en in . 
7 D a is yer re. 
aS : 7 a 7 - 4 - 
\ I ’ a i ™) i 
ae ; eo na ve - 
; YT ae 3 a 
: a) 4 %@ i a ees is nn 1 ce a) oh 7 
_ ‘ 1 ae d _ i v4 , m4 
i v 1 v Tite oy "] 7 iyi 7 A : i - a 
‘ ¥ 0 a i “N a % 7 _ _ oe oo 1¢ A 7 oF 
‘ rr a oo, Dee ot oe : 
- iD ; it Wy ' hi : : 7 - : 
Gul ; By ss - - a i : - ae : 
nak A! " s i 7 i ; 
oo _' ‘ * a. = - n iM a ‘ 7 : , int i U 


er 


. 


, 


a 


igleed 


ae 
1 if 


tw | Ait 


en a 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


AN ECOLOGICAL STUDY OF THE 
VEGETATION OF THE BIG SALT 
MARSH, STAFFORD COUNTY, 
KANSAS 


By 
Irwin A. Ungar 


VoL. XLVI Paces | To 99 May 1, 1965 No. 1 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces | To 99 May 1, 1965 No. 1 


An Ecological Study Of The Vegetation Of The 
Big Salt Marsh, Stafford County, Kansas 
By 


Irwin A. UNGAR* 


»  Axsrract. This paper contains a qualitative and quantitative description of 
the vegetation of the Big Salt Marsh, Stafford County, Kansas. An analysis of 
the soil and water is included and correlations are shown between the edaphic 
conditions and plant distribution. 

There is also a discussion of the salt tolerance of marsh species and their suc- 
cessional and climax relations. 

Seed germination studies with Bromus japonicus, Haplopappus phyllocephalus 
subsp. annuus and Suaeda depressa indicate their salt tolerances. 


INTRODUCTION 


During the 1958, 1959, and 1960 growing seasons an ecological study was 
‘carried out on the Big Salt Marsh in the northeastern corner of Stafford 
County, Kansas. The study was made to determine the types of vegetation 
‘and the factors which affected plant distribution in this area. 

The only investigation on the distribution of salt marsh vegetation in 
Kansas was by Schaffner (1898), who studied some northern Kansas salt 
‘marshes. Hitchcock (1898) made a cursory list of nine species he found on 
the Big Salt Marsh. 

. In these earlier works no quantitative methods were used in studying 
the vegetation and only a very small amount of data were presented on the 
“salt tolerance of the species observed. The present work submits quantitative 
data and attempts to fill other gaps which are present in our knowledge of 

the vegetation and salinity tolerance of species in inland salt marshes. 

No attempt has been made in this paper to review the entire literature on 
coastal halophytic vegetation and the physiology of halophytes in general. 
, The reviews of Uphof (1941), Magistad (1945), Hayward and Wadleigh 

, (1949), Hayward and Bernstein (1958) and Chapman (1960) provide a 


*Assistant Professor of Biological Sciences, Quincy College, Quincy, Illinois. 


2 Tue Universiry ScrENCE BULLETIN 


fairly complete bibliography and discussions of the literature on these sub- 
jects. 


ACKNOWLEDGEMENTS 


The author wishes to express his gratitude to Dr. W. H. Horr for his 
assistance throughout this research. Thanks go to Dr. R. H. Thompson for 
identification of Charophyceae, Dr. A. A. Cridland for moss determinations, 
Dr. W. H. Horr and Dr. R. L. McGregor for checking determinations of 
the more difficult groups of vascular plants and Dr. B. L. Wagenknecht for 
sending me a list of species which he had collected in this area. Thanks also 
to Dr. E. L. Richards and Ana C. Ungar for assistance in collecting. 

Special appreciation goes to Dr. W. H. Horr, Dr. R. W. Lichtwardt, Dr. 
R. L. McGregor, Dr. R. W. Baxter, Dr. R. H. Thompson and Dr. G. Byers 
for reading and editing the manuscript. 

This paper is part of a thesis submitted in partial fulfillment of the re- 
quirements for the degree of Doctor of Philosophy at the University of 


Kansas, Lawrence, Kansas. The study was aided by a grant from the Sigma 
Xi-RESA Research Fund. / 
LOCATION 


According to Schoewe (1949) the Big Salt Marsh is located in the Great 
Bend Lowland of the Central Lowland physiographic province. The marsh, 
situated in the northeastern part of Stafford County, Kansas (T.21S., R.11W., 
and the northern row of sections of T.22S., R.11W.), covers parts of 12 sec- 
tions (Fig. 1) and has an elevation of 1,737 feet above sea level. The total 
relief in this general area is less than 200 feet and on the marsh it is 40 feet. 
Sand-dune topography is characteristic of this general area except for rela- 
tively small sites such as the salt marsh. 


ORIGIN OF SALT 


Latta (1950) stated that waters in the Pleistocene, Meade Formation, in 
the vicinity of the Big Salt Marsh, were highly mineralized and contained 
high concentrations of chloride. He believed that the source of this water 
is in the Permian and Cretaceous rocks which are known to be highly min- 
eralized. Hay (1891) estimated that the thickness of the salt masses ranged 
from one inch to one hundred feet. In a later study Bass (1926) estimated 
from oil well logs that the bed of rock salt was 150-200 feet thick and lay 
1,000 feet below the surface. Kulstad (1959), using data from well samples, 
drillers logs and electric logs, calculated that the salt bearing rocks in this 
area were 300-350 feet thick and contained 60-80°% salt. 

A high bedrock ridge trending approximately perpendicular to the direc- 
tion of ground water movement, according to Latta (1950), causes the 
highly mineralized waters at the base of the Meade Formation to be forced 


Tue VEGETATION Or Tue Bic Sart Marsu 3 


upward. These waters are discharged at the surface of the marsh where the 
salts are further concentrated by evaporation. A report of the ion content 
of the ground water by Latta (1950) at three locations in the vicinity of the 
Big Salt Marsh is cited in Table 1. 


Taste 1. Ground water relations at three locations in the Meade Formation in 

the vicinity of the Big Salt Marsh. Sample A was at T.21S., RI1W., SW cor. sec. 

24, 160-165 ft., sample B in T.21S., R.11W., NW cor. sec. 25, 207-212 ft. and 

sample C at T.21S., R.11W., NE sec. 27, 32 ft. All reports are recorded in P.P.M. 
Water analysis taken from Latta (1950). 


Dissolved Na and 
Site solids Fe Ca Mg K HCO; SO, Cl F NOs S102 COs 
eee 24,167 96". 4335 o> 8,377. 372 1,803 13,050 6 Soh kee 305 
Bae: 44,660 2.1 Hie Bik AGTH uly IS ees OD were 254 
C= 2,150 lll gts 10 664 382 116 SO5m) 10) 142) 313 
CLIMATE 


Weather Bureau reports at Hudson, Kansas, eight miles southwest of 
the Big Salt Marsh, give the annual precipitation in this area from 1931-1959 
as an average of 23.99 inches, from a low of 14.17 inches in 1936 to a high of 
39.61 inches in 1957. The average monthly precipitation for 1931-1959 was 
as follows: January, 0.62; February, 0.88; March, 1.31; April, 2.27; May, 
4:11; June, 3.80; July, 3.19; August, 2.55; September, 2.14; October, 1.62; 
November, 0.78: December, 0.72. According to Park and Dennis (1947) the 
average growing season in this area lasts 172 days, April 28 through October 
20, and from a calculation of the data above it was found that 72°% of the 
total precipitation occurred during this period. Rainfall in this area is sporadic 
and the total for a given month may fall in a single rainstorm. 


This marsh lies between the 98th and 99th meridians and within the 23 
inch rainfall zone, which Weaver and Albertson (1956) report as being 
part of the Mixed Prairie Region of Kansas. 


The average yearly temperature (°F), as recorded at the Hudson Weather 
Bureau from 1951-1958, was 56.5 with a range of 55.2 in 1958 to a high of 
58.4 in 1953. The average monthly temperatures for 1951-1958 were as 
follows: January, 32.6; February, 37.6; March, 41.9; April, 55.0; May, 65.3; 
June, 78.5; July, 81.1; August, 80.6; September, 71.7; October, 59.3; Novem- 
ber, 43.7; December, 35.2. The high air temperatures and low relative 
humidity during the summer months causes rapid evaporation of moisture 
from the soil and plants. 


Data which Flora (1948) reported from the Wichita Weather Bureau 


indicates that the prevailing winds are from the south, except for the month 


4 Tue University Science BuLLETIN 


of February. The wind velocity at Wichita for 1932-1935 averaged 12 miles 
per hour annually, and at Dodge City for the same period it was 12.9 miles 
per hour. 

This area has a yearly average of 140-180 clear days and receives 70-80°/, 
of the total possible sunshine occurring from June through August. Accord- 
ing to Borchert (1950), this grassland region, compared to forested areas, is 
typified by fewer days with precipitation, less cloud cover, and a lower rela- 
tive humidity on the average during July and August. He also mentions 
that there is a great chance of a rainfall deficit during the summer and that 
low rainfall and snowfall are typical of winters. 


HISTORY AND MANAGEMENT 


According to Schoewe (1953), the Big Salt Marsh is one of the original 
12 granted by congress to Kansas on its admission to statehood in 1861. 
Andreas (1883) has stated that no attempt at settlement was made in this 
portion of Stafford County prior to 1876. In 1876, however, a few people 
did move into the area, and, according to Sheridan (1956), a company was 
orgamized for the purpose of manufacturing salt on the marsh. Shortly 
after, it was found that this operation would be unremunerative and the 
undertaking was abandoned. Another factor which may have affected the 
vegetation in early times was grazing. A report in the Hutchinson News 
(1886) mentioned that there was an excellent area for stock pasturage along 
the salt marsh. Hay (1890) has also mentioned that salt marshes were used 
by native animals as a source of salt in early times. 

Two original surveys in this county covered parts of the marsh area. The 
first, by Wilcox (1870), reported that the land was generally poor and more 
or less sandy and that there were no stones or timber in the area. Another 
survey by Thompson (1871) indicated that willows and cottonwoods were 
present. He stated also that waters of the salt marsh were quite salty and 
the drying up of the area caused the flats to be covered with a thin coating 
of salt. 

At present, the most southern part of the study area, containing sandhills 
of an old beach ridge in sections 3, 4, 5, and 6, T.22S., R.11W., is heavily 
grazed throughout the growing season. Other areas which have been dis- 
turbed by grazing are the Scirpus americanus-Eleocharis rostellata areas 
along the western margin of the marsh. Some of these are heavily grazed 
throughout the growing season; however, the one quantitatively analyzed 
in this study has been only lightly grazed in April of some years. The sand- 
hill prairie areas on the eastern margin of the marsh, south of the east-west 
road which bisects it, are being grazed at the present time. The prairie area 
analyzed has not been grazed in the recent past, but both it and the S. ameri- 
canus-E. rostellata Community are mowed in late summer. The remainder 


YI 


Tue VEGETATION OF Tue Bic Sart Marsu 


of the communities, which make up the largest part of the area, are rela- 
tively undisturbed. 

In the past some ridges were made on the marsh when shallow lakes 
were constructed, and these still persist. Another factor altering natural 
conditions is that several small ditches have been dug to increase the rate 
of drainage, and in recent years the drilling of oil wells on the marsh area 
has caused some disturbance, but none has produced oil. In the future, the 
marsh should be less disturbed than in previous periods, since it has been 
included in the Quivira Wildlife Refuge which is being administered by 
the Fish and Wildlife Service. 


WATER RELATIONS 


WatTER TABLE AND DRAINAGE 


The marsh contains a complex of several shallow lakes, which are filled 
during wet periods when the water table is high. During the summer months 
there is a great deal of evaporation and a lowering of the water table, causing 
the surface of some of the lake beds to be covered by a white precipitate of 
salt. 

The only outlet from the marsh is in the northeastern corner (Fig. 1) 
where a short tributary carries the marsh waters into Rattlesnake Creek. The 
waters on the marsh rarely reach depths greater than 2-3 feet because 
man-made ditches drain parts of the area. 


METHODS AND RESULTS OF WATER ANALYSIS 


Water samples were taken from lakes, small ponds, and ditches through- 
out the growing season. The saturation extract technique for determining 
total salts and titration with silver nitrate for chloride determination were 
used as described in Richards et al. (1954). The pH was determined with a 
Beckman model Hz pH meter. 

Water samples collected by the author in 1959 and 1960 indicated some 
seasonal fluctuation in salinity. The fluctuation in salt content which occurs 
in inland saline lakes was indicated by Flowers (1934) in his work on the 
Great Salt Lake, where he found a 1°% increase in salt concentration with 
each one foot decrease in water level. The regularity of seasonal salt fluc- 
tuation in the waters of the Big Salt Marsh is much less than that reported by 
Purer (1942), Rawson and Moore (1944), and other authors. This can be 
explained by the fact that a regular fluctuation would occur only if there 
were definite dry and wet periods, and although there is a tendency for 
drying during July and August in this area, extremes in local weather condi- 
tions occur over short periods during the growing season. These fluctuations 
in weather, usually sporadic rains, cause fluctuation in salt content during 
any one month. 


6 Tue Untiversiry SciENCE BULLETIN 


As mentioned by Penfound and Hathaway (1938), surface waters can- 
not be considered of primary significance in fixing the actual distribution of 
plants except as to their influence on ground water. This holds true for 
most plants since roots usually penetrate into the soil and do not trail along 
the surface; however, the condition of the surface water does directly affect 
the distribution of submerged or floating aquatics found in these areas. 


It was found in this water analysis, as well as in the soil analysis which 
follows, that the chloride ion concentration fluctuated more or less directly 
with the total salinity. Table 2 shows that there was a general increase in the 
salt content during the month of July. This can be accounted for by a more 
or less uniform drying up of the marsh during this period, causing a con- 
centration of salts and in some areas leading to a complete evaporation of 
surface water. In a study of the ground water relations in this general region 
Latta (1950) took a few surface water samples on the Big Salt Marsh. One 
sample of special interest was taken from site one (see Table 2) in July, 
1944. It had a total salt content of 1.08° and a chloride content of 0.59%. 
This compares directly with the present author’s record of July 1960 in which 
the total salts was 1.17% and the chlorides 0.65%, indicating that the higher 
July salinity may not be just an isolated incident but of general occurrence 
during each growing season. Exceptions to this trend of a higher July 
salinity are sites 5 and 11, which had only a negligible salt content, and site 9, 
which was brackish but followed no clear pattern. A second explanation for 
the data presented in Table 2 could be that there was an irregular fluctuation 
in salt content throughout the growing season and that drier periods pro- 
duced the higher salinities. 


The pH relations in the various ponds give no clue to the salinity relations 
nor do they indicate what vegetation might occur in a pond. The 11 ponds 
studied had a pH of 7.5-8.2, but in no way could it be correlated with salinity 
in any of these areas. Site 11, a fresh water seep, had a pH of 7.2-8.7, and 
an average of 8.0, while the area which reached the highest salinity, site 7, 
had a pH which averaged 7.5 and a range of 6.7-8.1. 


The following is a list of the sites from which water samples were col- 
lected and tested for total salinity, chloride ion concentration, and pH 
throughout the growing season. The numbers given to the sites listed corre- 


spond with the numbers used in Table 2 and Fig. 2. 


1. A small pool north of the east-west road bisecting the marsh in the SE% sec. 22, 
aT 21Si) Rell We 

2. The main drainage ditch connecting the south and north areas of the marsh which 1s 
located under a small bridge along the east-west road in the NE'%4 sec. 27, T.21S., 
R.11W. 


. A small shallow pool north of the east-west road in the SEY% sec. 21, T.21S., R.11W. 


1oS) 


4. A roadside ditch on the north side of the east-west road in the SE% sec. 21, T.21S., 
R.1L1W. 


5 


Tue VEGETATION OF Tue Bic Sarr Marsu Th 


_ A small pool just north of the east-west road located .4 miles east of the north-south 
road bordering the marsh on the west in the SE% sec. 20a ZS eRe e 

_ A large shallow lake in the northeast corner of the marsh in the SE% sec. 15, T.21S., 
R.11W. 

_ The northeastern drainage outlet of the marsh in the NE% sec. 22, AZ (S Ree 

_ A shallow lake, south of hunting cabin, in the S% sec. 28, T.21S., R.11W. 

. A small pond along a man made hummock in the NW sec. 33, T.21S., R.11W. 

_ A large shallow lake in the southern part of the marsh in sec. 33, T.21S., R.11W. 

. A seepage area north of the east-west road which forms the southern boundary of the 
marsh in the SE% sec. 4, T.22S., R.11W. 


Taste 2. An analysis of the total salinity (T.S.), chloride ion concentration (Cl.) 


and 


pH of waters located on the Big Salt Marsh. Total salts and chloride are 
expressed as a percentage. 


Site May June July Aug. Sept. Avg. 
er ‘lS: 48 70 Istl7/ 2 45 66 
Gi: .20 233 .65 24 All 32 
pH 8.3 eal 8.5 Hell Tes: 7.7 
2, <r aS: 39 ws lS) 48 SA .60 
Cl: .16 34 61 22 a2 “29 
pH 8.2 We? 8.3 Tel Tie: 7.6 
3 re aes: 39 23 2 ie ee a2) 30 
Ck alg al Dry Dry .10 als 
pH 8.5 TET de NU eee he Jf 8.0 
Aen TS | oe Pee 30 By/ 30 
Cl. .08 als Dry 2 2h 14 
pH 8.2 (em es Te 8.2 Hell 
5) eS: .05 19 .16 .05 .18 “112 
@l O01 .09 06 .02 .07 05 
pH 8.6 8.4 8.0 7.9 8.1 8.2 
6. flies: 48 39 46 22 2D 5) 
Cl .28 .18 21 10 .10 pll7/ 
pH TET. 3 8.7 8.3 8.0 8.0 
 ——— aS: 62 2D) 2.96 36 29) 295 
Ck 30 5) 1.65 Alb) al 49 
pH 8.1 6.7 7.6 8.0 Va Td 
Gh a alieS: 79 54 1.26 5) 28 .64 
(Gl, 42 22 50 “li5 16 29 
pH iS 7.0 7.4 8.3 Wed US 
2) ite sea 1S: 152 40 42 38 5) 43 
Gh 25 .18 .28 16 5) oo. 
pH 7.6 Tee 8.2 8.1 TA Hel! 
Oi aS 76 70 80 35%, Me 65 
Cl 38 30 43 iy ee 31 
pH 8.1 Tes 8.2 S22) a ees 7.9 
11 eS 02 04 .03 03 05 03 
Cl 01 009 009 008 01 009 


8 Tue Universiry ScrENcE BULLETIN 


METHODS OF SOIL ANALYSIS 


Total salinity was determined by the saturation extract method described 
in Richards et al. (1954). Chloride ion content and saturation percentage 
were also obtained by methods described in Richards et al. (1954). pH was 
determined by a Beckman model Hz pH meter from a saturated soil paste. 
Mechanical analysis data were obtained following the method of Bouyoucos 
(1936), and for determining the organic content of soils the ignition method, 
described in Kurz and Wagner (1957), was used. Tables 3-6 contain the 
soil analyses data. 

Soil samples were collected at two different depths. The first at 0-10 cm. 
and the second at 60-70 cm. The following is a list of sites from which soils 
were collected throughout the growing season and upon which tests were 
made for the various soil factors. The numbers given to the sites correspond 
with the site numbers used in Tables 3, 4,5, and 6. 


1..A Sandhill Mixed Prairie area 25 meters north of the east-west road bisecting the marsh 
and 25 meters east of a north-south road in the SE% sec. 22, T.21S., R.L1W. 

2. A Sporobolus airoides area 25 meters north of the east-west road and 50 meters west of 
the north-south road in the SE% sec. 22, T.21S., R.11W. 

3. A Tall Distichilis stricta area 25 meters west of site 2 in the SE% sec. 22, T.21S., R.11W. 

4. A Tall Distichilis stricta area 45 meters north of the east-west road and 25 meters east 

of the north-south man-made ridge in the SE% sec. 22, T.21S., R.L1W. 

5, Bare ground 50 meters north of the east-west road and 350 meters east of the north- 

south road in the SW'% sec. 22, T.21S., R.11W. 

. A Suaeda depressa area 30) meters west of the north-south road and 100 meters north 

of the east-west road in the SE% sec. 21, T.21S., R.11W. 

. A Suaeda depressa area 300 meters west of site 6 in the SE% sec. 21, T.21S., R.L1W. 

8. A Scirpus paludosus-Distichlis stricta area 400 meters west of the north-south road and 
139 meters north of the east-west road in the SE% sec. 21, T.21S., R.11W. 

9. A Distichlis stricta-Suaeda depressa area 100 meters north of the east-west road and 150 
meters west of the short false road in the SW% sec. 21, T.21S., R.11W. 

10. A Spartina pectinata-Distichlis stricta area 100 meters north of the east-west road, 
directly north of the south road in the SW% sec. 21, T.21S, R.11W. 

11. A Dustichlis stricta-Suaeda depressa area 400 meters west of site 10 in the SW% sec. 21, 
MANS. Rowe 

12. A Spartina pectinata-Distichlis stricta area 400 meters west of site 11 in the SW% 
seen d/l We 21'S Rew: 

13. A Distichlis stricta-Suaeda depressa area 30 meters south of the east-west road bisecting 
the marsh and 200 meters west of the south road to a hunting cabin in the NW\% sec. 
28 le 2Se Rel We 

14. A Scirpus americanus-Eleocharts rostellata area 100 meters north of the east-west road 
and 300 meters east from the north-south road bordering the marsh on the west in 
the SEY sec..205 1.21S., Rall Ww. 

15. A Distichlis stricta-Suaeda depressa area 775 meters south of the east-west road and 200 
meters west in the SW4 sec. 28, T.21S., R.11W. 


ON 


NI 


Resu_Lts oF Som, ANALYSIS 


SOIL TEXTURE 


In an early survey of the soil in the marsh, Coffey, Rice, et al. (1912) 


Tue VEGETATION Or Tue Bic Sart Marsu 9 


classified the surface soils as a silty clay and also mentioned that there were 
small spots of sand, sandy loam, and loam scattered throughout the area. 

Park and Dennis (1947) stated that the soils of this area have been formed 
from outwash sands and silt which were mixed with loess deposits. They 
further mentioned that the soils were sandier on the tops of knolls and 
heavier and occasionally saline in the flats and depressions. 

In this study the textural classes of the soils were arrived at by applying 
the actual percentages obtained in the mechanical analysis given in Table 3 
to the triangular chart for soil textures in Lyon, Buckman, and Brady (1952). 

In the present survey, the surface soils, 0-10 cm., of the prairie and S. 
airoides communities, sites 1 and 2, were classified as sand. The eastern part 
of the tall D. stricta Community was loamy sand, whereas the western part 
of this zone and all other soil sample sites except site 14 was classified as a 
sandy, clay loam. Site 14, dominated by a S. americanus-E, rostellata Com- 
munity had a sandy loam. 

Although the soils underlying the marsh were classified by Coffey et al. 
(1912) as a silty clay, in the present analysis at the 60 cm. level all sites except 
11 and 14 were sand. Sites 11 and 14 were classified as a loamy sand. 


ORGANIC MATTER 

The organic content relations of the marsh were similar to other areas 
in that there was a general reduction in organic content from the surface soil 
downward. Table 3 shows that only in site 2 was there an exception, and 
the soils in this area had an extremely low organic content. The lowest 
organic content occurred in the sandy soils of the prairie and S. airordes areas. 
Site 14, located in the S. americanus-E. rostellata area, had the highest organic 
content, 23.4°/, of any soil on the marsh. 


SATURATION PERCENTAGE 

The saturation percentage of the various soils was determined at all sites 
and is recorded in Table 3 as an average saturation percentage for the five 
months sampled. As would be expected, the sandiest soils, located in sites 
1 and 2, had the lowest saturation percentages, 28°, and 34°%, of the surface 
soils. The sandy clay loam soils had a higher saturation percentage of 40-75%, 
in the surface 10 cm., and the highest saturation percentage, 151%, was found 
in a sandy, clay loam soil of the S. americanus-E. rostellata Community. This 
soil had the highest organic content on the marsh, and, correspondingly, other 
sites with high organic content and clay content had high water-holding 
capacities. At the 60 cm. level the soils were generally sandy, and in all 
cases except sites 1 and 2 the saturation percentage was lower in these sub- 
surface areas. 


PH RELATIONS 
Table 4 shows that there is an irregular fluctuation in the pH of the sur- 


10 Tue Universiry SciENcE BULLETIN 


face as well as the subsoil in each of the community types during the growing 
season. In general the pH in the surface 10 cm. averaged lower or equal to 
the soil pH at the 60 cm. level. Daubenmire (1959) stated that surface soils 
have a tendency to be more acidic than sub-soils because of the greater quan- 
tities of acid-forming organic matter and the stronger leaching action in the 
upper levels. In only two cases, sites 5 and 13, did the surface pH average 
higher, and in both cases there was only 0.1 pH difference. The pH of the 


prairie area was 6.5, the S. airoides area 7.2, and in the remaining communi- 


Taste 3. An analysis of the saturation percentage, expressed as an average for 

the five months sampled, organic content, expressed as a percent of the organic 

matter in the September collection and the soil texture, expressed as a percent of 

the different size particles in the September collection. A equals 0-10 cm. and 
B equals 60-70 cm. 


Saturation Organic Soil Texture 
Site Percentage Content Sand Silt Clay 
1s aN ar eects ean 8 se ey a ee 28 Jel 89.3 1.4 93 
BGS eee eek i 4 91.3 1.4 HS, 
Dae LACED: ean Ee oe oe ee 34 22 85.9 5.0 Il 
Bop eee eee 34 Bef, 84.9 6.0 wil 
a eee ee ee 43 25 78.9 6.5 14.5 
Baer rere eae eae 40 8 89.3 2.4 8.3 
AOAC. Upeeatt sha a. orc eee 75 6.2 07 10.0 SKS 
Bee pe ee. eee 39 1.6 89.7 “at 9.9 
SRA aoe ac eee: DEES Seem 40 Ded 69.9 D4 Niel 
Bare eee 30 1.0 87.7 0 123 
Onm AG tes See ee 40 3.6 Veni 7.0 20.3 
18h, Makerere 2 gna 2a ean mane 34 2.6 83.9 3.6 12.5 
ate NOM oe on 40 6.8 66.9 10.6 22.5 
Baga = <2 eee ee 32 el 88.7 0 mleS 
Rohit das Weyl see eee See See ee 47 9.2 64.3 8.8 26.9 
Beg a)... taheene cen ae ) 36 Tel 85.3 3.6 titel 
Shh: low Rigs cepecke Be Sabaenpta sents Me 4] 93 65.9 10.0 24.1 
Beene peta eas: 2 el ee 1.7 89.9 0 10.1 
Oe Aseameae enon. ee eas eee a XS 17.9 49.9 20.0 30.1 
1S}, ata bie on ote ee . Bf 4.6 91.9 0 8.1 
DS CAGE Beastie ene. = 44 ez 68.3 8.6 23 
|S a a iy OS 31 aaa 78.9 6.0 15.1 
1 ONS Soe a ee Ot 16.1 68.3 8.0 230i 
Bae ee ere ce eee 36 1.0 88.9 3.0 8.1 
HLS} JANG. (peeks 5 ee aparece 44 6.6 ile Sea S47, 
IBA i goes a Me Peete rere Seo 32 Ppp) 84.5 Ue 8.1 
Lie AW ang sero eset pee lS)! 23.4 61.5 26.8 11.7 
SE coe eee 2 ee ee 40) Bet ZED 8.8 15 ez, 
Sy BAN ae Sc ee eer. 62 25 Oe 22.4* Aporll 
Be eee et Soke ee 33 led 91.5 0 8.5 


* The organic content and texture of soils from site 15 were obtained from the August collection. 


Tur VEGETATION Or Tue Bic Sat MarsH 11 


ties averaged 7.7-8.4 in the surface soils. Keith (1958) mentioned in his 
cursory study of salinity relations in a Canadian salt marsh that pH may be 
an important factor in plant distribution after a certain salinity is reached. 
In this study as in those by Evans (1953) and Kurz and Wagner (1957) no 
correlation could be made between the various vegetation zones of the salt 
marsh proper and the pH values. 


Tora SALINITY AND CHLORIDE Ion CoNTENT 


As stated by Coffey et al. (1912), the factor which distinguishes the salt 
marsh soils from the other alluvial and aeolian soils of this area is the 
accumulation of salts, principally sodium chloride. The chief source of this 
sodium chloride is, as mentioned earlier, the rock salt deposits in the under- 
lying Permian and Cretaceous rocks. 

From the analysis of total salinity and chloride ion concentration given 
‘1 Tables 5 and 6, it is clear that in general the total salinity and chloride 
‘on concentration averages higher in the surface soils than in the subsoil 
sample. These findings also show that there is a variation in salinity and 
chlorinity between the various vegetation zones and also within each zone 
throughout the growing season. Here, as in the case of the ponds studied, 
the variation within a zone is not so great that one cannot separate the saline 
from nonsaline soils. 

The main factors affecting the fluctuation of salinity in these soils appears 
to be rainfall and the evaporative capacity of the air. In accord with constant 
fluctuations in weather during the growing season there are corresponding 
Auctuations in the soil salinity. There is a tendency for the marsh area to 
dry up during July and August, and along with this there is a rise in salinity, 
but sudden showers during these months cause rapid salinity changes. Kurz 
and Wagner (1957), in their work on coastal marshes, reported that chlor- 
inity fluctuates considerably within short periods of time, even at the same 
site, and they also mention that conditions of rain, drought, and seasonal 
temperatures are most important in modifying salinity relations. This same 
conclusion was reached by Stocker (1928) who stated, “These (salt) con- 
centrations represent no constants; they show great fluctuations in the 
smallest space and especially also at different times.” Novikoff (1958), in his 
work on the vegetation of the saline soils of Tunisia, also reported a temporal 
variation in soil salinity. 

In comparing Tables 5 and 6 it should be noted, as one would expect, 
that the chloride ion concentration varies more or less directly with the total 
salinity. A similar correlation was made by Evans (1953) in his study of 
the halophytic vegetation of Lake Ellesmere, New Zealand. The salinity and 
chlorinity in the surface soil of the prairie was the lowest, with the salinity 
0,003-0.02°/, and averaging 0.01°.. As one approaches the salt flats from the 
eastern prairie region there is a continuous increase in salinity, as illustrated 


12 Tue University SciENCE BULLETIN 


Taste 4. An analysis of the pH of the soils sampled from the Big Marsh. A 
equals 0-10 cm. and B equals 60-70 cm. 


Site May June July Aug. Sept. Avg. 
Vi AWS ech eS eae ee 6.9 6.5 6.3 6.4 6.7 6.5 
Pe eee nc Bec e 6.2 7.0 6.5 7D 6.9 6.8 
DAR Je Ne sere arene ie eee Seen 1S) Hep 6.8 6.5 7.6 Up: 
|S dienes free wae eet Sant eee ee 8.2 8.3 8.9 TS) 8.5 8.2 
3 RRS pre NMI or ol eta 8.3 8.6 Tee 7.8 8.0 79 
|S tear ses set AREA Ses ae Pe OS 8.6 8.3 GS 8.1 8.4 8.2 
ARAM ay eetcn. seaemiee Late Meee Le 8.6 8.3 8.0 8.2 8.2 8.2 
|B}. ALGAE OOSE Notre ROBE 3 See ENON 8.0 8.2 8.0 8.8 8.1 8.2 
DI aN hy ee ee ee eR ees Ee Se2 8.3 8.4 8.4 8.2 8.3 
Beer Lee en a 8.4 8.0 8.3 8.2 8.1 8.2 
CA Se ee eae ee et 8.8 8.1 8.2 8.3 8.0 8.2 
|B Falta i Men el tle a 8.4 8.4 8.1 8.1 8.1 8.2 
pe Nee ens es Ee poe 8.7 8.0 8.1 8.2 8.1 8.2 
| eR ee aa CRE PE Nowe daar see 8.6 8.2 8.2 8.3 8.2 8.3 
SEA a ees eerne so ebeee nen Thee 8.3 8.6 8.4 8.4 8.4 8.4 
Bie eee ene Ree ea dae 9.2 8.9 9.0 8.8 8.7 8.9 
OA re ce ee, Wee US 7.9 8.5 7.8 8.6 8.0 
|S PARA apoE Ue San eee cae 8.4 8.4 8.4 8.6 8.4 8.4 
l OAC eee Seen aw Pen ete 8.0 8.0 8.2 8.5 8.1 8.1 
ye) ere se ee ee 8.4 8.8 8.3 8.4 8.7 8.5, 
TRI AY yee eae ee ee 8.7 8.3 8.4 7.6 8.3 8.2 
Bike tte Ae ene Sees BE Scio) 8.1 8.4 8.2 8.3 8.3 
De AR eee Ae Pe ern 8.6 8.8 8.6 8.3 8.1 8.4 
1B ese an eee REE de SEE) 8.6 8.6 8.4 8.6 8.2 8.4 
NS Ag ase ee on eee Cee mane 8.0 8.4 8.5 8.3 8.1 8.2 
BS ages oe ee oe eee Hell 8.0 8.4 8.2 8.6 8.1 
Meee UN abeern al Pa Bieter tree gel Ta o8 Tel. 7.9 7.6 7) 7.8 Uff 
Bigg phic sree ee a 7.4 8.6 8.5 8.4 8.0 8.1 
LSERC Ys pn Dome tea ey RCE Ets Sere, 8.3 8.6 8.1 8.2 ae 8.3 
|S eee eee Say PRR Tee Z 8.3 8.2 8.3 8.4 ”, 8.3 


in sites 2, 3, and 4. This culminated in the area of highest salt content, the 
barren salt flats, where the surface soil salinity was 2.12-2.96°%. Sites 6 and 7 
had a cover of S. depressa and sites 9, 11, 13, and 15 a D. stricta-S. depressa 
Community. The soils at all of these sites were definitely saline, having sea- 
sonal averages of 1.04-2.15°.. The soils of site 8, which contain an S. paludo- 
sus-D. stricta Community, were surrounded by highly saline soils. This area 
was most saline on its margins, where the May sample indicated a salinity 
of 1.10°%, but was reduced toward its center as indicated by the 0.22-0.40% 
range for the remaining months. Sites 10 and 12 support a S. pectinata-D. 
stricta Community, and the surface soil salinity was 0.20-0.63°, and 0.27- 
1.00%, at each site respectively. The soils at site 14 maintain an S. americanus- 
E. rostellata Community and had a salinity in the surface 10 cm. of 0.23- 


Tue VEGETATION Or Tue Bic Sart Marsu 13 


0.68°.. This area of low salinity and chlorinity was on the western border 
of the study area. 

The extreme change in salinity that can occur in an area due to local 
weather conditions is clearly noted at site 9. The September soil collection 
at this site was made the day after an extremely heavy rain, and the salinity 
and chlorinity in the surface soils were greatly reduced due to its diluting and 
leaching action. The 0.49%, salt content at this site was almost 1°% below the 
average salinity of 1.42%, of the four preceding samples. 

The importance of these salinity and chlorinity measurements is that 
they provide some information as to the extremes in variation that can occur 
in the various vegetation zones. It is usually not the average but the extremes 


Taste 5. An analysis of the total salinity of the soils sampled from the Big Marsh 
expressed as a percentage of the salts in the soil. A equals 0-10 cm. and B equals 
60-70 cm. 


Site May June July Aug. Sept. Avg. 
IL, Gigs CSCI area ene eee 02 01 003 .02 005 01 
BRE OE ei ee De .009 005 003 005 002 .004 
DIMA rr res De eA ee Sy 04 ail) 10 14 a2 
| BR 0 ees, eer ea I rate 08 04 a 01 O01 05 
See eae beers ree csp 34 29 63 39 31 39 
Bh) eee ee ee cee 18 18 5) -49 31 28 
Ate SUNG = See eee oer oa See 73 ae) 4 19 74 fs) 
1 BAS cee ee ee OR 16 ) 3 65 42 42 
RAS Ree ee SNe ee ae, 2.78 2.90 2.96 DAZ 2.47 2.64 
Bee ee ee, ee See 1.08 98 81 he) 65 85 
eA er erratic oy as) weet 2.43 1.76 2.40 1.70 S38 1.92 
13) eae a eee Sas | pee ee eee 98 91 80 1.29 115 1.02 
TANS eee eee ee ae AVA 1.16 2719 1E29 1.59 1.70 
Bi se ee ee ee 1.01 /4 67 1.07 655) 76 
SRE ACE rete tr Wyn. eee ee ello yp) Boi, 30 40 45 
ee ee ee ee ee ee eee 32 28 29 BoD 45 33 
OTM Ne ee ren ee eee een 1.43 127, 1255 1.45 49 1223 
Bee ee err eee Sy 67 35 96 1.09 2 
NO NPAVRA SS cok: Se ee .20 24 3h) 63 62 41 
13}, See oy pee at a 07 16 40 49 36 29 
IMIR Att pyceencee sO Ur es. 58 1.46 .67 ey 1.01 1.04 
eee ee 56 76 27 70 .67 29 
Zh 2 ENG ie Se 59 47 Dy 1.00 JD 61 
1B et eae ste ae ae 34 a2 06 LD) 38 2 
LES Ne tee Beech Oe ee ee LSM 2.47 2.24 2.03 DZ PHS) 
1B fa Soy sc) at oe re ee a HD 51 56 48 41 23 
os SRN Poe tee ae oe 23 18 38 53 68 40 
ete ee ALES ree hs 07 03 04 05 05 .04 
SAC NErse.s bei B iat oT. 1.18 82 1334) 25g eee 144 
1B (Seen ere ne er a ee ee ee a) .67 94 eS) ee 69 


14 Tue University ScIENCE BULLETIN 


Taste 6. An analysis of the chloride ion concentration of the soils sampled from 
the Big Marsh expressed as a percentage of the chloride in the soil. A equals 
0-10 cm. and B equals 60-70 cm. 


Site May June July Aug. Sept. Avg. 
ge NO ares Nearer, SA cette MARU ee en 007 .002 001 007 .003 004 
BS ese BAe 5 i PR EE 003 .0009 0009 001 -0007 001 
Die Atay e 2 eRe ce tek Rh 06 .02 09 04 .06 .05 
ta nae woe Rant owt ke 03 01 05 003 004 01 
So! Talis re ee rate eerie Ve 18 10 31 sll) 18 19 
1 Do Ce eae 10 10 ale 28 iI lS 
hs GN  Mnarnreest a teers <P ANE Neto 34 hy) 43 30 Ly) 39 
13) ee Ee Lae oe, bo Ae A alll sot 24 37 20 22 
SP a eee ee ae Eee ee 1.38 234 1.77 1.30 1.67 1.69 
12 = = ee ee ae ee ee 65 63 44 40 39 50 
Gig At Hse eich. Ra hn od 1.20 eZ ley 78 96 1.11 
Bit gee ee See ee ee eee 29 58 .47 74 74 62 
ed NE ea See nr ee re ns 1.05 .88 72 .88 LS 1.05 
|S i fed nye See teers Ce gene eye oy) 44 36 4 als 41 
CUNA GS, «easement tebe as weer eee 68 .09 ails) alu 19 24 
1 Bs Aap LP AT it A aR ed De sil7/ al2 5) als 2D 16 
Ogre tid Aures Mes ee UA aris Lee 96 1.20 93 94 16 83 
Ie eee Pe ae aay ReaD 0 “37 45 16 4 69 He 
HOV ge At ected eee ee 09 14 all7/ 36 29 21 
|B ee sa a een! aly .03 .08 18 28 22 5ll5) 
UNS (At ae eee eee ole a eee 5) 56 38 98 60 DD) 
Be ee ee ee 24 43 sll 234 so) 30 
LO AT eR ale Se Santee ee ae 35 3) 514 46 sal sil 
1S Rg ie a ePnrcies Rpesee Spe 18 07 04 alll .20 a2 
1 ae a ee ee go I a eR eae Ges Pe 1.04 1.20 .80 1.16 1.10 1.06 
13 eee ee ne ene & Prone eS 43 J] 38 27 alt sill 
PORE AS 9 ese oop RE Seer eer ace 09 .03 a12 ally 3} sl 
BA feo Ree eee Pe eee 04 008 01 02 02 01 
PS EBAG | Bee ee oe nee eee al 46 70 Se .87 
Be gee ae ae ee Sea S5) 40 .48 24: Saeeees 36 


in salinity which are most limiting to plant distribution. The important 
question is not whether a plant can survive in an area such as site 9 after a 
heavy rainstorm, but whether it could survive throughout the growing sea- 
son when at times the salinity surpassed 1°... 


VEGETATION*® 


SEASONAL ASPECTS AND ZONAL DISTRIBUTION 


Throughout the growing season notes were taken on the seasonal aspect 
and distribution of the vegetation. The following section is a résumé of this 


*An attempt was made to follow the nomenclature of Hitchcock (1950) for the classification 
of grasses and Fernald (1950) for the remainder of the vascular plants. 


Tue VEGETATION OF Tue Bic Sart Marsu 15 


information, indicating phenology and zonal distribution (Table 8). 

The central area of the marsh is a barren salt flat, and the various vegeta- 
tion zones radiate out from it. Vegetational relations in this marsh differ from 
many coastal ones and the Kansas salt marshes described by Schaffner (1898) 
in that there is no definite sequence of community zonation which always 
occurs from the bare area outward. There are definite patterns of vegetation, 
but these are affected by variations in local topography, drainage, and salinity. 
To illustrate the lack of definite concentric zones one can find a prostrate 
Suaeda Community, an upright Swaeda Community, a Sporobolus airoides 
Community, a Distichlis stricta-SSuaeda depressa or possibly a Distichlis 
stricta Community bordering or entering upon the barren salt flats. 


Tat Distichlis stricta MEADOW CoMMUNITY 


The first major community of the salt marsh discussed here, the tall 
D. stricta, is located on the eastern border of the salt flats in the SE% sec. 22, 
T.21S., R.11W., south of this location on the eastern border and in wetter 
saline areas throughout the marsh. In this vegetation type there were 13 
species of vascular plants, and only three of these, Plachea purpurascens, 
Rumex maritimus var. fueginus, Polygonum hydropiperoides, appeared 
solely in this community. 

The most characteristic and abundant plant in this zone was D. stricta; 
others flowering and fruiting in the spring, were Poa arida, Hordeum juba- 
tum, and Scirpus paludosus. The flowering period of the last mentioned 
species continued through the summer months. 

Distichlis stricta flowers throughout the summer months, and fruiting 
plants can be found from July through October. Bassia hyssopifolia and 
Atriplex patula var. hastata flower during the summer, while Aster exilis, 
Aster ericoides, and Suaeda depressa were found in anthesis in late summer 
and fall. These summer and fall species were of extremely rare occurrence 
in this community. 

In this area one also finds extensions of prairie vegetation on hummocks 
of higher ground. The dense cover in this community is illustrated in Fig. 3. 


Sporobolus airoides COMMUNITY 


This community was located in the SE% sec. 22, T.21S., R.11W., between 
the Tall D. stricta community and the Sandhill Mixed Prairie community. 
It contained only 11 species of vascular plants, all of which were found in 
other communities on the marsh. This vegetation type also occurred on small, 
self-produced hummocks formed by the extensive root and rhizome systems 
of S. atroides, which was 1-3 feet in height. Harris (1920) and Hilgard 
(1914) have also mentioned such raised areas produced by this species. 

The vegetation in this zone was dominated by S. airoides, which bloomed 
from June through August. In early spring Poa arida was the only plant 


16 Tue University ScrENcCE BULLETIN 


in anthesis. Elymus canadensis was found blooming in late spring and early 
summer. 

Flowering during the summer months were Ambrosia psilostachya var. 
coronopifolia, Atriplex patula var. hastata, Conyza canadensis, Desmanthus 
illinoense, Euphorbia marginata, and Suaeda depressa. D. stricta was found 
blooming also throughout the summer months but was much rarer than in 
the previously described community. 


SANDHILL MIxep PRAIRIE COMMUNITY 


The eastern border of the marsh, directly east of the S. airoides Commu- 
nity, in the SE% sec. 22, T.21S., R.11W., and southward along the eastern 
border was occupied by this community. It was found also on the higher, 
sandy hummocks in moister and more saline zones. Of the 110 species oc- 
curring in this mixed prairie community, 41 were not found elsewhere on 
the marsh. 

In early spring, Lithospermum incisum, Androsace occidentalis, Ceras- 
tium brachypodum, Lappula Redowsku, Veronica peregrina var. xalapensis 
and Viola kitatbeliana var. rafinesquiu were found in the flowering state. 

From late spring through early summer, among others, the following 
species were in flower: Achillea lanulosa, Agropyron smithu, Callirhoe in- 
volucrata, Elymus canadensts, Erigeron strigosis var. beyrichu, Festuca octo- 
flora, Monarda citriodora, Panicum lanuginosum var. fasciculatum, Pant- 
cum scribnerianum, and Plantago purshu. Also found in some abundance 
during this period were Bromus japonicus and B. tectorum, which are not 
common in undisturbed prairie. 

The abundance of B. japonicus must be ascribed to the disturbance caused 
by present mowing during the summer and possibly to grazing in the past 
history of the area. Herbel and Anderson (1959) have reported annual 
brome grasses to be the major invaders due to disturbance by grazing in 
the Flint Hills of Kansas. 

During the summer months Artemesia ludoviciana, Cassia fasciculata, 
Coreopsis tinctoria, Cyperus filiculmis, Oenothera rhombipetala, Setaria 
geniculata, Solidago missouriensis, and Strophostyles leiosperma were in the 
flowering state. 

In the late summer and fall the tall grasses dominated the vegetation. 
These included Andropogon gerard, Andropogon scoparius, Panicum virga- 
tum, and Sorghastrum nutans. Other plants flowering at this time were 
Ambrosia psilostachya var. coronoptfolia, Bouteloua gracilis, Eragrostis 
trichodes, Froelichia campestris, Solidago altissima, Sporobolus asper, and 
Sporobolus cryptandrus. 

Moister areas in this zone contained species common to the S. pectinata- 
D. stricta and S. americanus-E. rostellata communities, described later. An 
early summer aspect of this community is shown in Fig. 5. 


Tue VEGETATION OF Tue Bic Sarr Marsu 17. 


Suaeda depressa COMMUNITY 


This community was composed of widely scattered individuals of S. 
depressa in both its depressed and erect (S. erecta (S. Wats) A. Nels.) forms 
(Figs. 7, 8). Along with the two forms of Suaeda, succulents such as Sesza- 
vium verrucossum and, in one location, Salicornia rubra were found. This 
community type was scattered everywhere in and directly bordering the open 
salt flats in sections 21, 22, 27, 28, and 29, T.21S., R.11W. 

Suaeda depressa was flowering and fruiting in late September and Octo- 
ber. It was present, however, in this community from early May through 
October in its vegetative form. Salicornia rubra was in vegetative condition 
from July through October, and Sesuvium verrucosum was flowering and 
fruiting in late August and September. 

As mentioned earlier, S. depressa has two distinct growth forms in this 
area, the depressed form, generally found in the more saline areas, and the 
upright form situated at slightly less saline locations. However, continuous 
fluctuations in salinity over short periods of time seemed to allow both forms 
to occur at the same location. Even the usually barren salt flat sometime had 
a very widely scattered cover of the depressed form. The occurrence of these 
two forms at the same location seems to leave no doubt whether or not the 
growth form has any significance in the distribution of this species or any 
real survival value. It also appears as if specimens which develop later in the 
season are all of the depressed form. 


Distichlis stricta-“Suaeda depressa COMMUNITY 


This community covered the largest area of the marsh and was composed 
almost entirely of D. stricta and the upright growth form of S. depressa 
(ie. 9). Large areas of sections 21, 22, 27, 28, and 29, T.21S., R.1]W. were 
covered by this vegetation type. 

Other species found here comprised only an extremely small part of the 
total vegetational cover, but they were occasionally of some local importance. 
Species of secondary importance included Poa arida, Scirpus paludosus, 
Polygonum ramoisissimum, Sesuvium verrucossum, Sporobolus atrotdes, 
Tamarix gallica, Spartina pectinata, and Heliotropium curassavicum. 


Scirpus paludosus-Distichlis stricta COMMUNITY 


This community type covered relatively small areas of the Big Salt Marsh 
and was primarily limited to the SE’% section 21 and the SE% section 28, 
T.21S., R.11W. It contained a total of nine species, all of which were found 
in other areas, except for Atriplex argentea. 

The dominants in this vegetation type (fig. 4) were D. stricta which 
bloomed through the summer months and S. paludosus which started bloom- 
ing in late spring and continued through part of the summer. Other species 
in this area included Polygonum ramosissumum, Suaeda depressa, Poa arida, 


18 Tue University ScreENcE BULLETIN 


Hordeum jubatum, Heliotropium curassavicum, Atriplex argentea, and 
Atriplex patula var. hastata. Many of these species occurred also in the pre- 
viously mentioned D. stricta-S. depressa Community, but in this, as in the 
latter community, with the exception of S. depressa, they played a very small 
role in the cover relations. 

Polygonum ramosissimum, sometimes locally abundant in this and the 
previously discussed community, flowered throughout the summer months. 
H. curassavicum, which was extremely rare, flowered throughout the grow- 
ing season. 


Spartina pectinata-Distichlis stricta COMMUNITY 


This community was located primarily in the SW’ section 21, T.21S., 
R.11W. It was also found bordering the Scirpus americanus-Eleocharts 
rostellata Community and in the southwestern part of the marsh, and small 
colonies were found scattered throughout the D. stricta-S. depressa Com- 
munity when the salinity was lowered. 

Spartina pectinata, the tall dominant here (Fig. 12) was 70-130 cm. tall, 
and D. stricta, the short co-dominant, was 20-50 cm. tall. There were 24 taxa 
in this community, one of which, Eustoma grandiflora f. fischeri, seems to 
be limited to this area. 

In the spring such species as Amorpha fruticosa, Apocynum sibiricum, 
Hordeum jubatum, Juncus torrey1, and Juncus interior were flowering. Dur- 
ing the summer months the most prominent flowering plants were D. stricta, 
Eustoma grandiflora, Asclepias incarnata, S. pectinata, Verbena hastata and 
Vernonia fasciculata. In late summer and fall, species such as Aster exilis, 
Iva annua, and Flaveria campestris were in anthesis in this zone. Spartina 
pectinata, towered over the remaining species, and, as mentioned earlier, it 
should be considered the most characteristic species in the upper stratum of 
the community. 


Scirpus americanus-Eleocharts rostellata COMMUNITY 


This community contained 89 species, 41 of which were limited to this 
sedge-meadow type area located in sections 20, 29, 30 and 31, T.21S., R.11W. 
and bordering the marsh along its western edge. 

The dominants in this area were S. americanus, the tall dominant, (70- 
160 cm.) and E. rostellata, the short dominant (30-70 cm.). These two species 
bloomed from late May through June. Besides the dominants, some of the 
characteristic flowering species during the spring (Fig. 13) were Carex an- 
nectans, Carex lanuginosa, Carex praegracilis, Juncus interior, Juncus torreyt, 
Scirpus paludosus, Eleocharis tenuis, Scirpus lineatus, and Sphenopholis 
obtusata. 

During the summer months, Agrostis alba, Asclepias incarnata, Cicuta 
maculata, Helianthus maximilliant, Lippia lanceolata var. recognita, and 


Tue VEGETATION Or Tue Bic Sart Marsu 19 


Verbena hastata were in bloom (Fig. 14). Flowering in late summer and 
fall were Vernonia fasciculata, Aster ericoides, A. exilis, Lobelia siphilitica 
and Spiranthes cernua. 


FresH-W ATER-SEEP COMMUNITY 


This area was south of the Grazed Sandhill Community and north of the 
east-west road forming the southern boundary of this study. It was situated 
a half mile west of an oil pump in the center of the region south of the sand- 


hills in the SW sec. 4, T.22S., R.11W. 


The vegetation was in the center of the seepage area where a fresh water 
pond was located (Fig. 15), and was dominated by Typha latifolia and 
Sagittaria latifolia. The area contained 30 species, five of which were not 
found in any other area studied. Species limited to this area were Berula 
pusilla, Mimulus glabratus var. fremontu, Myosurus minimus, Ranunculus 
sceleratus, and Spirodela polyrhiza. Surrounding the pond on slightly higher 
ground were found the following species typical of the Sedge-Meadow Com- 
munity described earlier: Scirpus americanus, S. paludosts, Eleocharis macro- 
stachya, E. rostellata, E. tenuis, Cicuta maculata, Flaverta campestris, Carex 
annectans, and C. praegracilts. 


Taste 7. Distribution of Charophyceae. 


Ponds 

Species 1 3 8 9 
CNET ODES TAOS, Baa a = x 
GGL CUNCUOLULG # ree an EE oe ere : 24 a Xx _ 
CUNGIREL, GUO ES: Ses a ees. xX x X 
CINEREA NO RIALT TD eee ee X é ».4 
INVIAGHIIDD. OPN LTE ee eR ee = - X 
TG DANE DELICE IN ace ae X X 


M. glabratus var. fremontii, M. minimus, and R. sceleratus were the 
earliest species in the pond area undergoing anthesis during early spring. 
Typha latifolia was blooming in the early summer months, S. latifolia trom 
middle to late summer, and Lycopus americanus in mid-summer. Flaveria 
campestris flowered during late summer and fall and was one of the last 
species to flower in the area. 

The water at this site was tested for pH, total salinity, and chloride ion 
concentration throughout the growing season. It had a very low salinity of 
0.02-0.05°%, a chloride ion concentration of 0.008-0.019., and a pH of 7.2-8.7. 
Details of the exact monthly variations in these factors during the growing 
season are in Table 2. 


20 Tue University ScteNce BULLETIN 


SALINE Ponp CoMMUNITY 


An aquatic community differing markedly from the Fresh-Water-Seep 
was located in the shallow saline lakes scattered throughout the marsh. Two 
species of fowering plants were found growing in these lakes. In June and 
July Eleocharis acicularis was flowering and Potamogeton foliosus was in 
anthesis from June through August. 

The group of plants making up the greatest density (Fig. 16), in these 
brackish and saline lakes belongs to the class Charophyceae. Jewell (1927), 
in a study of the fauna of salt marshes in this county, mentioned that the 
shallow ponds were literally choked with Chara. The six species of Charophy- 
ceae found on the Big Salt Marsh include Chara canescens, C. evoluta, C. 
globularis, C. hornemani, Nitella opaca, and Tolypella intricata. These 
species are most abundant during the spring and early summer months. 
Later on in the summer when many of the shallow lakes dried up, these 
algae disappeared and only the thick walled zygotes, which fall to the bottom 
of the ponds, remained. The zygotes overwinter as such and germinate 
during the following spring. Table 7 provides data on the distribution of 
Charophyceae in four ponds. 

The salinity in the ponds in which these plants grew was 0.48-1.17°4 in 
pond 1, 0.25-0.39%, in pond 3, 0.28-1.26°% in pond 8, and 0.38-0.52°% in pond 9. 
The exact locations of these ponds are shown in Fig. 2, and more detailed 
data as to the properties of their waters are given in Table 2. 

Rawson and Moore (1944) studied brackish waters in Canada and re- 
ported that Chara spp. were prevalent at 0.01-0.78°, salinity. On the Little 
Salt Marsh in Kansas which Jewell (1927) studied, the salinity was 0.40- 
0.81°.. Moore and Jewell clearly indicated that members of the Charophyceae 
are tolerant of a wide range of salinity, and certain species such as those 
mentioned in this study can withstand great fluctuations. 


GRAZED SANDHILL CoMMUNITY 


According to Latta (1950), this grazed sandhill area occurs on an old 
beach ridge which in past ages formed the border of an ancient lake much 
larger than the present one. The ridge is 10-15 ft. high, 200-900 ft. wide and 
located in the grazed portion of the marsh, in sections 3, 4, 5, and 6, T.21S., 
R.11W., which form the southern border of the salt flats, and also the eastern 
border of the marsh. 


A total of 74 species of vascular plants were found in this area, 35 of which 
are limited to this locality. Some species blooming in the spring (Fig. 17) 
were Callirhoe involucrata, Cryptantha crassisepala, Festuca octoflora, Mira- 
bilis albida, Pyrrhopappus grandiflorus, Hordeum pusillum, Monarda citrio- 
dora, and Oenothera laciniata var. grandiflora. 


During the early summer Argemone polyanthemos was the most con- 


THe VEGETATION OF Tue Bic Sart Marsu 21 


spicuous flowering plant (Fig. 18) and provided a definite seasonal aspect. 
Other plants present at this time and characteristic of a later aspect were 
Eriogonum anuum and Conyza canadensis. 

The following species bloomed through the summer and early fall: 
Haplopappus ciliatus, Haplopappus divaricatus, Amaranthus tamartscinus, 
Ambrosia psilostachya var. Coronopifolia, Cenchrus longispinus, Eragrostis 
cilianensts, Salsola kali var. tenuifolia, Cycloloma atriplicifolium, and Pas- 
palum ciliatifolium. 


Hummock CoMMUNITIES 


Hummock vegetation occurred in many of the community types dis- 
cussed earlier, but since in most cases the majority of species found growing 
on hummocks in the various communities were similar, all will be discussed 
in this section. The hummocks contained a combined total of 56 species, six 
of these only on raised areas. One point to be made about the hummocks is 
that some were natural and others man-made, but this had little effect on 
the vegetation which finally occupied them. 

After studying the various plant communities on the marsh, it was 1m- 
mediately apparent that besides the horizontal distribution described earlier, 
there was also a vertical one. This vertical distribution was not very exact 
for any of a multitude of 56 species were found on the hummocks. One ob- 
vious fact is that if the hummock were present in a non-saline or slightly 
saline area, only sandhill species occupied it. If it was in a more saline area, 
the species closest to the margin were those more tolerant of high salinities. 

Two low hummocks of 1-3 feet occupying areas close to or on the open 
salt flats, will be mentioned first. One was dominated by Sporobolus airoides 
which occasionally occurred on the open salt flats. Fig. 20 illustrates this 
hummock type and a full discussion of the soil relations at this site are in 
the quantitative description of this community. A second type, also of interest, 
occurred in an area where D. stricta-S. depressa was the main vegetation 
type, but species such as S. pectinata, H. yubatum, P. arida, and S. paludosus 
were present also but scattered. This hummock type, found in only two 
places on the marsh, was completely dominated by Baccharts salicina (Fig. 
19). Soil analysis here indicated a salinity of 0.50°, in the surface 10 cm. at 
the top of the hummock and 0.20%, at 60 cm.; the chloride content of these 
two samples were 0.23° and 0.11°% respectively. This, as well as the analysis 
of the S. atroides hummock described later, indicates that in such low hum- 
mocks the salinity factor is still an important element in plant distribution. 
Tamarix gallica was found occasionally also occupying low, sandy hum- 
mocks. 

On the higher sandy hummocks, where the salinity range was the same 
as that of the Beach Ridge and Prairie communities, there was a slightly 
higher salinity at the lowest margins. A community of S. airoides was the 


22 Tue Universiry ScteNcE BULLETIN 


commonest vegetation surrounding the base of these, and only when a hum- 
mock was in a completely non-saline area was this species usually absent. 
Upward on the hummocks were found species tolerant of only a slight 
amount of salinity, such as Aster ericoides, A. exilis, Atriplex patula var. 
hastata, Haplopappus phyllocephalus subsp. annuus, Bassia hyssopifolia, 
Muhlenbergia asperifolia, Polygonum ramosissimum, S. pectinata, Setaria 
geniculata, and Sporobolus texanus. 

The remainder of species on the top of the raised areas are for the most 
part intolerant of high salinities and include such non-halophytic taxa as 
Achillea lanulosa, Amaranthus tamariscinus, Artemesia ludoviciana, Ascle- 
pias verticillata, Bromus japonicus, Callirhoe involucrata, Celtis occidentalis, 
Cenchrus longispinus, Euphorbia dentata, Gaura parviflora, Helianthus an- 
nuus, Lepidium virginicum, Panicum scribnerianum, Panicum capillare, 
Panicum virgatum, and Sorghastrum nutans. A more complete listing of 
species on the hummocks and their phenological relations are in Table 8. 

The species on the tops of hummocks had no regular distribution, and 
any of them might be found on any hummock studied. A species such as 
Euphorbia dentata might have little or no significance on most hummocks, 
but then for some unexplained reason it comprised the greatest part of the 
vegetation on others. This was true for many of the species discussed, and 
it would be extremely difficult to state exactly which species should be con- 
sidered the dominants of this community type. 

The following is a list of the species limited to the hummock sites, but 
these should not be considered community indicators because in each case 
they occurred on only a few hummocks: Baccharis salicina, Celtis occt- 
dentalis, Scutellaria latifolia, Solanum nigrum, Sporobolus flexuosus, and 
Sporobolus texanus. 


LIFE FORM 


The life-form spectra provided in Table 9 for the various marsh com- 
munities is based on the percentage of species representing each life-form. 
As mentioned by Chapman (1960) and Cain and Castro (1959) this method 
provides an indication of the life-form groups present but not the quantita- 
tive relations of species within the various communities. Since the next 
section of this paper is devoted to a quantitative study of the marsh com- 
munities, no mention of these relations is necessary here. In all communi- 
ties mentioned in Table 9 for which species names are given in the com- 
munity title, the species cited are the most abundant and their life-forms 
dominate. 

As Raunkiaer (1934) found in his study of Fano, the therophyte life-form 
appears to be dominant in the pioneer stage of the marsh, but in areas of 
lower salinity, hemicryptophytes, geophytes, and other life-forms are of 


Tue VEGETATION Or Tue Bic Sarr Marsu 23 


Taste 8. Season and zonal distribution of the vegetation of the Big Salt Marsh. 
A—Prairie, B—Tall D. stricta Meadow, C—D. stricta-S. depressa, D—S. pectinata-D, stricta, 


E—S. americanus-E. rostellata, F—Grazed Sandhill, G—S. airoides, H—Hummock, I—Fresh 

Water Seep, J—Saline Pond, K—S. depressa, L—S. paludosus-D. stricta. Life Form: Th— 

Therophyte, G—Geophyte, H—Hemicryptophyte, Hh—Hydro-helophyte, P—Phanerophyte, 
L—Liana. Phenology: A—Flowering or Sporulating, F—Fruiting, X—Presence. 


PLANT. DISTRIBUTION 


Life SEASONAL ZONAL 

SPECIES Form M JN JY Age S 0 ACB CG DEE EG HI skeie 
Achillea lanulosa .|........-------- H 7 eG RM i ie || > nee ree RR Caney RM eee 
Agropyron smithit .....-...+----- H ) ey Ue eee, we ||", ae ne, CER, © Fen Br 
AS ROGLS. (AGT Tere H PAG e Ar wey 5 eee Rar, ere ao te SE Soe 
Agrostis hiemalis ..........------ le Negi Pap S: < — EUMRe «See a) Xe eee So is ee 
PAUISTIVGMITIUIGIC) -22 css acncece-cee He Sih er cA 2cAU eR oe Par a2? Se Kae ee mee eeneee eee 
Allium mutabile ..............- G AVE = ae xXee oes 
Amaranthus tamariscinus .... Th ee. ee og) can! | oe Een Ee ed ae a 
Ambrosia psilostachya .......... G SP Ans FAC ce tA TE XG Ne Nec Nie 
var. coronopifolia 
Ambrosia trifida ..........-..-.-- Th ee, Geeey Ae A 2 ||P EXC lee We. anes Ae teeters nes 
Amorpha fruticosa ....--....--..-- Ph AS ATER) AES cas Mane Oe RK cera eee eins 
Andropogon gerard ...........- H 2k et, CAC AGBAIR (HEX Re eX ee eee 
Andropogon hallit ............ H eh TF cS oe pt ae ets BARS | Xie Fe 2c, Rea ge tes ne 
Andropogon scoparius ......... H Pog Be BOR at Ae VAR OXGS oe, Ee Ooi ey cS gan 
Androsace occidentalis _........ Wht VARY © 2 ey a cs, ee PERE eee 
Apocynum sibiricum _...... H IND ING IN Be OG 2K Saeco a 
Argemone polyanthemos ...... Th BA “AC AWAP AE Se, Joby 35 oes ae ne ieee 
Aristida oligantha ................ Th . AF AF F ae xX 
Artemesia ludoviciana .......... H Av’A A A Xe eee xX 
Asclepias incarnata .............. He rig Awe Amer Ares eed. 2D .6 me 
Asclepias speciosa ........--.------ H Aw Aly Ay VAR EF xee XG 
Asclepias tuberosa ...........---- H A a Xo these 2 
Asclepias verticillata ............ A BeAr, APTPeAR eA mee ce. (XG ee ree ee cca Nee ee coe ee 
Asclepiodora viridis .............. H Ate ke wees ee is os aR TG le Neal Ae PE 
ASter EVICOIDES 22.2... .oeeoeen-nee H IN IN INE ID en OKO 
AMSVEP CEE ED Th PO AeA AED XG eg BAIN, Grays tt 
Atiiplex argentea ...............-.. Th Pies ee ee. Se se OE ete Ane 8. Ten 
PAEEDI EX PALUNG <..0.022<.00..-<+00- Th oie ame, CANE PSU ATE TE IGG NG oe a eG 
var. hastata 
Baccharis salicina ...............--- Ph See ee AS a Khe eee SPE etree Sela soe, Cee ae 
Bass:a hyssopifolia ..............-. Th ee A ARS Eire BO), Cmte” omnes Go); Gee ae Sra: 
WBCLULOADUSINIG: Sco o occ scc cee He eo NEA Pale PR ey eee ee eae ne eer 
Bouteloua curtipendula ........ H ee Aig ihe Se Se ht Cree SR emeee PT 
Bouteloua gracilis ............--.. H re 8 TAB CAR SAR gx. Be Ek Tee ares 
Bromus japonicus ..........-------- Meh AGAVE ORE gh eX ae a ar NG ates 
Bromus tectorum .........--..---- th Ak TE = ee, |. eee eee. ee eer eee 
Calamouilfa gigantea ...........- H oy ee th eA ABRs TE XG oe es ee Sins, ee uate 
Callirhoe alcaeoides .............- H AUTEAG = X . 
Callirhoe involucrata ........--.- H AVA AR EF Xx a Sa < 
Callitriche heterophylla ....... Hh F = 23>. Ge: 


24 Tue University SciENCE BULLETIN 
PLANT DISTRIBUTION 
Life SEASONAL ZONAL 

SPECIES Form M JN JY Az S © || A.B GD EE Gieia eke 
Capsella bursa-pastoris .......- ‘Tela NE = Oe 6: 
GOL CLAD NCCLAN Sere ee H Ix 8 x = Ok es 
var. xanthocarpa 
GarexuDTeyi0le H ACE a Tage 2S obs 
Garexbuxbaumi AeA ead aXe 
Carex lanuginosa __.........---.- Hy ARE XeE er 
COLCATICA ie H ae AX ae 
Carex praegracilis ..............-- lal UAE Ee 7» eee Ee, Gack 
Carecmstipataae e H F ae a eS 
Gassia fasciculata 22.2.2 Th oat, ENA WA A 3 | OXe 7: Beds 
Celtis*ocadentalis, =. Ph XO SE OPA PX EX -. ee 
Cenchrus longispinus ......... Th = ACAI) (AR Ey |X oe XX 
Cerastium brachypodum  ..... Toe ANI Xi: as 
Chenopodium leptophyllum .. Th SARs Geseee aC 
Chloris verticillata ............--. H . AF AF AF AF a Xer 
Cicuta maculata _..........--.... Th 2 ACA Bp. xX 
Cirsium undulatum _............ H A AF F XxX XOXGRIEX 
Cirsium undulatum .|............. H AWE oe 
f. album 
Cleome serrulata ................-- Th . A A AR aes 
Cleomella angustifolia _........ Th AAS SALA Fee deg. 
Commelina erecta ........---.-.-- H AG AGA mL xe X 
var. angustifolia 
Conyza canadensis ........-------- Th me AWARD 8B a exe Xi EXOXG 
Coreopsis tinctoria .........-..--- Th eft: TAS AKie y= > ae Pee 
Cornus drummondu _.......... Ph XX xxx Xe 
Groton capitatus ==. Th Aten ae Xai 
Croton glandulosus .............. Th A xX . Xx 
var. septentrionales 
Cryptantha crassisepala —.... Th . AF LGR 
Cucurbita foetidissima ........ H Are GAT = xe 
Cuscuta glomerata _............. Th ee) ee SEXES: 
Cycloloma atriplicifola —..... Th aes F ae ASX 
Cyperus acuminatus ...........- Th A AF X 
Cyperus filiculmis _............... G A AE = x} Pe 
Cyperus strigosus ....--2.-------+- H SAR CAESAr Me ¢ XX: eae 
Delphinium verescens ......... H A A > ae aes 
Descurainia pinnata .........-...- Th AF ee pie, - %! Xe 28 
Desmanthus illinoense ......... H By ASAI EES XG X X 
Digitaria sanguinalis _........... Th ZN NY NIE AN | ORs 
Distichlis” stricta vecaececce----:2- G A A AF AF AF X KX KX Xeree x 
Echinochloa crusgalli ............ Th AF AF AF Xo Xer 
var. mites 
Eleocharis actcularis ............ H . A AF F = EE 
Eleocharis macrostachya ...... H A ADE Aas Xe 
Eleocharis rostellata. ............ H A AF s)he ae REX eX ae 
Eleocharis tenuis .............----- H A AF F X ag OX See 


Tue VEGETATION Or Tue Bic Satt Marsu 


25 


PLANT DISTRIBUTION 


Life SEASONAL ZONAL 

SPECIES Form MeN JY. AisoS Om PAT B CDE FIG Eloy eK Ie 
Elymus canadensis. ............-- H A AF AF 2 met... tet er feed 
Elymus virginicus .....-.-------- H A AF F D Ng cree 
Equisetum laevigatum ........ A a OXG g. 
subsp. funstoni 
Eragrostis cilianensis .....-....- Th . AF AF Xe: 
Eragrostis oxylepis ..........------ H - AF P.S 
Eragostis spectabilis _............. H AR See i, Soe 
Eragrostis trichodes .............. H = 2 PAE PAIR eX. ear ee 
Erigeron philadelphicus ...... qDy AN a ES. ee 
Erigeron strigosus ....-...------- ‘gy eX AN 1D Xa rah aX 
var. beyrichi 
Eriogonum annuum _.........- Th A AF AF Xz eX ee 
Eupatorium perfoliatum ...... H in de a ¢ EK 
Euphorbia dentata .............. Th 5 ge ut ue o. Pa, Coe 
Euphorbia glyptosperma ...... Th ACT AE ee |e Ey eX ees 
Euphorbia marginata ......... Th . A A AF AF |X .. XOX 
Euphorbia missurica ........... Th . AF Xie 
var. intermedia 
Euphorbia obtusata ............. Th Nie alee >.> ae 
Eustoma grandiflora _.......... H A -A AF AF AF . XX 
Eustoma grandiflora ............ H A Xe 
f. fischeri 
Festuca octoflora .................. Th AC are oe = xX . BED Gee 
Fimbrystilis castanea ..........-- H ce A iy Le ee ee Pe 
Flaveria campestris .............. Th a Ae! ce XX .. Xx 
Frochlichia campestris .......... Th 2 IX Sete JN FAN | OC Br 
Galium GPATINe —......-c.--------- Mohs AWE) 2s: ot ENG Fete ea 
Gaura parviflora _..........-.-... Th A <A AF AF Xe yea eX = 
Gerardia tenutfolia .............. Th . AF is DG 
var. macrophylla 
Glycyrrhiza lepidota ............ H AVE i Xa Nee seme eX OX 
Gnaphalium obtusifolium ... Th 1 a eNie 
Haplopappus ciliatus ......... Th A <A AF AF | X . KOXe 2: 
Haplopappus divaricatus —.... Th A CA VAE AB ex = XEEXG ee 
Haplopappus phyllocephalus Th SAR AE >a 
subsp. annuus 
Hedeoma hispida .................- Th INS IN = sa ee 
Helianthus annuus —............ Th AP UN ME || OK aXe: 
Helianthus maximilliani ...... H “ A A AF Xs Soke 
Helianthus petiolaris ........... Th A A A A AF AF eet, Oh. ae an? 
Heliotropium curassavicum .. Th AAS AAS aA Fc? Gn © 5 ox 
Heterotheca latifolia ............ Th A oe ars 
var. mcgregoris 
Hordeum jubatum _.............. H A AF F Xx X X eXGeX mx 
Hordeum pusillum __.........- Thi Ay Need || se eens =5 OK 
NBO) CPL) a eee Th Py Ae AR En eX ee eX 
WURCUS ANLETION 2-2-2. -o----- He A AF AF F Xie. XN ».4 


26 Tue University ScIENCE BULLETIN 


PLANT DISTRIBUTION 
Life SEASONAL ZONAL 


SPECIES Form M (JN. JY -Ag2'S ‘O q-A BC DIE sEIGSEIe a ere 


UCRCUS OTN Cy ep emer eee He Be Ee ALLE pe ere we ae ae Dk ON 
Lactuca ludoviciana ............ Th : .. A AFCAF  .. |X .. 4... =e ee 
£. campestris 
EGCIUCGSCAV1O| Ga Th ey 22, Pee. t co es es 
Lappula redowskit —............... Th- AF’ ow ow = «=. = |X... 2... SR 
LEC CRSIGMORVZOLG Ciena H as oe gt lees we sete at cok en 
Lepidium virginicum .......... Th -AF AF. = = .. . | XK... .2 2. exe 
Leptoloma cognatum ..........-- H ee Ae ss ze ae are Or 
[iatnismelabratay pease G 2 «+A oe a cn EX 2. eee 
Daatrislan cola G Lee 2 2A ee: We dees ee ES 
Linaria canadensis ..............-- Th A F 2. & «. 2. | &... = 2 ae 
var. texana 

Lippia lanceolata ...........- pe al ee NN NN LX en. in OS we ye 


var. recognita 


Lithospermum incisum —...... H AF 2. 2. 2 %2 «1 | -X 2. 2353 ee 
Lobelia siphilitica .............. H Be SP een ee we er as RE 
var. ludoviciana 

Lobelia splendens _.............. H ea oe oe SLA Te Sa a A See 
Lotus americanus = AGT 7x sei ves ben SUL 
Lycopus americanus ............ H 2 ie WAL SATE fh ae we ese, Gn Oe 
Lythrum californicum .......... H a NENG) 
Marsileasuestita ass ee ee A ns fo Re? oe Ge ee 
Mellotus alba 7 2. Th os A AL) a Poa Ly VK 2 oe Se 
Mimulus glabratus ............ Pe tA 7 Ware, OMe ee ree Css 
var. fremontit 

Mirabilisalbida pe H = AF OF. 94.) 2 || % 22.2252 eee 
Mollugo verticillata .............. Th sot we AC A a Aes, ERS. Sees, 6. 
Monarda citriodora .........--..- Th = GA GF 9... 02 \||2& 225. 22 eo Xe 
Muhlenbergia asperifolia ...... H 2 oi. JAF AF UAF |i uk, 2. 2,202, Skee 
Myosurus minimus ............. DE RAUE ope eee ok ae ep oe, SSeS ee 
Oenothera lacintata .............. Th 
Ocenothera laciniata .............. Th 
var. grandiflora 

Ocenothera rhombipetala ...... Th oh Oe BS A Re EK SEX ee 
Oxglisestricia gee H A, OB ee ey eX Sa ee 
Panicum agrostoides ..... see H 3 eee BRAT 2 Sa. A ton 
Panicum capillare _.............. Th ~ «A CA AF AF IX =. .. «2 . xe eee 
Panicum lanuginosum .......... Hi GAR GF oc. eo NEM 2 oe ee 


> > 
> SP 
> 
ba 


var. fasciculatum 

Pamcum lanuginosum .......... H Fok 3) a. ©. Ree ERS... 3 Ke eee 
var. lindheimeri 

Panicum scribnerianum. .......- Fg CAGE SE St ee ae le — kes 
Panicum virgatum «0.00.0... G ~  « ~ACAFS (F° 2. | X -2-. 2X) See 
Parthenocissus quinquefoha .. L ee Oe Foe ae re 22> go et ae deeeewes 
Paspalum ciliatifolium  ........ H . AF AF AF AF AF |X 2.5 SX eS Se 
Petalostemum purpureum ... H =) ow BA «2 “ee 4 EX 2S eee eee 
Phleum pratense .........--..--0- H ARS Le. Le = to eat ce es OX aes 


Tue VEGETATION OF Tue Bic Sart Marsu 


27 


PLANT DISTRIBUTION 


Life SEASONAL ZONAL 

SPECIES Fom M JN JY Ag S 0 | ABCDEFGHIJ KL 
Physalis virginiana .......-..--.- H AAG XY wan. vlop ts Wert oon Pega ene 
var. hispida 
PIGntagO PUTSAIL <...----.--2.----- thy ASABE > eae 2p. 
Pluchea purpurascens .......... Th Se A cK ee Ea eae eer are 
Pon CRUE a eee H AF F SKE IN DRG KG ye XG 
IROWUMPIGLCNSIS: <2: 5.2 2e.c2e-22e--- G A - eee 
Polygonum convolvulus ...... Th A AF A OK 
Polygonum hydropiperoides He . AF = Sop Ga ae: 
Polygonum lapathifolium .... Th be oe A Re at EY 
Polygonum pensylvanicum .. Th . AF x L 
var. laevigatum 
Polygonum ramosissimum . Th = . ABR ABCA ag. Oe nese haa). ee .¢ 
Populus sargentii .............--.-- Ph A ie eX 
Portulaca oleracea ...........-.--- Th A F SP Sree 
Portulaca parvula ..........--.--- Th epee, Acme: XS 
Potamogeton foliosus ......... Hy AAT AGE iar 
Prunus angustifolia —............ Ph Xx xe: a 
Psoralea digitatum ............-..- H Tels oo ek bone Ee Xnae 
Pyrrhopappus grandiflorus .. H ACE Xor 3 PE eS 
Ranunculus sceleratus ......... Th A Lee EX 
Ratibida columnifera —..........H ale OB xX Eee 
IURUS TACICANS, —..co..c.--o--cenn-n-- 1b A AF F exe 
Rosa arkansana ........2..2------ Ph A ae 
var. suffulta 
IRaumex alluvius .......-.0------- H ae rr: FOr neha 
IR CLEIGTUSDUS _—<..-ccnneeeeecnnen2 H A AF AF .. ae OS. PX 
Rumex maritimus .........-..... Th F Xia 
var. fueginus 
Sagittaria cuneata ...............- He AS Sian Bee 1. ae Peete 3 
Sagittaria latifolia ...............- He A AF AF Xam EX) a et 
WGCOTNIa TUDIA ...........-.------- Th - XG Gee x aS ae. ie 
Salix amygdaloides .............. Ph A apc 
SARS? ORATIO ee Ph A ae 
SUES CIP er eee ee ne Ph A > th EVEXGh Se 
Sal gall, 12GB Th . AF ee 
var. tenuifolia 
SVL AMCES Tt H A xe 
var. grandiflora 
Sambucus canadensis ............ Ph OUR HES, St Pe ee Xa oe 
Scirpus americanus .............- He A AF AF F XOX. me XG 
Scirpus atrovirens .........-...--- He = ANE ope 
Scirpus lineatus ._..........----.-- He + WN AN Pe er Ce Len oe 
Scirpus paludosus ................ He A AF AF F ID SED.E AOR Me Xe oi se 
ISGIGDUS UGUAUS 2... ccc2cece-n2e ists NINE TR SREY oe Xe 
Scutellaria latifolia ............... H ae A = me ee > Geoseas ae 
Sesuvium verrucosum ......... Th A A A Pie ex 


28 Tue University SciENCE BULLETIN 


PLANT DISTRIBUTION 


Life SEASONAL 
SPECIES Form IM aNe Ares 70 

Setaria geniculata ......... ...... G A AF AF AF 

Sclaviah glauca ere Th AS 18 

Silene santirr hing eee Th =) aE 

Sisyrinchium angustifolium .. H IN EN 

Solanum nigrum. |....--..--------- Th Sire mee™, Fe 

Solanum rostraturt .......-.-- Th IX Yt Ne 18 

Solidago altissima ............... H ss A AF 

Solidago mussourtensis _..... H AVAB SAR = 

Sorghastrum nutans —.......... H = JAAR AE 

Spartina pectinata ...............- H _ A AF ARF 

Specularia perfoliata —....... Th . A 

Spermolepis inermis ...... libeeeA ee 

Sphenopholis obtusata ......... H . AF to MA: 

Spiranthes cernua ...............- H = A A 

Spirodela polyrhiza Hy Pe .G enka 

Sporobolus airoides .......... H AC CAR CAR OR == 

Sporobolus asper ....-.------------ H AF AF 

Sporobolus cryptandrus —...... H - . AF 

Sporobolus flexuosus —........ H ae ee 

Sporobolus pyramidatus ...... H SAS ee 

Sporobolus texanus ............. H . AF AF AF 

Strophostyles letosperma ...... Th AAA 

SUGCUG A EPTESSA ee Th Sys GAR AS CA CAE 

Mamarnxapallicage- = Phy eA AA 

Taraxacum officinale .......... H A 

Teucrium canadense ...........- H EPA 

Thelesperma megapotamicum H Ai. Ee -2 

Tradescantia occidentalis _.. H A AR SE 

Tragopogon major ..............-- Th AF ne As 

Tribulus terrestris .......-...---- Th = SAR 

Tripsacum dactyloides .......... H ASR 

Typha angustifolia .............. He A AF 

Typha latifolia _ 2-22-22. He A AF e 

Witicamdioica ee eee H = = AF 

Verbena bracteata .............--- Th AAW 

Verbena fastata 2 svrc.cce! H . A <A AF 

Verbena ‘stricta geste H A A A AF .. 

Vernomia fasciculata ...........- H # AeA AA 

Veronica peregrina ........----- Th AF 

var. xalapensis 

Vicia angustifolia ...........-..- Th = A 

Viola kitatbeliana .........2....-- Th AF 

var. rafinesquit 

Viola papilionacea ............---- HAE = 

Vals riparia es I NS 18 oS seer ee 

Xanthium pensylvanicum .... Th SA eer 


mM PPM: 


MP 


ZONAL 
ABC DE F GHmmmeen 
xX... 
<2 
a 
Xt ape 
_ ee 
. Xe 
ke 
_ ok 
X Kel 4 =e 
xe? 
ee 
BPcr AS 
mee ee 
ee... 
eee 
Xe 
Perey <a" 
kiana 
ee 
a ne 
XX XK eer 
ok 2 ose 
Le 
PG. 
Ge 
oe 
ex gabe 
XxX. eee 
1X ae 
ee 
XX. 
= ae 
Xe 
ee 
BO. eu 
Xe 
Oo acne 


THe VEGETATION OF Tue Bic Sarr Marsu 29 


importance. The Suaeda depressa Community is the pioneer on the open 
salt flats and, as noted in Table 9, was made up entirely of therophytes. In 
the Spartina pectinata-Distichlis stricta area, the hemicryptophytes and the 
hydro-helophytes comprised the largest number of species, 46 and 29%, for 
each group respectively. In all of the remaining communities in which 
Distichlis stricta, a geophyte, was one of the dominants, the largest number 
of species belong to the hemicryptophyte and therophyte life-forms. The 
Saline Pond Community contained only two vascular plants, Eleocharis 
acicularis, a hemicryptophyte, and Potamogeton foliosus, a hydrophyte. 
Sporobolus airoides, also tolerant of high salinities, is a hemicryptophyte 
which dominated a community containing a 46° hemicryptophyte and 36%, 
therophyte complement. 

In the dry non-saline areas such as the Sandhill Mixed Prairie, Hummock, 
and Grazed Sandhill communities (Table 9) the hemicryptophytes and 
therophytes dominated. The high proportion of therophytes, 59° in the 
Grazed Sandhill Community, was due probably to the disturbance caused 
by grazing, which opens new areas for these annuals to invade. In the 
Scirpus americanus-Eleocharis rostellata Community the two dominants be- 
long to the hemicryptophyte and helophyte life-form groups and the largest 
number of species, 537., are hemicryptophytes. The Fresh Water Seep, quite 
similar to the S. americanus-E. rostellata Community, was dominated also 
by hemicryptophytes. 

A summary of the life-form relations on the entire marsh indicated that 
the largest number of species belong to the hemicryptophyte and therophyte 
life-forms. 


QUANTITATIVE ANALYSIS 


The characteristic communities on the marsh were more intensively 
studied by use of a modification of the line intercept method of Canfield 
(1941) in which only the plants hitting the line were counted, and the basal 
area was measured as a percentage of the line covered. These transects pro- 
vided quantitative information as to relative density, basal area, and fre- 
quency of the species. Transects were set up in all major vegetation zones 
and in transition zones between them. After several experimental attempts, 
a three meter line transect size was chosen for use in this study. In order to 
check the accuracy of this method, the transect length was always doubled 
to see if any new species would be added. In none of the transects was any 
new species added by doubling or tripling the length. 

The transects were distributed more or less uniformly over the study 
area. This method, employing a regular distribution of transects, was 
chosen because an attempt was being made to describe the entire area, and 
it was felt that a random distribution of samples might give a bias to one part 
of the study area over another. 


30 Tue University Science BULLETIN 


Tasve 9. Life form spectra of the marsh communities studied. 


Hemi- Hydro- 
Community No. of | Panero-  crypto- Geo- Helo- Thero- —_Lianas 
Species — phytes phytes phytes phytes phytes Ys 
%o Ws %o % %o 
Mall aectricia: =e 13 2 23 8 15) 54 
S701 CS eel e. 46 18 > 36 
Sandhill Mixed Prairie .... 110 1 47 if 1 44 
Sed CDi CSSG) ee 3 s os Bs: = 100 
D. stricta-S. depressa ........ 10 10 30 10 10 40 
S. paludosus-D. stricta ...... 9 &. 22 11 11 56 
S. pectinata-D. stricta ..... 24 4 46 8 29 13 _ 
S. americanus-E. rostellata 88 9 48 6 ly 17 3 
Fresh Water Seep ............ 30 ig 53 2 30 17 
Saline” Pond a=. ey ee, = 50 ee 50 
Graded ‘Sandhill — 2-2. 74 1 37 3 ie 59 
Elummock ee 56 5 45 5 6 45 
Total Marsh Flora ...... 241 5) all 4 8 41 1 


To check the quantitative analysis, the abundance estimate analysis of 
Braun-Blaunquet (1932) was also used, but this only corroborated the 
transect analyses. In most cases species with a rare or very rare ranking were 
those that did not occur on the transects, whereas all other estimate classes 
were represented. 


Suaeda depressa COMMUNITY 


This community borders and invades the bare salt flats in many areas 
of the marsh. The principle species, S. depressa, was found in both its de- 
pressed and erect forms (Fig. 7, 8). The ten transect analysis made in July 
are summarized in Table 10. In other parts of the community, succulents 
such as Sesuvium verrucosum and, in one isolated area, Salicornia rubra were 
encountered. 

This area had the lowest basal area, 0.06%, of any community on the 
marsh, and the plants were very widely scattered, being 1-40 meters apart. 
In 70°, of the transects no vegetation could be found on the three meter 
transect line; however, in these areas the vegetation was widely scattered and 
15-40 meters separated individual plants. 


Taste 10. Analysis of Suaeda depressa community. Ten transect summary. 


Density Basal Area Frequency Height 
Species No. wh Sb cm. 


SV LAO GIG HRASTO) Ba ee oe 8 06 30 1-30 


Tue VEGETATION Or Tue Bic Sart Marsu Sl 


The transects which were made in this community at soil sample sites 


6 and 7 are delineated in tables 11 and 12. 


Tasce 11. Transect analysis at soil sample site 6. 


Density Basal Area Growth 
Species No. OF Form 
Swonedla GERGOSGD, seeccep acco se oe 2 16 depressed 


At site 6 the soils in the surface 10 cm. had a pH of 8.0-8.8, an average 
saturation percentage of 40°, a low organic content of 3.67., and a soil 
texture classified as a sandy, clay loam, whereas the soil texture at the 60 cm. 
level was sand. The total salinity at the site averaged 1.92%, and the chlor- 
inity averaged 1.11%, in the surface 10 cm. 


Tasce 12. Transect analysis at soil sample site 7. 


Density Basal Area Growth 
Species No. es Form 
SABO OG GROSSE, rece eee Pa por 4 40 erect 


At site 7 the pH was 8.0-8.7, the saturation percentage 40%, the organic 
content 6.8°%, and the soil texture a sandy, clay loam in the surface 10 cm. 
and sand at the 60 cm. level. The total salinity in this area averaged 1.70% 
and chlorinity 1.05°%. Tables 3-6 provide a fuller account of the soil relations. 

The relation between the prostrate and erect form of Saaeda is still ob- 
scure. Through a large part of the growing season it appears, as evidenced 
by the soil samples, that the prostrate form occurs in the more saline areas; 
however, an observation in September 1960 modified this assumption. At this 
time recently germinated forms of Swaeda had begun development on the 
open salt flats and in the open areas of its former distribution, but all forms 
were found to be prostrate whether they developed in an area where pre- 
viously there were upright forms or on the formerly barren salt flats. 


Tatu Distichlis stricta MEADOW COMMUNITY 


Bordering the eastern margin of the salt flats were three major zones of 
vegetation. The first, a Tall D. stricta zone was separated from the flats by a 
man-made ridge in the northeastern part of the marsh, but it bordered the 
flats or the S. depressa Community in the southeastern and southern parts of 
the marsh. The ridge, 3-5 ft. high, supported many weedy species including 
Tamarix gallica, Bromus japonicus, Cenchrus longispinus, Melilotus alba, 
Gaura parviflora, Chenopodium leptophyllum, Lactuca ludoviciana and a 


Us 
No 


Tue University ScieNcE BULLETIN 


host of others, whereas, the Tall D. stricta Community contained none of the 
above species and had a much smaller number of species. 

Actually this community contained only 13 species of vascular plants, 
five of which were present in the spring and early summer, including D. 
stricta, Scirpus paludosus, Poa arida, Hordeum jubatum, and Suaeda de- 
pressa, whereas the remainder included eight species of extremely rare oc- 
currence which appeared in late summer. Even though this community had 
only 13 species, it had the highest basal cover, 12.25°., of any community on 
the Big Salt Marsh. 

A ten transect analysis of the vegetational relations made in June in this 
community is in Table 13. In this zone (Tables 13 and 15), D. stricta was 
very evenly distributed and occurred at all transect sites. It comprised 94.46%, 
of the total vegetation and had a basal area of 11.35°% or 92.65%, of the total. 
S. paludosus, found in scattered colonies, had a basal area of 0.89°% or 7.27% 
of the total, a frequency of 30°, and a relative density of 5.38%. 


Taste 13. An analysis of the Tall Distichlis stricta Meadow Community. Ten 
transect summary. Total density, 2529. 


Relative Density Basal Area B.A. % Total Frequency 
Species Yo YS, Hs Ye 
Distichlis Sir1ciay =. ee 94.46 NILES 35) 92.65 100.00 
Scirpus paluidosus 2 5.38 89 72d, 30.00 
ROURAGHTG Cena e e eee 16 01 08 10.00 
oS 


In a single transect, S. paludosus was the codominant and had a relative 
density of 48°, and a basal area of 5.80%, 75.30% of the total. This transect 
is described in Table 14 and the basal area relations are cited in Table 15. 


Taste 14. Transect analysis of area with S. paludosus-D, stricta codominance. 
Total density, 79. 


Relative Density Basal Area B. A. % Total Height 
Species Ve We Yo cm. 
ID AG HAAS (GHGS he 52.00 1.90 24.70 20-60 
Scinnus PaludosUs) =e 48.00 5.80 75.30 75-90 
7.70 


Poa arida comprised less than one percent of the total vegetation in this 
zone and with H. jubatum was very widely scattered to rare. 

The comparison of the basal area of species at the various transect sites 
(Table 15) indicates clearly that D. stricta was the dominant, having the 


5 


highest basal area at 90° of the sites sampled. Only at site 3 was another 


1S) 
Ur 


Tue VEGETATION OF Tue Bic Sart Marsu 


species important in the cover relations. Table 15 indicates S. paludosus was 
dominant and traces of it were found also in transects 6 and 9. The remain- 
ing species provided very little additional cover. 

The soils in the surface 10 cm. in the eastern section of the community 
at site 3 had a pH of 7.28.6, an average saturation percentage of 43%, an 
organic content of 2.5%, and a soil texture classified as loamy sand and as 
sand at the 60 cm. level. The total salinity was 0.29-0.63%, and the chlorinity 
0.10-0.31°% in the surface 10 cm. The vegetational relations at site 3 is indi- 
cated in Table 16. 

At site 4, in the western part of this community, the soils at the 10 cm. 
level had a pH of 8.0-8.6, a saturation percentage of 75%., an organic content 
of 6.2%, and a soil texture classified as a sandy, clay loam and at the 60 cm. 
level as a sand. The total salinity was 0.73-0.94° and the chlorinity 0.34- 
0.47%. Table 17 delineates the vegetation relations at site 4. Tables 3-6 pro- 
vide more complete data on the soil relations in the area, 

Soils in the area were covered with 2-6 inches of water during wetter 
periods of the growing season, and the combination of salinity and inunda- 
tion tended to prevent the invasion of species from the prairie, hummocks, 
and the adjoining western ridge. Dry land plants of these regions are elim- 
inated by the standing water, and the salinity eliminates non-halophytic 
aquatic plants. 


Tare 15. A comparison of the basal area of species at ten transect sites within 
the Tall D. stricta Meadow community. B. A. is expressed as a percentage of 
the actual cover. 


Transects 
Species 1 4) 3 4 5 6 7 8 9 10 
Distichlis stricta .....- 1907 90m 1900 7-20-1380) 15.00 18:00) 82020 970 owt 
WaIipuUs PUIUAOSUS =. 9 cess aces 528 (hae eee ee ace 190) PACE i ores 


Poet. GCI ee ee ee eee SQ, Gees. gc Aas ee ee ee eee 
7.30 13.80 16.90 18.00 20.20 10.90 6.74 


motalk, fee es 112908 930" 7270 


The fairly high basal area, 12.25°%, in the area provided a dense cover of 
living and dead plant material (Fig. 3) which did not allow invaders to 
enter, making it a more or less closed community. 


Sporobolus airoides COMMUNITY 


The main body of this community was on the eastern portion of the 
marsh between the Tall D. stricta Community on the west and the Sand- 
hill Mixed Prairie Community on the east. It also occurred on small, self- 
produced hummocks on the open salt flats (Fig. 20). There was a sharp 1-2 
ft. rise between this and the D. stricta Community and a very sharp change 
in the character of the vegetation. 


34 Tue University ScrENcE BULLETIN 


Tase 16. Transect analysis at soil sample site 3. 


Density Basal Area Height 
Species No. Ws cm. 
LD SStiGh ames ETi Cl pee 188 11.9 35-50 


Sporobolus airoides was the dominant species in the area, making up 
90.50%. of the density and covered 96.04°% of the total basal area. It had a 
very wide distribution and a frequency of 100°4. The remainder of the 
species in this community comprised only 3.96% of the total vegetation cover 
and only 9.50°% of the relative density. D. stricta also occurred in this zone, 
but it had a relative density of only 5.85%, a relative basal area of only 1.98%, 
and was found in only 40% of the transect sites. This was an abrupt change 
from the Tall D. stricta Community described earlier where it made up 
94.46% of the relative density, 92.65°% of the total basal area, and had a fre- 
quency of 100%. 


Tasce 17. Transect analysis at soil sample site 4. 


Density Basal Area Height 
Species No. We Cm. 
Distichlisustricta. sete ee a ea 251 13.8 40-60 


Another species in the area was Ambrosia psilostachya var. coronopifolia 
which had a frequency of 40°, covered 0.99°% of the total vegetated area, 
and had a relative density of 2.08°%. Other species were Aster ericoides which 
had a frequency of 30% and Elymus canadensis with a frequency of 10°. 
A complete analysis of the transects made during July and September is in 
Table 18. 

The basal area relations at all transect sites within this community is 
delineated in Table 20. The data clearly show the complete dominance of 


Taste 18. Analysis of the Sporobolus airoides community. Ten transect sum- 
mary. Density, 266. 


Relative Density Basal Area B. A. % Total Frequency 
Species We % %o %o 

Sporobolus atroides .............--..-.- 90.50 4.86 96.04 100.00 
DiSiChiSeSTtiCiG a = ee Sep) 10 1.98 40.00 
Ambrosia psilostachya ............... 2.08 .05 By) 40.00 
var. coronopifolia 

ASTCLMICTICOIA ES a 1.16 04 19 30.00 
Elymus canadensis ...............-.....- 41 01 20 10.00 


5.06 


Tue VEGETATION Or Tue Bic Sart Marsu 5) 


S. aroides at all transect sites. This species made up an average of 96.04%, 
of the total basal area whereas the remaining taxa covered only a negligible 
portion of the area and had a total combined cover of only 3.96% of the total. 

The vegetation relations at soil sample site 2 are given in Table 19. The 
surface 10 cm. of soil had a pH of 6.5-7.9, a saturation percentage of 34%, a 
loss on ignition of 2.2%, and a soil texture classified as sand. The salinity 
was 0.04-0.19°5 and the chlorinity 0.02-0.09°, in the surface 10 cm. More 
complete data on seasonal soil relations in this area are in Tables 3-6. 

At the site of Fig. 20, an example of this community invading the open 
salt flats, the salinity was 1.60% and the chlorinity 1.16° in the surface 10 
cm. of soil. The soil salinity and chlorinity analyses were made in July. The 


Taste 19. Transect analysis at soil sample site 2. Density, 30. 


Relative Density Basal Area BA. % Votal Height 
Species is Sh, es cm. 
Sporobolus airoides _............--.... 90.04 334 95.36 80-90 
Ambrosia psilostachya ................ 6.64 alg} 2.32 18-20 
var. coronopifolia 
EV INUS. CANGACNSIS, 222. ..0c.-nonnnneene-e 5,32 13 2.32 75-80 
5.60 


salinity and chlorinity content at this site were much higher than in the 
site described earlier where S. airoides occupied larger areas. Other species 
on the hummock were Poa arida and Suaeda depressa. 

The lack of a great deal of lateral translocation of salts from the Tall 
D. stricta Community to the S. aroides and Prairie communities is evident, 
for, as noted in Table 5, the former had an average salinity of 0.39%, the 
next of 0.12%, and the latter of 0.019%. 


SANDHILL Mrxep Prairie COMMUNITY 


This area formed the eastern border of the marsh and was covered with 
a Sandhill Mixed Prairie vegetation. The average basal area from eight 
transects made in June was 5.97% and varied from 3.04-9.13% (Table 22). 
Panicum scribnerianum, a short grass, made up 19.78°, of the total density, 
had a frequency of 100%, and a basal area of 0.89°% or 14.88°% of the total. 

Elymus canadensis, Agropyron smithu, and Sporobolus airoides were 
found in 25° of the transects made. A. smith had a relative density of 
7.28%, S. atroides, 6.24%, and E. canadensis, 0.48%. These species had a 
combined cover which accounted for 22.75°, of the total vegetated area. 

Tall grasses that play an important role in this community had the fol- 
lowing relative densities: Sorghastrum nutans, 847°; Panicum virgatum, 
7.28%; Andropogon gerard, 853°. In the same order these species com- 
prised 8.70%, 11.88%, and 16.73% of the total basal area. 


36 Tue University ScIENCE BULLETIN 


Tasie 20. A comparison of the basal area of species at ten transect sites within 
the S. airoides community. B. A. is expressed as a percentage of the actual cover. 


Transects 

Species 1 2 3 4 5 6 7} 8 9 10 
Soporobolus airoides 5.34 435 4.74 4:95 5.40 3.86 5.94 4:01 ) 4:95) 5el4 
Ambrosia psilostachya 13 oe aS) Cees: lle ea 13°90 
var. coronopifolia 
Distichlis esirictg = ee al 2 Pe 41 8 ee eee 32) | one 
NGO? CRUTCHES ome cue Rill Sas Sees AQ. yaa Alle epee 
Elymusscanadensis = 3m 0 02. (an ee ee ee eee 

(Potala ree 00) 46: 4.7 eZ eel) 4516) (6159) 1425 Ces Omer 


Bromus japonicus, an annual which invades mowed and grazed prairies 
in this area of Kansas, made up 27.97%, of the total density, had a frequency 
of 87.50°%, and a basal area of 0.86% or 14.38°% of the total. 

The taxa mentioned above along with other widely scattered grasses, in- 
cluding Eragrostis trichodes and Andropogen scoparius, had a combined 
relative density of 89.40°% and a cumulative basal area of 5.39°% or 90.15% 
of the total. The remainder of the vegetation was composed of other widely 
scattered grasses and forbs. Table 21 provides an indication of the quantita- 
tive relations of the vegetation in this area obtained from eight transects. 

A comparison of the basal area at all transect sites is in Table 22. It can 
be seen that the species making up the highest amount of cover varied at 
different transect sites, and no species played a dominant role in more than 
two. Panicum scribnerianum had a frequency of 100°%, and a basal area of 
0.26-1.85°, but in no transect did it make up the largest portion of the vege- 
tation. This species, however, must be considered one of the chief indicators 
of the lower strata of this community type because of its wide distribution 
throughout the area. Weaver (1954), in his discussion of upland species, 
stated that “because of its wide distribution and rather common occurrence, 
it is one of the most important secondary species.” 


The vegetational relations at soil sample site 1 are delineated in Table 23. 
At this site the surfac 10 cm. of soil had a pH of 6.3-6.9, a saturation per- 
centage of 28°, the lowest marsh organic content of 1.1%, and a soil texture 
classified as sand and also as sand at the 60 cm. level. The salinity in the 
surface soil was 0.003-0.02°% and the chlorinity 0.001-0.007°.. More complete 


data concerning the soil relations are in Tables 3-6. 


As mentioned earlier, the translocation of salts into this community from 
adjoining more saline ones seemed to be slight or almost nonexistent. One 
reason for this is that the sandy prairie soils allow rapid infiltation of rain- 
water, so that even if there were any lateral translocation the salts would be 
leached out. 


THE VEGETATION OF Tue Bic Satt Marsu 37 


Taste 21. Analysis of the Sandhill Mixed Prairie community. Eight transect 
summary. Density, 638. 


Relative Density Basal Area B. Ac) Dotall Frequency 
Species Te V6 We % 

BROMUS J4PONICUS. — <a a2 -22e--ennee- 27-97. 86 14.38 87.50 
Panicum scribnerianum _.....- 19.78 .89 14.88 100.00 
Andropogon gerardi _............-.. 8.53 1.00 16.73 12.50 
Sorghastrum nutans 8.47 52 8.70 25.00 
Agropyron smithit ...........----------- 7.28 67 teIE241 25.00 
Panicum virgatum .........-.--...-- 7.28 e/a 11.88 25.00 
Sporobolus airoides 6.24 67 EZ A 25.00 
Artemesia ludoviciana _. 439 BD 5.85 37.50 
Plantago purshit _..........---...2...-- 3.63 09 1.50 75.00 
Agrostis Aiemalis: 2... c----<-- 3.37 05 .83 12.50 
Ambrosia psilostachya —.... 79 05 .83 12.50 
var. coronopifolia 

Elymus canadensis .......2..2-...----- 48 02 335) 25.00 
Monarda citriodora —......... 41 008 13 25.00 
Achillea lanulosa _........--.--c--.------ 35 05 83 25.00 
Tradescantia occidentalis _........ 29 02 BE) 12.50 
Lepidium virginicum _......... 29 004 06 12.50 
O<GISRSITICIG® =a 25 01 16 25.00 
yigerOn StrigOSus _.........-2..-----.- 14 O07 aul 12.50 
var. beyrichu 

Wancius: interior... 2... sl2 003 05 12.50 


Scirpus paludosus-Distichlis stricta COMMUNITY 


In the north-central portion, and scattered irregularly in other more or 
less open areas of the marsh were colonies of S. paludosus with a D. stricta 
codominant. A complete summary of the ten transect analysis made in this 
community during June is in Table 24. This summary shows the average 
basal area in this community was 2.17°.. 


Distichlis stricta had a relative density of 63.6%, a frequency of 100%, 
and a relative basal area of 43.04°... The tall dominant in this area, S. paludo- 
sus, had a relative density of 31.80°., a frequency of 100°., and a basal area 
of 54.99°%, of the total vegetational cover. These two species accounted for 
95.40% of the total density and covered 98.03°. of the total vegetated area. 


Another species was Polygonum ramosissimum with a relative density of 
3.8%, a frequency of 60°%, and covered 1.84°% of the total vegetated area. An 
erect form of Suaeda depressa had a relative density of 0.8°., a frequency of 
10%, and a basal area of 0.13% of the total. Other rarer species were Poa 
arida, Hordeum jubatum, Heliotropium curassavicum, Atriplex argentea 
and A. patula var. hastata. 


38 Tue University SciENcE BULLETIN 


Tasie 22. A comparison of the basal area of species at eight transect sites within 
the Sandhill Mixed Prairie Community. B.A. is expressed as a percentage of the 
actual cover. 


Transects 


5 6 7 8 


Nm 
Ww 


Species 1 


Andropogon gerard: @e ee. CC Sl OEE eee 
Panicum scribnerianum ‘ 
Bromus japonicus ............. 10 26. © =: 33 1.31 L/D 36 2.80 
Agropyron smithii .............- ofS) 
Sporobolus airoides ............ 
Songhastrum: mutans 22. 293 3323) eee 
Panicum virgatum 
Artemesia ludoviciana ....... ...... 
Plantago purshit 0... all} .10 .03 5) Oe eee al 03; 
AprostiseAlemalist 2082.2 nue of ee I eee bobs 
Ambrosia ipsilostachya) 2-22) ee ee eee 46 
var. coronoptfolia 

Achilleaalanulosae =e Lee eee ee 10 sh) 
Elymus canadensis .............. aig} (06 Wate Re.” 


dhwadescantia occidentalis’ == 234) ee eee 
Oxalisinstricta, 2.20 Oe ee Ee rn ee 
Monarda cttriodora ............ ~..-- .03 03 05 


Erigeron strigosus ............ 


var. Beyrichit 
Lepidum virginicum 


Juncus interior 


The basal relations listed in Table 25 show the clear dominance of S. 
paludosus and D. stricta at each transect site over the remaining species. The 
basal area at the various sites was 1.41-3.06%, that of D. stricta was 0.35-1.20%, 
and the basal area of S. paludosus was 0.70-1.93%. P. ramosissimum had a 


Tasie 23. Transect analysis at soil sample site 1. Density, 43. 


Relative Density Basal Area B.A. % Total 
fe} oO 


Species é % 6 
A STOLL ONMSINIE hime ee 27.92 ‘> 24.75 
SOP ZRASITUIM EN ULATSae e e 23.29 93 30.00 
Panicumescribnerianum = 25.60 73 24.09 
PIGNLAZ OM DUT SA Ite ee 9.38 13 4.29 
BYOTBUS IG BONICU Sere eee 6.91 10 3.30 
Wradescantia occidentalis: 2 233 23 159) 
Elymus (canadensisa se ae ee en BedS aS A829 
TC DIdtUTURUINGI NICHI ee ee 2733 03 99 


Tue VEGETATION Or Tue Bic Sarr Marsu 39 


Taste 24. Analysis of Scirpus paludosus-Distichlis stricta community. Ten tran- 
sect summary. Density, 341. 


Relative Density Basal Area B. A. % Total Frequency 
Species es Te %o % 
DAS ANT ORO HET pean 63.60 94 43.04 100.00 
SEInPUS PAlUdOSUS <2...c2-0c2-eno-2--- 31.80 1.19 54.99 100.00 
Polygonum ramosissimum. ........ 3.80 04 1.84 60.00 
SHRALEE CHAT eee ee 80 003 13 10.00 
ZAWS 


Taste 25. A comparison of the basal area of species at ten transect sites within 
the S. paludosus-D. stricta community. B.A. is expressed as a percentage of 
the actual cover. 


Transects 
Species 1 p) 3 4 5 6 7 8 9 10 
SCI PUS: PGLUAOSUS .--.2..--2t-s-=- 1.93 1.06 .76 1.04 .70 1.07 1.60 .96 1.10 1.61 
IDISHAAG “Gigtg]® rere 113°) 235" FON 1 258e 1.20 12 20 ale eno 
Polygonum ramosissiMUM We ve ee OF 060 Gae lO 2 -10 
Songila CUROSI ao eS es eee 103° (fie) BE. eee ee eee 
1K tell be eeetene are eee ree 3.06. 1.40 297 1:72, 1.99) 2.25" 2.80) 2.46, 15852267 


fairly broad distribution in this area, but its basal area was low, 0.06-0.10°, 
at the various sites. 

The vegetational relations at soil sample site 8 are described in Table 26. 
The surface 10 cm. of soil had a pH of 8.3-8.6, a saturation percentage of 
47°/,, an organic content of 9.2%, and a soil texture classified as sandy, clay 
loam and as sand at the 60 cm. level. The salinity in the surface soil was 
0.27-1.10°% and the chlorinity 0.09-0.68°%. More complete data concerning the 


soil relations are in Tables 3-6. 


Distichlis stricta-Suaeda depressa COMMUNITY 

To the west of the salt flats and in large areas of the southern section of 
the marsh the vegetation was composed mainly of dwarfed forms of D. stricta 
and S. depressa. Figure 9 illustrates the dwarfed form and low basal area 
in this community. This community was broken up into five separate zones, 


Taste 26. Transect analysis of soil sample site 8. Density, 46. 


Relative Density Basal Area B. A. % Total Height 

Species Tés oe Tb cm. 
IDiStiCRIIS StriCh@) 202 2--oe-nne-e- 54.36 1313 36.93 12-30 
Scirpus paludosus .......-.--.------- 45.64 1:95 63.07 25-45 


3.06 


40 Tue University SciENcE BULLETIN 


and Tables 27-31 provide the information found in these analyses. Table 32 
contains a 70 transect summary of the vegetational relations. 

Distichlis was the most abundant species, having a relative density of 
86.56%, a frequency of 96.82%, and a basal area of 1.07% or 86.53% of the 
total. Swaeda depressa, the codominant, had a relative density of 13.13%, a fre- 
quency of 62.37%, and a basal area of 0.13% or 11.50% of the total. The re- 
maining species in the area, including Scirpus paludosus, Spartina pectinata, 
Sesuvium verrucossum, Poa arida, Polygonum ramosissimum, and others 
mentioned earlier comprised less than 1°% of the total density, had a basal 
area of less than 2% of the total, and a very limited distribution. 

Even though this community contained many of the same species as the 
Tall D. stricta Community, it could never be confused with it. The main 
reason for the distinction between the two is that this one had a greatly re- 
duced basal area and the species had a dwarfed growth form which also clear- 
ly distinguished it from all others. 


Soil samples were taken in four parts of the community. Table 33 pro- 


Taste 27. Analysis of D. stricta-S. depressa community. An 11 transect sum- 
mary. Zone A. Density, 295. 


Relative Density Basal Area B. A. % Total Frequency 
Species % ve So %o 
Disiichlis siricla ys 78.95 79 87.00 90.00 
SUCCUURU CDI ESSi 21.05 ali 13.00 63.00 
OL 


Tasie 28. Analysis of D. stricta-S. depressa community. A 12 transect summary. 
Zone B. Density, 320. 


Relative Density Basal Area B. A. % Total Frequency 
Species Ses Te eo % 
Distichliss stricta = 85.22 99 85.10 100.00 
SUGCUG™ CEPI ESSa 14.60 16 13.80 83.30 
Spartina pectinata ..................-.- 18 01 1.10 8.30 
1.16 


Tape 29. Analysis of D. stricta-S. depressa community. A 9 transect summary. 
Zone Cy Density, 342: 


Relative Density Basal Area Be An, otal Frequency 
Species Wes He %o %o 
Distichlts stricta 22.2 96.04 1.59 97.30 100.00 
Suacdandepressa) =e 3.96 .04 2.70 66.00 


1.63 


Tue VEGETATION Or Tue Bic Sarr Marsu 4] 


Taste 30. Analysis of D. stricta-S. depressa community. A 16 transect summary. 
Zone D. Density, 497. 


Relative Density —_ Basal Area B. A. % Total Frequency 
Species % vo %o % 
DEGRA: Gai een ee 78.53 1.18 84.85 93.00 
SAGE TUG URCOUE eee 20.84 16 12.02 75.00 
Scirpus: paludOsus, -......2.-..-2s00---- .63 04 Rall’ 12.50 
1.38 


Tasce 31. Analysis of D. stricta-S. depressa community. A 22 transect summary. 
Zone B. Density, 638. 


Relative Density Basal Area By Ay Wotal Frequency 
Species Ne Ho % % 
DD ISTIGH IIS *SETICLA\ .cscetee eee eee =na oe 93.04 99 83.90 100.00 
SHAUL CIRCE, Serene 6.52 aS Qe 7a 40.00 
Sainpus: PaludOSUs, <222.2.2.22-----=- Beta 04 339 9.09 
1.18 


Taste 32. A summary of transect data obtained from five separate transect 
analyses of this community. A 70 transect summary. Density, 2092. 


Relative Density Basal Area BUAS%,, Lotal Frequency 
Species We of, of, % 
ID iStiGhlas, SEVICEA® <..--.-c-.2222-2----c-2- 86.56 1.07 86.53 96.82 
Shae OG AAR 13.13 13 11.50 62.37 
Scirpus paludosus ......--.------------ 28 02 1.78 5.71 
Spartina Pectinata __...2.-.-00---+--- 03 001 19 1.42 
12241 


Tasie 33. Transect analysis at soil sample site 9. Density, 18. 


Relative Density Basal Area By Au, Lotal Height 
Species % we 6 cm. 
DDiStichlis StTiCtd ....--.c...-22-.-20--<--- 77.76 70 70.00 15-20 
SUGCHG (CEPTESSG. 2.0 teee teen. sncnnss 2222 30 30.00 25-28 
1.00 


vides the vegetational relations at sample site 9, in the eastern part of Zone A 
of the D. stricta-S. depressa Community. 

The soils at site 9 had a pH of 7.5-8.6, an average saturation percentage 
of 41°, an organic content of 9.3", and a soil texture in the surface 10 cm. 


of sandy, clay loam and sand at the 60 cm. level. The total salinity was 
0.49-1.55°% and the chlorinity 0.16-1.20%, in the surface 10 cm. 


42 Tue Universiry SciENcE BULLETIN 


Soils in the western part of Zone D were studied at soil sample site 11, 
and the vegetational relations are described in Table 34. The soils had a pH 
of 7.6-8.7, an average saturation percentage of 44°%, an organic content of 
11.2°., and a soil texture in the surface 10 cm. which was a sandy clay loam 
and a loamy sand at the 60 cm. level. The total salinity was 0.58-1.52% and 
the chlorinity 0.14-0.46%, in the surface 10 cm. 


Taste 34. Transect analysis at soil sample site 11. Density, 53. 


Relative Density Basal Area B; A. Y; Wotall Height 
Species Wes ie Yo cm. 
Distichlis stricta) pe. 86.72 1.84 88.89 17-21 
SUGCADNUCDI CSS pees aa ee S225 23 Te 15-20 
2.0 


Soils of zone B were sampled at site 15 and the vegetational analysis in 
this area is recorded in Table 35. The pH of the soils was 8.1-8.6, the satura- 
tion percentage 62°, the organic content 17.5°%, and the soil texture in the 
surface 10 cm. a sandy clay loam and sand at the 60 cm. level. The total 
salinity was 0.82-2.45°% and the chlorinity 0.46-1.57°% in the surface 10 cm. 


Tasie 35. Transect analysis at soil sample site 15. Density, 26. 


Relative Density Basal Area B.A. % Total Height 
Species of, of, VE cm. 
Distichls stricta 2 26 0.99 100.00 18-22 


The fourth soil analysis was made in zone E at site 13. The vegetational 
analysis 1s recorded in Table 36. The pH of the soils was 8.0-8.5, the satura- 
tion percentage 44°, the organic content 6.6%, and the soil texture in the 
surface 10 cm. a sandy, clay loam and sand at the 60 cm. level. The total 
salinity aws 1.51-2.52°, and the chlorinity 0.80-1.20%, in the surface 10 cm. 


A summary of the soil conditions in the area indicated that pH was 7.5- 
8.7, the saturation percentage 41-62°., the organic content 6.6-17.5%, and the 


soil texture in the surface 10 cm. a sandy, clay loam and sand to sandy loam 


Tas_e 36. Transect analysis at soil sample site 13. Density, 32. 


Relative Density Basal Area B. A. % Total Height 

Species oh o/s Wb cm. 
Disiiohiis™sizictaa es 90.63 1.06 91.38 18-25 
Suacdasdepressa: ae 9.37 10 .62 10-15 


Tue VEGETATION Or Tue Bic Sarr Marsu 43 


Tape 37. Results of a transect on the southwest border of zone E of the D. 
stricta-S. depressa community. Density, 92. 


Relative Density Basal Area B. A. % Total 
/ ° 


Species Yo % 7o 
IDESGBINDS \GEGID. eee ee 44.56 1.64 26/4 
SEARS LI IG UPS eee 36.96 3.40 53.37 
SU AOGNARIG THOGTAITIEE. een cere cee eee 10.87 40 6.29 
SRGAEDG PHORDOIG oesseec 7.61 93 14.60 

6.37 


at the 60 cm. level. The salinity, which appeared to be the controlling factor 
in determining the vegetation in this community, was 0.49-2.52%, and the 
chlorinity 0.14-1.57°%. Further data concerning soil relations are in Tables 3-6. 

The comparison of basal area (Table 38) for 12 transects in zone B of the 
D. stricta-S. depressa area indicate their overall dominance. The rhizomatous 
growth habit (Fig. 10), gives D. stricta a great advantage in spreading into 
the more saline areas. S. pectinata was at only one location indicating that 
this transect was in a slightly moister situation than was typical for this com- 
munity. 

A single transect, run southwest of zone E (Table 37), appeared to have 
the typical vegetation of the S. pectinata-D. stricta Community merging with 
that of S. americanus-E. rostellata. To the southeast of zone E a five transect 
summary (Table 39) related a vegetation type similar to the Tall D. stricta 
Community. It was typical for this and the previously mentioned community 
to enter or border the D. stricta-S. depressa area when there was a lowering 
of salinity and some added source of fresh water. Besides the species listed 
in Tables 37 and 39, Bassia hyssopifolia, Eustoma grandiflora, Scirpus validus 
and Typha latifolia were widely scattered in the general area. 


Spartina pectinata-Distichlis stricta COMMUNITY 


Species making up this community formed a dense sword of vegetation. 
In one location on the marsh, as mentioned earlier, this community was ex- 


Taste 38. A comparison of the basal area of species at transect sites in Zone B 
of the D. stricta-S. depressa community. B.A. 1s expressed as a percentage of 
the actual cover. 


Transects 
Species 1 2 3 4 5 6 7 8 9 10 11 12 
ue rita 160 1A7~ 88 1.06 1.08 .86 .55 110) 96 253 41.13 90) 
Suaeda depressa .... .25 10... 06-13 221 O50 20.6 3 OZ Cee 


Spartina pectinata V6.0 cece steee eae th eee eee tee eet eee Sees 
meal... QO 27) 88 12 1.21 107 65 1.60 102 83 Eth. 


44 Tue University ScIENCE BULLETIN 


Taser 39. Results of a five transect study on this southeast border of zone E of 
the D. stricta-S. depressa Community. Density, 752. 


Relative Density Basal Area BaA® Y, Wotal Frequency 
Species We of os wh 
Distichlis stricta __..... SIC Se 95.60 5.98 94.32 100.00 
Roawaridaee ee 4.10 Li 4.26 30.00 
SGipUSepalud Oss, ees ees 30 09 1.42 10.00 
6.34 


tensive, and in other areas small patches of the S. pectinata-D. stricta vegeta- 
tion type were found. The transect analysis made in this community during 
July is summarized in Table 40, and Fig. 12 illustrates the vegetation. 


The short dominant, D. stricta, was 20-50 cm. tall, had a relative density 
of 50.77°%, a basal area of 27.46% of the total, and a frequency of 100%. The 
tall (70-130 cm.) dominant, S. pectinata, was the most characteristic species 
of the area. It had a relative density of 18.34%, a basal area of 45.80°, of the 
total and a frequency of 100°. 


On its western margin this community bordered one dominated by 
Scirpus americanus and Eleocharis rostellata. These two species infiltrated 
the S. pectinata-D. stricta Community, had relative densities of 6.36%, and 
18.92% respectively and a combined basal area of 1.28% or 21.94% of the 
total. Poa arida was scattered through the community, and had a frequency 
of 50°%, a relative density of 3.80%, and a basal area of 0.13% or 2.24% of 
the total. Other species, including Juncus torreyi, ]. interior, Scirpus paludo- 
sus and Apocynum sibiricum, had frequencies (Table 40) of 10-20%, relative 
densities of less than 1°4, and a combined basal area of 0.15°% or 2.56% of the 
total area covered. 


Tasre 40. Analysis of Spartina pectinata-Distichlis stricta community. Ten tran- 
sect summary. Density, 771. 


Relative Density Basal Area B.A. % Total Frequency 
Species be To Yo % 
Distichlasmstriclaw eee 50.77 1.66 27.46 100.00 
Eleocharis rostellata ..................-- 18.92 61 10.45 30.00 
Spartina pectinata 18.34 2.67 45.80 100.00 
Scirpus americanus ........------- 6.36 .67 ea 50.00 
POG G11 Gite net ee es 3.80 13 Deoet 50.00 
uncusatorrey ie 56 03 2) I 20.00 
NUNCUSNI TnI CriOV ae 49 03 =) 10.00 
Scirpus: paludOsus) === ee 45 07 1.20 20.00 
Apocynum sibiricum _.---- il .02 sot 20.00 


Tue VEGETATION OF Tue Bic Sarr Marsu 45 


Two species not occurring on the transects, Hordeum jubatum and 
Eustoma grandiflora, were rare to widely scattered through this area. 

Table 42 provides a comparison of the basal area of species at the various 
transect sites. S. pectinata had the highest cover at 60°, and the combination 
of S. pectinata and D. stricta had the highest cover at all sites, making up 
73.26% of the total area covered by this community. One can see also that 
S. americanus and E., rostellata formed an important part of the cover only 
locally. 

The comparison of basal area of species described in Table 42 gives a 
clearer account of vegetation relations than would a comparison of densities. 
The reason is that D. stricta had a relative density of 50.779 and S. pectinata 
only 18.34°%, but the cover of D. stricta was much lower because it is a much 
smaller plant and has smaller culms than S. pectinata. 

The soils in the eastern part of the area at site 10 had a pH of 8.0-8.5, an 
average saturation percentage of 56°, a fairly high organic content of 17.9°/, 
and a soil texture classified as a sandy, clay loam in the surface 10 cm. and 
sand at the 60 cm. level. The total salinity was 0.20-0.63°. and the chlorinity 
0.09-0.36%% in the surface 10 cm. The vegetation relations are in Table 41. 


Tase 41. Transect analysis at soil sample site 10. Density, 87. 


Relative Density Basal Area B. A. % Total Height 
Species Ne %o % cm. 
[DESEA GG 7161) 69.00 2.10 B5e5 35-42 
Spartina Peclinatas 2.2.2.2... 28.70 Sy) 62771 90-100 
Sampus paludOsus, ....:-<-.2---c:0--<--- 2.30 13 2ely, 68-73 
5.98 


Taste 42. A comparison of the basal area of species at ten transect sites within 
the Spartina pectinata-Distichlis stricta Community. B. A. is expressed as a per- 
centage of the actual cover. 


Transects 

Species 1 2 3 4 5 6 I 8 9 10 
Spanina peciinaia =. 375 4.90) 3:86 3.93 2.33 220)” 1.95 1:80 0 157 
Distichlis stricta ........2.10 3.74 1.24 1.68 Hells 56 96> 110) 05) 3207, 
Scirpus americanus .... ...... DEA) ae2 OO ee Ae ee eee 07 93 66 
Bilcocharis vostellaia.. 3-0 tee. ee es Peay) PAs) lO) ee a: 
BOGMAITAG et ee : 23 07 WE ae 63 33 
Scirpus paludosus ... 13 secece  ceenee ses dS. ee eee ee ee 
WERCUSMNOTYCYT a ent ee 230 eee ere 07 ae: 
Juncus anterior 2.0... -...-. Dee | ee Gen a Ae eee: ewes 
Apocynum sibiricum .. ...... 10 A Seem Saree ee ere ied: ees: 


‘ofall eee Seles ION AAO Saxo 4105) 5:23) e045 4-748 S-6l 5.63 


46 Tue University ScrENCE BULLETIN 


At site 12 in the western part of the community the soils at the 10 cm. 
level had a pH of 8.1-8.6, a saturation percentage of 64%, an organic content 
of 10.194, and a soil texture classified as sandy, clay loam and as a sand at 
the 60 cm. level. The total salinity was 0.27-1.00% and the chlorinity 0.14 
1.46°%. Tables 3-6 provide more complete data on soil relations, and Table 
43 gives the vegetational relations at soil sample site 12. 


Tasie 43. Transect analysis at soil sample site 12. Density, 147. 


Relative Density Basal Area B. A. % Total Height 
Species Ves We Ys cm. 
DDISHGRIES SITICI Ge 62.16 374 33.72 40- 50 
Spartina PECUNGIG o.csisctscseeeee 19.28 4.50 40.58 100-130 
Scirpus AMEPICANUS __...2..----------- 12.85 2.40 21.64 100-120 
Apocynum sibiricum .........-..----- wall 10 90 42- 44 
JUNCUS: Inte7107, pe 5.00 235) 3.16 76- 80 
11.09 


Taste 44. Analysis of the Scirpus americanus-Eleocharis rostellata Community. 
A 15 transect summary. Density, 1599. 


Relative Density Basal Area Be Ae, Lota Frequency 
Species % % % So 

Eleocharis rostellata ................---- 60.22 3.99 42.35 O3r24 
WCUPUSGINCTIGANUS, ste Pop apghe 4.05 42.99 100.00 
Eleocharis tenuis ......21..202--02------ 5.46 28 297, 13.32 
Dischiis stricta: ea ee 4.07 21 2.23 6.66 
VA CTO STIS Cee 2.83 alla 1.81 59194 
Carex) Prggoracilis: 2 ea 1.06 sila 1.49 13.32 
I UNGUSE TOLL Cy 1k .87 06 64 39.96 
Panicuc lanuginosum |.........-.-..-- 65 .04 a3 6.66 
Garexnlanucino saa 64 05 5 26.64 
SCUSPUSIMIITCGLICS, pan een 50 07 74 26.64 
Fimbristylis castanea ..........-..-..-- 40 02 Zl 6.66 
NUGUSMIniCTiOT ae 30 08 85 33.30 
Apocynum § sibiricum  .........22.------ 25 a2 1327, 33.30 
CichiQumaculatae 18 06 64 13.32 
Agropyron smithit 22.22.2020. 10 02 21 6.66 
Sphenopholis obtusata .............--- 07 003 03 6.66 
PGRIGUTN: WIN OGLUT Va 05 01 all 6.66 
Lippia lanceolata ........2...2....---..-- 05 006 06 6.66 
var. recognita 

Gareu GNREClans ee 05 004 04 6.66 
var. xanthocarpa 

Vernonia fasciculata .............-...- .04 008 .08 6.66 
Lainis: lanciioliae= ee .02 03 “2 6.66 


THe VEGETATION OF THe Bic Satt Marsu 47 


Scirpus americanus-Eleocharts rostellata COMMUNITY 


The most western margin of the marsh contained a sedge-meadow type 
vegetation in which the dominants were S. americanus and E. rostellata. 
This community was bordered on the east either by a D. stricta-S. depressa 
or a S. pectinata-D. stricta Community type. To the west of the sedge- 
meadow, forming the western limits of the marsh, were the typical sand- 
hills of the Great Bend Prairie. There was some infiltration of sandhill 
species on the western border, some of which were Rosa arkansana, Vitis 
riparia, Polygonum convolvulus, Cucurbita foetidissima, Galium aparine, 
Lactuca ludoviciana, and Oenothera rhombipetala. 

A 15 transect survey, made in the north-western corner of the community 
during June and July, indicated (Table 44) that the average basal area was 
9.42°.,. Eleocharis rostellata, the short dominant in this zone, was 30-70 cm. 
tall, had a relative density of 60.22%, a frequency of 93.24%, and a basal area 
of 3.999% or 42.35% of the total. Scirpus americanus, the tall dominant, was 
70-160 cm. tall, had a relative density of 22.24°9;, a frequency of 100.007, and 
a basal area of 4.05%, or 42.99% of the total. These two species had a com- 
bined relative density of 82.46% and covered 85.34%, of the total vegetated 
area, clearly indicating their dominance in this vegetation type. 

Distichlis stricta, which was not one of the characteristic species, had a 
relative density of 62.68°. in the most north-eastern transect where the com- 
munity bordered upon the D. stricta-S. depressa Community. This transect 
was probably through an ectone between the latter, which had a higher soil 
salinity, and thé Sedge-Meadow Community. D. stricta was not present in 
any other transect. 


Another species of some local importance in the northern part of the 
transected area was Eleocharis tenuis, which had at one site a relative density 
of 77.88% and a basal area of 48% of the total. Other species with a wide 
distribution and a fairly high frequency included Agrostis alba, 59.94%, 
Juncus torreyt, 39.96%, Juncus interior, 33.30°,, Apocynum sibiricum, 33.30%, 
Carex lanuginosa, 26.64%, and Scirpus lineatus, 26.64°. Data on the re- 
mainder of the species with lower frequencies are in Table 44, which sum- 
marizes the vegetational relations. 


A comparison of the basal area relations of the species in Table 45 clearly 
indicates the dominance of S. americanus and E. rostellata. The total basal 
area at all transect sites in this zone was 6.88-13.37°%, while that of FE. rostellata 
was 0).36-6.27% and S. americanus 1.50-8.00°.. Only two other species com- 
prised any appreciable part of the basal area, and these were only locally 
important. The first of these, D. stricta, had a basal area of 3.23°% in transect 
10 and was codominant there, whereas E. tenuts had a basal area of 3.81°% 
at site 9 and was one of the controlling species at that location. 

The vegetational relations at soil sample site 14 are in Table 46. Surface 


48 


Tue University ScIENCE BULLETIN 


Taste 45. A comparison of the basal area of species at ten transect sites within 
the S. americanus-E. rostellata Community. B. A. is expressed as a percentage 


of the actual cover. 


Species ] 


= 18:00 
1.87 


Scirpus americanus 
Eleocharis rostellata .. 
Eleocharis tenuis 
Distichlis stricta 
Agrostis alba 


Carex praegracilis —.. 25 
Apocynum sibiricum — 10 
Juncus Torrey 


Juncus interior 
Scirpus lineatus 


Cicuta maculata 


Carex lanuginosa 
Panicum lanuginosum 
Fimbristylis castanea .. 


Liatris lancifolia ...... 
Agropyron smithi _. 


Panicum virgatum _. 
Vernoma fasciculata. ...... 
Lippia lanceolata 
var. xanthocarpa 


Carex annectans 


var. xanthocarpa 
Sphenopholis obtusata 


Total 


Transects 

4 5) 6 7 
3.10 2.93 rol) Gly 
S419 Sesid) Gills} 4.76 

25 23 16 16 

[See Meroe Beans 
10 10 vA eee 
0 On poe =o oS) ee 
pees 16 oe nes 
2.51 Ale ED 50 aire 
POA Ie, Se 40 fee 
oa 13 ar Pe 
7.07 6.88 10.22 9.04 


8 9 10 
3102) 3:23. 
5.36 36 = 
—_. 3.81 See 
a 3225 
.07 35 

33? OS a 
.05 03 
0) Sy 
30 22 
saws 05 ae 
10.43 7.88 8.93 


soils in the area were the darkest of any on the marsh, due to their high 
organic content of 23.4%. The pH in the surface 10 cm. was 7.6-7.9, the 
saturation percentage 151°, and the soil texture was a sandy loam and a 
loamy sand at the 60 cm. level. The salinity in the surface soil was 0.18- 


Tasre 46. Transect analysis at soil sample site 14. Density, 144. 


Species 


Relative Density 


° 


Eleocharis rostellata 
Scirpus americanus 
Agrostis alba 

Juncus interior 
Apocynum § sibiricum 
Juncus torreyt 


Basal Area B. A. % Total Height 
Hp 6 cm. 
3.43 48.54 30- 50 
3.10 43.86 100-120 

5 3.52 80- 82 
06 84 ie 
13 1.83 54 
10 1.41 WS 


Tue VEGETATION OF Tue Bic Sarr Marsu 49 


0.68° and the chlorinity 0.03-0.23. One reason for the extremely high water 
holding capacity of this soil is the high organic content. In addition, the ex- 
tensive root and rhizome systems of the sedges made the soil surface highly 
absorptive. 


Grazed-sandhill COMMUNITY 

The portion of the Grazed-sandhill Community studied quantitatively 
was located on a sandy beach ridge which formed the southern border of the 
salt flats. A ten transect analysis of this area during July indicated that the 
average basal area was 2.88% and that forbs, common to sandy soil, reached 
their peak in diversity in this community. A total of 36 species were found, 
19 of which had a frequency of only 10%, indicating that they occurred at 
only one location and were not widely distributed. 

The species maintaining the highest frequency percentages were Pas- 
palum ciliatiforium, Eriogonum annuum, Strophostyles leiosperma, Am- 
brosia psilostachya var. coronopifolia, Monarda citriodora, each with a 60°, 
frequency, and Festuca octoflora and Argemone polyanthemos with fre- 
quencies of 50%. 

The species which had the highest relative densities were F. octoflora, P. 
ciliatifolium, E. annuum, A. psilostachya var. coronopifolia, M. citriodora, 
Setaria geniculata, Conyza canadensis, Cenchrus longispinus, and A. poly- 
anthemos in order of abundance. This group of nine species comprised 
67.59%, of the total density and had a combined basal area of 1.51%, or 
65.99°,, of the total. 

Table 47 provides a more detailed analysis of the species relations, and 
Fig. 17 and 18 illustrate two aspects of the community. 

A summary of the basal area relations in this community (Table 48) in- 
dicates that no one or two species made up the greatest part at all transect 
sites. A combination of two or three species at each site, such as P. ciliati- 
folium and F. octoflora at site 5, comprised the greatest part of the basal 
cover, 67°%, but at other sites other combinations of species formed the 
greater portion. 

The vegetation at the soil sample site in this community is described in 
Table 49. The soils in this area are made up of loose sand and had a pH 
of 6.5, a total salinity of .007°% and a chlorinity of .001°,. This analysis, made 
in July, is quite similar to the soil relations found in the Sandhill Mixed 
Prairie Community. The soils in the latter community are also sandy. They 
had a pH of 6.5 and a low salinity and chlorinity. The fact that the domi- 
nant prairie grasses play such a small part in the species relations in this 
Grazed-Sandhill Community must not be attributed to soil relations but to 
the heavy grazing which occurs throughout the growing season. 

Transects BorDERING THE SALT FLats AND SALINE Ponps 
North of the east-west road bisecting the marsh a series of transects were 


50 Tue Universiry ScrENcE BULLETIN 


Taste 47. Analysis of Grazed Sandhill Community. Ten transect summary. 
Density, 312. 


Relative Density Basal Area BaAq 7) Lotal Frequency 

Species We Wb OY, 65 
Festuca octoflor@ 2.22..22..20.-0----.a2 18.21 18 7.87 50.00 
Paspalum ciliatifolium ..... eee OPAll “oD 15.30 60.00 
Briogonum annuum _.............. 7.68 33 14.43 60.00 
Ambrosia psilostachya .............. 7.02 16 6.99 60.00 
var. coronoptfolia 
Monarda citriodora .................- 6.82 13 5.68 60.00 
WELInIGYGCNIGUI ALA ee 6.45 =12 5.24 20.00 
Conyza canadensis _.....---..-..---- 4.70 .09 3:93 30.00 
Cenchrus longispinus _............- 3.95 alli 4.81 20.00 
Boutelouam enacts po 3.66 03 13 10.00 
Strophostyles leiosperma ...........- 3.64 05 2.19 60.00 
Argemone polyanthemos ........... Sy5)) alle 6.12 50.00 
Ocenothera laciniata .................-- 2.38 08 3.50 30.00 
var. grandiflora 
Hordeum =pusillum) pe. 2.05 03 13H 10.00 
Andropogon gerardi —.......... 1.76 03 1e3il 10.00 
BrOMUsmLECLOTUTM: 32 eee 1.64 02 .87 10.00 
Euphorbia dentata ....................- 1.46 02 .87 10.00 
Ghiows werticllaiay = 1.39 06 2.62 20.00 
Chenopodium leptophyllum __.. 1.28 01 eat 10.00 
S150] GG eee 1.23 06 2.62 10.00 
var. tenuifolia 
Euphorbia glyptosperma —.......... 1:23 .02 .87 10.00 
Plantago: purshit, 22.) Lsl7 01 tet 20.00 
Haplopappus divaricatus ......... Pl2 05 DAO 20.00 
Bouteloua curtipendula _............ 1.06 02 .87 10.00 
Leptoloma cognatum ................ 292. 03 1.31 10.00 
Lepidium virginicum. ................ 92 01 44 20.00 
Croton glandulosus —_.......-...... .82 .009 39 20.00 
var. septentrionalis 
Panicum scribnerianum _........ .80 02 .87 10.00 
CGusium andulaium = Jf 04 1275 10.00 
Amaranthus tamariscinus .......... 29 02 ES 7 10.00 
Cycloloma atriplicifolium ........ 5) 01 a4 10.00 
Asclepias: “Speciosa: _- 22. 48 01 ata 10.00 
Commelina erecta .......-......-------- 36 01 ott 10.00 
var. angustifolia 
Helianthus petiolaris ................. ) 01 matial 10.00 
Sporobolus asper 22.2.2... 30 006 26 10.00 
Solanum rostratum  ............-...---- 28 01 ai 20.00 
CY PCrUussSirigOSUs = 27 003 a3 10.00 


2.288 


Tue VEGETATION OF Tue Bic Sart Marsu 


51 


Tasie 48. A comparison of the basal area of species at ten transect sites within 
the Grazed Sandhill Community. B. A. is expressed as a percentage of the 
actual cover. 


Species 1 


Eriogonum annuum 
Paspalum ciliatifolium  _...-.... 
Festuca octoflora 
Ambrosia psilostachya 
var. coronopifolia 

Monarda citriodora 
Setaria geniculata 


Cenchrus longispinus 


Conyza canadensis 
Argemone_ polyanthemos 
Ocenothera laciniata 
Chloris verticillata 
Salsola kali 
var. pestifer 
Haplopappus divaricatus 


Strophostyles leiosperma 
Cirsium undulatum 


Bouteloua gracilis 
Hordeum  pusillum 


Andropogon gerardi _.........---- ------ 


Leptoloma cognatum 


Cyperus strigosus 


Bromus tectorum 


Euphorbia dentata 


Euphorbia glyptosperma —...... —--- 


Bouteloua curtipendula 


Panicum scribnerianum 


Amaranthus tamariscinus 


Chenopodium leptophyllum 


Plantago purshu 
Lepidium 


virginicum 


Cycloloma_ atriplicifolium 


Asclepias speciosa 


Commelina erecta 


var. angustifolia 


Helianthus petiolarts 


Solanum rostratum 


Croton gladulosus 


var, septentrionalis 


Sporobolus asper 


Total 


Transects 
5 6 
Ee ae 

lO 
ro eee 

‘06 eee 
ss Al 

ue 78 

ae 30 

ea .06 

aerks, 40 

pa nos 

aces .20 
Oe 


1.86 


2.81 


Ose es 
a .03 
OG 
D223 


7 8 9 
cee 6 26 
Ee e235 ar 
poe JB 22: 
ee S38 
3.0) 038i 
ae 40 
NO) eee 73 
lO 2. B= 
eee eee 46 
DOE Gees eae 
re ee See .20 
ce oll3} 18 
SOR 
mart AN) eee. 
ae ee ee oll 
Se pees) eee 
ae AG: tee 
13) ee 
2201 et O 2267, 


a Tue Universiry ScIENCE BULLETIN 


set up. The first station was located 29 meters east of the road leading north 
from the east-west road in the center of the marsh and each of the following 
three was 100 meters east of the preceding. The first station ten meters north 
of the road was on bare ground. A transect one meter south had a Distichlis 
stricta basal area of 2.60°., and a second one three meters south had a basal 
area of 6.00°%. 

The second transect station was located 100 meters east of the first and 
five and one-half meters north of the east-west road, on bare ground. A half 
meter south ot the barren area the basal area of D. stricta was 1.50°%, and 
one and one-half meters south the basal area was 5.20°%. In this series, 
salinity tests were made and the total salt content was 2.46%, on bare ground, 
2.12°, a half meter south, and 1.05°% one and one-half meters south in the 
upper 10 cm. of soil. A chloride analysis showed closed correlation with the 
total salt analysis and was 1.13°%, 1.01°%., and 0.61° for each of these sites 
respectively. 

The third series of transects was initiated on bare ground 100 meters 
east of the second and 12.5 meters north of the east-west road. A transect one 
meter south of the bare ground had a D. stricta cover of 0.89% ., and three 
meters south was a transect with a cover of 9.86, with D. stricta making up 


9.80°% and P. arida 0.06%. 


Tasre 49. Transect analysis at site of soil sample. Density, 52. 


Relative Density — Basal Area B. A. % Total 


Species So Yo So 
CSEULCUMOCLO]L OT a ete ee 48.06 69 18.74 
JAG RAE. GH EDHOUTTO. ore eee ee eet eto ZMesy? 1.40 38.05 
Ghlogus@eperncillaia y= ee 9.62 Se) earl 
Monarda ci7i0d 07g) ee 5.60 il aie 
AL ZCMONE POlyant hCmosn ase ee 5.60 46 12.50 
Ambrosia pstlostach qs ee ee 12 .08 Bell) 
var. coronopifolia 
SOU gO Lies striae ee eee eee eo, 06 1.63 
ECPI UTUCL UTI CTT ee 1.92 03 81 
Solanmmirostratum 22. Se ca See B 1:92 06 1.63 
OenorheramlacinGta = 1.92 16 4.35 
var. grandiflora 

3.68 


The fourth station was 100 meters east of the third and 19 meters north 
of the east-west road on bare ground. One meter south the D. stricta cover 
was 0.16%, and six meters south it was 7.237,. In this series, salinity and 
chloride analyses were also made in the surface 10 cm. of soil and the results 
were similar to those described earlier. The salinity of the barren salt flat 
was 2.74%, 1.45% one meter south and 0.72% six meters south. Again the 


Wi 
Oo 


THE VEGETATION OF Tue Bic Sart Marsu 


chlorinity analysis showed a close correlation with the total salt analysis 
and was 1.68%, 0.78% and 0.40% for each of the sites respectively. 

These transects indicate the great fluctuation in vegetation which oc- 
curred at short distances from the bare salt flats. The salinity relations of the 
two series of transects show a close correlation with vegetational cover. 
There was little variation in the species cover, but the basal area of D. stricta 
was greatly reduced in the area of highest salinity, close to the salt flats. 
The reduced salinity in areas closest to the road is caused by the leaching 
action of water flowing through drainage ditches. 

Two transects were run on the eastern border of pond site 8 in a dense 
D. stricta Community, and Table 50 presents the results of these transects. 
The salinity in this area (0.28-1.26°.) was at least partially determined by 
the lake waters since they flood this area during wet periods. The transect 
data indicate that the vegetation of this area is closely associated with that of 
the Tall D. stricta Community. 

An area bordering pond site 9 was studied with a series of 12 transects. 
This area was very variable in composition, and had all types of vegetation 


Tasce 50. Results of a two transect analysis of the eastern border of pond site 8. 


Density, 288. 


Relative Density Basal Area B. A. % Total Height 
Species of. of, Ye cm. 
WD ISIIGHIDSISETICLE 2222-2 ooo e nes zun 90.68 6.39 96.23 50-62 
ISCIGPUS) PGLUAOSUS <c2.c.2.-2-52--02-00- 9.32 25 Sid, 60-70 
6.64 


from bare areas to the Tall D. stricta type represented. This last community 
developed in areas of greater moisture and lower salinity. A summary of the 
vegetation in this community is in Table 51. 


Tasce 51. Results of a twelve transect analysis of the vegetation bordering pond 
site 9. Density, 871. 


Relative Density Basal Area BA. % Total Frequency 
Species Y of, cf, of, 
DEG a AGN 1741 ra 95.29 3.80 95.00 S1FGS 
Sampus paludosus, ......--.......----.- 83 09 DDS 24.99 
Suaeda Aepressd —..-..200.0--c2ceen----- 1.60 02 50 24.99 
OWE TLG Gia soe ee Bee ors 228 09 2:25 8.33 
00 


Salicornia rubra was found in an open salt flat in the southwestern part 
of this area. Some other species, besides those mentioned in Table 51, found 


54 Tue Universtry SciENCE BULLETIN 


in areas of high moisture and lowered salinity were Polygonum hydropiper- 
oides, Polygonum lapathifolium and Xanthium pennsylvanicum. 


SUMMARY OF QUANTITATIVE RELATIONS 


The following pages provide a brief, comparative description of the 
quantitative relations of species in the major plant communities on the 
marsh. From a comparison of the relative density, basal area, and frequency 
of species within the areas (Tables 52-54), it can be clearly seen that many 
species were found in only one transect area, some in two, but only a very 
small number in more than two communities. The actual occurrence of 
species at transect sites within the various plant communities were one 
species in six communities, one species in four communities, three species 
in three communities, 15 species in two communities, and 52 species in one 
community. Species listed in Table 8, which describes the seasonal and 
zonal aspect of the vegetation, were either extremely rare or in some cases 
did not occur during the period of these quantitative studies. 

Distichlis stricta was the most widely spread and one of the more im- 
portant species in four of the seven communities in which it occurred. It was 
found in soil which varied in salinity from a low of 0.04°, in the S. azroides 
area to a high of 2.52% in the D. stricta-S. depressa Community. It reached 
its best growth in the Tall D. stricta Community where it had a relative 
density of 94.46%, a frequency of 100%, and a basal area of 92.65% of the 
total. The salinity in this area was 0.29-0.94°%,. D. stricta did not occur in 
the Sandhill Mixed Prairie or Grazed Sandhill communities which had 
lower salinities than the marsh communities, nor did it occur in the S. de- 
pressa area which occupied parts of the open salt flats. 

In the D. stricta and D. stricta-S. depressa zones, D. stricta, can be con- 
sidered the dominant of a unistratal community, whereas, in the S. paludosus- 
D. stricta and S. pectinata-D. stricta areas it must be considered the dominant 
of the lower strata of a bistratal community. S. paladosus in the former and 
S. pectinata in the latter made up the greater part of the upper strata in 
these areas. D. stricta was also found in the lower strata of the S. americanus- 
E. rostellata and the S. airoides communities, but in it was of much less 
importance than in those previously described. Billings (1945) also men- 
tions the ability of D. stricta to exist in a unistratal or multistratal community 
and that it could be found in almost any saline soil except the very driest. 

Another species of fairly wide distribution was Scirpus paludosus. This 
species seemed to favor the moist and slightly saline areas but sometimes was 
found in areas where the salinity reached slightly over 1°.. The range of 
salinity under which it ocurred was 0.22-1.10°% in the surface 10 cm. of soil. 
It was one of the dominants of the S. paludosus-D. stricta Community, with 
a relative density of 31.80%, a frequency of 100%, and a basal area which was 
54.999, of the total. S. paludosus occurred also in the Tall D. stricta Com- 


Tue VEGETATION OF Tue Bic Sart Marsu 55 


munity, where it was only of local importance, and in the D. stricta-S. 
depressa Community it was very widely scattered and not of local importance. 
In all of the communities it was found in, except S. pectinata-D. stricta, it was 
a member of the upper strata. In this zone and in its rare occurrence in the 
S. americanus-E. rostellata Community, it was a member of one of the lower 
strata. S. paludosus did not occur in the most saline salt flats or in the very 
dry communities. 


The next three species were each found in three of the major communi- 
ties on the marsh. The first of these, Suwaeda depressa, reached its greatest 
abundance in the more saline areas of the marsh. In the S. depressa commu- 
nity it had a relative density of 100°, a frequency of 30°%, and a basal area 
which accounted for 100° of the total. This community had the lowest 
basal area on the marsh, 0.06°., and S. depressa was the only species found 
on the transects. Two other species, Sesuviam verrucossum and Salicornia 
rubra, were found also locally in this general area, but they composed an 
extremely small portion of the total basal area. S. depressa was also of some 
importance in the D. stricta-S. depressa Community, where it was quite 
prevalent and had a relative density of 13.13°., a frequency of 62.37%, and 
a basal area which was 13.50% of the total. It was also found in small quan- 
tities in the Tall D. stricta, S. airoides, S. americanus-E. rostellata and S. 
pectinata-D. stricta communities. 


The remaining two species occurring in three communities, Ambrosia 
psilostachya var. coronopifolia and Juncus interior, were not of major im- 
portance in any community in which they occurred. The former species was 
found on relatively dry, nonsaline soils, while the latter occupied the moister 
portions of the communities in which it occurred. 


Of the 15 species found in two communities only A. gerardt, E. rostellata, 
S. americanus, P. scribnerianum, S. pectinata and S. airoides were dominants 
in at least one of the communities in which they occurred. 


Andropogon gerardi and Panicum scribnerianum were most important 
in the Sandhill Mixed Prairie Community, but they also occurred in the 
Grazed Sandhill Community, where they comprised only a very small part 
of the total vegetation. These two species were limited to the drier sandy 
areas of the marsh and were not found in wet or saline areas. In the Prairie 
Community, A. gerardi had a relative density of 8.53°., a frequency of 
12.50%, and a basal area which was 16.73°% of the total, whereas, in the 
Grazed Sandhill Community it composed only 1.76%, of the total density, 
had a frequency of 10°, and a basal area which was 1.31%, of the total. The 
basal area covered by this species and other tall grasses such as S. nutans and 
P. virgatum was greatly reduced in the heavily grazed areas. P. scribnerianum 
had a relative density of 19.78%, a basal area of 14.88°% of the total, and a 
frequency of 100°% in the prairie area. It must be considered one of the major 


56 Tue Universiry SciENcE BULLETIN 


dominants in the lower strata of this community, whereas, in the grazed 
sandhill area it was of much less importance and had a relative density of 
only 0.80°%, a frequency of 10%, and a basal area which was 0.87% of the 
total. 

Eleocharis rostellata and Scirpus americanus were codominants in the 
Sedge-Meadow Community and were also present and of some local im- 
portance in the S. pectinata-D. stricta Community. These species reached 
their peak abundance in the S. americanus-E. rostellata area, where the soil 
retains a great deal of moisture and the salinity was 0.18-.68°,. In this area 
the two species had a combined relative density of 82.46°., a frequency of 
93.24°, and 100%, and a basal area which comprised 85.34°. of the total, 
whereas, in the S. pectinata-D. stricta Community they were of less im- 
portance, having a combined relative density of 25.28°., a basal area of 
21.94%, of the total, and frequencies of 30% and 50°, for each species respec- 
tively. The great reduction in basal area and relative density and frequency 
which these species had in the S. pectinata-D. stricta Community showed 
they were of less importance here than in the S. americanus-E. rostellata 
Community. The reason for this is probably the higher salinities reached 
in the former community. 


Spartina pectinata, as mentioned earlier, was the tall dominant in the 
S. pectinata-D. stricta Community. In this area it had a relative density of 
18.34°., a frequency of 100°, and a basal area which accounted for 45.08% 
of the total. Its wide distribution and high basal cover were the factors which 
made it the most characteristic plant of a community. It also occurred scat- 
tered in wetter parts of the Sandhill Mixed Prairie, the S. americanus-E. 
rostellata and in less saline areas of the D. stricta-S. depressa communities, 
but in these areas it was only of minor importance and made up only a very 
small part of the total cover. 


Sporobolus airoides was found in two main community types and was 
also scattered in hummocks on the open salt flats. It reached its greatest 
abundance in the S. airoides Community, where it had a relative density 
of 90.50%, a frequency of 100%, and a basal area which accounted for 88.21% 
of the total. It also occurred on the western border of the prairie, but here 
it comprised only 6.24% of the total density, had a basal area of 11.21% of 
the total, and a frequency of 25°. This species appeared to make its best 
growth in those areas which had only moderate salinity and low moisture, 
but it was also found on the open salt flats in the areas of highest salinity 
and higher moisture. 


The species found in only one community and of more general im- 
portance where they occurred include Bromus japonicus and Festuca octo- 
flora. B. japonicus was found in areas of low salinity and moisture, includ- 
ing the sandy hummocks and the Sandhill Mixed Prairie Community. In 


THe VEGETATION OF Tue Bic Sart Marsu a7 


the prairie it had a relative density of 27.79°%., a frequency of 87.50%., and a 
basal area which accounted for 14.387, of the total. Unexpectedly, as dis- 
cussed earlier, this species was one of the more conspicuous plants in the 
lower strata of both the prairie and hummock communities. 

Festuca octoflora had a relative density of 18.21%, a frequency of 50%, 
and a basal area which made up 7.87°% of the total in the Grazed Sandhill 
Community. It was also found in the Prairie Community but was never 
very important there. This species was most characteristic of dry, sandy, 
non-saline areas, and during the early spring it was fairly characteristic of 
the lower strata of the Grazed Sandhill Community. 

Species in the prairie, grazed sandhill, sedge-meadow and other areas of 
the Big Salt Marsh which were of some importance but not mentioned here 
are described in more detail in some of the earlier sections. An account of 
their quantitative relations are in Tables 52-54, and their seasonal and zonal 
distribution are in Table 8. 


Salt Tolerance of Marsh Species 


Table 55 contains a list of plants found on the Big Salt Marsh which 
either were growing on saline soils or had been listed by Fernald (1950), 
Rydberg (1932), or others as occurring on them. In addition to the species 
list, Table 55 contains the range of salinity, in the upper 10 cm. and at the 
60 cm. level of the soil, under which the various species has been found 
growing. The majority of species listed for the S. americanus-E. rostellata 
(Sedge-Meadow) and S. pectinata-D. stricta communities have not been 
included in this list because their salinity relations were described earlier 
and many of these species, as cited in Table 8, reached their highest salinity 
tolerance in these areas, indicating that they were typical of fresh water areas 
or those of only the lowest salinities. 

Agropyron smithu, which occurred in the prairie, sedge-meadow, and 
on sandy hummocks, has been listed by Hitchcock (1898) and Weaver and 
Albertson (1956) as being tolerant of brackish soils. In this study it was 
found to have reached its greatest abundance in the Prairie Community 
which had the lowest salinity, 0.003-.02°, in the surface 10 cm. As noted in 
Tables 8, 44, and 45, it was also found in the Sedge-Meadow Community 
where the salinity was 0.18-0.68°.. A. smithi did not occur in areas of higher 
salinities on the marsh and, though it is tolerant of traces of salt, it should 
not be considered an indicator of saline soils. 

Aster ericoides (A. multiflorus Ait.) has been listed by Schaffner (1898) 
as a successful invader of saline soils. This species was found in a wide 
variety of vegetation types, including Prairie, Tall D. stricta Meadow, Sandy 
Hummock, and S. airoides communities. The salinity in the S. arotides 
Community was 0.04-0.19°., and the total range of salinity in the surface 10 


58 Tue University ScieENcE BULLETIN 


Taste 52. A comparison of the relative density (°) of species within the major 


1—Suaeda_ depressa, 


study areas. 


2—Tall Dustichhs stricta, 


4—Sandhill Mixed 


3—Sporobolus airoides, 


Prairie, 5—Scirpus paludosus-Distichlis stricta, 6—Distichlis stricta-Suaeda depressa, 7—Spar- 


tina pectinata-Distichlis stricta, 8—Scirpus americanus-Bleocharis rostellata, 9—Grade Sandhill. 


Species 1 


Achillecmlanulosg eee 
Agropyron smithit 2 see 
NTRS CHAD ett, ee 
ARTOSS MLCT GS) amen 
Amaranthus tamariscinus 0... 
Ambrosia psilostachya 0 ..... 
Andropogon gerardi ......... 
Apocynum sibiricum wo 
Argemone polyanthemos .... _.... 
Artemesia ludoviciana _....  ~..... 
Asclepias “spectosa 2 ee 
ASICRMCTICOLL CSA ee ee 
Bouteloua curtipendula _...  -.... 
Bouteloua gracilis 200000000. se. 
BFOMUS? JAPONICUS) == 
Bromus tectorum, 2. 
CGrEXManneclan's ae ee ae, 
Carex lanuginosa = 
QAR BE FRAG AGG) see 
Cenchrus longispinus 
Chenopodium leptophyllum — ...... 
Chloris verticillata .0000 


Conyza canadensis i. n-- 
Groton clandulosa ee 
Cycloloma atriplicifolium .. ...... 
Gy perds *SitigOstisa pe ee 
Distichlismstiiciage = ey eee 


ENCOGRGYTS CULL ane ee 
Elymus canadensis 0.000. 2----- 
Erigerom strigosus, 2.2 ee 
Eriogonum annuum _........... 

Euphorbiacdenigiag = ee 
Euphorbia glyptosperma .... ...... 
Restuca, OCLOfOra eee 
Fimbrystilis castanea .....--..  ------ 
Haplopappus divaricatus ...  _.... 
Helianthus “penolaris: - 222 


bo 


Transect Area 


es 30) tet. 25 —See ste 
ee VS oe a4 ae, LO). Ss 
See On ee ee 2:83. ee 
aes B37 ee eee eer 
ee ee ee er DS) 
2.08 19) ee ea 7.02 
st! 8.53 ac be | Le 
toa. aceeee ene pe eee 31 25) pee 
A Bee, as Gi enn 32D 
ene 35 Bie jee rrr 
St ees Ua 48 
16 eee Ee 
Pe ae Ges ces 1.06 
fee ee | ee, 3.66 
ae DTT Wee. we 
ee eee bo | fle.” 1.64 
es een ae oe ee 105.) iae= 
fe) ee ga ee cs OS Pe) eecee 
Ate Oe La, ee 0G 
dee ee ee ar 3.95 
ee wee es 1.28 
cap rote Ghee? oc. | (ae 139) 
PEED Wem Geet es ony Eee 13h, ee 
Stes) estes) Pmcviecin Go mcdise: (9 ee ee al 
Pe Ee 356 
SEE caine pests 5 Sst 4.70 
eee Se er 82 
See) ee” unk! ee 2s) 
el ee?) ee LH 
BHO) cee 63.60 86.56. 50.77, 407 ieee 
Se en ee 18:92) 16027 
pee Foe oy eee Oe 
41 Mee tee, sts ee nn 
Pee ee | 2” re 
ati See SEO Ste 4... Gee 7.68 
oko gees aeses6) |. ie 1.46 
<A 123 
ee eats 18.21 
a ey a ee AO” eee 
ny eee eee 1.12 
Wee Met Wis GO 35 


Tue VEGETATION OF Tue Bic Satt Marsu 59 


Transect Area 


Species 1 2 3 4 5 6* 7 8 9 
Hl@nteniae FCI ereereee eee eee ee ee ee eee 2.05 
WUE CPHETIORY coset Me a eee HOt 2 ee: -49 50 
NgTECDS BOG ONE eo eee ene oe ee ee eee 56 Oe eee 
MRCPLGIUIN VITBINICUT cnn cnceee veneers AS meet, = Pee Sere 2 
NEG PLOLOMLGUICORTIALUII scree cee cccczc) ects | ecclS. EE eres) «| ees 392 
Wiatris lanctfolia ..........-... rr re Ee xe AR a 102; == 
IED PIQMMIANCCOIGLD arse ach Seis!) a. ected © ESE 05" sae 
MOGERHO. “GHITOHORG capers eee) eee ces: pee Nae SPY =: 6.82 
Ocnothera laciniata _....... SY 62 oe reer ns pCO «tes 2.38 


var. grandiflora 
CDCI SESUN ICL ap een nen Pee) ee, keeS 225) ea ae Me eS ae 
ETGUIN ONUSINIOSUT =e ee eens oe ee ees "65? 9 pies 
IBARICUID  SCHIPNCTIGNUIM, sesce ccccze nde en 19.78 
GUE DURYOO. eee | ee eee 7.28 
PSPAIUIN: (CLGTY{OLLUTD 2c ceccne neces eee ee 
MOGQHERO: FATGIUD, ae eee eee eee 3303. sae See Bees aiy/ 
WDOUMCTIC Ufa sere ne devas Geese eee 3:00" ) 22= a 
Polygonum ramosissimum .. ee 3.80 
S\AUSQUEE WAGE en eerere eee 
SPONSE GADGETS Bay ee ee eee 6.36 22.24 
SaVOS [EROTETS ean ee Eee 
ISGUIPUS PALUGOSUS <...ccccccee 2 SYo hole eee pent te 31.80 28 45 
SAE (AADAC fo entree ees | eee eee! © 


SQUALL ROSIAGL Le ee ee ee See 
SOAWACGTTD (HEALS eee ee B47: 1 Se ase ee ee ee 
SREAG PORTER a Beer Be epee ADE UCSHy cere ee 
SVCHOPHOMS MOULUSAIG) see ee ee ea ee) eee pry apts HOP eset 
Sporobolus airoides 20000000 cee one DOS ON 6.2 4 eee 

SHOLOUOIUS GSPEP cone cese Ne Se eat 
SUZOPNOSINICS ELOSPCTING! 2022 ee cc) eee 3.64 
SAGE NG NOOO ee ee SO PIS 213 Mite ee 

Wivadescantia occidentalis 000 i stesso PL! ree Wee! Ree Toe 


WEEN ONT GME] GSCICUIGL Lg ete aM Me a Pee ee OAGs Bere 


* Data for this community have been compiled as an average of five separate analyses within 
this vegetation type. 


cm. was 0.003-0.94°,. Its broad range of salinity tolerance indicates that it 
might be an important species in areas with a fairly wide salinity range. In 
face, however, this species always was widely scattered to rare in occurrence 
and was never found invading the open salt flats or any communities directly 
adjacent to it. Its only occurrence in these areas was on the margins of sandy 
hummocks where the salinity was greatly reduced. Schaffner (1898) men- 
tions that A. ericoides was only occasional in occurrence in the northern 
Kansas salt marshes which he studied. 

Aster exilis was found in the Prairie, Tall D. stricta Meadow, Sedge- 


60 


Tue UNiversiry SCIENCE BULLETIN 


Taste 53. A comparison of the percentage of total basal area covered by species 
found in the major study areas. 


1—Suaeda depressa, 2—Yall 


Distichlis stricta, 3—Sporobolus airoides, 


4—Sandhill Mixed 


Prairie, 5—Scirpus paludosus-Distichlis stricta, 6—Distichl’s stricta-Suaeda depressa, 7—Spar- 


tina pectinata-Distichlis stricta, 8—Scirpus americanus-Eleocharis rostellata, 9—Grazed Sandhill. 


Species 


Transect Area 


Achillea lanulosa 


Agropyron smith 


Agrostis hiemalis 


Agrostis alba 


Amaranthus tamariscinus 
Ambrosia psilostachya 


Andropogon gerardi 


Apocynum  sibiricum 
Argemone polyanthemos _.. 
Artemesia ludoviciana 


Asclepias speciosa 


Aster ericoides 


Bouteloua curtipendula 
Bouteloua gracilis 


Bromus japonicus 


Bromus tectorum 


Carex annectans 


Carex lanuginosa 
Carex praegracilis _.............-- 


Cenchrus longispinus 
Chenopodium leptophyllum 


Chloris verticillata 
Cicuta 
Cirsium 


maculata 


undulatum 


Commelina erecta 


Conyza canadensis 
Croton glandulosa _........... 
Cycloloma  atriplicifolium 

Cyperus strigosus 
Distichlis stricta 


Eleocharis rostellata 


Eleocharis tenuis 


Elymus canadensis 


ErigerOn strigosus 2=2.-.----- 


Eriogonum annuum 
Euphorbia dentata 


Euphorbia glyptosperma 


Festuca octoflora 


Fimbrystilis castanea 


Haplopappus divaricatus .... 


Helianthus petiolaris 


27.46 


2.23 


10:45. 742335 


2.97 


THe VEGETATION OF Tue Bic Sart Marsu 61 


Transect Area 


W 
eS 
VW 


Species 1 2 6* 7 8 9 


DOndeUT DUSIIIUIN Coes aces een ene By oe age he ee 
VERELS VLTGRIC Pe ee 
NQEAGLS- OTHE Wee ere oes eee) ee ek eee =I 04K, a ee 
IDGPUATUM, UIFSINICUTA —oonncce  enenne neces wee 06 
L@PIOUCGOE. CORUHTTD eras ae ee 
IOGATISPRIGMICLIOILG wee ec asec a : 
(Lijppnee UROGOUOY eee ee eee ae ee ee | eee ee 002s ee 
WHOMAYAGACHTIOGOLG sere eect ees ssc BS: eee eee 5.68 


OER OLRCL PMIACINIGID Ieee serene At eR 7 A eee B}5)0) 
var. grandiflora 

ORADS: CGE See ee ae, eee Pa en ne: pee ae S 
Panicum lanuginosum 0 
IPGNIGUM SCTIDNETIANUM 2. eeeece eens wee 14.88 
MEQRIGUIN ULV QAIUTD cece eee ee 
PASAT GINUVORIT se ee eee eee ee Oe 15.30 
PUUPEERO! ALPE. ere e ene ee ee 150 
[PROVE LAL Raa ee 08 
Polygonum ramosissimum ee Beetles fete re 
Sasol An ee ee ee ee 


Wau DUS ANICTICATIUS| mae eee ee ee ee 


Sans UEGARDS ree 
ESCU;PUS PALUAOSUS!) <.occecneneee anno TAI i ean ee ee leith We peek eke 
IGLONIG SCNIGUIGIG eee ne 
SOUAPLGD GOGTTUTUD tin a, ESS A ene | Le = a 
ISOUDHASITUMM NULGNS ne eee ane 
Spartinata pectinadta nn. ence 
Sphenopholis obtusata 0... 0 ee. oenne RE) ese gr er 0S ieee 
Sporobolus atroides .....  .....- 
ISPORODOLUS ASPET eeeeencnecencee  nenens nase 
SURO PRUSIVICSMmICLOSPErING meas ee ee) Oe eee eee 
IUGCUG GEPTESSA  -nccnec----------- NOOMORM See) ee Es lS ATED 0 eee pee 
Tradescantia occidentalis .... 


Vernonia fasciculata ............ 


* Data for this community have been compiled as an average of five separate transect analyses 
within this vegetation type. 


Meadow, and Hummock communities. It occurred in a broad range of 
salinity relationships of 0.003-0.949, in the surface 10 cm., just as in A. 
ericoides. Hitchcock (1898), in his brief survey of the Big Salt Marsh, men- 
tioned this species as being salt tolerant, and Keith (1958) cites it as occurring 
in an area where the salinity was 0.26-0.50°... Information accumulated on 
its distribution in this study indicates that it is able to tolerate salinities below 
1%, in the moister areas. A. exilis was not of great abundance on the Big 
Salt Marsh, varying from rare to widely scattered, and it should not be con- 
sidered a species indicative of areas of high salinity. 

Atriplex argentea (A. expansa Watson) was listed by Hitchcock (1898) 


62 


THE University ScrENCE BULLETIN 


Taste 54. A comparison of the frequency (°%) of species within the major 
study areas. 


1—Suaeda depressa, 


2—Tall Distichlis stricta, 


3—Sporobolus airoides, 


4—Sandhill Mixed 


Prairie, 5—Scirpus paludosus-Distichlis stricta, 6—Distichlis stricta-Suaeda depressa, 7—Spar- 


tina pectinata-Distichls stricta, 8—Scirpus americanus-Eleocharis rostellata, 9—Grazed Sandhill. 


Species 


Achillea lanulosa 


Agropyron smithii 


Agrostis hiemalis 
ASTOSTISMAIOG: ae 
Amaranthus tamariscinus 

Ambrosia psilostachya 


Andropogon gerard 


Apocynum sibiricum _.......... 
Argemone polyanthemos .... 
Artemesia ludoviciana ........ 
Asclepias speciosa 
Aster 


Bouteloua curtipendula 


ericoides 


Bouteloua gracilis 


Bromus japonicus 


Bromus tectorum 


Carex annectans 


Carex lanuginosa, =e 
Garex, pracgracilis’ 
Cenchrus longispinus ........ 


Chenopodium leptophyllum 
Chloris verticillata 


Commelina erecta 


Conyza canadensis 
Croton glandulosa _............ 
Cycloloma atriplicifolium _.. 
Cyperus strigosus, t= 
Distichlis ‘stricta. 2 


Elymus canadensis 
Erigeron strigosus ..........------ 


Eriogonum annuum 
Euphorbia dentata 


Euphorbia glyptosperma .... 
HCStUCEOCLOIONG ese 
Fimbrystilis castanea 


Haplopappus divaricatus .... 
Helianthus petiolaris 


Transect Area 


3 4 6* 7 8 
Bee 25 ae 2 ee 
ae: 25 eee a ee HES 
ek 7255)0) ee ae ee 
Sie 2, eal, ee ae Se) 

40 12750) eeic 2) er 
ae 125 Oe ee EE se 
ee ee 20 33.30 
Ree 3750) eee eee estes — 

S09 RE fs 2: ae 
mee Sip. 2 pe SS 
eo per eee SS 6.66 
Sees ee re Dat: 26.64 
ey Pe eee te 13.32 
ee eee estes) Biceeen eeeees 13.32 

<1) Se 100 96.82 100 6.66 
ces ees) eee 30 93.24 
ee 13.32 

10 25a 0 ee 
ere 12250) -2..28 (Ase a 
Be a fee ee ane eae 2 6.66 


Tue VEGETATION OF Tue Bic Sart Marsu 


Transect Area 


Species ] 2 3 
IORHCOS 7 ATI! eee ee ees 
[LOGOS SREGTOP eee eee eee 12.50 
ELRGUSMELOUT Ch Lmenete ne eee eee eee © ececen | feeeee, lees 
IBEBIGIUIN VITZINICUMN nce cece neces wee 12.50 


Leptoloma cognatum 


(LagiinG UGHGAIUE somes | pees See eee ee 


Lippia lanceolata 
NIORERUG! GLROLOAL. ae ee eee 25 


Wepotheramlaciniaidpe se ee eee 


var. grandiflora 

OMG RSENIGL gece tees coe eeneal) ere Seen cee WS 
Panicum lanuginosum 
Panicum scribnerianum 0.00 ee 100 
Panicum virgatum 
Paspalum ciliatifolium 
[WECORO {ROGIER cece “bees | ee 75 


POCMAIIC deme wee eee ee ds ORAS oe. Ree 
Polygonum ramosissimum cee 
NZIS OLG MR CL] Lap rea ere ee et I es 
SCID SAINCTICANUS ee tere ee ee 
NCURROS EDAD es ee Ee ae ee Dee 


Scirpus paludosus .............-. 5 ee ee 100 
Wefurial peniculatd, 22... 2 


SOLGIUINETO SIT ALUN ee ee, pen @ LE 


SORQHETDD IMEDO eee ee ee 25 


PanING Ww PCCHNGian ee 


Sphenopholis obtusata 


Sporobolus atroides 202. cece ne 100 25 

SPO,OQU0IUSROS per ene en ee 
Winophostyles: lerosperma i... seccc2 cennce,easene se ee 
Suaeda depressa .............--..... SOMME Mn |. ee. 
Tradescantia occidentalis 0.00 0. eee 12.50 


Wernonia fasciculata 2.20.0 <0: 


6* 7 8 
—— 10 33.30 
ae 20 39.96 
ee See 6.66 
pur 6.66 
oo SAE 6.66 
Beieok \ Sse: 6.66 
eee 50 ee 
eee 50 100 
pe eee 26.64 

2:0) eee eee 
124255110 OS eres 
ee: 6.66 
623370 Wi ee ee 
BESS eer 6.66 


* Data for this community have been compiled as an average of five separate transect analyses 


within this area. 


and Schaffner (1898) as being a successful invader of saline soils. Schaffner 
(1898) mentioned that it was rare in occurrence in the more saline areas 
but occasional in less saline ones. This species was extremely rare on the 
Big Salt Marsh, being found only once bordering a small salt flat north of 
the Tall D. stricta area. It was associated with S. airoides, S. paludosus, P. 
ramosissimum and D. stricta. No salinity measurement was made in this 
area, but areas with the same vegetation complement had a salinity of 


0:22-1°%,. 


Atriplex patula var. hastata has been mentioned by Coupland (1950) and 
Rawson and Moore (1944) as being found in the most saline places border- 


64 


TaBLeE 55. 


Tue University Science BULLETIN 


Species 


Min. % 


Agropyron smithu 


Aster ericoides 
Aster exilis 
Atriplex argentea 
Atriplex patula 
var. hastata 

Baccharis salicina 


Cycloloma_ atriplicifolium 


Distichlis stricta 


Eleocharis 
Eleocharis 


Eustoma_ grandiflora 


acicularis 
rostellata 


Hustomamerandiflora: = : 


f. Fischeri 


Fimbristylis castanea 


Flaveria c 


Haplopappus phyllocephalus 


subsp. ann 


Heliotropium curassavicum 


Hordeum 
Hordeum 


Iva annua 


Lippia lanceolata 


am pestris 
uUus 


jubatum 
pusillum 


var. recognita 


Muhlenbergia asperifolia 


Myosurus 
Panicum 1 
Pluchea p 
Poa arida 


Polygonum lapathifolium 
Polygonum ramosissimum 


minimus 
urgatum 
urpurascens 


Potamogeton foliosus 


Rumex m 


aritimus 


var. fueginus 


Salicornia 


Scirpus americanus 
Scirpus paludosus 


Sesuvium 


Setaria geniculata 
Spartina pectinata 
Sporobolus airoides 
Sporobolus pyramidatus 
Sporobolus texanus 


rubra 


verrucossum 


Suaeda depressa 
Tamarix gallica 


Typha angustifolia 


Typha latifolia 


.003 
-003 
.003 
.22 

003 


10 cm. 


Max. % 


.68 
eat 
oth 
1.10 
1.10 


50 

.007 
2.52 
1.26 
1.00 
1.00 
1.00 


1.00 


1.10 


1.10 


1.00 


Min. % 
.002 
002 
002 
28 
28 


60 cm. 


Range of salt tolerance of species on the Big Salt Marsh. 


Max. % 
.07 


009 


Tue VEGETATION OF Tue Bic Sart Marsu 65 


ing saline lakes in Canada. This species has been mentioned in many works 
on coastal marshes (Harshberger, 1911; Ganong, 1903) as having a wide 
range of salinity tolerance. On the Big Salt Marsh it was found in a wide 
variety of communities, including the Prairie, Sandy Hummock, Tall D. 
stricta, S. airoides, and the S. paludosus-D. stricta, A. patula var. hastata was 
very rare to rare in occurrence, but as its salinity range, 0.003-1.10%, indi- 
cates, it is capable of a fairly broad distribution. In this area it could not be 
considered one of the primary indicators of high salinity since it was rarely 
found invading the open salt flats or areas of highest salinity. 


Schaffner (1898) and Penfound (1953) did not mention this species as 
being one of the important invaders of areas of high salinity. On this 
marsh, however, when the S. paludosus-D. stricta or one of the other com- 
munities in which it occurred bordered the open salt flats, some greatly 
dwarfed forms (up to 10 cm.) which appeared to be A. patula var. hastata 
were found. These dwarfed forms never developed flowers and died soon 
after germination. 


Baccharis salicina was found only on two low hummocks on the marsh. 
Gates (1940) and Harrington (1954) among others have listed this species 
as occurring on saline soil. The surface soil on one of the hummocks in 
which it occurred had a salt content of 0.50°% during July, but this value 
probably fluctuates throughout the growing season. Since this species was 
found in only two extremely small areas, it should not be considered one of 
the primary indicators of saline soils. It was never found invading or en- 
croaching upon the bare salt flats, and, as discussed earlier, the vegetation 
in the area in which it occurred (Fig. 19) was of the D. stricta-S. depressa 
type. 

Cycloloma atriplicifolium (C. platyphyllum (Michx.) Mog.) has been 
reported by Harshberger (1911) as being prevalent in saline areas in south- 
ern Kansas. This species was found growing in the area but only in widely 
scattered patches in the Grazed Sandhill Community. The soils in this area 
were sandy with a salinity of 0.007%., and were among the least saline areas 
on the marsh. 


Distichlis stricta (D. spicata (L.)) has been cited by numerous authors 
including Harshberger (1911), Kearney et al. (1914), Flowers (1934), and 
Billings (1945), as one of the controlling species in large parts of inland salt 
flats and salt marshes. Many workers including Hilgard (1914), Harris 
(1920), Richards et al. (1954) and some of those mentioned above have 
pointed out that this species has a very wide range of salt tolerance. On the 
Big Salt Marsh D. stricta was found in the wetter parts of the Prairie, the 
Tall D. stricta, Sedge-Meadow, S. pectinata-D. stricta, S. aroides, S. paludo- 
sus-D. stricta, and D. stricta-S. depressa communities. The total salinity 
range under which it existed, as noted in Table 55, was 0.04-252%.. 


66 Tue UNiversity SCIENCE BULLETIN 


It reached its most robust growth and highest basal area in the Tall D. 
stricta Community where the salinity was 0.29-0.94%,. In the D. stricta-S. 
depressa Community, where this species entered areas of salinity of 0.49- 
2.52°., it was definitely stunted in growth. Schaffner (1898) mentioned this 
stunted growth form of D. stricta in the more saline zones of vegetation he 
described. The mere presence of this species, as other authors have stated, 
does not necessarily indicate high salinities, but when it is found covering 
large areas, with a low basal area and a stunted growth form, the salinity 
of the soil is invariably high. 

This species spreads easily by underground rhizomes (Fig. 10). These 
are shallow, usually occurring within the surface 10 cm. of soil; however, 
the roots coming off the rhizomes are deeper penetrating and during soil 
sampling, root material was found at the 60 cm. level. This point is of some 
interest for in the D. stricta-S. depressa Community one finds the deeper 
rooted D. stricta and its codominant S. depressa which has a short taproot 
that rarely penetrates much below the 15 cm. level. These facts would seem 
to negate the ideas of some authors that plants tend to have their roots sit- 
uated above or below the layers of greatest salt concentration because D. 
stricta had its rhizomes in the saline area and roots at a lower depth where 
the salinity was still fairly high, 0.27-1.09°.. At germination, however, both 
species would be subjected to the higher range of soil salinity which occurs 
at the surface. 

Shantz and Piemiesel (1924), Aldous and Shantz (1924), and Weaver 
(1954) among others have mentioned that D. stricta grows well in an area 
containing a shallow water table as well as a high salt content. This is also 
true on the Big Salt Marsh where the water table is particularly shallow, 
and during wet periods parts of the Tall D. stricta area, where this species 
makes its best growth, is under as much as 15 cm. of water. 


Eleocharis acicularis was found growing at the bottom of shallow saline 
lakes. It is not found listed as a salt tolerant species, but Gates (1942), 
Flowers (1934), and Rydberg (1932) did list it as occurring in wet areas. 
As the lakes dry up large numbers of this species are observed covering ex- 
tensive areas, but the plants are so small, 1-10 cm. tall on this marsh, that 
it seems insignificant and can easily be overlooked. The salt content in the 
two lakes in which it was definitely observed was 0.22-1.26°.. This clearly 
indicates that E. acicularis must be tolerant of at least fairly high salinities 
or it could not exist 1n these saline lakes. 


Eleocharts rostellata has not been cited as a common species of inland salt 
marshes, but Fernald (1950) mentioned it as occurring in brackish or saline 
lakes, and Miller and Egler (1950) cite it for a coastal marsh in Connecticut. 
This species is the codominant in the Sedge-Meadow Community of this 
marsh, and it is also prevalent, but in greatly reduced numbers, in the S. 


Tue VEGETATION OF Tue Bic Sart Marsu 67 


pectinata-D. stricta Community. The salinity of the former locality was 0.18- 
0.68% and in the latter 0.20-1.00°., but if it appears to occupy more saline 
areas it can always be associated with increased local moisture or drainage 
ditches where the salinity is greatly reduced. From its distribution it could 
be said that E. rostellata is tolerant of salinities below 1°%, but was most 
abundant between 0.18% and 0.68°.. When present at higher salinities, the 
plants did not flower or fruit. 

Eustoma grandiflora and E. grandiflora f. fishert have not been men- 
tioned in the literature as being particularly tolerant of high salinities. This 
species was found, however, under similar conditions as Flaveria campestria. 
The bulk of this species was located in the western part of the S. pectinata- 
D. stricta Community, where the salinity was 0.27-1.00°., while some speci- 
mens were found in the Sedge-Meadow and other moist non-saline areas of 
the marsh. E. grandiflora and its white form were never found in any areas 
more saline than that mentioned above, and one should not consider it an 
inhabitant of the most saline areas but rather a species which reaches its 
maximum salt tolerance in the S. pectinata-D. stricta Community. 


Penfound and Hathaway (1938) and Kurz and Wagner (1957) have 
indicated that Fimbristylis castanea has been found growing in a wide range 
of salinity conditions on certain coastal marshes, but no record of its occur- 
rence on inland salt marshes has been found. On this marsh it was limited 
to the Sedge-Meadow Community, where the salinity was 0.18-0.68°.. It was 
not found invading more saline areas and can not be considered as indicator 
of the more highly saline soils. 


Flaveria campestris (F. angustifolia of manuals) has been listed by Hitch- 
cock (1898), Harirs (1920), Harrington (1954), and Jantzen (1960) as being 
present in saline soils. On the Big Salt Marsh F. campestris was found in 
the Sedge-Meadow, Fresh Water Seep, S. pectinata-D. stricta communities, 
and in other areas bordering drainage ditches. In actuality, this species does 
not invade the open salt flats of this marsh as mentioned by Jantzen (1960), 
but it does border these places along drainage ditches and other moist sites 
with a lowered salinity. F. campestris is widely scattered in these areas 
during the late summer and fall months. The actual salinity in the areas 
of its occurrence was 0).18-1.00°%, with 1.00°. being about the maximum in 
any area in which it might occur. It is definitely capable of entering areas 
with a salinity below 1.00°, but should not be considered one of the primary 
invaders of the bare salt flats. 


Haplopappus phyllocephalus subsp. annuus is cited by Hall (1928) as a 
taxon which occurs in saline areas. This species occurred only in a single 
location on the Big Salt Marsh at the base of a sandy hummock, where it 
was associated with A. exilis and A. ericoides. The majority of specimens 
were high up on the sandy hummock, but a few plants were found lower 


68 Tue University SciENcE BULLETIN 


down and their bases were covered with a salt crust. The salinity in this 
area was 0.003-0.05°.. Seed germination studies, described later, indicate 
that this species can germinate in salinities up to 27% at 20°C. 

Heliotropium curassacivum, a succulent with a shallow taproot, was 
located in the S. paludosus-D. stricta Community and was found also along 
the east-west road bisecting the marsh. This species is cited by Fernald 
(1950) as occurring in saline marshes. On this marsh its chief occurrence 
was in the S. paludosus-D. stricta Community where the salinity was 0.22- 
1.10%. It was tound also in the more saline D. stricta-S. depressa area. This 
species is tolerant of brackish conditions, but does not usually occur in situa- 
tions where the salinity is appreciably over 1%. 


Hordeum jubatum, has been cited by Schaffner (1898), Kearney et al. 
(1914), Coupland (1950), Ganong (1903), and Weaver and Albertson 
(1956) as being present in certain saline soils. Kearney et al. (1914) describe 
it as occurring in a community dominated by D. stricta, and Schaffner 
(1898) reports it as being found in a community in which species less char- 
acteristic of halophytic regions are trying to gain a foothold. On the Big 
Salt Marsh it occurred in many communities including the Sandhill Mixed 
Prairie, Tall D. stricta, Hummock, Sedge-Meadow, S. pectinata-D. stricta 
and S. paludosus-D. stricta. The total range of salinity under which it was 
found was 0.003-1.10°, but it was most abundant in the Tall D. stricta Com- 
munity and wetter saline areas where the salinity was 0.29-0.94°.,. Figure 6 
shows a colony of H. jubatum scattered through an area south of the sand- 
hill region occupied by D. stricta and S. airoides, where the salinity on 
August 19, 1960 was 0.85% in the surface 10 cm., 34% at 60 cm., and the 
chlorinity was 0.53% and 0.19%,. This species was widely scattered through 
wetter saline areas, but it was never particularly abundant in any one place 
and did not invade communities with salinities much over 1°. 


Hordeum pusilum is considered a member of slightly saline communities 
by Parodi (1930), Ragonese and Covas (1947), and Hitchcock (1950). It 
was found in the Sandhill Mixed Prairie, Grazed Sandhill, and Hummock 
communities of the Big Salt Marsh, where the salinity range was extremely 
low. In the Grazed Sandhill Community where it was most prevalent the 
salinity was 0.007%, whereas in the prairie area it was 0.003-0.02%. The 
occurrence of H. pusillum in only the least saline communities of this marsh 
indicated that, at least in this local area, the species was not tolerant of a 
wide range of salinity. 

Iva annua (Iva ciliata Willd.) has been cited by Schaffner (1898) as 
occurring along with D. stricta and S. depressa in saline communities, but 
only in rare instances. On this marsh it was found bordering the Prairie 
Community, in the S. pectinata-D. stricta Community, and bordering the 
D. stricta-S. depressa Community in one area where the former comes in 


Tue VEGETATION OF THe Bic Sart Marsu 69 


~contact with it. The total salinity in the areas in which this species exists 


was 0.003-1.00%, but it was found mainly in the part of the S. pectinata-D. 
stricta area in which the salinity was 0.27-1.00°.. This species was extremely 
rare on the marsh and only a few patches of 10-20 individuals were found. 
It was not commonly present in areas with a salinity higher than 1°%% and 
should not be considered one of the important indicators of high salinity 
in this area. 

Muhlenbergia asperifolia has been cited by Gates (1936) as being found 
in saline areas. This species was located only in the sandy hummock areas, 
seepage area, and along roadsides where the salinity was 0.003-0.50°. in the 
surface 10 cm. It was definitely not an indicator of highly saline ground,. 
and its extremely rare occurrence did not provide a great amount of infor- 
mation to judge its entire range of distribution. 

Stone (1959) reported that Myosurus minimus exists in soils which are 
from moderately to highly alkaline. It was found on this marsh only in 
the seepage area on the sides of small, 15-30 cm. high mud ridges. The 
waters in this area of the marsh have the lowest salinity, 0.02-0.05°%, of any 
pond site. The low salinity of these waters indicates that M. minimus grows 
well in areas of fresh water and the lack of its appearance in or around any 
of the more saline ponds demonstrates that, at least in this area, it has not 
been successful in spreading into saline soils. 


Panicum virgatum has been listed by Penfound and Hathaway (1938) 
and Kurz and Wagner (1957) as being found under a wide range of salinity 
conditions on certain coastal marshes. However, there are no reports of it 
in saline inland marshes. On the Big Salt Marsh, it was found in the 
Prairie, Sandy Hummocks, and Sedge-Meadow communities. It reached its 
greatest abundance in the Prairie and Sandy Hummocks where the salinity 
was 0).003-0.02°%, and was sparsely scattered through the sedge-meadow area 
where the salinity was 0.18-0.68°,. This species showed no signs of being 
especially salt tolerant and was not found entering highly saline areas on this 
marsh. 


Pluchea purpurascens has been considered an inhabitant of brackish and 
saline marshes by Fernald (1950) and Gates (1940). This species was found 
in the northern sector of the Tall D. stricta Meadow Community, bordering 
a large lake. The lake waters which overflow into this area had a salinity 
of 0.22-0.48°, and the soils 0.29-0.63°.. Only a single plant was found at this 
location and since it was observed in only one other place along a drainage 
ditch, this species must be considered tolerant of only the lowest salinities. 

Poa arida has been cited by Hitchcock (1950) as a member of saline com- 
munities. On this marsh it occurred in the Tall D. stricta, S. pectinata-D. 
stricta, Sedge-Meadow, S. airoides, S. paludosus-D. stricta and the D. stricta- 
S. depressa communities. It was not very abundant and only of local im- 


70 Tue University SCIENCE BULLETIN 


portance even though it occurred in many communities and in a fairly 
wide salinity range. As noted in Table 55, the total range of salinity in the 
surface 10 cm. which it tolerated was 0.04-1.60°%; however, at the location 
where it was most abundant, the Tall D. stricta Community, the salinity 
was ().29-0.94°%. 

This plant is a perennial and is the first species to be found flowering in 
the saline areas of the marsh. After flowering in May, it again begins vege- 
tative growth and by fall new culms develop. This is probably advantageous 
and allows it to flower early in the spring. P. artda is of secondary importance 
in most of the areas in which it occurs and, depending on the species which 
accompany it, it can be considered a fair indicator of moderate salinities. 

Polygonum lapathifolium has not been reported as being particularly 
salt tolerant, but it did occur in some wet, saline areas of this marsh. It was 
found in the sedge-meadow area and in a wet D. stricta Community in the 
southern part of the marsh, where the salinity was 0.18-0.68%. This species 
is definitely not an indicator of high salinities, but, as with many others 
which occurred in the Sedge-Meadow Community, it reached its peak of 
salinity tolerance in this area. 

The next species, Polygonum ramosissimum, has been cited by Schaffner 
(1898) as a characteristic species of the barren salt flat. On the Big Salt 
Marsh it occurred in the Sandy Hummock, S. paludosus-D. stricta, and D. 
stricta-S. depressa communities, and at times when these bordered the open 
salt flats it was associated with S. depressa. In the main areas in which it was 
growing, the salinity was 0.27-1.10°., and at the sites where it entered more 
open areas the salinity was slightly higher. P. ramosissimum should not be 
considered one of the main invaders of the open salt flats in this area be- 
cause it occurred chiefly in the communities mentioned above and only in 
one area did it invade the S. depressa community. 


Potamogeton foliosus occurred in a few of the shallow saline lakes on the 
marsh. Only a few widely scattered plants were found in the lake bottoms 
and it was usually rare. It is cited by Muenscher (1944) as occurring in 
brackish lakes and ponds. The waters it occurred in on this marsh had a 
salinity of 0.22-1.26%, which indicates that it can tolerate fairly high 
salinities. 

Rumex maritimus var. fueginus has been cited by Fernald (1950) as being 
an inhabitant of salt marshes. This species was found in the same locality as 
P. purpurascens, bordering a large lake in the northern part of the Tall D. 
stricta Community. As in the case of P. purpurascens, it was extremely rare. 
The salinity in this area was 0.22-0.63%, and, as mentioned earlier, the 
salinity of the lake waters which flood it was 0.22-0.48°.. Since R. maritimus 
var. fueginus was not found entering more saline areas it must be con- 
sidered tolerant of only the lowest salinities. 


H EGETATION Or Tue Bic Sa ARSH 
THE VEGETATION OF Tue Bic Satt M 71 


Salicornia rubra has been cited by numerous authors, including Aldous 
and Shantz (1924), Kearney et al. (1914), Flowers (1934), Weaver and 
Clements (1938), Shreve (1942), Rawson and Moore (1944), Billings (1945), 
and Coupland (1950), as occurring in inland areas containing extremely high 
salinities. Flowers (1934) described it as one of the pioneer species of the 
halosere of the Great Salt Lake and Kearney et al. (1914) cite it as one of 
the most salt resistant species in Tooele Valley, Utah. Schaffner (1898), 
Hitchcock (1898), and Penfound (1953) do not mention it as occurring in 
the Kansas or Oklahoma marshes they studied. Prior to this study the only 
known collection of this species in Kansas was made by Carleton in 1892 
in Stafford County, but no habitat or exact locality data were found with 
the herbarium specimen. S. rubra was not listed by Waterfall (1952) as 
occurring in Oklahoma, indicating, as is also cited in Muenscher (1944), 
that in the central states it reached its southern limits of distribution at this 
collection site in Kansas. 


If one considers that this species is one of the primary invaders of open 
salt flats in Utah and Canada (Flowers (1934); Coupland (1950) ), and while 
in this area it is found in only one extremely small locality, it becomes clear 
that it cannot spread as rapidly as it does in other areas. 


As mentioned by Kearney et al. (1914) S. rabra is a shallow-rooted an- 
nual with a taproot that usually does not penetrate deeper than 30 cm. It 
was found at only one location on the marsh where the salinity was 1.16- 
2.757. Accompanying it were two other halophytic species, Swaeda depressa 
and Sesuvium verrucosum. The high salinities in these salt flat areas indi- 
cate clearly that species located in them are definitely halophytic. The fact 
that S. rabra does not spread can be explained by the fact that no viable seed 
are produced and none were found as late as September 1959. If this is 
always true, S. rubra must be brought into this area anew, at intervals, by 
birds in their migration flights. Another reason for the latter assumption 1s 
that the 1959 colony could not be found again in 1960. A second possibility 
might be that germination of seeds is delayed and only during certain years 
are conditions just right for germination. 


Scirpus americanus has been reported by Rydberg (1932), Flowers (1934), 
Taylor (1938), and Penfound (1953) as occurring in saline marshes. Miller 
and Egler (1950) mention it as forming a community with E. rostellata in 
a Connecticut coastal marsh, just as it does in Kansas. This species was found 
in the Sedge-Meadow and S. pectinata-D. stricta communities, along drainage 
ditches, and in patches on wetter, less saline, parts bordering the D. stricta- 
S. depressa Community. The range of salinity under which it was growing 
was ().18-1.00°% in the surface 10 cm. and the range in the Sedge-Meadow 
Community where it was one of the major dominants was 0.18-.68°.. At 
times when this species was found in salinities above 0.80%, the plants were 


72 Tue University SciIENCE BULLETIN 


greatly stunted and produced no fruit, indicating that it can exist in areas 
of moderate but not high salinity. 


Scirpus paludosus (S. campestris Britt.), has been described by Penfound 
(1953), Schaffner (1898), and Flowers (1934) as occurring in slightly saline 
areas. On the Big Salt Marsh this species occurred in the Tall D. stricta, 
Sedge-Meadow, S. pectinata-D. stricta, D. stricta-S. depressa, and the S. 
paludosus-D. stricta communities. It was found also along drainage ditches 
and bordering some of the saline lakes. As mentioned by Flowers (1934), 
this species usually occurs in areas where the salinities are low or reduced. 
The total range of salinity under which S. paludosus grew was 0.18-1.10%, 
and it reached its highest basal area in a single transect of the Tall D. stricta 
Community (Table 14), where the salinity was 0.29-0.947,. As mentioned 
above, S. paladosus was found also in the more saline D. stricta-S. depressa 
Community, but when found there it was usually associated with some 
source of added moisture that tended to alleviate the high salinities. In some 
cases it was found in areas of higher salinity in this community, but no fruit 
was produced and the plants were greatly stunted. This species, as well as the 
previously mentioned S. americanus, spreads by underground rhizomes 
which gives it some advantage in occupying saline areas. 


Sesuvium verrucosum has been cited by Gates (1940) and Penfound 
(1953) as occurring on highly saline soils. This species is succulent in habit 
and is one of the primary invaders of the open salt flats. It was considerably 
more abundant and more widely scattered than Salicornia rubra but must 
be considered rare in comparison to Suaeda depressa. The total salinity 
range in the salt flat areas in which it grew was 1.16-2.75°.. Occasionally a 
few scattered individuals were in parts of the D. stricta-S. depressa Com- 
munity where the salinity was 0.82-2.45°,. The salinities which this species 
tolerated clearly indicate that it is halophytic. It is listed as a perennial by 
Rydberg (1932), but for all practical purposes it acts as an annual when 
occupying the open salt flats. It, as well as S. rubra and S. depressa, have a 
shallow taproot which does not survive from one season to the next. One 
factor that tends to eliminate S. verrucossum is that it can not stand sub- 
mergence for long periods, whereas, in the areas in which it occurs it has to 
contend with this factor at some time during the growing season. 


Hitchcock (1936) and Gates (1950) have indicated that Setaria geniculata 
occurs in salt marshes. On this marsh it was always rare, but it was found 


in the Sandhill Mixed Prairie, Sedge-Meadow, Grazed Sandhill, and Sandy | 


Hummock communities. The total salinity in the surface 10 cm. of the 


areas in which it occurred was 0.003-0.68°% and in the area where it was most 
prevalent, the Grazed Sandhill, the salinity was 0.003-0.02°.. The appearance 
of S. geniculata in many non-saline communities indicated that 0.68°, was 
probably the highest salinity tolerated by this species on the marsh. Since 


Tue VEGETATION Or THe Bic Sarr Marsu 73 


it did not occur in areas of higher salinity, it should not be considered an 
indicator of highly saline soils in this locality. 

Spartina pectinata has been mentioned by Hitchcock (1950), Muenscher 
(1944), and Fassett (1957) as being a member of brackish coastal communi- 
ties. It was found mainly in the S. pectinata-D. stricta Community, but it 
occurred also in the Sedge-Meadow Community, on sandy hummocks, in 
wetter parts of the prairie, bordering the most northern saline lake along 
drainage ditches, and in the wetter and less saline portions of the D. stricta- 
S. depressa Community. The salinity in the areas where it occurred was 
0.18-1.00°.. All salinities over 1°5 the plants were greatly stunted and 
usually did not produce fruit. S. pectinata was characteristic of a wet, slightly 
saline community, but in no instances did it occur in the bare salt flats or 
areas with a continuously high salinity throughout the growing season. 

Sporobolus airoides has been cited as an occupant of inland saline areas 
by many authors, including Kearney e¢ al. (1914), Harris (1920), Shreve 
(1942), Penfound (1953), Aldous and Shantz (1924), and Hilgard (1914). 
Its distribution on the Big Salt Marsh was limited to the Sandhill Mixed 
Prairie, S. airoides, open salt flats, and the grazed area south of the sand- 
hills. The total range of salinity under which this species grew was 0.003- 
1.60°.. In the S. avroides Community where it was very abundant and had 
a high basal area, the salinity was 0.04-0.19°.. It was common also in parts 
of the southern grazed area where the salinity was 0.8595 on August 19, 
1960. A small portion of this area is illustrated in Fig. 6. 


The fact that this species was capable of establishing itself on the open 
salt flats indicates that it is tolerant of high salinities. Fig. 20 illustrates one 
of the hummocks formed by S. arordes in an open salt flat habitat. 


Another species on the marsh, Sporobolus pyramidatus, has been cited 
by Ragonese and Covas (1947) and Hitchcock (1950) as occurring in saline 
soils. This species occurred in one location in the southern grazed area, 
just south of the sandhills proper. It was extremely rare and found only 
at the site of Fig. 6, where it was associated with S. airoides, D. stricta and 
H. jubatum. The salinity in this area was 0.85°, in the upper 10 cm. A fuller 
account of other soil relations are in Table 55 and in the description of H. 
jubatum in this section. It is definitely tolerant of moderate salinities, but its 
rare occurrence and complete absence from any of the ungrazed saline com- 
munities indicate that it would not be useful as a primary indicator of saline 
conditions in this area. 

Sporobolus texanus has been cited by Schaffner (1898) as being a success- 
ful invader of saline areas in the northern Kansas marshes which he studied. 
It was found only in the sandy hummock area and bordering roadsides on 
the Big Salt Marsh. The salinity in these areas usually was 0.003-0.50°%. 
This low salinity range combined with the fact of its low abundance demon- 


74 Tue University SciENcE BULLETIN 


strates that this species should not be considered an indicator of salinity on 
the marsh. 

Suaeda depressa (S. erecta (S. Wats.) A. Nels.) has been reported by 
many authors, including Kearney et al. (1914), Flowers (1934), Gates 
(1940), Rawson and Moore (1944), Coupland (1950), and Keith (1958) as 
being a salt tolerant species. On this marsh it was found in the Tall D. stricta 
Meadow, S. depressa, D. stricta-S. depressa, 8. paludosus-D. stricta, S. pecti- 
nata-D. stricta, Sedge Meadow, and S. airoides communities. The total range 
of salinity in the surface 10 cm. of the area it grew in was 0.04-2.75°%, where- 
as, in the S. depressa Community, where it was one of the primary invaders 
of the open salt flats, the salinity was 1.16-2.75°. 

According to Flowers (1934) and Kearney et al. (1914) S. depressa is less 
tolerant of alkali than S. rabra, but in this marsh it comprised the greatest 
part of the vegetation in the areas of highest salinity, indicating that it is as 
salt tolerant as S. rebra in this locality. 

A seed germination study, discussed in fuller detail later, indicated that 
seeds of S. depressa could germinate in salinities ranging from distilled water 
to 4% sodium chloride. This may explain its wide range of distribution and 
its occurrence in the highly saline communities. One possible explanation 
for its lack of ability to spread out in the less saline communities might be 
its inability to compete for light and moisture in these areas. Schimper (1903) 
has stated that the competition of more vigorous plants excludes halophytes 
from all localities except the saline. 

A study of root penetration was made on 100 specimens of the depressed 
form of S. depressa on September 23, 1960. The depth of penetration at this 
time averaged 10.59 cm., varying within the colony from 6-15 cm. and at no 
time during the course of this study was root material of this species found 
penetrating below the 30 cm. level. The specimens found in this colony 
seemed to have developed from recently germinated seed, and they appeared 
less robust and slightly smaller than the depressed forms which appeared 
earlier in the season. These data and observations made earlier during the 
growing season indicate that S. depressa has a shallow rooting taproot. 

As cited in Table 5, the extensive areas covered by the D. stricta-S. 
depressa vegetation type had a salinity of 0.49-2.52°.. These species have two 
distinct habits of growth. S. depressa is a succulent with a shallow taproot, 
while D. stricta is a deeper penetrating grass which spreads by rhizomes. 
The broad distribution of the two species demonstrates that these two life 
forms are exceptionally favorable for widespread distribution in highly saline 
soils. 

Suaeda depressa wilts when submerged and finally dies off after long 
periods of submergence. This factor as well as the extremely high salinities 
which develop are probably the main reasons why it does not establish itself 
on certain portions of the open salt flats. 


Tue VEGETATION OF Tue Bic Sart Marsu iD 


Tamarix gallica has been cited as being tolerant of soils containing some 
salinity (Gates, 1940; Penfound, 1953; Hopkins and Tomanek, 1957). On the 
Big Salt Marsh it occurs on certain hummocks, and along with D. stricta 
(Fig. 11), bordering the barren salt flats. The soil salinity in the latter area 
was 0.72-1.45%. 

When Hitchcock (1898) visited this area he did not report T. gallica, a 
fairly conspicuous species, as growing in saline soils, and conversations with 
local residents indicate that it is of more recent introduction in the area. It 
was not found actively spreading until the summer of 1960, when a group 
of 10-30 seedlings were found along a natural drainage channel in the north- 
eastern part of the marsh. 

Seed germination studies directed by Tomanek (1957) indicate that this 
species can germinate almost as well at salinities of 0.50-1.00°, as it did at 
0.01-0.05°.. Germination at high salinities allows T. gallica to spread out 
through this area; however, it has not spread rapidly and if one considers 
the entire marsh area it must be considered rare in occurrence. As Merkel 
and Hopkins (1957) have stated, this species can produce both a superficial 
lateral root system and a deeply penetrating primary root which would allow 
it to reach all the available water at both levels, but even with these ad- 
vantages it exhibited a low abundance on this marsh. 

McMillan (1959), Penfound and Hathaway (1938), Taylor (1939), and 
Gates (1940) have indicated that Typha angustifolia occurs in saline areas. 
On the Big Salt Marsh it was limited to drainage ditches, including those 
which bordered the S. pectinata-D. stricta and Sedge-Meadow communities, 
and it was found growing also along the margins of one or two of the saline 
ponds. The salinity in the areas in which it occurred was 0.25-0.52%,. T. 
angustifolia was not widely distributed on the marsh and occurred in areas 
of only moderate salinity, as reported by Flowers (1934) in his work in 
Utah, so it should not be considered a characteristic species of highly saline 
areas. McMillan (1959) has indicated that the species is more restricted in its 
distribution to saline areas than T. /atifolia. In this area the two species 
occurred very close together and for most of their distribution had a similar 
range of salinity tolerance. 

Typha latifolia has been described as a member of saline communities by 
Rawson and Moore (1944), Penfound and Hathaway (1938), and McMillan 
(1959). This species occurred along drainage ditches bordering the Sedge- 
Meadow and S. pectinata-D. stricta communities, on the margin of saline 
ponds, and also in the seepage area of the marsh. It was found in areas where 
the salinity was 0.02-0.52°, which indicates that it has a slightly broader 
range of distribution than T. angustifolia. McMillan (1959) has reported 
that T. latifolia has a broad distribution in eastern Nebraska, whereas, T. 
angustifolia seems to be chiefly confined to saline areas. The areas on the 
marsh occupied by these species were low to non-saline. 


76 Tue Universtry SciENcE BULLETIN 


SEED GERMINATION 
Seeds of Bromus japonicus, Suaeda depressa, and Haplopappus phylo- 


cephalus subsp. annuus were grown in petri dishes on two sheets of filter 
paper. Two dishes containing 50 seeds or a total of 100 seeds were used for 
each salt concentration in each study. The dishes containing the seeds were 
placed in a dark room and allowed to germinate for thirty days. No special 
treatment was given the seed of B. japonicus or those of H. phyllocephalus 
subsp. annuus, but it was found from preliminary studies that the seeds of 
S. depressa would only germinate after their seed coats were ruptured. Seeds 
of all species studied were considered to have germinated when the plumule 
reached one centimeter in length or when the plumule and radicle had both 
developed. 

The species chosen for this study occupied different habitats. S. depressa 
was almost always found in the most saline areas, H. phyllocephalus subsp. 
annuus was found on the border of a sandy hummock which was marginal 
in position between saline and non-saline soil, and B. japonicus was never 
found on any but the most non-saline soils. The salinity in the upper 10 cm. 
of soil in which these species were found was 0.04-2.75° for S. depressa, 0.003- 
0.50% for H. phyllocephalus subsp. annuus, and 0.003-0.02°,, for B. japont- 
cum. Table 56 shows the germination ability of these species at different salt 
concentrations. 


Tasre 56. Percent germination after thirty day period.* 


Sodium chloride concentration 


Species 0.0 a> 50 IB) | 2 3 4 5 
S, depressad 20RC@, 2s. 27 3 53 37 38 23 21 19 — 
Rooms em pages eee 48 52 72 46 48 28 30 x — 
B: yaponicus) Ry i ee 63 46 31 4 — _— — xX xX 
Hi. phyllocephalus. 20... 40 36 25 14 12 5 — Xx D4 


subsp. annuus 20°C. 


* X = no test made, — = no germination. 


The results of these studies indicated that S. depressa reached its optimum 
germination percentages, 53 and 72°%, in a 0.50°. sodium chloride solution 
and that seeds of this species were able to germinate at all salt concentrations 
up to 4%. The seeds of B. japonicus and H. phyllocephalus subsp. annuus 
acted like most glycophytes in that they had the optimum germination in 
distilled water and at all other salt concentrations there was a continuous 
drop in germination with a rise in salinity. This reduction in germination 
with increased salinity was also reported by Poma (1922) and others. 

The seeds of B. japonicus which did not germinate in a 1, 2 or 3% salt 
solution were placed in distilled water and after thirty days these yielded 


Tue VEGETATION Or Tue Bic Sart Marsu ry 


germination values of 83, 89, and 90°, indicating that the high salt con- 
centrations main effect was osmotic and not a toxic effect of the sodium or 
chlorine ions. The same procedure was carried out with seeds of S. depressa 
which did not germinate in a 5°, sodium chloride solution, and in this 
case 50° of the seeds germinated. MacKay and Chapman (1954), in their 
study of seed germination of Swaeda australis Moq. var. novazelandica and 
Mesambryanthemum australe Sol. ex Forst., have reported also that high 
salt concentrations were not toxic and that a transfer of seeds to tap water, 
from high salt concentrations, would allow them to germinate. 

S. depressa, one of the primary invaders of the bare salt flats, could ger- 
minate in salinities ranging from that of distilled water up to 4% sodium 
chloride, but its actual distribution was limited mainly to areas where the 
salinity was 0.49-2.75°., and only in extremely rare cases was it found at 
lower salinities. 

The germination rates obtained for S. depressa at all salt concentrations 
up to 4% indicated that this species can germinate in a much wider range of 
salinities than other species of Suaeda studied by Chapman (1947, 1947a) and 
MacKay and Chapman (1954). Chapman (1942, 1947, 1947a, 1960) reached 
the conclusion that many marsh species did not germinate until the salinity 
was reduced. 


PIoNEER HaLopuytTic VEGETATION IN INLAND SaLt MarsHEs 


Chapman (1960) has stated that areas which have a Distichlis or Suaeda 
community occupying the most saline areas in the interior of North America 
should be considered salt marsh, whereas the drier areas in the interior con- 
taining a primary community of Salicornia or Allenrolfea should be con- 
sidered salt desert. In the study of the literature concerning these communi- 
ties, and as Chapman himself has indicated, it seems that this division is more 
arbitrary than real. 

The saline areas of north-central Kansas described by Schaffner (1898) 
were bordered by a Distichlis stricta area and he cited D. stricta, Suaeda de- 
pressa and Polygonum ramosissimum as the characteristic plants of bare 
areas. On the Big Salt Marsh S. depressa, Sesuvium verrucosum, and rarely 
Salicornia rubra occurred on the barren salt flats, while the next most saline 
areas were dominated by a D. stricta-S. depressa Community. In Oklahoma 
marshes, Penfound (1953) listed S. verrucosum as forming the pioneer com- 
munity on the salt flats and D. stricta as marginal. Pound and Clements 
(1900) have stated that D. stricta, Suaeda depressa, Salicornia rubra, Atriplex 
hastata, A. argentea and P. ramosissimum occurred in saline areas of Nebras- 
ka. Rawson and Moore (1944), Coupland (1950) and Keith (1958) all have 
cited D. stricta, S. depressa, and S. rubra as occurring in the most saline 
areas of the Canadian marshes which they studied. Flowers (1934) and 
Kearney et al. (1914) have listed S. rabra as one of the most important 


78 Tue Universiry SctENcE BULLETIN 


invaders of the bare salt flats and D. stricta and S. depressa in slightly less 
saline communities. This would seem to indicate that in the northern and 
western areas studied, S. rubra is always one of the most important invaders 
of the bare salt flats. In Kansas this species is of little importance, but S. 
depressa becomes extremely important, while in Oklahoma S. verrucosum 
seems to be the pioneer species. 

In all the studies reviewed, D. stricta covers a large part of the highly 
saline areas and should be considered one of the most characteristic species 
of the inland salt marshes and deserts of the United States and Canada. The 
fact that many workers such as Billings (1945), Hilgard (1914), Keith 
(1958), Flowers (1958), and others, including this author, have cited it as 
occurring under a wide range of salinity conditions indicate that it is not 
always a good indicator of highly saline soils. However, as mentioned earlier 
in this paper and by Schaffner (1898), this plant is dwarfed under conditions 
of high salinity, and in this study it was found to have a very low basal area 
under these conditions. Another helpful factor in determining whether the 
area is highly saline is the associated species. If these are succulents such as 
S. depressa and S. verrucosum, this indicates a high salinity. Kearney et al. 
(1914) also diagrammed Swaeda as being a dwarfed form when occupying 
salt flat areas. 

Sporobolus aroides has been cited as occurring in moderately saline areas 
by Flower (1934), Kearney (1914) in Utah, Penfound (1953) in Oklahoma, 
and by the present author in this study, but no mention of it has been made 
in the works on Canadian marshes. Other species listed as being of some 
importance in Oklahoma marshes are Tamarix gallica, Scirpus americanus 
and S. paludosus. The first of these was of little importance on the Big Salt 
Marsh, but it did occur in areas where the salinity reached 1.45°%. S. amert- 
canus was limited mainly to fresh water or only brackish areas on the marsh 
and S. paludosus was found in slightly more saline areas, but, as Flowers 
(1934) and Penfound (1953) have mentioned, it is also usually limited to 
brackish areas and is not found in localities of the highest salinity. Other 
species play important parts locally in these areas, but the taxa mentioned 
above and in the former section on the salt tolerance of species have the 
broadest distribution in the United States and Canada. 


SUCCESSIONAL RELATIONS 


From the vegetational relations discussed earlier, possible trends of suc- 
cession can be described, based on the present conditions of vegetation dis- 
tribution. 

One can see from the successional relations of the vegetation suggested in 
Fig. 21 and in the previous analyses that two main trends appear to exist. 
One leads toward a line with a reduced salinity, but a retention of a high 


THE VEGETATION OF Tue Bic Sart Marsu 79 


amount of soil moisture and a Scirpus americanus- Eleocharis rostellata Com- 
munity. The second main trend occurs with a building up of soil, a reduc- 
tion in moisture and salinity and the formation of a mesic state in which the 
Sandhill Mixed Prairie Community is predominant. According to the cli- 
matic conditions in this area of Kansas one would expect this type of com- 
munity to develop. However, for this to occur in the area of the salt marsh 
the soil must be built up and the salinity and moisture content reduced con- 
siderably. 

A building up of soil occurs when vegetation occupies the more saline 
or wetter flat areas, causing an accumulation of organic matter and wind 
blown soil particles. One species which is tolerant of high salt content 
and is important in raising the ground level is Sporobolus airoides. The 
raised mounds formed by this and other species accumulate blowing soil 
from the surrounding sand-dunes and raise the general level of the land. 

It must be remembered that the successional trends delineated in Fig. 21 
may never occur. ‘The reason for this, as Polunin (1960) has mentioned, is 
that the conditions in these saline areas may never change and if they do the 
tendency would probably be toward an increase in salinity. This would 
make these communities static successional units since no successional 
changes could take place without a lowering of salinity or soil moisture. The 
reason for this is that the sedge-meadow plants can not tolerate the extremely 
high salinities of the salt flats and also by and large the prairie species can 
not generally occupy areas of higher moisture or higher salinities. 

Flowers (1934) has stated that Salicornia rubra tolerated higher salt con- 
centrations than most other species and that Swaeda depressa is also promi- 
nent in invading the strand. On the Big Salt Marsh S. rubra was extremely 
rare and found only in a single colony of 50-100 individuals, associated with 
S. depressa and Sesuvium verrucosum. It would seem, then, that in this 
area S. depressa in both its erect and depressed forms is definitely the primary 
invader of the bare salt flats, where the salinity was 1.16-2.75°%,. Another 
succulent, S. verrucosum, is of secondary importance and formed small, 
widely scattered colonies on the open salt flats, sometimes being associated 
with S. depressa and in only one instance with S. rubra. 

Distichhis stricta, as mentioned by Flowers (1934), Kearney et al. (1914), 
Hilgard (1914), and Billings (1954) is very versatile, being able to live under 
a wide variety of salinity relations, 0.29-2.52°., on the Big Salt Marsh. Flowers 
(1934) has stated that D. stricta does not seem to allow many other plants to 
enter areas which it dominates. The Tall D. stricta Meadow Community on 
this moderately saline and highly moist marsh was almost completely dom- 
inated by dense stands of the species and only a few scattered plants of other 
species were present. The D. stricta-S. depressa Community occupied very 
large areas of the marsh with a salinity averaging lower than the barren salt 
flats but in all cases over 1°. 


80 Tue University ScIENCE BULLETIN 


All other marsh communities seemed to radiate out from the S. depressa 
Community mentioned above, but D. stricta and not S. depressa was one of 
the major vegetational elements in many of these communities. As mentioned 
in the quantitative discussion, an S. airoides Community developed along 
the eastern margin of the marsh, forming a rather abrupt boundary to the 
Tall D. stricta community. The change can be best expressed by the quan- 
titative results, for, in the Tall D. stricta area, D. stricta had a relative density 
of 94.46% and a basal area of 11.35% or 92.65% of the total, whereas, in the 
adjoining S. airoides Community it was much sparser, having a relative 
density of 5.85°% and a basal area of 0.10% or 1.81% of the total. This S. 
airoides Community then blends into the Sandhill Mixed Prairie area, which 
indicates, as noted in Fig. 21, that the development of the prairie need not 
pass through a S. americanus-E. rostellata Community. 

The D. stricta Community, besides having the relations discussed above, 
seemed to be transitional to many of the other major communities. As illus- 
trated in Fig. 21, it leads to a S. americanus-E. rostellata area on the western 
margin of the marsh or to a Spartina pectinata-Distichlis stricta community 
first and then to the former. 

In one area it leads directly to a prairie community and in other areas lo- 
cated on the salt flats, a D. stricta or a D. stricta-S. depressa Community 
blends into a Scirpus paludosus-Distichlis stricta Community. 


CLIMAX RELATIONS 


In the Big Bend area of Kansas the dominant natural vegetation is of the 
Sandhill Mixed Prairie type. Under normal conditions, the climate of Staf- 
ford County, Kansas, with low rainfall, low humidity and drying winds 
would produce a Mixed Prairie Climax Association. In this localized area, 
however, a salt marsh vegetation develops which has several communities 
occupying different salinity ranges. 

Under the system of plant community classification described by Clements 
(1920), the salt marsh communities would not be considered on an equal 
level with the Sandhill Mixed Prairie Climax Association but must be con- 
sidered a sub-climax since the Sandhill Mixed Prairie was the one con- 
sidered to be in harmony with the climate in the area. The subclimax com- 
munities are produced by the local deposits of excess salts. 

An application of the continuum approach described by Curtis (1955, 
1959) to the communities on the Big Salt Marsh reveals that one can find 
a gradual change in species composition between two distinct areas in many 
cases, but in other instances the changes in community structure are abrupt 
and the distribution is definitely discontinuous. An example of a continuous 
relation exists between the species of the Scirpus americanus-Eleocharis 
rostellata and the Spartina pectinata-Distichlis stricta communities, whereas 


Tue VEGETATION Or Tue Bic Sarr Marsu SI] 


an extreme discontinuity exists the Sandhill Mixed Prairie or the Sporobolus 
aroides and Tall Distichlis stricta communities. Goodall’s (1954) statement 
that all vegetational classifications are more or less arbitrary may be true, 
but as he himself stated, some areas must be considered discontinuous and if 
no attempt is made to separate areas of different structure the results of 
vegetational analysis would lead to great confusion. 

The evidence seems to indicate that this should not be considered a Salt 
Marsh Climax Association, but rather a series of subclimax associes caused 
by the more or less permanent presence of excess salts. It seems clear that 
if the salts and excess water were removed, by changes in one or more of the 
edaphic factors discussed, the salt marsh vegetation would disappear and the 
Sandhill Mixed Prairie type would predominate. 


Facrors ControLtLiInG ComMMUNITY DistrRIBUTION 


A good introduction to this subject is a statement advanced by Clements 
(1920): “Every plant is a measure of the conditions under which it grows. 
To this extent it is an index of the soil and climate, and consequently an in- 
dicator of the behavior of other plants and animals in the same spot.” This 
statement is undoubtedly true and an attempt will be made here to explain 
the controlling factors in each of the marsh communities. The distribution 
of communities on this marsh clearly indicates that it was the upper limit 
of salinity that determined the species complement. 

For the most highly saline areas which included bare ground, the Swaeda 
depressa stands, and the Distichlis stricta-Suaeda depressa Community where 
the salinities reached 2.96°%, 2.75°% and 2.52°%, there was a direct correlation 
between decreased salinity and increase in species number and cover. The 
community with the next highest salinity was the Saline Pond which has a 
high 1.17%, and only two species of vascular plants, but in this case the 
moisture factor was decisive, since many species which are not aquatic, 
could not enter. However, saline conditions are also important in deciding 
species presence, as evidenced by the fact that the fresh water seep, with the 
lowest average salinity (0.03%), of any pond tested, had a completely differ- 
ent complement. 

The next three communities were fairly closely related by salinity condi- 
tions. The Tall D. stricta Community had an average salinity of 0.599% 
and a high of 0.949... There was no great increase in species number in this 
area; however, there was a great increase in basal area, and this community 
had the highest, averaging 12.25°., of any community on the marsh. The 
Scirpus paludosus-Distichlis stricta Community had an average salinity of 
0.45% and a high of 1.10%.-It contained only nine species, all of which must 
be considered at least slightly halophytic, indicating that the salt concentra- 
tion was the factor which prevented the invasion of the area by glycophytic 
species. The final community in this group, the Spartina pectinata-Distichlis 


82 Tue University ScrENCE BULLETIN 


stricta, had an average salinity of 0.51%, and high of 0.63°% and 1.00%. This 
community had an increase in species and cover over all the more saline 
areas except the Tall D. stricta which had a higher cover, indicating that 
the reduction of salinity allowed for the invasion of glycophytes and species 
of low salt tolerance. The former area was bordered by the highly saline 
soils of the D. stricta-S. depressa Community, and this accounts for the high 
salinity reached at the margin. The S. pectinata-D. stricta area also had a 
higher organic content than the S. paludosus-D. stricta area. 

The Scirpus americanus-Eleocharis rostellata Community indicated a low 
to non-saline wet habitat. Furthermore, there was an abrupt reduction in 
species number as well as cover when this community bordered a more saline 
area. In this area the average salinity was 0.40% and the highest 0.68%, in- 
dicating that the salinity was low. 

The Sporobolus airoides community reached its best development in an 
area where the salinity averaged 0.127 and the high was 0.19%, but in the 
southern grazed area of the marsh it was found growing perfectly well in 
dense colonies at a location where the salinity was 0.85%. and on the salt 
flats in an area where the salinity reached 1.60%. This community is only a 
good indicator of high salinity when S. airoides 1s associated with other 
halophytes. 

The prairie and grazed sandhill areas are a culmination of this line of 
development because they have the lowest salinity and moisture of any com- 
munity on the marsh. The differences in species composition between these 
two communities must be ascribed chiefly to grazing and not to any soil 
factor. 


This brief summary indicates that for the more saline communities 1n 
this area there is no difficulty in distinguishing the limiting factor as salinity, 
but in areas of low salinity, soil moisture and the associated factor of aeration 
play an important role. An example of this would be the completely different 
floras which develop in the prairie and sedge-meadow communities; soils 
in both communities are low in salinity, but the sedge-meadow has a fairly 
high water content, whereas the prairie is comparatively well drained. 


Good (1953), Steiner (1934), Harshberger (1909), and many others have 
also pointed out that salinity is the most important factor in determining the 
distribution of plants in a salt marsh habitat. In many communities, includ- 
ing the Sandhill Mixed Prairie, S. americanus-E. rostellata and S. pectinata- 
D. stricta, a reduction in salinity was accompanied by a large increase in 
species number. 

Others, as mentioned earlier in this paper, have mentioned that pH 
and soil moisture may be important factors controlling halophyte dis- 
tribution. On this marsh all areas were fairly moist, and it would be ex- 
tremely difficult to use this as a separating factor. Standing water and lakes 


THe VEGETATION OF Tue Bic Sart Marsu 83 


do play some part in preventing the succulent halophytes from invading 
certain highly saline areas, but when this surface moisture is gone these 
plants are able to enter. The pH factor, as mentioned earlier, showed little 
variation in the saline communities and from this study it can not be con- 
sidered very important. 


Adjustments to great variations in osmotic concentration of the soil 
solution must be made by the roots of species such as S. depressa, Sesuvium 
verrucossum, Salicornia rubra and D. stricta to allow them to survive. Some 
of these species as mentioned by Frey-Wyssling (1935) and Warming 
(1925) have special mechanisms which aid them. In the case of Saaeda, 
Salicornia and Sesuvium it is succulence, whereas, in Distichlis and some 
other grasses salts are excreted through specialized glands. 


SUMMARY 


1. The soil and water relations of the Big Salt Marsh, Stafford, Kansas have 
been described and discussed. 


2. A complete description of the seasonal and zonal distribution of the 
marsh species of plants was worked out. 


3. Quantitative studies of the marsh plant communities indicated density, 
basal area, and frequency relations within the various communities. 


4. It was found that the principal factor affecting the qualitative and quan- 
titative relations of distribution was soil salinity, but in the low and non- 
saline areas soil moisture played an important role. 


5. The salt tolerance of marsh species has been indicated. This demonstrated 
that it was the upper limit of salinity that was most potent in species dis- 
tribution. 


6. Seed germination studies with three species, Bromus japonicus, Haplo- 
pappus phyllocephalus subsp. annuus, and Suaeda depressa indicated that 
the first two species behaved as glycophytes and the latter as a halophyte. 
All species were found to germinate in a wider range of salinity conditions 
than their distribution indicated. 


7. The species which were found to have the greatest salt tolerance were 
Suaeda depressa, Sesuvium verrucosum, Salicornia rubra, and D1stichlis 
stricta. The latter species was the most versatile on the marsh. Of the 
remaining, S. depressa was the most abundant on the open salt flat, 
S. verrucosum was widely scattered and only of local importance, and 
S. rubra was found in only one very localized area. 


84 Tue UNiversity ScIENCE BULLETIN 


LITERAGURE, CITED 


Avpous, A. E. anp H. L. SHantrz. 1924. Types of vegetation in the semiarid portion of the 
United States and their economic significance. Jour. Agr. Res. 28:99-128. 

Anpreas, A. T. 1883. History of the state of Kansas. A. J. Cox and Co., Chicago. 

Bass, N. W. 1926. Geologic investigations in western Kansas. Kans. Geol. Sury. Bull. 11:90-95. 

Bittincs, W. D. 1945. The plant associations of the Carson desert region, western Nevada. 
Butler Univ. Bot. Studies 7:89-123. 

BorcHert, J. R., 1950. The climate of the central North American grassland. Ann. Assoc. 
Amer. Geog. 40:1-39. 

Bouyoucos, G. J. 1936. Directions for making mechanical analyses of soils by the hydrometer 
method. Soil Sci. 42:225-230. 

Braun-Biaunauet, J. 1932. Plant Sociology. McGraw-Hill Book Company, Inc., New York. 

Cain, S. A. ano G. M. Castro. 1959. Manual of vegetation analysis. Harper and Brothers, 
New York. 

CanrieLp, R. H. 1941. Application of the line interception method in sampling range vegeta- 
tion. Jour. For. 39:388-394, 

CuHapMan, V. J. 1942. The new perspective in the halophytes. Quart. Rev. Biol. 17:291-311. 

. 1947. Suaeda maritima (L.) Dum. Jour. Ecol. 35:293-302. 

.1947a. Suaeda fruticosa Forsk. Jour. Ecol. 35:303-310. 

. 1960. Salt marshes and salt deserts of the world. Interscience Publishers Inc., New York. 

CLEMENTS, F. E. 1928. Plant succession and indicators. H. H. Wilson Co., New York. 

Correy, G. N., T. D. Rice et al. 1912. Reconnoissance soil survey of western Kansas U.S. Dept. 
of Agric. Bur. of Soils. 

Coupxanp, R. T. 1950. Ecology of the mixed prairie in Canada. Ecol. Monog. 20:271-315. 

Curtis, J. T. 1955. A prairie continuum in Wisconsin. Ecol. 36:558-566. 

. 1959. The vegetation of Wisconsin. Uniy. of Wisconsin Press, Madison. 

Daupenmire, R. F. 1959. Plants and environment. John Wiley and Sons Inc., New York. 

Evans, L. T. 1953. The ecology of halophytic vegetation at Lake Ellesmere, New Zealand. 
Jour. Ecol. 41:106-122. 

Fasserr, N. C. 1957. A manual of aquatic plants. The Univ. of Wisconsin Press, Madison. 

FerNaLp, M. L. 1950. Gray’s Manual of Botany. American Book Co., New York. 

Fiora, S. D. 1948. Climate of Kansas. Rept. Kans. St. Bd. Agric. 67:1-320. 

Fiowers, S. 1934, Vegetation of the great salt lake region. Bor. Gaz. 95:353-418. 

Frey-Wyssiine, A. 1935. Die Stoffauscheidung der hoheren Pflanzen. Berlin. 

Ganonc, W. F. 1903. The vegetation of the Bay of Fundy salt and diked marshes. Bot. Gaz. 
36:161-186, 280-302, 349-367, 429-455. 

Gates, F. C. 1933. Grasses in Kansas. Rept. Kans. St. Bd. Agric. 51:1-226. 

. 1940. Flora of Kansas. Contribution No, 291, Dept. Bot., Kans. St. Coll. 1-226. 

Goop, R. 1953. The Geography of the flowering plants. Longmans, Green and Co., London. 

GoopaLit, D. W. 1954. Vegetational classification and vegetational continua. Festschrift fur 
Erwin Aichinger 1:168-182. Springer, Vienna. 

Hari, H. M. 1928. The genus Haplopappus. Carnegie Inst. Wash. Publ. 389. 

Harrinoton, H. D., 1954. Manual of the plants of Colorado. Sage Books, Denver. 

Harris, F. S. 1920. Soil Alkali. John Wiley and Sons, New York. 

HarsuBerGer, J. W. 1909. The vegetation of the salt marshes and of the salt and fresh water 

ponds of northern coastal New rJesey. Proc. Acad. Nat. Sci. Phil. 61:373-400. 

. 1911. Phytogeographic survey of North America. G. E. Stechert and Co., New York. 

.191la. An hydrometric investigation of the influence of sea water on the distribution 

of salt marsh and estuarine plants. Proc. Amer. Phil. Soc. 50:457-496. 

Hay, Rosert. 1890. Notes on some Kansas salt marshes. Trans. Kans. Acad. Sci. 12:97-100. 

. 1891. Geology of Kansas salt. Seventh biennial report of the state board of agriculture 

Kansas. 12:83-96. 

Haywarp, H. E. anp L. Bernstern. 1958. Plant growth relationships on salt affected soils. 

Bot. Rev. 24:584-635. 

, AND C. H. Wapteicu. 1949, Plant growth on saline and alkaline soils. Advances in 

Agronomy 1:1-35. 

Herser, C. H. ano K. L. Anperson. 1959. Response of the true prairie vegetation on major 
Flint Hills range sites to grazing treatment. Ecol. Monog. 29:171-186. 

Hircarp, E. W. 1914. Soils, their formations, properties, composition, and relations to climate 
and plant growth. Macmillan Co., New York. 

Hircucock, A. S. 1898. Ecological plant geography of Kansas. Trans. Acad. Sci., St. Louis 
8:55-69. 


Tue VEGETATION OF Tue Bic Sart Marsu 85 


. 1950. Manual of the grasses of the United States, U.S. Govt. Print. Office, Washington. 

Misc. Publ. 220. 

Hopkins, H. H. anp G. W. Tomanek. 1957. A study of the woody vegetation of Cedar Bluff 
reservoir. Trans. Kans. Acad. Eci. 60:351-359. 

HurtcuHinson News. 1886. October 6. 

Jantzen, P. G. 1960. The ecology of a boggy marsh in Stafford County, Kansas. Emporia State 
Res. St. 9:1-46. 

Jewett, M. E. 1927. Aquatic biology of the prairie. Ecol. 8:289-298. 

Kearney, T. H. et al. 1914. Indicator significance of vegetation in Toeele valley, Utah. Jour. 
Agr. Res. 1:365-417. 

Keirn, L. B. 1958. Some effects of increasing soil salinity on plants communities. Can. Jour. 
Bot. 36:79-89. 

Kutstap, R. O. 1959. Thickness and salt percentage of the Hutchinson salt. St. Geol. Sur. 
Kans. Bull. 137:241-247. 

Kurz, H. anp K. Wacner. 1957. Tidal marshes of the Gulf and Atlantic coasts of North 
Florida and Charleston, South Carolina. Fl. St. Univ. St. 24:1-168. 

Larra, B. F. 1950. Geology and ground water resources of Barton and Stafford Counties, 
Kansas. Univ. Publ. St. Geol. Sur. Kans. 88:1-228. 

Lyon, T. L., H. O. Buckman anno N. C. Brapy. 1952. The nature and properties of soils. 
MacMillan Co., New York. 

MacKay, J. B. anp V. J. CHapmMan. 1954. Some notes on Suaeda australis Mog. var. nova 
zelandica var. noy. and Mesembryanthemum australe Sol. ex. Forst. f. Trans. Roy. Soc. 
New Zealand 82:41-47. 

McMiiian, Carvin. 1959. Salt tolerance within a Typha population. Amer. Jour. Bot. 
46:521-526. 

Macistrap, O. C. 1945. Plant growth relations on saline and alkali soils. Bot. Rev. 11:181-230. 

MerkKeEL, D. L. ann H. H. Hopkins. 1957. Life history of salt cedar (Tamarix gallica L.) 
Trans. Kans. Acad. Sci. 60:360-369. 

Mitter, W. R. ano F. E. Ecter. 1950. Vegetation of the Wequetequock-Paweatuck tidal 
marshes, Connecticut. Ecol. Monog. 20:143-172. 

Muenscuer, W. C. 1944. Aquatic plants of the United States. Comstock Publ. Co. Inc., Ithaca, 
New York. 

Novikorr, G. 1958. Relations entre la structure de la vegetation et la stratification des milieux 
halophiles en Tunisie. Bulletin Du Service De La Carte Phytogeographique 1:69.83. 

Oostinc, H. J. 1956. The study of plant communities. W. H. Freeman and Co., San Francisco. 

Park, R. S. anp C. B. DeNNis. 1947. Physical land conditions affecting use, conservation and 
management of land resources. Stafford County, Kansas. U.S.D.A. Soil Conserv. Serv. 
and the Kans. Agric. Expt. Sta. 

Paropr, L. R. 1930. Ensayo fitogeographico sobre el partido de Pergamino. Rev. Fac. Agron. 
B. Aires 7:65-271. 

Penrounp, W. T. 1953. Plant communities of Oklahoma lakes. Ecol. 34:561-583. 

,AND E. S. Haruaway. 1938. Plant communities in the marshlands of southeastern 

Louisiana. Ecol. Monog. 8:1-56. 

Potunin, N. 1960. Introduction to plant geography. McGraw-Hill Book Co., New York. 

Poma, G. 1922. L’influence de la salinite de l’eau sur la germination et la croissance des 
plantes halophyles. Bull. Acad. Roy. Belg. Sci. 8:81-92. 

Pounp, R. ano F, E. CLeMents. 1900. The phytogeography of Nebraska. Jacob North and Co., 
Lincoln. 

Raconese, A. E. anp G. Covas. 1947. La flora halofila del sur de la provincia de Sante Fe 
(Republica Argentina). Darwiniana 7:401-496. 

Raunkiaer, C. 1934. The life forms of plants and statistical plant geography. Clarendon Press, 
Oxford. 

Rawson, D. S. anv J. E. Moore. 1944. The saline lakes of Saskatchewan. Can. Jour. Res. 
22D:141-201. 

Ricnarps, L. A. et al. 1954. Diagnosis and improvement of saline and alkali soils. U.S. Dept. 
Agric. Handbook 60. 

Rypperc, P. A. 1932. Flora of the prairies and plains of central North America. N.Y. Botani- 
cal Garden, New York. 

St. Jouns News. 1935. October 17. 

SCHAFENER, J. H. 1898. Notes on the salt marsh plants of northern Kansas. Bot. Gaz. 25: 
255-260. 

Scuimper, A. F. W. 1903. Plant geography upon a physiological basis. Clarendon Press, 

Oxford. 


86 Tue University ScrENcE BULLETIN 


ScHorwr, W. H. 1949. The geography of Kansas, Part II physical geography. Trans. Kans. 
Acad. Sci. 52:261-333. 

SHantz, H. L. anp R. L. Premetser. 1924. Indicator significance of the natural vegetation of 
the southwestern desert region. Jour. Agr. Res. 28:721-802. 

SHerman, R. 1956. Economic development in south central Kansas. School of Business, Bur. 
Bus. Res., Univ. Kans., Lawrence. 

Sureve, F. 1942. The desert vegetation of North America. Bot. Rev. 8:195-246. 

Sremner, M. 1934. Zur Okologie der Salzmarschen der Nordéstichen Vereinigten Staaten von 
Nordamerika. Jahrb. Wiss. Bot. 81:94-202. 

Stocker, O. 1928. Das Halophytenproblem. Erg. der Biol. 3:265-353. 

Stone, D. E. 1959. A unique balanced breeding system in the vernal pool mouse-tails. Evolu- 
tion 13:151-174. 

Taytor, N. 1938. A preliminary report on the salt marsh vegetation of Long Island, New 

York. N.Y. St. Mus. Bull. 316:21-84. 

. 1939. Salt tolerance of Long Island salt marsh plants. N.Y. St. Mus. Cire. 23:1-42. 

Tuompson, R. A. 1871. Original survey data located at the State Auditors Office, Topeka, 
Kansas. 

Tomanek, G. W. 1957. Research of salt cedar and other vegetation primarily at Cedar Bluff 
Reservoir, Kansas. Dept. of Botany, Div. of Bio. Sciences, Fort Hays Kansas State 
College. 

Upnor, J. C. To. 1941. Halophytes. Bot. Rev. 7:1-58. 

WarMiING, E. 1925. Oecology of plants. Oxford Univ. Press, London. 

WaTERFALL, U. T. 1952. A catalogue of the flora of Oklahoma. Oklahoma A. & M. College, 
Stillwater, Oklahoma. 

Weaver, J. E. 1954. North American Prairie. Johnsen Publ. Co., Lincoln, Nebraska. 

,AND F. W. ALBERTSON. 1956. Grasslands of the Great Plains. Johnsen Publ. Co., Lin- 

coln, Nebraska. 

, AND F. E. CLements. 1938. Plant Ecology. McGraw Hill Book Co., New York. 

Witcox, I. B. 1870. Original survey data located at the State Auditors Office, Topeka, Kansas. 


87 


THe VEGETATION OF Tue Bic Satt Marsu 


ic aS ena sapecaeeecee 
ae 


SERR MS GRR RS Sete 


(State 


iG, il. 


Location of the study area in the northeastern corner of Stafford County. 


Geological Survey of Kansas Map). 


mile square. 


County is divided into sections which are a 


88 Tue University ScrENcCE BULLETIN 


Fic. 2. Aerial photograph showing sites of quantitative vegetation sampling and water 
samples. 1-11 Water sample sites (see Table 2). A-L Vegetation sites (see Tables 
8, 27, 28, 29, 30 and 31). (U.S. Dept. Agric. photograph). One mile—41 milli- 
meters 


THE VEGETATION OF Tue Bic Sart Marsu 89 


Fic. 3. Tall Distichiis stricta community. This area has the highest basal cover on the 
marsh, 


Fic. 4. Scirpus paludosus-Distichlis stricta community. The tall species in this figure is 
S. paludosus. 


9() Tue University SciENcE BULLETIN 


Fic. 5. Early summer view of the Sandhill Mixed Prairie Community. Tall grasses are 
beginning to develop. 


Fic. 6. Saline area south of sandhills. Species present include Hordeum jubatum, Distich- 
lis stricta, Sporobolus pyramidatus and S. airoides. 


THe VEGETATION OF Tue Bic Satt Marsu 9] 


Fic. 7. Upright form of Suaeda depressa on the open salt flats. Light meter case is 1] cm. 
high. 


Fic. 8. Depressed form of Suaeda depressa on the open salt flats. 


2 


Tue University ScrENcE BULLETIN 


ae 


Fic. 9, Distichlis stricta-Suaeda depressa community. The cover is very low and the 
plants are dwarfed. 


Fic. 10. This excavated section shows the rhizomatous growth habit of Distichlis stricta. 


Tue VEGETATION OF THE Bic Satt Marsu 


See 


Fic. 11. Tamarix gallica is shown here invading a highly saline soil. The undergrowth 
is a dense cover of Distichlis stricta. 


Fic. 12. Spartina pectinata-Distichlis stricta community. 


94 THe Untiversiry ScrENCE BULLETIN 


Fic. 13. Spring aspect of the Scirpus americanus-Eleocharts rostellata community. 


Fic. 14. Summer aspect of the Scirpus americanus-Eleocharis rostellata community. 
Cicuta maculata is the most characteristic plant of this period. 


THe VEGETATION OF Tue Bic Satt Marsu 95 


Fic. 15. Fresh Water seep community. Some of the more characteristic species of this 
summer aspect are Typha latifolia, Sagittaria latifolia, and Scirpus validus. 


Fic. 16. A saline pond filled with a dense growth of Chara spp. 


6 Tue University ScrENcE BULLETIN 


Fic. 17. Late spring aspect of parts of the Grazed Sandhill community. The most char- 
acteristic species in this area is Monarda citricdora. 


Fic. 18. Early summer aspect of the Grazed Sandhill community. Flowering is Argemone 
polyanthemos. Other species in this vicinity are Eriogonum annuum and Conyza 
canadensis, but these are still in vegetative condition, 


THe VEGETATION OF THE Bic Sart MarsHu 97 


Fic. 19. A low hummock covered with a dense colony of Baccharis salicina. 


Fic. 20. A Sporobolus airoides hummock on the open salt flat. The dense root and 
rhizome system of this species forms the raised area. 


98 


Tue UNIversiry SCIENCE BULLETIN 


Re Grazed 
Bramic Sandhill 
Fresh-water S. americanus- 
Seep E rostellata 
S. pectinata ae 
D stricta > alroides 
nce ae 
D. stricta 
S. paludosus - 
a » strieta 
Saline D stricta- 
Pond S. depressa 


S. depressa 


Bare Ground 


Fic. 21. A hypothetical scheme of the major successional trends on the Big Salt Marsh. 


| 
| 


- = 
- _ ‘ - 
i " -o 
= 
= 7 
er eed 
& 
. ‘ 
r cs 
7 ‘ 
, 7 
a 
= - <= 
- 


we oy bea ‘1% 
me ' ! 
7 
a -_ 
7 
= ets 
7 7 
7 ae 
oF 
s 
a ; - 
' 
- : : 
cs = 
‘ 
; < é — 
ae - 
- : * a 
= i 5 7 
a oe . z 
os 7 7 


> 
- 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


FOSSORIAL ADAPTATIONS IN THE 
BANK SWALLOW, RIPARIA RIPARIA 
(LINNAEUS) 


By 
Abbott §. Gaunt 


VoL. XLVI Paces 99 To 146 May 1, 1965 No. 2 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 99 To 146 May 1, 1965 No. 2 


Fossorial Adaptations in the Bank Swallow, Riparia riparia 
(Linnacus)' 


By 


AssoTt S. GAUNT” 


ABSTRACT 


The Bank Swallow is the only North American swallow that characteristically 
excavates a nesting cavity. This habit seemingly is incongruous with the delicate 
structure that swallows have evolved, partly as an adaptation to aerial foraging. 
Behavioral and morphological studies were made to determine what, if any, 
modifications for digging have evolved in Bank Swallows. 

A key adaptation is behavior in burrowing. Unlike many burrowing birds, 
Bank Swallows do not excavate by thrusting a strong bill into the substrate in 
drilllike fashion, but instead they use a rapid, lateral slashing motion. This 
technique engenders torque around the long axis of the bill and around the 
midpoint of the frontonasal hinge. 

Various proportions based on measurements of the Bank Swallow’s skull were 
compared with similar proportions in skulls of six other species of North Ameri- 
can swallows. The Bank Swallow’s bill was shown to be smaller than bills of 
other swallows relative to cranial and sternal measurements, and more nearly 
circular in cross-section. The supports of the bill, though not enlarged, are func- 
tionally larger by virtue of the bill’s small size. Length of the skull compared to 
dry weights of the five paired muscles that control digging motions showed that 
these muscles are about as massive as would be expected in a swallow the size of 
a Bank Swallow. Structure and placement of these muscles do not vary signifi- 
cantly among the swallows studied. 

The small size and conical shape of the Bank Swallow’s bill probably repre- 
sent departure from the ancestral condition. Such changes confer the following 
advantages: (1) reduction of stress upon supports of the bill, (2) reduction of 
fractural stress, (3) creation of a structure capable of slashing into substrata in a 
variety of directions, and (4) slight increase in mechanical advantage. Seeming 
lack of change in either mass or structure of the muscles studied indicates that 


a 
"Submitted for publication in May, 1963. 
"Present address: Department of Biology, Middlebury College, Middlebury, Vermont. 


100 Ture Universiry ScrENCE BULLETIN 


these are preadapted to a slashing mode of excavation; that is to say, the action 
of slashing is judged to require the same muscles as those used in feeding motions. 

For the Bank Swallow, energy considerations alone demand that modifications 
of the bill or associated structures be such as to allow continued efficiency in 
feeding; such modifications meanwhile must also tend toward increasing efficiency 
in digging. Thus, these swallows, notably successful as burrowers, must either 
have evolved modifications that confer substantial gain in efficiency of digging 
with a minimum of change in the original structure, or they must have been 
preadapted to burrowing. The Bank Swallow shows both osteologic change and 
myologic preadaptation. 

The Rough-winged Swallow also nests in burrows but is restricted to use of 
pre-existing burrows or natural cavities. It is capable of modifying these but 
seems not to be sufficiently effective in digging to justify the expenditure of 
energy required to excavate an entire burrow. 

Swallows in general are faced with a paucity of suitable nesting sites. The use 
of burrows (probably beginning with use of burrows abandoned by other ant- 
mals), seems to be an adaptation to increase the numbers of available nesting 
sites. The evolution of burrowing species is a consequence of this adaptation. 


INTRODUCTION 


The past two years have seen a blossoming of work in avian functional 
anatomy. In 1961, Bowman published his opws magnus on adaptations 1n 
the Galapagos finches. The year 1962 saw the publication of both Zusi’s work 
on the head and neck of the Black Skimmer and the studies of Goodman 
and Fisher on the feeding apparatus in waterfowl. Taken together with 
Forbes’ (1882) classic study on the neck of darters, Burt's (1930) studies of 
modifications in woodpeckers, and Beecher’s work on American orioles, 
American blackbirds, and honey creepers (1950; 1951a; 1951b), these papers 
represent almost every possible approach to the field. Forbes’ paper presents 
the simplest approach, analysis of the functional adaptation of a given struc- 
ture, the Donitz bridge. The work of Zusi is an extension of Forbes’ ap- 
proach. Here the functional import of several structures is studied in relation 
to the highly specialized feeding behavior of the Black Skimmer. Burt 
studied the adaptations of a group in which all members are more-or-less 
specialized for the same ecologic role, that is, drilling into a solid substrate. 
Bowman’s interest lay in the evolution of divergent adaptations within a 
small, closely related group. Goodman and Fisher undertook an analysis 
of adaptations for a general function within an extremely large group. 
Beecher’s studies have a distinctly phylogenetic orientation. 

This paper presents an approach differing from all of the above. I have 
attempted an analysis of adaptations pertinent to a specific, seemingly im- 
plausible, pattern of nesting behavior in a single species of a highly spe- 
cialized group. 


FossortaL ADAPTATIONS IN THE BANK SwALLow 101 


The diverse genera and species of the family Hirundinidae may be re- 
garded as minor variations on a single morphological theme. They are all, 
without exception, highly specialized for feeding upon airborne arthropods. 
So stringent are the adaptive requirements for occupancy of this niche that 
the resemblance in external morphology of all the birds exploiting it is re- 
markable. The great degree of morphological similarity amongst the swal- 
lows and the taxonomic difficulties attendant thereto have been remarked by 
Mayr and Bond (1943). In their attempt to order the genera in this family, 
they were forced to rely on patterns of nest construction as much as, if not 
more than, on morphological characters. 


Among the adaptations shared by airborne aerial feeders are long wings, 
short neck, short legs, relatively weak feet, large gape, and short, flat bill. 
None of these seems conspicuously well-suited to the task of burrowing. 
Yet, the trait of burrowing nesting cavities into soil has arisen independently 
in three genera of swallows: Riparia (4 species, 2 confined to Africa, 1 Afri- 
can and Asiatic, 1 African and Holarctic), Pseadhirundo (1 African species), 
and Cheramoeca (1 Australian species). In addition, it is probable that at 
least some of the species of Psalidoprocne (11 African species) dig their nest- 
ing burrows (Serle, 1950). It is evident that any fossorial modifications must 
have evolved within the rigid morphological restrictions set by aerial feeding. 
The advantages of maintaining light weight seemingly have precluded such 
possible adaptations as massive bills or feet. Therefore, the question of just 
what modifications may be developed presents an intriguing problem to the 
functional anatomist and student of evolution. 

The fact that no extensive modifications are probable, if, indeed, possible, 
suggests that the digging swallows may have evolved behavioral techniques 
for the utilization of existing anatomical structures with little or no morpho- 
logical modification. If so, then analysis of the situation might provide addi- 
tional information about morphological preadaptation. The problem en- 
countered with the Bank Swallow closely parallels the situation suggested 
for study by Davis (1949): “The most promising situation for detecting 
morphological preadaptation should be in a form or group of related forms 
showing some specialized habit or other environmental relation that would 
be reflected in its morphology. The generic or family levels are most likely. 
... A further requirement is that the adapted type belong to a taxonomic 
group large enough to exhibit a range of morphological variation, i.e., large 
enough that the basic morphological plan of the group as a whole can be 
determined. Finally, knowledge of the ecological relations of both the 
adapted type and its unadapted or differently adapted relatives must be 
sufhciently intimate to reveal the functional relations of the preadaptation.” 
The last criterion is, perhaps, not met as well as might be wished. This fail- 


102 Tue University ScrENCE BULLETIN 


ing is balanced by the fact that “the basic morphological plan” of the swal- 
lows is both obvious and restricted. 

As an object of study in relation to the posed problem, the North Ameri- 
can Bank Swallow, Riparia riparia riparia, presents a number of advantages. 
Its natural history has been intensively studied by a number of authors, and, 
consequently, considerable information about its digging habits is available; 
it is common in Kansas, thereby facilitating the collection both of study 
specimens and of additional data on digging; moreover, the Museum of 
Natural History at The University of Kansas possesses a good series of 
skeletons of Riparia and other North American swallows. 


ACKNOWLEDGEMENTS 


I would like to extend my sincere appreciation to Profs. Richard F. John- 
ston, A. Byron Leonard, and Raymond C. Jackson for their counsel during 
this study. Dr. Johnston was my principal advisor and served as a sensitive 
sounding board for many of my ideas. Dr. Leonard was of great assistance 
in the preparation of photographic material. Special thanks are due to Dr. 
Walter J. Bock for volunteering his time and advice and critically reading 
the anatomical portion of the paper. Thanks are due also to Dr. Robert M. 
Mengel for his drawing of the swallows’ heads, to Mr. Tom Swearingen for 
preparing illustrations of the muscles and other figures, to Mr. Cathcart of 
the Holliday Sandpits for permission to do virtually as I pleased on his 
property, to Jon C. Barlow and Erwin E. Klaas for assistance in field studies, 
and to Hobson Wildenthal for his help with the physical aspects of the 
problem. Specimens examined were from the collection of the Museum of 
Natural History at The University of Kansas. Dr. Kenneth Armitage kindly 
permitted me the use of his laboratory equipment for obtaining the weights 
of muscles. Finally, many thanks to my fiancée, Miss Sandra Lovett, for as- 
sistance in field work, typing drafts, and general moral support. 


FIELD STUDIES 


The natural history and behavior of the temperate zone swallows have 
been generally well documented. Concerning the Bank Swallow, the obser- 
vations of Petersen (1955), coupled with the voluminous and detailed work 
of Stoner (esp. 1936), could provide virtually all the field information requi- 
site to the present study. In view of these extensive works, a detailed report 
of my field work would be redundant here, but a discussion of certain points 
is vital to the theme of this work. 

The field work in connection with this study was undertaken with three 
objectives in view: (1) to gain a personal acquaintance with the behavior of 
the birds: (2) to corroborate observations of previous authors in the light of 


Fossor1aL ADAPTATIONS IN THE BANK SwaLLow 103 


a different emphasis; (3) to compare the Bank Swallow with the Rough- 
winged Swallow (Stelgidopteryx ruficollis ). Observations were made in the 
springs and summers of 1960 to 1962. 


CoLonires STUDIED 


Two nesting sites were studied, one along the banks of a sand pit, the 
other in a bank of the Kansas River. The former site provided nesting area 
for both Riparia and Stelgidopteryx; the latter was occupied primarily, if not 
exclusively, by Riparia. 

The Holliday Sandpits comprise a roughly kidney-shaped pond, the 
north-west end of which is the site of sand dredging operations. The “hilus” 
of the pond is bordered by sand banks from one to eight feet high. All nests 
of Riparia and the majority of those of Stelgidopteryx were tunneled into 
these banks. The eastern and southern margins of the pond are partially 
composed of earthen banks. A few Stelgidopteryx nests were found there. 
This site provided nesting area for about 100 pairs of birds in the years 1960 
and 1961. In 1962, it was largely destroyed by flooding and was abandoned 
by the birds. 

Three and one-half miles east of Lawrence and about one and one-half 
miles southeast of the sand pits there is a long curve in the bed of the Kansas 
River. The south shore here is a sand bank about twelve feet high at normal 
spring and early summer river levels. Of this, about eight to ten feet are 
vertical and suitable for nesting. In the years 1959 to 1961, this bank afforded 
nesting sites for a Bank Swallow colony of about 100 to 300 pairs. In 1962, 
possibly as a result of the destruction of the sand pit site, two colonies were 
established at the river. Combined, these colonies included about 700 breed- 
ing pairs. 

I observed Rough-winged Swallows at the river site several times, but 
none was ever found nesting there. However, it is possible that some nests 
were built in abandoned Bank Swallow burrows. On several occasions, birds 
flushed from the banks showed the flashing display of white under-tail coverts 
characteristic of Stelgidopteryx. 


Burrows 


The Bank Swallow and the Rough-winged Swallow are the only North 
American swallows that characteristically nest in burrows, although the 
habit is not uncommon in certain other hirundinids. Despite the similarity 
in physical appearance, the nesting habits of the two species differ markedly. 
There are few credible reports of North American Bank Swallows nesting 
in other than burrows of their own construction, although the European 
Sand Martin is reported by Rennie (/n Burns, 1924) to nest in crevices, 


104 THe UNIversiry ScrENCE BULLETIN 


masonry, and tree cavities, by Henson and Johnson (1955) to occupy artifi- 
cial burrows, and by Hickling (1959) to nest in drainage pipes. Stockard 
(1905) reports finding a Bank Swallow's nest in Mississippi that was “placed 
in a deserted Kingfisher’s tunnel.” Considering that the rough-wing is more 
common in Mississippi, the great similarity of the eggs of the two species, 
and the fact that the Bank Swallow is highly colonial, this may be a case of 
mistaken identity. Such is almost certainly the case of the bird reported by 
Hammond (Coues, 1876) as nesting behind a loosened weather board of a 
saw mill. 

The Rough-winged Swallow is known to nest in a wide variety of sites. 
Lunk (1962) lists 11 categories of nesting sites used by this bird. 

The question of whether or not the rough-wing digs its own hole has 
been open for many years. ee ranged from Dingle’s (1942) flat af- 
firmative (“Burrows, excavated ... by the birds themselves, are the usual 
nesting sites of the roughwing.’ to Allen’s (1933) equally positive denial 
(“the Roustweiies ... never dig burrows.”). In relatively few reports is 
there any indication that the author actually saw the birds dig. Bailey (1913) 
describes the “drilling of the hole, which is made by the birds using their 
feet to scratch with, and push the dirt backward out of the tunnel. Unlike 
the Kingfisher, their beak plays a secondary part in the drilling of their 
home.” Burns (1924) also mentions that the rough-wing “is perhaps more 
apt to use its feet in scratching out the dirt” than the Bank Swallow. Eliot 
(1932) reported a pair which “began to excavate a hole .. . but gave up.” 
The hollowing of decayed roots by the rough-wing is described by van Fleet 
(1876). Sutton and Pettingill (1942) rather ambiguously reported the be- 
ginning of “nest-digging.” Weydemeyer (1933) bluntly stated, “I have ob- 
served them |rough-wings| digging burrows in banks of clay, of sand, and 
of gravel.” Lunk (1962) recorded rough-wings “kicking out a few spurts 
of sand” from Bank Swallow holes. He then continues: “Otherwise, how- 
ever, my evidence is entirely negative. The activity just described was quite 
out of the ordinary, and of a most desultory character. It appears that during 
my studies no Rough-wing nest, of nearly a hundred built in the area, was in 
a hole excavated by the birds themselves. Without exception, burrows were 
known, or could be presumed, to be unused or abandoned holes of other 
species. At certain times and places suitable holes were evidently at a pre- 
mium—yet not even then did the Rough-wings exhibit the slightest tendency 
to do any burrowing for themselves. I have no evidence that they even 
deepened or enlarged any existing burrow before building.” 

Most authors seem to have assumed that any bird found nesting in a hole 
must have dug the hole. The majority of rough-wing nests I have observed 
were in evidently abandoned Bank Swallow burrows. I have never observed 
rough-wings digging, although two pairs nested in banks avoided by Bank 


FossortaL ADAPTATIONS IN THE BANK SWALLOW 105 


Swallows. While these tunnels may have been excavated by rough-wings, it 
is here assumed that some hole already existed. It is most probable that Bank 
Swallows had used these banks in previous years and for some reason, pos- 
sibly texture of the substrate, had abandoned them. Also, slumping of these 
banks frequently exposed rodent burrows, and other holes were left by the 
falling of clay nodules that suffuse the area. Such holes might easily be 
adapted by rough-wings; Dawson (1923) remarked that the rough-wings 
“prefer some natural lead,—an old Kingfisher’s burrow, a hole left by a rot- 
ten root or a fallen stone, or something of that sort.” Tyler (1913) stated: “I 
believe that only in exceptional cases do the birds excavate their own nest 
cavity, as the hard formation of the [San Joaquin River} banks would seem 
to make such a task very difficult.” Strangely, the burrows I observed that 
were most likely dug or modified by rough-wings were in harder, less sandy 
banks than those in which I have found Bank Swallows digging. 

In summary, I concur with Lunk that it would be improvident to state 
that rough-wings never perform any extensive excavation of their own. 
However, the activity would appear to be unusual. For this reason the bird 
is treated as a non-burrowing swallow in later analyses. 

There have been several attempts to distinguish between holes of Riparia 
and Stelgidopteryx by the shape of the entrance. In reality, there is no 
question that both species may use either round or elliptical entrances. In the 
event that the rough-wing has used a previously existing tunnel, the problem 
becomes meaningless. I have noticed, however, that Riparia burrows that 
were probably enlarged by Stelgidopteryx tended to be modified more ex- 
tensively laterally than vertically. I will show that the rough-wing’s bill and 
musculature are, if anything, more suited to lateral than to vertical excava- 
tion. 

Two methods, neither completely satisfactory, were used by Dr. R. F. 
Johnston and myself to measure the depths of several hundred Bank Swal- 
low burrows. The first method, simply digging them out, was usually em- 
ployed when it was desirous to obtain data on clutch size. This method has 
the advantage of revealing which holes contain nests, but is tiring, slow, and 
necessitates an estimate rather than an accurate measurement of the depth. 
Moreover, it 1s occasionally impossible to dig out the deepest holes. The 
mean depth obtained by this method for 28 tunnels containing nests was 27.7 
inches with a range of 15 to 48. 

The second method involved the use of a calibrated aluminum rod. This 
method is fast, accurate, and permits measurement of both very deep and 
hard to reach holes. Unfortunately, it does not reveal which holes contain 
nests. Therefore, it was necessary to pick some minimum depth for purposes 
of calculating the mean depth. Stoner (1936) reported a minimum of 14 
inches and Petersen (1955) a minimum of 16 inches for completed burrows 


106 Tue UNiversity ScriENCE BULLETIN 


in the colonies they studied. Cooper (1955) reported a Sand Martin burrow 
only 3.5 inches deep and with no entrance tunnel. He was “quite certain that 
the hole had not been interfered with by human beings.” One wonders if a 
natural slump may not have been involved. I have observed swallows con- 
tinuing to use nests the tunnels to which had been completely eliminated by 
minor slumping of the bank. The minimum depth that I have observed for 
an entire, completed burrow was 15 inches. As this depth is exactly inter- 
mediate between those reported by Stoner and Petersen, I chose it as the 
minimum depth for calculations. It is obvious, however, that many incom- 
plete holes were measured, and the data, therefore, are biased toward shal- 
lowness. Using the rod technique, 122 tunnels yielded a mean of 27.2 inches 
and a range of 15 to 51 inches. Combining the measurements of all tunnels 
15 inches or more in depth, including some that were dug out but did not 
contain nests, | found 161 burrows yielded a mean of 27.3 inches and a range 


of 15 to 51 inches. 


EXCAVATION 


The excavatory behavior of the Bank Swallow is central to the theme of 
this study. Most of the salient points have been suitably described by Stoner 
(1936) as follows: 


“When starting a burrow it appears that the birds first cling to some slight 
projection on the face of the bank from which they can reach the point of attack, 
either with the claws or bill or both. After a time a slight concavity is formed, its 
sides creased with the marks of claws and bill. As the work proceeds, the ceiling 
takes on a distinctly arched form, while the floor is practically flat. 

“Observations indicate that the more deeply scarred appearance of the inner 
or apical part of the cavity is due to the use of the bill. The bird clings to the 
walls and dislodges the particles of sand or gravel by pecking with a rapid side- 
to-side movement of the head. As soon as a shelf has resulted the feet with their 
long sharp claws are brought into action. Both bill and claws of captured in- 
dividuals were sometimes caked with moist earth. In digging, the tail is frequent- 
ly used as a support, as in woodpeckers.” 


He continues: 


“With the deepening of the excavation the dislodged materials fall to the 
floor of the shallow burrow whence they are ejected by frequent vigorous kicks 
accompanied by a kind of wriggling movement of the body. In this action the 
wings also are ‘shuffled’ rapidly in a backward and forward direction as well as 
from side to side, thus aiding in whipping the sand out behind the bird.” 


A number of important points are to be noted in this description. Para- 
mount is the lateral motion of the bill. The Bank Swallow does not probe 
into the substrate in the manner of a woodpecker or kingfisher. Rather, the 


FossortaL ADAPTATIONS IN THE BANK SWALLOW 107 


head is moved through a series of arcs to alternate sides. The resulting 
cavity, thus, is carved, rather than drilled. In suitable substrate, the roof of 
the tunnel and nesting cavity bears a scalloped design. In 1962, a rise in the 
level of the river undercut the bank below the colonies. The subsequent 
slumping neatly sectioned several completed cavities, clearly revealing the 
tell-tale slashing pattern. The same pattern may be seen in newly begun 
holes (Figs. 8, 9). 

After the entrance to the tunnel has been formed, the beak becomes the 
major, if not the sole, instrument of digging. The feet may be used in exca- 
vation, especially in the lower portion of the tunnel; but, aided by the wings, 
they are more important in kicking the loosened sand out of the tunnel. In 
consequence, the swinging motion of the head prescribes an arched roof, 
while the wings and feet scrape out a floor that is usually flat or slightly con- 
cave. Sometimes the bird may kick so vigorously that twin grooves are cut 
into the floor. The entire process defines a tunnel that varies from almost 
round to elliptical in cross-section. 

Although he mentions it elsewhere, Stoner (1925) says nothing in the 
above account to substantiate the fable that the Bank Swallow carries 
loosened sand from its burrow in its bill. To anyone who has seen the sand 
jet from the opening of a burrow, the relative inefficiency of the by-the-billful 
method is obvious. Perhaps, as Petersen suggested (1955), the legend may 
be attributed to observations of the birds picking up bits of sand, a habit in 
which they, as well as other swallows, frequently indulge. Petersen also 
mentioned that the bird has a predilection to pick up something upon leay- 
ing the burrow. If a fecal sack is not available, a grain of sand or small 
pebble may be picked up. This habit probably explains Stoner’s earlier 
observations. 

In contradistinction to the situation with the Bank Swallow, there is no 
detailed description of a rough-wing using its bill to dig. Rather, all men- 
tion the use of the feet, in the same manner as those of the Bank Swallow, to 
remove loose sand from the burrow. Whether the loose sand is of natural 
origin or a product of scratching or pecking is not clear. The best that can 
be found concerning the use of the bill are the remarks of Bailey and Burns 
cited previously. This is further evidence that the skull of the rough-wing 
is appropriately considered with the skulls of non-burrowing swallows. 

Before leaving the subject of excavation, there is one more point to be 
considered. Hickling (1959) described a community digging display of the 
Sand Martin at the time of burrow-excavation. He described the birds 
circling and hovering in front of the cliff face in a “progressive series of 
‘movements’ or elements building up to a climax. ... The actual digging 
occurs only at this climax.” He describes the communal nature of the display 
as follows: 


108 Tue Universiry SciENCE BULLETIN 


“The communal nature of these displays is best seen if a well-defined group 
of holes is watched closely. At one display a group of birds—three or four com- 
monly, but quite often five or six—may be seen to be active at one hole while 
neighbouring holes are left unattended. During the next display quite different 
holes may be used for excavation. Up to three birds often enter one hole to dig, 
and there may be other birds sitting at the entrance ready to take the place of 
those already in the hole. Again, birds may be watched moving from hole to 
hole, excavating at several in succession. Yet fighting and aggressive posturings 
are sporadic only, and seem to me to be accidental squabbling arising out of the 
jostling in a crowd, rather than strictly territorial defense.” 


Such an approach to the problem of community housing is, to say the 
least, unusual in swallows. The North American Bank Swallows appear to 
be clearly territorial, and their behavior in this respect has been well docu- 
mented by Petersen (1955). He observed that “groups of several or a dozen 
or more birds move along the bank in hovering flight,” but interpreted this 
behavior as a part of territory formation. Hickling’s observations were of 
birds at a sand pit, and he remarked that “the behaviour described may not 
be typical of colonies of other kinds.” He noted that Petersen’s observations 
were made at river colonies. I have observed periodic or cyclic hovering 
similar to that described at both types of sites. I have not, however, observed 
any instance in which more than the members of a presumably mated pair 
were involved in excavating a given tunnel. Digging is performed also in 
the absence of any such “display.” I have also witnessed Cliff Swallows 
hovering before Bank Swallow burrows in the described manner, both in 
the presence and absence of the owners. Such “displays” as I have observed 
appear to be hovering by birds that are strongly attracted to the holes but for 
some reason cannot or will not enter them. Among the Bank Swallows, 
many hovering birds may be unemployed members of mated pairs awaiting 
their turns to dig. Tooby (1947) suggested that hovering might be “an 
advertisement flight of the male Sand Martin.” While the performance may 
serve some social function, I doubt that it has any direct sexual significance 
or that only males are involved. If the behavior described by Hickling is 
usual for English Sand Martins, then they would seem to be behaviorally 
distinct from their North American relatives. The English birds would seem 
to have evolved closer than the American members of the species to true 
community nesting. 


OSTEOLOGY 


Various osteological measurements were made from specimens of seven 
species: Tachycineta thalassina (20: 11 6 8,5 2 2,4 ?), Indoprocne bi- 
color (26: 8 8 6,17 9 9,1 ?), Riparia riparia (16: 7 $ 6,9 2 2), Stelem 


5 


dopteryx ruficollis serripennis (8: 3 66,4 22,1 ?), Hirundo rustica 


Fossor1AL ADAPTATIONS IN THE BANK SWALLOW 109 


wei2) 6 6,11 2 2,12); Petrocheldon pyrrhonota (33: 13 6 6,17 2 8, 
3 ?), and Progne subis (8: 3 6 6,5 2 2). Three additional specimens, 
which were not suitable for measuring, were used for studies of the process 
of ossification in each of the three following species: 1. bicolor, H. rustica, 
and P. pyrrhonota. Two specimens of R. riparia from New Mexico and two 
of S. r. psammochrous from Arizona were found to differ markedly in size 
from the specimens from Kansas. These, too, were used for studies of ossifi- 
cation, but were not included in the calculations. 

Except for Tachycineta, the majority of specimens used for calculations 
were from Kansas. The exceptions are as follows: Tachycineta, 1 from 
Washington, 19 from New Mexico; Riparia, 1 from Colorado; Hirundo, 2 
from Nebraska, 1 from Missouri, and 1 from New Mexico; Petrochelidon, 
1 from California; and Progne, 1 from Missouri, 1 from Arizona. The 
Arizona specimen of Progne was from Pima County and may represent 
P. s. hesperia. However, the measurements of this specimen cannot be dis- 
tinguished from those of P. s. subis, and it was treated as such. 


CHARACTERS STUDIED 


The following is a list of the characters studied and of the linear meas- 
urements used in the analysis with a brief definition of each (see Fig. 1 
for detail). 

Skull Length: from the tip of the bill to the most posterior point of the 
cranium (does not include the rhamphotheca). 


Fic. 1. Characters studied. Dorsal and lateral views of skull of Riparia, slightly diagrammatic. 
a-b, Skull Length; a-c, Cranial Length; c-b, Bill Length; d-e, Frontonasal Width; f-g, 
Maxillary Width; h-i, Interquadratic Width j-k, Cranial Width c-l, Bill Depth m-n, 
Cranial Depth. 


110 THe University ScrENcE BULLETIN 


_Force 


QS 
YooBey 


Fic. 2. Pattern of forces impinging on bill supports in Riparza when digging. Lateral force will 
create torque around midpoint of frontonasal hinge (T). Arrows indicate direction of 
net stress. 


Bill Length: from the tip of the bill to a transverse line across the most 
anterior portion of the rostral border of the frontals (does not include the 
rhamphotheca). 

Cranial Length: Skull Length less Bill Length. 

Frontonasal Width: transverse distance across the frontonasal hinge. 

Maxillary Width: greatest width between the external margins of the 
maxillaries. This measurement is approximately equal to the gape of the 
bird. 

Interquadratic Width: greatest width between the external margins of 
the quadratojugals. 

Cranial Width: greatest width of cranium not including the circum- 
orbital ridge. 

Bill Depth: from the top of the bill at the level of the frontonasal hinge 
to a transverse line across the bottom of the maxillaries. 

Cranial Depth: from the highest point on the cranium to a median sagit- 
tal line along the parasphenoid rostrum. 

Sternal Plate Length: in the midline from the anterior border of the 
dorsal lip of the coracoid sulcus to the posterior edge of the sternal plate. 


ie Se. 


TS. 


Fic. 3. Pattern of forces on bill of Riparia during slashing in an arc. Torque is created around 
Small arrows on lateral skulls indicate direction of stress and torque around midpoint of 
frontonasal hinge; large arrows direction of torque around long axis of bill. 


FossortaAL ADAPTATIONS IN THE BANK SWALLOW 111 


MeErTHop 


Before an analysis can be made, it is necessary to consider briefly the 
forces acting upon the skull of the Bank Swallow in the process of digging. 
Stress will result from the resistance of the substrate to the cutting edge of 
the bill in a direction opposite to that of the head’s movement. Since the 
Bank Swallow excavates by carving, primarily with a lateral or dorso-lateral 
motion, the forces impinging upon the bill will contain a large lateral com- 
ponent. Vertical and posterior components are also present (Figs. 2, 3). 
Stresses resulting from digging will be directly upon the bill and the points 
of its attachment to the cranium. Therefore, if any modifications to resist 
stress are present, they should be found in one or more of three critical sites: 
(1) the bill itself; (2) the frontonasal hinge; and (3) the quadratojugal 
articulation. 

Analysis of the skull must deal with proportions of one measurement 
relative to another, for it is obvious that linear measurements cannot be com- 
pared directly across generic lines. The use of proportions involves several 
risks. One is that both structures used for measurements may have been 
modified, which could obscure indications of those very modifications. 
Another is that the organism that is being used as a standard may itself be 
modified from the “basic morphological plan” of the group in one or more 
characters. Both these hazards may be avoided if one of the characters used 
as a basis for comparison is of a structure not likely to have been modified by 
selection for digging with the bill, and which, moreover, is likely to maintain 
proportional stability throughout the group. The discovery of such a charac- 
ter might well prove worthy of a study in its own right. However, one can 
choose certain characters that meet the first criterion and appear, @ priori, to 
satisfy the second to a reasonable degree. The length of the sternal plate 
appears to be such a character. It is divorced from the influences pertinent 
to the skull and is not so likely to be influenced by differences in flight tech- 
nique as is the depth or length of the keel. Another check is to look for con- 
sistency in the direction of the deviation. A character that consistently 
appears larger (or smaller) in comparison with a series of other characters, 
or in comparison with the same character in several species, is more likely to 
have been modified than one that deviates extensively but in only a few 
instances. Again, comparison with a series of species or genera is more likely 
to show a real deviation from “swallowness” than comparison with a single 
standard species or genus. Both these checks have been used in this study. 

Proportions were compared directly by setting the mean proportion of 
measurements of two characters (N.B.: not the proportion of the means of 
two measurements) of Riparia equal to 100 and dividing this into the mean 
proportions of the same characters obtained in the other genera. The results 
are expressed as a percentage of the proportion obtained for Rrparia. 


112 Tue University ScrENcE BULLETIN 


Two sources of possible error have been introduced into this analysis as a 
consequence of lumping the data. All specimens of a given species were con- 
sidered together, regardless of possible age or sex differences. The error in- 
troduced through differences in age is probably minimal. In all skulls used 
for measurements, the cranial sutures had completed fusion and there was 
extensive pneumatization of the cranial vault. Fusion of cranial sutures may 
be considered as the indication of termination of linear growth. Differences 
in the degree of pneumatization might well affect measurements of skull 
weight or volume, but have little significance in regard to linear measure- 
ments. In none of the species studied was there clear-cut sexual dimorphism 
so far as the features under consideration are concerned; in all instances there 
was extensive overlap in the range of measurements of both sexes. Females 
generally averaged somewhat smaller, but in most instances the difference 
was extremely slight and never approached statistical significance. Even in 
Stelgidopteryx, in which there is noticeable sexual dimorphism in a number 
of external measurements, the sample size at hand did not permit detection 
of any significant differences between the skulls of males and females. In- 
deed, in some measurements, the females averaged slightly larger. In view 
of this, it seemed permissible to lump all specimens in order to obtain larger 
samples. Considering the small size of some samples, the lumping may have, 
at worst, slightly increased the variance of the sample. 


ANALYSIS 


Bill—Table 1 shows the relationship of the Bill Length to four other 
measurements of the skull and to the Sternal Plate Length. In four of the 
five comparisons, the mean proportions of measurements for Riparia are con- 
sistently equal to or smaller than those of the other species. The constant 


Tasre 1. Percentage Comparisons of Bill Length.* 


Bill Length Bill Length Bill Length Bill Length Bill Length 


Cranial Sternal Plate Interquadratic Frontonasal Maxillary 

Length Length Width Width Width 
Tachycineta .......-.. 103.7 102.9 104.5 113.0 107.9 
Iridoprocne ........-. 105.9 98.7 102.5 112.1 105.5 
Stelgidopteryx _..... 106.5 102.6 105.2 100.0 96.8 
Er 36 0 121.8 ve7 113.0 124.2 98.6 
Petrochelidon __...... 103.5 102.4 101.2 98.6 94.1 
PROSne ee 140.1 ils 116.7 NAS 105.9 
Mean, 22 ees 113.6 107.6 107.2 110.1 101.5 
Repay Ge 100.0 100.0 100.0 100.0 100.0 


*In this table the figures presented represent a percentage of the mean proportion for Riparia. 
Except where noted, the measurement to be tested is used as the numerator, and percentages 
greater than 100 indicate that this measurement is smailer in Riparia. 


FossortAL ADAPTATIONS IN THE BANK SWALLOW 113 


Taste 2. Percentage Comparisons of Bill Depth.* 


Bill Depth Bill Depth Bill Depth Bill Depth Ball Depth 

Cranial Sternal Plate Frontonasal Bill Maxillary 
Depth Length Width Length Width 
Tachycineta _........ 106.2 107.0 120.1 102.0 117.6 
Iridoprocne _........ 108.0 100.6 116.2 98.6 111.6 
Stelgidopteryx _.... 98.2 93.0 OBR 89.0 64.8 
MOTURAO 220 nnn 110.3 103.2 1its}33 88.0 90.4 
Petrochelidon __. 104.5 102.5 103.1 100.7 97.2 
TG ae eas 115)3},11 127.2 120.1 103.8 113.6 
Meanie a5. 34 105.6 111.0 97.0 99:2 
Ri 100.0 100.0 100.0 100.0 100.0 


* For explanation of figures see Table 1. 


factor being the length of the bill, it may be concluded that, compared with 
the other species, the bill is relatively shorter in Riparia. The column in 
which Riparia is not consistently smaller is that comparing Bill Length with 
Maxillary Width, another measurement of the bill. Here the value obtained 
for Riparia is seen to be median in the series, with three of the other species 
being larger and three smaller. 

If comparisons are now taken by species, we see that the greatest devia- 
tion from Riparia lies in comparison with Hirundo and Progne, two species 
with relatively long bills. Note that the discrepancy of Riparia with Petro- 
chelidon \essens, and even changes direction in measurements involving 
width. The Cliff Swallow’s skull is noticeably short and wide; this can be 
detected by simple visual comparison of the skull with that of any other 
species. 

Table A-1 in the Appendix presents much the same information. In this 
table, the ratios, along with other data, are presented without conversion to 


Tasie 3. Percentage Comparisons of Maxillary Width.* 


Cranial Sternal Plate Frontonasal Bill Bill Interquadratic 
Width Length Width Length Depth Width 
Maxillary Maxillary Maxillary Maxillary Maxillary Maxillary 

Width Width Width Width Width Width 
Tachycineta .... 106.8 105.5 Maik 107.9 117.6 108.8 
Iridoprocne _.... 101.3 106.2 94.7 105.5 111.6 108.2 
Stelgidopteryx .. 89.7 93.7 96.6 96.8 64.8 89.2 
ATUNdO _......-.. 79.7 83.9 79.1 98.6 90.4 83.6 
Petrochelidon . 93.1 90.5 9352 94.1 97.2 103.5 
Byogne _...... 72h 85.6 93.2 105.9 113.6 93.0 
Beams oo 90.4 94-2 92.3 101.5 99.2 Wen 
WEpayioe 63s 100.0 100.0 100.0 100.0 100.0 100.0 


*Note in this table that the measurement to be tested is used as the denominator. The figures 
still represent a percentage of the mean proportion for Riparia, but percentages greater than 
100 indicate that Maxillary Width is larger in Riparia. 


114 Tue University ScrENcE BULLETIN 


percentages of the value for Riparia. A similar table will be found in the 
Appendix for each of the textual tables. 

Bill Depth at the level of the frontonasal hinge is considered in Table 2. 
Here it can be seen that the relative depth of the Bank Swallow’s bill is con- 
sistently less than five of six species in three of five columns. Stelgidopteryx 
presents an outstanding exception, being considerably smaller than Riparia 
in all comparisons. Again, in those columns that do not show a consistent 
deviation, there is a comparison of two measurements of the bill. The rela- 
tive flatness of the rough-wing’s bill is probably significant and will be dis- 
cussed later. 

Table 3, in which Maxillary Widths are compared, can serve as a sum- 
mary for all the tables concerned with characters of the bill. Maxillary Width 
has been used as the denominator rather than the numerator in this table, as 


Fic. 4. Dorsal and lateral views of heads of Riparia, Iridoprocne bicolor, Stelgidopteryx, and 
Hirundo. Note the progressive widening and lengthening of bill in this series. Plumage 
patterns only suggested. Ca. life size. 


Fossor1aAL ADAPTATIONS IN THE BANK SWALLOW 115 


well as in others. Thus, an increase in the value of the ratio signified either 
a decrease in Maxillary Width or an increase in the character used as the 
numerator. Therefore, a percentage of less than 100 indicates that Maxillary 
Width in that species is relatively greater than that of Rrparia. As the situa- 
tion in this table is the opposite of that in other tables, some care should be 
taken in the interpretation of these data. 

In terms of Skull Width, Sternal Plate Length, and Frontonasal Width, 
Maxillary Width in Riparia is seen to be relatively smaller than in most other 
swallows. The pattern here, and in the preceding two tables, is clear, When- 
ever a character of the bill is compared with a non-bill character, the value 
of the proportion for Riparia is shown to be small. This relationship does 
not hold between any two characters of the bill. In such comparisons, the 
value for Riparia is found to be median within the range established for 
swallows. Thus, while maintaining its intrinsic proportions, the Bank Swal- 
low’s bill as a unit is relatively the smallest of those studied. 

The tendency toward small bill-size in Rrparia is least conspicuous in the 
first two columns of Table 3. Here two species, the Violet-green Swallow 
and the Tree Swallow, are seen to have values smaller than that of the Bank 
Swallow. The relatively small values obtained for Tachycineta and Irido- 
procne may be the result of modifications of Maxillary Width and/or Skull 
Width and Sternal Plate Length in these species. It is also possible that this 
character is relatively less modified in Riparia than other characters of the 
bill. The bill of the Bank Swallow begins to widen at approximately the 
level of the frontonasal hinge. At the level of Interquadratic Width, the 
widening is sufficient to eliminate a clear pattern of exceptionally small size. 

Analysis of the data presented in the above tables conveys little indication 
as to the over-all shape of the bill. The Bank Swallow’s bill (Fig. 4) tends to 


be more nearly circular in cross-section, especially craniad, than the bills of 


Taste 4. Percentage Comparisons of Frontonasal Width.* 


Frontonasal Frontonasal Frontonasal Frontonasal Bill Bill 
Width Width Width Width Length Depth 
Sternal Plate Cranial Cranial Maxillary Frontonasal  Frontonasal 
Length Width Length Width Width Width 
Hachycneta ... 92.4 92.5 OTE? Val 113.0 120.1 
Iridoprocne ...... 88.7 94.0 93.8 94.7 ile 116.2 
Stelgidopteryx .. 102.8 107.8 105.5 96.6 100.0 93.2 
EMgrUndO -......--- 94.0 100.0 96.7 79.1 124.2 113.3 
Petrochelidon .. 102.5 100.6 101.8 93.2 98.6 103.1 
WROQNO: 2.2 0veceusee 108.8 117.2 123.4 93.2 MAS 120.1 
Meanweres-.--. I8e2 102.0 102.1 9233 110.1 111.0 
RU es 100.0 100.0 100.0 100.0 100.0 100.0 


*Note that in this table the measurement to be tested is used as the numerator in the first 
four columns, but as the denominator in the last two columns. 


116 THe University ScrENcE BULLETIN 


Tasie 5. Percentage Comparisons of Interquadratic Width. 


Interquadratic Interquadratic Interquadratic Interquadratic Interquadratic 


Width Width Width Width Width 

Cranial Sternal Plate Skull Cranial Maxillary 

Width Length Length Length Width 
Tachycineta .........- Or 102.6 100.0 100.0 108.8 
Inidoprocne ..... «103.4 101.1 102.0 105.0 108.2 
Stelgidopteryx .....- 98.1 92.6 93.6 96.4 89.2 
EIGHT Oe 101.9 99.2 92.9 100.2 83.6 
Petrochelidon ........ 107.6 113.6 109.4 111.0 103.5 
PRORnCHR Eee Wer) 108.4 105.9 ate 93.0 
Meant) 222. 2aan3 103.9 102.9 100.6 105.6 97.7 
Ripa ee 100.0 100.0 100.0 100.0 100.0 


Iridoprocne, Stelgidopteryx, and Hirundo. The difference in shape is least 
noticeable between Riparia and Iridoprocne, which, as noted, also has a rela- 
tively narrow gape. In both of these birds, the apparent rounding of the bill 
can be attributed to a relatively small flare in its proximal portion. 

Frontonasal Hinge.—T able 4 presents much the same kind of information 
as Table 3. When compared with Sternal Plate Length or with other cranial 
characters, Frontonasal Width shows no tendency to deviate in any con- 
sistent manner. When compared with measurements of the bill, however, 
Frontonasal Width is consistently large (note that in the last two columns 
Frontonasal Width is the denominator). Thus, while the Bank Swallow’s 
frontonasal hinge is no wider than would be expected in a swallow of that 
size, it is effectively larger in relation to the size of the bill than it is in 
other species. 

Quadrates.—The Interquadratic Width represents a character that forms 
a bridge between measurements of the bill and of the cranium. This meas- 
urement tends to be relatively small in Riparia (Table 5). However, the 
tendency is not as distinct either in magnitude or in consistency as it is in 
other measurements of the bill. The tendency may not even be as strong as 
it appears. Cranial Width and Cranial Length in Riparia are relatively 
large (Table 6). Thus, the values in columns 1 and 4 of Table 5 may not be 
truly representative of the Interquadratic Width. Smaller values for Cranial 
Width and Cranial Length would increase the value of the ratios for Riparia, 
thereby lowering all the values in columns 1 and 4. It is, therefore, probable 
that Interquadratic Width for Riparia is small, but not exceptionally so for 
a species of its dimensions. 


Miscellaneous characters——Table 6 presents comparisons of Cranial 
Width, Cranial Length, and Skull Length. The data indicate that the skull 
of Riparia, including Bill Length, is about as long as would be expected, but 


FossortaL ADAPTATIONS IN THE BANK SWALLOW 117 


that the cranium is wider than all but that of Petrochelidon, and tends to be 
somewhat elongate. 

Cranial Width was used as a basis of comparison in Tables 3, 4, and 5. 
Its effect upon the last of these has already been discussed. The effect of 
narrower Cranial Width in column 1 of Table 3 would be to give a low 
value for the ratio found in Riparia, thereby emphasizing the apparent large 
size of Tachycineta and Iridoprocne, but bringing all other values more in 
line with those of columns 2 and 3. No change in interpretation is necessary. 
In column 2 of Table 4, the value for Riparia would be higher, and all per- 
centages, consequently, lower. This would probably change the “sign” of 
the value for Petrochelidon and would make Hirundo somewhat smaller 
than, rather than exactly equal to, Riparia. Considering the small sample 
size for Riparia (13) in this comparison, and the fact that the standard error 
of the mean for Riparia (.032) is 16 times as great as that of Hirundo and 
eight times as great as that of Petrochelidon, I do not believe that any change 
in interpretation is necessary. The differences between these three species, 
in regard to this proportion, are extremely small (Table A-4). 

Weights.—I have presented no data concerning comparative weights of 
the skulls of the seven species. Such data as are available are crude and sub- 
ject to considerable error. They do indicate that, relatively, the skull of 
Riparia is among the more massive of swallows’ skulls. Increased mass 
would be useful in the absorption of either sudden shock or continuous pres- 
sure. It must be remembered, however, that the avian skull completes its 
ossification rather slowly, owing to the process of pneumatization. Skulls at 
the Museum of Natural History indicate that the process is not wholly com- 
pleted in swallows until midway through the second (first breeding) season. 
It is hardly probable that ossification of the sternum, which must withstand 
stress as soon as the birds begin to fly, is equally delayed. This introduces an 
age factor for which it is difficult to compensate. Moreover, slight damage to 
the skull, often unnoticed, can cause rather considerable change in the 


Tasie 6. Miscellaneous Percentage Comparisons. 


Cranial Width Skull Length Cranial Length 
Sternal Plate Length Sternal Plate Length Sternal Plate Length 
DOG CURCL eee 99.8 101.8 100.5 
WRMIODROCNE <....-2 5.02. A csancaceec.-assecess ee 94.6 96.1 94.0 
BCIOIMOPIET YN 22. oe eae encase 94.6 OO 97.0 
GUO: oe a 94.2 104.3 96.8 
Petrochelidon ..........-.... Bae ei OF. - a 102.3 100.8 100.1 
Mice ea ee 84.4 99.1 87.4 
POM ee 95.0 100.2 96.0 


ECURGAYGL «Se at 100.0 100.0 100.0 


118 THe UNtiversiry SciENCE BULLETIN 


weight. To have any real value, comparisons would have to be made among 
large series of skulls of equal degrees of pneumatization. 


MYOLOGY 


Myological studies were made of at least one specimen of eight species: 
Tachycineta thalassina (3 examined, only 1 suitable for dissection and weigh- 
ing), Iridoprocne albilinea (2 dissected, both weighed), I. bicolor (2 dis- 
sected, both weighed, but Skull Length available for only 1), Riparia riparia 
(12 dissected, 9 weighed), Stelgidopteryx ruficollis (9 dissected and weighed), 
Hirundo rustica (6 dissected and weighed), Petrochelidon pyrrhonota (11 
dissected and weighed), and Progne subis (8 dissected, 5 weighed). Greatest 
attention was given to five paired muscles: M. cucullaris, M. biventer cer- 
vicis, M. splenius capitis, M. rectus capitis lateralis, and M. rectus capitis 
ventralis. For reasons to be discussed below, the M. cucullaris and M. rectus 
capitis lateralis were treated as a single unit. These five pairs of muscles offer 
two advantages to the investigator. First, they are all large enough to be 
dissected free with reasonable assurance that the entire muscle has been 
obtained. Second, they are all involved in motions of the head that are criti- 
cal to any digging movements. 


MeTHopD 


Each pair of muscles was dissected free as cleanly and completely as pos- 
sible and stored in 70% alcohol in individual, labelled, two-dram vials. Upon 


fascia 
asci ae ye 


Fic. 5. Dorsal view of cervical musculature inserting on skull in Riparia. Skull tipped ventrad, 
muscles extended. Superficial muscles shown on left, dissected away on right. Biventer 
cervicis shown cut across anterior portion of second belly. Geniohyoidious removed. 
biv., biventer cervicis; cuc., cucullaris; dep., mandibular depressor, lat. rec., lateral rectus; 
spl., splenius cervicis. Ca. 3X. 


Fossor1aAL ADAPTATIONS IN THE BANK SWALLOW 119 


Fic. 6. Ventral view of cervical musculature inserting on skull in Riparia. Ventral rectus 
dissected always on right. car., carotid artery; |. v. rec., lateral portion of ventral rectus; 
m. v. rec., medial portion of ventral rectus. Other abbreviations as in Fig. 5. Ca. 3X. 


completion of all dissections, the muscles were transferred for 1 week to a 
1:1 mixture of methanol and ether at 7° C in order to reduce fat content. 
The mixture was then poured off and the muscles dried at 124° C for 24 
hours. Each pair was then weighed on a Mettler model H-16 scale, weigh- 
ing to 0.05 milligram and allowing a reasonably accurate estimate to 0.01 
milligram. 

The actions described for each muscle were determined from the mor- 
phology of the muscles and not from observations on living birds. It is 
entirely possible that the muscles may have slightly different actions under 
the influence of differential nerve impulses. Also, the action of a muscle 
may be modified in different ways when it acts in concert with various com- 
binations of other muscles. Figures 5, 6, and 7 of Riparia may be used for 
reference. 


dep. 
lat. rec. 


Fic. 7. Insertions of cervical musculature on skull in Riparia. Somewhat diagrammatic. Abbre- 
viations as in Figs. 5 and 6. Ca. 4X. 


120 THe University ScrENcE BULLETIN 


DescriPTION OF Muscies 


M. cucuLLaris, caput portion 


Origin: This muscle originates as fleshy fibers from the diapophyses of 
the fourth cervical vertebra. There is also extensive fascial involvement with 
other muscles in this area. 

Insertion: Attachment is fleshy or by a very short aponeurosis across the 
dorsal edge of the occipital. Insertion begins at or near the midline and 
extends directly laterad or dorsad, then laterad, to the insertion of the M. 
rectus capitis lateralis, approximately one-half to one-third the distance from 
the midline to the median edge of M. depressor mandibulae. There is ex- 
tensive fascial connection across the midline, especially in the most cephalad 
portion, between the two Mm. cucullarii. 

Structure: The belly is broad, flat, and thin. The fibers diverge slightly 
from the origin, then remain parallel throughout the greater portion of the 
belly. 

Action: Contraction of both members of this pair tilts the head upwards. 
Contraction of either unit singly also swings the head slightly to the side 
contracted. 

Variation: The insertion of the M. cucullaris shows some variation. The 
insertions of the Mm. splenii capiti and Mm. biventer cervici fit into depres- 
sions on either side of the midline. The median portion of the occiput, there- 
fore, appears as an inverted, tear-shaped bulge. In Progne and Hirundo, the 
insertions of the Mm. cucullarii extend across the bulge to meet, or nearly 
meet, at the midline. Before dissection, therefore, the insertion appears,con- 
tinuous and flush with the back of the skull. In Iridoprocne, the insertion 
appears to be continuous but lower on the occiput. In Riparia, and to a lesser 
extent in Stelgidopteryx, the two muscles insert beside the bulge and do not 
extend across it to the midline. 

Remarks: Avian cervical musculature historically has been plagued with 
a plethora of names. Fortunately, a considerable degree of uniformity is 
found in most recent publications. Two names are in common usage for the 
muscle just described. I have followed the lead of Fisher and Goodman 
(1955) and Fisher (1958, 1961, 1962) in the use of M. cucullaris. It should 
be noted, however, that M. complexus is equally acceptable (Zusi, 1962). 


M. RECTUS CAPITIS LATERALIS 


Origin: This muscle arises from the hypopophyses of the second, third, 
and fourth cervical vertebrae. The attachment is aponeurotic or by numer- 
ous, short tendons. 


Insertion: Attachment is fleshy or by a very short aponeurosis. The inser-_ 
tion begins at the lateral edge of the insertion of M. cucullaris, runs laterad_ 


| 


FossortaL ADAPTATIONS IN THE BANK SwaLLow 12) 


to the medial border of M. depressor mandibulae, then turns ventrad to run 
along or just under the margin of the depressor as far as the exoccipital 
process. The lateral border of the insertion lies just above the insertion of 
M. splenius capitis. 

Structure: The fibers are essentially parallel. The body of the muscle is 
broad and flat. The upper one-half to one-third of the muscle is closely 
associated with the M. cucullaris. At the point of insertion, the association 
is extremely intimate, and, in many specimens, the two muscles can be dis- 
tinguished only with great difficulty. 

Action: Contraction of both members of the pair tilts the head upward. 
Action of a single member will turn the head laterally and cause some 
dorso-ventral rotation of the head. 

Variation: No significant variation was noted. 

Remarks: According to Fisher and Goodman (1955:13 and Fig. 4, p. 16), 
the insertion of the lateral rectus in Gras americana is by a tendon to the 
exoccipital process and remote from the cucullaris. Burt (1930:515, Fig. 28) 
shows the lateral rectus extending well out onto the occiput of the Pileated 
Woodpecker. However, the arrangement of the other muscles that insert in 
the area is so different from the arrangement in passerines that the entire 
situation is questionably comparable. Boas (1929) describes the lateral rectus 
inserting lateral to the cucullaris on the exoccipital ridge in Larus, Tetrao, 
and Anser. His Plate 10 depicts a dorso-ventral orientation for the insertion 
with no lateral extension onto the occiput. Zusi (1962:76, Fig. 35; 86) figures 
and describes a similar insertion in Thalasseus maximus, Geohelidon nilo- 
tica, Larus atricilla, and Rynchops nigra. Schufeldt (1890:19, Fig. 4) shows 
the insertion of the rectus lateralis in Corvus corax meeting the edge of 
the cucullaris, but not extending onto the dorsal portion of the occiput. 
Palmgren (1949) describes the insertion in the seven species he studied 
(Acrocephalus schoenobaenus, Carduelis flammea, Certhia familiarts, Parus 
atricapillus, Phoenicurus phoencurus, Pyrrhula pyrrhula, and Regulus regu- 
lus) as running from the insertion of the biventer to the neighborhood of the 
temporo-maxillary joint. His Figs. 1 to 4 show an insertion like that in the 
Raven for Phoenicurus, Certhia, and Parus but well out on the dorsal occipi- 
tal ridge in Carduelis. It therefore appears that in some of the passerines, and 
perhaps the woodpeckers, the insertion of the lateral rectus has moved dor- 
sad and laterad, perhaps assuming some of the function of the cucullaris and 
certainly increasing its ability to rotate the head. 


M. BIVENTER CERVICIS 


Origin: This muscle originates as a tendon arising from the neighbor- 
hood of the neural crest of the 14th (second thoracic) vertebra. The actual 
attachment is to the tendon of the M. spinalis cervicis, 


122 Tue Universiry ScrENcE BULLETIN 


Insertion: Fleshy fibers attach to the occipital directly beneath the inser- 
tion of the M. cucullaris. 

Structure: This muscle is composed of two bellies connected by a tendon. 
The anterior belly extends from the insertion to the neighborhood of the 
fourth vertebra, approximately on a line with the posterior portion of the 
origin of the cucullaris. The posterior belly begins in the vicinity of the 
ninth vertebra and extends to the 13th vertebra. The fibers are parallel except 
for some pennateness in the regions of the tendons. 

Action: Contraction of this muscle tilts the head upwards and draws the 
anterior portion of the neck toward the thorax. The posterior half of the 
central tendon and all of the posterior belly are covered by a thick fascia that 
acts as a pulley. Thus, the pull of the biventer is along the curve of the neck 
rather than directly back toward the thorax. 

Variation: No significant variation was noted. 


M. sPLENIUS CAPITIS 


Origin: Fleshy fibers arise from the dorsal and lateral surfaces of the 
neural crest of the axis. 

Insertion: The insertion is broad and fleshy, extending from the occipital 
bulge laterad beneath the insertions of M. biventer cervicis and M. rectus 
capitis lateralis to beneath the exoccipital process at the basitemporal process. 

Structure: This muscle forms an almost perfect fan across the entire oc- 
cipital area. The fibers diverge from the origin, but are never bipinnate. 

Action: Again, combined action of both members of the pair results in a 
backward tilting of the head. It is probable that this is the lesser function. 
Action of a single member of the pair swings the head to the side. 

Variation: No significant variation was noted. 


M. RECTUS CAPITIS VENTRALIS 


This muscle is conveniently divided into lateral and medial portions and 
should, perhaps, be considered as a complex. The dividing line approximates 
the line of the right and left dorsal carotid arteries except at the insertion, 
where the median portion spreads laterad. The relation between the two 
portions is only slightly more intimate than that between the dorsal portions 
of the cucullaris and lateral rectus. 

Origin: The origin of the median belly is partially fleshy, partially ten- 
dinous. It arises in three heads from the hypopophyses of the axis, atlas, and 
third vertebra. The lateral belly arises by a short tendon from the hypopo- 
physis of the fourth and sometimes fifth vertebra. The origins of all but the 
most anterior head of the medial portion are closely associated with the 
origin of M. rectus capitis lateralis. 


FossortaL ADAPTATIONS IN THE BANK SWALLOW 123 


Insertion: The median portion inserts fleshily over most of the area of 
the basitemporal plate. The anterior portion of the insertion forms an arc 
along the anterior and lateral borders of the basitemporal plate. The lateral 
portion inserts fleshily at the extreme latero-anterior corner of the arch 
formed by the median portion. 

Structure: The fibers are rather loosely-associated, and the muscle tends 
to crumble readily. The fibers of the medial portion diverge from the points 
of origin, but are not bipinnate. The fibers of the lateral portion are essen- 
tially parallel except at the point of origin, where they converge to form the 
tendon. 

Action: The medial portion of this muscle serves to tilt the head down- 
ward. The lateral portion serves this function also, but, in addition, can 
swing the head to the side if contracted singly. 

Variation: Fisher and Goodman (1955:14 and Fig. 7, p. 22) describe the 
rectus ventralis as an asymmetrical muscle, with the left side extending 
farther caudad than the right. They attribute the asymmetry to the change 
in position of the trachea and pharynx. The situation in the swallows is 
somewhat ambiguous. There is no apparent difference amongst the species 
studied in the medial portion. The origins of the lateral portion seem to vary 
at the level of the individual. For those specimens inspected for asymmetry, 
the results are: 


I. albilinea: \eft equal in both; 

I. bicolor: left longer in one; 

Stelgidopteryx: left longer in four, equal in one; 
Riparia: \eft longer in three, equal in four; 
Hirundo: \eft longer in one, equal in two; 
Petrochelidon: \eft longer in two, equal in five; and 
Progne: left equal in three. 


In many cases in which the left was equal to the right, it appeared more 
robust. In most cases in which the left was longer, it originated from the 
fifth vertebra. 

The area of insertion of the medial portion appears to be smaller in J. 
bicolor. In both specimens, it was confined to the area between the branches 
of the dorsal carotid arteries. 


ANALYSIS 


In analysing the relative size of muscles, it is necessary to find a measure- 
ment that will reflect expansion into three dimensions. Volume is, per- 
haps, the best measurement, but determining the volume of very small, ir- 
regularly shaped, absorbent structures presents rather complex technical 
problems, especially when a large portion of the volume to be measured con- 


124 Tue UNIversiry ScIENCE BULLETIN 


sists of diffusible fluids. Weight appears to be the next best choice of meas- 
urement to document relative size. The major portion of the weight of a 
muscle consists of fluid, much of which is extracellular. Since the amount of 
fluid is quite variable, even from time to time in one muscle, it is necessary 
to eliminate this variable by thorough drying before weighing. Because 
weight varies as the cube of linear measurements, the cube root is used in 
comparisons of the weights with a linear measurement. 

Note also that the mass (=weight) of a muscle represents only a crude 
approximation of its strength. Mass can be used as an index of strength only 
if there is little difference in the size and shape of the muscles compared. It 
has already been shown that there is little difference in the shapes of muscles 
in the species studied. The differences in mass also are small. The mean of 
the mass of any muscle in Progne, the largest species, is approximately five 
times the mean for the smallest species. In no other swallow is the mean 
mass for any muscle as much as twice as large as the mean for the smallest 
species. 

Skull Length was chosen as the non-muscular measurement for use in 
comparisons. This character has been shown (Table 7) not to deviate pro- 
portionally in Riparia. It reflects, to some extent, the mass to be moved by 
the muscles studied and is easily measured on a partially dissected specimen. 
Unfortunately, when taking the measurement from alcoholic specimens, the 
rhamphotheca must be included. I have only one Riparia and no Stelgidop- 
teryx skulls from Kansas with the rhamphotheca intact. Proportions based 
on the one Riparia skull indicate no change in direction from the proportions 
of the skulls without rhamphothecae. ; 

Table 7 presents the myological data in the same manner in which the 
osteological data previously were presented. The data for Tachycineta and 
I. bicolor in Table 7 are from one specimen each. The Violet-green Swallow 
was mature. The Tree Swallow was probably a first-year bird. The data 


Tasie 7. Comparisons of Cube Roots of Muscle Weights With Skull Lengths. 


Ventral Rectus CLR-Complex Splenius Biventer Total Mass 

Skull Length Skull Length Skull Length Skull Length Skull Length 
Tachycineta .......... 98.2 94.6 97.0 98.1 97.0 
I albilined 101.8 97.3 100.0 103.7 100.0 
[ehicolo7 94.5 90.5 93.9 103.7 95.0 
Stelgidopteryx ...... 945) Sik 100.0 98.1 96.0 
Elinundo ee 945 94.6 100.0 100.0 98.0 
Petrochelidon ........ 107.3 104.0 103.0 103.7 105.1 
PEO Rn Ciena 112.7 108.1 107.6 109.2 109.8 
Meant eee 100.5 97.3 100.2 102.4 100.1 


Riparigg 100.0 100.0 100.0 100.0 100.0 


Fossor1AL ADAPTATIONS IN THE BANK SWALLOW 125 


Taste 8. Summary of Critical Proportions. 


Bill Bill Bull Bull Frontonasal 

Length Length Length Depth Width 

Cranial Interquadratic | Frontonasal Cranial Maxillary 

Length Width Width Depth Width 
ihachycineta .......... 103.7 104.5 113.0 106.2 O7 a 
Iridoprocne .......... 105.9 102.5 112.1 108.0 94.7 
Stelgidopteryx _.... 106.5 105.2 100.0 98.2 96.6 
FE BOO eos. - 121.8 113.0 124.2 110.3 79.1 
Petrochelidon _...... 103.5 101.2 98.6 104.5 93.2 
POG a 140.1 116.7 AS 5335 93m 
i MICE eae ee 113.6 107.2 110.1 WBE 925 
IREVSTO RY re 100.0 100.0 100.0 100.0 100.0 


pertaining to these two birds differ markedly from those for the other swal- 
lows only in relation to M. splenius capitis, for which they appear somewhat 
small. 

One of the intriguing aspects of the myological data, as seen in Table 7, 
is the absence of any clear pattern beyond the fact that both Petrochelidon 
and Progne are more heavily muscled than Riparia. There seems to be con- 
siderable variation in the proportions of the muscles relative both to Skull 
Length and to each other. This is especially true for Stelgidopteryx, Hirun- 
do, Petrochelidon, and Progne, for which the samples are largest. 

Although there is no over-all pattern, some pertinent facts are revealed by 
examination of individual columns. Perhaps most significant are the com- 
parisons involving the total muscle mass. It appears that Riparia possesses 
cranial musculature that, although it approaches the limits of range of the 
muscle to skull length ratio for swallows, is not exceptionally large. Assum- 
ing that the first three values in this column do represent the species to a 
reasonable degree, then the total muscle mass for Rzparia is relatively larger 
than four, smaller than two, and about equal to one of the other species. The 
data for the ventral rectus show about the same situation. The biventer 
appears to be somewhat small. There is apparently very little difference be- 
tween the relative size of the splenius capitis in Riparia and the other swal- 
lows. The cucullaris-lateral rectus complex appears to be relatively large in 
Riparia, for it is comparatively larger than those of five of the seven other 
species, and in comparison with Progne it is the second largest muscle. 


DISCUSSION 


OsTEOLOGY 
The most apparent fossorial modifications of the Bank Swallow are 
found in the structure of its bill. This structure, as a unit, is clearly smaller 
than the bills of other swallows relative to the size of the cranium and Ster- 


126 THe UNIversity ScIENCE BULLETIN 


nal Plate Length. Those proportions that most clearly show this relationship 
are summarized in Table 8. Also, with the rhamphotheca in place, the Bank 
Swallow’s bill is more conical than those of most other swallows (Fig. 4). 
As in other swallows, the bill of Riparia tapers to a slightly hooked point at 
the tip, but flares less proximally. Thus, in the intact bird, it appears to be 
more centrally placed relative to the shape of the head than in other swallows 
and resembles a short awl or nail projecting from the skull. 

The shortening and reshaping of the bill confer a number of advantages. 
In the order of their probable importance, these are: 

1. Reduction of tension and compression stress upon the supports of the 

bill, especially the frontonasal hinge; 

. Reduction of shearing stress across the bill; 

. Creation of a structure capable either of probing or of scratching in 
several directions with almost equal facility; and 

4. A slight increase in mechanical advantage. 

In the kinetic avian skull, the frontonasal hinge, as a flexible joint of thin 
bone, must be considered a primary weak spot. This is especially true in 
those cases in which strong forces must be transmitted through the joint 
from the bill to the cranium, or vice versa. In most avian activities, pressure 
through the hinge into the skull is directed either posteriorly from the tip of 
the bill (pecking) or dorsally from the tomial edges (biting). The digging 
motions of the Bank Swallow exert lateral, dorso-lateral, and postero-lateral 
pressures on the frontonasal hinge (Figs. 2, 3). For all practical purposes, 
the applied force may be considered the same as those applied to the tomium 
in biting, that is, at an angle to the axis of the bill—with three critical excep- 
tions. First, the lateral orientation of the applied force exerts a tearing or 
shearing action along the axis of the hinge. Second, both the bill and hinge 
are relatively inflexible in a lateral direction. Third, often in the process of 
digging, the bill is subjected to forces tending to rotate it around its longi- 
tudinal axis as well as around the hinge or any point in the hinge. 

In digging, the bill forms a portion of a lever system. The “Force” of 
Fig. 2 is actually the resistance of the substrate to the motion of the bird’s 
head. For the sake of simplicity, the force is considered to be applied at the 
tip of the bill and at a right angle to the axis of the bill. Torque at any point 
along the bill is equal to the product of the applied force times the distance 
along the axis from the point of application to the point in question. It is 
important to note that, in digging, torque is applied not in the plane of rota- 
tion of the frontonasal hinge, but around a point in the center of the hinge 
(technically, around any point in the hinge) in a plane in which it is inflexi- 
ble. Thus, the shorter the bill, the less torque around this point. 

Increased relative width of the hinge enables it to withstand increased 
torque around its center. The hinge can be considered the base of an equi- 


Os bdo 


Fossor1aAL ADAPTATIONS IN THE BANK SWALLOW 27, 


lateral triangle with its apex at the tip of the bill (Fig. 2). If the triangle is 
to remain stable, each half of the base must be able to compensate for at least 
half of the impinging force. The situation may be expressed mathematically 
by the formula: 
GCE Oe NW). CnC 2 Ni) 

where F is the applied force, T the torque, f and f’ the resistance to compres- 
sion or tension, BL the Bill Length, and FNW the Frontonasal Width. 
Given no differences in structural strength, a relatively wide base will with- 
stand more torque than a relatively narrow one. Table 4 shows that the 
width of the Bank Swallow’s frontonasal hinge falls into the middle of the 
hirundinine range when compared with other non-bill characters. Thus, it is 
about as wide as would be predicted for a swallow the size of Riparia, or 
perhaps slightly larger, as the cranium is slightly larger than might be ex- 
pected (Table 6). The hinge becomes functionally wider than in other 
swallows by virtue of the fact that the bill is shorter, and the height of the 
triangle is, therefore, relatively small. 

The situation in regard to the shearing strain, that is, strain along the axis 
of the hinge, is somewhat different. In this case, the applied force is direct 
and independent of the length of the bill. Therefore, it is not subject to in- 
crease by lever action. Resistance to shearing can come only from an increase 
in the actual, not relative, area of supporting bone in the frontonasal hinge. 
There is little evidence for such an increase in my data, unless one considers 
that the slight enlargement of the cranium as indicated in Table 6 is sufficient 
to warrant considering the hinge a little wider than expected. It is perhaps 
best to adopt a conservative attitude and consider such an increase insignifi- 
cant. On the other hand, if the Frontonasal Width were found to be reduced 
in accordance with measurements of the bill, resistance to shear would be 
lost in proportion to the lost area. Such a loss might well be critical in view 
of the fact that bone is adapted to withstand the stresses of compression and 
tension, but not shear. 

By slashing primarily in a lateral direction, the Bank Swallow also re- 
duces the possibility of having the bill forced open. The frontonasal hinge is 
an avian adaptation that permits the upper mandible to be raised. The 
ramphotheca of the upper mandible of Riparia, as in all swallows, slightly 
overlaps that of the lower mandible and extends slightly beyond it at the tip. 
Therefore, if slashing were oriented mainly in a vertical direction, the bill 
might occasionally be forced open, and the mouth filled with sand. The rela- 
tively great width of the hirundinine frontonasal hinge resists lateral bend- 
ing. This resistance, combined with the overlap of the lower rhamphotheca 
by the upper, insures that the bill will not be forced open in lateral, or even 
dorso-lateral, slashing. 


128 Tue Universiry SciENcE BULLETIN 


The situation concerning Interquadratic Width is similar, but not quite 
the same. Table 5 shows Interquadratic Width in Riparia is small, but not 
the smallest among the swallows. .The narrow width reflects the small bill, 
but is more in line with characters of the cranium. It, too, can be considered 
the base of a triangle. This is especially important because the quadrates 
form the sole arthrological junction between the cranium and the lower 
mandible. As shown by Bowman (1961), the wider the Interquadratic 
Width, the greater its ability to withstand lateral pressure. 

The problem of fractural strain across the bill is similar to that of lateral 
pressure on the frontonasal hinge. In digging, force is applied to the bill at 
an angle to its long axis, thereby inducing a risk of fracture. Such risk con- 
ceivably could be reduced by a broad bill with flat edges, forming a double 
wedge, much like the head of a double-bitted axe or the blade of a broad 
sword. As will be shown, considerable disadvantage would accrue from such 
a structure. An alternate structure would be one with a short distance be- 
tween the point at which pressure is applied to the point at which it can be 
absorbed or dispersed. Such a structure also would reduce risk of fracture 
owing to excessive torque at any point along the bill. 

In shape, the Bank Swallow’s bill does not differ greatly from the distal 
portion of the bills of several species of swallows. In many respects, it re- 


Fic. 8. Entrances to burrows of Riparia, south bank Kansas River, Douglas County, Kansas, 
June, 1962. Note, in roof of cavity, scalloped-pattern resulting from slashing action of 


bill. 


FossortaL ADAPTATIONS IN THE BANK SWALLOW 129 


sembles the usual hirundinine bill from which the broad, proximal portion 
has been removed. It is narrowness of the proximal portion that confers 
upon the bill its rounded cross-sectional appearance. 

The rounded bill permits the Bank Swallow to slash into the substrate in 
any direction. While the primary motion may be lateral, the birds in fact 
slash in a variety of directions, including, albeit rarely, directly vertical. Such 
action is amply evidenced by slash marks left in the substrate (Figs. 8, 9). 
These slash marks also indicate that the head is frequently moved through 
vertical as well as horizontal arcs. Thus, pressures may impinge upon the 
tip of the bill from several directions in the course of a single stroke. During 
some strokes, the bill could be subjected to severe torque around its axis as 
well as around the center of the frontonasal hinge. Were the bill shaped as a 
double wedge, the torque around the axis would be increased by added re- 
sistance to the substrate, and the thin margins would be liable to fracture 
and/or tearing. Such is not true for a rounded edge in which streamlining is 
topologically uniform. Moreover, torque around the axis is a product of the 
applied force times the radius of the bill. In slashing, pressure is applied not 
only to the tip of the bill, but along varying portions of its length. It is, 
therefore, of considerable advantage to the Bank Swallow to possess a bill 
that remains relatively narrow for most of its length. This advantage is lost 


Fic. 9. Fresh marks made by slashing action of bill of Riparia near old burrow. Note that some 
of these form arcs. South bank of Kansas River, Douglas County, Kansas, June, 1962. 


130 THe University ScrENCE BULLETIN 


at the level of the cranium, thereby permitting a broadening of Maxillary 
and Interquadratic Width. 

The shortening of the bill also increases its mechanical advantage to some 
extent. The skull can be considered as a lever with the fulcrum at the oc- 
cipital condyle, the moment arm being posterior and the resistance arm an- 
terior to the condyle. As such, the mechanical advantage (M.A.=length 
moment arm/length resistance arm) is less than one. The lever is adapted 
for increasing speed rather than force. Any decrease in length of the re- 
sistance arm relative to the moment arm will increase the mechanical ad- 
vantage. In the case at hand, the increase is exceedingly small—possibly too 
small to be of selective advantage. Selection for a short bill is to be attributed 
more to a reduction of stress than to a gain in leverage. 

An idealized perfect tool for excavating with the techniques of the Bank 
Swallow would be a short awl or cone suspended by a relatively large, sturdy 
base. The Bank Swallow’s bill is by no means this perfect. It is hooked 
rather than pointed at the tip, although not so much as in some other swal- 
lows, and is not perfectly round in cross-section. However, it approaches the 
ideal more closely than the bill of any other swallow, especially in its small 
dimensions. There is nothing in the Bank Swallow’s ecology that suggests 
any advantage for a relatively small bill other than those accruing during 
excavation. As has been shown, these advantages are several. Thus, the 
excavatory techniques of the Bank Swallow have engendered selective pres- 
sures for the evolution of a relatively small bill. Such could be attained either 
by a direct reduction of the bill or by an increase in the size of the cranium. 
The figures in Table 6 indicate that the cranium of Riparia is large relative 
to the Sternal Plate Length. This, however, does not fully account for the 
differences between the bill of Riparia and those of other swallows. It ap- 
pears that both processes have taken place. Thus, the Bank Swallow’s bill 
has probably been reduced in size from the ancestral condition. 

Although reduced, the bill maintains its intrinsic proportions. This is, of 
course, necessary if it is to remain functional as part of a feeding apparatus. 
It does not seem necessary to postulate a series of mutations that individually 
and independently reduced length, width, and depth. It is more likely that 
the bill has been reduced as a unit through the action of one or more muta- 
tions affecting the bill “field” (Davis, 1949; Rensch, 1960). 

In reduction of the bill we find a single modification that confers several 
adaptive advantages. This happy solution cannot be fortuitous. It is a basic 
tenet of evolutionary theory that mutations are random and that there can 
be no a prior: direction of evolutionary change. The genesis of the raw ma- 
terial of any adaptive solution to a problem is strictly accidental. The prob- 
ability of an organism adopting any given solution is a function of the 
probability of given mutations arising and being selected for in the popula- 


Fossor1AL ADAPTATIONS IN THE BANK SWALLOW nS 


tion prior to the occurrence of genetic changes (mutations and/or recombi- 
nations) leading to any other solution. Thus, there appear in nature adap- 
tive patterns that seem unduly complicated and/or inefficient from the stand- 
point of what we judge might have been developed from the structure of 
the supposed ancestral organism. In the Bank Swallow, however, any modi- 
fication must have been one that fitted within the rigid specifications de- 
limited by the bird’s major specialization, aerial feeding. It is theoretically 
possible that the genetic constitution of Riparia could provide mutations 
leading toward development of a massive skull and chisel-like bill, such as is 
found in woodpeckers. It seems likely, however, that such development 
would be disadvantageous for aerial foraging. Maintenance of delicately- 
balanced genetic co-adaptations is such that the probability of the organism 
acquiring a given adaptation is increased. It is important to realize that, in 
the case of a highly specialized organism, modifications of specialized struc- 
tures for the advantage of exploiting a new niche must be efficient at the 
outset. The greater the initial degree of specialization, the less opportunity 
the organism will have to evolve modifications that, initially, are relatively 
inefficient, and the greater the probability that a successful modification will 
be one that simultaneously confers several advantages. If the restrictions set 
by the specialization are sufficiently rigid, then any exploitation of a new 
niche will require a relatively efhcient modification. The one alternative 
possibility is that the organism can exploit the new niche with no modifica- 
tion of its structure; that is, it is morphologically preadapted to the niche. 
The situation can be viewed diagrammatically (Fig. 10). Assume (1) that 
it is advantageous for an organism to exploit a niche (burrowing nest cavi- 
ties) but that its structure permits little or no use of that niche; (2) that the 
organism can become adapted to fuller exploitation of the niche through any 
of several possible modifications of its structure (A-E); (3) that any in- 
creased exploitation of the niche requires an increased output of energy. In 
such a situation, the advantage gained by use of the niche is balanced by the 
increased output of energy. Selection will favor continued modification 
along any line of adaptation that increases efficiency, thereby reducing the 
energy required. Full exploitation (self-excavation of an entire burrow) is 
justified only when the line of adaptation reaches a sufficient degree of efh- 
ciency. It is plain that some lines of modification will require less change to 
reach this degree of efficiency than others. In other words, some modifica- 
tions confer a greater degree of efficiency per unit change than others. Thus, 
modification along line “A” would require twice as much change as along 
line “B” to reach the same level of efficiency. If there are no restrictions on 
the amount of change, then any number of solutions to the same problem 
can be reached from the same basic structure (Simpson, 1949). If, however, 
the original structure of the organism is already specialized for the exploita- 


132 Tue University SciENcE BULLETIN 


tion of another niche (aerial feeding), there will be resistance to any change 
and a limit to the amount of change in any line. In Figure 10, this limitation 
has been set, arbitrarily, at three units of change. Only modification along 
line “E” will confer sufficient increase in efhciency per unit change to permit 
full exploitation of the niche. Thus, it is to be expected that any burrowing 
swallow will show modification along a relatively efficient line of adaptation. 
The conferring of several advantages in one change is one way that the 
efficiency of an adaptation is increased relative to other possible changes. 

In reality, the situation is not as simple as presented. The extant char- 
acters of an organism may already permit varying degrees of exploitation 
along the various lines of modification (preadaptation). Also, resistance to 
change will vary in the different lines. The effects of these complications are 


i FUE EXPCOMATION 
Al4 Be 


Original 
I Structure 


Fic. 10. Model of comparative efhciencies of various lines of evolution and effects of preadapta- 
tion and restriction by specialization. Line I represents original condition. Line II repre- 
sents level at which efficiency of modification is sufficient to justify full exploitation. 
Numbers at Line II indicate units of change necessary for full exploitation. Lower border 
of hatched area indicates degree of preadaptation; upper border indicates tolerance for 
change. 


FossortaAL ADAPTATIONS IN THE BANK SWALLOW 133 


exemplified by the hatched area in Figure 10. Thus, line “A” shows a rela- 
tively great amount of preadaptation, and line “C,” though less efficient than 
modifications along lines “D” or “E,” shows a relatively greater tolerance 
for change than these latter lines. 

The idea of “unit of change” is an abstraction involving the nature and 
number of structures modified, the nature of the modifications, the number 
of pertinent mutations, etc. Obviously, this concept is not subject to rigorous 
definition. I judge, however, that this does not reduce its validity for pur- 
poses of theoretical discussion. 

The structure and excavatory techniques of the Bank Swallow have at- 
tained a level of efficiency that justifies full exploitation of a cavity-burrowing 
and the absence of a 
complete fossil record make it difficult to draw a line that precisely represents 
Riparia in Figure 10. The data presented indicate that, as expected, line “E” 
fairly represents Riparia insofar as modification of the bill is concerned. The 
cranium of the Bank Swallow is slightly larger, the bill slightly smaller, than 
one would expect to find in a swallow of its size. In only three instances, 
Beak Length/Cranial Length, Beak Length/Interquadratic Width, and 
Frontonasal Width/Maxillary Width (Table 8) do any of the mean propor- 
tions of the Bank Swallow’s bill fall completely beyond the range found in 


’ 


niche. The imprecision of the term “unit of change’ 


other swallows. There are no rearrangements of structures and no addi- 
tional mass or supporting structures. Yet, the slight change in mean propor- 
tions of the bill relative to the cranium confer several distinct advantages. 

It is probable that the osteology of the skull in all swallows and, therefore, 
the ancestral form of Riparia, is preadapted to a slashing mode of excavation. 
The hirundinine bill is short, flat, and broad. It has been mentioned that a 
short, flat, broad blade could be used as a slashing tool, but that such a tool 
is limited to slashing in one plane. The hirundinine bill would be most 
effectively used in a lateral plane, which is the plane of the majority of 
strokes made by the Bank Swallow. Also, the frontonasal hinge of all swal- 
lows is relatively wide. Seemingly, any swallow would have the basic physi- 
cal equipment necessary for excavation of the type performed by Rzparia. 
The reduction of the Bank Swallow’s bill is a postadaptation increasing the 
versatility of the structure and decreasing the risk of fracture and the effect 
of torque on the frontonasal hinge. 

The term “preadaptation” has had a varied and sometimes confusing 
usage during its brief history. Throughout this paper, it is used in reference 
to structures that have been evolved to fulfill one function but are capable of 
performing another should the need arise (Bock, 1959). This use of the 
term is distinct from the mutational approach criticized by Simpson (1953a: 
134-135) and closer to his later use of the term (747d :191-193). It seems clear 
that the structure of the skulls of most swallows would permit some exploita- 


134 Tue University SciENCE BULLETIN 


tion of a burrowing niche. However, full exploitation is evidently impossible, 
or at least impractical, without some modification (postadaptation). 


Myo .ocy 


The lack of distinct pattern of modification in the muscles studied has 
already been mentioned. However, Table 7 presents data indicating that the 
cucullaris-lateral rectus complex (hereafter referred to as the CLR-complex) 
is relatively large in Riparia. A large CLR-complex is well-suited to the 
needs of the Bank Swallow. As mentioned above, the two members of the 
complex can turn the head up and toward the side. While carving the arched 
roof of the tunnel and dome of the nest chamber, the birds must perform a 
considerable amount of slashing at a level above the plane in which the head 
is normally held. Strong musculature for tilting the head dorsally and 
laterally is, therefore, of advantage. It is interesting to note in this regard 
that the tunnel of the burrow frequently slopes slightly upward from the 
mouth (Stoner, 1936). The slope aids in drainage of the nesting cavity. By 
balancing a tendency toward strong dorsal slashes with an attempt to keep 
the body horizontal, or with strong scratching by the feet, the bird would 
achieve a slanted rather than a level or upward curving tunnel. Therefore, 
it is possible, even probable, that there has been selection for greater emphasis 
on dorsal strokes, a selection that would favor increase of the CLR-complex. 

Besides drawing the head dorsally and laterally, contraction of the CLR- 
complex on one side only will cause the head to rotate on its axis. Rotary 
movement may be augmented in the swallows by medial extension of the 
insertion of the lateral rectus. Thus, there is in all swallows a well-developed 
ability to move the tip of the bill through both lateral and vertical arcs. To 


this ability—a function of “swallowness”—can be attributed both the shape | 


of the tunnel and the arc-shaped grooves mentioned above. 


The small size of the biventer may be related, in part, to the increase of — 


the CLR-complex. Simultaneous contraction of both pairs of the CLR- 
complex effects the same movement as contraction of the biventers. It is 
reasonable on grounds of conservation of energy that, given no increased 
necessity to tilt the head straight up, a genetic increase in the mass of one 
set of effectors will permit a proportional decrease of the other set. 

The seemingly unmodified condition of the ventral rectus and of the 
splenius capitis (Table 7) is rather startling. On an a priort basis, one would 
expect these muscles, and especially the latter, to be enlarged. 

The figures in Table 7 indicate that the ventral rectus in Riparia is among 
the largest in the swallows, relative to the length of the skull. In fact, pro- 
portions for this muscle closely agree with proportions obtained for total 
muscle mass. Although Riparia does little directly vertical slashing, contrac- 


Fossor1aL ADAPTATIONS IN THE BANK SWALLOW 185 


tion of one side of the ventral rectus probably aids the CLR-complex in 
swinging the head through a dorso-ventral arc. The splenius capitis is al- 
most exactly the size one would expect relative to length of skull. In view of 
the lateral digging motion of Riparia, the lack of a distinctly large splenius 
capitis, accompanied by no noticeable change in origin, insertion, or struc- 
ture, is puzzling. This observation, combined with the fact that the Bank 
Swallow’s musculature is generally not large relative to the size of the skull, 
indicates that there is no selective advantage to an increase of the effects of 
the muscles beyond what is already present in swallows as a group. Again, 
the Bank Swallow is seen to be preadapted, by virtue of adaptations found 
throughout the hirundinids, to its mode of excavation. 

The motor patterns of foraging of swallows have not, to my knowledge, 
been studied in detail. Certain facts, however, are evident, and others can 
be inferred. It is obvious that swallows cannot be filter feeders that strain 
insects from a continuous airstream. Nor is it at all probable that swallows 
trap insects in pockets formed by the wing and tail in the manner of some 
bats (McCue, 1961). Although I have not observed them do so, it is possible 
that some feeding is done by a form of aerial skimming in which the bill is 
gaped until one or more insects has struck it and/or become lodged in the 
buccal cavity. This manner of feeding would be most effective when the 
birds are preying upon small, swarming insects such as gnats or mosquitoes. 
Such food is not always available, and the birds must, most of the time, prey 
upon insects that have to be captured individually. These insects are plucked 
from the air with a snapping motion. If it were necessary for the birds to 
center each insect directly in front of the bill before snapping, their flight 
would be slower and more capricious than it is. The birds can, and do, snap 
at angles to the line of flight. I have seen foraging Bank Swallows “strike” 
laterally with great vigor. Significantly, the Bank Swallow, with small bill, 
glides less and follows a more irregular course than other swallows (Blake, 
1948). 

It has already been mentioned that the head of a bird can be considered 
as a lever with a mechanical advantage less than one, specifically, a Class I 
lever. This type of lever is useful whenever it is desirous to have the re- 
sistance arm move rapidly through an arc, and circumstances prevent use of 
a Class HI lever (Brown and Schwachtgen, 1949). Obviously, for any given 
lever, the greater the force applied to the moment arm, the faster will the arc 
be traversed by the resistance arm. It is evident that there is selective ad- 
vantage to a swallow in being able to swing its head rapidly. 

Another factor that may lend selective advantage to heavy neck muscula- 
ture is collision shock at the moment of capture. With small insects, this 
shock is probably insignificant, but with larger insects, such as bees, it may 
be considerable, especially if the capture is head-on rather than overtaking. If 


136 Tue Universiry ScrENcE BULLETIN 


the insect is in a direct line with the body of the bird, the shock can be passed 
through the skull and neck to be absorbed by the body, or can be mitigated 
by recoiling of the neck. If the insect is not directly in line with the body, 
increased musculature may be necessary either to absorb the shock directly 
or to permit the bird an elastic follow-through of the head and neck. 

The slight magnitude of the differences between the Bank Swallow’s 
musculature and that of other swallows indicates that the excavatory habits 
of Riparia require little myological change from the usual hirundinid condi- 
tion. In this sense, the swallows are better preadapted myologically than 
osteologically to the task of slashing into a substrate. 

No matter how perfectly the extant physical characters of the swallows 
may lend themselves to any technique of excavation, the behavioral patterns 
requisite to that technique must be developed if the bird is to be an effective 
miner. The adoption of suitable behavioral patterns is, therefore, a key modi- 
fication. There are no data relevant to the evolution of the digging move- 
ments of the Bank Swallow. Conceivably, such movements arose from nest- 
building movements, possibly in connection with the enlargement or modif- 
cation of pre-existing holes. A careful study of the nest-building movements 
of the entire genus Riparia might provide useful information in this regard. 

Although slashing behavior may be the key to effective use of hirundinine 
adaptations, it is not necessary to assume that appropriate behavior was de- 
veloped prior to or even simultaneously with the physical modifications. 
Many of the arguments pertaining to reduction of the bill can be applied to 
a pecking mode of excavation. As I have shown, physical modifications are 
severely limited by the general specialization of the swallows for aerial forag- 
ing. The only limits imposed on behavior are those placed upon it by the 
structure of the organism. Behavior is easily modified by the process of 
learning. In fact, there is no indication that excavation behavior of the Bank 
Swallow is presently genetically controlled. Even supposing that the general 
behavior pattern of slashing rather than pecking is now genetically con- 
trolled, it could have been a learned process at one point in the evolution of 
digging techniques. Learned behavior can be assimilated into the genotype, 
if it is selectively advantageous to do so, possibly via the Baldwin effect 
(Simpson, 1953b; Waddington, 1953). 

At this point, it is necessary to insert a word of caution concerning many 
of the major points of this study. Even given a substantial increase in sizes 
of samples, certain reservations would be wise. Virtually nothing is known 
of the selective forces operating on the skulls of swallows, either generally or 
specifically. As previously mentioned, I know of no detailed analysis of the 
feeding motions of the various swallows. The differences found in their 
flight patterns (Blake, 1948) suggest that their styles of feeding differ. Dif- 
ferences, particularly if large, would be the results of slightly different selec- 


FossoriAL ADAPTATIONS IN THE BANK SWALLOW lei 


tive forces in the past, and would modify those now operating. Again, se- 
lective forces on the skull must differ between birds with such a wide range 
of nest types and nesting sites. There are undoubtedly selective forces that 
have not been mentioned, and some that might not be readily anticipated. 

The action of varying selective pressures, mostly unexplained, can be 
seen in even the limited number of specimens and species at my disposal. 
Certain variations in myology have been mentioned. The skulls of Tachy- 
cineta thalassina, Iridoprocne bicolor, and Stelgidopteryx ruficollis are all 
much alike and may be taken to represent a hypothetical “basic pattern.” 
Riparia riparia differs from this pattern only in the relative dimensions of 
the bill. Hirundo rustica seems to represent a slightly narrowed and elongate 
version of the pattern, and Petrochelidon pyrrhonota is the opposite, being 
short and broad. Progne subis is quite different from the pattern in many 
features, especially in mass. 


The Rough-winged Swallow 


The paucity of detailed information on whatever excavation techniques 
the rough-wing may employ in modifying Bank Swallow burrows invali- 
dates any attempt at a thorough analysis of its skull. In view of what has 
been said about the Bank Swallow, however, there are enough data present 
for certain speculations. They indicate (Tables 2, 7) that the rough-wing 
has an extremely flat bill, a small CLR-complex, and relatively large splenius 
capitis and biventer muscles. I have noted that the burrows of Bank Swal- 
lows that appeared to have been modified by rough-wings were frequently 
modified more extensively in a lateral than a vertical direction. The flatten- 
ing of the bill and large size of the splenius capitis strongly suggest that, 
when and if the rough-wing does use its bill for excavation, it does so by 
slashing in a horizontal plane. Its adaptations, then, would follow the pat- 
tern of a sword blade. Since this induces severe risk of fracture at the mar- 
gins of the bill in the event of any rotary motion, it is of advantage to reduce 
the possibilities of such movement. It has been mentioned that the CLR- 
complex can impart rotary motion to the skull, and this is the group of 
muscles that appears smallest in the rough-wing relative to both the total 
muscle mass and the length of the skull. The smallness of the CLR-complex 
would reduce somewhat the ability of the bird to tilt its head upward. Pos- 
sibly in compensation for this, the biventer appears to be relatively large. It 
would seem that Stelgidopteryx has evolved one of the possible, but seem- 
ingly inefhcient, modifications for digging. If so, this may explain why the 
rough-wing does not excavate its own burrow but only modifies existing 
holes to its own needs. In terms of Figure 10, the rough-wing may be rep- 
resented by line “A” or “B.” It has begun to modify, but has been prevented 
by prior specialization from achieving an effective solution. 


138 Tue Universiry ScrENcE BULLETIN 


Evolution ef Burrowing in Swallows 


It is perhaps not beyond the scope of this study to offer some comments 
on the evolution of burrowing in the swallows. The major question is, 
“What advantages accrue to those swallows nesting in burrows, either pre- 
existing or self-excavated?” The answer is actually rather obvious. The evi- 
dence clearly suggests an advantage of differential productivity gained 
through increasing the number of available nesting sites. 

The swallows as a group seem to be severely restricted in the types of 
nesting sites that they may choose. All swallow nesting sites have one factor 
in common. They are all so situated that, immediately upon leaving the 
nest, the birds are in open airspace. Thus, we find swallows nesting in cavi- 
ties in trees and in the ground, in artificial nest-boxes (but usually not if the 
box is close under leafy branches), in mud-cup nests on real or artificial 
ledges, and in closed mud nests that are plastered to vertical walls. 

A paucity of nesting sites seems to be a major limiting factor on popula- 
tions of North American swallows. Local populations can be increased 
dramatically if appropriate nesting sites are provided (Buss, 1942; Goodsell, 
1919; McCanne, 1936). Austin and Low (1932) comment, “The relative 
abundance of Tree Swallows on Cape Cod during its breeding season de- 
pends directly on the availability of nesting sites.” Low (1933) further re- 
marks on the success of artificial nest-boxes. Johnston and Hardy (1962) 
attribute the early arrival (up to two months prior to egg-laying) of the 
Purple Martin in the breeding area to “intraspecific ‘competition’ for nesting 
cavities in past time, prior to the relatively recent availability of man-made 
colony houses.” Similar, though not so dramatic, patterns of early arrival 
have been reported for the Violet-green Swallow (Gullion, 1947), the Tree 
Swallow (Paynter, 1954), and the Rough-winged Swallow (Lunk, 1962). 
McCanne (1936) remarks that Barn and Cliff Swallows are able to delay 
nesting until mud is available for the construction of their nests. The delay 
of nesting in the Rough-winged, Barn, and Cliff Swallows can be attributed 
to a lack of immediately available sites or materials. For the Purple Martin, 
the early arrival date insures sufficient time in which to search out nesting 
cavities before the onset of those conditions necessary for the rearing of a 
brood. 

I have noted that Bank Swallows that nest at the river site usually com- 
mence excavation somewhat later than those at the sand pits. In 1961, birds 
delayed the onset of excavation at the river until late May and early June. 
The reasons for these delays, especially the latter, are obscure but may be 
related to the flood cycle of the river or associated phenomena. I have been 
unable to determine any correlation between the onset of excavation and any 
single cause or obvious combination of phenomena. 


FossortAL ADAPTATIONS IN THE BANK SWALLow 139 


Many swallows have evolved behavioral patterns that tend to increase the 
number of available sites. Some, as Stelgidopteryx, will nest in a wide va- 
riety of places (Lunk, 1962). Most are colonial. Emlen (1952) correctly 
points out that colonialism and reduced territorial behavior allow the Cliff 
Swallow to make maximum use of available sites. The same argument can 
be applied to the majority of swallows. Even those species of North Ameri- 
can swallows that are only weakly colonial show reduced territorial be- 
havior. Thus, if the opportunity presents itself, a colony may be formed at a 
suitable nesting site. The Purple Martin is a prime example. Before the 
advent of European man, Purple Martins nested in separate pairs or small 
colonies, as they still do in less settled portions of North America. Today, it 
is rare to find a single pair east of the Rocky Mountains. Here the birds take 
advantage of colonial houses and are considered colonial birds. In view of 
this evidence, it seems probable that any modification that would permit a 
swallow to exploit new or relatively unused nesting sites would confer a 
distinct selective advantage. 

It is difficult to say if the habit of nesting in earthen cavities developed 
before, with, or after cavity-nesting in trees, or even if the use of cavities of 
any sort is primitive within the swallows. Mayr and Bond (1943) suggest 
that Phedinia, a cup-nest builder, may be the “least specialized” of the Old 
World swallows, but also refer to Riparia as “Apparently a rather primitive 
genus.” In any case, the problem is somewhat academic. The first ground- 
nesting swallows undoubtedly used pre-existing burrows, as many species 
still do. Seemingly, there would be selective advantage in being able to per- 
form more and more extensive modifications on less and less suitable holes. 
Eventually, the birds would be able to construct their own burrows. 

The present center of distribution of the genus Riparia is in Africa, where 
it probably evolved, although Darlington (1957) remarks that “swallows 
have plainly undergone successive as well as multiple dispersal, and I hesi- 
tate to deduce any simple history from their present distribution.” As there 
are no burrowing swallows in Europe, northern Asia, or North America 
other than Riparia riparia, it is probable that at one time a vast area awaited 
any of the burrowing African swallows that developed the physiological 
adaptations requisite to a northward extension of range. It is, of course, im- 
possible to state just what factors are responsible for the ability of an organism 
to occupy an extensive range, unless detailed information concerning that 
organism’s ecology, physiology, and history is available. While the ecology of 
the Bank Swallow has been well studied, its physiology is but incompletely 
known and there are no data available as to the history of the species. Un- 
questionably, however, a successful solution to the housing problems has 
been a major factor in the occupancy of the relatively enormous range en- 


joyed by this bird. 


140 Tue University ScrENcE BULLETIN 


SUMMARY 


Studies of fossorial modifications in the Bank Swallow (Riparia riparia) 
are described and discussed in this paper. 

The members of the family Hirundinidae are all highly specialized for 
aerial foraging. They are characterized by generally light construction with 
short, flat bills, small legs and feet, and long, narrow wings. In spite of the 
delicacy of their structure, members of three not closely related genera dig 
nesting burrows, a behavior frequently associated with sturdy anatomy. 

The Bank Swallow is the only North American swallow that characteris- 
tically performs extensive excavation. The burrows are dug not by pecking 
or drilling but by lateral or dorso-lateral slashing at the substrate with the bill. 

Comparisons were made of proportions of the skull between Riparia 
riparia and six other North American swallows, Tachycineta thalassina, 
Iridoprocne bicolor, Stelgidopteryx ruficollis, Hirundo rustica, Petrochelidon 


pyrrhonota, and Progne subis. These comparisons show that the Bank Swal- | 


low’s bill is small relative to cranial measurements and the Sternal Plate 
Length. In addition, the Bank Swallow’s bill is more nearly round in cross- 
section than those of other swallows. 

Similar comparisons were made of the weights of five pairs of muscles 
that control the approprifhate slashing movements, M. cucullaris, M. biventer 
cervicis, M. splenius capitis, M. rectus capitis lateralis, and M. rectus capitis 
ventralis. [ridoprocne albilinea was used in these comparisons in addition to 
the other six species. The comparisons indicate that, with the exception of a 
relatively large cucullaris-lateral rectus complex, the weights of muscles of 
the Bank Swallow are appropriate to a swallow of its size. 

The small size of the Bank Swallow’s bill decreases the effects of torque 
upon the bill and the frontonasal hinge, reduces the risk of fracture across 
the bill, and slightly increases the mechanical advantage of the excavating 
structures. The rounding of the bill reduces its resistance to the substrate, 
thereby reducing torque and permitting slashing in other than a strictly 
lateral plane. 

In a specialized organism, resistance is imposed against modification by 
the co-adapted system. Therefore, successful modifications are likely to be 
those that initially impart a high degree of efficiency with a minimum of 
change. The small size of the bill in the Bank Swallow can be considered 
to be reduced from that of its ancestors, and this represents an efficient solu- 
tion by virtue of the fact that it confers several advantages. 

The lack of extensive modification of the muscles studied suggests that 


the swallows as a group are myologically preadapted to a slashing mode of | 


excavation. The muscular requirements for aerial foraging seem to be quite 
similar to those for lateral or dorso-lateral slashing. 


FossortaL ADAPTATIONS IN THE BANK SWALLOW 141 


The Rough-winged Swallow (Stelgidopteryx ruficollis ) also nests in bur- 
rows, but it does not excavate them. In comparison with the Bank Swallow, 
the rough-wing seems to be but slightly modified for digging, and such 
modifications as it does possess appear to be quite inefficient. 

The Hirundinidae as a group are pressed for available nesting sites. The 
use of pre-excavated cavities is an adaptation to increase the numbers of 
available sites. Burrowing probably evolved from attempts to modify al- 
ready existing holes of various types. 

The successful development of fossorial adaptations is undoubtedly a 
contributing factor to the wide range of the Bank Swallow. 


LITERATURE, ChlED 


Auten, A. A. 1933. The Bank Swallow’s story. Bird Lore 35:116-125. 

Austin, O. L., Jk. AND Low, S. H. 1932. Notes on the breeding of the Tree Swallow. Bird- 
Band. 3:39-44. 

Baitey, H. H. 1913. The birds of Virginia. Lynchburg, J. P. Bell Co. 

Brecuer, W. J. 1950. Convergent evolution in the American orioles. Wilson Bull. 62:50-86. 

195la. Adaptations for food getting in the American blackbirds. Auk 68:411-440. 

1951b. Convergence in the Coerebidae. Wilson Bull. 63:274-287. 

Brake, C. H. 1948. The flight of swallows. Auk 65:54-62. 

Boas, J. E. V. 1929. Biologisch-anatomische Studien tber den Hals der Vogel. Kongl. Danske 
Vidensk. Selsk. Skrifter, Naturv. Math. Ser. 9, 1:102-222. 

Bock, W. J. 1959. Preadaptation and multiple evolutionary pathways. Evol. 15:194-211. 

Bowman, R. I. 1961. Morphological differentiation and adaptation in the Galapagos finches. 
Univ. Cal. Publ. Zool. 58. 

Brown, H. E. anp ScHWACHTGEN, E. C. 1949. Physics, the story of energy. Boston, D. C. Meath 
and Company. 

Burns, F. L. 1924. The philosophy of birds’ nests and comparative caliology in consideration 
of some local nidicolous birds [Part I]. Wilson Bull. 36:78-88. 

Burt, W. H. 1930. Adaptive modifications in the woodpeckers. Univ. Cal. Publ. Zool. 32:455- 


524. 
Buss, I. O. 1942. A managed Cliff Swallow colony in southern Wisconsin. Wilson Bull. 54: 
153-161. 


Cooper, R. 1955. Unusually exposed Sand Martin’s nest. Brit. Birds 48:179. 

Cours, E. 1876. Peculiar nesting site of the Bank Swallow. Bull. Nutt. Ornith. Club 1:96. 

Daruincton, P. J., Jr. 1957. Zoogeography: the geographical distribution of animals. New 
York, John Wiley and Sons, Inc. 

Davis, D. D. 1949. Comparative anatomy and the evolution of vertebrates. In: Genetics, pale- 
ontology and evolution. Jepsen, Mayr, and Simpson (eds.), Princeton Univ. Press. 

Dawson, W. L. 1923. The birds of California. Vol. I. San Diego, South Moulton Company. 

Dincte, E. v. S. 1942. Rough-winged Swallow. In: Life histories of North American fiy- 
catchers, larks, swallows, and their allies. A. C. Bent. U.S. Nat. Mus. Bull. 179. 

Exior, S. A., Jr. 1932. Notes from the Connecticut Valley of Massachusetts. Auk 49:101-103. 

Emten, J. T., Jr. 1952. Social behavior in nesting Cliff Swallows. Condor 54:177-199. 

Fisner, H. I. 1958. The “hatching muscle” in the chick. Auk 75:391-399., 

1961. The hatching muscle in North American grebes. Condor 63:227-233. 

1962. The hatching muscle in Franklin’s Gull. Wilson Bull. 74:166-172. 

Fisher, H. I. anp GoopMan, D. C. 1955. The myology of the Whooping Crane, Grus amert- 
cana. lll. Biol. Mono. 24. The Univ. of Ill. Press. 

vAN FLEET, W. 1876. Notes on the Rough-winged Swallow (Hirundo serripennis), in Penn- 
sylvania. Bull. Nutt. Ornith. Club /:9-11. 

Forses, W. A. 1882. On some points in the anatomy of the Indian Darter (Plotus melano- 
gaster), and on the mechanism of the neck in the darter (Plotus), in connection with 
their habits. Proc. Zool. Soc. Lond. 208-212. 

GoopMan, D. C. ano FisHer, H. I. 1962. Functional anatomy of the feeding apparatus in 
waterfowl (Aves: Anatidae). Carbondale, Southern Ill. Univ. Press. 


142 THe University ScrENCE BULLETIN 


GoopseL_, W. 1919. Helping Barn and Cliff Swallows to nest. Bird Lore 21:175-176. 

Guituion, G. W. 1947. The use of artificial nesting sites by Violet-green and Tree Swallows. 
Auk 64:411-415. 

Henson, B. R. AND JoHNson, J. D. 1955. Nest boxes designed for Sand Martins. Brit. Birds 
48:180. 

Hickuine, R. A. O. 1959. The burrow-excavation phase in the breeding of the Sand Martin 
Riparia riparia. \bis 101:497-500. 

Jounston, R. F. anp Harpy, J. W. 1962. Behavior of the Purple Martin. Wilson Bull. 74: 
243-262. 

Low, S. H. 1933. Further notes on the nesting of Tree Swallows. Bird-Band. 4:76-87. 

Lunx, W. A. 1962. The Rough-winged Swallow, a study based on its breeding biology in 
Michigan. Cambridge, Publ. Nutt. Ornith. Club 4. 

Mayr, E. anp Bonn, J. 1943. Notes on the generic classification of the swallows, Hirundinidae. 
Ibis 85:334-341. 

McCanne, H. D. 1936. A colony of Cliff Swallows in Chester County, Pa. Auk 53:84-85. 

McCue, J. J. G. 1961. How bats hunt with sound. Nat'l. Geographic 119:570-578. 

Pacmcren, P. 1949. Zur biologischen Anatomie der Halsmuskulatur der Singvogel. Im; Orni- 
thologie als biologische Wissenschaft. Heidelberg, Carl Winter. 

Paynter, R. A. 1954. Interrelations between clutch size, brood size, prefledgling survival, and 
weight in Kent Island Tree Swallows [Part I]. Bird-Band. 25:35-38. 

Perersen, A. J. 1955. The breeding cycle in the Bank Swallow. Wilson Bull. 67:235-286. 

Renscu, B. 1960. Evolution above the species level. New York. Columbia Univ. Press. 

ScHUFELDT, R. W. 1890. The myology of the Raven. London, MacMillan and Co. 

SerLte, W. 1950. A contribution to the ornithology of the British Cameroons. Ibis 92:602-638. 

Simpson, G. G. 1949. The meaning of evolution. New Haven, Yale Univ. Press. 

1953a. The major features of evolution. New York, Columbia Univ. Press. 

1953b. The Baldwin effect. Evol. 7:110-117. 

Srockarb, C. R. 1905. Nesting habits of birds in Mississippi [Part II]. Auk 22:273-288. 

Sroner, D. 1925. Observations and banding notes on the Bank Swallow—I. Auk 42:86-94. 

1936. Studies on the Bank Swallow, Riparia riparia riparia (Linnaeus). Roosevelt 

Wildlife Annals 4:121-233. 

Surron, G. M. anp Pertinaiti, O. S., Jr. 1942. Birds of the Gomes Farias region, southwest- 
ern Tamaulipas. Auk 59:1-34. 

Toosy, J. 1947. Notes on Sand Martins. Brit. Birds 40:290-297. 

Tyter, J. G. 1913. Some birds of the Fresno district, California. Pacif. Coast Avi. 9. 

WappincTon, C. H. 1953. Genetic assimilation of an acquired character. Evol. 7:118-126. 

WepEMEYER, W. 1933. Nesting of the Rough-winged Swallow in Montana. Auk 50:362-363. 

Zust, R. L. 1962. Structural adaptations in the head and neck of the Black Skimmer. Cam- 
bridge, Publ. Nutt. Ornith. Club 5. 


| 


Fossor1aL ADAPTATIONS IN THE BANK SWALLOW 


APPENDIX A 


The tables in this appendix provide additional data concerning the pro- 
portions discussed in text. Table A-1 concerns proportions first appearing in 
Table 1, Table A-2 those first appearing in Table 2, etc. The data for each 
proportion are presented only once, in the table corresponding to the first 
text table in which the proportion appears. The data are presented in the 
following pattern: 


Riparia 
Tachycineta 
Iridoprocne 
Stelgidopteryx 


Hirundo 


Petrechelidon .... 


Mean = 2 Standard Errors of the Mean 
Range; Number. 


Tape A-l. Bill Length, Data. 


Bill Length — Bill Length Bill Length Bill Length Bill Length 
Cranial Sternal Plate Interquadratic Frontonasal Maxillary 
Length Length Width Width Width 

-459==.020 543.030 401.008 1.707.102 894.060 
.412—.503;10 .517—.612;9 .377—.430;6 1.500—1.976;9 .797—1.097;9 
476==.012 pean iW 419+.016 1929225128 965.012 
.448—.509:5 .528—.582:5 .410—.431;5 1.738—2.102;5 .937—.987;5 
486.022 536.020 411.012 1.914.066 .943+.048 
.420—.531;16 .477—.596;16 .364—.446;16 1.673—2.065;16 .795—1.097;15 
.489+.028 557+.034 asd is: 1.707.142 865.058 
411—.548;8 .474—.622;8 .371—.453;8 1.358—1.957;8 .734—.989;8 
59 ==. 014 639.020 453.008 22 VS=:064 .882+.020 
517—.615;14 .589—.700;13 .428—.470;13 1.937—2.382;14 .823—.973;14 
475.008 556.010 406.006 1.684.044 .841+.028 
.437—.514;20 .506—.613;20 .376—.446;20 1.500—1.934;20 .766—.922;18 
643 660 468 1.920 947 
.636—.650;2 .650—.671;2 .466—.470;2  1.907—1.933;2  .947;2 


144 Tue University Science BULLETIN : 
Tape A-2. Bill Depth, Data. 
Bill Depth Bill Depth Bill Depth Bill Depth Bill Depth 
Cranial Sternal Plate Frontonasal —*é&Bll:— “Maxillary” 
Depth Length Width Length Width 
TRADI: poset cc 224.012 .158+.010 .482+.050 .292+.014 .250+.028 
-189—.2523;12 .130—.178;12 .428—.571;11 .214—.333;9 .212—.319;12 
Tachycineta ........ .238+.008 .169+.004 579.020 298+.016 29 4S 022 
.216—.262;16 .145—.215;19 .534—.704;19 .268—.328;5 .256—.369;19 
Iridoprocne ......... .242+.028 159.004 560+.016 .288+.012 .279+.008 
.209—.294;23 :135—.186;24 .488—.652;24 .235—.350;14 .242—.316;24 
Stelgidopteryx .220+.012 147.004 449+ 028 .260+.008 162.008 
.198—.244;8 .132—.162;8 .377—.510;8 .235-+.277;8 .148—.176;7 
Eland Ome .247+.010 163.008 546.020 Ds fata .226+.008 
.207—.300;22 .138—.194;21 .469—.652;23 .221—.303;14 .190—.275;23 
Petrochelidon .234+.006 .162+.006 497==012 .294+.008 .243+.006 
.203—.274;28 .140—.181;28 .442—.549;29 .266—.325;16 .215—.272;28 
PrOgne ee 343.012 .201+.004 07 9'=:024 303 .284+.004 
322—.375;7. .183—.220;8 .532—.640;8 .296—.310;2 .270—.308;8 
Tasie A-3. Maxillary Width, Data. - 
Cranial Width Sternal Plate Length 
Maxillary Width Maxillary Width 
Ripanigs 22 22. itis ee ee) ee eee 1.576+.064 1.657+.090 
1.415—1.806;13 1.483—2.076;13 
DiQCRY GINCLG: 50 sacxs, ahah 5 nen ne 1.671.022 1.748+.038 
1.557—1.747;18 1.612—1.852;19 
THdODEOCNE. het, = Wn Aut: SS PRE A Me PAS ES 5 1.597+.026 1.760.042 
1.469—1.726;25 1.552—1.976;25 
SHCI STAG PICHIA Ins oe Ve i eR OO eB 1.413+.036 1.553.048 
1.360—1.494;7 1.469—1.663;8 
ET IUH GOMER oe cate SN OR eens stern h LU de 1.256+.016 1.390+.018 
1.183—1.346;22 1.292—1.491;22 
Remochclido pe we sae ae hee LN eS Tee 1.468+.026 eae (034i 
1.357—1.652;29 1.357—1.764;29 
TROLL Pes ce Sn SR a OR Be AAD RP ee 1.142.036 1.418.058 


1.061—1.189;8 


1.252—1.547;8 


Roparig  _.....- 
Tachycineta 

Iridoprocne . 
Stelgidopteryx 
Hirundo _ ..... 


Petrochelidon 


Riparia 


Tachycineta 


Iridoprocne . 


Stelgidopteryx 


Hirundo 


FossorrAL ADAPTATIONS IN THE BANK SWALLOW 


Tasie A-4. Frontonasal Width, Data. 


145 


Frontonasal Frontonasal Frontonasal Frontonasal 
Width Width Width Width 
Length Width Length Width 

318+.014 58251032 .273+.006 2 O=t=. 024 

.262—.356;13 311—.360;13 .253—.291;12 .477—.611;12 

294+ .006 307.004 249.006 py lilse( 07, 

.268—.323 ;20 .282—.325;20 .234—.274;19 .469—.558;19 

.282+.004 312.002 256.004 498.006 

.259—.304 326 .286—.335 ;26 .239—.276;26 -447—.541;25 

327.014 358+.018 288+.006 508.060 

300—.353;8 330—.392;8 .262—.306;8 .480—.540;8 

299.004 352 =t=,002 264.004 416.004 

.269—.327 ;22 .312—.356;22 .246—.284;24 389—.452;24 

326.006 334+.004 .278+.004 490.010 

304—.364;32 3 15—.364;32 .262—.298 333 -450—.550;30 

346.010 389+.006 337+.008 E48 Osten (G 

.336—.372;8 375—.402;8 326—.345;8 .463—.527;8 
Tasie A-5. Interquadratic Width, Data. 

Interquadratic Interquadratic Interquadratic Interquadratic Interquadratic 
Width Width Width Width Width 
Cranial Sternal Plate Skull Cranial Maxillary 
Width Length Length Length Width 

785+.016 727.036 -438+.006 634.008 IIS 9S=I056 

.755—.830;9 .664—.818;9 .409—.455;6 .619—.651;8 1.095—1.379;9 

780+.018 746+.014 438.008 634.008 1.304.028 

.733—.8833;19 .693—.826;19 .427—.464;5 .598—.691;19 1.231—1.461;19 

812.010 Saas 04 447.008 666.002 1.298.020 

.765—.840;26 .660—.785;26 .429—.460;16 .620—.704;26 1.200—1.393;25 

.770+.016 .673+.018 410.018 611.006 1.070.018 

.720—.807;6 .644—.729;7 .376—.441;7 .583—.626;7 1.029—1.112;7 

800.012 221014 407.002 635.008 1.003.018 

.736—.866;21 .642—.768;21 .397—.427;14 .589—.676;23 .913—1.069;23 

845.008 826.006 .479+.004 704+.004 1.241+.024 

.810—.900;32 .786—.870;32 .461—.498;20 .681—.733;33 1.144—1.404;30 

885.020 .788+.014 464 .768+.010 TEMS e036 

.821—.931;8 .757—.823;8 .459—.469;2 .741—.798;8 1.030—1.184;8 


146 Tue University ScreNcE BULLETIN 
' 
Taste A-6. Miscellaneous Measurements, Data. 
Cranial Width Skull Length Cranial Length 
Sternal Plate Sternal Plate Sternal Plate 
Length Length Length 
Riparia (2 2s ae ee ee: 956+.040 1.704+.014 Lli7oe=084 
.817—1.056;14 1.470—1.871;11 1.056—1.282;13 — 
ACHYCINCIG. = = 954+.014 1.735.054 1.181.018 
.900—1.038;19 1.683—1.815;5 1.121—1.271;19 
ltd OPrOCHhC een nee ee 904+.014 1.637 +.036 1.105.016 
.830—.973 326 1.506—1.738;16 1.029—1.185;26 
SELOTdLG PICK Ir rere ne eee 904.028 1.694.058 1.140.010 
.840—.964 ;7 1.601—1.833;8 1.217—1.067;8 
Eirando. 12 ab a aek sees I01+.008 1.778+.032 als 7014 
.862—.937 320 1.706—1.863;13 1.09 1—1.187 ;22 
Perrocheidop, 22 978+.010 1.718.028 1.176.008 
.906—1.032;31 1.576—1.839;20 1.120—1.242;32 
PYOONE. cD eee SE Hee 807.048 1.688 1.027.024 
-768—.863 ;8 1.650—1.727;2 .987—1.088;8 
Taste A-7. Muscle Weights (Cube Roots), Data. 
Ventral Rectus CLR-Complex Splenius Biventer Total Mass 
Skull Length Skull Length Skull Length Skull Length — Skull Length 
IRIPAyiGme ee eee 055.001 074.002 066.002 054.002 100.003 
.052—.057:7 .068—.078;9 .061—-.069;9 .044—.059;9 .094—.104;7 
Tachycineta ......------ 054 070 l 053 .097 
[emaloilincam 056 072 056 100 
056—.057:2 .070—.073:2 .065—.066;2 .054—.057;2 .100—.101;2 
M5 WHA QUO ccncnorcecteee .052 .067 .056 .095 
Stelgidopteryx — ......- .052+.001 068.003 065.002 053.002 096.004 
0.48—.058:9 .063—.076:9 .060—.072;9 .050—.058;9 .089—.106;9 
Hirundo 2.2222. .052+.002 070.002 066.002 054.003 .098+.003 
048—.055:6 .066—.074:6 .062—.069;6 .049—.059;6 .091—.102;6 
Petrochelidon _..... 059.003 077.001 068.002 056.002 -105+.002 
.052—.067:11 .072—.080;11 .059—.073;11 .052—.062;11 .099—.112;@ 
PORNO) ease cate ee .062+.001 080.005 .07 1.001 059.002 110.002 
061—.064:5 .078—.081;5 .068—.074;5 .054—.062;5 .105—.112;5 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


A REVISION OF MICRONECTA OF 
AUSTRALIA AND MELANESIA 
(HETEROPTERA: CORIXIDAE) 


By 
Ling-chu Chen 


Voi. XLVI Paces 147-165 June 1, 1965 No. 3 


te i ¥ ook ER ; 

4 aan 4 N° j LA , ‘wea f 

Us Seba Ag KASEY | i Re 

pep ye eM ; Paiste ri! 
Tey vA aT Le wr \ aa f Y 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 147-165 June 1, 1965 No. 3 


A Revision of Micronecta of Australia and Melanesia 
(Heteroptera: Corixidae )' 


By 


Linc-cHU CHEN 


INTRODUCTION 


It has been more than 40 years since the publication of Hale’s (1922) 
paper, “Studies in the Australian Aquatic Hemiptera.” There has been no 
other work concerning the Micronecta of the Australian region except one 
on the Melanesian species by Wrdblewski (1962), which contains redescrip- 
tions of two species and a statement on the synonymy of another. 

The structural differences among species in the genus have been more 
fully recognized in recent decades. It is, therefore, necessary to redescribe 
those species which have not been reviewed since Hale’s work. Besides 
redescriptions, eight new species and a key to the Australian and Melanesian 
species of Micronecta are included in this paper. Kirkaldy described three 
species from Australia, but unfortunately only one has been recognized for 
certain, and his type specimens cannot be located. The original descriptions 
of his two unrecognized species are reproduced verbatim. 

I wish to express my appreciation to all who have assisted me in this 
study. Dr. H. B. Hungerford placed at my disposal all the Micronecta in 
the Snow Entomological Museum plus some types lent from the Hungarian 
National Museum and the undetermined Micronecta that had been sent to 
him from the following museums: British Museum (Natural History), 
California Academy of Science, Chicago Natural History Museum, Hun- 
garian National Museum, Museum of Comparative Zoology (Harvard Uni- 
versity), South Australian Museum, and the United States National Mu- 
seum. Dr. C. D. Michener and Miss Ellen Ordway have helped in revising 
the manuscript. 

"Contribution number 1208 from the Department of Entomology, The University of Kansas. 


This study was made possible by a grant from the National Science Foundation to the 
University of Kansas for research directed by the late Dr. H. B. Hungerford. 


148 


Tue University SCIENCE BULLETIN 


KEY TO THE MICRONECTA OF AUSTRALIA 
AND MELANESIA (MALES) 


1. Male genitalia with sinistral asymmetry; free lobe of eighth tergite 

with two outer, lower and upper, setigerous angles (Fig. 14) M. sinistra 

Male genitalia with dextral or sinistral asymmetry; free lobe of eighth 

tergite only with outer, lower, setigerous angle -..........--..---::-:-s:2:s:eeso-eneonnomnee 2 
2. Body less than 2.5 mum. lomg -....-..--------------esseeesceeeeee eee eee ctee tena eteneccemecemnaas 3 

Body more than 2.5 mmm. long -...-.---..-----+--++--:+-ce ee cscs eeec erect tee eteestentneecencens 5 
3. Interocular space as wide as an eye OF MALOWE -...------------ ee eee 4 

Interocular space wider than an eye; parameres as in Figs. 52-53 _..... 

PI Og OR DN Oca ee AND at Dae EE EE re rec M. australiensis 
4. Clavus with three dark stripes; interocular space narrower than an 

VC cs oe ae eee ee eee M. ludibunda 

Clavus with two dark stripes; interocular space about as wide as an 

eye; parameres as in Pigs) 1-4 22222 odie M. micra 
5. Body 3.5 mam. to 5 mim. lomg ...-c-oee-c- cece eesseeeeces cence sence 6 

Body 2.5 mm. to 3.5 mam. lomg —....-------------sseecseenneeeeeccceceeceeeettne tenses 7 
6. Vertex roundly produced; palar claw with one margin greatly ex- 

panded as in Fig. 44 ......---.cc-.sccccccseccsseces cscs cece eeeeeececatenetrnesttneseeecenaa M. robusta 

Vertex conically produced; palar claw elongate (Fig. 36)... M. major 
7. Hemelytra of uniform color .....--.-------------esceecceeee eects cee 8 

Hemelytra with dark maculations ........-..-----2--------------- eer 9 
8. Hemelytra black -........2.-..-2---cssccesccscceececesssesnseecesceseenenesenstereectenenseees M. carbonaria 

Hemelytra yellowish brown; parameres as Figs. 80-82 ~ + M. windi 
9. Corium with many diffuse blotches centrally; pronotum with a dis- 

tinct ridge behind posterior margin of head; parameres as in Figs. 

SHB | Wie ef ta eR Se aa i ee oo a epee cleo Lon M. carinata 

Corium with four broken, dark, longitudinal stripes; pronotum with- 

out a distinct ridge behind posterior margin of head... 10 
10. Interocular space about 1.60 times as wide as an eye; middle femur 

long, about 41°% of the body length eee M. virgata 

Interocular space less than 1.55 times as wide as an eye; middle femur 

less than 40°% of the body lemgth .......--.-----.-------s::-e--ecceeeeeeeetestettecteceeetceea 11 
11. Free lobe of eighth tergite with inner angle obsolete and outer angle 

distinctly prolonged, completely margined with bristle-like hairs as in 

Bi 7B oo acecc cece anseceectaancceeececcessecesnpere esse aseenetteenseeeeesorsecestos othe tta 12 

Free lobe of eighth tergite with a well developed inner angle and an 

outer setigerous angle as in Fig. 28 ......-..-..--:---------e-ec 13 
12. Second hair of the posterior lower flexor margin of pala slightly en- 

larged, palar claw with a distinct notch distally (Fig. 79); parameres 

as ity Figs, 73275 cccccscce-sncccesecceeecosccennesccrcneeonseeenseseeteeneccencceanarsse M. queenslandica 

No special enlarged hair on the posterior flexor margin of pala, palar 

claw completely margined —.........--.--.--::-1------- eee M. quadristrigata 


13. 


Palar claw with one margin greatly expanded (Fig. 29); parameres as 
im Pigs, 24225) coo setecseesce cc ecesse rece eee ene eeee recta eee M. gracilis 


A Revision oF Micronecta or AUSTRALIA AND MELANESIA 149 


Halareclaw evenly expanded (Fig. 72) .:2.... akc etc etesctec dence 14 
14. Distal end of palar claw swollen (Fig. 72); parameres as in Figs. 

(ADC pe IES Mee re cee a ey a oe ee M. hale: 

Misavend of palar claw monmal (Fig. 23)) 2c occcctcceccc ec eececcecteececceecececeseceeeenes 15 


15. Vertex strongly produced, head as long as (macropterous form) or 
longer (brachypterous form) than pronotum; parameres as in Figs. 


LSD: ce nner, Oe Gees LO es ee, Th SO M. batilla 
Vertex slightly produced, head shorter than pronotum; parameres as 
“Rt FAS i ee En Oe Pd OE .... M. adelaidae 


Micronecta micra Kirkaldy 
Micronecta micra Kirkarpy, 1905, p. 26; Harz, 1922, p: 328: 

Size. Brachypterous form, length 1.9 mm. to 2.0 mm. 

Color. Yellowish brown, pattern in seven specimens studied almost com- 
pletely effaced. Venter and legs grayish yellow. 

Structural characteristics. Head about twice as long as pronotum, vertex 
roundly produced in front of eyes; interocular space about as wide as eye; 
length to width of pronotal disk as 1.5:5.5, lateral margins of pronotum very 
short, making head almost contiguous with corium, posterior margin of 
pronotum almost truncate. Hemelytra with numerous minute setae. Wings 
reduced, extending to third abdominal tergite. Prestrigilar flap (fig. 5) broad 
and short with round tip. Submedian process of seventh abdominal sternite 
(fig. 6) broad basally, abruptly pointed at tip, with four extremely elongate, 
enlarged bristles subbasally. Free lobe of eighth tergite (fig. 7) with large, 
round inner angle and rather small, pointed, setigerous outer angle, posterior 
margin between angles slightly concave. Parameres well chitinized and 
shaped as in Figs. 1-4. Foreleg similar to that of M. australiensts. 

Remarks. Kirkaldy’s types are unknown, but the Snow Entomological 
Museum has seven specimens, two males and five females, from the type 
locality, Kuranda, northern Queensland. These specimens agree well with 
the original description, especially considering the body size and posteriorly 
truncated pronotum. 


Collection data. Two males and five females, Kuranda, northern Queens- 
land, Aug. 12, 1938 (R. G. Wind). 


Micronecta batilla Hale 
Micronecta batilla Hare, 1922, p. 323. 


Size. Macropterous form, length 2.8 mm. to 3.3 mm. Brachypterous form, 
length 2.6 mm. to 3.1 mm. 

Color. Grayish or yellowish to dark brown; vertex sometimes uniformly 
colored, usually with three parallel, reddish, longitudinal stripes. Pattern on 


hemelytra as in M. robusta. Venter and legs usually pale. 
| 


| 


Tue University SCIENCE BULLETIN 


Fics. | to 7. M. micra. 1-2, left paramere; 3-4, right paramere; 5, prestrigilar flap; 6, sub- 
median process of seventh abdominal sternite; 7, free lobe of eighth abdominal tergite. Fics 
8 to 15. M. sinistra. 8-9, left paramere; 10-11, right paramere; 
submedian process of seventh abdominal sternite; 

15, foreleg of male. 


12, prestrigilar flap; 13, 
14, free lobe of eighth abdominal tergite; 


A Revision oF Micronecta oF AUSTRALIA AND MELANESIA 151 


Structural characteristics. Head about one and one half times as long as 
pronotum, width to length of pronotal disk as 3.1:0.9 in brachypterous form. 
In macropterous form, head as long as pronotum, width to length of pro- 
notal disk as 3.3:1.2. Vertex noticeably produced beyond anterior margins of 
eyes, more so in brachypterous form. Interocular space as wide as eye or 
slightly wider; posterior margin of pronotum rounded. Hemelytra with 
scattered minute hairs. In brachypterous form, wings extending to seventh 
abdominal tergum. Prestrigilar flap (fig. 20) and submedian process of 
seventh abdominal sternite (fig. 21) similar to those of M. robusta. Free lobe 
of eighth tergite (fig. 22) with well developed rounded inner angle and 
slightly produced setigerous outer angle. Right paramere (fig. 19) slightly 
dilated before tip and gradually narrowed to apex; left paramere (figs. 16-18) 
denticulate with many small conical barbs, broad basally and constricted 
distally to bent apex. Foreleg (fig. 23) as in M. robusta but the spines on 
tibia uniform in size and palar claw dilated evenly. 

Remarks. In the original description, the variation of body size of this 
species is indicated as 2.75 mm. to 5 mm. In view of the strong similarities 
between M. batilla and M. major n. sp., it seems that the large individuals 
of Hale’s M. batilla actually were M. major. | have examined more than 250 
specimens, including syntypes of M. batilla, and find no intergradation with 
M. major. 

Collection data. South Australia: Adelaide (identified by H. M. Hale). 
Victoria: Bacchus Marsh, Jan., 1904. New South Wales: Pine Island, Federal 
Capital Territory (J. W. Evans); Mt. Kosciusko, 5-7000 ft., Dec. 13, 1931 
(2: J. Darlington, Harvard Exp.); Dorrigo, 3000 ft. Feb., 1932; Valley 
Heights, near Katoamba, May 23, 1954 (E. S. Brown). Queensland: Bris- 
bane, Dec., 1932 (H. Hacker); Fresh Water Creek, Redlynch, Sept. 27, 1938 
(R. G. Wind). Western Australia: Yancheyarra, June 24, 1933 (N. B. Tin- 
dale). Tasmania: Lake Leake, 1937 (J. W. Evans). 


Micronecta gracilis Hale 


Micronecta gracilis Hare, 1922, p. 326. 

Size. Macropterous form, length, 3.26 mm. to 3.53 mm. 

Color. Grayish to dull brown; vertex much lighter than rest of body, 
with dark central stripe; pattern of pronotum and hemelytra as in M. 
robusta. Venter and legs pale. 

Structural characteristics. Body elongate, length more than twice as long 
as widest part (3.4:15). Head shorter than pronotum (1.2:14), vertex 
slightly produced beyond anterior margins of eyes; interocular space wider 
than eye (1.5:1.3); width to length of pronotal disk as 3.4:1.4; pronotum 
spindle shaped. Hemelytra with relatively long, fine hairs. Prestrigilar flap 


152 Tue University ScrENcE BULLETIN 


Fics. 16 to 23. M. barilla. 16-18, left paramere; 19, right paramere; 20, prestrigilar flap; 
21, submedian process of seventh abdominal sternite; 22, free lobe of eighth tergite; 23, foreleg 
of male. Fics. 24 ro 29. M. gracilis. 24, left paramere; 25, right paramere; 26, prestrigilar 
flap; 27, submedian process of seventh abdominal sternite; 28, free lobe of eighth tergite; 
29, foreleg of male. Fics. 30 to 36. M. major. 30-31, left paramere; 32, right paramere; 
33, prestrigilar flap; 34, submedian process of seventh abdominal sternite; 35, free lobe of 
eighth tergite; 36, foreleg of male. Fics. 37 ro 44. M. robusta. 37-39, left paramere; 40, right 
paramere; 41, prestrigilar flap; 42, submedian process of seventh abdominal sternite; 43, free 
lobe of eighth tergite; 44, foreleg of male. 


A Revision oF Micronecta o— AUSTRALIA AND MELANESIA 153 


of fifth tergum (fig. 26) broad and short. Subequilateral, submedian process 
of seventh sternite with six slightly enlarged bristles subbasally (fig. 27). 
Free lobe of eighth tergite with well developed postero-external setigerous 
and antero-internal rectangular angles (fig. 28). Right paramere (fig. 25) 
evenly curved, narrowing distally to blunt tip; left paramere (fig. 24) with 
middle part of shaft almost straight, distal part sharply curved in sickle-shape, 
denticulate beyond two parts. Palar claw (fig. 29) as in M. robusta Hale. 

Remarks. The redescription is based on syntypes and specimens identi- 
fied by Hale. This species, in general appearance, is close to M. guadristri- 
gata Breddin. However, in the denticulate left paramere, the characteristic 
palar claw and the submedian process of the seventh sternite, M. gracilis 
seems more closely allied to M. robusta. 

Collection data. South Australia: Quorn (A. H. Elston) (syntype); My- 
ponga (A. H. Elston) (syntypes); Neales River at Algebuckina, May 10, 
1953 (R. A. Stirton and R. H. Tedford) Lake Callabonna, June 3, 1953 
(R. A. Stirton and R. H. Tedford); White Crossing, Coopers Creek, June 
18, 1953 (R. A. Stirton); Mt. Serle, N. Flinder Range (Hale and Tindale); 
Well 4 m. E. Oraparinna (at light), Feb. 1956 (G. F. Gross); Everard 
Ranges (A. Brumby). Queensland: Cairns district (A. M. Lea); Kings 
Creek, Aug. 25, 1954 (R. A. Stirton); St .George, 1923 (G. H. Wilkins). 
New South Wales: Carbamatta, Sydney, Jan. 31, 1958 (M. J. Nikitin); Bo- 
gan River (Mr. Beane); Valley Heights, near Katoomba, May 23, 1954. 
Western Australia: 20 m. South of Erlinda Station, Oct. 23, 1953 (N. B. 
Tindale); Yeeda Station, Aug. 25, 1953 (N. B. Tindale); Warburton Ranges 
(A. Brumby). Victoria (no detailed data on specimens). 


Micronecta robusta Hale 


Micronecta robusta Harr, 1922, p. 325. 


Size. Macropterous form, length 3.5 mm. to 4.2 mm. 

Color. Head pale to dark brown either with uniform color throughout 
or with a central longitudinal yellowish brown stripe, sometimes with light 
spot between eye and central stripe. Pronotum and hemelytra usually much 
darker than head, pronotum crossed by a pronounced, broken, transverse 
band visible only in specimens with light background. Clavus margined 
with stripe along inner and outer edges and corium marked with four longi- 
tudinal broken stripes. Venter and legs grayish yellow to grayish fuscous. 

Structural characteristics. Body stout, length less than twice as long as 
Widest part (4.2:2.0); head noticeably shorter than pronotum; vertex not 
produced but rounded in front of eyes; interocular space broad, usually one 
and one half times as wide as an eye; width to length of pronotal disk as 
4.2:1.5; many specimens with posterior margin of pronotum, in front of 


154 Tue Universiry SclENCE BULLETIN 


Fics. 45 ro 51. M. adelaidac. 45-46, left paramere: 47, right paramere; 48, prestrigilar 
flap; 49, submedian process of seventh abdominal sternite; 50, free lobe of eighth tergite; 
Fics. 52 1o 57. M. australiensis. 52, left paramere; 53, right paramere; 
54, prestrigilar flap; 55, submedian process of seventh abdominal sternite; 56, free lobe of 
eighth tergite; 57, foreleg of male. Fics. 58 to 64. M. carinata. 58, left paramere; 59-60, — 
right paramere; 61, prestrigilar flap; 62, submedian process of seventh abdominal sternite; — 
63, free lobe of eighth tergite; 64, foreleg of male. 


51, foreleg of male. 


A Reviston oF Micronecta or AUSTRALIA AND MELANESIA 155 


scutellum, somewhat straight. Hemelytra shiny and smooth, with numerous 
hairs. Prestrigilar flap broad and short (fig. 41). Submedian process of 
seventh abdominal sternite well produced, with four to six enlarged bristles 
subbasally (fig. 42). Free lobe of eighth tergite (fig. 43) as in M. major. 
Right paramere (fig. 40) distinctly dilated just before tip and suddenly con- 
stricted distally; left paramere (figs. 37-39) somewhat straight at middle, 
bent toward poined apex. Femur of foreleg (fig. 44) with two spinelike 
setae on lower side and three apical spinelike hairs near outer margin; tibia 
with large apical spine on inner margin and subapical spine on flexor mar- 
gin; palar claw greatly dilated near base. 

Remarks. This redescription is based on eight specimens from Murray 
Bridge, South Australia, which were determined by Dr. Hale, and on the 
syntypical specimens which I have also examined. In addition, more than 
150 specimens of this species from various places were studied. The large 
body size and the shape of the free lobe of the eighth tergite are similar to 
those of M. major but they differ in the relative sizes of the interocular space 
and the eye, the shape of the palar claw, and the genitalia. 

Collection data. South Australia: Adelaide (H. M. Hale); Murray 
Bridge. Western Australia: Seaforth, June, 1952 (Mrs. B. Y. Main); Rott- 
nest Island, near Perth, Oct. (P. J. Darlington, Harvard Exp.); Mt. Sterling, 
macy 19, 1957 (J. A. L. Watson); Bickley Swamp, Sept. 29, 1954 (E. P. 
Hodgkin). New South Wales: Carbramatta, Sydney, Feb., 1958 (M. J. 
Nikitin); Pine Island, near Canberra, Fed. Cap. Terr. (J. W. Evans); Val- 
ley Heights, near Katoomba, May 23, 1954. Tasmania: Hobart, April, 1937 
(J. W. Evans). 


Micronecta adelaidae nu. sp. 


Size. Macropterous form, length 2.5 mm. to 2.7 mm. 

Color. General coloration dark; head yellow with three contrasting paral- 
lel, longitudinal, brown stripes on vertex, one median, other two close to 
inner margin of each eye; pronotum, scutellum and hemelytra brown, with 
pattern of hemelytra as in M. batilla but broken stripes on corium more pro- 
nounced. Venter and legs yellowish gray. 

Structural characteristics. Head shorter than pronotum, vertex slightly 
produced in front of eyes; posterior margin of eyes approximate posterior 
margin of head; interocular space about one and one-thirds times as wide as 
an eye; length to width of pronotal disk as 1.1:2.8; pronotum spindle shaped, 
with anterior and posterior margins rounded. Hemelytra shiny and smooth, 
with sparsely scattered, fine, short hairs. Prestrigilar flag as in Fig. 48. Sub- 
median process of seventh sternite (fig. 49) moderately produced, apex blunt. 
Free lobe of eighth tergite as in Fig. 50. Right paramere (fig. 47) evenly 


156 Tue University SciENCE BULLETIN 


curved and pointed apically; left paramere (figs. 45-46) smooth with knob- 
like expanded apex. Foreleg (fig. 51) as in M. carinata. 

Comparative notes. In general appearance, this species is similar to M. 
gueenslandica, but M. adelaidae has a narrower interocular space. The well 
sclerotized left paramere of M. adelaidae is similar to those of M. wind: and 
M. hale. 

Holotype. Male, Adelaide River 70 mi. South of Darwin, Northern Ter- 
ritory, Mar. 25, 1954 (B. Malkin) in the United States National Museum. 

Allotype. Female, same data as for holotype. 

Paratypes. One female and two males with same data as above, one male 
in the United States National Museum, and one male and one female in the 
Snow Entomological Museum of The University of Kansas. 


Micronecta australiensts 0. sp. 


Size. Brachypterous form, length 2.03 mm. to 2.33 mm. 

Color. Light to dark brown, pattern partially or completely effaced; 
when present, masculations as in M. batilla. Head, legs and thoracic venter 
usually pale; abdominal venter usually black. 

Structural characteristics. Head longer than pronotum, vertex rounded 
in front of eyes; interocular space wider than eye (3.1:2.2); width to length 
of pronotal disk as 6.5:1.4; posterior margin of pronotum almost straight; 
pronotum, scutellum and hemelytra slightly rugulose, hemelytra with. scat- 
tered pale hairs. Wings reaching sixth abdominal tergum. Prestrigilar flap 
(fig. 54) and submedian process of seventh sternite (fig. 55) both similar in 
outline of those of M. micra. Free lobe of eighth tergite (fig. 56) with 
rounded, setigerous inner angle, rectangular outer angle, posterior margin 
between angles almost straight. Right paramere (fig. 53) simply curved to 
blunt point; left paramere (fig. 52) styliform, unevenly narrowed toward 
blunt point. Foreleg as in Fig. 57. 

Camparative notes. This species is similar in general appearance and 
structure to M. micra, but differs in the body size, the relative sizes of the 
interocular space and the eye, and the genitalia. 

Holotype. Male, Federal Capital Territory (near Canberra), Australia, 
Mar. 25, 1931 (J. Evans) in the Snow Entomological Museum of The Uni- 
versity of Kansas. 

Allotype. Female, same data as for holotype. 

Paratypes. Nineteen males and twenty-two females, same data as holo- 
type; one male, Alexandra, Victoria, F. L. Billinghurst, all in the Snow 
Entomological Museum of The University of Kansas, except three males 


and three females in the South Australian Museum. 


A Revision oF Micronecta or AUSTRALIA AND MELANESIA 157 


Micronecta carinata n. sp. 


Size. Brachypterous form, length 2.5 mm. to 2.6 mm. 

Color. Ground color of dorsum yellow to light brown, with dark brown 
maculations; head with a central longitudinal stripe on vertex; pronotum 
without maculations; clavus with irregular dark blotches basally and distal- 
ly; corium with four longitudinal broken stripes, tending to fuse near center. 
Venter and legs dull gray. 

Structural characteristics. Head longer than pronotum, vertex roundly 
produced in front of eyes; pronotum with noticeable ridge, along hind mar- 
gin of head; interocular space narrow but slightly wider than eye "(2:8297)- 
width to length of pronotal disk as 7.2:1.9; posterior margin of pronotum 
somewhat straight or hardly curved. Hemelytra with relatively long and 
dense hairs. Wings reaching seventh abdominal tergum. Prestrigilar flap 
(fig. 61) about twice as wide as long. Submedian process of seventh sternite 
(fig. 62) broad basally and gradually narrowing to short, acute point, with- 
out enlarged bristles subbasally. Free lobe of eighth tergite (fig. 63) with 
distinct, round inner angle, rectangular outer angle with numerous long 
bristles uniformly arranged along lower exterior margin. Right and left 
parameres broadly styliform and somewhat twisted (figs. 58-60). Foreleg as 
in Fig. 64. 

Comparative notes. M. carinata differs from other Australian species in 
having four central, fused, broken, longitudinal stripes on the hemelytra, 
and in the broadly styliform parameres of the male. These distinct charac- 
teristics indicate that M. carinata is a specialized species. 

Holotype. Male, Dorrigo, New South Wales, Australia (J. P. Darling- 
ton) in the Museum of Comparative Zoology (Harvard University). 

Allotype. Female, same data as holotype. 

Paratype. One female, same data as holotype, in the Snow Entomologi- 
cal Museum of The University of Kansas. 


Micronecta halet un. sp. 


Size. Macropterous form, length 3.0 mm. to 3.3 mm. 

Color. Dorsum generally yellowish brown, usually without color pattern, 
or with one transverse band on pronotum and four broken longitudinal 
stripes on corium. Venter and legs lighter than dorsum. 

Structural characteristics. Body elongate, length more than twice greatest 
width (5.0:2.2); head shorter than pronotum, vertex roundly produced be- 
yond eye margins; interocular space slightly wider than eye (3.5:3.2); width 
to length of pronotal disk as 8.6:3.3; pronotum spindle shaped. Hemelytra 
with sparsely scattered fine hairs. Prestrigilar flag as in Fig. 69. Submedian 
process of seventh abdominal sternite (fig. 70) moderately produced, apex 


158 THe University SciENcE BULLETIN 


Fics. 65 ro 72. M. halei. 65-57, left paramere; 68, right paramere; 69, prestrigilar flap; 
7(), submedian process of seventh abdominal sternite; 71, free lobe of eighth tergite; 72, 
foreleg of male. Fics. 73 to 79. M. queenslandica. 73-74, left paramere; 75, right paramere; 


A Revision oF Micronecta or AUSTRALIA AND MELANESIA 159 


subacute, with four enlarged bristles subbasally. Free lobe of eighth tergite 
(fig. 71) with rounded inner angle and rather prominent setigerous outer 
angle. Right paramere (fig. 68) simply curved and pointed; left paramere 
(figs. 65-67) without denticulations, broad basally and twisted slightly to- 
ward narrow distal part. Palar claw (fig. 72) elongate and expanded apically. 

Comparative notes. In general appearance, M. halei has no specific diag- 
nostic feature. However, the sclerotized paramere and the shape of the sub- 
median process of the seventh abdominal sternite indicate a close alliance to 
M. windi, from which M. hale differs in the twisted distal end of the left 
paramere and the expanded tip of the palar claw. 

Holotype. Male, De Grey at Yarrie Station, Western Australia, July 10, 
1953 (N. B. Tindale) in the South Australian Museum. 

Allotype. Female, same data as holotype. 

Paratypes. Three males and twenty females, same data as holotype. One 
male and twelve females in the South Australian Museum and two males 
and eight females in the Snow Entomological Museum of The University 
of Kansas. 


Micronecta queenslandica n. sp. 


Size. Macropterous form, length 2.6 mm. to 3.0 mm. 

Color. Grayish brown above and pale yellow below. Ground color of 
vertex mostly grayish yellow contrasting with three parallel longitudinal 
reddish brown stripes arranged as in M. batilla; pattern on pronotum and 
hemelytra as in M. robusta but lighter. Venter of thorax and abdomen yel- 
lowish, slightly suffused with gray. Legs mostly yellow, darker on tarsi. 

Structural characteristics. Head shorter than pronotum; vertex rounded 
beyond anterior margins of eyes; interocular space slightly wider than eye 
(3.5:3.2); width to length of pronotal disk as 8.7:3.0; posterior margin of 
pronotum somewhat straight in front of scutellum. Hemelytra shiny and 
smooth, with numerous fine, rather long hairs. Prestrigilar flap (fig. 76) 
narrow and elongate, unlike those of most other Australian species. Sub- 
median process of seventh sternite (fig. 77) slightly prolonged, without en- 
larged bristles subbasally. Free lobe of eighth tergite (fig. 78) with inner 
angle obsolete, outer angle distinctly prolonged and completely margined 
with bristles. Right paramere (fig. 75) arcuate with pointed apex; left para- 
mere (figs. 73-74) somewhat straight, denticulate with a few conical barbs, 
ventral margin of distal part with shallow depression, rounded apically. 
Palar claw with distinct notch (fig. 79). 


76, prestrigilar flap; 77, submedian process of seventh abdominal sternite; 78, free lobe of 
eighth tergite; 79, foreleg of male. Fics. 80 ro 86. M. wind:. 80-81, left paramere; 82, right 
paramere; 83, prestrigilar flap; 84, submedian process of seventh abdominal sternite; 85, free 
lobe of eighth tergite; 86, foreleg of male. 


160 THe University SciENCE BULLETIN 


Comparative notes. The unusually shaped palar claw and the prominent 
prestrigilar flap as well as the remarkable free lobe of the eighth tergite of 
the male distinguish this species. The denticulate left paramere seems to 
indicate a close alliance to M. virgata, although M. queenslandica has the 
general facies of M. adelaidae. 

Holotype. Male, northern Queensland, Australia, Marshall Laird, June, 
1954, in the Snow Entomological Museum of The University of Kansas. 

Allotype. Female, same data as holotype. 

Paratypes. Five males and three females, same data as holotype; one male 
and three females, Townsville, northern Queensland, Australia, 1920 (G. H. 
Hill), in the Snow Entomological Museum of The University of Kansas 
except one male and two females in the South Australian Museum. 


Micronecta major n. sp. 


Size. Macropterous form, length 4.6 mm. Brachypterous form, length 
4 mm. 

Color. Medium to light brown, lineations on hemelytra as in M. robusta 
but more distinct. Venter and legs uniformly yellowish. 

Structural characteristics. In brachypterous form, head longer than pro- 
notum (1.4:1.2), width to length of pronotal disk as 3.7:1.2. In macropterous 
form, head as long as pronotum, width to length of pronotal disk as 4.2:1.5, 
Vertex strongly produced beyond anterior margins of eyes; interocular space 
as wide as eye or slightly wider; pronotum rounded at posterior margin. 
Hemelytra shiny with numerous fine hairs. Wings reaching eighth abdomi- 
nal tergite in brachypterous form. Prestrigilar flap (fig. 33) broad and short, 
with rounded apex. Submedian process of seventh sternite (fig. 34) pro- 
longer to blunt point, without enlarged bristles subbasally. Free lobe of 
eighth tergite as in Fig. 35. Right paramere (fig. 32) evenly curved and 
rounded apically; left paramere stout, slightly curved and denticulate with 
many conical barbs (figs. 30-31). Femur of foreleg (hg. 36) with two spine- 
like setae on lower side, palar claw elongate. 

Comparative notes. The general appearance of this species and, particu- 
larly, the distinctly produced vertex and the relative sizes of the interocular 
space and the eye, are similar to those of M. barilla, but M. major differs from 
M. batilla in the parameres, which have a stouter shaft and a rounded distal 
end, and in the larger body size. 

Holotype. Male, Coolabah, New South Wales, Nov. 15, 1900 (W. Frog- 
gatt) in the Hungarian National Museum, Budapest. 

Allotype. Female, same data as for holotype. 

Paratypes. Four females, two in the Hungarian National Museum, Buda- 
pest and two in the Snow Entomological Museum of The University of 
Kansas. 


A Revision oF Micronecta or AUSTRALIA AND MELANESIA 161 


Micronecta wind n. sp. 


Size. Macropterous form, length 2.9 mm. 

Color. Yellowish brown throughout, pattern completely faded. 

Structural characteristics. Head about as long as pronotum, vertex tri- 
angularly produced in front of eyes; interocular space wider than eye 
(3.8:2.9); pronotum spindle shaped; width to length of pronotal disk as 
8.0:3.0. Hemelytra shiny and smooth, with sparsely scattered hairs. Prestrigi- 
lar flap as in Fig. 83. Submedian process of seventh sternite (fig. 84) notice- 
ably prolonged, with four enlarged bristles subbasally. Free lobe of eighth 
tergite( fig. 85) with well developed inner rounded angle and pointed se- 
tigerous outer angle, posterior margin between angles concave. Right para- 
mere (fig. 82) arcuate, gradually narrowing toward distal point; left para- 
mere (figs. 80-81) slightly curved, free portion emarginate ventrally, distal 
margin somewhat folded. Palar claw elongate and slightly expanded apical- 
ly (fig. 86). 

Comparative notes. The color pattern of the two available specimens is 
completely effaced. M. windi has the general facies of M. gracilis except the 
former is much smaller. M. windi is also similar to M. Aalez; the chief dif- 
ferences are cited under that species. 

Holotype. Male, Kuranda, northern Queensland, Australia, Aug. 12, 
1938 (R. G. Wind) in the Snow Entomological Museum of The University 
of Kansas. | 

Paratype. One male, same data as holotype. 


Micronecta sinistra n. sp. 


Size. Macropterous form, length 1.7 mm. 

Color. Medium brown, head yellowish, pronotum and hemelytra darker; 
pattern almost completely effaced except for one dark stripe running along 
outer margin of corium. Venter blackish. 

Structural characteristics. Head about as long as pronotum; vertex mod- 
erately produced in front; interocular space wider than eye (0.8:0.7). Heme- 
lytra smooth and shiny, with scattered very fine hairs, almost invisible. 
Abdomen with usual asymmetry reversed, having strigil on the left. Sub- 
marginal bristles on left side of fifth abdominal tergite absent (as in most 
species of minutissima group), tergite and prestrigilar flap as in Fig. 12. Sub- 
median process of seventh sternite (fig. 13) broad basally and pointed dis- 
tally, with four slightly enlarged bristles subbasally. Free lobe of the right, 
eighth tergite (fig. 14) different from that of dextral species, having two 
outer, lower and upper, setigerous angles and one inner rounded angle. 
Right and left parameres (figs. 8-11) similar, free portions curved and well 
chininized with extremely expanded distal lobe. Foreleg as in Fig. 15. 


162 Tue University Science BULLETIN 


Comparative notes. The single specimen has sinistral abdominal asym- 
metry, which deviates from other species of Micronecta, but has been re- 
corded once before by Wrdblewski (1962) who described a sinistral male 
individual of M. quadristrigata Breddin from VietNam. M. sinistra might 
represent an independent offshoot from the original stock of Micronecta be- 
cause the structural characteristics of the male parameres and the free lobe 
of the eighth tergite show many differences from other taxa. 

Holotype. Male, Astrolabe Bay, Stephansort, New Guinea (Biré 97), in 
the Hungarian National Museum, Budapest. 


NOTES ON ADDITIONAL SPECIES 


Micronecta annae Kirkaldy 
Micronecta annae Kirkatpy, 1905, p. 262. 


On the basis of the original description, copied below, it is impossible to 
recognize this species: 

“M. annae sp. n—Head pallid. Pronotum dark fuscous brown, with 
darker transverse median line. Tegmina fuscuous brown (the margins of 
the areas narrowly darker), somewhat superficially punctured. Head a little 
longer than pronotum, rounded in front. Pronotum elongate ellipitical, 
lateral margins very short, much less than half the width of the posterior 
margin of an eye. Mesoxyphus acutely triangular. Terminal segment of 
antenna elongate, somewhat thickened. Intermediate femur equal in length 
to tibia, tarsus and claw together; tarsus one-half longer than a claw, which 
is equal in length to the tibia, subcostal furrow much as in M. vanduzeet. 
Length 34% mm. 

“Australia, Victoria (my collection).” 


Micronecta annae Kirkaldy var. Pallida Kirkaldy 
Micronecta annae KirKkaupy var. pallida Kirxacpy, 1907, p. 788. 


Recognition of a variety of any Micronecta species on the basis of colora- 
tion alone, without discussion of the individual variation within the species, 
seems unacceptable. The following is a copy of the original description: 

“M. annae Kirkaldy var. pallida nov. 

No transverse line on pronotum; tegmina with a pale castereous basal 
band. 


ce a 7 P *; = ” 
Hab.—Q.: Kuranda (Aug.;Perkins). 


Micronecta erato Kirkaldy 
Micronecta erato Kirkarpy, 1905, p. 263. 


The original description of this species is mostly based on coloration and 
almost agrees with all the Australian Micronecta. Kirkaldy did give certain 


A Revision oF Micronecta or AUSTRALIA AND MELANESIA 163 


structural information on M. erato (head, pronotum and body size) which 
suggests M. datilla, but this synonymy is still far from established. The fol- 
lowing is a copy of the original description: 

“M. erato sp. n—Head and underside pale stramineous. Pronotum pale 
sordid yellow, with a broad blackish brown median transverse stripe which 
does not reach the lateral margins. Tegmina sordid stramineous; clavus 
with two narrow dark brown lines running parallel to interior and corial 
margins, uniting at the apex of clavus. Corium with two elongate suboval 
areas narrowly dark-brown-bordered, and the exterior lateral margins also 
brownish black. Pronotum, scutellum and tegmina somewhat superficially 
punctured. Head rounded in front, longer than the pronotum; lateral mar- 
gins of pronotum obsolescent; membrane apically angulate. Length about 
3 mm. 

“Australia, Victoria (my collection).” 


Micronecta carbonaria Horvath 


Micronecta carbonaria Horvatn, 1904, p. 595; Wrostewskt, 1962b, p. 319. 


Among the materials lent from the National Hungarian Museum, Buda- 
pest, is a specimen of M. carbonaria bearing a type label beside the collection 
label (Berlinhafen, Lemien, New Guinea, 1896 (Bird), although Horvath 
did not designate a type specimen in his original description. Since Wro- 
blewski recently (1962b) has given a complete description of this species 
based on the materials from the type series, redescription is not necessary 
here. Because of the similarities of the parameres, free lobe of eighth tergite, 
and the palar claw of the foreleg, M. carbonaria, M. virgata and M. batilla 
seem to be closely related species. 

Collection data. Berlinhafen, Lemien, New Guinea, 1896 (Bird) Type; 
Sepik River, New Guinea (K. P. Schmidt). 


Micronecta ludibunda Breddin 


Micronecta ludibunda Breppin, 1905, p. 57; Cuen, 1960b, p. 115; Wrdsiewskt, 
1962b, p. 323. 

Micronecta inconspicua LunpsLap, 1933, p. 96. 

Micronecta striatella LuNpBLAD, 1933, p. 98. 

Micronecta graphiptera Horvatn, 1918, p. 146. 


One female type of M. graphiptera from the Hungarian National Mu- 
seum, Budapest, is available for this study; it confirms Wrodblewski’s state- 
ments (1962b) that M. graphiptera is a synonym of M. ludibunda. 

M. ludibunda is a widely spread species, known from Thailand, Sumatra, 
New Guinea and the Solomon Islands. 


164 Tue Universiry SciENCE BULLETIN 


Micronecta quadristrigata Breddin 
Micronecta quadristrigata Breppin, 1905, p. 57; Lunpsvap, 1933, p. 87; Hurcnin- 
son, 1940, p. 376; WrésLewsk1, 1960, p. 301; WrosLewskt, 1962a, p. 176. 
Micronecta minthe Distant, 1910, p. 347. 


Two specimens, one male and one female, labelled Astrolabe Bay, Steph- 
ansort, New Guinea (Biré 97) from the Hungarian National Museum be- 
long to this species. The male is 2.4 mm. and female 2.5 mm. long; most 
other features of these two specimens agree well with the previous descrip- 
tions by Lundblad, Hutchinson, and Wroblewski except for the relative 
widths of the interocular space and the eye. The ratio of the interocular 
space to an eye, in the New Guinea specimens, is 1.03, smaller than the same 
ratio for specimens from Hong Kong and Celebes (Wroblewski, 1960). 

Micronecta quadristrigata, because of many unusual morphological struc- 
tures, was considered a primitive species by Wroblewski, 1960. In 1962, he 
found one aberrant male from Viet-Nam that showed sinistral abdominal 
asymmetry. In my opinion, this is additional evidence of the primitiveness 
of this species. Another species from New Guinea, M. sinistra, also has the 
abdomen sinistrally asymmetrical; at the same time it has many primitive 
structures such as the undifferentiated, enlarged bristles on the seventh 
abdominal sternite (fig. 13) and parameres (figs. 8-11) as in some Tenagobia 
species (Deay, 1935). The left and right parameres of M. sinistra both have 
a narrow median shaft and a large expanded lobelike distal end which is 
almost as large as the basal end, while most other Micronecta species have a 
stout shaft, narrower or pointed distal end, and a broad base. It is possible 
that dextral and sinistral asymmetrical forms were both common in the early 
members of this genus. Through the long course of the evolution the most 
favored form, in this case the dextral, has become dominant. Since only two 
sinistrally asymmetrical specimens are known (the type of M. simistra and 
the aberrant M. guadristrigata), they may be phylogenetic relics. 

In 1962, Wrdblewski placed M. quadristrigata in a new monotypic sub- 
genus Sigmonecta, based on the sigmoid shape of the free lobe of eighth 
tergite. However, M. queenslandica of Australia and M. eupompe Hutchin- 
son (1930) of Abyssinia also have such a characteristically shaped free lobe. 
Yet there is no other structure which shows their affinity. The left parameres 
of the three species (often used for indicating phyletic relationships among 
species of Micronecta) are strongly different. It seems doubtful that the 
sigmoid free lobe of eighth tergite alone can have much value for establish- 
ing a subgenus. 


Micronecta virgata Hale 
Micronecta virgata Hare, 1922, p. 327; Wrdsiewsk1, 1962b, p. 320. 


This species is very easy to distinguish from other Australian taxa since it 


A Revision oF Micronecta oF AUSTRALIA AND MELANESIA 165 


has the largest ratio of the interocular space to the eye width and the longest 
middle femur of all the known Australian species. After comparing the 
female syntype from the South Australian Museum with other specimens 
listed in the following paragraph, I found no special difference between 
them. In 1962, Wroblewski redescribed this species from specimens from the 
Solomon Islands; therefore, a redescription is omitted here. 

Collection data. Queensland: Carins district (A. M. Lea) (syntype); 
Townsville, 1920 (G. F. Hill); Townsville, March, 1932 (P. J. Darlington, 
Harvard Expedition). Solomon Islands: Russell Island, Ufa, Sept. 1955 
(E. S. Brown); San Cristobal, Ugi, April, 1955 (E. S. Brown). 


BIBLIOGRAPHY 


Breppin, G. W. 1905. Uebersicht der javanischen Micronecta-Arten (Rhynchota). Soc. Ent. 
[Zurich], 20:57. 

CHEN, L. C. 1960. A study of the genus Micronecta of India, Japan, Taiwan and adjacent 
Regions. Jour. Kansas Ent. Soc., 33:99-118. 

Deay, H. O. 1935. The genus Tenagobia Bergroth (Corixidae, Hemiptera). Univ. Kansas Sci. 
Bull., 22:403-477. 

Distant, W. L. 1910. The Fauna of British India, Rhynchota, 5. London. 

Hare, H. M. 1922. Studies in Australian aquatic Hemiptera. Rec. South Australian Mus., 
2:309-330. 

HorvatH, G. 1904. Hydrovorisae tres novae. Ann. Mus. Nat. Hungarici, 2:594-595, 

. 1918. De Hydrocorisis nonnullis extraeuropaeis. Ann. Mus. Nat. Hungarici, 16:140-146. 

Hurcuinson, G. E. 1930. Report on Notonectidae, Pleidae, and Corixidae (Hemiptera), Mr. 

Omer-Cooper’s investigation of the Abyssinian fresh water. (Dr. Hugh Scott’s expedi- 

tion.) Proc. Zool. Soc., London, (29)437-466. 

. 1940. A revision of the Corixidae of India and adjacent regions. Trans. Connecticut 

Acad. Arts. Sci., 33:339-476. 

KirKatpy, G. W. 1905. Five new species of Micronecta Kirkaldy. Ent. News, 16:260-263. 

. 1907. Memoir on a few Heteropterous Hemiptera from eastern Australia. Proc. Linn. 

Soc. New South Wales, 17:768-786. 

LunpgLap, O. 1933. Zur Kenntnis der aquatilen und semiaquatilen Hemipteren von Sumatra, 
Java und Bali. Arch. Hydrobiol., Suppl. to Bd. 12:1-489. 

Wroscewskl, A. 1960. Notes on some Asiatic species of the genus Micronecta Kirk. (Hemi- 

ptera, Corixidae.) I, Ann. Zool. [Warszawa], 18:301-332. 

. 1962a. Notes on Muicronecta from Viet-Nam (Heteroptera, Corixidae). Bull. Acad. 

Polon. Sci., 10:175-180. 

. 1962b. Notes on Micronectinae from Melanesia (Heteroptera, Corixidae). Bull. Acad. 

Polon. Sci., 10:319-324. 


a | a 
P Ce th 
be : ’ we oS = ere 


se? ae) fue 


ale © Ym 
a ee ae ee 


Ds ae an.29 18Ue 
wet ince = oad >, Tie 
& ,)e-y)" )aall 4e8, 
> ae a meth 
<r 
5 
, 
a 
ae 
—— 
or 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


MICROGEOGRAPHIC VARIATION AND 
COVARIATION IN 
PEMPHIGUS POPULI-TRANSVERSUS 


By 
Richard C. Rinkel 


| rn ee 
ia. 
‘dvine Biological ie ae ay 
= iGry 


VoL. XLVI Paces 167-200 Ocroser 28, 1965 No. 4 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Vou, XLVI Paces 167-200 Ocroser 28, 1965 No. 4 


Microgeographic Variation and Covariation in 


Pemphigus populi-transversus’” 
By 


Ricuarp C. RinKeEv® 
Department of Entomology, The University of Kansas 


ABS URAC T 


This study is a quantitative analysis of the microgeographic variation and 
covariation of 24 characters of two forms of the gall forming aphid, Pemphigus 
populi-transversus. The variation and covariation are separately studied at three 
levels: within galls, among galls within localities, and among localities. Covaria- 
tion is studied by means of correlations among the characters and restated in 
terms of fewer dimensions (factors) by means of factor analysis. The analysis is 
based on 15 localities of 15 galls each, from Douglas and Leavenworth counties, 
Kansas. 

The two central problems considered are (1) the comparison of the amounts 
of variation present in this study with corresponding amounts present in previous 
studies, and (2) a consideration of character covariation within a limited geo- 
graphic area compared with character covariation over a much larger area. 

To determine the amounts of variation present in this study for the characters 
involved, 24 separate hierarchic analyses of variance were calculated for each of 
the three levels of variation. It was found that variation among localities forms a 
much smaller portion of the total variance than in the previous studies, which 


‘Contribution No. 1202 from the Department of Entomology, The University of Kansas, 
Lawrence. This paper is part of a thesis presented in partial fulfilment of the requirements for 
the M.A. degree in Entomology. 

*Part 4 of a study of variation in the aphid genus Pemphigus. Parts 1 to 3, respectively, are 
the following publications: Sokal (1952), Sokal (1962), Sokal and Rinkel (1963); part 5 is 
Sokal and Thomas (1965). 

* The author wishes to express his gratitude to Dr. Robert R. Sokal, Professor of Statistical 
Biology, without whose assistance this study would have been impossible. Thanks are also duc 
Drs. Charles D. Michener and George W. Byers for their appropriate and helpful comments on 
the manuscript. Mrs. Betty Lou Deffley and Mrs. Maxine L. Howe assisted in the preparation 
of the material for study and in the preparation of the data for analysis. The University of 
Kansas Computation Center furnished computer time to carry out the numerous computations 
of this study. 


168 Tue University ScIENCE BULLETIN 


involved a much larger geographic area. Variation among galls within localities 
and variation within galls were found to be of the same general magnitude in 
the present study as before. 

In order to study character covariation, matrices of correlation coelhicients 
among the 24 characters for each level of variation were computed. These were 
compared with the corresponding matrices from the previous studies, and factor 
analyses were performed in an attempt to reduce the complexity of the covariation 
patterns. 

At the within-galllevel of variation, the wing, head, and thorax variables 
seem to be influenced by the same factor as in the previous studies. The remain- 
ing veriables appear to be influenced by somewhat different factors in the present 
study. At the intergall level of variation, the patterns of covariation consist of 
general size factors for each of the two forms studied. At the inter-locality level 
of variation the factors extracted bear little resemblance to the factors at this level 
in the previous studies. This would seem to indicate that covariational patterns at 
the interlocality level in the present study are almost entirely determined by local 
influences. 

In an attempt to describe geographic variation over the area studied, maps of 
locality means for each cf the characters were plotted. A multiple range test was 
employed on these means to find groups of significantly different locality means. 
These tests, as expected from the above results, reveal less differentiation among 
localities in the present study than in the study for eastern North America. 


INTRODUCTION 


This study is a quantitative analysis of microgeographic variation and 
covariation of 24 characters of two forms of the gall-forming aphid, Pem- 
phigus populi-transversus Riley. It will form part four of a general study of 
the geographic variation and covariation of morphological characters of 
this organism. 

Part one of this general study (Sokal, 1952) dealt with variation and 
covariation within a local population at Ripley, Ohio. The 21 characters in 
this initial study were analyzed with a view, to selecting those most useful 
for a study of intraspecific variation. The characters so analyzed were 
studied singly by means of the analysis of variance and in combination by 
means of the analysis of covariance and partitioning of the correlation co- 
efficients. 

The second part of the general study (Sokal, 1962) dealt principally with 
a quantitative assessment of the variation and covariation of 18 of the above 
characters of the alate forms over eastern North America. The analysis of 
covariance was abandoned in favor of factor analysis as a means of sum- 
marizing patterns of covariation among the characters. 

The third part of the general study (Sokal and Rinkel, 1963) was pri- 
marily concerned with the description of the geographic variation of the 


MicroceocraPHic VARIATION AND CovaRIATION 169 


characters studied. Meaningful patterns of variation of the 18 alate charac- 
ters studied over eastern North America were sought, and various techniques 
were used in an attempt to clarify the variational trends. The use of multiple 
range tests (such as the Student-Newman-Keuls test) on locality means of 
the various characters was introduced. 

This study is based on 15 localities from northeastern Douglas County 
and western Leavenworth County, Kansas. Two main problems are in- 
vestigated. The first is a comparison of the amounts of variation present at 
the various levels of this study with those of previous studies. Consideration 
of this problem should lead to a better understanding of amounts of varia- 
tion in populations of the organism in areas of various sizes. Allied with a 
comparison of the amounts of variation present is a comparison of the locality 
means of the present study with those of the larger studies (Sokal and Rin- 
kel, 1962; Sokal and Thomas, 1963). This should give added information 
about the patterns of geographic variation of characters in this species. 

The second problem is a comparison of character covariation of a limited 
area with a much larger area. Patterns of covariation at various levels among 
the characters studied are likely to be somewhat different within a restricted 
geographic area, as indeed proved to be the case. 

Covariation among the stem mother characters, and among stem mother 
and alate characters combined, are here examined at the interlocality level for 
the first time. 

The variation and covariation of the 24 characters are separately studied 
at three levels of variation; within galls, among galls within localities, and 
among localities for the alate characters; among galls within localities and 
among localities for the stem mother characters. Covariation is studied by 
means of correlations among the characters and restated in terms of fewer 
dimensions (factors) by means of factor analysis. Results of the various 
analyses performed are compared with results from the previous studies. 


MATERIALS AND METHODS 


The biological material used for this study consisted of both alate funda- 
trigeniae and apterous fundatrices or stem mothers of P. populi-transversus. 
The fundatrigeniae are the parthenogenetic offspring of the stem mothers, 
and both occur in the petiole gall formed around the stem mother on the cot- 
tonwood, Populus deltoides, in the spring. The stem mother hatches from an 
overwintering egg. The terms alate and stem mother, whenever used below, 
refer to the fundatrigeniae and the fundatrices, respectively. 

The localities from Douglas County and Leavenworth County, Kansas, 
were chosen because they were sources of 15 or more galls containing a stem 
mother and at least two alates each. Where more than 15 such galls were 


170 Tue University ScIENCE BULLETIN 


available at any one locality, 15 were selected at random from those col- 
lected. The localities usually consisted of two or sometimes three cotton- 
woods. At locality number 15, three samples of 15 galls each were taken 
from each of three trees within one locality. 

Localities within the study were chosen to give as comprehensive cover- 
age of the area as was possible within the ecological limitations of the host 
plant. A list of localities is given in Table 1. 

All material for this study was collected during September, 1961. Galls 
were snipped from the cottonwoods and placed immediately into containers 
of 70 per cent ethyl alcohol. 

The galls were then sorted and the stem mothers and alates contained in 
each were removed and placed in separate, numbered vials. 

Both alates and stem mothers were prepared for measurement by clearing 
in successive solutions of dioxane, dioxane-xylene, and pure xylene. They 
were then mounted in balsam on microscope slides. If more than three alates 
per gall were available, three were chosen at random from among these and 
mounted. During measurement, two out of these three were chosen at ran- 
dom for actual measurement. 

The alate characters were the same as those used by Sokal (1962); the 
stem mother characters the same as those employed by Sokal (1952), except 
for total length and last rostral segment length, which had proved to be 
unreliable. Of the 18 alate characters measured, 12 were continuous and the 
last six were meristic. All six stem mother characters were continuous. 

In the interest of brevity, abbreviated names, such as antenna III for 
length of antennal segment HI, sensoria II for number of sensoria on anten- 
nal segment III, and islands V for number of sclerotic islands in the primary 
sensorium of antennal segment V, will be used below. 

All measurements were made by means of a Reichert microprojector 
fitted with a vernier scale to correct the difference between real and apparent 
length of any structure which did not lie level on a slide. Whenever such 
was the case, the difference in elevation of the two ends of the structure was 
measured. The true length of the structure was calculated as the hypothenuse 
of a triangle whose sides were the apparent length of the structure and the 
difference in elevation of the two ends of the structure, respectively. 


ANALYSES OF VARIANCE 


Twenty-four separate hierarchic analyses of variance were performed, 12 
for the continuous alate characters, six for the meristic alate characters in 


coded form, and six for the stem mother characters, which were all continu- 


ous. Coding of the meristic alate characters was accomplished by adding 0.5 
to each of the original values and taking the square root of this value. The 


MicrocEocraPHic VARIATION AND CovaARIATION 17h 


coding of the meristic variables was performed in an attempt to normalize 
their distribution. 

For the alate characters, the three levels of variation were between alates 
within galls, among galls within localities, and among localities, with 225, 
210, and 14 degrees of freedom, respectively. For the stem mother characters, 
the two levels of variation were among galls (individuals) within localities, 
and among localities, with 210 and 14 degrees of freedom, respectively. 

These analyses are interpreted here as model II analyses of variance. The 
localities can be thought of as being chosen at random from those available 
in the general area, and the aim here is to estimate the variance components 
at the various levels of variation. 

The results of these analyses are shown in Table 2 and 3, along with some 
of the results of Sokal (1962) and Sokal and Thomas (1965). 

The first two columns of Table 2 show the variances of individuals with- 
in galls for the 18 alate characters of both the present study and that of 
Sokal (1962). These figures represent an average within-gall variance over 
all localities in the two studies. Since, presumably, all individuals within 
any one gall are genetically identical, these variances are due entirely to en- 
vironmental influences. A comparison of these variances between the two 
columns reveals that they are of nearly equal magnitude except for the char- 
acters tibia length and sensoria V. A two-tailed F test of the ratios of the 
variances for each of these two characters showed no significant difference 
between them. 

The next two columns of Table 2 show the sum of the variance com- 
ponents within galls and among galls for the present study and a comparable 
quantity, the variances among galls within a locality found by Sokal and 
Thomas (1965). Only eight alate characters and five stem mother characters 
were used by Sokal and Thomas. Comparison of the two columns again 
shows nearly equal values. The one exception is for the stem mother charac- 
ter, tibia length. Here the ratio of the two variances shows significant differ- 
ence at the 0.01 level when tested. 

The above comparisons establish that with few exceptions the variances 
at the two lower levels of variation of the three studies are of the same order 
of magnitude. Comparisons of the variances at the interlocality level should 
therefore give a clear picture of the contrast in the amounts of variation 
present at this level in the various studies. 

The next four columns of Table 2 show the mean square of galls within a 
locality, the mean square of localities, and the significance levels for the F 
ratios calculated for the present study. All the alate continuous characters 
except head length and antenna V and VI show significant variation among 
galls and localities. Head length exhibits no added variance at all, antenna 
V shows significant variation among localities and antenna VI does so among 


172 Tue Universiry SciENcE BULLETIN 


galls. None of the meristic alate characters varies significantly among locali- 
ties and only one, islands VI, shows added variance among galls. Only two 
of the stem mother characters, head width and antenna III, show significant 
variance among localities. 

At locality 15, three trees within 50 feet of one another were selected, and 
15 galls, each containing a stem mother and at least two alates, were col- 
lected from each tree. Measurements were taken of the same 18 alate char- 
acters and 18 more analyses of variances were performed. 

None of the characters exhibited a significant added variance component 
among trees at the 1 per cent level, and only one (islands VI) showed signifi- 
cance at the 5 per cent level. None of the meristic variables showed an added 
variance component among galls within trees; however, the same continuous 
characters that proved significant at this level for the entire area under study 
are again significant. 

Table 3 contains various ratios of variance components from the four 
studies under consideration. The first four columns of this table contain the 
ratios of variance among galls within localities to variance among individuals 
within galls. The first column is derived from Sokal’s (1952) study of one 
locality near Ripley, Ohio. The second and third columns are from the 
present study and are the ratios for the main study and for the separate study 
of three trees, respectively. The fourth column gives the same statistic from 
Sokal’s (1962) study of 23 localities in eastern North America. The ratios 
from the present study are in general much closer to those of Sokal’s later 
study (1962) than to the earlier one (Sokal, 1952). Since it had been found 
that the galls of the Ripley, Ohio, locality were unusually well differentiated 
(Sokal, 1962), these results were to be expected and probably reflect the 
actual relationship between the two variance components. The fact that no 
variance components exist at the intergall level for several of the characters 
which did exhibit small components in the 1962 study, is probably due to the 
smaller sample sizes of the present study. The same phenomenon is ex- 
hibited in the ratios from the separate study of three trees. Since all of the 
ratios from both the present study and Sokal (1962) are near to or less than 
one, variance among galls within a locality, in general, accounts for less of 
the total variance than does variance among individuals within galls. 

The next two columns of Table 3 show the ratios of variance among 
localities to variance among individuals within galls for the present study 
and for that of Sokal (1962), respectively. It will be noted that in every case 
the ratios from this study are smaller than the corresponding ratios from 
Sokal. These results confirm the idea that a much smaller portion of the 
total variance is due to variation among localities in the present study than 
in the previous ones. Several of the characters studied, notably the meristic 
ones, show no added variance component at this level (among localities), 


MicroceocrRaPHic VARIATION AND CovARIATION 173 


while all the characters but one did so over eastern North America. These 
results are not surprising in view of the minute geographic area of the 
present study. The lack of variance components for several of the antennal 
sensoria characters tends to support the hypothesis that they are in general 
less variable than the continuous characters at the intergall and interlocality 
levels. 

The interlocality ratios just discussed could not be directly compared with 
the findings of Sokal and Thomas (1965), since the study of these authors 
was based on but one alate per gall and thus lacked intragall variance. The 
last three columns of Table 3 show the ratio of the variance component 
among localities divided by the sum of the variance components within and 
among galls for the present study and that of Sokal (1962), as well as a 
comparable ratio, the variance component of localities divided by the vari- 
ance within localities, for the data of Sokal and Thomas (1965). The ratios 
from the present study are again in all cases smaller than those from the 
studies of the wider geographical area. The columns of the present study 
and that of Sokal and Thomas (1965) also contain ratios for several stem 
mother characters. For the two stem mother characters (head width and 
tarsus length) which exhibit significant variation among localities, the ratios 
of variance among localities to variance among galls within localities are 
smaller than the comparable ratios obtained by Sokal and Thomas (1965). 
This again indicates that, for the limited area under study, variance among 
localities contributes only a small portion of the total variance. 


CORRELATIONS 


The cross-products for the 24 variables studied were obtained simultane- 
ously with the analyses of variance by matrix operations carried out by pro- 
grams written for the IBM 650 digital computer. These programs are writ- 
ten in the FORTRAN interpretive language and hence are available for use 
on almost all IBM and many other computers. 

The three levels of alate variation yielded three covariance matrices, in- 
tragall, intergall, and interlocality. The two levels of stem mother variation 
yielded two covariance matrices. In each case the covariances obtained cor- 
respond to the partitioning of the variance for that particular level. Covari- 
ances for the alate-stem mother character correlations were obtained by using 
gall sums of the alate characters and proceeding as above. 

Product-moment correlations were obtained from all of the above mat- 
rices in the usual manner; that is, by dividing each covariance by the geo- 
metric mean of the appropriate variances. Component correlations (Sokal, 
1962) were computed for the two higher levels of variation for the alate 
characters and for the interlocality level of the stem mother characters. 


174 Tue University SCIENCE BULLETIN 


Tue INTRAGALL CorRELATIONS 


The intragall correlation coefficients (for alate characters only) are shown 
in the half matrix below the diagonal in Table 4. In general these correla- 
tions are low. On the average they are lower than the intragall correlations 
found by Sokal (1962), but essentially the same pattern of significant 
(P < 0.01) coefficients obtains in the present study as in the previous one. 
Several differences will, however, be noted. As in the previous study, corre- 
lations among the meristic characters are practically nonexistent. The con- 
tinuous characters are significantly correlated, with several exceptions. Head 
length shows significant correlation with only three other characters. Anten- 
na VI shows significant correlation with only half of the other continuous 
variables and several of the other continuous characters are significantly cor- 
related with only one-half to two-thirds of the others. This pattern would 
seem to agree more closely with the results of Sokal (1952, table +) on a local 
population than with his study of eastern North America in which all the 
continuous characters were significantly correlated. This may in part be due 
to the fewer degrees of freedom employed in the present study. 

A cluster analysis of all correlations above 0.350 in the matrix reveals a 
major cluster of wing, head, thorax, and leg variables and also includes, 
rather strongly, antenna V. This cluster is similar to that obtained by Sokal 
(1962), for the corresponding matrix of his study. Antenna V was not, how- 
ever, included in that cluster, nor were other antennal variables. The cluster 
of highly correlated variables obtained in the present study at the intragall 
level seems to be similar to the intergall component cluster of Sokal (1962) 
in that it does include an antennal variable. 


THe INTERGALL CORRELATIONS 


The intergall product-moment correlations for both alate and stem 
mother characters are shown in the half matrix below the diagonal in Table 
5. For the alate characters, this matrix is in general quite similar to the 
corresponding matrix of Sokal (1962), both in the magnitude of the corre- 
lations and in the arrangement of the significant correlations within the 
matrix. Almost all the continuous alate variables are rather highly inter- 
correlated. Head length, however, is not as highly correlated with the other 
continuous characters as in Sokal’s (1962) intergall matrix and in general 
fewer of the correlations among the continuous variables are significant than 
in that matrix. This again may be due to the fact that fewer degrees of 
freedom were available in the present study. 

Almost without exception the meristic alate characters show no appre- 
ciable correlation, either among themselves or with the continuous charac- 
ters. Sensoria III, however, is significantly correlated with seven of the con- 


MicroGEOGRAPHIC VARIATION AND CovARIATION 175 


tinuous alate characters, and sensoria IV is significantly correlated with tar- 
sus length as is islands VI with thorax width. 

Most of the stem mother characters are significantly intercorrelated, with 
the exception of tibia length with tarsus length, and tibia length with 
antenna IV. 

None of the correlations between alate and stem mother characters proved 
to be significant at the intergall level. This confirms Sokal’s (1952) finding 
of no covariation of alate and stem mother characters within a locality. 

The intergall component correlation coefficients are shown in the half 
matrix above the diagonal in Table 6. Intergall component correlations could 
not be calculated for head length, antenna V, and all of the meristic charac- 
ters except islands VI, since these characters showed no added variance com- 
ponent at this level of variation. 

The component correlations shown are, in general, higher than the cor- 
responding product-moment correlations. All the applicable continuous 
characters are highly correlated and islands VI is significantly correlated 
with almost half the other characters in contrast to correlation with only one, 
in the case of the product-moment correlations. Thus by eliminating con- 
tributions from lower levels of variation by means of component correla- 
tions, we arrive at a clearer picture of the character relationships at this level. 

On comparing the intergall component matrix with the intragall matrix, 

we find that the major differences are that the correlations are uniformly 
higher and that islands VI shows a much closer relationship with some of 
the continuous variables. 
. A cluster analysis of all component correlations greater than 0.600 reveals 
a cluster of wing, head, thorax, leg, and antennal variables similar to the 
cluster derived for the intragall correlations of this study. However, the 
antennal variable involved here is antenna VI rather than antenna V. 

Cluster analysis of the corresponding matrix of Sokal (1962) reveals 
much the same cluster except for the antennal variables involved. Antenna 
IV is included in the cluster of the present study, while it and antenna V are 
not in the previous study. Since it was impossible to calculate component 
correlations for the variable antenna V in this study, it is not known whether 
it would have entered this cluster. Inspection of the product-moment corre- 


lations for this variable seem to indicate that it would have been included as 
in the cluster in Sokal (1962). 


Tue INTERLOCALITY CORRELATIONS 


The interlocality correlation coefficients are shown in Table 6 with the 
product-moment correlations below and the component correlations above 
the diagonal of the matrix. 


176 Tue University SCIENCE BULLETIN 


In comparing the alate interlocality product-moment correlations with 
the analogous matrices for the two lower levels of variation, a different pat- 
tern of relationships emerges. Although the significant correlations are 
almost without exception among the continuous variables, they are consider- 
ably fewer in number. In general the locality product-moment matrix of 
correlations among alate characters agrees rather closely with the correspond- 
ing matrix of Sokal (1962). 

There is only one significant product-moment correlation coefhcient be- 
tween alate ard stem mother characters. Sensoria VI and stem mother an- 
tenna IV are highly correlated. Unfortunately neither of these variables 
showed significant variation among localities and thus it is impossible to 
follow this trend in the component correlations. Several of the other alate 
variables show appreciable, if nonsignificant, correlations with stem mother 
characters and this covariational trend at the interlocality level emerges in the 
factor analysis of the interlocality component matrix. 

Among stem mother characters, head width, femur length, and antenna 
II] are highly correlated. 

In the matrix of locality component correlations head length, antenna VI, 
and all of the meristic alate characters except islands V, plus all of the stem 
mother characters except head width and antenna III, are missing because 
all of these variables lacked significant variance components at the inter- 
locality level. In this matrix the pattern evident in the product-moment cor- 
relations becomes more pronounced. Among the alate characters there is a 
cluster of wing, head, thorax, and leg variables and a connection between 
this cluster and the character antenna IV. These results differ from those of 
Sokal (1962) in that the antennal variables of that matrix formed a cluster 
that was separate from the cluster of thorax and leg variables. The antennal 
variables of the present study do not cluster with the head variables, but since 
antenna VI is missing in this study it is impossible to say if it would have 
been highly correlated with head length as was the case in the previous 
study. Wing length in this study is not independent of the leg and thorax 
characters; it was not included in the interlocality cluster of Sokal (1962). 

There are no significant component correlations of the alate characters 
and the two stem mother characters present in this matrix. Among the stem 
mother characters, head width and antenna III are highly correlated. 

Correlations within one locality were also computed among gall size, leaf 
size, stem mother head width, and antenna III, since it was suspected that 
there might be a relationship among these variables within a locality. The 
two stem mother characters that were chosen from the six available were 
those with the highest loadings on the general size factor for stem mother 
characters which was extracted from the gall component matrix. Rather 
surprisingly, none of these variables were significantly correlated at this level, 


MicroGEOGRAPHIC VARIATION AND CovaRIATION W/FE 


although stem mother size and gall size, and leaf size and gall size, had 
proved to be positively correlated on an interlocality basis in an earlier study 


(Sokal, unpublished results). 


BACTOR ANALYSES 


The use of factor analysis in the interpretation of matrices of correlations 
among characters at several levels of variation has been developed by Sokal 
(1962). 7 

The three correlation matrices (intragall, intergall component, and inter- 
locality component) subjected to factor analysis in this study were treated in 
the following manner: factors were first extracted from the matrices by 
Thurstone’s complete centroid method. The residual matrices were then 
tested for completeness of factor extraction by several criteria (see Sokal, 
1959). The factors were then re-extracted until estimates of the communali- 
ties stabilized. The resulting centroid factor matrices were rotated to simple 
structure by Sokal’s (1958) mass modification of Thurstone’s analytical 
method (MTAM). All of the above computational steps were performed on 
an IBM 650 digital computer with floating decimal arithmetic and indexing 
accumulators. 

The results of these analyses, the three simple structure matrices (intergall 
component, intragall, and interlocality component), are shown in Table 7. 
These matrices represent the correlations between each variable and each 
factor. 

In the intragall matrix, four factors were necessary to account for the ma- 
jor portion of the covariation, while in the intergall component and the inter- 
locality component matrices, three were sufficient. 

In the intragall simple structure matrix, factor I is a wing, head width, 
thorax, and femur length factor, which also affects tibia length and tarsus 
length to a lesser degree. This factor corresponds in general to factor I of Sokal 
(1962), except that tibia length and tarsus length are not so strongly affected 
in the present case. Factor II is an antennal factor affecting antenna III and 
less strongly, antenna IV, sensoria III, islands V, and islands VI. This factor 
seems to correspond to factor II of Sokal (1962), except that the latter factor 
does not affect islands V and VI. Factor III is another antennal factor acting 
upon antenna V and antenna VI. Factor HI of Sokal (1962) does affect 
antenna VI but other than this the two factors have little in common. Factor 
IV of the present study is a leg factor affecting tibia lengh and tarsus length, 
and, to a much lesser extent, sensoria IV and V. Tibia length and tarsus 
length seem to vary much more independently of the other alate variables 
at this level than was the case in Sokal’s (1962) material. It will be noted 
that head length and sensoria VI have no high loadings on any of the four 


178 Tue Universiry SCIENCE BULLETIN 


factors and therefore presumably represent independent dimensions of vari- 
ation at this level. 

The intergall component correlation matrix involves both alate and stem 
mother characters. In the simple structure matrix for this level, factor I 
appears to be a general size factor involving all of the alate characters in- 
cluded in the intergall component matrix. Factor II affects islands VI and, 
weakly, antenna VI. Factor III involves only the stem mother characters. 

The three patterns of covariation present at the intergall level are very 
elegantly summarized by factor analysis. Covariation of the alate characters 
resolves into two clusters, the continuous and the meristic characters, re- 
spectively. The stem mother characters vary together as a group quite inde- 
pendently of the others. Of course this information is simply a summariza- 
tion of that to be derived from inspection of the matrix of correlations; how- 
ever, it does seem that these results provide an additional example of the 
validity of simple structure solutions for this type of analysis. 

In comparing the intergall component simple structure matrix of the 
present study with the same matrix in Sokal (1962), we find that while the 
covariation of the alate characters is divided between a wing-leg-antenna 


factor and a head-thorax factor in the latter matrix, here most of the covaria-— 


tion of alate characters is accounted for by a general size factor. Comparison 
of the intergall simple structure matrix of the present study may also be 
made with a similar matrix from the study of alate and stem mother charac- 
ter variation and covariation over eastern North America by Sokal and 
Thomas (1965). In this work covariation of characters at the intergall level 
is summarized by four factors. Two of these affect alate characters, while the 
remaining two involve the stem mother characters. These results confirm 
the independence of covariational patterns of alate and stem mother charac- 
ters at the intergall level. 

At the interlocality level, the component correlation matrix results in 
three factors. Factor I is a wing-leg-antennal factor (not including antenna 
V). Factor I is a head width-thorax-femur length factor. Factor HI involves 
stem mother head width and antenna III as well as alate antenna V. This 
factor indicates the first evidence of covariation between alate and stem 
mother characters. The factor analysis of the interlocality component corre- 
lation matrix from Sokal and Thomas (1965) reinforces this indication al- 
though not involving the same characters. In the latter study, four factors 
account for most of the covariation. Two of these factors show influence on 
stem mother and alate characters. Unfortunately alate antenna V is not in- 
cluded in the latter study and therefore direct comparison of the two analyses 
for this character is impossible. 

Factor III of Sokal and Thomas resembles factor I of the present study in 
that wing length is affected along with the head, thorax, and leg variables by 


MicrocGEoGRAPHIC VARIATION AND CovaARIATION 179 


both factors. This was not the case in Sokal (1962) where wing length was 
more highly loaded on another factor. With the exceptions of the similarities 
noted above, the interlocality factors of the present study bear few close 
resemblances to the corresponding factors of the other two studies. 


DESCRIPTION AND ANALYSIS OF GEOGRAPHIC VARIATION 


Figure 1 is a map of the localities and their code numbers. As a first step 
in analyzing the geographic variation of the characters the descriptive facts 
are best summarized by separate maps for each of the characters which ex- 
hibited significant variation among localities. These maps, showing the 
means of the populations at the various localities, are based on the data in 
Table 8 stated as mm. for continuous characters and counts for meristic 
characters. The average standard error within localities is furnished at the 
right of each row instead of attaching a separate standard error to each mean. 
This average standard error suffices for any significance testing desired, since 
the means are all based on the same number of individuals. 

To simplify comparison of distribution maps of different characters the 
conventions adopted by Sokal and Rinkel (1963) were again adopted. The 
original scale of measurement of the data was transformed by standardiza- 
tion. Standardization of each locality mean was carried out by dividing its 
difference from the grand mean of localities by the standard deviation of 
locality means. This procedure results in negative standard scores for approx- 
imately half of the localities. To avoid mapping of negative quantities, the 
standard scores are transformed by adding 5.0 to each score. Thus a locality 
with a mean identical to the grand mean of localities will score 5.0, a locality 
with a smaller mean will score below 5.0, and a locality with a larger mean 
will score above 5.0. As a final step in simplifying the maps, all scores were 
multiplied by 10 to avoid decimal points. Thus a locality with a score of 5.2 
is mapped as 52. 

In studies of geographic variation the analysis of variance may be con- 
sidered from two different but legitimate points of view as was pointed out 
by Sokal and Rinkel (1963). In the earlier portion of this study it was inter- 
preted as a model II analysis of variance when it was desired to measure the 
contribution of the various sources of variation to the overall variance. In 
the present part of this study, however, the analysis of variance is interpreted 
as model I since we wish to make comparisons among the locality means. 
The primary concerns at this point are the patterns rather than the amounts 
of variation present. 

The model II analyses of variance yielded 12 characters that exhibited sig- 
nificant variation among localities. We wish now to estimate differences in 
means between localities and test these for significance. 


180 Tue University SCIENCE BULLETIN 


The application of ordinary ¢ tests to such data is not statistically legiti- 
mate (Sokal and Rinkel, 1963). Therefore tests between means at any two 
localities were carried out by means of a multiple range test, the Student- 
Newman-Keuls (SNK) test (see Steel and Torrie, 1960). Multiple range 
tests have been used previously in studies of this type by Cross (1955) in a 
study of geographic variation of bees of the subgenus Epinomia, by Ehrlich 
(1955) in a study of the butterfly, Erebia epipsodea, and by Sokal and Rinkel 
(1963) in the study of geographic variation of Pemphigus popult-transversus 
in eastern North America. All multiple range tests array the means in order 
of magnitude and delimit either mutually exclusive or overlapping sets of 
means. Those means contained in any one such set are not significantly dif 
ferent among themselves. The Student-Newman-Keuls (SNK) test was 
adopted for the present study because of previous satisfactory results (Sokal 
and Rinkel, 1963), and because it yielded relatively less overlap and hence 
more differentiation among sets than other such tests. 


| 


‘ 


Results of the SNK test are shown graphically along the righ hand mar- — 


gins of Figures 2 through 5 in the manner of Sokal and Rinkel (1963). 
These diagrams show the 15 locality means for any one character arrayed in 
order of magnitude from lowest to highest. Lines to the right of the array 
define sets of means not significantly different. Any two means not con- 


nected by a single line can be considered to be significantly different from 


one another at P < 0.01. 


The conventions used for differentiating levels of means on the distribu-— 


tion maps (Figures 2-5) were based on the SNK tests (see Sokal and Rinkel, 
1963). The low group was dotted, the intermediate group hatched with 
diagonal lines, and the high group cross-hatched. If significance tests are 
desired they may be easily performed by comparing the means on the map 


with the SNK diagram alongside it. Several cases were found in this study 


where the SNK test did not differentiate subsets among the means in spite of 
significant variance ratios in the analysis of variance. In such cases all means 
were arbitrarily coded as belonging to a homogeneous low group. 

To conserve space, a map for only one character, alate thorax length, is 
shown in Figure 2. This character shows a group of high scores in the 
center of the area studied (localities 3, 4 and 5). Intermediate level scores 
are shown at localities 7, 1 and 15. Low areas are largely in the south and 
east, but also at locality 6 in the west of the area. Maps for the eleven other 
significantly differentiated characters can be consulted in Rinkel (1963). 

As a result of the factor analyses discussed previously, three patterns of 
character covariation were shown at the interlocality level of variation. To 
represent these factors, three characters for each factor were chosen. Those 
characters were the ones having the highest three loadings on the factor to be 
represented. Wing length, tibia length, and tarsus length were chosen to 


MicrocEocRAPHIC VARIATION AND CovARIATION 18] 


represent factor I. Head width, thorax length, and femur length were 
chosen for factor II, as were antenna V, stem mother head width, and stem 
mother antenna III for factor II. Averages of the standardized scores of the 
three characters for each factor were computed for each locality and plotted 
on maps (Figures 3, 4 and 5). Estimated variances of galls within these 
localities were computed, based on expected variances of sums of variables 
and allowing for the correlations between the variables composing the aver- 
age score for each factor. Using these estimates of the variance, SNK tests 
were applied to the average scores for the localities. Diagrams for these SNK 
tests are furnished at the right hand margin of Figures 3 to 5. The five per 
cent significance level was used in these SNK tests to provide as much dif- 
ferentiation as possible. Factor I (see Figure 3) produces an area of high 
means in the center of the area studied. Locality means decrease in magni- 
tude outward from the center of the area. The lowest mean score is at 
locality 14. Factor II also produces an area of high central scores surrounded 
by intermediate values (Figure 4). The area of lowest scores however, in- 
cludes localities 9, 10 and 13. Unfortunately factor II] produced no signifi- 
cant differences when the SNK test was applied. However it may be seen 
from inspection of the map (Figure 5) that the area of highest scores is here 
centered at locality 1. In general three patterns of geographic variation of 
the factors may be seen. These patterns are not so distinct as those of Sokal 
and Rinkel (1963). This may in part be due to the smaller area and sample 
sizes of the present study having not produced as great differentiation among 
localities as did the previous work. 

In comparing the area of the present study with that analyzed by Sokal 
and Rinkel (1963), locality means for the various characters must be com- 
pared in millimeter units and counts since the standardized scores from the 
two studies are based on different standard errors and grand means of char- 
acters. Table 9 shows the grand means for all localities of the characters of 
the present study along with the locality means for the same characters from 
the four closest localities from Sokal and Rinkel (1963), and Sokal and 
Thomas (1965). These localities were Hutchinson, Reno Co., Kansas: 
Arkansas City, Cowley Co., Kansas; Kansas City, Jackson Co., Missouri; 
and Delawan, Morris Co., Kansas. Also shown in Table 10 are the ranges 
of the locality means from both the present study and Sokal and Rinkel 
(1963). 

It is interesting to note that, while variation among localities is much 
‘greater over eastern North America than is variation among localities in the 
‘Present study, the differences between the four localities closest to the area 
under study here are rather large. For instance the mean value for wing 
length of the present study is much closer to the locality mean for the same 
| 

| 


| 


182 Tue UNiversity SCIENCE BULLETIN 


character at New Trenton, Indiana (see Sokal, 1962, table 2) than it is to 
any of the four locality means from Table 9 of this study. 

Comparison of the ranges of locality means of the two studies shows that 
for most of the characters the ranges found in the present study are contained 
within those found in Sokal and Rinkel (1963). However, important excep- 
tions exist. For the alate characters, wing length, head width, thorax width, 
tarsus length, antenna V, and sensoria IV, the extreme high value of the 
present study is greater than the high value from Sokal and Rinkel (1963). 
For only two characters, head length and islands VI is the extreme low mean 
value lower than in the previous study. 


DISCUSSION AND CONCLUSIONS 


THe ANALYSES OF VARIANCE 


In general the analyses of variance performed in the present study are of 
most interest when compared with similar analyses of the same characters 
over geographic areas of both smaller and much greater size. The most im- 
portant conclusion to be derived from such comparisons is, as might be 
suspected, that variation among localities forms a much smaller portion of 
the total variance for any one character in the present study than was the 
case in the study of eastern North America (Sokal, 1962). The fact that 
fewer characters exhibited significant variation among localities in the pres- 
ent study tends to bear out such a conclusion. The meristic characters in 
general tend to be less variable than continuous characters over a geographic 
area and also within the localities studied. That stem mother characters tend 
to confirm the above conclusions is evidenced by the fact that only two such 
characters showed significant variation among localities in the present study, 
while in another study of the same characters over eastern North America 


(Sokal and Thomas, 1965) all stem mother characters studied yielded sig-— 


nificant variation among localities. 

The amount of variation within galls is very nearly the same for the 
three studies. There are two exceptions to this statement. The variance 
within galls for the alate character sensoria V is much smaller than the com- 


parable variance from Sokal (1962). It may indeed be the case that the | 


variance within galls for this character is less in this area. In the case of the 
stem mother character, tibia length, however, the large error variance 
(among galls within localities, since there was only one stem mother per 
gall) may be due to inaccuracy in measurement, since in many cases this 
structure was obscured by the large, soft body of the stem mother. 

The variance within galls of the alate measurements is presumably due to 


environmental influences only, since all of the alates within any one gall are” 


genetically identical, barring mutations. This environmentally induced vari- 


5 


MicroGEoGRAPHIC VARIATION AND CovaARIATION 183 


ance is very nearly equal in this study to the variance shown for eastern 
North America by most of the characters. Thus the factors acting to pro- 
duce this variation are probably similar. 

The amount of variation present among galls within localities for the 
three studies seems to be of the same general magnitude. As was pointed out 
earlier, the Ripley, Ohio, locality studied by Sokal (1952) seems to have been 
unusually well differentiated among galls. Hence the ratios of variance 
among galls to variance within galls for that study are considerably larger in 
magnitude than the comparable ratios in the present study and that of Sokal 
(1962). Ratios of the same quantities from the substudy of three trees in the 
present work are of the same magnitude as the ratios of Sokal (1962), and 
the ratios over the whole area of the present study. 

Variation among galls within localities is due to both genetic and environ- 
mental influences. Thus, considering the above results, these influences must 
be of relatively the same magnitude over the area studied as they are on the 
average over the whole of eastern North America. 

Variation among localities must also be due to environmental and 
genetic causes. As mentioned previously, this variation is much smaller in 
the present study than in the previous studies. Inspection of the ratios of 
variance among localities to variance among individuals within galls (Table 
3) reveals, however, that variation among localities is not a linear function of 
the distance between localities. For instance the above ratio for the alate 
character wing length is 0.43 in the present study. In Sokal (1962) the same 
ratio is 2.98. Thus the latter ratio is roughly seven times the former while 
the average distance between localities of Sokal’s study is certainly greater 
than seven times the average distance between localities for this study. 

Unfortunately it is impossible to partition the interlocality variance into 
environmental and genetic components. However, is seems that if the 
genetic differentiation were a function purely of the distance between locali- 
ties, the interlocality variances of the Sokal (1962) study are less than would 
be expected. Perhaps similar environmental combinations result in similar 
gene complexes. 

The comparatively large amount of interlocality variance present in the 
limited area studied suggests either that the population structure of this 
organism consists of many small, extremely localized populations or that 
microgeographically differentiated ecological factors are very important in 
causing phenotypic differentiation. Most likely both explanations are par- 
tially true. 

The separate study of three trees within one locality revealed very little 
significant variation among trees within a locality. These results justify the 
use of the sampling procedures in which several trees within a small area are 
sampled and counted as one locality. 


184 Tue University SCIENCE BULLETIN 


Tue CorrELATIONS AND THE Facror ANALYSES 


As mentioned by Sokal in his study of eastern North America, intragall 
correlations are based on common factors of environmental origin which 
must act during the development of the organisms, since, presumably, 
genetic differences do not exist among individuals within galls. The intragall 
correlations among the alate characters of the present study are of the same 
general magnitude as those found by Sokal (1962). A cluster of high correla- 
tions is again found among the wing, head, thorax, and leg variables and in 
the present study, one antennal variable, antenna V, is included in this 
cluster. The patterns of covariation at this level, as reflected by a comparison 
of the factor analyses, are, however, somewhat different. 

Factor I of both studies involves wing length, head width, thorax length, 
thorax width, and femur length. In the present study, tibia length and tarsus 
length, however, are not affected by factor I. Tibia length is not strongly 
influenced by any of the four factors of the present study and may represent 
an independent dimension of variation. Tarsus length is most strongly in- 
fluenced in the present work by factor IV. In the previous study the anten- 
nal variables were most strongly affected by factor II, while in the present 
study they are influenced in groups of two or less by all of the last three 
factors. 

It appears then that in the present study the wing, head, and thorax vari- 
ables are influenced by the same general environmental factor as in the 
previous study. The remainder of the alate variables are, however, in- 
fluenced by different factors which come into play within the geographic 
area under study. The two leg variables, tibia length and tarsus length are 
influenced by a factor which also has an effect upon sensoria TV and V. 
Tibia length is, however, influenced in the opposite direction from tarsus 
length and, as noted above, not too strongly. Antenna III is influenced by 
factor Il which has very little influence on any of the other variables. How- 
ever factor II is highly correlated with factor I and thus represents a very 
similar dimension of variation. Antenna V and VI share factor II which, 
again, has little effect on the other variables. 

Factor I of the present study may represent a general size factor as it did 
in the previous study. Factor III found here may bear some relation to factor 
IL of the previous study but this is far from clear cut. The remainder of the 
factors are perhaps best interpreted as being of local origin. 

At the intergall level of variation, the component correlations again re- 
flect a cluster of the wing, head, thorax, leg, and some of the antennal vari- 
ables for the alate characters. Here, however, the patterns of covariation are 
somewhat different from those of the previous level of variation. Factor I is 
again a general size factor for the alate characters. All of the alate variables 


MicroceocraPHic VARIATION AND CovARIATION 185 


included at this level are highly loaded on factor I except antenna VI and 
islands VI which are more strongly influenced by factor I, which appears to 
be an antennal extremity factor. The influence of a leg character-wing length 
factor such as factor II of Sokal (1962), does not appear in the present study. 
The stem mother variables, which are present in the correlation matrix at 
this level, are all highly influenced by factor III. Factor II, therefore, ap- 
pears to be a general size factor for the stem mother characters of the 
present study. In summary, the patterns of covariation at this level of varia- 
tion seem to consist of characters influenced by the general size factors for 
both the alates and the stem mothers. The exception to this is the appearance 
of an antennal extremity factor for the alate antennal characters. 

The fact that no group of covarying stem mother and alate characters 
appears is, of course, not surprising since no high correlations involving these 
characters were found. In general, the patterns of covariation at this level 
are different from the Sokal (1962) study. 

At the interlocality component level of variation, the correlation matrix 
gives rise to three factors. The first two of these factors represent covaria- 
tional patterns among the alate characters. The third factor is of special 
interest since it reveals a group of both alate and stem mother characters. 
This is supported by the findings of Sokal and Thomas (1965) and may 
represent a common genetic influence at the interlocality level of variation. 

In summary, it may be said that while the amounts of variation present 
at the respective levels in this study reflect what might be expected from a 
consideration of and a comparison with, previous studies, the patterns of 
covariation present in the material studied here differ considerably from 
previous results. These differences are most striking at the interlocality level 
where the influence of different local factors within the area studied come 
into play. It should also be added that the data on interlocality covariation 
are based on only 15 localities, hence are not too reliable. Differences in 
covariational patterns are, however, present at the lower levels of variation 
and might reflect the influence of factors present in the area studied and 
absent in most areas that were included in the study of eastern North 
America. One might expect the influence of factors of local importance in 
one area to be obscured by the presence of other factors in other areas in 
such a study as that of Sokal (1962). Investigation of the factors influencing 
covariation in such a study as the present one may at least give insight into 
the types of patterns which may be present within a microgeographic area. 


186 Tue University SCIENCE BULLETIN 


LITERATURE CITED 


Cross, E. A. 1955. Studies of systematics and variation in the subgenus Epinomia (Hymenop- 
tera, Halictidae). M.A. Thesis Univ. Kansas. 

Euruicu, P. R. 1955. The distribution and subspeciation of Erebia epipsodea Butler (Lepidop- 
tera: Satyridae). Univ. Kansas Sci. Bull. By ia -9 4. 

Rinker, R. C. 1963. Microgeographic variation and covariation in Pemphigus populi-transver- 
sus. M.A. Thesis. Univ. of Kansas. 

Soka, R. R. 1952. Variation in a local population of Pemphigus. Evolution 6:296-315. 

_ 1958. Thurstone’s analytical method for simple structure and a mass modification 

thereof. Psychometrika 23:237-257. 


———. 1959. A comparison of five tests for completeness of factor extraction. Trans. Kansas 
Acad. Sci. 62:141-152. 

962, Variation and covariation of characters of alate Pemphigus populi-transversus in 
Eastern North America. Evolution 16:227-245. 

SokAL, R. R. anv R. C. Rinker. 1963. Geographic variation of alate Pemphigus populi- 
transversus in eastern North America. Univ. Kansas Sci. Bull. 44:467-507. 

Soa, R. R. anp P. A. THomas. 1965. Geographic variation of Pemphigus populi-transversus 
in eastern North America: Stem mothers and new data on alates. Univ. Kansas Sci. 
Bull. 46:201-252. 


STEEL, R. G. D. ano J. H. Torrie. 1960. Principles and procedures of statistics. McGraw-Hill, 
New York. 481 p. 


MicrocEocrRaPHIC VARIATION AND COVARIATION 


Taste 1. List of localities employed in this study. 


187 


Locality 
Code Date of 
Number Locality Collection 
Ie Douglas Co. Kan. U.S.G.S. Lawrence East map, quadrat 25, on boating 
ground; west bank of Kansas River. <0... secec cece 15-IX-61 
Is Douglas Co. Kan. U.S.G.S. Midland Section map, quadrat 18, 14 mile 
west of junction of U.S. Highways 40 and 24, 1 mile north of Lawrence, 
LISTS te ep es aon ee PR sc coe ae 27-IX-61 
ie Douglas Co. Kan. U.S.G.S. Lawrence East map, quadrat 20, east of 
BSI Gk s GONG rk A eee ee Ritmo ITS ee hy, 8-IX-61 
4. Douglas Co. Kan. U.S.G.S. Midland Section map, quadrat 7, 4 mile 
east and %4 mile north of U.S. Highway 24-59. ca 1-IX-61 
Be Douglas Co. Kan. U.S.G.S. Midland Section map, quadrat 17, '4 mile 
north of U.S. Highway 40, east of Lawrence, Kan. aI PORta sere 1 -IX-61 
6. Douglas Co. Kan. U.S.G.S. Williamstown Section map, quadrat 15, along 
creek north of east-west county road due south of Lakeview recreation 
CURE «cece AE aeons a eine a ead ins er ee Et sere Tee 2-IX-61 
Te Douglas Co. Kan. U.S.G.S. Williamstown Section map, quadrat 15, due 
east of Lakeview Lake and % mile southeast of Lakeview School. 2-IX-61 
8. Douglas Co. Kan. U.S.G.S. Lawrence East map, quadrat 36, along south 
bamnicsobKcnsasw Rivers taht 1c ee, VCE es ee ATR ce 2. 6-IX-61 
wy. Douglas Co. Kan. U.S.G.S. Lawrence East map, quadrat 36, between 
Kansas River and county road intersection No. 811... 6-1X-61 
10. Douglas Co. Kan. U.S.G.S. Lawrence East map, quadrat 33, % mile 
north of county road intersection No. 817. .0.00-0----0--ccccceeeceeceeeeeee cece ; 6-1X-61 
nT: Douglas Co. Kan. U.S.G.S. Lawrence East map, quadrat 32, % mile east 
olycounty, bridge No. (BIB. ee ee 6-IX-61 
12 Douglas Co. Kan. U.S.G.S. Lawrence East map, quadrat 33, % mile east 
@t county: Toad: intersection Nos 822. 2... 2-0. 8-IX-61 
11335 Leavenworth Co. Kan. U.S.G.S. Lawrence East map, quadrat 27, at 
countyetoad intersection, No, Olli, S22 en eee ee 8-IX-61 
14. Leavenworth Co. Kan. U.S.G.S. Midland Section map, quadrat 15, north 
of Kansas Highway 32 at county road intersection No. 879. 11-IX-61 
15. Douglas Co. Kansas. U.S.G.S._ Lawrence East map, quadrat 32, % mile 
east of county road intersection No. 822. 0 --ecceeeeeeecee 15-IX-61 


188 Tue University SCIENCE BULLETIN 


Tasce 2. Results of analyses of variance. 


F ratios 
Mean Squares (Present study) 
ey) inde oe as = g 
eis an < So 5 ‘5/2 
fe a6 2 £2 3.. 2 cee 
Fe 32 46 28 23-08 sole 
2 ¥ OM B eie AB 2 3 ool c oo| S 
Bes Sis age ee Sc Wages s|6 s§/& 
Sree meee 2s 3 £5 9 Shea 
= Sie Ge ae a RE Sano ne 
C= w= Ov = =e $8 wpe outs 
Characters Eee SEN 6S OL Che ie SoS 
Alates: 
Wing Length —.. .01567 .01090 0418 .0247 06786 .28711 +t 44 
Head Length —.. .00022 .00025 xX X 00024 .00035 ——- —— 
Head Width ...... .00034 .00040 00655 .00061 .00076 —.00267 ++ +4 
Thorax Length. .00148 .00182 00307  .00269 00465  .02556 ++ ++ 
Thorax Width .. .00274 .00259 X Xx .00827  .05210 ++ 4+ 
Femur Length . .00170 .00103 X X 00380  .00983 ++ ++ 
Tibia Length —. .00323 .00199 90382 .00283 00541 .01320 ++ +4 
Tarsus Length .. .00014 .00007 xX xX 00024 .00085 ++ +4 
Antenna III ...... 00019  .00015 00030 .00026 00042  .00099 + t+ 
Antenna IV _... .000035 .000030 X xX .000080 .000385 ++ —++4 
Antenna V ........ 000083 .000040 X X 000102 .000214 —— +4 
Antenna VI 22: 000108 .000120 00016 .00019 .000214 .000374 + + === 
Sensoria III —.... 9658  .7797 9658 1.1356 07960 12131 ——- —— 
Sensoria IV -...- sissy lO .1137 — -.1600 .02340 .03352 ——- 
Sensonia §Vi 22 0311  .0739 X xX 00371  .00343 ——- —— 
Sensoria VI _... 0156 = .0188 X X 00201 .00110 —-—- — 
Island senV; ee 4667 ~—-.6261 xX xX 05152 .00713 —-—- —— 
Islands VI ...-..-- 6422 .8421 X X 07852 .12014 ++ —— 
Stem Mothers: 
Head Width X X 00094 = .00120 .00094 .00330 X + + 
Femur Length .. X X 00112 .00110 00112 00159 X —— 
dubia) Lengthy. xX X 00495 .00079 00495 .00344 xX —— 
Tarsus Length .. X xX X X 000096 .000080 X —— 
Antenna III ...... xX X 000275 .000270 = .000275 .000531 x + 
Antenna IV ..... x xX 000084 .000090 =.000084 .000060 xX —_—— 
Explanation: X signs denote either variables not used or not applicable in the correspond- 


ing category. 
+ + denotes significance at the P 
“fb denotes significance at the P 
—— denotes no significance (P > 


= 1 level. 


0.0 
0.05 level. 
) 


ollAll 


05 


MicrockocraPHic VARIATION AND CovaRIATION 189 


Taste 3. Comparison of ratios of variance components. 


o oN 
= s g ae 
> = ~ alse es =e 
nn nna nD nm~ Cn eS fe S fh (kee cies Gt Pid loa & 
Sse se cae) Sle SS Sees eee 
25 ,l28 «28 @iSq Ses Slee 2l28 s|26 alfa 
bibs (Fe WES ES S52 SPS thee Vee SEs 
Characters he Ge ao eo he (ec 4 6 a) 2 
Alates: 
Wing Length 1.86 1.66 1.58 1.54 0.43 2.98 0.14 loil7 1.39 
- Head Length 0.64 0 0 0.21 0 0.19 Xx Xx xX 
Head Width 1.64 0.62 1.14 0.50 0.15 0.40 0.09 0.27 0.47 
Thorax Length 4.81 1.07 0.71 0.86 0.44 0.97 0.21 052 0.67 
Thorax Width 3.13 1.01 0.56 0.63 0.50 0.55 xX X X 
Femur Length 2.21 0.62 0.60 0.75 0.08 0.83 X X X 
Tibia Length 1.82 0.71 0.70 0.65 0.08 0.83 0.05 0.48 0.59 
Tarsus Length (0.92 0.36 0.98 0.58 0.11 0.82 xX X X 
Antenna III 1.07 0.61 1.50 0.76 0.07 1221 0.04 0.69 0.84 
Antenna IV 0.95 0.63 135 0.72 0.26 0.85 X XxX X 
Antenna V 0:52 0 0 0.51 0.04 0.35 xX x Xx 
Antenna VI 0.88 0.49 0.96 0.39 0 0.31 0 0.22 0.12 
Sensoria_ III 0) 0 0 0.37 0 0.28 0 0.20 OF 
Sensoria IV 0 0 0 0 0 0.32 0 0.32 0.17 
Sensoria V 0 0 0 0 0 0.05 Xx Xx x 
Sensoria VI 0 0 0 0.14 0 0 Xx Xx XxX 
Islands V 0 0 0 0 0 0.08 X X X 
Islands VI 0.03 0.23 0 0 0 0.06 X X X 
stem Mothers: 
Head Width Xx xX xX x X X 0.17 Xx 0.29 
Femur Length xX X X X X X 0 X 0.41 
Tibia Length Xx XxX xX xX XxX XxX 0) X 0.25 
Tarsus Length  X X Xx X X X 0 X X 
Antenna III Xx Xx xX X X X 0.06 X 0.26 
Antenna IV Xx XxX x Xx xX X 0 X 0.19 


ixplanation: 


X signs denote either variables not used or inapplicable in the corresponding category. 


oAeY sjutod [eutd9q 


‘addy uewlol ul ore (/['J=+) [0° 


xe JEpE 00- 
a X £0 
Z & Anes 
w 
Sj : ns 
= 
a) --- - 
w 
oO = S 
Zi 
Oo 
Ex 
Z ---- ee <= 
a a ee 
zz 
a o--- acne noes 
2B, 
i. ) ee. 
_ — ie) 
ro) co) 5 
22 Gap I 
a 7 Ete 
0) 
8 gee 


A Piosuag 


AJ eosuas 


II 
LI 
tI- 
+0 


II] Bt10suag 


IA ebuuajUuy 


A buuaquy 


=> 


*peytwuo U39q 


d We JULIYTUSIS 9souy :SJUSTIYYJIO9 uone[at41o5 yuswour-jonpoid oIe SJUSTIYJIOI IV :uoneurldxy 


90 40 10- £0 CO 0m 
als CGO) CL $0n Ve 
SOO 10 Gi W0- §= 
(IS £40) Coe len SO. jet 
OUm “Ef bb lime OE 81 
TO ce Gi al 90 Gi 
bC 1Z i (OE eo aa 
Te 0¢ cl ce C7 ce 
x ZZ 20 Ie OT 02 
ai Kijae. ee ral Os 
a: Gi (Ie bb 


“s > 8C CE 

; : x Or 

; x 

> > cael 4 ps) ia 
=) =} = > 3 => 
a a x me = 2 
5 5 & e 5 
5} 3} ee = a x 
Be sig ke es A ee | Se 
a = 5 ae 5 = 
% vel > ag) ee 

SP eo is 


Lom 
9 
Lit 
60 
PI 
ST 
81 
Ir 
60 
he 


ysuaT xeroy 


COm ano GOne Git en are IA spurjs] 
9C £0 1 ed alte oe aE ca A spurjs] 
> Ci Ce IA Ptosuag 
00 S0= Lt tg aa A. PBiosuag 
SI O0=* 61% seeneiomaiee aa AI Pltosusg 
81 BQ “yO II] eltosuas 
OL 00 (Ge oh a ea IA euuajuYy 
cl Qi) Pe ie) ee ae een A vuuajuy 
60 07 Ci" al ecient AI euuoluy 
O€ C= AG Sia ee II] euucjuy 
bE BO CQ yisuaT snsie 
eC Gil Gis ae ae ee ysuaT eiqhy 
€C (CH eae (G(R aN ae yasuay muy 
9S Sp" AS open ae pra xesroy 
IS El an SOU Cee ee yagsuaTy xer0y L, 
xX ZO SSm eee aT aang YIpIM PPeH 
ce) x GD oF area rae wasus7T prop] 
7 ac b Gears aa mna ts yasueT Surpy 
<5: (== te) 
i es 
ie 5 


190 


“SUOTIP[IIIOD [jesenuy ‘> aldv 7 


191 


MicrocGEoGRAPHIC VARIATION AND CovVARIATION 


-YIUSIS Sues dy] SUMOUYUN de SUONR[III0D JUSUOdUIOD JOJ s[aAd| BDUKIYTUSIS 9dUTG ‘ddA UROL UT aie (6° 


sJuaToyya0> yusWOW-}INpoid [PY ‘suoHe[s110) yUsUOdWIOD [eUOBeIP dA0qe 


*saJIeIPYD JayOU WI}s a1e WS Aq papadsid ssaydeIeyO ‘asaid 919M syUsUOdWIOD dIURTIBA JURIYTUBISs OU aduIs aInd 
-WI0D 0} a]qIssodut suONR[aII0I JUZUOdWIOD JUasaIdaI systIasy “paytwo usIq aAeY syuTod jeuIIIq ‘“A[LeNIGIe UdsOYyd Ud2q Sey [PAd] 29UeD 


=} 4“) 10 S d 3 JWeoytusts 
:SUONP[IIIOD JUSUIOW-JINpoOId die [eUOSEIP MO[Iq s]UdIDYJZI0g 


:uoneurldxy 


Sanaa Cane Con CO = mie = 20-10 aGhoeZ0 80°50. SO a80="20 ~0- 20— S0— 80= 20. T0- ~ AI euuauy WS 
EGC CCS GuY VO e/0un 20) 10- -S0- ZI S02 "Sl “SC YP <9 “Ci Si “TE LO OL ZO Zr II] euuauy WS 
rr ee 0p oe) c0N 50-50) 20— €0- ci “SO 20 IC If 20 80 Of 80 “80 60 80 $1 ~~~ ysuaT snsiey, WS 
Fee ee CCG 20, 90 c0-" S0— S0— 0° 40 SO) £0 “0 90 §0 50 60) “OF OF “S0) Bi ~~ ysuaT eIqry, WS 
eee ee XG 90— S0- J0= €0— S1- 20° S0="S0 Cr -FI. 90 20 20°20 “40 <CF “£0 “7 yisuaTy inway WS 
ne ee ne eX 00] S0200= 90— £0—= 60 “CO 60 £1 Zr 90" SO. GI 60 V0l “OF iSO OF ~~ = WPI PeeH WS 
ee ee ee lS) F0) C0 80— Gl 90) 80 “£0= T0= go. OF 06 Gk “=9f FOn Si = IA. Spurjs] 
+ +¢ *£ * & # eC 50-700" 90 90! “Ole Ol 460 SO. SE SE 10 “10. O° COM LO = A spurs] 
* * * * *  * Se XoeCU= 090) SS0= les C0= 2O-—60 Oe 20 "00. 90= L0= W0e=l0= 6° = ae IA blosuag 
oe ce «ele es a eer OlGN60 60 eee 90h Si = Sie £0= 90. Ch Sit. Sie 0 O= aoa be A elosuag 
ee) os) ee se es er COL Camera Cle aL Ce P= 90 Cl 0 Sl Gm Uae ee ae AI Blosuag 
ee ee ee! tos ee X Peal. he Mik CF 8S. Sa Sh 8S ST 90) GEG. a II] etrosuas 
pe ee ee GL ae kk ele XM ES OE CE Se Ob PE Te 60S «6b 8 8h IA euuauy 
ee + ee eke MOXG TS) la bw: x oe 2 XX 19 6S $8 C6 SS Sb 9b 7h fe G0 A euuauy 
eee cece ee Cl eee 6s oa a C9 4 XK 6S Ib «0S: 2b IS <2¢% TS) 81 99. AI euuauy 
cee ee Pree Ola en ne 8x ee 9 £9 4 68 X° 16S" 96 “9S 09: G9" 69 “9G TL II] euuauy 
ree ae +e le MnO wae oe ee a (OS ie 928°. 8S ~X: Ob? 2b CbS 1S (26> Sk “Gp Se ysusT snsie yp 
Fee Gees Ol ek ek a6 el | CO. Cw «| SG 88 SG CX ED 09 09° Sc CE OL yisusT eIqrys 
ee eee ee OG ee Se ee 2S 6 9/9 $6 92, 26. X% 89-599 8G. fF TL ysuay inuis4 
Se eee eee eee er CS ee ee Oe we Ok ID we. CL GL 99 LE 88 X 3084 82 06 8h MPLA Xesoy 
tr eee ete ka OS x ee x a fe «SL. <4 985 88 ~/9° TS 68 98 xX “IL OR. Ze ee yasuay xeroy 
Stee ee ee ee NY aie ee ce fe C8 om C8 -b8 CZ Of 98. 26 $8 XK. Lt 89 (PIAA _peeH 
* * * * * * * * * * * * * * * * * * # * * XG ell Cae eee ones yysusT pray 
*  * * # *  * 96 x *% * «* « © « IT HS AKA GO SS GR yasuaT Bury 
eee eke a BE eS ae Fy Boe RFs 

oe og oe ee | 8 CP She Se ue a ee Pita Ga a aa ies ie a als 

ST Res ES Ie ee ee ae ee ee ee eae 

2 iis. ie = % eS os SE sige ot a tak SRB pa eke 

2 6 8 « 8 & = S&S & & = 

ae eae a Ee 


= 


ssoduit suone]os 


ay ‘suONe]I1109 JUDUOdWIOD JOY UMOUYUN dIe SI>IAIT aouroyIusis 3uIg “sdA} ueWoI ut oe (79° 


SAMI CACYY ASYIUE Cress cae me Rm MOP vee ene gy su oma ote 
qu quasaidar sysuiasy “pavtwo usoq 2aeY SyUTOd PeUDI9q. “AjLeNIqse UDdsOYD UVZE2Bq sey [PAdg|_ adUe> 
“) 10° S d i aueoytusts 


MUSULOULUOD SIUELIEA 4ULIYLUGIS OU QUIS 9]nNG 


SWUIIYJe09 JuswuoUlyNpord |[Y “suoNvjes10d JUsUodWIOD aie [eUOseIP 2aoqe {suONe]II109 YUSWOW-IoNpord aie [eUOseIP MoOjaq syUsYI0q :uoNLUL|dxY 


Kee CUm i NOUS ari sla C2 ve  S0- 6r LP §& 61— TI- 60 #0—- SO ZI— If 2% Ze O0= 000 AI euuaUuy WS 
FO i hme (CO Cle Tisch SC SS S0=8 LC “Pl "CS “OF “2 W— 8t0 20> T2= 00= 2O SI— “o- IJ PuuaUyY WS 
* tee CS Cee ia sil “mel LO ES to= Zi— é— S0— f= S&> OS c= Ol- OF #0 SO. === (psuaT snsiey, WS 
* * Cm CCU SO aC 60m l= aC Cl= 90— Si £a —L0- 60) T0- bl 60. 80) 6) Gb ss (ysuaTy eIqly, WS 
* * * Fe UG OG 00m (SON tel 60 2G “86 65 GC Ck “Si ZS. SI 80 He Sf 90) Zz yisueTy Inuajy WS 
* G6 « * Ce SS eS Cont Cue Lamm Cue SC) i NOS 9G eo MSE ICS 9850) S0= Gl (60m (Gla aol. wan mae MptMm PpeeH WS 
2 * * * * * * X CG Wl 02> Ei BA SE ESS GY iia NO HO 0 WO Ga i= IS OS IA Spurs] 
a * * * * * * * XS @ Oe A WE GE Wee WR YES CS H- We ha KS Ve Oe A Spurs] 
is * * * * * * * OG IG TS SO 0 NS ISIS OS Si 0) an AO) AO SSG IA blosuag 
5 * * * * * * * * oO 0 GOS ON ie US WP GCE SO ON GN Ge BC GR WE Bae A PLIOSu9g 
om * * * * * * * * * OHO GG HS AS NE eS) NG OD), AGS KG AI Btosuag 
3) * * . * * * * + * * * 2G QM) “GES Oe SO 1 IE SSG OG II] el4osuag 
a * * * * * * * * * * * Sy OS MS Ge GS GS YS EES OD AGS GAS OY IA euUusjUuy 
Z * cS x * *« te ¥* * * * * * eS HD OE ee tae OS GIS ISS rae he (Gre A euusjuy 
Ss * £0 * * * 0& > * * * * * * Li x BAUS OES SZ ye OS SA CI) GE, irr A] Puujuy 
= * 9F «* * * OF * * * * * * * OP (OD. 6, AS ASO) AS EKG AS VAS, A NG [J] euuajuy 
Fs * Cote * * PO x * * * * * * IKE = AS) WAS Oe HL, ES YAS IS) AS AAG GE Pa yisuaT snsiv yp, 
i * +0- x * * €S «* * * * * * * OG GE AUS WB OS GG SL HE KS YE GI (ysuaTy viqhys 
5 AS er" * * SO x * * * * * eeecces (69) NiO Za 9S SSX 6) aConeCs) 19 eo an eo ania ene qsuay Inwio4 
3 * OF x * * (61>. 2 * * * * * * Gms 9 OS Oey Cem Xen Gee Cinna Qipray xesoy | 
es * FO- * * ¥e 21 * * * + * * . (A NSN LSA PMO RO NOVA YE DC ANY GE | Meelis) Se se yisuaT xei0y |, 
* bO- x * * 80 * * * * * * * ONS NG RS re NS SS AE ASHE GS) MPIM Pree 
* * * * * * * * * * * * * * * + * * * * * * DN OI agar (ySuaT prop] 
* 6I[- * * * 61 * * * * * * * GS AS 99 £9 09 19 ¢L LCL x >, iia qisuay] SurAy 
oes ese 6 2 Pee sg foe 2 eee 
cee ee eee ss ee UP Ee EM U8 lS 
ee ee ee ee 
eee er ae Se ee ee ee ee 
qe ieee oa ee ee = ee ees PAU 
es cha tye = 
a a 


“suOT]e]IIIOS AVPeIO]IINIUT] ‘9 ATAV 


5 


MicroGEOGRAPHIC VARIATION AND CovARIATION 193 


Taste 7. Factor loadings for the intragall product-moment, intergall component 


and interlocality component correlation matrices. 


Intragall Intergall Component Interlocality Component 

Factors I II Ill IV I II Ill I II Ill 

Characters: 
Alates: 

Wing Length ++ = sinain ate =e 
Head Length D4 Xx x Xx ».4 Xx 
Head Width + ++ fal te 
Thorax Length ++ -+ 4+ 4+ aL 
Thorax Width +++ +4 afk iL th 
Femur Length + -- 4--- te 

Tibia Length — +++ +++ 
Tarsus Length +E fe 4+ 4+ 4 
Antenna III ++ + -+- + + 
Antenna IV +++ 
Antenna V +--+ Xx xX XxX -|- 
Antenna VI 4+ Xx x x 
Sensoria III xX X X X x Xx 
Sensoria IV oo xX Xx xX x Xx Xx 
Sensoria V x Xx Xx x XxX XxX 
Sensoria VI Xx Xx xX x Xx x 
Islands V X x X X x X 
Islands VI +--+ Xx xX Xx 

Stem Mothers: 

Head Width X x X xX + +++ 
Femur Length X xX xX me ++ x Xx X 
Tibia Length X Xx Xx xX xX x xX 
Tarsus Length X x x x Xx x Xx 
Antenna III xX XxX xX xX ++ +-+-+ 
Antenna IV Xx xX Xx x x X xX 


Explanation: 


This table represents simple structure factor loadings which are correlations of 
reference vectors with the variables. They are proportional to the primary patterns 
(standard partial regression coefficients on primary factors) which are not shown 
here to conserve space, but together with the correlations between these factors are 
available from the author upon request. Plus or minus signs represent the follow- 
ing magnitudes of factor loadings (positive or negative, respectively): Intragall, 
3 signs =|0.65|, 2 signs =]0.50|, 1 signs =]0.35); intergall component, 3 signs 
=|0.75|, 2 signs =|0.65|, 1 sign =|0.55]; interlocality component, 3 signs =|0.65|, 
2 signs =|0.55|, 1 sign =|0.45). X-signs denote variables not used in the study 
indicated. 


194 


Tue Universiry ScrENcE BULLETIN 


Tasie 8. Means of all characters for all localities. 


Characters 


Alates: 
Wing Length 
Head Length 
Head Width 
Thorax Length 
Thorax Width 
Femur Length 
Tibia Length 
Tarsus Length 
Antenna III 
Antenna IV 
Antenna V 
Antenna VI 
Sensoria III 
Sensoria IV 
Sensora’ V 
Sensoria VI 
Islands V 
Islands VI 

Stem Mothers: 
Head Width 
Femur Length 
Tibia Length 
Tarsus Length 
Antenna III 
Antenna IV 


Locality Code Numbers 


6 


oh 


8 


y) 


10 


11 


2.09 2:16 2:24 2.33 2.15 2:04 2.22 2.09 1.99 2.04 2:09 2.18 206 ROS e2ais 


18 
38 
56 
-60 


18 
38 
=P) 
oy) 
45 
.66 
18 
16 
.07 
10 
aly 
3.70 
.03 
1.03 
1.00 
2.23 
253 


sa! 
iy? 
eat 
alle 
16 
<li 


18 
38 
2) 
.67 
49 
.69 
alls) 
all) 
.08 
ailel 
ali, 


slY) 
40 
61 
70 
50 
.69 
all) 
.18 
08 
oll 
.18 


3.87 3.50 


10 
1.10 
1.00 
Bese) 
2.10 


46 
ed 
34 
alley 
16 
sili 


1.33 
1.03 
1.00 
233 
ZS 


44 
41 
34 
13 
15 
Al 


18 
33 
58 
62 
47 
.69 
18 
.16 
.07 
10 
ally/ 
3.87 
.07 
1.00 
1.03 
2.07 
2.47 


Ra 
al 
oD 
a3 
als 
alal 


18 
=3/ 
5 
62 
47 
.66 
.18 
16 
.07 
10 
a7 
3007 
.07 
1.00 
1.03 
2.60 
2.63 


46 
342 
aot 
allay 
16 
mall 


sil) 
39 
By 
67 
48 
.66 
.18 
all 
08 
10 
a7 
3.43 
1.00 
1.03 
1.03 
2149 
2.60 


43 
40 
39 
14 
15 
alii 


18 
37 
hit 
-60 
Ls) 
.63 
al7, 
16 
.07 
10 
ally 
Sie) 
.03 
1.07 
1.00 
27, 
233 


4 
40 
34 
13 
all> 
lull 


18 
36 
py 
Sy) 
43 
.63 
lly 
16 
.07 
10 
16 
3.97, 
0.00 
1.00 
1.00 
Beli 
2.63 


“43 
40 
“ont 
at 
15 
oli 


aly 
37 
2S) 
>» 
cS) 
.63 
18 
oly 
.07 
sili 
a7 
4.07 
.03 
1.07 
1.00 
2.40 
227, 


53 
40 
39 
16 
18 
sls) 


18 
38 
5 
.62 
47 
66 
ll 
16 
.07 
10 
16 
3297 
aly, 
1.00 
1.03 
Pal 
2.47 


46 
Hil 
Sat 
14 
16 
ll 


19 
39 
ey) 
.67 
49 
.66 
18 
.16 
08 
.10 
a7 
3.53 
28 
1.03 
1.00 
Dey) 
2.83 


45 
Anil 
32 
.14 
16 
alli 


18 
34 
ey? 
-60 
46 
.67 
18 
16 
.08 
10 
.16 
3.30 
03 
1.03 
1.00 
2.53 
3.03 


Ai 
40 
oh 
18 
alld 
10 


18 
38 
>e) 
61 
Gil 
64 
aly 
all5) 
07 
10 
7, 
3.50 
0.00 
1.00 
1.03 
2.63 
2.70 


a9) 
40 
34 
oS 
SII) 
slit 


All continuous variables are given in millimeters; all meristic ones as counts. 


oll) 
38 
56 
.65 
48 
67 
18 
16 
.07 
-10 
mili 
3.50 
0.00 
1.00 
1.03 
2:28 
2.50 


ais 
41 
33 
ale 
AS 
sll 


MicrockoGRaPHic VARIATION AND CovVARIATION 195 
Taste 9. Comparison of locality means with related studies. 
a # io, 
See. Og. Ee ‘g 
ees 20) eee > z 
eas) AS 2S ogf Se aS) s 
eee wee sero Ser é fe 
Eye 224 qs a~-u O28 re % 
9} 26 iS a ~ adi S$ a = aN 8) o os ~ 
aa & £6S £68 die aoe ag 2s 
Characters Owe aio Bea a=) wes Se SLES 
Alates: 
Wing Length 2.12 1.55 ly? 155X0) 1.52 1.98-2.33 1.52-2.10 
Head Length .18 fl LAll XxX 10 .18- .19 19- .21 
Head Width 38 soy 36 30 34 36- 40 34- .39 
Thorax Length .56 a5? By 45 44 D2= 61 44- .63 
Thorox Width .62 54 52 xX 46 ))- ./0 46- .65 
Femur Length .47 42 41 Xx 40 .43- 50 40- .52 
Tibia Length  .66 2) 56 56 Syl .63- .69 Se off 
Tarsus Length .18 13 a3 X 13 sls Ie) .13- .16 
Antenna III 16 14 Alla ois 14 .15- .18 14- .18 
Antenna IV 07 06 06 xX 06 .07- .08 06- .08 
Antenna V 10 .09 09 XxX 10 O= el .09- 10 
Antenna VI aly sil SH/ S17, ale .16- 18 eiG=" altS 
Sensoria HI 3.63 D2 3733) 3.00 3.47 3.30-4.07 D2 TEF AT. 
Sensoria IV 25 00 00 00 00 .00-1.33 .00- .83 
Sensoria V 1.03 1.03 1.00 xX 1.03 1.00-1.07 1.00-1.27 
Sensoria VI 1.01 1.03 1.03 xX 1.00 1.00-1.03 1.00-1.07 
Islands V 2.44 Das eA. xX 2.13 2.07-2.77 1.73-2.80 
Islands VI PSI 3.20 ei I Xx 3.00 2.10-3.03 2.20-3.30 
Stem Mothers: 
Head Width 45 XxX XxX Al XxX .43- .53 xX 
Femur Length .41 xX xX 37 X 40- .43 X 
Tibia Length  .34 XxX xX iS) XxX 32- .39 xX 
Tarsus Length .14 Xx x X X lls “11; X 
Antenna III 16 x Xx eile xX .15- .18 xX 
Antenna IV lll xX XxX 12 XxX .10- .13 X 


Explanation: X-signs denote variables not used in the study indicated. 
All continuous variables are given in millimeters; all meristic ones as counts. 


Tue University ScrENCE BULLETIN 


196 


"NV 
(00) HLYOMNSAVSI 


‘SToquunu Aq] RIOT pue SITTPBIOT jo dey ‘T “94 


L, 


NY = 
JONIUMY 


i 
r-— 


“"NVY ‘OO SVI9NNOA 


ARIATION AND CovARIATION 197 


MicrocEocraPHic V 


“Y15UI] xD Ol Jo}DeIPYS oy IOJ sueoul APIO] jo dey GG “OL 


EN MAD alk 


Nvy b-- 
JONIYMV 1 


NVX ‘OO SV1IDNOG 


LSad " 
YNS SNVGAN NW» 


} 
(00 HLYOMN3AYSI | 
| 
| 
| 


188 Tue University ScteNcE BULLETIN 


EST 


MEANS SNK 


LEAVENWORTH CO, 
KAN 


x 
eS 
SS 


LAWRENCE 
—-7 KAN 


AY 
! RaQ 
Ls. 
DAY 


Yj 
Q 


Fic. 3. Map of average locality means representing factor I (see text for explanation). 


MicroGEOGRAPHIC VARIATION AND CovVARIATION 199 


a 
2S A 3 
O et 
op) 
pales 2 wHaernroa 
=A ov isis 
a 
(e) 
Ss 
Zee 2 
Loy (| 
= . 2 
=< = 
ui E 
os SS Fa 
| Vv 
| Z 
@ 
| ~ : : 
| \N KX ia 
I ‘es & 
N Y G 
| SSS ASRS OE 
SSE S 5 
iL s NSS Zs 2 
RX AG Se Ww = g 
SW \\ S < = = 
SS Qe a -d g 
Ss —n > 
| et 
Z J @ 
<x ov 
SZ 5 
fo) rs) 
O = 
n” a = 
< = ee 
= 
eo) Q 
= a | & 
Oo 
a Il 
Co | 


Tue University ScrENcE BULLETIN 


200 


*(uoneur[dx9 IoF 3x9} 998) [J] 10}9vF Sutussoido1 suvsur AW][eIo] Vde19Aev Jo dey “¢ ‘oly 


SW be =, 


“NV Ea 
JONIUMV 1 


c 
r-—- 


“NYY “OO SV1IDNOG 


ISaL “NV 
MNS SNVAN 
‘(00 HLYOMNSAVS1 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


GEOGRAPHIC VARIATION OF 
PEMPHIGUS POPULI-TRANSVERSUS 
IN EASTERN NORTH AMERICA: 


STEM MOTHERS AND NEW DATA 
ON ALATES 


By 
Robert R. Sokal and Paul A. Thomas 


' api r ; i Ti ee ue ‘ris i 
Marine Biorogica: LavOsdiwry 
Ll BRA rR 


Vout. XLVI Paces 201-252 OcroseErR 28, 1965 No. 5 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 201-252 Ocroser 28, 1965 No. 5 


Geographic Variation of Pemphigus populi-transversus in 
Eastern North America: Stem mothers and new data on alates'’” 


By 


Ropert R. SoKAL aNp Paut A. THomas? 
Department of Entomology, The University of Kansas, Lawrence 


ABSTRACT 


This paper continues earlier studies of the geographic variation and covaria- 
tion of morphological characters of the gall-making aphid Pemphigus populi- 
transversus in eastern North America (see footnote 1). Samples of galls from 
74 localities ranging from Massachusetts to Kansas and Wisconsin to Florida 
were employed in the study. Alate aphids were found in 57 of these localities. 
Eight alate characters, chosen to represent the three interlocality factors obtained 
by Sokal (1962), were measured together with six stem mother characters. 
Analyses of variance yielded significant interlocality variance components for 
all characters. The proportions of interlocality and intergall variance agree with 
previous results. As expected there is considerably more interlocality variance 
than in a study of microgeographic variation by Rinkel (1965). 

Separate correlation matrices were computed at the intergall and interlocality 
levels of covariation. The intergall correlations of alate characters agree with pre- 
vious findings. There are no significant intergall correlations between alates and 
stem mothers, a remarkable finding in view of their genetical identity. High 


"Part 5 of a study of variation in the aphid genus Pemphigus. Parts 1 to 4 are, respectively, 
the following publications: Sokal (1952), Sokal (1962), Sokal and Rinkel (1963), and Rinkel 
(1965). 

* Contribution No. 1218 from the Department of Entomology of The University of Kansas, 
Lawrence, Kansas. 

* This paper is based on aphid material collected and measured between 1949 and 1951 when 
the senior author was a graduate student in the Department of Zoology at the University of 
Chicago. Computations for the various analyses were carried out by the junior author. 

The authors acknowledge the assistance of Dr. F. James Rohlf and Mr. Richard C. Rinkel 
with computer programs employed in this work. Mrs. Julie C. Sokal drew the maps. The 
computations were carried out at The University of Kansas Computation Center. 

This research has been supported by The University of Kansas General Research Fund. The 
senior author completed this study while supported by a Public Health Service research career 
program award (No. 3-K3-GM-22, 021-01S1) from the National Institute of General Medical 
Sciences. The junior author completed this study during his tenure of a predoctoral fellowship 
from the Division of General Medical Sciences, United States Public Health Service. 

The authors wish to thank Dr. Raymond C. Jackson for a careful reading and criticism of 
the manuscript. 


202 Tue University ScIENCE BULLETIN 


interlocality correlations are found among stem mother and alate characters, 
and there are some important cross correlations between the two morphotypes. 
Multiple factor analysis with rotation to simple structure was carried out on 
both correlation matrices. Partial results are shown in Table 4. Four common 
factors account for most of the observed correlations at both levels of covariation. 
It is possible to homologize all but one of the factors affecting alates in the 
present study with those found in earlier work. New factors affecting stem 
mother characters emerge from this study. Maps of the geographic variation 
of the 14 characters investigated are shown in Figures 2-15. Locality means have 
been transformed to probits for ease of comparison. The differences among 
localities are tested by means of a multiple comparisons test (SNK test), the re- 
sults of which are shown in tables at the right hand margin of the figures. The 
method of representation of the present data is modified from the technique of 
Sokal and Rinkel (1963). The multiple factor analyses of interlocality correla- 
tions permitted the geographic variation to be summarized in four geographic 
variation pattern maps, each showing one interlocality factor (Figures 17-20). 
Techniques for categorization of infraspecific taxa are examined and illustrated 
in Figure 21. The implications of this study are discussed with reference to (1) 
consistency of geographic variation patterns based on earlier, more restricted 
samples, and (2) the correlation and absence thereof between stem mother and 
alate characters and possible genetic and ecological reasons to account for the 
observed phenomena. 


INTRODUCTION 


This paper continues the study of variation and covariation of mor- 
phological characters of the aphid, Pemphigus populi-transversus Riley, pub- 
lished in earlier papers of this series. The present study extends the analysis 
of geographic variation to 57 localities in eastern North America for eight 
morphological characters of the alates. These were selected by factor analysis 
from the 18 characters studied by Sokal (1962) and Sokal and Rinkel (1963)? 
in a smaller sample of localities. The study also introduces for the first time 
an analysis of the geographic variation of stem mother characters, six of 
which were measured. The number of localities with stem mothers is 74, 
which includes all those localities from which alates were also obtained, plus 
others in which only stem mothers were found. The method of representa- 
tion of geographic variation has been altered somewhat from that of Sokal 
and Rinkel (1963), a step taken for several reasons discussed below. Studies 
of variation and covariation of characters were again undertaken. In view 
of the extensive previous analyses of variation and covariation within galls, it 
did not seem necessary to repeat such work here; thus the basic unit of 
sampling and measurement was a gall, represented by one stem mother and 
one alate individual. 


' See erratum following Literature Cited. 


+ 


GeocraPHic VARIATION OF Pemphigus populi-transversus 203 


The general significance and implications of this work have been suf- 
ficiently discussed in the previous papers of this series (cf. Sokal, 1962; 
Sokal and Rinkel, 1963), and can be referred to there. The special points 
of interest and emphasis in the present paper follow. 

1. The study affords us an opportunity to compare variances in the 
earlier sampled 23 localities with the more extensive results presented below. 
We shall investigate whether the components of variance established in 
earlier work are similar in magnitude to those found here. 

2. A second point relates to comparison of the covariation pattern pre- 
viously established with that found below. This is done by comparing the 
previous and present results of factor analysis of correlations among alate 
characters. From the point of view of biological application of factor 
analysis, it is important to know whether the results of factor analytic work 
are replicable from study to study. This problem of factorial invariance is 
also of considerable importance in the field of factor analysis itself. 

3. We wish to examine the effect of increasing the number of locality 
samples on the stability of geographic variation patterns established earlier 
by Sokal and Rinkel (1963). 

4. An entirely new aspect of this study is the analysis of geographic varia- 
tion of the stem mothers. In addition to the intrinsic interest of such an 
analysis, its major importance is that it will permit the comparison of 
geographic variation in stem mothers and alates. It will be remembered that 
these two forms are isogenic (the alate is a diploid parthenogenetic offspring 
of the stem mother in each case). A priori it might be thought that char- 
acter correlations between the two forms should be as high as those within 
each form and that their geographic variation patterns should be concordant. 
It will be shown that this is not so, raising interesting problems of evolu- 
tionary adaptation as well as delineation of subspecific populations. We be- 
lieve this to be the first study of concurrent geographic variation and covaria- 
tion of characters of two distinct morphotypes in any species. 


MATERIAL 


A brief description of the complex life history of these aphids has been 
given in Sokal (1952). They form galls on petioles of the cottonwood, 
Populus deltoides, as the primary host and migrate to the roots of herbaceous 
plants in the summer. In the present study we employ both the fundatrix or 
stem mother and the alate fundatrigeniae found in the same gall with her. 
The term alate, whenever employed, refers to such forms only. 

The present study extends over 74 localities, based on a minimum of two 
and a maximum of 25 galls each. Each gall contains at least a stem mother 
and many of them contain alates as well. Since the number of galls per 


locality was unequal, some complications, to be discussed below, arose in the 


204 Tue University SCIENCE BULLETIN 


analysis of the data. A further complication was caused by the fact that some 
localities contained only stem mothers and no alates, while others contained 
both morphotypes. In order to obtain maximum information on stem 
mothers, all localities and all galls containing them were employed in the 
computation of means and variances of stem mother characters. These “com- 
plete” data were also employed in calculating the correlations among stem 
mother characters. However, for the computation of correlations among 
stem mother and alate characters all galls containing only stem mothers had 
to be eliminated from the study. Therefore such correlations could be 
computed only over the 57 localities which contained both stem mothers and 
alates. 

Figure 1 is a map of the eastern United States on which are plotted all 
studied localities. The localities and the dates of collection are listed in 
Table 1. The code numbers of the first 23 localities correspond with those 
of the earlier study (Sokal and Rinkel, 1963). In that study, however, 
localities 18, 19 and 20 each contained several small samples from the South- 
east, which had been combined to obtain localities of 15 galls each. Since 
other localities in the present study have unequal sample sizes, it was not 
necessary here to combine these localities and they are listed separately. In 
order to retain a consistent numbering system they have been numbered by 
adding a letter to the locality number, e.g. 18a, 18b, ete. The number of galls 
containing alates and stem mothers, respectively, are in Tables 5 and 6 of 
this paper. The number of alates measured per locality is generally less or 
at most equal to the number of stem mothers measured. Galls containing 
alates, but no stem mothers were excluded from the study. Locality 9 of 
the previous study was excluded since data on stem mothers were not avail- 
able. The means of alate characters in the present study will not agree 
exactly for any given locality with those given by Sokal and Rinkel (1963). 
This is due to the discrepancy in the number of galls employed in both 
studies and to the fact that only one alate was measured per gall in the 
present study as compared with two alates in the previous one. 

The single alate measured per gall was chosen at random from among 
those found in the gall. In the 23 localities of the previous study, one of the 
two alates was recorded at random from the data. 

Eight alate characters were chosen to represent the three interlocality fac- 
tors obtained by Sokal (1962). Their names and those of the stem mother 
characters are listed in the section on Description and Analysis of Geographic 
Variation and Covariation. Their definition has been given in the previous 
papers (Sokal, 1952, 1962). For brevity we shall use abbreviated names 
such as antenna III for antennal segment III length, and sensoria II for 
number of sensoria on antennal segment II]. Two of the stem mother 
characters described in Sokal (1952), tarsus length and last rostral segment 


GerocrapHic Variation oF Pemphigus populi-transversus 205 


length, were not employed in the present analysis. Of the eight alate char- 
acters, two are meristic and the others are continuous. All stem mother 
characters are continuous. The entire study involved 10,780 measurements 
of characters. 


ANALYSES OF VARIANCE 


The data were subjected to 14 analyses of variance, one for each character. 
The structure of the sample permitted computation of variance among 
localities and variance within localities (=variance among galls), with 73 
and 1036 degrees of freedom, respectively, for stem mother characters (56 and 
458 d.f. for the alate characters). At this stage of the study the analyses were 
considered as model II analyses of variance, and variance components were 
estimated for them. For the two meristic alate characters, sensoria III and 
sensoria IV, square root transformations were attempted to see whether these 
would affect the outcome of the analyses. No important differences in pro- 
portions of mean squares or F values were found; the data presented here 
are based on untransformed values. 

The results of the analyses of variance are in Table 2. Mean squares 
among localities are significant (P <0.01) for all characters, indicating the 
presence of a significant variance component among localities for every 
character measured. The first two columns in Table 2 show the percentages 
of the total variation attributable to the two levels of variation. The char- 
acters grade from forewing length, which has as much as 58°/ of its variance 
among localities, to antenna VI with only 11° interlocality variance. In 
general, alate characters, other than antenna VI and the meristic ones, have 
a high proportion of interlocality variance, while the stem mother characters 
are not as differentiated among the localities of the study, 13.5°/% for total 
length to 28.99% for femur length. 

In order to compare the present findings with earlier ones, it is necessary 
to study ratios of the variance components because in the earlier study three 
hierarchic levels had been investigated and the total percentage was thus 
divided three ways. Column 3 of Table 2 shows the ratio of the locality 
variance component over the within localities variance component of the 
present study. Since in the present study only one alate was measured per 
gall, an estimate comparable to the present within localities variance can be 
constructed from the components obtained by Sokal (1962) by adding the 
component of galls to that of within galls. The comparable ratio is given in 
column 4 of Table 2. The two sets of ratios are quite similar as might have 
been expected since the present study is based in part upon the data of Sokal 
(1962). However, it is reassuring to find that the sample of the smaller 
study was quite representative of the larger domain. The only noticeable 
discrepancy is in alate head width, which has a ratio twice that of the 


206 Tue Universiry SCIENCE BULLETIN 


previous study. The comparison cannot be made for stem mothers since 
these had not been studied previously. In the final column of Table 2 are 
shown similar estimated ratios based on the study of Rinkel (1965), in which 
variance components were calculated based on samples from a local area 
where the greatest distance between any two localities was eight miles. We 
note that the ratio in this study of microgeographic variation is much lower 
and that some of the interlocality variance components were not significant, 
as might have been expected in samples so close to each other. 


CORRELATIONS 


Correlations among the 14 variables of this study were computed on an 
IBM 650 digital computer. The structuring of the sample permitted the com- 
putation of correlations among galls within localities (intergall correlations) 
as well as among localities. Since the lowest sampling unit was galls within 


localities, the intergall correlations are product-moment coeficients. The — 


| 


interlocality correlations were computed both as product-moment and as_ 
component correlations (Sokal, 1962). Only the component correlations 


are shown here to conserve space. Subsequent analyses at the interlocality 
level are all based on the component correlations. 

The computations of the correlation coefficients for the stem mothers are 
based only on those galls which also had alates. Correlations among all the 
stem mothers were also computed but were not appreciably different from 
those of the more limited sample, hence are not given here. 

The present data permitted for the first time a correlation of stem mother 
characters with alate characters, both on an intergall and an interlocality 
basis. The only earlier attempt at such correlations was based on a single 
locality (Sokal, 1952) and was therefore not very representative. 

Table 3 shows the correlation coefficients with the intergall product- 
moment correlations below the principal diagonal and the interlocality com- 
ponent correlations above the principal diagonal of the matrix. 


INTERGALL CORRELATIONS 


The intergall correlations are not very high in absolute magnitude. Most 
high ones involve continuous alate variables, but not including antenna VI. 
Correlations among stem mother characters are only moderately high, except 
femur against tibia length. There are no high correlations between stem 
mother and alate characters. 

In comparing these correlations with the intergall product moment cor- 
relations found by Sokal (1962), we find a considerable amount of agreement. 
An average difference of -0.05 is found between the corresponding correla- 
tions of the earlier and the present study, indicating that the earlier correla- 
tions were slightly higher. However, this difference is only slightly greater 


GEOGRAPHIC VARIATION OF Pemphigus populi-transversus 207 
gy 


than one standard error of the correlation coefficients. It should be pointed 
out that the intergall product-moment correlations of the earlier study were 
based on galls containing two alates each, while galls in the present study 
contain only one alate each. Thus the intragall variance component in the 
earlier study was only half of that in the present one. The thorax-head width 
cluster of intergall correlations illustrated in Sokal (1962) is again shown in 
these data, although in a more limited manner since fewer variables are 
employed. 
INTERLOCALITY CorRELATIONS 


The interlocality component correlations are shown above the diagonal 
in Table 3. They are generally much higher than the intergall correlations 
and comprise the continuous variables of the alates, excluding antenna VI 
which has high correlations only with tibia length and antenna II. High 
correlations are also found among most stem mother characters, although 
stem mother total length is correlated with the other characters only in a 
limited way, and antenna IV (which is homologous to antenna VI in the 
alates) is quite independent from the rest of the measurements and nega- 
tively correlated with most (except with antenna VI with which it has a 
correlation of 0.60). There are some interesting high cross correlations be- 
tween stem mother and alate characters in addition to those involving stem 
mother antenna IV, which is also highly negatively correlated with forewing 
length (r=-0.71). Most important cross correlations involve stem mother 
head width which is highly and positively correlated with the continuous 
variables of the alates. 

Comparing the interlocality component correlations with those computed 
by Sokal (1962), we find only minor differences, the average being only 
0.08 which in view of the lower number of degrees of freedom is not a sig- 
nificant difference. 

A clustering of the correlation coefficients similar to that carried out by 
Sokal (1962) yielded a wing-thorax-antenna III cluster, more or less cor- 
responding to the pattern found by Sokal except that these patterns were not 
too distinct since there were fewer characters employed. A marked change 
was noted in the correlation between sensoria II and IV which has a value 
of 0.97 in the present study. 

The relations among the variables will be put into clearer focus and also 
simplified as a result of the factor analyses. 


FACTOR ANALYSES 


The intergall product-moment and _interlocality component correlation 
matrices were next subjected to multiple factor analysis with rotation to 
simple structure. A brief explanation of factor analysis, an account of its 


208 Tue University SciENCE BULLETIN 


application to biology, and a description of the computational steps carried 
out are given in Sokal (1962), where identical techniques were employed. 
The results are expressed in the form of primary pattern matrices of pattern 
coeficients (Harman, 1960), representing the standard partial regression 
coefhcients of each variable (character) on each factor. The two primary 
pattern matrices are not shown here to conserve space. They have been re- 
produced in mimeographed form, together with the correlation matrices 
among primary factors, and can be obtained by writing to the senior author. 
Table 4 gives the primary pattern matrices as a scheme of + and — signs 
representing only the important pattern coefhcients. This permits a relatively 
simple and instructive picture of the nature of the factors. We note that four 
common factors account for covariation in both of the correlation matrices. 
In Table 4 the factors are first of all numbered in sequence, but underneath 
these numbers are given other numbers homologizing the factors with those 
of the earlier study (Sokal, 1962) or a new symbol if they have not been 
encountered before. 

The intergall product moment correlations yielded a satisfactory simple 
structure. Factors I and II affect only alates and do not seem to affect stem 
mother characters at all. In trying to homologize factor I, we note that its 
loadings are an amalgam of factors I and II of the intergall component and 
the intergall product-moment primary pattern matrices obtained by Sokal 
(1962). These two factors of the earlier study appear to have collapsed here 
into a single factor. This should not occasion surprise since the earlier factors 
I and II were extremely highly correlated (r=0.93 in the intergall product 
moment matrix and r=0.71 in the intergall component matrix). Factor I 
therefore represents a size factor affecting the continuous, but not the meristic, 
variables of the alates. The second factor affecting alates cannot be easily 
identified. It is clearly an antennal factor, strongly influencing antenna Il 
and sensoria II, a combination which did not occur before; it also weakly 
affects sensoria IV (not shown in Table 4). We shall call it intergall 
factor IV. 

The other two intergall factors exclusively influence stem mother char- 
acters. Factor II (henceforth called F-1, meaning intergall fundatrix factor 
1) moderately affects stem mother total length and head width. Factor IV 
(henceforth called F-2) affects the appendages and antennae of stem moth- 
ers. There are no important correlations involving the primary factors. 

The interlocality component factors can be more easily homologized than 
the intergall factors with the results of the previous analysis by Sokal (1962). 
Factor I affects sensoria III and sensoria IV strongly and antenna III mildly. 
This is a similar pattern to that of interlocality factor I of the previous study. 
Interlocality factor II affects antenna VI strongly and mildly affects head 
width, thorax length and tibia length (not shown in Table 4). This re- 


GerocraPHic VARIATION OF Pemphigus popult-transversus 209 


sembles interlocality factor II in the previous study, except that in that study 
head width loaded higher on the factor than did antenna VI. A new feature 
in this factor is its strong loading on stem mother antenna IV. This char- 
acter is homologous to antenna VI in the alates (being in each case the 
terminal segment of the antenna), and it is interesting that geographic varia- 
tion patterns for these two characters would coincide when no similar inter- 
gall correlations between homologous body structures in stem mothers and 
alates are found. Interlocality factor II strongly affects forewing length, 
head width, thorax length, tibia length and also antenna VI. Except for 
the latter loading this factor represents factor HI of the 1962 study. Factor 
HI also affects total length of stem mother weakly (not shown in Table 4). 
There is appreciable negative correlation between interlocality factors I and 
HI (r= —0.68), which was not the case in the previous study. Interlocality 
factor IV (henceforth called interlocality factor F-1) is limited to stem 
mothers and controls stem mother femur, tibia and antenna IIT length and 
mildly also stem mother total length and head width (not shown in Table 
4). This factor is quite uncorrelated with the other factors. 

Thus interlocality component correlations can be accounted for to a large 
degree on the basis of four factors, three of which are mainly alate factors 
(but two of these affecting stem mother characters also), with the fourth 
being exclusively a stem mother factor. 

The interlocality communalities of all variables save stem mother total 
length are very high so that these factors account well for the geographic 
variation at hand. This means also that four independent trends will ade- 
quately describe the geographic variation of all alate and stem mother 
characters of this study. 


DESCRIPTION AND ANALYSIS OF GEOGRAPHIC VARIATION 
AND COVARIATION 


The means of the alate and stem mother characters for all the localities 
studied are given in Tables 5 and 6, respectively. Means are stated in milli- 
meters for continuous variables and as counts for meristic characters. Sample 
sizes for the stem mother means are generally greater than those for alates 
because some galls contained stem mothers but no alates. Furthermore there 
are more localities given in Table 6 than in Table 5 since 17 localities in the 
study had no alates at all. The standard deviation at the foot of each char- 
acter column is the average standard deviation within localities. It is not 


_the standard error of the means. An average standard error of the means is 


difhcult to compute because of the variation in sample size. For a working 
average standard error for any character one might divide the standard 
deviation given in the table by the square root of the average number of 
galls per locality, calculated as shown in the explanations to Tables 5 and 6. 


210 Tue University ScrENCE BULLETIN 


In addition to tabulating the means, graphic descriptions of geographic 
variation on distribution maps are necessary. Maps depicting geographic 
variation patterns should furnish information on two issues. They should 
communicate the magnitude of the changes of character means over a given 
area, and this is achieved by different symbols for different levels of the 
means or by the drawing of isophenes based on a more or less exhaustive 
sampling of the area. The second consideration is some measure of the 
reliability of the means and concomitantly some measure of testing signifi- 
cance of differences between means. These two statistical demands on a 
map, description and hypothesis testing, are to some degree in conflict, mak- 
ing the presentation of a suitable map something of a compromise. In their 
earlier study of Pemphigus, Sokal and Rinkel (1963) used probit scores to 
indicate the magnitude of the means and used shading as well as a diagram- 
matic representation of a multiple comparisons test to give some idea of the 
statistical differentiation among the means. 

The probit scores for the means are obtained by dividing the difference 
of each mean from the grand mean of localities by the standard deviation of 
locality means and adding 5.0. Localities whose means are less than the 
grand mean will have scores of less than 5.0; those with a mean larger 
than the average will score above 5.0. Decimal points have been omitted for 
simplicity. Since the standard deviations differ, depending on the number 
and kinds of means included in the study, the probit scores of the previous 
23 localities in the present study are not identical to those calculated by 
Sokal and Rinkel (1963). However, their relative magnitudes remain the 
same. 

In the present study some changes in the representation of geographic 
variation were instituted for several reasons. We are in a period of experi- 
mentation in this type of work, and it is desirable to attempt a number of 
different approaches in order to arouse the interest of students of geographic 
variation in these problems. We abandoned the shading system employed 
earlier for two reasons. We had some qualms from the very beginning about 
the reality of the areas defined by the relatively few points of that study, 
although these were more numerous than in many studies of geographic 
variation in which the drawing of isophenes is therefore likely to be even 
more open to doubt. Some of the boundaries for areas of supposedly uniform 
character means are thus likely to be quite arbitrary. Secondly, since the 
shading of the maps by Sokal and Rinkel was based on the results of multiple 
comparisons tests, i.e. test of significance, the division into separate shades 
often made for quite unequal classes either in frequency or in range of the 
variable. On the present maps we decided to represent the magnitude of the 
mean by a probit score, as before, and to reinforce this measure of the mag- 
nitude of the character by a circle covering the geographic locality from 


GeocrapHic VARIATION oF Pemphigus populi-transversus 211 


which the sample was taken, shading this circle in one of four shades repre- 
senting four groups of probit scores. These groups do not have equal class 
intervals, being set up so that scores greater than 63 are colored black, those 
between 51 and 63 are dark gray, between 37 and 50 are light gray and those 
less than 37 are white. These classes are based on an expected normal distribu- 
tion of the probit scores in which the extreme classes represent 10°% of the 
distribution at each tail and the two central classes make up 40°7, each. Since 
the means are not always normally distributed, the proportions in these 
classes are sometimes not as expected. By drawing a circle with a radius of 
27 miles over the collection site but refraining from drawing any isophenes, 
we feel that we are presenting a less biased picture of the geographic varia- 
tion than before. The large number of localities on the map form a pattern 
as the entire map is inspected and areas of high and low means become 
apparent to the viewer. 

The following are brief accounts of the geographic variation patterns 
of each character as illustrated by Figures 2-15. Alate characters are dis- 
cussed first, followed by stem mother characters. 

Forewing length (Figure 2). In general an area of low means occurs in 
the western half of the area studied, with some low means also in New 
England, New York and Pennsylvania. High means are most prominent in 
southern Ohio, and in the Southeast but some high means occur west of the 
Mississippi also. 

Head width (Figure 3). The lowest means are found in a triangle 
bounded by east-central Kansas, northern Missouri, and eastern Arkansas. 
High values cluster in Ohio, Indiana, northern Illinois, and extend southeast 
through Kentucky and Tennessee into Georgia. More high values seem to 
occur in the northern part of the distribution. 

Thorax length (Figure 4). The highest means cluster in Ohio, Indiana, 
Kentucky, and Tennessee. In general all localities east of the Lower Missis- 
sippi and the Wabash valleys are high, while those west of this line are low. 
A single high locality is found on the border between Iowa and Missouri. 

Tibia length (Figure 5). This resembles the pattern in Figure 4 with 
high means to the east of the Wabash-Lower Mississippi line and low means 
to the west of this line. Locality 13 in Decatur County, Iowa, high for the 
last character, is even higher for tibia length. 

Antennal segment III length (Figure 6). This pattern is related to the 
previous ones but less clear in that some high localities extend into Illinois 
and Missouri, while some lower ones occur in the South, northern Pennsyl- 
vania, and Ohio. 

Antennal segment VI length (Figure 7). This character shows a general 
north-south trend with high means in the North reaching south as far as 
Kentucky and Tennessee, and low means generally in the South and South- 
_ west although some low means reach as far north as central Indiana. 


212 Tue University SciENCE BULLETIN 


Number of sensoria on antennal segment III (Figure 8). This variable 
shows a peculiar distribution with high values in Ohio, some extending into 
Illinois, Indiana and Iowa, and low values concentrated in the South and 
Southeast and extreme North. There is an isolated very high mean in south- 
east Missouri. 

Number of sensoria on antennal segment IV (Figure 9). The pattern 
here resembles that of the previous character, but because of a strong skew to 
the right there are fewer high localities. The New England means are 
higher than before, and the locality in Wayne County, Missouri, high for 
the previous character, is quite low for this one. 

The stem mother characters are described in Figures 10-15. 

Stem mother total length (Figure 10). This character has a difhicult-to- 
interpret pattern of variation. The highest means are found along a north- 
south axis from central Indiana to Mississippi with an extension into Florida, 
but the low means are scattered through various regions without any mean- 
ingful pattern. The distribution pattern resembles somewhat that of stem 
mother tibia length reported below. 

Stem mother head width (Figure 11). This character features low means 
in the West, high means in the Mississippi Valley and generally east of the 
Mississippi, although low means are again found in New England. 

Stem mother femur length (Figure 12). This resembles the previous 
character with some low means in New England, some in western Kansas, 
but on the whole a fairly confusing pattern in which the highs and lows are 
distributed almost in crazy quilt fashion over the area of study. 

Stem mother tibia length (Figure 13). This pattern resembles closely the 
one just studied and is rather confusing. The highest means are found along 
a north-south axis through the middle of the map with an extension into 
the Great Lakes and northern Pennsylvania area. New England is generally 
low, as is Kansas and northern Missouri. 

Stem mother antennal segment III length (Figure 14). The pattern re- 
sembles that of the two previous characters. 

Stem mother antennal segment IV length (Figure 15). This character 
is distributed quite differently from the stem mother characters discussed so 
far. There is an area of high means in the North with the highest ones in 
the northwest of the area. With very few exceptions, means south of the 
level of central Ilinois are low. There is a single high locality in Georgia. 

Sokal and Rinkel (1963) employed the Student-Newman-Keuls (SNK) 
test to compare the means of the localities after the analysis of variance had 
been carried out. Because the present data are based on unequal sample 
sizes and also because of the large number of means involved, systematic test- 
ing of differences between means proved to be a very tedious procedure. For 
this reason we developed a computer program for the IBM 1620 computer 
(written in FORTRAN and available upon request from the authors) which 


GeocraPHic VaRIATION OF Pemphigus populi-transversus 216 


makes multiple comparisons among means according to the SNK test as well 
as by Duncan’s multiple range test, for both equal and unequal sample sizes. 
The formulae for these tests are in Steel and Torrie (1960). To gain an 
understanding of the workings of the SNK tests and to compare the effects 
of using unequal sample sizes with employing an average sample size for all 
localities, we processed all data by both methods. Performing the tests with 
unequal sample sizes proved unsatisfactory because it broke the continuity 
of ranges enclosing nonsignificant sets. With equal sample sizes, once a 
higher mean B has been shown significantly different from a lower mean A, 
a mean C (where C > B) will be significantly different from the lower 
mean A. There are exceptions to this rule, but they are rarely encountered in 
practice. With unequal sample sizes, however, it is quite possible that mean 
B (B> A) can be significantly different from A and yet mean C (C > B, but 
based on a smaller sample) would not be significantly different from mean 
A. Such relations are, of course, warranted by statistical theory; a mean based 
on a small sample is not easily proven significantly different from another 
mean. Yet the arraying of means and the creation of overlapping ordered 
groups is made impossible by such relations. For this reason we investigated 
the amount of error which would arise if we were to base the SNK test on an 
average sample size and employ the formula for equal sample sizes. Stem 
mother total length was analyzed by SNK tests for equal size employing 
the formula for 7» in the explanation of Table 5. Parallel tests were carried 
out using the actual sample sizes mi for each locality. Figure 16 illustrates 
the results of this comparison. It is a half matrix which shows the probit 
scores of the means for stem mother total length along the principal diagonal 
and the left hand margin. The sample sizes of each locality are given at the 
bottom of the matrix. In order to find out whether any given mean is sig- 
nificantly different from the other we look up the lower of the means along 
the principal diagonal and run down the column until we reach the level 
of the higher mean along the left hand margin of the matrix. In case of 
several probit means of equal score employ the mean with the appropriate 
sample size. If both ends are marked by an X the two means are not sig- 
nificantly different (based on an SNK test with unequal sample sizes) ; if the 
higher mean (the lower end of the column) lacks an X the two means are 
significantly different (at P <0.01). The black line which cuts across the 
graph in zig-zag fashion is the boundary of not significantly different sets of 
- means based on equal sample sizes and average sample size 70; if the higher 
mean is within this boundary it is not significantly different, if beyond it is 
different from the mean at the head of the column (P <0.01). In general 
the boundary line assuming equal sample sizes is similar to decisions based 
on unequal ones. If the per cent error of using equal sample sizes is computed 
_as the percentage of comparisons wrongly included and wrongly excluded 
_ by the boundary (analogous to combining type I plus type II error) we find 


| 
| 


214 Tue University SciENCE BULLETIN 


that 6.96% of all possible comparisons between means would be in error. ‘This 
is clearly higher than the 1°% type I error assumed in the tests, but is not 
prohibitively large in view of the heavy cost in extra work and complexity 
which would be exacted by using unequal sample sizes. 

The standard error used in the SNK tests was the square root of the 
error mean square obtained in the analysis of variance. Since the q tables 
in Pearson and Hartley (1958) do not go beyond a 7 value of 20, we had to 
extrapolate values for the higher degrees of freedom necessary in this 
analysis.’ 

Since our computer program calculated Duncan’s multiple range test 
along with the SNK, we were able to compare the results of the two tests 
for a number of variables. We confirmed the findings of Sokal and Rinkel 
(1963), that the Duncan test provides a greater amount of differentiation 
among means than is desirable from the point of view of our analysis. 

The significance of differences between pairs of means is not evident 
from the maps proper, but can be learned from their right hand margin. In 
the previous paper Sokal and Rinkel (1963) used the common convention 
of overlapping lines to depict the results of the multiple comparison tests. 
When we attempted these on the present data, we found that the number of 
lines necessary for such a large sample was so great that it became almost 
impossible for the eye to follow any one line along the measurement scale. 
The convention adopted here seems more useful. The ranges provided indi- 
cate sets of not significantly different means at a probability level of 0.01. 
Thus if we wish to know in Figure 2 whether two localities, one with a mean 
score of 33, the other with a mean score of 57, are significantly different from 
each other, we find in the margin that no such inclusive range exists, hence 
the two means are significantly different at P <0.01. 

In order to summarize the information on geographic variation presented 
in the 14 maps for separate characters, we obtained factor scores for the 
four interlocality factors described above by averaging the scores of the two 
characters best representing each factor. The rationale behind this technique 
and details of computation are given in Sokal and Rinkel (1963). To repre- 
sent factor I we used sensoria III and sensoria IV; to represent factor II we 
used antenna VI and stem mother antenna IV; factor III was represented by 
head width and thorax length, while the stem mother factor F-1 is represented 
by stem mother femur length and stem mother tibia length. Averages of 
the standardized scores of the two characters for each factor were computed 
for each locality and plotted on maps (Figures 17-20). 

Examination of the geographic distribution of the average scores reveals 
four distinct patterns of variation. 

Interlocality factor I (Figure 17). The distribution of this factor is 


' Subsequently we discovered the table by Harter (1960) which furmishes 7 values up to 100. 


Gerocrapuic Variation oF Pemphigus popult-transversus 215 


peculiar in having high means in Ohio, Indiana, and Illinois and also three 
high localities in New England. The low localities are found in the far north 
and generally in the south, southeast, and southwest of the area studied. 

[nterlocality factor II (Figure 18). This shows a general north-south 
gradient with high values in the North reaching as far south as Kentucky and 
Tennessee, and low values in the south, southwest, and southeast of the area 
with low means extending northward along the Mississippi, Wabash, and 
Ohio River valleys. 

[nterlocality factor III (Figure 19). The low means are generally found 
west of the Lower Mississippi and Wabash with lowest values in Kansas, 
Missouri, and Arkansas. High values are generally in the East with highest 
means found in Indiana, Ohio, Kentucky, Tennessee, and Georgia. The 
northern and eastern portions of the area seem to have the higher means in 
general. 

[nterlocality stem mother factor F-1 (Figure 20). This factor provides 
a very confusing picture, as do the stem mother characters on which it is 
based. No clear trends over the entire geographic area can be seen but there 
are patches of reasonably homogeneous means. In general, means are low 
west of the Mississippi. High means cluster in the Mississippi River Valley, 
but there are also high means in southern Ohio, Alabama, and North Caro- 
lina. A second low group is found in the Northeast. 


DISCUSSION 
"TECHNIQUES 


This study provides a further illustration of the application of analysis of 
variance and multiple comparisons tests to the study of geographic variation. 
New problems introduced by the data at hand include the great number of 
localities to be analyzed and compared as well as unequal sample sizes. The 
numerous localities make multiple comparisons among means rather labori- 
ous although not nearly so tedious as separate comparisons by ¢ tests between 

all possible pairs of means would have been. Representation of the results 
required abandonment of the earlier scheme of representing nonsignificant 
sets by a line (Sokal and Rinkel, 1963), however, the system adopted in this 
paper has proved quite useful in checking the significances of differences 
between means as was done during the preparation of this paper. 

The analysis of variance of numerous localities does require heavy com- 
putation but not more so than the separate computation of means and vari- 
ances for each locality would involve. If, as in this study, many characters are 
to be measured and analyzed, the use of a digital computer is recommended 
to insure speed, accuracy, and economy in computation. 

The unequal sample sizes do not present difficulties during the computa- 


216 Tur University SCIENCE BULLETIN 


tion of the analysis of variance; simple changes in the computational formulae 
allow for such differences. However, the computation of multiple compari- 
sons tests is, as we have seen, considerably slowed down and furthermore 
becomes difficult to interpret. Our experience in the present case has shown 
that substituting average sample size mo for n in the SNK technique for equal 
sample sizes does not result in conclusions so wide of the mark as to be 
useless. The present data are a rather severe test in this regard because several 
localities in the study were based on only two galls (see, for instance, Figure 
16 where the locality with the score of 71 based on two galls cannot be shown 
to be significantly different from any other one). In data such as the present 
ones, where sample sizes differ drastically, certain critical comparisons be- 
tween means may have to be tested by separate comparisons tests based on 
the formula for unequal sample sizes. 

The employment of factor analysis to summarize covariation of char- 
acters among localities has again proved to be an adequate way of represent- 
ing the variation patterns. The four interlocality factors are quite different 
in nature, although I and IH, and II and III showed substantial correlations. 
They presumably represent different adaptational trends. 

There is little evidence of specific geographic variation trends shown 
only by a single character. Even stem mother total length, the variable with 
the lowest interlocality communality, shows no distinct geographic variation 
pattern. 

ConsISTENCY OF ALATE CHARACTERISTICS 


In describing patterns of geographic variation, the degree to which the 
sample studied represents the area covered is always of interest. Comparison 
of geographic variation of alate characters in the present and previous studies 
affords an opportunity to gauge the consistency of pattern. The new patterns 
for six of the alate characters are quite consistent with their previous de- 
scription. The addition of further localities has brought more intergradations 
to the new maps, yet, as a rule, localities with high means have not been 
placed in areas of low means, or vice versa. Even some peculiarities of 
pattern such as the tongue of low means for forewing length protruding 
into western Ohio has been maintained and reinforced in the present study 
(cf. Figure 3 in Sokal and Rinkel, 1963, with Figure 2 in this paper). Such 
differences as exist between the patterns of the previous and present studies 
largely refer to the so-called area samples of Sokal and Rinkel (1963) which 
were composite localities based on several small subsamples. These have 
been separated in the present study, and in at least one case, the Kentucky and 
Tennessee area sample, the separate localities show quite different means. 
Additional localities have also demonstrated that hypothetical narrow “land 
bridges” bounded by isophenes, such as the one in Sokal and Rinkel (1963, 
Figure 9), extending west from Ohio into northern Indiana and southern 


Geocrapuic Variation or Pemphigus populi-transversus 217, 


Iowa are not correct. In Figure 5 of this paper we can see that the high 
mean for tibia length in Decatur County, Iowa (locality 13), represents an 
isolated pocket or perhaps the easternmost representation of a hypothetical 
area of high means extending through Nebraska and Minnesota from which 
samples have so far not been available. 

Two of the characters, sensoria III and sensoria IV, show appreciable 
changes over their previous distribution patterns. The present pattern of 
variation of sensoria HI is somewhat clearer and more meaningful than that 
published by Sokal and Rinkel (1963). The former narrow band of high 
means extending westward has expanded into a general area of highs in the 
Midwest with a thin belt of low means north of this area and lows to the 
south and southwest. An unexpected high locality in southeast Missouri is 
encountered. The variation pattern of sensoria IV which previously had two 
high mean enclaves in Illinois and northern Ohio (Figure 16 in Sokal and 
Rinkel, 1963) now has a more extensive high area in the Midwest with a low 
belt in the North and a large low area in the South. It is interesting that 
locality 37 in Wayne County, Missouri, with an extraordinarily high mean 
for the previous character has a low score for this character, although in 
general the two are quite highly correlated. 

Comparing the geographic variation of factors in the previous and present 
studies, we find that interlocality factor I has changed the most, as have 
the two characters on which it is based (sensoria III and sensoria IV dis- 
cussed above). Still this is only a relative change and the high area in Ohio 
extending west through Illinois into Iowa is again represented. The low 
area in the far West and the Southeast is also clearly shown. The variational 
pattern of interlocality factor II is very consistent with that of the previous 
study, showing a high area in the North, an intermediate area in the South- 
west and Southeast, and a low area in the Mississippi River Basin. Similarly, 
interlocality factor III shows high values through Ohio, southern Indiana 
and the Southeast and an area of lows largely west of the Mississippi. Here 
again the two patterns are quite compatible. 

We have already seen that the correlation coefficient matrices based on the 
alate characters show considerable resemblance to those found in the earlier 
study (Sokal, 1962). This supports the reliability of the findings presented 
here. 

AGREEMENT BETWEEN ALATES AND STEM Mortners 


Of special interest in the present study is the simultaneous analysis of two 
separate morphotypes, the stem mothers and their alates. This comparison 
is carried out at two levels of variation and covariation. The contents of one 
gall represent one genotype, since the stem mother and her alate offspring 
are isogenic. Intergall variance can be attributed to genetic differences among 
founding stem mothers as well as ecological differences among the galls of 


218 Tue Unrversiry SciENCE BULLETIN 


a given locality. One might expect that the characters of the stem mothers 
and of the alates should be highly correlated on an intergall basis. In fact, 
however, not a single appreciable intergall correlation coefficient was found 
between the two sets of characters, the highest r value being 0.26. We must 
therefore conclude that whatever factors cause the characters of stem 
mothers to vary from gall to gall do not affect the studied characters of 
alates (and vice versa). 

What theoretical basis can we assume for such a phenomenon? Two 
explanations come to mind: (1) To the extent that intergall variance is 
genetic, different loci may come into play to affect the two suites of char- 
acters chosen. Thus although the two morphotypes are isogenic, the varia- 
tion patterns observed would be due to different sets of active loci. (2) The 
ecological factors differ from the time that morphogenetic forces are at work 
in the stem mother to the time that such forces affect the morphogenesis of 
the alates. We shall now examine these hypotheses in more detail. 

Evidence is mounting from a study of several organisms that gene action 
is selective both in site as well as in time of action. The phenomenon of 
puffing of chromosome bands (e.g., Clever, 1961), indicates that different loci 
come into play during various stages of an organism’s development. It would 
therefore not be too surprising to find a similar phenomenon operating in 
this instance. Supposedly homologous characters in the stem mother and 
alates might respond to quite different gene products. 

The environments to which these two morphotypes are subjected will also 
be considerably different. The stem mother is formed much earlier in the 
year than the alates, resulting in grossly different climatic experiences. 
Derived from a fertilized egg which hatches early in the spring, the stem 
mother makes her way to the young leaf bud on a cottonwood twig. Her 
early history is thus quite different from the alate born parthenogenetically 
within the considerably more homeostatic confines of a gall. In the very early 
history cf the stem mother (first and possibly second instars) the term inter- 
gall environmental factor has no real meaning because the gall has not yet 
been established and such a level of variation reflects only the differences in 
the individual histories of the stem mother nymphs. After the stem mother 
nymph has settled on the petiole of an unfolding leat and has begun to 
induce the gall, we can talk meaningfully of intergall variation. The early 
environment of the alates by contrast is within the body of the stem mother, 
while later development takes place in the mature galls. 

In saying that the genetic or environmental factors affecting one stage 
do not act on the second stage, we do not mean to imply categorically that 
these factors have no effect on the other morphotype, but only that the char- 
acters examined by us in the second morphotype are not under control of 
these factors. We cannot at this time distinguish between the two hypotheses. 


Grocrapnic Variation oF Pemphigus populi-transversus 219 


It is more than likely that both genetic and environmental influences come 
into play. 

At the interlocality level we again cannot separate genetic from environ- 
mental covariation. Genetic differences among populations are likely to be 
adaptive and thus may be a secondary consequence of climatic environmental 
differences which, however, may affect the phenotype of the organisms in 
a direct primary way. More likely the organisms will be ecotypically different 
and thus be responding to a complex interaction of genetic and environmental 
differences presumably established at some time past by a mechanism such 
as genetic assimilation. 

At the interlocality level a number of large and significant correlation 
coefhicients are shown. These involve stem mother head width and stem 
mother antenna IV more than the other four stem mother characters 
measured. Rather than discuss these correlations in detail we might refer 
to the primary pattern matrix for interlocality components (Table 4) which 
shows only the most important of these cross correlations, namely the one 
between stem mother antenna IV and alate antenna VI. The actual pattern 
coefhcients show other appreciable correlations as well. While antenna IV 
in the stem mother and antenna VI in the alate are homologous segments, it 
seems unlikely that these correlations occur because of morphological cor- 
respondence. We have seen that such relations do not hold at the intergall 
level where morphogenetic patterns of this sort should have considerably 
more influence. The stem mother-alate correlations must mean that the 
geographical differences to which these characters are responding are likely 
to be climatic factors, adaptation to which occurs independently in both 
morphotypes of these populations. Thus the sensory organs of the last seg- 
ment of the antenna in the stem mother may be responding to the same 
selective forces affecting the last alate segment. In noting this correlation 
we should not lose sight of the fact that there is considerable difference be- 
tween the patterns of geographic variation of stem mothers and alates for 
other characters. 


CATEGORIZATION OF INFRASPECIFIC TAXA 


At the conclusion of their study Sokal and Rinkel (1963) endeavored to 
characterize infraspecific populations by combinations of factor scores for 
the three interlocality factors which they found in their study. Their method 
has been simplified here. In lieu of plotting all localities with appropriate 
shading representing the characterization, we have simply drawn a relatively 
arbitrary boundary between the high and low means of an area. In drawing 
this boundary we were primarily guided by the solid black and the white 
circles in the factor variation maps (Figures 17-20), but we also considered 
the general clustering of localities with intermediate shades. Thus the lines 
drawn in Figure 21 represent approximate boundaries between high and low 


220 Tue University ScrENCE BULLETIN 


areas. While undoubtedly the great majority of localities on the two sides 
of the boundary are not statistically significant, the highest and lowest locali- 
ties are so in each case, as can be seen from an inspection of the SNK test re- 
sults in the right hand margins of Figures 17-20. The boundaries shown in 
Figure 21 are for interlocality factors I, Il, and II. The crazy quilt distribu- 
tion of interlocality factor F-1 makes it impossible to draw a single line 
dividing high from low localities. We have therefore copied the extreme 
means (black and white circles) from Figure 20 onto this map. Ignoring 
the areas for factor F-1 we arrive at six distinguishable areas identified by 
lower case or capital Roman letters to indicate the low or high states of the 
factors, respectively. Each of these populations is significantly different by 
at least one factor score from every other one and some are different by as 
many as three factor scores (as for example the population in northern Ohio 
from the population in southeastern Missouri and Arkansas). 

The implications of such a categorization for the so-called subspecies 
controversy are discussed elsewhere (Sokal, 1965). It is obvious that further 
study of the organism involving more characters would result in a greater 
number of factors and yet more drastic subdivision of the area. Also since 
Pemphigus populi-transversus is supposedly distributed over much of the 
North American continent (an extensive study of this is in the planning 
stage) the amount of diversification and possibilities for defining distinguish- 
able subpopulations are likely to be great. 


ERRATUM 


A regrettable systematic error has been discovered in Table 2 of the 
earlier paper in this series (Sokal and Rinkel, 1963). A corrected version 
of the table is reproduced after Table 6 of the present paper. 


LIGERATURE CHEED 


Ciever, U. 1961. Genaktivititen in den Riesenchromosomen von Chironomus tentans und 
ihre Beziehung zur Entwicklung. I. Genaktivierungen durch Ecdyson. Chromosoma 
(Berl.) 12:607-675. 

Harman, H. H. 1960. Modern factor analysis. Univ. Chicago Press. 469 p. 

Harter, H. L. 1960. Tables of range and studentized range. Ann. Math. Stat. 31:1122-1147. 

Pearson, E. S., AND H. O. Harttey. [ed.], 1958. Biometrika tables for statisticians. Vol I. 
Cambridge Univ. Press, Cambridge. 240 p. 

RINKEL, R. C. 1965. Microgeographic variation and covariation in Pemphigus populi-transversus. 
Univ. Kansas Sci. Bull. 46:167-200. 

SoxaL, R. R. 1952. Variation in a local population of Pemphigus. Evolution 6:296-315. 

1962. Variation and covariation of characters of alate Pemphigus populi-transversus 

in Eastern North America. Evolution 16:227-245. 

1965. Pemphigus and the subspecies controversy. (Manuscript in preparation.) 

ano R. C. Rinket. 1963. Geographic variation of alate Pemphigus populi-transversus 

in Eastern North America. Univ. Kansas Sci. Bull. 44:467-507. 

SterL, R. G. D., ano J. H. Torrie. 1960. Principals and procedures of statistics. McGraw-Hill, 
New York. 481 p. 

Supwortu, G. B. 1934. Poplars, principal tree willows and walnuts of the Rocky Mountain 
region. U.S. Dept. Agr. Tech. Bull. 420. 111 p. 


GeEocrAPHIC VARIATION OF Pemphigus populi-transversus pf 


Tasce |. List of the Localities Employed in This Study. 


Locality 

Code 

Number Locality Date 

1 WEStlivwlevSema SEO Walpole; Norfollei@ors Massie ee eee 12-VIII-50 
2 U.S Ihe GO; Z toms WY we Gopnen, Olenrye: (Co NANG eee ee 11-VIII-50 
3 Wass Ely lOs) 5) me Nieot Wyalusing, Bradford ‘€o-) Pan 2 ee ee 10-VI1f-50 
4 INfortin “Tomato, INitever ial (Crees INI Ae ee ee 15-VIII-50 
5 IP ene SNVaTObeleapontes Weorain’ Go: Ohio) 22 ee eee 10-VHI-50 
6 IPlesriniyeloumuen, Wworerel. (Chota G) 00 oye ae eee re PS eee 16-VIII-50 
7 Mishra kasies tag) OSE pine @orsme lin clan wet ee ene ee ee ee 16-VIL-50 
8 Wis. Hy. 50, 5 m. E of New Munster, Kenosha Co., Wis. ...........------------ 15-VII-49 
10 Ohio Hy. 364 along Lake St. Marys, 3 m. SW of St. Marys, 

JENIN Ney A Chae ©) 0 os ea ee ee 29-VIII-50 
11 WES Mulan 2eme cot MastalPeona, MazewellliGo., Ill se eeeserees eee 11-VII-49 
12 il tims Teese (CGhetivelinnvoyovavely (Chhivonweyay (Choy MENS ee eee ee ee 23-VII-50 
13 HL cern, «ID Yee ACOH eat Yah 1) Gee eee ee ea ee rere ee 24-VII-50 
14 WES Salvans Ose ems Wivotubetullyy Bike Cos) My = eee eee 25-VII-50 
15 WE SHaliyeg2 440) my on iansas City, Jackson’ (Go;, Mo. -.2. 22 eee 23-VII-50 
16 Seme SHeor tdutchinson,, RenorGo:, Kans, 2222222... Spemiaes So ie rss 2-IX-51 
17 WESeeelya lOGs lilisms sot Arkansas (City, Cowley Co; Kans. 2s oe ce es 6-1X-51 
Ram Saitye 44 205. me Not Ocala, Mariom Co., Fla. 2. ..2.ccee ee 22-VIII-49 
fey Te aptnl leat AYA / octet nt Gro ant Ce yee se ee ne ra Pr ener ee 29-VIII-49 
Bes CHES TICAG Se AGISO Mig Ove MEL ag reso 0 epee cepa an ccc m eet oe ee cae Sea suceeoessa ares = dove eee 21-VIH-49 
18d U.S. Hy. 19,1% m.WN of Albany, Dougherty Co., Ga. ......... 2 eae 30-VII-49 
SAMOS. Hye 705 VOsme SW of Brinkley, Montoe Co., Ark: _ 2222.2: sce 3-VIII-50 
19b ~~ Flood Plain of Arkansas River at Forth Smith, Sebastion Co., Ark. —........ 3-VHI-50 
IO CWESS Eve 71058 me Wi of Lonoke; Wonoke ColjvAnk: 2.22. c eee eee 3-VIII-50 
19d ~~ Mississippi River bed, 1 m. W of the river and 8 m. E of 
Wiest Mempinisa: @rittemdem™ Cory vAw Key eee rereces coe eens se 3-VIII-50 

cmUtSa ye /le 1 m> IN of) Wexarkana, Bowie @o:, Wexas ..ceee ee 3-VIII-50 
aie -Seulty: 40. 1 m: NW of Adams; Robertson ‘Co., Wenn. 2-22-22 eters. 15-VIII-49 
SiO eS srl yenca ln Guithries ModdaCoy, Nye cect pecs weet occas sare eee 15-VIII-49 
PcmUES. Fy. 605 4 mm. W of Kevil, Ballard’ Go:, Keys 222s. sccceee-cccseeeeees ernest 31-VII-49 
iGeenW:S: tdy. S155 m. Sof Covington, Tiptom Co,, Menn: 2.2. 3-VIII-50 
PU PSepilys Oil: lameesnok dinimble,. Dyer Cor, (emmys ees ener 3-VIII-50 
21 WES Salyer? = 4m SE voteRipley.) Browm (Cov, @ lion reer eee eens 9-TX-49 
22 WSs Ely. 5256 m. SE of New Trenton; Franklin ‘Co., Indi. 10-IX-49 
23 (GviyntnevalleseSirelbyye (Gorges es eee oe wean aE 10-IX-49 
24 WESRoElye 43.00) Sof Hutaw.. Greene, Cor Alla. esr sesee ee eee ener 17-VIII-49 
25 WES ilya.43.5 lim: N of Fayette, Fayette Co: Alay 222 Eat sales Paneer 16-VII-49 
26 WeSe Ely. 845914 m. E. of Waurel; Jones Coz, Miss) 2222 see Poca lte 9-VIII-49 
27 WES Eyer lee3am. NiofiGarlisle, Sullivan) (or, lmceges = eee 14-VII-49 
28 Beato nit Periyar Co ry ilVI1SSe, yee eens cee ee 19-VIII-49 
ay) U.S. Hy. 52, 4 m. NW of New Richmond, Clermont Co., Ohio __............. 9-IX-49 
30 WESaelys, 435 4)m. Ni of Rockdale; Maury@o:) Wenn) 222 22. 15-VIII-49 
31 WSs Fly. 411, 5 m. N of Evansville, Vanderburgh Co., Ind: 22. 14-VIII-49 
32 WES terial 427, ms S of Rockvalle;yPanke (Cov Inds =e eee 24-VII-49 


33 Mimi 4a Same of Pichon; klarcime (Coss: eeeeees teens ee 30-VII-49 


222 Tue University ScrENcCE BULLETIN 


Taste 1. List of the Localities Employed in This Study (Continued). 

Locality 
Code 

Number Locality Date 
34 WESe lye 0S 2h Bvot Butnamvdlleseatnamm@o salina een 24-VII-49 
35 NUL, aby WANG; Otomo IN| Ge WSs lehy 455 Wecsae Col, IM, 30-VII-49 
36 HI Elyse co) me SE sof waternlooy Montoen Cory ll] hpees ase en ee 18-VII-50 
37 Riedimont:) Waynes Cor Moist... oo BAe eet oe 1-VII-50 
38 North Provid ences Providence | Gor, hk] eeess ee eee ee 12-VIII-50 
39 WES Ely. G1Zemreenotvalkertons ieaPonte Cosy nds esse 9-VIII-50 
40 WESs Hebe (one Gr ins, SM we IDEN Soyords, Georde (Clo, Ie 19-VII-50 
41 Wiysoxss Bradford tCose bayer. moe tte coe va eee ne eee ee er 10-VIII-50 
42 West end of Jefferson Barracks Bridge between Columbia, Monroe Co., 

le ancdeeh valle taneous ors Mone es eens ene eee 18-VII-50 
43 ort, Kaskaskia sMemmeaState Park) Randolph Gos .eilllba eee 18-VII-50 
44 WES» Ely 65. 2ameW) of \Wawaka\ INoblei@orulinds 22.) =e 9-VIII-50 
45 MO me el eee kircamllcsl irre © coset VC sp ese ee 18-VII-50 
46 WES El y= eZ 0 kisperances Schoharie GoryeN pipes se eee ee 12-VIII-50 
47 Yam. N of Missouri R. on Mo. Hy. 19, Montgomery Co., Mo. ...........-.--.---- 18-VII-50 
48 5 m. E of Jefferson Barracks Bridge between Columbia, Monroe Co., 

HE andes Mehivillesss tenleouis) Cons Moses = eo 18-VII-50 
49 Mo. Hy. 30, Little Meramec R., E of St. Clair, Franklin Co., Mo. ............ 18-VII-50 
50 U.S. Hy. 6, 2 m. E of Vermillion-on-the-Lake, Lorain Co., Ohio ........... 16-VIII-50 
51 Mo. Hy. 34,6 m: Wi of Jackson, (Cape ‘Giradeau Co:, Mo, =. 1-VIII-50 
sy) Mor ElysrcilesS mes Npotikaho kas: Clarke Coss NO sees ee 19-VII-50 
53 Cape Hear R= Fayetteville Cumberland {Cons Ni Gs 1-IX-50 
54 WS: Elys 1675 Oris: S\wieom Prescott; NevadauCormAks 2a ae 3-VIHI-50 
55 Palmyra, Mariony\ Gos; Mon ssc. <2. tse ree Bee ere ae 19-VII-50 
56 WES: Hy. 35; 4m: NWeot Pranktont, Rossi Go @Ohionee ss eee 29-VIII-50 
57 U.S. Hy. 20, 2% m. E of Springfield, Hampden Co., Mass. ............-.-------- 12-VIII-50 
58 INichols:; Muscatine (Cobjeilaie. 0 Asan. << c,cae eee oe ee ee 19-VII-50 
59 WES: ly. 24455) im.) of leeroys Medinal Cons @Inioy se = 3-1X-50 
60 Us. Hy. 20; 4% m. SE of North Madison, Lake (Co:, Ohio. 10-VIII-50 
61 U:S, Hy. 65 Isms Weok Woodside; Wood" Cos Ohio = = = eee 9-VIII-50 
62 Ohio. Hy: 7, 1 im: E of Mipp City, Miamui(Co.,, Ohio ee eee 29-VIII-50 
63 WES siya 52s slronton awaences@o...@ io) eee 29-VITI-50 
64 WES Tel SIN IDSlbxmin, Ikoyeans (Clo NG, ee eee ee eeeocnce: 1-IX-50 


“uuIn]OS yey Aq pauasaidai Apnys oy) ul Ppepnypsur JOU J19M SO[QPLIVA BY} SoyRITpUT KY 


(sonteso] ,2) 


Jojoereys jo a5] UIDIIg 


610 LSI ae) AI evuuauy 
970 C02 C6L Il] euuajjuy 
S70 C02 8°62 ysus] eIqLhs 
Tr'0 6°87 any: yisueay inuay 
670 GC? ey jpn proypy 
910 Ce] C98 ysuay [Rol 
ISIIIOJ UTIIS 
£10 Sr 7S8 AI bli0suag 
£10 bbl 9°S8 II] elosuag 
cl'0 OTT 0°68 IA vuuajjuy 
680 Lb Ch Il] euuajuy 
6$°0 GLE 8°79 yisus] eqn 
£90 U0r 66S ysuay xes0y L, 
Ly 0 oz 0°89 yp prop 
6e'l Gs Sib yisu2] Surma104 
:sae[V 
Apnys Sant] eo0'T AY][RIOT & s1oqv1eyO 
quasoid WOly UI IM s[[eD 
(sotqt]ed0] 
uIgIIM _9) O} paNqiyAe SdUPLIvA 


& 

2 0 x 

& . 

: 90°0 x 

S 0 X 

S % x 

= L1'0 x 

x De 

NN 

2 0 Z£'0 

= 0 07°0 

eS 

< 0 7Z'0 

s +0°0 69'0 

QQ c0'0 8t'0 

5 1Z'0 7S'0 

Z 60'0 £70 

= FLO Alal 

= 

[4 

SS (S961) (7961) 
[SyUYy wot [B4OS wot 

g (UTI 0 (UItRIM _0 

SS +13 ,9) + + [eB 0) + 

= (santeoo] 9) (santpeso] .9) 

< . 

3 s}usu0dur09) SIUPLIBA jo soney 

(©) 


“QOUPTIE A jo sash]euy JO SYNSIY *Z XTdv 7 


224 Tue University ScrENCE BULLETIN 


Tasce 3. Correlation Coefficients. 


la cs 
: 2 a) eye = H Ee 
= op <= pe o iS co) 3 os 
ees eee ee 8 eee 
4a oF x# += &€ € 2 2 8 & 5 See 
2 z g . g g Z Z Be oo am Ey < 
St 6 6 <2 2 8 & 2B ie 2eua meee 
Wing Length DF S32 AGH Ol SORO7" i 37 AS a Cee 45 58 -71 
Head Width 51 xX 81 70 57 50 32 20 40 47 (452 
Thorax Length 70 61 x 92 76 49 60 36 48 69 26° SceSse—oe 
Tibia Length 63° 46° 56 Xo 589) 50) 725 49 77 38 34 49 —-37 
Antenna _ III DO Se. 46) 855 X 26 83 75 35. 7/5 042 46m Soee 
Antenna VI 2 840) S35e39'e 30 X 47 16 -03 -15 -15 -30 -24 60 
Sensoria_ III 125 VO SOR 05= 35 —05 X 97 16 49 50 40 S56>=% 
Sensoria IV 10> O35 05) 01 18 -06 22 X 07 6! © 22) lesa 
SM Total 
Length 14 08 O08 06 09 02 -02 02 xX 62 18 (sis 
SM Head 
Width 26 14 12 13 205 307 125 00 30 X 64 65 66 =f 
SM Femur 
Length 26 15 17 ho 20 ea SOS 02: 18 38 X 85 73 =05 
SM Tibia 
Length 20 OSE ey 14 21 (04> 12 © 08" S16; 35. 2165 xX 89 -I4 
SM Antenna 
Ill 22 O08 10 121 09 eG eee 2, aS, 50 X -30 
SM Antenna 
IV 05 10 04 (04 10 20 02 (04 6 25. 355s 60a xX 


Explanation: Coefficients below diagonal are intergall product-moment correlations, those above 
the diagonal are interlocality component correlations. All intergall product-moment coefficients 
significant at P 0.01 are in roman type (7 0.12). Coefficients below this level are in italics. 
Since significance levels of component correlations are not known, the significance level per- 
taining to product-moment coefficients of 55 degrees of freedom have been chosen arbitrarily 
and applied to the interlocality component coefficients. The conventions of roman type for sig- 
nificant (P 0.01) coefficients (7 =0.34) and of italics for not significant ones are again 
observed. Decimal points have been omitted. The stem mother character correlations are based 
only on those galls which also contained alates. SM is an abbreviation for stem mother, 


GeocraPHic VARIATION OF Pemphigus popult-transversus 25 


Tasce 4. Primary Pattern Matrices for the Intergall and Interlocality Component 
Correlations. 


Intergall Factors Interlocality Component Factors 


Factor Number 
in this Study I II Ill IV I II Ill IV 


Permanent Factor 
Number I+Il IV F-1 F-2 I II Ill F-1 


Characters 
Wing Length ae -E au tL 
Head Width +++ slochesh 
Thorax Length +++ Jb 
Tibia Length + + +- eee 
Antenna III ++ -b 
Antenna VI + ao Ee 
Sensoria III + Jha, Je. 
Sensoria IV Jee 
SM Total Length 
SM Head Width 
SM Femur Length +++ aeetoete 
SM Tibia Length Teatn ae sPSRar 
SM Antenna III ee aia. 
SM Antenna IV at JEL IL 


++ 


Explanation: Plus signs represent the following magnitudes of pattern coefficients. Intergall, 
444142 0.65, ++ = 055, + = 0.45; interlocality component, +++ = 0.85, 
++ = 0.75, + = 0.65. The account of these data in the text is based on the complete 
primary pattern matrices (not shown here, but available upon request from the authors). Thus 


not all the results are evident from this table. SM is an abbreviation for stem mother. 


226 Tue UNiversiry ScrENcE BULLETIN 


Tasie 5. Means of Alate Characters for All Localities. 


= eslhe og _- aes eS fe = as as 
ocality Os ae™) Ss oS 2 a = 
Cae Ex 2 a Ez E a 2 z g cs 2 2 
Numbers Zo (em amy ies HL BY <5 a> aS AZ 
Ol 13 1.70 363 57 62 161 180 3.46 0.23 
02 14 1.65 371 55 62 158 174 3.14 0.29 
03 15 1.63 379 55 .60 149 174 3.27 0.20 
04 13 1.67 361 a») 63 154 180 6)-5)1| 0.08 
05 15 2.00 374 59 68 189 171 4.93 0.87 
06 15 1.90 Si) .60 .64 .158 y/7 3.87 0.07 
07 14 1.75 378 56 66 A153 181 D/A 0.14 
08 5) 1.66 378 3 62 146 Ay 2.93 0.00 
10 15 He7Al 384 56 62 156 182 335 0.07 
11 15 1.61 369 50 -60 .169 .178 4.00 0.40 
2 14 1.65 367 34 63 149 lla/7/ 3.86 0.14 
13 11 1.78 375 59 .69 by.) 185 3.91 0.27 
14 14 153 352 48 ai], 133 .168 3390 0.00 
15 13 sy? 339 34 58 136 174 Bp 0.00 
16 12 1.54 368 il 54 134 163 DAS) 0.00 
17 15 153 359 Dl 56 134 170 3.20 0.00 
18a 2 1.90 361 56 .64 lls 160 3.00 0.00 
18b ep, 2.08 356 il .64 168 156 2.00 0.00 
18c 2 1.92 361 56 .64 150 162 3.50 0.00 
18d > 2.06 388 60 66 .167 .166 2.60 0.00 
19a 3 E43 336 44 Dl 15 sII5)3} 235 0.00 
19b 6 1.95 376 54 .64 162 168 3.33 0.00 
19c 2 1.40 328 48 53 AVA 164 p,5)1| 0.00 
19d 3 1.76 358 48 Di 140 158 1.33 0.00 
20a 4 2.09 384 62 68 179 ai, 3.76 0.00 
20b 5) 1.94 381 .62 66 ally 180 3.80 0.20 
20c 3 Ie 345 48 59 2158 158 3:33 0.00 
20d 2 1.92 365 5) 62 156 148 3350) 0.00 
20¢e 2 1.82 37> 54 162 150 168 3.00 0.00 
21 25 2.04 383 58 69 a5 .174 3.52 0.20 
22 15 Pail 397 165 fl LYS) 174 3.87 0.13 
23 1] 1.95 S18 61 .67 156 174 3.46 0.09 
27 3 1.76 364 54 58 146 55) 3.00 0.00 
28 2 1.74 356 ae} 63 52 .162 3.00 0.00 
29 8 Zal> 386 63 aif 191 BD 3.76 0.38 
32 3 eS y/ 342 Sy? 59 157 168 3.67 0.00 
34 5 1.81 344 Dl 62 a155 159 3.40 0.20 
Sf 2 1.78 347 Dil 63 168 .164 5.00 0.00 
38 7, 1.69 soi} 59 63 160 li72 37 0.14 
39 3 1.67 385 ul 66 158 186 4.67 0.67 
40 13 oy? 347 50 Sy 135 slyi72 Bell5} 0.00 
43 11 1.49 535i 49 Sy) 143 Stil 3.00 0.09 
44 7 1.78 363 DY -64 161 al Al 3.86 0.43 
50 5) 1.83 350 54 .62 165 168 3.93 0.60 
51 8 1.73 345 50 56 146 157 Ball 0.12 


GeocraPHic VARIATION oF Pemphigus populi-transversus 27. 


Taste 5. Means of Alate Characters for All Localities (Continued ). 


: Pa gi x Ee} a g = 3 
Code 3% ES ee 2s 2S =e ain, c _ Sc 
Numbers Ao foe] Sole a bo <a i> w= ae 
2 12 1255 306 49 29 138 164 3.08 0.08 
53 7 1.98 365 59 63 163 167 3.86 0.00 
55 5 1.60 346 49 58 146 160 3.80 0.40 
56 8 2.03 37> 59 .68 176 178 4.25 0.12 
Dif 6 1.91 381 58 65 174 169 3.83 0.50 
58 6 1.61 367 23 .60 147 78 3.50 0.17 
D9 13 1.95 374 Sy) 67 186 174 4.46 0.31 
60 15 1.83 359 55 66 ally sil 7al 4.53 0.67 
61 15 1.69 364 56 63 146 179 333 0.07 
62 10 1.69 347 4 .63 lSy/ “73 3.90 0.10 
63 8 2.11 369 Sih 68 184 178 4.00 0.00 
64 8 1.50 304 45 56 142 72 3.00 0.00 
Standard Deviation aly 024 052 .053 0163 .0138 1.065 400 


Explanation: The number of galls per locality is generally less than those in Table 6, be- 
cause some galls had stem mothers but not alates. Furthermore, there are only 57 localities listed 
in the table, since 17 localities had stem mothers but no alates. All continuous variables are given 
in millimeters; all meristic ones as counts. The standard deviation at the foot of the table 1s 
the average standard deviation within localities. It is mot the standard error of the means. In 
order to obtain an average standard error for the study, one might multiply it by 1/2.9965 which 
is the reciprocal of the square root of the average number of galls with alates per locality, cal- 


a 


E Vn 2 
l a an; Ms - ne : 
culated as nj = ——— Sieg where a is the number of localities and n, is the 
a- cael a 
=n 


number of galls in the zth locality. 


228 Tue University SciENcE BULLETIN 
Tasce 6. Means of Stem Mother Characters for All Localities. 
2 s cm 

se Sp = e % = > 

ss E = = E S g 

Locality S = ® 5 z: S| S 

Code E E g 3 E 6 g £ 

Numbers Zn B 0, am B < < 
Ol 13 22 428 Spy? Bly/ .137 116 
02 20 2.40 432 365 S21 .142 US 
03 20 2.30 435 “apy soils} 138 oll Il 1 
04 20 2-32 424 365 307 .149 114 
05 20 2.54 468 390 344 .165 .109 
06 20 2.58 454 404 328 1538 110 
07 18 2.93 449 366 327 154 116 
08 19 2.64 448 374 Sly 149 119 
10 20 237 429 381 B23} 147 sil 
11 7 Pehle} 409 384 323 152 .114 
12 20 2.49 424 399 337 156 119 
13 20 2.68 420 387 S27) aloyll 120 
14 20 237 410 348 305 143 108 
15 18 2.56 405 354 319 5115x0) .109 
16 20 2259 384 349 305 140 AS 
17 18 Wa\\5) 399 364 S15 139 SITS} 
18a 6 3.00 426 384 319 .148 101 
18b 17 2256 460 356 334 BlSy/ sls 
18c 12 2:55 446 385 318 165 .107 
18d 10 2.48 446 399 322 156 .104 
19a 17 2.58 445 385 325 156 101 
19b 20 2.80 443 .382 333 156 .107 
19¢ 1, 3.06 455 -410 350 .162 .109 
19d 13 2.54 438 385 Soil .160 108 
19e 7 22 422 386 yy? 149 103 
20a 17 3.14 477 384 328 .167 .109 
20b 12 2.90 425 ES) 307 .160 105 
20c 12 2.62 .439 379 5325 BISy/ 102 
20d 20 2.96 464 413 Sy 164 110 
20e 20 2.24 412 384 £335 150 107 
P| V5) 2.23 455 411 338 lS .109 
22 17 2.66 460 400 332 .165 .106 
23 20 3.03 435 360 304 140 110 
24 20 2.74 447 395 351 156 .108 
25 16 Slly/ 469 384 337 .164 .108 
26 14 2.66 460 397 325 161 105 
Qi 9 3.09 437 340 BY) 154 105 
28 2 E22 462 412 370 WZ .100 
29 8 ay | 454 353 297 138 101 
30 14 3.38 446 398 338 161 .102 
31 5 2.58 .409 379 326 148 .099 
32 8 2.62 413 392 336 146 .109 
35 12 2.92 457 412 356 .168 .106 


GeocraPpHic VARIATION oF Pemphigus populi-transversus 229 


Taste 6. Means of Stem Mother Characters for All Localities (Continued ). 


u c = | 

& Y Y so} s 

Locality 2 é i ie 5 es e a 

Code ets g S E 5 2 2 

Numbers Zn B a0) am < < 
34 17 2S 453 415 352 B57 109 
35 20 2.38 424 396 341 156 108 
36 20 273 421 358 308 145 102 
37) 7 2.87 443 391 330 153 106 
38 7 2.04 429 395 soil 7/ 140 a2 
39 6 2n33, 450 382 308 142 105 
40 15 2.45 409 389 319 144 7 
41 5 2.74 442 411 353 161 114 
42 18 2.90 466 412 341 161 107 
43 20 255 442 392 05) 147 109 
44 9 Defpd 462 411 348 al53 110 
45 4 2.38 460 386 328 149 104 
46 4 233 430 342 278 116 aS 
47 20 27 463 402 pil 165 106 
48 20 2.63 455 398 344 163 110 
49 20 2D 446 389 328 154 106 
50 20 2.28 436 374 334 146 105 
51 20 2.64 443 415 BDz 158 108 
52 16 2.01 421 31) 311 143 allalil 
53 20 2.76 458 418 3D 161 108 
54 15 2.79 458 397 349 166 107 
DS) 19 2.45 442 386 328 150 105 
56 10 2.62 435 412 a6)! 159 .107 
By) 12 2.50 432 370 310 141 103 
58 7 2.49 420 410 338 “153 121 
59 20 2.98 475 421 346 .170 110 
60 20 2.84 473 405 357 mlW72 15 
61 20 2.67 440 382 Sey 152 118 
62 10 1.87 430 387 Sy 143 1e 
63 8 2.29 429 397 342 158 110 
64 8 23 406 371 326 148 116 

Standard Deviations 379 0341 0338 0283 0165 0098 


Explanation: These means are based on all available stem mothers regardless of whether the 
stem mother had alates associated with her or not. All continuous variables are given in milli- 
meters. The standard deviation at the foot of the table is the average standard deviation within 
localities. It is not the standard error of the means. In order to obtain an average standard error 
for the study, one might multiply it by 1/3.8693 which is the reciprocal of the square root of the 


a 


: | =n,” 
average number of galls per locality, calculated as n,—= —— oo 
a-l a a 
=n; 


where a is the number of localities and n, is the number of galls in the th locality. 


Se *S1UNOD SE souO SStIOUT [[E fssaOUUT][TU UT UdATS av so|qeIueEA snonuNUCD llv 


230 


991" vel 820" ZS0" 190" ZZ = 9200" €L100" 99100" 9€00° ZZ00' ZZI0’ 600° OFTO’ 8zI0 IS00° FE00"° 6¢0' suvoul 
Aqtpeo0] JO S1O119 
piepuris odeiay 
LE SiCu LORC £0'T Stal £0° fre Als 9S60° 0990° 9ST" cr” 799° LLY IT9” 109° O8e 807° 06'T €~ 
EG leer I £O'T LOM Or Ly CLs £660" T0Z0° OLT- Ist £0L cls L6S° €€9° C6e O17 OVC CC 
€8°C ANG (OI OO'T Or €8°¢ G/A\ STO.’ bIL0 OL Srl WZ9; Lr £69 b9S° SLE COC 86'l IZ 
Lye GG (00 £0'T £0° fre 991° 0660° 0290 v9OT OST 6c9° 19p" OS L9¢° 89° COC 06'T OZ 
LOC Cac. 00M OT 00° BSG Ee 1760" Ce90° Ip 8el° C8¢° [cr 61¢ bOS 8Se° 86. (6 I 61 
< OTC £0°C £0'T £0'T 00° 00°E b9OT £00 00L0° 691 Lvl 699° C6r €8¢° 6LS° cle COC 00°C 8I 
5 YEE HE £O'T O0'T 00° ete: OVATE [760° 86S0° iS) Of 9S¢ €Ip 91I¢ bOS COS" 807 cS I LI 
5 OTE eG £0'T £0'T 00° LOG col" €Z80° CSCO" Daly Lol I¢¢ Icy £p¢ ES: LOS Iie: cc'l 9 
jaa) O0'€ See. e OOM £0'T 00° Lye CLs 1S60° OZS0° oeT pel bLS bOr COV" Cry 9E¢° 16 6ST cI 
8 CCieee One OORT £O'T L0° JAS te 89T 0680° 6SS()" Ort bel 98¢° ely 609° 88h Loe 661 bo T bl 
a SS €SC O0T tet |I fet f6'¢ bs O0£60° 8£90° bol vale £S9° OL 66S 8oS° ele LOC GE el 
3) ANAS EG £0'T OMT Or £9'¢€ 621 8br60° L190 6b br oc" Grr BES IES OLE lie b9'T cl 
se esiGe L6G, 00 LOT Ly 06'€ Ls OL1OT 8120" 99K cel L6S° Orb 6rS ITS o9e" 607 09°T IT 
Et Og € LEC £0'T Seal £0 EGE 6L1 Cr60" 9r90° €sl" Ort 979 Cor l6¢ £9¢ p8e LMG IZ 1 Or 
es LG {LEG £0'T £0'T 0g £0 6ZL1° 6r01 09ZL0° pst 6ST CUES EUS: [c9° LC9° 88o° COC 60°T 6 
z £0 L0G 00k1 00'T 00° LOE SLT £680" C8S0" Sa EE S19 6br COS ces CLES 607 COT 8 
5 ei (ASG (0 00'T el €8c = O8T Cf60° L£90° est 9b coo" C8h b8S° CLS ble VIC CLT i 
- OG “Loe  (O0rtt LOT alee (oka SA 6960" O890° O9T Srl €S9° LLY [19° 009° OLE SIG co | 9 
a Gre OAs = (El ele £8" €8'°p OL CS60° €S£0° C8T cS" ¢L9° O8bh £6¢° 6S" ELE bOC 66 1 ¢ 
i LNGS LG S001 LOT ale Lae. OL ly £760" OF90 CSL IbT c£9° 09b* £9¢° 9S C9e O17 99°T b 
OCes Z6nT £0'T eI LG Og'€ DAL 6680 6090" (ASI Ie Orr’ clo" Loy OSS OPS v8e ENG o9'T € 
Lie a LOKCe Lal (36) 0g Lye 8ZT c£60° 8S9()" 8ST 6£T bho" LOv €9¢° COS OLS; 6070 OL T G 
€6°C ELT £O'T “vel Or LCE LEY Ch60" be90" 6ST Ort 1€9° 9Sb CSS b9¢ 89¢° 907 89'T I 
ee ee ee = eS ee 
oe os 8 3 3 6 g g S Ce ee soe fe es Se Se ee oe 
a a atc i. me Ee 5 5 5 5 ere = mew “ x = 7S. 
< < [<) > =) io) rt} rT) p> tod ie 
= S J Zs 4 < 2 = 
*SODTTBIO'T ITV Jo STOJIEIVYT) ITV jo suvay— (¢96] ) TAANIYY aNv TVMOS AO G ATav Ga LOAYAOT) 


Se 
N 


GEOGRAPHIC VARIATION OF Pemphigus populi-transversus 
g 


*padso[duua IIIM 9194) WIO1T sajduures Ou 9ouls poeyeotput jou IIe pieog 
-bas INUPNY ayy Ieou uonnqliystp pooMuoqo5 jo Sotre punoqg JOEXY “PpeYlioa u90q sey bore ueryoereddy oy wot SpOOMUO0}05 
jo aoUasqe 9UL ‘a]qeuonsenb ST dew sty} UO UMOYS se PMO] pue TINOsstyy ul seore uleyIo5 Wolf SpOOMUO0}O9 jo dUISqe eu L 
*paydaaae usoq sey i dew astdoid o1IOU PB jo Not] ut Inq 97yeINdDDe Ajaanua jou SI dew sty} JOYyINe IJOTUIS oui Aq Sp4IOSo1 uOT)d3[[0) 
woly “(LC61) ywompng Jo dew ey O7 Sutps10j9e “ried SIP Of]9p snjndog ‘pooMuojjoo ey) jo uOnNQLYSTp 941 SaqtOsuIMoITS 
ouly poysep o"uLL “Il AGP L ul UudAIS ST SONT[BIOT ey) jo Sunst] po teJ9p V *sToquinu our BUIMOT] [OJ SIDI] ISPD IMO] Aq peynuspt 
pue payeredas e194 oIe ‘sojdures AVRO] [[euus uO poseqg SONTPRIOT aqutsod uo sunuasoidos “sToyyne 9soy jo so[dures vole, IY LL 


“(CQ6T) JPYUrY pue [eyos Aq parpmys asoy o7e ¢7Z 01 [ SJaqUINU saNtTEIOT ‘sIoquiNnu apod AW][eIO] Moy) are Wey) 0} IX9dU SJOquINU 
fsayIs AWRIO] OY Wasasdar sa]dII9 MOT[OF{ *pe49]U2 sJaquINU apoO3 May. pUe soNTeso] YAM patpnys vse sy) Jo suTjINO *[ “Of 


CIENCE BULLETIN 


‘ 


THe UNIvErsITY 


N 
N 


“AIYIO YICI WIOAF JUIIIPIpP Apuroytusis oie Asya a0u94 
faqqei ayi Jo out, Aue Ur papnypout yIOq JOU oIe gC puke ZS “IOAIMOFY “G9-ZLG UT] 942 UT papnputT oie Ady) SyusIayIp APueoyrusis 
jou 21e gg puke [g snyy, ‘asuer sy uIyIM Woy) Jo YO sapnypsut s[qui YNS Py} Ul auT] OU Ft (TQS d Ww) Iwerayip Apues 
-YlusIs ae suRdUT OMY, “OINSY sy} Jo ulsieW pueY YS ay Ie UPAID are syso] YN ey} Jo syNsor sy], “] sans ur poure[dxs 
u22q aARY souT[INO sy, *safIID yeIq se ¢g<X pur saps9 Avs yep se ¢g OF [¢ UVa jaq ‘sap AvIS IYSI] se (Vg 01 LE UIIMY 
-9q SUvdLE ‘Sa[DIID AYA se UMOYS de LE >suRIY ‘9109s go1d sYyi JO spnyUsei ayy Jo kapt [esauad ev saatd Sutpeys ssoym 
sapomo Aq pajuasoidar are sonies07T ‘uostreduiod jo asva J0f Q] Aq patjduyntu pue sigoid oUt pouojsuen Us02q 9AvYy suUPOU 
ay], ‘sae Surureqos saniyeso] /¢ ey JOJ ,.ySua, Surmoasoj,, Ja}IvIeYD 9y2 Jo suvaut jo dew uonnqmystp sydess0ay *7 “OI 


695525 

89—ES 

DE SAE, 

ce Ie . 
¢9— OS é bis 
VI 8v 
GEES hy 
laa v 
Ome, 
8S— cv 
Ge ay, 
pG— Le 
7G— 9 
8p— ce 


_— fo 


=~ 7 / 


GrocrapHic Variation oF Pemphigus populi-transversus 


*saye[e SUTUTeJUOD saNt[eI0] /G¢ IY} JO 
gu ESO ee Seo eres 


oe 


‘7 dINSIy Ul se uONeURL|dxy 


(PPM pry, 


N os 


Jaqoe1eyd 94} Jo sueoul fo deur UOTINLIYsIp ITYC 


ff J i Wor 4 nw 7S 


| 


PIS0IH 


Xs ‘Ol 


Tue University SciENCE BULLETIN 


"N 


°Z ainsi ul se uoneurldxy 


*SoTel[E Sutureyuos SONTPeIOT £¢ IY IOF Ui sus] xePIOYI,, IoJIeIeYS 941 Jo sueout Jo dew uonnqgiystp o1ydeis0ay “p “SIA 


¢2— OG 
OL Vv 
Loma OY 
CSmes CV 
Ser = Oe, 
came teh 
OSSe9e 
GG— ce 
€S—.0¢ 
SIG 


dS cid 


® \ SG ; 3 ~ 


LN A * BR os ony 
v eee ee foe A: Si mK Ze if 
cus at 
oe 
(ay 


GeocraPHic VARIATION oF Pemphigus populi-transversus 


*T JINSILY ul se uoNeurl|dxy 
‘saqe|e BUIUILIUOD saneso] ZC PY} JO} ,.YIua, viqn,, JayeIeYyD 3y) jo suvour jo dew uonnqiuisip s1ydeis0ayn *¢ ‘org 


Tue University SciENCE BULLETIN 


‘7 dINSLYy ul se uoNvueldxy 
*saqe[e SULUTeJUOD satPEIo] /G yi OJ ,.YISUI| [J] Wourdos [euusue,, JayeIeYyd ayy Jo suvsut Jo dew uoNNqistp sYydesB0IH “g “Oly 


ClammeS 
O25 lis 
Soi aiy: 
29S LV 
Vom eV 
Coe Iv 
issn Oitz 
09> Se 
SSe— 9 
IVa Ba 
m= iiee 


dS Ble 
ANS 


GrocRAPHIC VARIATION OF Pemphigus popult-transversus 
fod 


*sa]e]P SUTUTE]UOD Sotqt[ed0] 


*Z nS ul se uoneurldxy 
jeuuaque,, 


LG 24) JOF ,.YaBu2] [A WIWsIs 


JOJIPIBYS 94} jo SUP. 


e) SSG 
Sean 7 —_—_~ 


JUL JO dew uOTNqLystp aydeis0a5 ‘f ‘OI 


yy Fe 
Lan 
oak ut / 


Tue University ScreNcE BULLETIN 


*Z IANS UI sv uONvUL[dxyY ‘sajeye Surureyu0s 
SANRIO, £¢ ay) IOJ , [J] JUIWISIs [euUDJUL UO LITOSUSS JO JoquINU,, Jo]vIvYD ay} JO suvow Jo deur uoNNqmstp s1ydes1s0aH °g ‘o1y 


R) a & 


. 


We 
Cir 


GeocrapHic Variation oF Pemphigus popult-transversus 


‘7 ons ul se uoNneur|dxy ‘soqeye SuruTeyUOD 
Sanedo[ £¢ 942 JOF ,. AT WowsIs [eUUDIUR UO eLIOSUas JO JoquNu,, Jaqoerey> ay) Jo suvour jo dew uonnqmastp s1ydes 


rei 


JIL) 


6 


‘OL 


Tue University ScrENCE BULLETIN 


240 


*Z oINSI UT sv uoneur[dxg ‘s19yjOul 
UW19}s SUTUILIUOD satyt]edo] $f IY OF .Qasuay [e107 JayOWT WId}s,, JoeIeYD 94} Jo suLsUT FO dew uoynnqimsip s1ydeiso0ay “(Q] “914 


—— 


Liam 6S 
emits 
OL 67 
895-27 
29m SV 
12S As 
c9— OV 
lOmenioe 
6G— LE 
Loa Ge 
GS— ve 
CS NS 


241 


GEOGRAPHIC VARIATION OF Pemphigus popult-transversus 
g 


‘7 AINSI UI se UONeURIdxY ‘s19yIOUT 
W93s SUTUTEIUOD Satyt[eIO] $/ IY} JOF ..YIpPIM pray JoyIoU Uajs,, JayeIeYD 9Y2 Jo suvaur Jo deur uoNNqistp sYydeIB0IH *[] ‘og 


Tue University SciENCE BULLETIN 


‘7 oINSIL UI se uoNvURldxY ‘sroyOU 
WIdIs BuTUTEIUOD saTy[RIO] Lf IYI JOF ,.YuI| IMWayz JYyIOW W}s,, I9}eIeYD sy) Jo suvou Jo deur uONNIsIp YydeiB0IyH *Z] “914 


(a= 


Cg 
ays 


se) 


GEOGRAPHIC VARIATION OF Pemphigus populi-transversus 


‘7 aunsty Ul se uoneuR[dxy ‘srayiow 
Jo}IeIeYS 9} jo sueout Jo deul uolnqdrstp a1ydeis0ay 


UWo4s SuTUIeyUOD SITTBIO] bl 24} I0j 4isu2| BIG JOYAJOU Us 


” 


SS 


re 


Let 


Tue Universiry ScrENCE BULLETIN 


244 


‘7 sins ul se uoNeur|dxg ‘ssoyour Ways SuruteqUos 
SALI] $Z YI JOJ ,.YISua] J] Wousas [euUauP JoyIW WoIs,, JaIvIPYD 2 JO suvSL Jo dew! uoynquastp s1ydeis0ay “p] “Oly 


LN 


5 


24 


GeocraPHic VARIATION OF Pemphigus popult-transversus 


aC ainsi ul se uoneur[dxy *sTOYIOU Ways sururejuos 
SATLIOT $/ IYI JOJ ,.YISuI] AT Wowsas [euusjuR JoyoW Wels, 


Jaqoeieys ey jo suevou Jo dru uonnqdiysip aydeis0a5 , 


246 Tue University ScIENCE BULLETIN 


Fr. Sede tetelelele te] 


Rate te Rete la ie ia! 

QM HH ee 
\Leletelteteletetatelalalal 
\cieleteleialelslelelelalalal. 
GS hE Ok 
Af teel alate ele lala late lial sis! 


Lt oteltetetetetet altel te telelelals! 

tote ele tee et ele late le lela lel] 

VW teletetsteleleleletsletelalalalalials| 
retell et ete tele eltelelelelelaltelts lis! 

Velo eteteleteteleleletetelelelelelallals| 
Aleit ettettetetate tal ala ia ital oil tal I. 
WNeteteteleteleletelelte late tetelel-alia belt all a! 
Vol elteletetetelelieteletalelalelalalalialieals! 


 Colletetelteteteteteteteleletet alee tele lalslalia| 
Vr ieleleleleleletsteleletetatstetsistetetatetatata! 


(QPS PS OS OS 2S OS 26 O8 0d 96 BS 426 DS OS OS 04 OS 94 28d aid PS IPS 

Vl oleteletetetetelelelelelelteleletelelalelteteleltatela! 

(3 SSO RH KK KO 

1 oleleleteleletelelelelteleleleieteleleleteteteltetetetalals, 

ol aleteteielelelelelelelelelelstetetelsisteietetstetetetals 
totoheteleieteleteeleletelstslelstetetalestsietetetelatsial 

(tot aiaieletetetelalelielelelelelstetetaleteietstelslslstalsiell 
etetaieletelnetelieleleieteleleielateietelstsletetel sls tets tale (al 

i tetetteteteteletelieleteteletetelelelelteltetetaltel elt ete telat iteltalta! ia) 

\ Solel ateletelelteteleleleleleleletetelelstetslieiatsistetetaiatotslal:| i) 
\tolotetetetateeteteletelelelelteleleletsleltelteleleletelsieletelala) 
QPS Ds Og Og OE Og Dg Og OS PS OE OE OE OS Od Od DS Od OE DE DE DE PE Og Od OE De OE Dg OK DE PE OK OK 
Gp Pe 26 dd Pg DE 4 Pg Od Dg DS PE OK Dd PE DE PE Dd Dd D4 Dd Dd Dg bg Dd Dd DE Dd Dd Dd Pd DG OE Dg Pe OX Od Pe ~. 
os HSK SK OOK OS OS SOS OS OS DK SO OS OS OS OS OS OO OO 


=F 
QPS HE Oe OS OK OE Og Od Og og OE OS PE Pg Od 94 OK Od DE DE Dg Od D4 DE Od PE DE Dd DK Od Dd DK Od Bg Dd Od Od 
QD PS Pd D6 4 Og D4 Od Og DE Og DE Od OS OM Ph Og OS Og DE OG OE OG DK Pd DS PE Og D4 Dd D4 DE D4 DE OE Dd Od D4 DE PIPE OE 
QD 2s Od Oe Pe Og Od De Oe Pe od Dd Dg DE Od OE bd Od Od OE OE Pd Dd DS D4 Od DE Dg Dd DM DE DE DE OE DE DE Od Od OE OS 


To teleleleleleleleletelsleleleietseteletetelelelelelsleltalsioialslietealslalatetatal aio! 


QD 2s OS De OE Od og OK Og DE DE Og Dd DE Og DE OE DE OE OE OE DE OE DE OE D4 D4 Dd D4 DG DE DE D4 DE D4 D4 Og Dd Dd Od Pd OE 
Moleleleleislelelsleielstalslslelelsiaisleleleletslslelsielelslisietstaistslelstalatala, 
iNotateleteteletetetetetatatatateltetelelssieleltetelelteleletetetelsiatetetsistotatatats! ta] 
MQ PS 2S Pe os oe Oe Oe og Oe Og oe me Pe Dg OK Oe og Od Og OE Dd Dd OS OE PE Od Dd OM Dd DE DE OE OE Od DE OE DK DE HE Dd Oe EO 
Tol et ela eteteleleletelateteltetetetetateleleleletele: 
+ Teleleleteleteteteleteltetstetateltetststetelelelstetalel 
WX XK MM KH KKK KKH KKK KM 


7 


MY 26 OS oS od 4 0 og og od od oe O¥ og de Oe Oe DE De dE DE De DE OK PE Od Og PE Od OK 


Sf) PS DS PS Dg DS Pg 2d OK Dg od Od Od Oe DE Od De DE OE DE DE Dd DE PE OE DE DM PE OE OE OE OE OE Od Od OE Od OE DE OS OE OS EOE 


tote leleleleteteteleteletetelelteltetetetetstetelelelisieisielateteleleltelelslsielelslielalitallia! 


Eel eleleleteteleletetelelteltetateteteltstatsietalelelteleltslelelelelelslelslslslslietsletelatolaiota alstsialaiola, 


i teleleleleleleleleteletetelelatetetatala, 


ito leteleleleleleteteleletsteteleletaiatata, 


MMMM MMM HMM KKH KO) 
Do DS OS DS OS OS KOSS OE OS OS DS OS OS OS OS EO OS 


teteteteleteleleleleltelelelsteleltetelsiatateletstsletslelsletsisteiesielelelslelslslieielolaialoioia alsleatslalalialo: 


0G 2S. 2S OF POMS 29 25 89 OF 92 98 29 888 


i Eeleleleleleleleleteleleleietslsteieleltetatststatstaletsteteleletelelelslataleleltsleltslelstieloleiota alslelalelaiala 


telelelelelelelelslelelstelealeleteleieletsieleielatelelstsleielalietalielaieliolelelelielaleialiiotal >| x x 
<tetoteteltteleteteltotete tate eltetet ete teteltetetetete tates teeta tetetele tele le tele l ele lelel al cial | = 
it eleleteleleleleleleleleltsielelsteletsletaliatetetstatataletetslalelelslatelelelolslelelaleletelelelaia alalaletalate; Et 

delet etelelteletotetelsleletetateletelelteleletelteteltetetetelteleteleteletetetelelelelalele tele l alo) Polted mS cal 

Nel etetelelelaleleteltstelelstelatststetatatetatett tela tatetalatetatelelelisielelelialslelaislelalelelaiela alelelelaiotal x 

Qos 26 DS OS 96 2S OK D6 96 Pd Od Od D4 OS OE DE OE DE OE OK OK DE OE OE DE DE OE OE DE OG OE DE OK OE DS OS ME DEE OR PEE DEE OE > OE EEO x x 

OD D4 24 24 D4 DE DS PS DE D4 D4 OS Pd OG DE DE OS OE PE PE DE Pd DE DE DK OG OE BS OE OE Dd DS Od Dd OS PE OS OS PR DROS Od OE PS OE Dd OO OS ~ x x 

OD De 26 Od DE OE OE Oe Dg OE OE OE DE OE DE D4 PE DE OE PE DE > OE OK OE OE DE DS DE DK DE DE OK DE OE PE DK OE OK DE DEE OK DEE DHE OE OPE > x > ~ 

(iOS Pe DS Pe OS De DS Od OS Dd DS Od Od Od OK PE Dd PE Od Bd Od OK De Od OE Bd DE Od OK DE OE Od OK OE OG OG Bg PE SO DE OS PS OS OS Oe Od OS OS x x ta 

Dy SDS 9K Og OE 6 Og Og OS OS Og Og OK OG OE PE DE OE OE OE OE DE OE DE OE OE DE DE DE OE DE OE DE DK DK DK PE DE OE DE DE DEDEDE OE OE OE OE OE DE DE OE Dee DE OE DE DS PE DS OOS DS ~ 

MQ PS PS > DE Og DE DE OS De Og DE De De Og Od Pd OK DE OK PS OE OS OG OE OG KO OO OE OE OO EO ES xz ~ x 
tolelelelelslelielaleialeleleleleielelelelelelelealsleisisisietelsieleleteltetstetslaielaiaia ~ bad cat 
TAK PS Pe Od De 2d 0g Dd Od DS Od Og Og Dt oe OS Od Od Bd OK OK De dE OK OS OE PP OO PO OO OO ~ x — O x 
DH MMMM MM MK MM MMM MMM MMMM MR OM OR x x * 
ielelelteteteletetelstelelteteeleteleletele alaleleleiictaiatita San nee > Ep = 
BRARAAARRAASS SIFT IIIDHIMOY GS SSSSPP SSIS AAAAAARAARARA AARABARASSSSIIOS ISHS RAE 


247 


Geocrapuic Variation oF Pemphigus populi-transversus 


‘1X91 99S UOISsNIsIp JOYUN§J JOY ‘poyjour ase 
-Taae oy YUM dn Moys jou pynom YITYM suvaut UaaMj0q s9dUdIa Ip 
JURIYIUSIS sie ATepuNog dyi UTM seame yuR[G Sew aBeIIAK Suisn 
U2IYM IURIYIUSIs pasapisuod aq pynom yYoIYM saduarayIp JuUeIYTUSIS 
jou oie Arepunog sty} aptsino s.x [py ‘uw JO} °u aziIs ajdures asei9ae 
Sunmmsqns pur sozis ajdures jenba Jo poyjou 24} Suisn jsoi YNS ue 
UO paseq sias JURIYIUSIs Jou Jo Arepunog ay} syussaidar xuMeUT ay) JO 
Wo}og 242 0} UOTYsey Sez-S1z ul ulsiew pury Ya] ey) wo Suruuns 
oul, OY, “8p URI WoIF JUsIDYIP ApuPoytusis st yng ‘7p ura WOT] 
usIIHIp Apuesyiusis jou st Q¢ uvsur s[durexs Joy snyy, “yuorIyIp 
Afuvoylusis ore sursur ay} KX Ue syDRT UIT JOYSIY ay SuNudsasdas 
pue ot) J] “oersyIp Apuroyiusis jou oie suvsur aya KX Aq paxyieUT 


are spua Yioq FJ ‘ozs ajdures ayetidosdde ayy ym ura ayy Aojduo 
atods Jenba jo suvatu yqord [esaaas Jo ased uy ‘ulsieur puey yoy 
ay? Suoje uveur JoYstIYy syi Jo [aaar_ ay) OY UNOS ay1 uMOp UNI pur 
Jeuoseip jediourd ayy Suoye uvsur samo aya dn yooy JuatoyIp Ayues 
-Yylusis ore suvaut om} Aue JOYIOYM 4s} 0} JOpIO UT *jeuoseip [eds 
-ulid 9y3 Suoje uvaw Sutpuodsais0d ayi Jo azis ajduues ay) a1edtput 
xtyeur F[eY sy} Jo Wonog I9y) Suoje siaquinu sy, “(QT ent osed 
“WOD) Ja}IvIvYD IYI OJ sasoos yqord uvsuT ay are xLeUT J]eY ay Jo 
Jeuoserp [edioursd ay) Suoje pur urssewr pury aay aya av ssoquinu 9y_ 
‘azis ajdures asevioae ue Sutdojdwa usyM synsazr ayi YM pareduios 
sazis ajduies jenbaun uo paseq yisuea] [BIO JayIOUT WAS TOF stsAyeuL 
MINS UP jo sijnsor 9y1 Jo uonviUdsaIdd1 ONeUIWIRISeICG] “QO, ‘OL 


Tue University SCIENCE BULLETIN 


248 


‘7 DUNS Ul se uoNeUR|dxy “] 10,9ey AW[eoo]JoUT Jo UONELA DYdeIBoIH */] “Oly 


GerocrapHic Variation oF Pemphigus populi-transversus 249 


18. Geographic variation of interlocality factor II. Explanation as in Figure 2. 


Fic. 


Tue University Science BuLetin 


250 


‘Z ANSI ur se uoneurydxy [py toyz AY[eoo]193Ut JO UONKIIEA MYydeIS0ag "61 “Old 


GrocraPuic VARIATION oF Pemphigus populi-transversus 


c 


ainsi ut 


se uoneuer[dxy *‘[-q foes AYpedO[ 191 UT jo UOTIPTIBA o1ydeis095) ‘O07? “OI 


Tue University ScreENcE BULLETIN 


*10IIEF IW9YI IOF sours.nxa a31soddo am wos 1U912,4Ip Apuvoyrusts are SMO] JO systy suiaNnxo Suyuasoider suvour [je avya Avs 
0 ajes st yt ‘sdnois 19yI0 ur ssoyi Wor; JUsIZYIP ApURoyTUsIs jou Ay[e1oUDs sie sdnois asoy] UL SULIT JTIY AK “S10IDeF Isayi 
JO Saas] YstY oy2 yuasordor PI] “TT ‘T A[FYA s10eF Jd1Yy1 YI Jo sjaAa_ MOT ay} Suneorpur ur pur ‘W ‘t ‘saroos 10}Ie} JO apnyiusew 
ayeoipur suoneindod symadsvazur ayy Surdynuapt sjoquidg ‘deur sy owwo QZ ammB1q wosy (so]o10 auYM pue Yyoe]q) suraur swat) 
-X9 942 potdod savy 2M “sOJLF SIYI JO UONRLILA dy} Jo vapt aWOS aATS OF, ‘suvaUT MO] pur ysty Ayed sey yorym ‘p-y ro wey 
OF astdap 0} a]qQIssoduit 919M satepuNog YONG *AJaaNdadso1 ‘souT] pros puke paysep-rop ‘panop Aq paztjoquics aie soliepunoq 
PSU “TIT Pub I] “[ Stoiv} JoF svore Moy pue ysty Sunesedas sotepunog ayeuntxoidde Aq paprarp ase suonejndod sytsedsvazut 


PUL “OZ-LT SINS UT UMOYsS sIO}IeF INOJ ayy JO siseq ay2 UO suONE[Ndod syIDadsesyUT IOUNSIP dzII0S91~9 0) JduDNY “[Z ‘oO 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


NEW ASIATIC AND AFRICAN CAECILIANS 
WITH REDESCRIPTIONS OF CERTAIN 
OTHER SPECIES 


By 
Edward H. ‘Taylor 


LAr ny 
Wi UI HO 


~oeomey 
i A “ 
ali GO 1ORE | 
, e & | Pa) OD : 
fNnns if Ane : 
Alf } OF ENE wih Te | 
Po rerserrrrs tigate Se SEE 
; . 


VoL. XLVI Paces 253-302 Ocroser 28, 1965 No. 6 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Voi. XLVI Paces 253-302 Ocroser 28, 1965 No. 6 


New Asiatic And African Caecilians With Redescriptions Of 
Certain Other Species 


By 


Epwarp H. Taytor 
This study was made under a grant (GB-1349) from the National Science Foundation. 


ABSTRACT 


This paper deals with caecilians of southeastern Asia and the African islands 
of Sao Thomé and Rolas in the Gulf of Guinea. Four species are redescribed: 
Ichthyophis glutinosus (Linnaeus), Ichthyophis hypocyaneus (Van Hasselt), 
Ichthyophis weberi Taylor, and Schistometopum brevirostris (Peters). A neo- 
type is designated for /. weberi. The following are described as new: Ichthyophis 
atricollaris (Borneo), 1. elongatus (Sumatra), 1. biangularis (Borneo), 1. pesudan- 
gularis (Ceylon), I. asplenius, Borneo, I. forcarti (Ceylon), 1. orthoplicatus (Cey- 
lon), 1. billitonensis (Billiton); and Schistometopum ephele Sao Thomé. 

Ichthyophis hypocyaneus and Schistometopum brevirostris are rehabilitated 
names removed from the synonymy of Ichthyophis glutinosus and Schistometopum 
thomense respectively. 


INTRODUCTION 


In the preparation of a taxonomic treatment of the Gymnophiona of the 
world, I have been reviewing the caecilians in many of the museums of the 
world and collecting in those areas where members of this order are to be 
found. As was to be expected certain forms, presumably undescribed, have 
been encountered and a part of them are herein described. 

A neotype is designated for Ichthyophis weberi whose type was destroyed 
in the Bureau of Science in Manila in World War II. 

I have had the fortune to study the type of Caecilia hypocyanea Van Has- 
selt preserved in the Leiden Museum. This is, I believe, a recognizable species 
and is removed from the synonymy of Ichthyophis glutinosus, a species that 
has been a most remarkable catch-all during taxonomic history. Another 
species, described as Siphonops brevirostris Peters and later placed in the 


254 Tue UNiversity ScrENCE BULLETIN 


synonymy of Schistometopum thomense, is resurrected and designated 
Schistometopum brevirostris. Unfortunately however the type of brevirostris 
was not available for examination when I was last in Berlin. 
A list of forms treated in this paper are: 
Ichthyophis glutinosus (Linnaeus ) 
Ichthyophis hypocyaneus Van Hasselt 
Ichthyophis atricollaris sp. nov. 
Ichthyophis elongatus sp. nov. 
Ichthyophis biangularis sp. nov. 
Ichthyophis pseudangularts sp. nov. 
Ichthyophis asplenius sp. nov. 
Ichthyophis forcarti sp. nov. 
Ichthyophis webert Taylor 
Ichthyophis orthoplicatus sp. nov. 
Ichthyophis billitonensis sp. nov. 
Schistometopum ephele sp. nov. 
Schistometopum brevirostris Peters 


ACKNOWLEDGMENTS 


In my caecilian studies of the past three years I have placed myself under 
obligation to many persons and institutions. This very considerable list will 
appear in the general work which I have in preparation. Those persons and 
institutions that have helped with this preliminary study or who have pro- 
vided specimens are mentioned here. 

I desire to acknowledge my deep obligation for information or for many 
kindnesses while visiting their museums; and my grateful thanks for the 
loan of specimens and the privilege of describing new forms: 

To Dr. Hialmar Rehndahl and Miss Greta Vestergren of the Royal 
Natural History Museum, Stockholm, Sweden, for pertinent information on 
the type of Caecilia glutinosa (Linnaeus). 

To Dr. L. D. Brongersma, Director of the Leiden Museum, who loaned 
the type of Caecilia hypocyanea Van Hasselt, and permitted me to describe 
the Bornean Ichthyophis atricollaris and Ichthyophis asplenius from the 
Leiden Museum collections, and who furnished much helpful information 
regarding the work of Kuhl and Van Hasselt in Java, and the publications 
on their collections; also to Dr. H. Boschma and Dr. M. Boeseman of 
Leiden for many courtesies at the Museum. 

To Dr. George S. Myers, Division of Systematic Biology, Stanford Uni- 
versity, and Dr. Alan E. Leviton, Associate Curator of Zoological collections 
for loan of material from the Stanford collections, especially the only known 
specimens of Ichthyophis weberi (practically topotypes), one of which has 
been named a neotype of the species. 

To Dr. Lothar Forcart of the Museum of Natural History, Basel, Switzer- 


New AstaTIC AND AFRICAN CAECILIANS 255 


land for the loan of significant specimens with the privilege of describing 
Ichthyophis forcarti, and Ichthyophis pseudangularts. 

To Dr. Josef Eiselt of the Natural History Museum in Vienna for loans 
and the privilege of describing Ichthyophis elongatus; and also for other 
specimens. 

To Miss Alice G. C. Grandison for the privilege of describing Ichthyophis 
biangularis; also for the loan of a number of other significant specimens. 

To Mr. Hellenius of the Amsterdam Zoological Museum for the loan of 
specimens, one of which, Ichthyophis billitonensis, is described as new. 

To Dr. Enrico Tortonesi, Director of the Civic Museum of Genoa for the 
loan of significant specimens, one of which has been named the type of a 
new species, Schistometopum ephele; and one a specimen of S. brevirostris 
Peters which convinces me that Peters’ species merits specific designation. 

To the Director of the Zoological Survey of India and Miss Mira B. 
Kirpalani for the loan of specimens, one of which is named the type of 
Ichthyophis orthoplicatus. 


Ichthyophis glutinosus (Linnaeus) 
Frese 1i22 5: 


? Serpens caecilia ceylonica Seba, Locupletissimi rerum naturalium thesauri ac- 
curata descriptio, et iconibus artificiosissimis expressio, per universam physices 
historiam, vol. 2, 1735, p. 26, pl. xxv, fig. 2 (Ceylon). 

Caecilia glutinosa Linnaeus, Museum S.R.M. Adolphi Friderici Regis Svecorum— 
in quo Animalia rariora imprimis et exotica: Quadrupedia Aves, Amphibia, 
Pisces, Insecta, Vermes describuntur et determinantur labine et svetice cum 
Iconibus jussu Reg. a. Car, Linnaeo, 1754, p. 19 ((“In Indus”); Caroli Linnaei 
Systema Naturae per Regna tria Naturae secundum Classes, Ordines, Genera, 
Species, cum Characteribus, Differentiis, Synonymis, Locus, Ed. 10, reformata, 
1758, vol. 1, p. 229 (type-locality “In Indiis”); Andersson, Bihang Till. K. 
Svenska Vet. Akad. Handlingar, Bd. 24, Afd. 4, No. 6, 1899, p. 6 (type of 
glutinosus). 

Ichthyophis glutinosus Taylor, Univ. Kansas Sci. Bull., vol. 40, Apr. 20, 1960, 
pp. 38-39. 

The name [chthyophis glutinosus* has been used in more than a hundred 
articles applying to perhaps ten or more different species. Some of these may 
actually deal wholly or in part with g/utinosus but most of them do not. To 
make an authoritative assignment of each, it would be necessary to examine 
the material studied or referred to by each author. 

In a paper dealing with Asiatic caecilians published in the University of 
Kansas Science Bulletin (Vol. 40, April. 20, 1960, pp. 37-120), I pointed out 
that the Linnean description of Caecilia glutinosa was somewhat inadequate 
to identify certainly this species among the several forms that exist in Asia. 


*See Bourret, Les Batraciens de |’Indochine (l'Institut Oceanographique de Indochine) 1942, 
p. 136, for a listing of articles. 


256 Tue University ScIENCE BULLETIN 


Thus Linnaeus states that he is unable to observe tentacles (“Cirrhos nullo 
observare potui” ); teeth are not mentioned as occurring on the lower jaw; no 
mention is made of the presence of scales; characteristics of the nuchal collars 
are not mentioned and the locality given “Habitat in Indus” might be any- 
where in India, Ceylon, southeastern Asia or Islands of the Indo-Australian 
Archipelago. 

Seemingly the type was not seen by subsequent writers on caecilians until 
studied by Andersson when he published his “Catalogue of Linnean type 
specimens of snakes” in 1899.* Andersson gives the length as 400 mm, 
(Linnaeus had said “Longitudo pedem superat’), and gives 355 as his count 
of transverse folds, Linnaeus’ count having been 350. 


In 1959 I appealed to Dr. Hialmar Rendahl of the Royal Natural History 
Museum, Stockholm, Sweden for further data on the type specimen. The 
matter was placed in the hands of Miss Greta Vestergren. She has had the 
kindness to forward certain data on the type which I present here: 

“T must call your attention to the fact, that the specimen in question is 
not in exceedingly good condition. The snout is broken, most of the teeth 
in the upper jaw are lost, and moreover the specimen seems to have been 
dried partially, on account of which the fold-limits on the anterior part of 
the ventral surface are rather difficult to distinguish.” 

“Measurements in mm.: 

“Total length 394; length from front of vent to tip of tail, 5.4; length of 
head from tip of snout to first groove, 14; width of head at first groove, 12.2; 
width of body at near middle, 17.2; distance between eye and tentacle, 2 mm. 
Tentacle to nostril? (snout broken); level of eyes to tip of snout, 6.2; diameter 
of eye, 0.7; distance between eyes, 7.5. 

“Scales begin in the first and second folds, each fold having two scalerows. 
The third fold has three scalerows, the fourth fold four rows; while through- 
out almost all of the remaining folds of the body there are six scalerows.” 
Thus the scalerows total something more than 2000. 

“The count of the transverse folds on the middle of the right side of 
dorsum (including tail) is 359; on the middle of the left side, 367; lower on 
sides, 355-365; count on venter 357 (?).** 

“Accurate count of the teeth was difficult to obtain and those that follow 
must be regarded as approximations: maxillary-premaxillary, total, 18-18; 
vomeropalatine, 16-17; dentary teeth, 20-17, spenial teeth, 11-12. 

“The details of the color-markings are as follows: 

A yellowish lateral stripe 4.5 mm. wide reaching to the tip of the tail; it 
does not connect with any yellowish spot at vent; the stripe is broken on the 


* Bihang Till K. Svenska Vet.-Akad. Handlingar Bant. 23, Afd. 4, No. 6, 1899, p. 6. 
** These differences in counts are due to splitting of the folds. The counts range from 355 to 
367 at various points. 


Uy 


New AsIATIC AND AFRICAN CAECILIANS 


‘ 


prey jo 


MITA [VIO] 45) speoy jo MoIA [eNU 


OA 


‘ 


“WLU C°Z] 


d 


‘ 


pesy fo 


‘ 


QA0018 [LYINU IsIY yv peasy JO WYpIA, “UOIsoI [epneoqns pue Apoq jo snutwsai ‘q 


MOIA [esIOp Sy ‘WO]AI_) 


i -Old 


if 


I 


‘ON 


uimasnyy joseq ‘“(snovuury) svsouiunps siydody1y2] |, ‘og 


258 Tue UNIversiry SCIENCE BULLETIN 


neck by dark color, but narrowed, reaches beyond this, at least to the 
tentacle. At the jaw-angle the yellow divides, one narrow branch passing 
along the edge of the lower jaw to the tip. The stripe widens somewhat on 
the neck. The color of the entire ventral surface is a lighter brown than 
the brown of the dorsum. The eye is distinct on the left side (scarcely so 
on the right), and surrounded by a light ring. There is no light anal spot. 
It is impossible to say whether or not light spots were present on tentacles 
and nostrils. No glands are evident at the sides of vent. The sex is in- 
determinable. 

“Concerning the folds it might be added that they are complete under 
the abdomen at least on the hinder two thirds. As I mentioned previously, 
the fold-limits are rather indistinct on the mid-ventral region of the anterior 
part of the body. The folds form a sharp angle on the abdomen pointing 
backwards except for an area in front of vent. On the anteriormost part of 
the back they may form an angle.” 

Diagnosis: A relatively short thick-bodied species, the body width in total 
length 19 to 21 times in adults; transverse primary and secondary body folds 
approximately from 348 to 369 ventral count, 354 to 382 dorsal count (rarely 
counts may be higher or lower). Color in life bluish gray to ultramarine 
often turning brown in preservatives. Eyes represented by grayish-white 
areas without visible details of lens or pupil; scalerows six or seven in each 
fold throughout most of the body, averaging a total of 2000 or more scale- 
rows. Splenial teeth 11-11 to 14-14. 

Description of species (from Musée d'Histoire Naturelle Bale (Suisse) 
No. 1, Ceylon, collected and studied by Paul and Fritz Sarasin): Body some- 
what flattened; head rather small, the eyes small, evidenced only by a 
slightly elevated gray area (no details of lens or pupil discernible); tentacle 
near edge of lip, separated from eye by a distance of 2 mm., from nostril by 
3.8 mm.; distance between eyes, 7 mm.; distance from level of eyes to tip 
of snout, 6.1 mm.; eye to nostril, 5 mm.; eye to mouth-angle 4 mm.; two 
nuchal collars, the first limited by the first groove going straight across throat 
and curving back a little and passing completely around occiput; second 
groove curving strongly forward on throat and up on sides of head to about 
level of jaw; third groove limiting second collar moderately distinct around 
neck; from tip of snout to first groove (lateral measurement), 13 mm. 
(median ventral, 11 mm.), (dorsal, 13 mm.); to second groove (lateral), 17 
mm.; (ventral, 14 mm.); to third groove (lateral), 20.3, (ventral, 21 mm.); 
(dorsal, 18 mm.). The first collar has a faint median transverse dorsal 
groove; the second collar has two posterior folds, first reaching down to | 
mouth level, the second reaching venter. 

Total transverse folds following the two collars, 342 (dorsal count), 328 
(ventral count); nine folds confined to tail. The grooves and folds on 


New ASIATIC AND AFRICAN CAECILIANS 259 


anteriormost dorsal part of body curve forward mesially and sometimes form 
a slight angle on dorsal part of body; but ventrally in anterior half of body 
the grooves are dim or interrupted, although the folds can be seen to cross 
completely and form a backward-directed angle; posteriorly the grooves 
are clearly seen and tend to go straight across venter for some distance in 
front of vent; two slight elevations present on sides of vent suggesting glands. 

One to four scalerows in the folds of the first three centimeters of body; 
beyond this, folds with five to seven (sometimes eight) scalerows, reaching 
a total of approximately 2000 scalerows. Anteriorly scales very small, trans- 
versely widened; posteriorly they are larger, more nearly cycloid, but in any 
fold the scales of one row may vary in size and overlap laterally, or in places 
they may not even touch. 

Dentition: Maxillary-premaxillary, 20-21; vomeropalatine, 20-20; dentary, 
18-19; splenial, 12-12. The teeth relatively small for the size of the species, 
the anterior dentaries a trifle larger than the premaxillaries; tongue broad, 
with some longitudinal plicae, covering splenial teeth completely, the edges 
free; choanae somewhat elongate, oval, the greatest width of one, in distance 
between them, approximately three times. 

Color: Above bluish-gray uniform on all dorsal surfaces; venter slightly 
lighter but almost same shade; a yellow (now faded) lateral stripe about 2.8 
to 3.2 mm. wide extending posteriorly to or almost to extreme tip of tail but 
not joining a dull cream spot on vent; anteriorly on collars the stripe is inter- 
rupted completely, but anterior to collar it widens and bifurcates, one upper 
part reaching a short distance along the upper jaw, and lower part passes 
on lower jaw for a distance; ill-defined light spots on tentacle and nostril. 

Measurements in mm.: Total length, 303; tail, 5.7; width of head at eye, 
9; width of head at first groove, 12.5; length of head to first groove (lateral 
measurement), 13; width of body, 16; width in length, 19 times. 

Variation: The specimen described agrees in most characters with the 
type. The change from the ultramarine or bluish gray color to the brown 
color of the type appears in some recently collected specimens. It is con- 
ditioned by the character of the preserving fluid and perhaps also to light 
exposure and slight desiccation. This is true also of many species of Ichthyo- 
phis that are of a violet or lilac color in life. However, in many if not most 
cases the original color or a part of it remains. 

In several species the three or four anterior folds following the collars may 
form a more or less distinct forward-directed angle. At least one species 
(herein described) has folds forming distinct dorsal angles throughout much 
of the body. 

Several species of the genus have the primary and also the secondary folds 
(of which there may be three or four to each primary) tending to split 


260 Tue University SciENCE BULLETIN 


FIG.2 


Fic. 2. Ichthyophis glutinosus (Linnaeus). Basel Museum No. 1, Ceylon. A, palatal region of 
mouth and upper jaw; B, lower jaw and tongue. 


dorsally or laterally. Counts made at different levels vary considerably, 
sometimes as much as a difference of twenty. 

The larvae of glutinosus are relatively wider than those belonging to most 
other species of Ichthyophis. The eyes are represented by circular milky- 
white spots, slightly elevated, and a whitish stripe tends to connect eye to 
tentacle. A light area is present below and about the nostril. A low fin 
begins at the 16th preterminal fold, and extends to the tip of the tail and 
slightly below it. Two separate lateral gill-slits are present on each side. The 
neuromast system of the head and neck is more or less complete in the fol- 
lowing specimen: (Natural History Museum Basel, No. 8, Ceylon): Total 
length, 121 mm.; width of head, 6.2; head length, snout tip to Ist nuchal 
groove, 7.2; width of body, 7.5; width in length, 16 times. 

The penis, developed in the posterior part of the gut, is extrusible. A 
figure is given here showing its general appearance in situ, unextruded (from 
Basel Museum, No. 4). 

Variation in measurements etc. are indicated in the following table of 
data. 

Remarks: Another species, of a violet to lilac color, occurs also in Ceylon. 
It reaches a length perhaps as great as glutinosus but is slenderer and has 
fewer transverse folds. 

As more than a single species was available to the Sarasin brothers, who 
seemingly were unaware of these differences, it is not impossible that their 
work is based on more than a single species.* However I have no certain 
evidence that this is the case. 


* As has been suggested by Dr. L. S. Ramaswami, Current Science, vol. 16, Jan., 1947, p. 8-10. 


New AsIATIC AND AFRICAN CAECILIANS 261 


Taste |. Measurements and data on Ichthyophis glutinosus. 


Museum Number Basel 4 (6124) K.U.31291+ K.U.31290+ K.U.31283+ K.U.31293} 
DGXS fe -secscescecs ics Bidecssisisdieai: a g 2 ) g 
otal wengthy y= 345" 320 300 275 263 
Manllilemethy eee Tie, 6.7 7. 6 6 
Width) of thead) 2.2... 2. = 10.6 10.2 9.6 8.8 
Wadthimote bodys 22 ee. 17 en 13.8 13}533 12 
Width in length, t mes -.._........ 20 2D. Zile7, 20.6 PS) 
Snout to) list groove -_-.:.. = 14 13.2 12 13 
Snout to 2nd groove —.. BH 18.2 16 16 I5}33) 
Snout toi 3rd) groove _...... ae 22.6 20 21 18.2 
[Byes HORNE Se eae oe 6.8 6.6 6.3 6.2 
Eye level to snout tip _.... ee 5.6 5.8 5.8 De 
iliemtacley to) eye. =... =p 1.8 1.8 1.8 1.7 
iientacle’to: mostnill ©... Sx: Soll 3.6 Shai 3.5) 
slotaletoldisy ease 355 349-361 348-260 352-354 369-382 
Teall iolKolg =2 ea eee ) 7.8 8 9 7-8 
Max:-premax., teeth! -.--... 23-23 23-23 26-27 20-21 
Womeropalatime  —_...-......- 24-25 23-23 24-24 25-26 
HD) GT CA Iny Pipes ieee ee ee 22-22 21-21 20-20 21-22 
Splemial ees cs eet. 12-12 14-14 11-13 13-14 
Posterior scalerows 7 6 6 ) 


6 6 


* Estimated length, head missing and perhaps certain anterior body folds. 
+ Kansas Univ. Specimens from Tonacumbe Estate, Numunukula, Ceylon. W. W. Phillips, col- 
lector. 


Ichthyophis hypocyaneus Van Hasselt (in Bote) 


Ichthyophis Hasselti Fitzinger, Neue Classification der Reptilien nach thren 
natirlichen Verwandtschaften, Wien, 1826, p. 63 (nomen nudum; Java). 
Caecilia—Van Hasselt, Algemeene Konst Letter-Bode voor Het Jaar 1823, No. 41, 
Vrijdag den 10 den October (Letter, Kuhl and Van Hasselt); (Reprint of 
above in! French), Bull; Sci. Nat. et \Geol.,. 2nd sec; vol. 2) Panis, p. =. 

(dated Ceram Province of Bantam, Feb. 1, 1823). 

Caecilia hypocyanea Van Hasselt, in H. Boie, in F. Boie, Isis, 1827, p. 565; I. 
Miller, Isis, vol. 22, 1829, p. 875; idzd., vol. 24, 1831, pp. 707-710; S. Muller, 
Zeitschr. fiir Phys., vol. 4, p. 195; Arch. Anat. Phys., 1835, p. 391, pl. 8, figs. 
12-14; Schlegel, Abbildungen neuer oder unvollstandig bekannter Amphibien 
nach der Natur oder dem Leben entworfen herausgegeben und mit erldutern- 
den 1837-1844, p. 119, pl. 39, fig. 1 (entire animal depicted). 

Epicrium glutinosum: Duméril and Bibron, Erpétologie générale, vol. 8, 1841, pp. 
286-287 (part.). 

Ichthyophis glutinosus: Boulenger, Catalogue of the Batrachia Gradientia s. 
Caudata and Batrachia Apoda in the collection of the British Museum, ed. 2, 
1882, pp. 89-91 (part. but not the figure). 

The first description of this species appeared under its generic name only 
and is a descriptio nudus. This was in a letter of Van Hasselt’s that was 


FIG. 3 


New AsIATIC AND AFRICAN CAECILIANS 263 


published (Algemeene Konst en Letter-Bode, voor Het Jaar 1823, No. 41, 
Vrijdag den 10 den October. Letter Kuhl and Van Hasselt) in which men- 
tion is made of a caecilian. I have not seen this publication but translated 
into French it was reprinted in the Bulletin des Sciences Naturelles et de 
Geologie, (deuxieme section) Bulletin Universal des Sciences et de l'Indus- 
trie, vol. 2, Paris. The title was “Fourth letter upon the reptiles of Java. 
Dated in Ceram | Serang| Province de Bantam.” 

I have seen the French translation, part of which, dealing with the 
caecilian, is here given in free English translation. 

“I found this reptile in the wet and marshy places on the northern coast 
of Bantam [Java]. The Malays call it Octur-doeél and do not fear it. Its 
tongue is not extensile or visible except when the mouth is open and must be 
regarded as rudimentary. The teeth are very small, curved backwards and 
arranged in several rows. The eyes are hidden under the skin and are very 
small. The head is equally small and of a width equal to the trunk. In front 
of the eyes there are small elongations of the skin (tentacles) one half-line in 
length, which the animal is able to evert. The anus is found near the ter- 
minus of a very short tail rounded to a point, which appears to have a great 
sensitiveness. The scales are invisible at least in a living (fresh) state, and 
the skin is smooth and viscous, ringed by transverse grooves quite similar 
to those of annelids. The rings in the anterior part of the body are inter- 
rupted below. The color above is a dark olive and below steel-blue. The sides 
of the body are decorated by two longitudinal lines of ochreous yellow spots. 
Length to vent, 0.78 sic;* of the tail, .02. 

A larval specimen of a caecilian in the Vienna Museum now bears the 
following data: “Epicrium hasselti, Wagler, Java, 1825, III, 68.” The present 
catalogue number is 9097. The catalogue shows that it was obtained from 
the Leiden Museum by Natterer in March 1825 and thus may have been seen 
by Fitzinger. A nomen nudum may not be considered as having a type, else 
this specimen would be so considered. The specimen seemingly was one 
sent from Java by Van Hasselt. 

The “legal” description of Caecilia hypocyanea was published in 1827 in 
a paper prepared largely if not wholly by H. Boie, who had access to Van 
Hasselt’s notes, but which was actually published by his brother, F. Boie. The 
fact that he has added Van Hasselt’s name after the scientific name makes it 
necessary that Van Hasselt be regarded as the author of the name. It is true 
that when his description was originally published (/oc. cit.) under the 
generic designation Caecilia it was a descriptio nudas and the type-descrip- 
tion must date from F. Boie’s work in 1827. Van Hasselt had given it a name 
in his unpublished notes. 


*In the original Dutch publication this number was given 0.98 and is presumably correct. 
T am uncertain as to the unit of measurement used. 


264 Tue University ScrENCE BULLETIN 


Schlegel, Joc. cit. gives certain pertinent data, stating that Van Hasselt 
saw two specimens from Bantam. One was about 10 inches long (taken on 
muddy soil). The back was brilliant black-olive while on the underside it 
was steel-colored. The lateral stripes were ochre-yellow. The other was a 
young specimen, somewhat red-brown on the dorsal side and olive on the 
venter, which had been caught in the Loudemanik River. He states that Van 
Hasselt does not mention gill-slits or a fin. The young breathe by gills but 
these are lost as the animals become developed. It is probable that Schlegel 
did not know of the specimen sent to Vienna. 

Now it would appear that the larva mentioned is the specimen in the 
Leiden Museum earlier studied by Miiller (Isis, 1831, vol. 24, p. 710). He 
had observed its gill-slits and noted its other characteristics such as the 
branchial arches. He concluded that despite the resemblance of these animals 
to reptiles, they were really amphibia and for the known caecilians he pro- 
posed the name Gymnophides as the first Order of the Amphibia. 

This small specimen is still extant (Leiden Museum number 2409).* It 
measures 115 mm. in length and has approximately 311 transverse body folds. 
A part of the neuromast system is still in evidence on the head. The dental 
formula: Maxillary-premaxillary, 13-14; vomeropalatine, 18-18; dentary, 16- 
17; splenial, 8-8. There are actually two gill-slits on each side, the anterior 
one the smaller. 

Van Hasselt’s notes sent from Java and preserved in the Leiden Museum 
have been copied for me by Dr. Brongersma, Director of the Leiden Museum: 
1. He refers to the preceding small specimen giving it a specific name stating 

that it was found in the Loudemanick River in July, and then a ques- 

tion—‘“the young of the following?” Then follows a short Latin descrip- 
tion of the specimen. 

2. Caecilia hypocyanea Ceram [Serang | January, Octur-Doeel | native name |. 
Then follows a statement in Dutch, “Lives in muddy regions and it is 
said to enter into the anus of chickens.” 

Then follows a description in Latin. 

In a previous paper I have stated that Van Hasselt described the species 
C. hypocyanea from the specimen in the Vienna Museum. This I believe now 
to be an error and must conclude that the type actually is a Javanese speci- 
men, now catalogued in the Leiden Museum as No. 2408. This specimen is 
redescribed here. 

Diagnosis: A species with a very narrow lateral stripe broken minutely 
in several places, appearing to be composed of a series of ochre-yellow spots; 
blackish olive dorsally, steel-blue ventrally in life. Transverse body folds, 
314-316*; distance of tentacle from eye about one half its distance from 
*In Van Hasselt’s notes the young specimen is given a name but he suggests that it may be 


the young of the following species C. Aypocyanea. Fortunately the name has not been 
published since it is indeed the young of C. hypocyanea. 


New AsIATIC AND AFRICAN CAECILIANS 265 


nostril. Maxillary-premaxillary teeth, approximately 20-20; splenials, 11-11. 
Folds incomplete anteriorly, passing completely across back posteriorly, form- 
ing an angle ventrally. The width in length about 26 times. 

Description of type (Leiden No. 2408): Head wider than neck but less 
than width of body; eye (lens and iris distinct) surrounded by a slightly 
darker ring and a minute outer ring consisting of a circle of yellowish-white 
glandules; tentacle close to edge of lip, preceding eye, its distance from eye 
(1.45 mm.) about half its distance from nostril (2.9 mm.); distance between 
eyes (5.8 mm.) greater than length of snout from level of eyes to tip. Two 
collars not strongly defined; first nuchal groove passes around back of head; 
second ascends sides to a point above mouth angle; second collar fuses with 
first dorsally and has one or two transverse folds on its posterior dorsal part. 
Anterior skin of head, neck and part of body glassy smooth. Following the 
collars, 314-316 transverse primary and secondary folds, incomplete dorsally 
on anterior third of body, and likewise incomplete on median ventral part; 
on middle third folds practically complete, those above slightly sinuous, those 
below forming an angle pointing backwards. On posterior third folds and 
grooves pass nearly directly around body, or more anteriorly may have an in- 
distinct angle below while in front of vent they may actually curve forward; 
six or seven folds confined to tail which terminates in a small pointed tip; 
vent longitudinal its sides denticulate, the area slightly swollen. 

Scales appear in the first or second fold, two rows being present in tenth 
fold; two rows present at middle of body with some small scattered scales; 
four rows in posterior folds. 

Dentition: Maxillary-premaxillary teeth, 20-20; vomeropalatine, 19-19; 
dentary, 19-19; splenial, 11-11; latter teeth as large as the dentary teeth, and 
at least a part of the vomeropalatine series as large as the maxiliary-pre- 
maxillaries. 

Color: In life, dark olive above, steel-blue below with broken ochre-yellow 
lateral lines. The specimen is now brown, the neck being darkest; head some- 
what olive-gray above; venter a lighter shade of brown except that on under 
part of tail and a little area in front of the vent the color is bluish gray. The 
narrow lateral yellowish lines are rather dim and the numerous breaks in it 
are hardly discernible. 

Measurements in mm.: Length snout to vent, 258; tail length, 5; width 
of head, 8; width of neck, 6.8; width at middle of bedy, 10; tip of snout to 
first groove (lateral), 10; to second groove, 12.8; to 3rd groove, 16.8. 

Remarks: The characteristics of the folds broken above and below suggest 
a relationship with the Bornean asplenius. However the presence of splenial 
teeth would seem to deny such a relationship. 


*Van Hasselt’s count of the folds was 320. This no doubt has included the two collars and 
the two folds across the second collar. 


Tue UNiversiry ScrENCcE BULLETIN 


266 


qns pue Apoq jo snuruuta, ‘q ‘pray ay Jo smota sosyi “OD “dq “V “80FC CN Whosny 


v ‘Sis 


(wut g ‘peoy Jo 
uaplay ‘adAy, “3fassepy 


YIpIm [emdy) “uoTsor [epneo 
uea snauvdzoddy siydosyiy7] “4 “Oy 


New AsIATIC AND AFRICAN CAECILIANS 267 


Ichthyophis atricollarts sp. nov. 
Fies. 5, 6. 


Type: Rijksmuseum van Natuurlijke Historie, Leiden, Holland, No. 
10684; Long Bloee, Boven Mahakkam, Borneo. 

Paratypes: Nos. 6912 D, Upper Mahakkam, and 10685, Long Bloee, 
Boven Mahakkam, Borneo. 

Diagnosis: A broad yellow stripe extending from second collar, terminat- 
ing at level of vent; head nearly uniform dark brownish violet lacking light 
spots at eye, tentacle and nostril; neck nearly uniformly dark above and 
below, the collars only dimly indicated; body dorsally uniform brownish 
violet slightly lighter below and on chin and neck. Primary and secondary 
folds, 263-300 (ventral count), 275-310 (dorsal count), complete above 
throughout, incomplete on anterior fifth of venter, the grooves not crossing 
above or below on anterior two thirds of body. Splenial teeth present, 8-8. 

Description of type: Head short, rather broadly oval anteriorly; eyes 
small, the distance between them (6 mm.) greater than length from anterior 
level of eyes to tip of snout; tentacle from mouth (0.25 mm.), the distance 
from eye (1.52 mm.), much less than its distance from nostril (3 mm.) ; snout 
projects one millimeter beyond the mouth. First and second collars distin- 
guishable on sides of neck but more or less fused together dorsally, less so 
ventrally. Primary and secondary folds together number 263 (ventral count), 
275 (dorsal count), the folds complete, except for those on anterior mid- 
ventral region, the grooves however distinct on sides and on latter third of 
body; eight folds ventrally, ten dorsally on tail, the vent interrupting only 
three folds below. Scales begin at about the 25th fold, one or two scales being 
present near middorsal line; at middle of body three or four rows varying 
somewhat in the same fold dorsally and laterally; posterior fifth of body with 
seven to eight rows. 

Dentition: Maxillary-premaxillary teeth, 22-22; vomeropalatine, 22-23; 
dentary, 19-20; splenial, 8-8. Dentary teeth largest. Choanae relatively small, 
the transverse diameter of one (.4 mm.) in distance between them (2.5 mm.), 
about 6 times; splenials on a high splenial ridge about on same plane as 
dentary teeth. 

Color: Above brownish violet, the edges of the transverse primary and 
secondary folds a little lighter, especially so in the latter half of body; ventral 
surface a lighter shade; head lighter than dorsum, nearly uniformly colored 
above and below, with a vague suggestion of a lighter eye circle and a scarcely 
discernible lighter region about nostril; a vague small yellowish spot near 
mouth-angle; a white spot surrounding vent; a lateral yellow stripe from 
second collar to level of vent, the edges barely visible from above, narrowing 
a little anteriorly. 


(WUT 676 ‘peey 9u) 
jO YAplar yenjy) ‘worso1 [epnroqns ‘q, $pesy] oy} Jo sMoIA Gomtp ‘DO “dq. V “+8901 “ON wWresny Usps] radd yz, *aou ‘ds seppjostgn stydodyiy 7] *¢ “Ol 


Z 
= 
a 
= 
4 
5) 
sa 
ww 
S 
Z 
= 
S) 
op) 
> 
= 
Z 
ss 
a 
5 
a 
iar, 
= 


New AsIATIC AND AFRICAN CAECILIANS 269 


A _ B 
FIG. 6 


Fic. 6. Ichthyophis atricollaris sp. nov. Type. Leiden Museum No. 10684, Long Bloee, Boven 
Mahakkam (river), Borneo. A, upper jaw and palatal region of mouth; B, lower jaw 
and tongue. (Actual width of head, 9.9 mm.) 


Variation: Variation in measurements and tooth numbers is given in the 
following table. In color there is a strong similarity. The dark neck is the 
same in all, the yellow streak stopping abruptly at the 2nd collar. Dorsal 
folds tend to cross the back in a straight line. 

In none of the specimens are the collars strongly marked, but the smallest 
shows the first plainly on the underside of the neck. No. 10685 has a short 
longitudinal whitish line under the first collar. This specimen also has the 
folds somewhat less distinct than the others. 


Taste 2. Table of measurements of [chthyophis atricollaris. 


Number 10684 10685 6912(D) 
Museum Leiden Leiden Leiden 
Diotaltelength 20s. 2 ee ee! 285 255 204 
Mrrclae emote smeecta Niet tegen calle <. 4.8 Dal 4 
Width of body Bo Set sete Oa ee eel tos Hee oe 11 915 8 
Wiactimotelreac es aes Sete oe a Tae 9.9 9 oll 
SMOUtMtIpNtO nist) @LOOVE! .- eecs ee ee 10 11 9 
DSMTOU ttl Py tOn2 MG STOO VE) a. -2eeeee ss eecse cere eee se7, 13.2 1] 
SHOUtetiyp to STC groove! 2M... s ee : M72 17 13.4 
BifemitaclemtOuey.clrs tert ner ee ese ee Ne 152 125) 1.25 
Mientacles top mostrill 222-2 ges ee, 3 2.5 De 
WB) AD) Vv 
Bite ola sien eet tue ree ee SR 263-275 300-310 293-303 
rolcsmonatailane een. Pee eee stele 8-10 8-9 8-8 
Maxillary-premaxillary teeth -....../.....-...-.-. 22-22 25-27 24-24 
Momlenopalatine, . - bie sei el 22-23 20-20 20-19 
MeO Cary geen ered et ee 19-20 19-19 18-19 
‘Seplkemiianl| oS PW so pe Ea 8-8 8-5-+ 8-8 


Madtheanslength (times)! -...2.2--.20222.2---- 26 26.8 25.5) 


270 Tue Universiry SciENcE BULLETIN 


In all the angle of the folds on the venter is less acute than in most other 
species now recognized in the genus Ichthyophis. 

The type specimen was taken on the Upper Mahakkam River, at Long 
Bloee, by the Nieuwenhuis Borneo Expedition. The paratype No. 10685 is 
from the same locality and No. 6912 (D) from Boven Mahakkam; the exact 
locality may or may not be known. 

Three other species were taken in the same general area. At least two 
other species seemingly were taken at the same place since No. 6912 is one 
of a series of five caecilians bearing this same number and locality. 


Ichthyophis elongatus sp. nov. 
IGS a/90: 


Holotype: Naturhistorische Museum Wien. No. 9094. Padang, Sumatra. 

Paratype: N.M.W. No. 9092, Domenik, Sumatra. 

Diagnosis: Head a little wider than body. Transverse folds, 274-290, 
three confined to tail; folds not complete across anterior two thirds of body; 
maxillary-premaxillary teeth, 28-32; splenial, 14-16; latter half of body with 
six scalerows in each fold; body width in length 37 to 40 times; a narrow 
lateral yellow stripe. Tip of tail acuminate. 

Description of type: Head rather wider than body; eyes distinct with a 
narrow white ring; distance of tentacle to eye (1.65 mm.) a little less than 
half its distance to nostril (3.45 mm.); eye to nostril, 4.5 mm.; tentacle from 
lip, 0.3 mm.; first nuchal groove distinct below and on sides, dimly visible, 
curving forward on back of head; second groove distinct below and on sides; 
third groove visible across back and on venter except in mid-ventral region; 
the two collars fused together dorsally, the second divided dorsally by a trans- 
verse groove visible below only at sides. 

Folds following collars, 274 (dorsal count), 277 (ventral count), the folds 
and grooves incomplete on the dorsal surface; anteriorly separated by a 
distance of three millimeters which gradually narrows until near middle of 
body, they become complete; except for two anterior centimeters, the folds 
are complete ventrally, forming an acute angle pointing backwards for about 
seven eights of the body length. The ventral grooves are not in evidence and 
the complete fold and angle often cannot be discerned unless the specimen is 
somewhat desiccated. Only three folds on tail, two and a half interrupted by 
the longitudinal vent; a pair of glandular swellings near anterior end of 
vent. 

Small scales begin on second collar; at middle of body seven rows present, 
the scales large, variable in size, and these continued to near vent; seven or 
eight rows posteriorly. 

Dentition (of type and paratype): Maxillary-premaxillary, 32-32, 29-28; 
vomeropalatine, 27-27, 30-30; dentary, 29-29, 27-27; spenial, 14-14, 16-16. 


New AsIATIC AND AFRICAN CAECILIANS 271 


2 via 


Fic. 7. Ichthyophis elongatus sp. nov. Type. Vienna Museum No. 9094, Padang, Sumatra. 
A,B,C, three views of the head; D, subcaudal area. (Actual width of head, circa 
10 mm.) 


iS) 
Pal 
bo 


Tue University SciENCE BULLETIN 


FIG. 8 


Fic. 8. Ichthyophis elongatus sp. noy. Type. Vienna Museum No. 9094, Padang, Sumatra. 
A, Upper jaw and palatal view of mouth. (Actual width of head, circa 10 mm.) 


Choanae semilunate, lateral, the diameter of one in distance between them, 
4% times; tongue narrowed anteriorly, not covering spenials. 

Color: Dark lilac dorsally, somewhat grayish lilac of a slightly lighter 
shade ventrally; a yellow to cream lateral stripe about one millimeter wide 
reaching to near eye, terminating anterior to vent, the stripe broken on second 
collar; a white spot surrounding vent; a light ring about eye; a light spot at 
nostril and tentacle, and a light area above tentacle; top of head somewhat 
brownish. 

Measurements in mm. (type and paratype): Total length, 300, 280; tail 
length, 3.4, 3.5; width of body, 7.6, 8; width in length (about), 40 times, about 
37 times; snout to first groove, 12, 11.2; to second groove, 14.9, 15.9; to third 
groove, 19, 19.8; eye to eye, 6, 6; eye level to snout tip, 5.4, 5. 

Variation: The transverse body folds of the paratype are 287 (dorsal 
count) and 290 (ventral count). The tail has three folds with a fourth in- 
distinctly indicated near the tip. In both, the body and tail are narrowed, the 
width of the tail at vent being three millimeters. It is flattened ventrally 
ending in a fine point. 

In the paratype the general lilac color is of a lighter shade and the lateral 
stripe is somewhat more distinct and a trifle wider at least at certain points; 
the head is somewhat lighter in color and the tip of the snout is cream. The 
second collar dorsally is about one half divided by a transverse groove. 

Remarks: Incomplete dorsal folds appear in another species, occurring in 
Borneo, which, however, lacks vomerine teeth; the number of dorsal folds is 
264-270; and the width in length is about 24 times instead of 37 to 40 times. 


[chthyophis biangularts sp. nov. 


Fics. 9, 10. 


273 


New AsiAaTIc AND AFRICAN CAECILIANS 


« 


peey JO sata 92Iy) ‘OD “g “WY ‘“OsuIOg “yeMeEIES 


W surie 


W 


(‘wut Z 
Vo OSsOl cich ON NesnINe SRM a 


6 92 


ad&y, ‘aou 


g ‘pray JO yIpIA\) “vate [epneoqns *q 


ds sapynsuvig siydodyry7] “6 ‘4 


274 Tue University ScrENCE BULLETIN 


Holotype: British Museum No. 72.2.19.59,A, Matang (Mt.) Sarawak, 
Borneo; Everett, Coll. 

Diagnosis: Eye distinct in socket; tentacle twice as close to eye as to 
nostril; total folds behind collars, 330-333, with ten confined to tail; transverse 
folds distinctly angular above as below, throughout most of body; body width 
in body length about 26 times. Scales begin in first transverse fold, four rows 
at middle of body, four rows posteriorly. Color blackish slate above and 
below with a yellow lateral line terminating anteriorly at second collar; yel- 
low spot at mouth angle not dividing. 

Description of type: (Skin partly removed from head.) Eye dimly dis- 
tinct with a slightly lighter ring about it; tentacle much closer to eye (1.2 
mm.) than to nostril (2.4 mm.); eye to nostril, 3.5 mm. First and second 
collars fused above; collars distinct below, the first groove seemingly inter- 
rupted on dorsal surface; second groove distinct below and on sides, scarcely 
visible from above, bordered below by a light line. Third groove distinct 
above interrupted mid-ventrally where the second collar fuses with the first 
and second transverse folds. 

Counts of transverse folds vary but little. Two counts show, dorsally, 333, 
ventrally, 330; ten folds on tail (from anterior edge of vent), four interrupted 
by vent, six more or less complete behind vent; folds forming a distinct 
angle dorsally and a somewhat more acute angle ventrally, except on the 
fifth of body preceding vent where dorsal and ventral grooves and folds run 
nearly directly across. 

Scales begin in the first folds following collars, at least two rows present on 
the dorsal surface; at mid-body there are three well-defined rows, and a 
fourth complete or incomplete; posteriorly there are four well-developed rows. 

Dentition: The teeth have been exposed and lost or removed. The fol- 
lowing counts are close approximations: Maxillary-premaxillary teeth, 25- 
26; prevomeropalatine, 24-24; dentary, 19-19; splenial, 2-2. 

Color: Generally blackish slate, the venter scarcely less dark than dorsum; 
a dull narrow, irregular-edged stripe on sides, beginning at second collar 
and terminating posteriorly behind level of vent; a yellowish-cream spot at 
jaw angle not branching; a white spot about vent, a very narrow dim light 
ring about eye, and a lighter spot at tentacle (skin in area about nostrils 
removed). 

Skull: The exposed parts of the skull show the eye in a circular socket 
completely separated from the fenestral opening of tentacle; no diastema 
between the squamosal and parietal bones; eye socket between maxillary, 
squamosal and a small supraocular; a preocular and a septomaxilla or “lateral 
nasal”; nasal fused with prefrontal; frontals and parietals distinct, no ethmoid 
visible; premaxillaries distinct with premaxillary teeth, 8-8, followed by 17 
or 18 maxillary teeth on each side; tentacular aperature pierced in maxillary. 


275 


New AsIATIC AND AFRICAN CAECILIANS 


sry ye Apoq JO WIptM) 
-IOp JO sMoIA 3914} ‘O 


if 


*sp[oy Apog asiaasueq 9Y} JO UONTpPUOD 9]v[NSue 
gq ‘V ‘oaulog ‘yemeviesg I Suey “Vy ‘6S 6] 


ol ‘5 


Iq SurMoys 


‘TTL CON Wnasnyy ysnug ‘adXy ‘acu ‘ds simmynsuviq siydodyiy2] “OQ “Ol 


ld 


“WU 76 BATS ‘yutiod 


Apoq jo yed JOIIajUL SY} JO SMITA [eIA\e] pur [eNUDA 


‘ 


[Ps 


276 Tue University ScIENCE BULLETIN 


Measurements in mm.: Total length, 258; tail, 5.8; width of body, 9.8; 
width of head, 8.2; tip of snout to first groove, 10; to second groove, 12.5; to 
third groove, 16.5. Width in length, 26.3 times. 

Remarks: The type number (BM 72.2.19.59) originally included three 
specimens. The other two are larvae and in my opinion do not belong to the 
species described here. The two larvae likewise represent two species. The 
larger larva, here designated A, has 261-256 transverse folds. The folds cross 
the back and are perhaps slightly angulate in the anterior folds. The smaller 
one, here called B, has the folds failing to cross the median dorsal part of 
the body. It may be the young of Ichthyophis asplenius herein described. 

The name diangularis is from Latin, bi=two, angulus=angle. 


[chthyophis pseudangularts sp. nov. 
Fics. 11512513: 


Holotype: Musee d'Histoire Naturelle Bale (Suisse) No. 4412. Collected 
in Ceylon by Paul and Fritz Sarasin. 

Diagnosis: A medium-sized species having a yellow lateral stripe extend- 
ing the length of the body, more or less broken on collars, and a yellow 
cream spot at jaw angle dividing and sending a branch onto lower jaw; 
transverse folds, 269-271, five confined to tail; tentacle twice as close to eye 
as to nostril; body width in total length about 26 times; splenial teeth 10-9; 
dorsal transverse folds curving forward on median line, somewhat angular 
anteriorly, strongly angular on venter except on posterior fifth of body. 

Description of type: Head somewhat bluntly conical; eyes small but 
distinct, in a socket; tentacle to eye, 1.5 mm., to nostril, 2.9 mm.; distance be- 
tween eyes (straight line), 5.3 mm.; level of eyes to tip of snout, 4.6 mm.; 
nostrils plainly visible from above; a slight median elevation on dorsum 
extending along body for about half its length; a dorsolateral ridge on sides 
of body extending to tail, making body somewhat quadrangular in cross- 
section; tentacle very close to edge of lip; eyes lateral. 

First nuchal groove passes completely around head; second groove dis- 
tinct across throat reaching up on sides to level of mouth-angle; third groove 
indistinct; the two collars fused dorsally, the second with two transverse 
folds on its posterior part, neither of which cross venter. 

Folds following collars 269-272 (variable at different body levels), the 
dorsal folds running somewhat forward and forming a distinct angle or 
sharp curve pointing forward becoming somewhat more obtuse towards 
middle of body; ventrally folds form an acute angle pointing backwards 
except on posterior fifth of body in which the folds pass nearly straight 
across dorsally and ventrally. A few folds following collars are incomplete 
ventrally. 

Scales appear on the second collar; in the tenth fold there are at least 


277 


New Asiatic AND AFRICAN CAECILIANS 


jemav) 


“Pp 


2. 


IC 


le 


pheoqns 


‘d :proay 


JO SAAOTA 9914 


‘ 


@) 


‘ 


d 


Vv 


uo] 


A396) 


Le 


‘TItb CON wmasny, jase 


old 


d 


od 


i 


TL, 


“AOU 


(‘uw ¢'g ‘proy Jo YIpIAr 
‘ds szuwynsuvpnasd stydosyiy2] “[] ‘old 


278 Tue UNtversity ScrENCE BULLETIN 


FIG. Iz 


Fic. 12. Ichthyophis pseudangularis sp. nov. Type. Basel Museum No. 4412, Ceylon. A, view 
of lower jaw and tongue. 


two incomplete scalerows. In latter half of body there are six more or less 
complete rows of scales in each fold. 

Dentition: Maxillary-premaxillary teeth, 18-19; vomeropalatine, 22-22; 
dentary, 18-18; splenial, 10-9. The anterolateral dentaries are the largest teeth. 
Tongue somewhat swollen laterally, the tip not covering the splenials. 

Color: Above slate to lavender slate; somewhat brownish slate on venter; 
a narrow yellow to cream lateral stripe more or less interrupted on collars, 
terminating posteriorly behind level of vent; posterior to angle of jaw a 
yellowish cream spot which bifurcates sending a branch along lower jaw, the 
yellow not reaching eye along upper jaw. Some indistinct cream marks on 
chin, and one asymmetrical spot above second collar: a small cream spot at 
vent, one at tentacle, and one at nostril. Tip of snout light cream. 

Measurements in mm.: Total length, 225; tail, 3.5; width of body, 8.6; 
width of head at first collar, 8.5; width in length, 26 times. 

Remarks: The name pseudangularis is from Greek, pseudo, false and 
Latin, angulus, angle. 


[chthyophis asplenius sp. nov. 
Figs. 14,15. 


Holotype: Rij\csmuseum van Natuurlijke Historie, Leiden, No. 6912 B. 


From, Boven Mahakkam, Borneo. 


Paratype: R.N.A. Leiden, No. 6912 A. Topotype. Museo Civico Genova 


No. 32195, Sarawak, Borneo. 


New Asiatic AND AFRICAN CAECILIANS 279 


"FIG. 13 


Fic. 13. Ichthyophis pseudangularis sp. nov. Type. Basel Museum No. 4412, Ceylon. Pemial 
organ seen from ventral view, lying in cloacal region. (Much enlarged.) 


Diagnosis: A moderately broad lateral yellow stripe broken or not on 
collar. Eye in socket, the tentacle much closer to eye than to nostril; width 


280 Tue University Scrence BULLETIN 


Fic. 14. Ichthyophis asplenius sp. nov. Type. Leiden Museum No. 6912 D. Boven Mahakkam 


(river), Borneo. A,B,C, three views of head; D, subcaudal area. (Actual width of | 
head, 6.5 mm.) 


New ASIATIC AND AFRICAN CAECILIANS 281 


in length, 23-26 times; total folds, 247-270; no splenial teeth in transformed 
specimens; grooves and folds do not cross the back except in the posterior 
half of body, some folds tending to form angles directed forward; through- 
out most of body, grooves do not cross venter but the folds are complete, 
forming a ventral angle pointing backwards except on latter fourth of body 
where they, with the grooves, pass directly across. 

Description of the type: Head moderate. Eye visible, somewhat elevated 
but dim; distance between eyes (4.4 mm.) considerably greater than distance 
from level of eyes to tip of snout; snout projecting about one millimeter 
beyond mouth; distance of tentacle from eye (1.2 mm.) much less than dis- 
tance to nostril (2.15 mm.); eye to nostril, 3 mm.; tentacle to edge of lip, 0.2 
mm.; nostrils nearly terminal but visible from directly above head. 

First and second collars fused above, distinct below; first nuchal groove 
complete below, broken in middorsal region of head; second groove pre- 
ceded below by a slight fold, visible laterally when seen from above; third 
groove vaguely marked above and on sides; second collar fused below with 
the following body fold. A fold, similar to a body fold is present dorsally on 
back of first collar. Primary and secondary folds following collars are 264- 
270, the folds and grooves incomplete across back on anterior half of the body 
being separated on median line anteriorly by a distance of three to three and 
a half millimeters, the distance between growing gradually shorter until 
folds become complete near the middle of body. The grooves are complete 
only on latter fourth of body; ventrally folds form a backward-pointing angle 
except on latter fifth of body where the grooves run straight across venter; 
five folds on tail, three of which are behind vent; vent longitudinal, short; 
glandular swellings at side of vent not evident. 


Fic. 15. Ichthyophis asplenius sp. nov. Type. Leiden Museum No. 6912 D, Boven Mahakkam 
(Upper Mahakkam river), Borneo. A, upper jaw and palatal region of the mouth; 
B, lower jaw and tongue. (Actual width of head, 6.5 mm.) 


282 Tue University SciENcCE BULLETIN 


Very small scales present in the 30th fold and there may be a few scattered 
scales preceding this; at middle of body about three rows where the scales 
can be traced to the mid-ventral line; posteriorly four rows present with 
occasional scales that may tend to form a fifth row. 

Dentition: (of type and paratype respectively): maxillary-premaxillary 
teeth, 26-26, 27-28; vomeropalatine, 26-27, 29-29; dentary, 25-25, 25-25; splenial, 
0-0, 0-0. The teeth all small, the dentary teeth perhaps the largest; the vom- 
eropalatine teeth minute. 

Color: Generally rather dark brownish lilac, very nearly the same shade 
above and below; a yellowish lateral stripe, with rather uneven edges, about 
1.5 mm. wide beginning on head at tentacle and extending to level of vent. 
Head slightly olive-brown; only a suggestion of a light spot at vent; a very 
faint ring about eye; tip of snout light. 

Measurements in mm. (of type and topotypic paratype): Total length, 
207, 202; tail length, 3, 3.2; width of body, 8.8, 8.2; width of head, 6.5, 6.6; 
between eyes, 4.8, 4.4; eye level to tip of snout, 3.5, 3.5; eye to tentacle, 1.2, 
1.25; tentacle to nostril, 2.2, 2.5. 

Variation: The Sarawak paratype in Genova measures 191 mm., the tail, 
3.3 mm.; it lacks splenial teeth and has the typical incomplete folds. Near 
the middle of the body where the folds are complete the folds tend to form a 
slight median angle directed forward; the incomplete folds tend to curve 
forward somewhat. The ventral count of the folds is 247, the dorsolateral 
count 254. The width of the body (7.2) in the total length is a little more 
than 26 times. 

The character of incomplete folds in the anterior part of the body occurs 
in a Sumatran form herein described. This latter form however differs in 
having a slightly higher average number of folds. The body width in length 
is nearly 40 times instead of 23-26 times, and there is present a series of 14-14, 
16-16 splenial teeth. The head is proportionally larger and wider. 

When the folds first meet on the back near the middle of the body they 
may tend to form an obtuse angle or a median curve. 

Remarks: Number 5 (Basel Natural History Museum), and number 
9090 (Vienna Museum) are two problematical specimens both purporting 
to be from Ceylon that seem to show a relationship with Ichthyophis as- 
plenius. They agree in two significant characters: the incomplete folds above 
and below on a considerable part of the body, and the complete absence of 
splenial teeth. There are certain small differences and a rather considerable 
difference in the other three tooth-series. Thus the maxillary-premaxillary 
teeth are 26-26 to 28-28; vomeropalatine, 26-27 to 29-29; the dentary, 25-25. 
If the localities of these two specimens are correct it shows two forms occu- 
pying ranges in the easternmost and westernmost points in the range of the 


genus, and lacking, as far as known, in any closely related forms in the 
area that separates them. 


New AsIATIC AND AFRICAN CAECILIANS 283 


Other data on these two specimens No. 5 and No. 9090 respectively 
are: Total length, 2202 and 203; tail length, 3.7, 3; head width, 7, 6.2; 
body width, 8.8, 7; width in length, 25, 29 times; tip of snout to first nuchal 
groove, 8, 7.8; to second groove, 11.8, 10; to third groove, 15, 13; eye to eye, 
5, 4.3; level of eyes to snout tip, 3.85, 3.45; tentacle to eye, 1.1, 1.1; tentacle to 
nostril, 2.1, 2.2; total folds ventral and dorsolateral counts, 277-288, 275-281; 
tail folds, 4, 5-6. There are four or five rows of scales in the posterior folds. 

It would not be impossible that each of these specimens bears an incorrect 
locality label. However I have no evidence that this is true. Rediscovery of 
this form in Ceylon is necessary before one can deal with these specimens 
with certainty. 

A series of five Malayan specimens from the National Museum of Sin- 
gapore (formerly Raffles Museum) are referred to this species. On the Loan 
Invoice these are listed: No. 2 (R 8928, 7.12.42, Forest Research Institute, 
Kepong; No. 3 (R 9246. 30. 12. 49.) Forest Research Institute, Bukit Lapong; 
No. 4 (R 8081. 269. 49.) Forest Research Institute, Kepong; No. 5, Forest 
Research Institute, Bukit Lapong, Selangor: No. 10 (28.5. 31.) Tg. Rambutan, 
Perak, Malaya. 

These specimens are from 209 to 242 mm. in length, the width in length 
27 to 30 times. There is but little difference in the dorsal and ventral counts 
of the folds (primaries and secondaries together which are not distinguish- 
able. The counts vary from 251 to 279. There is no trace of splenial teeth. 
All are of the same brown color with a rather broad yellow lateral stripe 
from head to tail. The grooves are present laterally but grooves and folds 
are broken dorsally on much of the anterior third of the body. I do not 
consider these as paratypes. 


[chthyophis forcarti sp. nov. 
Fics: 16, 17. 


Holotype: Musee d'Histoire Naturelle Bale (Suisse) No. 4411, Ceylon, 
Paul and Fritz Sarasin, collectors. 

Diagnosis: Transverse folds 346-348, forming a sharp median curve 
(rarely angulate), pointing forward on anterior half of body, posteriorly 
crossing body nearly in a straight line; ventrally folds form an angle pointing 
backwards except in posterior part where they pass directly across; splenial 
teeth, 5-5. Folds and grooves complete above and below; eye distinct, the 
tentacle twice as close to eye as to nostril; body width in length 27 times; a 
narrow irregular-edged yellow lateral line from the second collar to close to 
tip of tail. 

Description of type: Head not wider than neck and body, the eyes distinct 
in sockets, the distance between them (5.5 mm.) greater than distance from 
eye level to snout tip (4 mm.); tentacle much closer to eye (1.5 mm.) than to 


Tue University ScreNce BULLETIN 


284 


jO YPM [enjoy “ore [epneogns ‘q 


(‘uu ¢°g ‘pray 
speey Jo sata sary) ‘OD ‘g ‘WY ‘UO[AID “TI bh ‘ON winasny jaseg ‘adXz ‘aou -ds yawr4of stydorysy2] “9[ “Ol 


9! Sid 


New Asiatic AND AFRICAN CAECILIANS 285 


nostril (2.65 mm.); eye to nostril, 3.5 mm.; tentacle to lip, 0.2 mm.; first and 
second collars fused above, distinct below, the nuchal grooves shallow, 
bordered by a slightly lighter line; second collar with one transverse fold 
above on posterior part, below mesially fused to first transverse fold posterior 
to second collar; 346-348 transverse folds complete above and below except 
anterior three or four on median ventral surface; nine folds on tail, the two 
preceding vent forming an angle directed backward into vent; four or five 
complete folds behind vent. Five or six anterior folds form angles dorsally. 

Scales begin on back of second collar (the surface of body is slightly 
dehydrated and scales visible throughout) ; at middle of body maximum scale- 
rows in each fold, four or five; at posterior part, five rows, only a part of 
which are complete on sides and ventrally. 

Dentition: Maxillary-premaxillary teeth, 25-26; vomeropalatine, 24-24; 
dentary, 18-19; splenial 5-(?). The teeth are small without much contrast in 
size except splenials, which are much smaller and arise at a much lower point 
than dentaries; splenials scarcely reach level of the bases of dentary teeth. 
Choanae lateral, directed forward, the transverse diameter of one contained 
in the distance between them somewhat more than three times; tongue 
plump, rounded anteriorly, covering splenials which scarcely reach surface 
of gums. 

Color: Lilac-brown above and below, the head slightly olive-brown, some- 
what dimly mottled; a lateral yellow stripe beginning behind second collar 
and terminating behind vent; light yellow mark at mouth-angle; a small 
circular light mark above eye; a small light spot at tentacle, one below 
nostril and one surrounding vent, the latter narrowly separated from the 


A B 
FIG. I7 


Fic. 17. Ichthyophis forcarti sp. nov. Type. Basal Museum No. 4411, Ceylon. A, upper jaws 
and palatal region of mouth; B, !ower jaws and tongue. (Actual head width, 8.5 mm.) 


286 Tue University ScriENCE BULLETIN 


lateral stripes. In the present slightly dehydrated state each fold shows a 
narrow light line caused by the scales and glands showing through the skin, 

Measurements in mm.: Total length, 236; tail, 4; width of body, 8.6; 
width of head, 8.5; snout tip to first groove, 10.3; to second groove, 13.5; to 
third groove, 16.2; width in length, 27.4 times. 

Remarks: The exact locality at which this specimen was taken is seem- 
ingly no longer known. 

The species is named for Dr. Lothar Forcart of the Natural History 
Museum of Basel. 


Ichthyophis webert Taylor 
Fics. 18, 19. 


Ichthyophis weberi* Taylor, Philippine Journ. Sci., vol. 16, no. 3, March, 1920, pp. 
227-228 (type-locality, Malatgan River, Palawan, Philippine Islands, C. M. 
Weber, collector); Dept. Agri. Nat. Resour., Bureau of Science, Manila, publ. 
15, Dec. 15, 1921, pp. 26-27 (reprinting of the type-description); Univ. Kansas 
Sci. Bull., vol. 40, Apr. 20, 1960, pp. 43-44 (removed from synonymy). 

Ichthyophis monochrous: Van Kampen (part.), The Amphibia of the Indo- 
Australian Archipelago, Leyden, 1923, pp. 3-4, 282; Inger, (part.), Fieldiana: 
Zoology, vol. 33, no. 4, July 23, 1954, pp. 207, 209. 

Diagnosis: A small species reaching a length of 256 mm.; above uniform 
dark lilac to violet, the ventral surface a little lighter and showing a slightly 
brownish lilac shade; cream spot at vent; eye in a socket, visible through 
skin. Transverse folds (primaries, secondaris and tertiaries indistinguish- 
able from each other superficially), from 304-322, ventral count to 313-329, 
dorsal count; splenial teeth absent in adult, present in at least some larva. 
Body width in total length about 25 times. Vertebrae, 104-108. 

Description of the neotype: (Stanford University No. 21758): Head 
rather short (10 x 8.5 mm.), rather flattened; distance between eyes (in 
straight line), 5.85 mm.; length from level of eyes to tip of snout, 4.9 mm.; 
nostrils plainly visible from above, directed upward and slightly backward; 
tentacle, small rather conical, close to lip (0.25 mm.), its distance from eye 
(1.8 mm.) much less than its distance from nostril (3.2 mm.); snout project- 
ing about 1 mm. beyond mouth; eye from mouth, 1.4 mm., from nostril, 
3:9 mm: 

Two collars, first strongly defined on throat but first groove, while dis- 
tinct laterally, is incomplete above; second groove distinct laterally, scarcely 
visible from above; third groove limiting second collar is visible laterally but 
incomplete below and only dimly visible above; an indistinct groove sepa- 
rates a transverse fold on the back part of second collar. 

* The holotype of this species was destroyed in the Second World War during one of the final 


battles in Manila, which partially destroyed the Bureau of Science. I hereby designate Stanford 
University, No. 21758 from near Iwahig, Palawan (virtually a topotype) as a neotype. 


New AsIATIC AND AFRICAN CAECILIANS 287 


Fic. 18. Ichthyophis weberi Taylor. Neotype. Stanford University No. 21758, near Iwahig, 
Palawan, P.I. A,B,C, three views of head; D, terminal part of body and subcaudal 
region. (Actual width of head, 8.5 mm.) 


288 Tue University Sctence BuLLETIN 


The folds following second collar, 304 (ventral count), 313 (dorsolateral 
count); folds immediately following the second collar, incomplete below, 
but complete dorsally; on middle third of body folds incomplete or dim on 
middle of dorsum; on latter third both folds and grooves more or less distinct, 
definitely so on posterior part. All folds, except a number preceding vent, 
form a well-defined ventral angle; tail very short, six folds present, four or five 
interrupted below by vent; sides of vent with eight or nine denticulations; an 
anal gland on either side of the vent, slightly elevated. 

Scales begin immediately following the collars, one or two rows being 
present, the largest scales reaching a length of half a millimeter, the smaller 
ones 0.1 to 0.2 mm.; at middle of body three rows of larger scales present; on 
posterior fifth there are three or four rows (the fourth row not necessarily 
complete) and usually the scales not contiguous or overlapping, but generally 
large. 

Dentition: Maxillary-premaxillary teeth, 25-25, small, of nearly equal 
size; vomeropalatine, 27-27, slightly smaller; dentary, 22-22, the last six to 
eight of the series a little larger than all the otner teeth; no trace of splenials. 
Palate high, the choanae rather large, subtriangular directed outward and 
forward, the distance between them more than two and one-half times width 
of a choana. Tongue large, flat anteriorly, covering entire surface between 
the dentary series. 

Color: Above nearly uniform violet to lilac; on ventral surface lighter 
with a suggestion of brownish; a whitish mark on vent. Tip of snout 
vaguely lighter; no spot at eye or tentacle; lower jaw not or but vaguely 
lighter than chin. 


FIG. 19 


Fic. 19. Ichthyophis webert Taylor. Neotype. Stanford University Museum No. 21758, near 
Iwahig, Palawan, P.I. A, upper jaw and palatal region; B, lower jaw and_ tongue. 
(Width of head, 8.5 mm.) 


New AsIaTic AND AFRICAN CAECILIANS 289 


Measurements (see table) : 


Tase 3. Table of measurements and data on /. weber. 


21758 3 * 


Number Neotype 217602 217624* 21764 Larva Type 
slotaleslemgtin 222-5. aso 256 226 230 209 148.5 250 
“Teil These dn, eee ee 4 Soll 3.1 3.3 AD) DD) 
Widthiof "body. 222.) 2s 10 OF TS) 7.6 6.8 
Nyardthy (om sheady 2: 8.5 8 8.6 7.8 7 
Length of head to Ist groove — 10 9.6 10.2 OF 6 
SCM LOMCYC noe eee sere. cs 5.8 4.8 5.4 25) 4 oes 
ye level/to snout tip) .-..--...- ae) a 4.8 4.9 oes 5 
Mentaclemtomeye je 1.8 (335) 1:35 1.28 in eye spot i) 
slientacle toy nostril! 2-2. 32 2.75 3.1 2.8 
Snout to Ist groove. .............- 10 9.6 10.2 OF é 
Snout to 2nd groove ............ NS a7] 12.8 12.8 eS 3 
Snout to 3rd groove ............ 17 16.4 16.3 15-2 10 
Total folds: 
dorsolateralp ss. = 304 332 312 316 320 324 
WOT ee BY eae een 313 327 324 321 329 
“Trav OCI) Se oe ee ee 6 6 6 6 6 
Maxillary-premaxillary teeth _ = 25-25 25-25 25-26 24-25 15-14 
Womeropalatine 2.2.00... Dea 24-24 25-25 23-24 15-15 
IDGntanyy ten ee ws re la 22-22 18-18 21-22 21-21 13-13 oes 
Sole tiallipemnesee seers 720s 0-0 0-0 0-0 0-0 5-5 0-0 


WETteEDGAG: 0 eee Fete ee 107 108 105 105 105 


* Numerous small ventral papules present on folds, preceding vent. 


Variation: There are a few differences between the type and the neotype. 
The ventral groove mentioned in the type is probably due to preservation. 
Only the larva, No. 21759, shows a similar groove, obviously caused by 
preservation. The length of the tail of the type was said to be 25 mm. I 
suspect this measurement was made from the posterior rather than the 
anterior end of the vent. The color (above “yellowish brown”) was probably 
due to changes brought about by fixation. The individual glandules that 
appeared in the type as “minute rounded yellowish dots” were doubtless 
made apparent by a slight surface dehydration. In the specimen reported 
here, the rounded glandules are dimly visible under magnification and larger 
elongate glandules are dimly visible bordering the grooves. 

It would appear that the splenials are present in some larvae and that these 
are lost during transformation, as occurs in salamanders. One recently trans- 
formed specimen of the series, No. 21761, still shows two of the splenials still 
present. These I presume would soon have been lost. All others of the series 
of ten (except a larva) show a complete absence of splenials. The specimen 
with the highest number of teeth (No. 217662 measuring 258 mm. in 
length) is the largest of the series, the formula being, 28-28, 26-27, 28-28, 0-0. 


toy = 
The extruded tentacle is fattened, rather than conical. 


25() Tue University ScrENcE BULLETIN 


Remarks: The coloration of the larva is darker above than that of the 
two largest specimens listed but scarcely darker than the two medium-sized 
ones. These are slaty violet dorsally and somewhat darker above than the 
neotype; the ventral coloration however is very similar. All the specimens 
have the head bent down at an angle. Whether this is a result of fixation, I 
cannot say. 


Ichthyophis orthoplicatus sp. nov. 


Irie, A, ile 


Holotype: Zoological Survey of India, Calcutta, No. 17010, Pattipola, 
Central Province, Ceylon; F. H. Gravely, collector,* Aug. 2, 1915. 

Diagnosis: A short, relatively broad species, the width in length approxt- 
mately 19 times; unicolor lilac-slate, lacking any trace of a lateral stripe. Total 
folds on body and tail posterior to the collars, 291 to 295, seven confined to 
tail. Grooves and folds not angulate on venter; splenial teeth, 10-10; eye 
visible in a socket; tentacle a little more than twice as close to eye as to 
nostril; no diastema between squamosal and parietal bones. 

Description of the type: Head oval, the distance between eyes (straight 
line), 5.5 mm.; from eye-level to tip of snout, 4.7 mm.; distance of tentacle 
from eye, 1.25 mm.; from nostril, 2.9 mm.; tentacular opening a horseshoe- 
shaped groove, separated from the edge of lip by a distance of 0.55 mm.; 
nostrils visible from above; first collar vaguely indicated laterally but the 
first and second grooves cannot be discerned above or below (probably visible 
in well-preserved specimens); third groove limiting the second collar not or 
scarcely discernible (probably because of the dehydration of the specimen); 
evidence of lateral longitudinal swellings on each side of the combined 
collars; seemingly there is one or more folds (grooves) on the dorsal posterior 
part of second collar. Total number of folds behind collars varies from about 
291 to 295, of which six (or five) are confined to tail. Latter somewhat com- 
pressed laterally, flattened below, terminating in a point; anterior ventral 
folds incomplete for one fifth of body length; elsewhere folds and grooves 
above and below pass almost directly across dorsum and venter, no folds 
showing ventral angles. Vent longitudinal, the edges denticulate, interrupt- 
ing three or four folds. 

Dentition: Maxillary-premaxillary teeth, 23-22; prevomeropalatine, 20-20, 
smaller than preceding; dentary, 20-20, somewhat larger than the maxillary; 
splenial, 9 (or 10)-10, some equaling dentaries in size. Tongue rounded an- 
teriorly not covering splenial teeth; choanae rather large, the diameter of 
one in distance between them, about three times. 


*Ichthyophis monochrous ( part.) Deraniyagala, Ceylon J. Sci. (B), vol. 17, pt. 3 (May 19, 
1933). 


New Astatic AND AFRICAN CAECILIANS 291 


FIG. 20 


Fic. 20. Ichthyophis orthoplicatus sp. nov. Type. Zoological Survey of India, Calcutta No. 
17010, Pattipola, Ceylon. A,B,C, three views of head; D, terminal part of body and 
subcaudal area. (Actual width of head, 7.9 mm.) 


29? Tue University Scrence BULLETIN 


Color: Above and below lilac-slate (more brownish where dried), and 
seemingly not lighter on ventral surface; a very dim lighter ring about eye, 
and scarcely discernible lighter areas about tentacle and nostril; a light spot 
at vent, but no anal glands visible. Head somewhat olive above. 

Measurements in mm.: Present shrunken length about 222 (estimated 
true length 235); tail length, 4; head width, 7.9; approximate body width, 
12.5; width in length (approx.) 18 times. 

Remarks: The type specimen is strongly contracted, the vertebrae form- 
ing a strongly sinuous line. 

Rhinatrema, a South American genus of caecilians, which likewise has 
a species with a lateral stripe, differs from Ichthyophis in having all grooves 
and folds passing directly around the body. As far as I know, the present 
species is the only species of Ichthyophis known in which the folds (and 
grooves) fail to form a ventral angle throughout most of the body. How- 
ever, for a greater or lesser distance preceding the vent (one fourth or less of 
the length) the folds and grooves of all species of Ichthyophis pass straight 
across the venter. 

Deraniyagala, Ceylon J. Sci. (B), vol. 17, pt. 3, May 19, 1933, writes of the 
type specimen: 

“Ichthyophis monochrous. This is the only specimen in the Colombo 
museum and has 298 annul.” 

The few differences between Mr. Deraniyagala’s description and that 
given here are insignificant. His count of folds perhaps includes the two 
anterior collars. 

Kelaart (Prodromus Faunae Zeylanicae, 1852), mentions a mutilated 
Ichthyophis from Kandy “of a brown color above and a pale yellow brown 


FIG- 2l 


Fic. 21. Ichthyophis orthoplicatus sp. nov. Type. Zoological Survey of India, Calcutta, No. 
17010, Pattipola, Ceylon. A, upper jaw and palate of mouth; B, lower jaw and tongue. 
(Actual width of head, 7.9 mm.) 


2935 


New Asiatic AND AFRICAN CAECILIANS 


“TUT 


d 


peoy JO sMatA sari ‘DO ‘g SV 


"eag eARl 


REETSE 


went 


(uit g*¢ “peay JO YIpIM [enjoy) ‘Pore [epneoqns pue Apoqg jo qed [eu 


607$ “ON Whasnyw, Weplayisuy 


ee “Sls 


adk qt, ‘Aou ‘ds sisuauogyyiqg stydody1y2] *7Z “Old 


294 Tue University ScrENcE BULLETIN 


beneath, without the side streak of Ichthyophis glutinosus.” He also states 
that Dr. Templeton found a new species in the island. These specimens are 
probably no longer in existence. 


The species is named orthoplicatus, from Greek orthos=straight, plus 


Latin, plicatus, fold. 


Ichthyophis billitonensts sp. nov. 
Piesn225 25) 


Holotype: Zoodlogisch Museum, Amsterdam, No. 5209, Billiton Island, 
Indo-Australian Archipelago. 

Diagnosis: Probably a small species (type 135 mm.). Total transverse 
folds, 251-254, five on tail, angulate on venter except on last fifth of body and 
tail; dorsally sinuous, a few slightly angulate anteriorly, others curving 
forward somewhat; posteriorly folds pass directly around body; eye distinct; 
tentacle nearly twice as close to eye as to nostril; splenial teeth, 1-1, relatively 
large. Collars distinct ventrally, fused dorsally; second nuchal groove below 
preceded by a narrow curved white fold (groove); scales relatively large, 
about three rows in each posterior fold. 

Description of type: Head bluntly conical, the gape short; area above 
jaw-angles somewhat thickened; eye very distinct, probably functional; 
tentacle much closer to eye (1 mm.) than to nostril (2 mm.); eye to nostril, 
2.7 mm.; tentacle from edge of lip, 0.25 mm.; nostrils visible from above, 
directed upward. First collar distinct below, fused to second collar above; 
first and second transverse grooves very distinct below, the second tending 
to form a slight fold which is curved, and white in color; dorsally first nuchal 
groove is interrupted for three and one-half millimeters; second barely dis- 
cernible on sides when seen from above; third groove sharply marked 
dorsally, while ventrally it is indistinct; second collar fused posteriorly with 
the median ventral parts of two or three transverse folds; the collars, fused 
dorsally, have two longitudinal swellings separated mesially by a median 
longitudinal elevation. 

Following the collars, there are 251-254 transverse primary and secondary 
folds complete around the body, except first three or four anterior folds on 
ventral surface; five or six folds confined to tail, two or three of which are 
behind vent. Tongue somewhat oval, anterior part normally covering 
splenials which are slightly exposed above gums. 

Dentition: Maxillary-premaxillary teeth, 21-21; prevomeropalatine, 17-18; 
mandibular, 18-17; splenial, 1-1. Choanae large with somewhat elevated rims 
directed forward and laterally, the distance between them 1.4 mm., the trans- 
verse diameter of one (0.43 mm.), in this distance about three times; moder- 
ately large scales appear in first fold; at middle of body scales in two large 
and one small incomplete row; posteriorly there are three large rows and a 
few scattered smaller scales in each fold. 


New AsIATIC AND AFRICAN CAECILIANS 295 


Color: Generally brown; head rather olive-brown, the collars dark brown; 
body above brown, the ventral regions a lighter brown; a light spot at vent, 
tentacle and nostril. A minute lighter ring surrounds eye; an indistinct 
lighter area above jaw-angle. At the present time there is indication of a 
yellowish line around each fold, probably due to a slight dehydration, thus 
causing the edges of the scales to be visible (not impossibly visible in life). 

Measurements in mm.: Total length, 135; tail, 2.3; width of body, 6.7; 
width of head, 5.8; snout to first groove, laterally, 7; to second groove, 9.2; to 
third groove, 11; eye to eye, 3.3; eye level to tip of snout, 3; width in length, 
20 times. 

Remarks: Despite the small size of this specimen it gives evidence, in the 
distinctness of the folds and the development of the teeth and scales, of 
maturity. 

The absence of a lateral stripe suggests a relationship with the monochrous 
group but in many ways it resembles the g/atinosus forms. The retention of 
only two splenial teeth is unusual in this genus. 

The species is named for its place of origin, the island of Billiton situated 
in the northern part of the Java Sea, between Sumatra, Borneo, and Java. 
Sometimes this name is spelled Belitung. 


Schistometopum ephele* sp. nov. 
Fires. 24,25, 20. 


Type: Museo Civico di Storia Naturale, “G. Doria,” Genova, No. 8773; 
Agua Ize (400-700M.) Ihla Sao Thomé, Gulf of Guinea. 
Paratype: B.M. 1933.11.16. 1-4 ? Ihla Sao Thomé. 


A 8 
FIG. 23 


Fic. 23. Ichthyophis billitonensis sp. nov. Type. Amsterdam Museum No. 5209, Billiton Island, 


Java Sea. A, upper jaw and palate; B, lower jaw and tongue. (Actual width of head, 
5.8 mm.) 


* From Latin ephelis=freckled. 


Tue Untiversiry ScIENCE BULLETIN 


Fic. 24. Schistometopum ephele sp. nov. Type. Museo Civico Genova No. 8773, Agua Ize, Sao 
CML: I : ; 
Thomé, Gulf of Guinea. Photograph of type about natural size. 


New AsIATIC AND AFRICAN CAECILIANS 297 


Diagnosis: A Schistometopum with a proportionally smaller, more- 
pointed head than thomense. Color light yellowish-brown with dark lilac- 
brown flecks over dorsum and sides of body; fewer flecks on venter chin 
and throat; primary folds, 97-106; secondary, 40-52. 

Description of the type: Head relatively very small; eye small, distinct in 
socket; distance between eyes (3.1 mm.), equal to length of snout in front 
of eyes (3.1 mm.); tentacular aperture small, separated from eye by 0.6 
mm.; from nostril by 2.5 mm.; from mouth by 0.35 mm.; eye from nostril, 
3.2 mm.; nostrils barely visible from directly above, practically terminal. Two 
collars; first collar sharply defined by a deep groove fore and aft surrounding 
neck, curving very slightly forward both dorsally and ventrally; a short 
indefinite transverse groove visible above and below; a distinct fold across 
chin preceding first groove; second collar a little wider ventrally than dorsally, 
with a distinct transverse groove above as long as width of body, not present 
ventrally; third groove, limiting collar, complete. Primary folds 97, all 
complete above and below, the grooves likewise distinct; secondaries, 40; 
total folds, 137; at first the grooves marking the secondaries are short; eight 
are complete. 

Scales present, one or two appearing low on sides at about the 33rd fold; a 
single row complete around body where secondaries begin. In posterior por- 
tion where secondaries are complete, three scalerows present. 

Dentition: Maxillary-premaxillary teeth 16-16; vomeropalatine, 18-18; 
dentary, 11-11; splenial, 10-9; dentary teeth longest, all directed backwards 
somewhat; splenials elevated to nearly same level as dentaries. Tongue not 
covering splenials, and lacking lateral beadlike elevations and grooves. 

Color: Light yellowish brown with lilac-brown flecks and marks. 

Measurements in mm. (type and paratype): Total length, 186, 340; width 
of head at first groove, 6, 9.4; width of body (middle), 9.2, 15.5; tip of snout 
to first nuchal groove, 7.2, 11; to second groove, 9.2, 13; to third groove, 12, 
17; width near terminus of body, 6.9, 13. 

Variation: The dental formula of the paratype is approximately 16-15; 
18-18; 9-9; 8-9. There are 100 primary folds and 36 secondary; there is a 
slight lateral dorsal swelling (muscular) on each side of the head preceding 
the first groove. This is discernible also in the type. The snout projects very 
slightly beyond mouth. The vent is transverse, denticulated, but no anal 
glands are in evidence. 

Distribution: Known only on Thla Sao Thomé. 

Remarks: This type was collected by the Leonardo Fea Expedition. In 
the British Museum one specimen is designated a paratype. This is a large 
specimen but considerably desiccated. A second specimen bearing the same 
number may likewise belong here (348 mm. long and dissected partially). 
Two other specimens under this number are completely yellow and are re- 
ferred to S. thomense. 


Tue UNiversiry SciENCE BULLETIN 


298 


JeQuaa ‘Apoq jo yied [euTW2) ‘q :ptoy Jo sMatA 9214) ‘OD ‘g ‘VW 


(wu g ‘pray Jo YIpIM jenoy) “Mora 


"PaUIND JO J[NH ‘puryjs] gwoyy, ovg ‘ad4y, ‘Aou ‘ds 


gz 8d 


ajayda windojawmoyjsiyIg “CZ 


2) (| 


New AsIaTIC AND AFRICAN CAECILIANS 299 


B.M. Museum, No. 1927.2.10.1 from the same locality may also belong 
with this species. 


Schistometopum brevirostris Peters 
Fics. 27, 28. 


Siphonops brevirostris Peters, Monatsb. Akad. Wiss., Berlin, 1874, pp. 617-618, pl. 
1, fig. 2 (type-locality, Westkuste, Africa. Later reported by Peters as “aus 
Guinea”; Gorham, 1962, gives Rolas Is., Gulf of Guinea, but I have overlooked 
the source of this information). 

Dermophis brevirostris, Peters, Monatsb. Akad. Wiss. Berlin, 1879, p. 937 (aus 
Guinea); :bid., 1880, p. 223 (Dermophis brevirostris gleich Siphonops thomen- 
sis Bocage). 

Diagnosis: Head elongated somewhat; diameter of body in total length 

35 times. Color bluish gray, the grooves lighter; splenial teeth, 4-5. 
Description of species: (from Museo Civico di Storia Naturale “G. Doria,” 

Genova, Italy, No. 28881, from Sao Thomé): Body moderately slender and 

seemingly flattened somewhat throughout (rather than cylindrical); head 

slender, oval; the eyes distinct, minute, the distance between them equal 

to the length of snout from level of eye; snout extends beyond mouth, 0.6 

mm.; tentacular aperture about one third diameter of eye, its distance from 

eye, 0.78, from nostril, 3.35; from lip, 0.5 millimeters. Two distinct collars; 

the first nuchal groove curves forward dorsally passing under chin in ad- 
vance of the dorsal portion, the width of the first collar greater ventrally; 
first collar with a short transverse medial groove above and below; second 
collar nearly same width above and below, also with a dorsal transverse 


A B 
FIG. 26 


Fic. 26. Schistometopum ephele sp. nov. Type. Museo Civico Genova No. 8773, Sao Thomé 
Island, Gulf of Guinea. A, upper jaw and palatal region: B, lower jaw and tongue. 
(Actual width of head, 6 mm.) 


Tue University Science BuLLEtTIN 


300 


(uu 9 “pesy Jo YIpim yenyoy) “AIA [eQUAA “Apoq jo snuruia, ‘cq 


spray JO SMITA JaIyI ‘O ‘gq “yY ‘eoUIND JO FIND ‘purjsy guioyy, ovs ‘1 g9ggz ‘ON PAOUSD ODIAID Oasny *(sI9}0q) SLUSOMNIAG UNdOJIUWOISIYIS *[7 “IT 


[Ue 


New AsIATIC AND AFRICAN CAECILIANS 301 


median groove but none on ventral part; the third nuchal groove incomplete 
below; second collar fused to the following first primary fold ventrally. 

Total primary folds 101, complete above and below; secondaries, 38, the 
last nine complete; a pair of ventral longitudinal grooves beginning just in 
front of third transverse nuchal groove, cross the collars run forward on 
chin but fail to meet anteriorly; a small transverse ventral groove across back 
part of chin preceding the first nuchal groove. 

Scales begin about the 20th fold, where one or two very small transversely 
widened scales are present; at middle of body a short lateral row present; 
at beginning of the secondaries a nearly complete row around body, the 
largest scales nearly one mm. in length; in each fold posteriorly there are 
four somewhat irregular rows, the largest scales a little more than 1.75 mm. 
in width; some are subcircular in shape. Skin glands not conspicuous, but 
in primary grooves a row of glandules usually evident, and at least posteriorly 
some very minute pores are evident under a lens. Terminus of body tapering, 
the vent transverse, its two sides denticulate, with a pair of small anal glands 
preceding it; from vent to terminus, 1.6 mm. 

Dentition: Maxillary-premaxillary teeth, 15-16; prevomeropalatine, 20-20; 
dentary, 11-14; splenial, 9-9. 

Color: In preservative, body light gray, becoming brownish anteriorly 
when removed some minutes from liquid. Head above and on sides nearly 
uniform cream to fawn, contrasting strongly with color on first collar; lower 
jaws light, the area between them gray brown; posterior part of body per- 
haps a little lighter than anterior. 

Measurements in mm.: Total length, 236; width of body, 9.5; width of 


A B 
FIG. 28 


Fic. 28. Schistometopum brevirostris (Peters). Museo Civico Genova No. 28881, Sao Thomé 
Island, Gulf of Guinea. A, upper jaw and palate; B, lower jaw and tongue. (Actual 
width of head, 6 mm.) 


302 Tue University ScreNcE BULLETIN 


head at first nuchal groove, 6; tip of snout to Ist groove, 9; to second groove, 
11.2; to third groove, 14.9; eye to eye, 3.85; eye-level to tip of snout, 4.1; body 
width in length, 24.8 times. 

Variation: A specimen in the British Museum (No. 1927.2.10.1 from 
Sio Thomé) is somewhat darker gray than the specimen described and the 
grooves are somewhat lighter. The width in length (9.8 and 261 mm.) con- 
tained 26.6 times. There are 106 primary folds and 52 secondaries; the dental 
formula: m-p., 16-17; v.p., 19-18; d., 11-11; sp., 10-11. 

Remarks: 1 would interpret Peters’ counts, in my own terms of reference 
as: four folds for the two collars; followed by 132 primaries and secondaries 
of which the last 14 are complete; when the secondaries begin these folds 
are alternately incomplete and complete. Thus there are 96 primaries follow- 
ing the collars, and 36 secondaries. 

I have referred these specimens to Peters’ S. brevirostris, although I have 
not compared them with the type. Peters did not distinguish between pri- 
maries and secondaries, or between these and the two collars. The type, if 
still extant, was not available for examination in 1962. 


THE UNIVERSITY OF KANSAS 


SCIENCE ean) 


Marine Biolog ogical -al Lal abor. atory 
ji BRARY 


WOODS HOLE, MASS 


Nii 


CATALOGUE OF THE TYPES IN THE SNOW 
ENTOMOLOGICAL MUSEUM 
Part IV (Orthoptera) 


By 
George W. Byers and Lanny B. Carney 


CATALOGUE OF THE TYPES IN THE SNOW 
ENTOMOLOGICAL MUSEUM. 
Part V (Acarina) 


By 
George W. Byers and Calvin L. Wong 


Vor. XLVI Paces 303-315 Ocroser 28, 1965 No. 7,8 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Vor. XLVI Paces 303-307 Ocroser 28, 1965 No. 7 


Catalogue of the Types in the Snow Entomological Museum 
Part IV (Orthoptera) ' 


By 


GEorcE W. Byers AND LANNY B. CARNEY 


ABSTRACT 


Types of 69 species and subspecies of Orthoptera are catalogued, with refer- 
ences to original description (except for species represented by paratypes only), 
number and condition of specimens, and type numbers as recorded in the cata- 
logue of types in the Snow Entomological Museum. 


INTRODUCTION 


Although there has never been a staff member of the Snow Entomological 
Museum working primarily on the taxonomy of the Orthoptera, there has 
long been an interest in building a good representative collection of these 
insects. A few papers dealing with the Orthoptera have been published by 
Museum personnel, notably those on the Oedipodinae of Kansas, by R. H. 
Beamer, and the Melanopli of Kansas, by P. W. Claassen, which papers ap- 
peared in 1917, in volume 11 of the University of Kansas Science Bulletin. 

A number of specialists on the Orthoptera have, through the years, made 
use of the collection in the Snow Museum, returning to the Museum certain 
type specimens representing new species revealed by their studies. Outstand- 
ing among these specialists are James A. G. Rehn and Morgan Hebard, both 
of the Academy of Natural Sciences of Philadelphia. By similar means, or 
by exchange, the Museum has come into possession of types (principally syn- 
types or paratypes from extensive series) of species described by S. H. Scud- 
der, T. H. Hubbell, and others. 

In the following list of type specimens, we have arranged the families 
and suborders according to the system proposed by Rehn and Grant, in their 


1. Contribution number 1239 from the Department of Entomology, The University of Kan- 
sas, Lawrence, Kansas. 


5 


304 Tue University ScIENCE BULLETIN 


Monograph of the Orthoptera of North America (Volume 1, 1961, Monogr. | 
Acad. Nat. Sci. Philadelphia, 12). Beneath the family headings, the types are | 
arranged alphabetically by genus, then species, in two separate lists. The | 
first includes holotypes, allotypes, and syntypes (cotypes), as well as para- 
types if these are present in addition, and gives a reference to the original 
description of each species. The second list includes species represented by 
paratypes only and omits literature references. Both lists indicate the sex 
of the types and the condition, if the specimens are seriously damaged. 
(Missing parts are indicated only when absent from both sides of the speci- 
men.) The type number from the Catalogue of Types in the Snow Ento- 
mological Museum is given for each species. 

We have not included any secondary types, such as homotypes, plesiotypes 
or metatypes. The lists also include only those type specimens that are ac- 
tually present in the Museum as of 1964. In some cases an original descrip- 
tion indicated one or more types were deposited in the Snow Museum, but | 
none—or a lesser number than that indicated—was actually found in the’ 
collection. Of the types listed by Hunter (1913, Univ. Kansas Sci. Bull., 
8[1]:16), only that of Lactista arphoides was not found. | 

Only original combinations of generic and specific names have been used, | 
as it was deemed impracticable to attempt to give the current generic assign- 
ment or synonymic status of every species. | 


ACKNOWLEDGMENTS | 
We wish to thank Dr. Harold J. Grant, Jr., and Dr. Theodore J. Cohn’ 


for assistance with various details in the preparation of the catalogue. Finan-_ 
cial support from the General Research Fund of the University of Kansas, 

: . . . . 
which has made possible the completion of this work, is gratefully acknowl- 
edged. 


Suborder Caelifera 
Superfamily Acridoidea 
Tetrigidae—Paraty pes 


Apotettix eurycephalus brevipennis Hancock. 14, 192 (antennae missing from 
both). 5424. 

Paratettix hesperus Morse. 1 6,12 ( @ lacks antennae). 1295. 

Tettigidea lateralis cazieri Rehn & Grant. 4 6,29. 5421. 


Acrididae 


Aeoloplides rotundipennis Wallace, 1955, Ann. Ent. Soc. Amer., vol. 48, no. 6, pp. 
464-466. ¢ holotype, 2 allotype, 32 6,12 2 paratypes. 5685. 

Cordillacris pima Rehn, 1907, Proc. Acad. Nat. Sci. Phila., vol. 59, pp. 69-71. 
2 holotype (both hind tibiae and tarsi lacking). 1315. 


CATALOGUE OF Types IN SNow MusEuM—Part IV 305 


Melanoplus blatchleyi Scudder, 1898, Proc. U.S. Nat. Mus., vol. 20, no. 1124, pp. 
322-323. 14,19 syntypes (antennae missing in both). 1302. 

Melanoplus gladstoni Scudder, 1897, Proc. Amer. Philos. Soc., vol. 36, no. 154, p. 
23. 14 syntype. 1303. 

Melanoplus huroni Blatchley, 1898, Psyche, vol. 8, no. 264, pp. 195-196. 12 syn- 
type (five legs missing, wings damaged). 1314. 

Melanoplus insignis Hubbell, 1932, Misc. Publ., Mus. Zool. Univ. Mich., no. 23, 
pp- 56-58. ¢ holotype, 2 allotype. 2231. 

Melanoplus packard:* 

Melanoplus snowii Scudder, 1898, Proc. U.S. Nat. Mus., vol. 20, no. 1124, pp. 
274-275. & holotype (no legs or antennae), 9 allotype (no front or middle 
legs, no antennae). 1301. 

Trimerotropis snowi Rehn, 1905, Trans. Kans. Acad. Sci., vol. 19, pp. 223-224. 
? holotype. 1298. 


Acrididae—Paratypes 


Aerochoreutes carlinianus strepitus Rehn. 1 4 (hind legs missing), 1 9. 5412. 
A groecotettix modestus crypsidomus Hebard. 14,12. 1310. 
Appalachia arcana Hubbell & Cantrall. 1 6,19. 4550. 
Aptenopedes aptera borealis Hebard. 1 2. 3052. 

Aptenopedes aptera saturiba Hebard. 16,192. 3053. 
Aptenopedes aptera simplex Hebard. 2 6,22. 3051. 
Aptenopedes hubbelli Hebard. 2 6,292. 3054. 

Aptenopedes nigropicta Hebard. 2 6. 3055. 

A ptenopedes robusta Hebard. 1 6,22. 3056. 

Aptenopedes sphenarioides appalachee Hebard. 2 6 ,2 2. 3057. 
Barytettix cochisei Gurney. 1 6 (antennae missing). 5415. 
Bradynotes chilcotinae Hebard. 14,19. 1309. 

Circotettix crotalum Rehn.1 6,192. 1300. 

Circotettix nigrafasciatus Beamer. 14 6,692. 1299. 

Conalcaea cantralli Gurney. 1 6 . 5409. 

Conalcaea coyoterae Hebard. 12. 927. 

Eumorsea balli Hebard. 1 6. 2267. 

Melanoplus beameri Hebard. 164,172. 1313. 

Melanoplus bruneri Scudder. 1 8 ,2 2 (antennae lacking in 6,12 ). 1305. 
Melanoplus elongatus Scudder. 1 Q (antennae missing). 928. 
Melanoplus indigens digitifer Hebard. 5 6,5 2. 3058. 
Melanoplus intermedius Scudder. 1 2. 1306. 

Melanoplus macclung: Rehn. 46 6,51 2. 5418. 

Melanoplus mastigiphallus Strohecker. 14,19. 3641. 
Melanoplus oreophilus Hebard. 1 6,2 2. 933. 

Melanoplus viridipes eurycercus Hebard. 16,19. 1312. 


2. Two male and three female specimens labelled as part of Scudder’s type series are some 
of those listed by Scudder, 1898, in Proc. U. S. Nat. Mus., vol. 20, but are not regarded by us 
as original types of this species, which was described in 1878 (Proc. Boston Soc. Nat. Hist., 
vol. 19, p. 288). 


306 Tue Universiry SciENCE BULLETIN 


Oedaleonotus fratercula Hebard. 1 6,19 (antennae missing). 2266. 
Phrynotettix robustus manicola Rehn & Grant. 1 6. 5423. 
Phrynotettix robustus occultus Rehn & Grant. 1 6 . 5892. 

Psoloessa thamnogaea Rehn. 1 6 . 3807. 

Psychomastax psylla robusta Hebard. 1 6 . 5417. 

Schistocerca ceratiola Hubbell & Walker. 3 6,29. 4551. 
Stenobothrus olivaceus Morse. | 6 . 1804. 

Zapata salutator Rehn. 1 Q (antennae missing). 5426. 


Suborder Ensifera 
Tettigoniidae 


Plagiostira gracila Rehn, 1905, Trans. Kans. Acad. Sct., vol. 19, p. 227. 9 holo- 
types alle | 


Tettigoniidae—Paratypes 


Arethaea ambulator Hebard. 146,19. 2274. 

Arethaea coyotero Hebard. 16,19. 2272. 

Arethaea mescalero Hebard. 2 6 (one without legs). 2273. 
Brachyinsara hemiptera Hebard. 14,192. 3049. 

Decticita balli Hebard. 1 g . 3050. 

Dichopetala gladiator Rehn & Hebard. 1 6,12. 1307. 
Idionotus tehachapi Hebard. 1 6,12. 2271. 

Insara tessellata Hebard. 3 6 (all lacking antennae). 2270. 
Inscudderia walkeri Hebard. 16,12. 1308. 


Gryllacrididae 


Ceuthophilus paucispinosa Rehn, 1905, Trans. Kans. Acad. Sci., vol. 19, pp. 
227-228. 2 holotype. 1291. 

Ceuthophilus tuckeri Rehn, 1907, Ent. News, vol. 18, no. 10, pp. 445-446. 6 holo- 
type (both hind legs missing). 1292. 

Phrixocnemis franciscanus Rehn, 1905, Trans. Kans. Acad. Sci., vol. 19, pp. 228- 
229° 1651 2 syntypes. 1293. 

Phrixocnemis socorrensis Rehn, 1905, Trans. Kans. Acad. Sci., vol. 19, pp. 229- 
230. 13,12 syntypes ( 2 missing both antennae). 5431. 

Udeopsylla serrata Rehn, 1905, Trans. Kans. Acad. Sci., vol. 19, pp. 230-231. 
holotype. 1294. 

Gryllacrididae—Paratypes 


Ammobaenetes lariversi Strohecker. 1 6,192 (@ missing both antennae). 3642. 


Ceuthophilus wichitaensis Hubbell. 146,19. 5413. 
Stenopelmatus intermedius Davis & Smith. 1 6,19. 1290. 


Gryllidae—Paratypes 


Cycloptilum bidens Hebard. 3 6,32. 1803. 
Cycloptilum spectabile Strohecker. 1 6,19. 3640. 
Ocecanthus californicus pictipennis Hebard. 1 6 (antennae damaged). 5560. 


CATALOGUE OF Types IN SNow MusEuM—Part IV 307 


Gryllotalpidae—Paraty pes 
Scaptericus acletus Rehn & Hebard. 1 4,19. 3808. 


Suborder Dictyoptera 
Superfamily Mantodea 
Manteidae 
Stagmomantis gracilipes Rehn, 1907, Proc. Acad. Nat. Sci. Phila., vol. 59, pp. 67- 
68. 4 holotype (head and abdomen damaged). 1289. 
Superfamily Blattodea 
Blattidae 
Latblattella lucifrons Hebard, 1917, Mem. Amer. Ent. Soc., no. 2, pp. 43-46. 
$ holotype, ? allotype. 1287. 
Blattidae—Paratypes 
Pseudomops septentrionalis Hebard. 4 8,19. 1288. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Vor. XLVI Paces 309-315 OcroBeErR 28, 1965 No. 8 


Catalogue of the Types in the Snow Entomological Museum 
Part V (Acarina)’ 
By 


GeorcE W. Byers AND CaLvin L. Wonc 


ABSTRACT 


Types of 113 species of Acarina are catalogued, with references to original 
description (except for species represented by paratypes only), number and sex 
or stage of specimens, and type numbers as recorded in the catalogue of types in 
the Snow Entomological Museum. 


INTRODUCTION 


Most of the species of mites (Acarina) represented in the Snow Ento- 
mological Museum by type specimens were named and described during 
recent years by staff members and students or former students of the De- 
partment of Entomology of The University of Kansas. These acarologists 
mtic-W. 1. Atyeo, R. E. Beer, J. Baczek, J. G. Borland, J. M. Brennan, E. A. 
mioss, D: A. Crossley, B. Greenberg, C. C. Hall, E. W. Jameson, Jr., E. H. 
Kardos, D. S. Lang, L. J. Lipovsky, R. B. Loomis, D. S. Narayan, and K. A. 
Wolfenbarger. The few types representing other species were obtained by 
exchange or were received as gifts. 

In the following list of type specimens, the families are arranged alpha- 
betically. Under the family headings, the genera and then the species are 
likewise arranged alphabetically, in two lists. The first list includes species 
represented by holotypes, allotypes, syntypes or lectotypes and indicates para- 
types if these are present in addition. This list contains references to the 
original description of each species and to the designation of a lectotype, 
where applicable. The second list is of species represented by paratypes only 


1. Contribution No. 1228 from the Department of Entomology, The University of Kansas, 
Lawrence, Kansas. 


310 Tue University SCIENCE BULLETIN 


and omits literature references. Both lists indicate the sex or stage of the 
types. The type number from the Catalogue of Types in the Snow Ento- 
mological Museum is given for each species. On the labels of the micro- 
scope slides carrying certain of the types, there appears another number, 
which refers to a catalogue of specimens of Acarina prepared and slide- 
mounted at The University of Kansas. 

No kinds of types other than those already noted are included in this 
catalogue. Only those type specimens actually present in the collection as of 
1964 are listed. 

The combinations of generic and specific names appearing here are those 
originally used by the authors indicated; a few of the species have sub- 
sequently been removed to other genera, but it seems best, in the interest 
of stability, to record only the original combinations. In one instance of pri- 
mary homonymy, however, the main entry has been made under the re- 
placement name, although the original combination is listed, as well. 


ACKNOWLEDGMENTS 


We wish to thank The University of Kansas for financial support, pro- 
vided from the General Research Fund, which has made possible the prepara- 
tion of this catalogue. Assistance relating to certain included types was given 
by Dr. Richard B. Loomis of Long Beach State College, California, whose 
help is gratefully acknowledged. We also acknowledge with thanks the as- 
sistance of Miss Linda Hardee and Miss Judy Jobson in cataloguing, copy- 


reading, and typing the manuscript. 


Cheyletidae 


Cheletophyes knowltoni Beer & Dailey, 1956, Univ. Kans. Sci. Bull., vol. 38, pt. 1, 
no. 5, pp. 409-417. @ holotype, ¢ allotype, 34 6, 389 paratypes. 6251. 


Diplogyniidae—Paraty pes 


Neolobogynium lateriseta Hicks. 1 @. 6321. 


Eriophyidae 


Aceria slykhuisi Hall, 1958, Jour. Kans. Ent. Soc., vol. 31, no. 3, pp. 233-235. 9 
holotype, ¢ allotype, 4 4,162 paratypes. 5759. 

Catharinus axonopi Boczek, 1960, Jour. Kans. Ent. Soc., vol. 33, no. 1, pp. 11, 12. 
? holotype, ¢ allotype, 3 @ paratypes. 6252. 

Latonotus wegoreki Boczek, 1960, Jour. Kans. Ent. Soc., vol. 33, no. 1, pp. 9-11. 
? holotype, ¢ allotype, 3 Q paratypes. 6253. 

Tegonotus guavae Boczek, 1960, Jour. Kans. Ent. Soc., v@l. 33, no. 1, pp. 11-14. 
? holotype, ¢ allotype,5 @ paratypes. 6254. 


CATALOGUE OF TYPES IN SNow MusEuM—Part V 311 


Laelaptidae 


Laelaps stegemani Hefley, 1935, Jour. Kans. Ent. Soc., vol. 8, no. 1, pp. 22-24. @ 
holotype. 3789. 


Labidostomidae 


Labidostomma barbae Greenberg, 1952, Jour. N.Y. Ent. Soc., vol. 60, no. 4, pp. 
198-200. $ holotype, 2 é paratypes. 5405. 

Labidostomma caloderum Greenberg, 1952, Jour. N.Y. Ent. Soc., vol. 60, no. 4, 
pp- 197-198. 4 holotype, 14,19 paratypes. 5404. 


Labidostomidae—Paraty pes 


Labidostoma circinus Atyeo & Crossley. 2 2. 6255. 
Labidostoma fictiluteum Atyeo & Crossley. 1 6,12. 6256. 
Labidostoma malleolus Atyeo & Crossley. 2 2 . 6257. 
Labidostoma multifarium Atyeo & Crossley. 2 2. 6258. 


Macrochelidae—Paraty pes 
Macrocheles rodriguezi Oliver & Krantz. 16,19. 6322. 


Myobiidae—Paraty pes 
Radfordia bachai Howell & Elzinga. 1 6,12 2. 5686. 


Penthaleidae 


Penthaleus crinitus Narayan, 1957, Jour. Kans. Ent. Soc., vol. 30, no. 3, pp. 111- 
113. @ holotype, 4 2 paratypes. 6259. 


Podapolipodidae 
Podapolipus komareki Storkan, 1927, Zool. Anzeiger, vol. 71, pp. 19-22. 154, 
17 2 , 54 larval syntypes. 6260. 
Pterygosomidae 


Pimeliaphilus rapax Beer, 1960, Jour. Parasitol., vol. 46, pp. 434-440. @ holotype, 
8 allotype, 3 6,19, 2 larval, 17 nymphal paratypes. 6261. 


Scutacaridae 
Nasutiscutacarus ampliatus Beer & Cross, 1960, Jour. Kans. Ent. Soc., vol. 33, no. 
2, pp. 54-57. @ holotype, 3 2 paratypes. 6262. 
Nasutiscutacarus anthrenae Beer & Cross, 1960, Jour. Kans. Ent. Soc., vol. 33, no. 


2, pp. 51-54. @ holotype, 7 2 paratypes. 6263. 


Tarsonemidae 


Hemitarsonemus peregrinus Beer, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 
16, pp. 1300-1308. ¢ holotype, @ allotype, 24,39 paratypes. 5428. 


Si? Tue University SCIENCE BULLETIN 


Rhynchotarsonemus niger Beer, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 16, 
pp. 1221-1229. 8 holotype, @ allotype, 4 6,29, 3 larval paratypes. 5416. 
Steneotarsonemus fulgens Beer, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 16, 

pp. 1281-1285. g holotype, ? allotype, 2 ¢,109 paratypes. 5425. 
Steneotarsonemus hyaleos Beer, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 
16, pp. 1256-1261. ¢ holotype, @ allotype, 2 6,99, 1 larval paratypes. 5420. 
Steneotarsonemus kieferi Beer, 1958, Jour. N.Y. Ent. Soc., vol. 66, nos. 3-4, pp. 
153-156. ¢ holotype, @ allotype, 2 ¢, 339 paratypes. 6264. 
Tarsonemella beameri Beer, 1958, Jour. Kans. Ent. Soc., vol. 31, no. 2, pp. 189-192. 
2 holotype, 3 2 paratypes. 6265. 
Tarsonemus dispar Beer, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 16, pp. 
1148-1155. 4 holotype, ? allotype, 14,99 paratypes. 5406. 
Tarsonemus pritchardi Beer, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 16, 
pp. 1202-1210. 8 holotype, 9 allotype, 346,69, 1 larval paratypes. 5414. 
Tarsonemus sulcatus Beer, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 16, pp. 
1155-1162. ¢ holotype, 2 allotype. 5407. 

Tarsonemus viridis Ewing, 1939, U.S. Dept. Agr. Tech. Bull. 653, pp. 35-37. 1 ¢ 
syntype. 5432. 

Tarsonemus waite: Banks, 1913 (1912), Proc. Ent. Soc. Wash., vol. 14, no. 2, pp. 
98-99. 2 2 syntypes. 5411. 

Xenotarsonemus denmark: Beer, 1960, Florida Ent., vol. 43, no. 1, pp. 23-27. 6 
holotype, 2 6 paratypes. 6266. 


Tarsonemidae—Paratypes 


Pseudotarsonemoides cryptocephalus Ewing. 3 9 . 5048. 
Tarsonemus occidentalis Ewing. 3 6. 5430. 


Tetranychidae 


Aplonobia dyschima Beer & Lang, 1958, Univ. Kans. Sci. Bull., vol. 38, pt. 2, no. 
15, pp. 1234-1235. @ holotype, 9 @ paratypes. 6267. 
Aplonobia verrucosa Beer & Lang, 1958, Univ. Kans. Sci. Bull., vol. 38, pt. 2, no. 
15, pp. 1233-1234. 8 holotype, Q allotype, 146,169 paratypes. 6268. 
Eotetranychus oistus Beer & Lang, 1958, Univ. Kans. Sci. Bull., vol. 38, pt. 2, no. 
15, p. 1241. ¢ holotype, ? allotype, 4 ¢,19 paratypes. 6269. 
Neotetranychus flabellosetus Beer & Lang, 1958, Univ. Kans. Sci. Bull., vol. 38, 
pt. 2, no. 15, p. 1239. ¢ holotype, ? allotype, 19 paratype. 6270. 
Neotetranychus hamus Beer & Lang, 1958, Univ. Kans. Sci. Bull., vol. 38, pt. 2, 
no. 15, p. 1238. ¢ holotype, ? allotype, 74,69 paratypes. 6271. 
Neotetranychus hispidosetus Beer & Lang, 1958, Univ. Kans. Sci. Bull., vol. 38, 
pt. 2, no. 15, p. 1237. ¢ holotype, @ allotype,6 4,79 paratypes. 6272. 
Neotetranychus undulatus Beer & Lang, 1958, Univ. Kans. Sci. Bull., vol. 38, pt. 2, 
no. 15, pp. 1239-1240. ¢ holotype, @ allotype, 1 ¢ paratype. 6273. 
Oligonychus flexuosus Beer & Lang, 1958, Univ. Kans. Sci. Bull., vol. 38, pt. 2, 
no. 15, pp. 1243-1244. ¢ holotype, @ allotype, 2 ¢,69 paratypes. 6274. 


5 


CATALOGUE OF LypEs IN SNow MuseuM—Parr V 313 


Schizonobiella aeola Beer & Lang, 1957, Pan-Pacific Ent., vol. 33, no. 2, pp. 87-89. 
2 holotype,  allotype, 1 ¢ ,9 9,2 nymphal paratypes. 6275. 


Trombiculidae 


Acomatacarus angulatus Greenberg, 1952, Ann. Ent. Soc. Amer., vol. 45, no. 3, pp. 
485-488. Larval holotype, 4 larval paratypes. 5369. 

Acomatacarus micheneri Greenberg, 1952, Ann. Ent. Soc. Amer., vol. 45, no. 3, 
pp. 480-482. Larval holotype, 10 larval paratypes. 5366. 

Acomatacarus (Xenacarus) plumosus Greenberg, 1951, Jour. Parasitol., vol. 37, 
no. 6, pp. 525-527. Larval holotype, 4 larval paratypes. 5367. 

Acomatacarus polychaetus Greenberg, 1952, Ann. Ent. Soc. Amer., vol. 45, no. 3} 
pp. 488-489. Larval holotype. 5371. 

Acomatacarus senase Greenberg, 1952, Ann. Ent. Soc. Amer., vol. 45, no. 3, pp. 
484-485. Larval holotype, 2 larval paratypes. 5368. 

Acomatacarus whartoni Greenberg, 1952, Ann. Ent. Soc. Amer., vol. 45, no. 3, 
pp. 489-491. Larval holotype. 5370. 

Cheladonta crossi Lipovsky, Crossley & Loomis, 1955, Jour. Kans. Ent. Soc., vol. 
28, no. 4, p. 139. Larval holotype, 5 larval paratypes. 5580. 

Cheladonta micheneri Lipovsky, Crossley & Loomis, 1955, Jour. Kans. Ent. Soc., 
vol. 28, no. 4, pp. 137-139. Larval holotype, 12 larval paratypes. 5578. 

Cheladonta ouachitensis Lipovsky, Crossley & Loomis, 1955, Jour. Kans. Ent. Soc., 
vol. 28, no. 4, p .139. Larval holotype, 13 larval paratypes. 5410. 

Euschéngastia cynomyicola Crossley & Lipovsky, 1954, Proc. Ent. Soc. Wash., vol. 
56, no. 5, pp. 240-243. Larval holotype. 5564. 

Euschéngastia diversa Loomis, 1956, Univ. Kans. Sci. Bull., vol. 37, \pt.2, no. 19; 
pp. 1337-1340. Larval lectotype (Loomis, 1962, Jour. Parasitol., vol. 48, p. 
1545). 5571. 

Euschoengastia eadsi Loomis & Crossley, 1963, Acarologia, vol. 5, no. 3, pp. 372- 
374. Larval holotype, | larval paratype. 6325. 

Euschéngastia finleyi Crossley, 1955, Jour. Parasitol., vol. 41, no. 3, pp. 289-291. 
Larval holotype. 5566. 

Euschéngastia jonesi Lipovsky & Loomis, 1954, Jour. Parasitol., vol. 40, no. 4, 
pp. 407-410. Larval holotype. 5558. 

Euschéngastia loomisi Crossley & Lipovsky, 1954, Proc. Ent. Soc. Wash., vol. 56, 
no. 5, pp. 243-246. Larval holotype. 5565. 

Euschéngastia trigenuala Loomis, 1956, Univ. Kans. Sci. Bull., vol. 37, pt. 2, no. 
19, pp. 1343-1345. Larval lectotype (Loomis, 1962, Jour. Parasitol., vol. 48, p. 
154). 5570. 

Fonsecia palmella Brennan & Loomis, 1959, Jour. Parasitol., vol. 45, no. 1, p. 62. 
Larval holotype, 6 larval paratypes. 6323. 

Hannemania multifemorala Loomis, 1956, Univ. Kans. Sci. Bull., vol. 37, pt. 2, 
no. 19, pp. 1247-1250. Larval holotype, 23 larval paratypes. 6324. 

Neoschéngastia brennani Crossley & Loomis, 1955, Ent. News, vol. 66, pp. 114- 
117. Larval holotype. 5567. 

Pseudoschéngastia farneri Lipovsky, 1951, Jour. Kans. Ent. Soc., vol. 24, no. 3, 
pp. 100-102. Larval holotype, 8 larval paratypes. 5363. 


314 Tue University SciENCE BULLETIN 


Pseudoschéngastia hungerfordi Lipovsky, 1951, Jour. Kans. Ent. Soc., vol. 24, no. 
3, pp. 95-99. Larval holotype, 46 larval paratypes. 5362. 

Speleocola tadaridae Lipovsky, 1952, Jour. Kans. Ent. Soc., vol. 25, no. 4, pp. 
134-137. Larval holotype. 5559. 

Trombicula arenicola Loomis, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 13, 
pp- 930-933. Larval holotype, 29 larval paratypes. 5555. 

Trombicula breviseta Loomis & Crossley, 1963, Acarologia, vol. 5, no. 3, pp. 374- 
376. Larval holotype, | larval paratype. 6326. 

Trombicula crossleyi Loomis, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 13, 
pp. 920-922. Larval holotype, 24 larval paratypes. 5551. 

Trombicula (Neotrombicula) finleyi Kardos, 1954, Univ. Kans. Sci. Bull., vol. 36, 
pt. 1, no. 4, pp. 88-90. Larval holotype, | larval paratype. 5373. 

Trombicula fitchi Loomis, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 13, pp. 
926-928. Larval holotype, 53 larval paratypes. 5553. 

Trombicula gurneyi campestris Loomis, 1955, Univ. Kans. Sci. Bull., vol. 37, pt. 1, 
no. 9, pp. 258-260. Larval holotype, 22 larval paratypes. 5576. 

Trombicula (Euschéngastoides) hoplai Loomis, 1954, Univ. Kans. Sci. Bull., vol. 
36, pt. 2, no. 13, pp. 924-926. Larval holotype, 23 larval paratypes. 5552. 

Trombicula kansasensis Loomis, 1955, Univ. Kans. Sci. Bull., vol. 37, pt. 1, no. 9, 
pp- 260-262. Larval holotype. 5577. 

Trombicula kardosi Loomis, 1954, Univ. Kans. Sci. Bull., vol. 36, pt. 2, no. 13, pp. 
929-930. Larval holotype, 25 larval paratypes. 5554. 

Trombicula (Eutrombicula) lipovskyana Wolfenbarger, 1952, Ann. Ent. Soc. 
Amer., vol. 45, no. 4, pp. 660-666. Larval holotype, 10 larval paratypes. 5372. 

Trombicula (Neotrombicula) loomisi Kardos, 1954, Univ. Kans. Sci. Bull., vol. 
36, pt. 1, no. 4, pp. 85-87. Larval holotype, 3 larval paratypes. 5375. 

Trombicula merrihewi Loomis & Lipovsky, 1954, Jour. Kans. Ent. Soc., vol. 27, 
no. 2, pp. 51-53. Larval holotype, 40 larval paratypes. 5557. 

Trombicula ornata Loomis & Lipovsky, 1954, Jour. Kans. Ent. Soc., vol. 27, no. 
2, pp. 47-51. Larval holotype, 39 larval paratypes. 5556. 

Trombicula trisetica Loomis & Crossley, 1953, Jour. Kans. Ent. Soc., vol. 26, no. 
1, pp. 32-34. Larval holotype, 13 larval paratypes. 5374. 

Trombicula (Leptotrombidium) twentei Loomis, 1954, Univ. Kans. Sci. Bull. 
vol. 36, pt. 2, no. 13, pp. 922-924. Larval holotype, 8 larval paratypes. 5550. 


Trombiculidae—Paraty pes 


Acomatacarus arabicus Radford. 1 larval. 5562. 
Acomatacarus lawrencei Radford. 1 larval. 5561. 
Acomatacarus thallomyia Radford. 1 larval. 5585. 
Acomatacarus tubercularis Brennan. | larval. 5593. 
Chatia setosa Brennan. | larval. 5594. 

Euschéngastia africana Radford. | larval. 5575. 
Euschéngastia campi Brown & Brennan. 2 larval. 5589. 
Euschéngastia cordiremus Brennan. | larval. 5590. 
Euschongastia hamiltoni Brennan. | larval. 5588. 


CATALOGUE OF Types IN SNow MuszEumM—Part V 315 


Euschéngastia luteodema Brennan. | larval. 5591. 

Euschéngastia pipistrelli Brennan. 1 larval. 5599. 

Eutrombicula lumsdeni Radford. 1 larval. 5584. 

Hannemania hegeneri Hyland. 2 larval. 5681. 

Neoschéngastia kohlsi Brennan. See N. paenitens Brennan. 

Neoschéngastia moucheti Brennan. | larval. 6287. 

Neoschéngastia paenitens Brennan, new name for N. kohlsi Brennan, preoccupied. 

itlarval. 5592. 

Neotrombicula saperoi Radford. 1 larval. 5582. 

Pseudoschéngastia guatemalensis Brennan. | larval. 5587. 
seudoschongastia occidentalis Brennan. | larval. 5586. 

Schéngastia guyanensis Floch & Abonnenc. | larval. 6288. 

Schéngastia haddowi Radford. 1 larval. 5583. 

Trombicula aplodontiae Brennan. | larval. 5598. 

Trombicula brevitarsa Radford. 1 larval. 5563. 

Trombicula canis Floch & Abonnenc. | larval. 5568. 

Trombicula hoogstraali Radford. | larval. 5573. 

Trombicula jamesoni Brennan. | larval. 5597. 

Trombicula (Neotrombicula) jewetti Brennan & Wharton. 2 larval. 5596. 

Trombicula (Trombiculindus) kansat Jameson & Sasa. | larval. 6328. 

Trombicula knighti Radford. 1 larval. 5581. 

Trombicula montanensis Brennan. | larval. 5600. 

Trombicula mount: Radford. | larval. 5569. 

Trombicula scottae Brennan. 3 larval. 5595. 

Trombicula texana Loomis & Crossley. | larval. 6327. 


Trombidiidae 


Neotrombidium tricuspidum Borland, 1956, Jour. Kans. Ent. Soc., vol. 29, no. 1, 
pp. 30-35. Larval holotype. 5572. 


Tole aie. 


} 
meet) 7 cma WAH eee 


- 
: 


' g 
LA 
‘+ se Li ' er bane PA 


7" mm eyes 
;* ro 
‘ - wo) 17 
oc) _ 
-16) 2D Gols a 
2 
7 - & 


we 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


RANE Wa STARE 
i ’ . } +t ; wa 
| Marine Biological Laboratory 
LIBRAR Y 
ICTY 98 1065 


WOODS HOLE, MASS 


ee eae 


THE LIFE CYCLE AND SOCIAL 
ORGANIZATION OF BEES OF THE GENUS 
Exoneura AND THEIR PARASITE, Inquilina 

(Hymenoptera: Xylocopinae) 


By 
Charles D. Michener 


VoL. XLVI Paces 317-358 OcroserR 28, 1965 No. 9 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 317-358 OcrosBeER 28, 1965 No. 9 


The Life Cycle and Social Organization of Bees of the Genus 
Exoneura and their Parasite, Inquilina (Hymenoptera: 
Xylocopinae)' 


By 


Cuartes D. MIcHENER” 


This paper consists of an account of biological observations on several 
species of bees of the genus Exoneura and on their social parasite and close 
relative, Inquilina. These bees belong to the same section of the xylocopine 
tribe Ceratinini as the genera Allodape, Allodapula, Exoneuridia, etc. This 
group of genera is virtually restricted to Africa and the Indo-Australian 
region and is well known as the only group of bees other than Apis and 
Bombus which practices progressive feeding of the larvae. 


SeNERAL ACCOUNT OF THE LIFE HISTORY OF EXONEURA 


Sakagami (1960) has given a review of the biology of this group of genera, 
most of which make their nests in hollow or pithy dry stems. The nests of 
most Exoneura are burrows through pith or rotting stems that are as soft 
as pith. Sometimes they clean the pithlike frass out of beetle burrows in wood 
(Rayment, 1935) or nest in cavities in galls (Rayment, 1951) but probably 
most species do not habitually occupy beetle burrows as does Allodapula in 
Australia (Michener, 1962). However, nests of three species have been re- 
corded from beetle burrows in solid fence posts and spars (Rayment, 1954). 

‘Contribution number 1124 from the Department of Entomology, The University of 
Kansas, Lawrence. 

“Completion of this paper at this time is possible thanks to a grant (G 11967) from the 
National Science Foundation. It is a pleasure to acknowledge facilities provided by the Uni- 
versity of Queensland through the courtesy of Mr. F. A. Perkins, Reader in Entomology and 


head of the Department of Entomology of that Institution. The studies in Australia were pos- 
sible thanks to a research grant from the Fulbright program of the United States Government. 


318 Tue University ScIENCE BULLETIN 


Beetle burrows in branches and twigs are also used (e.g., Rayment, 1956) by | 


some species. Data on the related genus or subgenus Exoneurella are pre- 


sented elsewhere (Michener, 1964). 


Exoneura makes no separate cells, all young being reared together in the | 
burrow. Rayment’s early (1935) references to cell partitions are in error. I | 


never saw the threads on the walls of the nest tubes reported by Rayment 
(Gee wl95 ib): 

Nest entrances are regularly constricted, either being excavated in that 
way or narrowed by a constructed collar made of bits of the material removed 
from the burrow. The narrow entrance is important in defense of the nest. 
At a small disturbance, a female in the nest may bite at an intruder or a fine 
fiber held at the nest entrance, but usually she quickly turns and blocks the 
hole firmly with the flattened dorso-apical part of the abdomen. The females 
also secrete from the mouth region, seemingly from the mandibular glands, 
a brown liquid having the odor of tenebrionid or carabid beetles. They emit 
this material either at the nest or, if captured or greatly disturbed, away from 
the nest. On one occasion I found females of E. bicolor so common (11 
miles south of Uralla, New South Wales, on November 5, 1958) visiting the 
flowers of Daviesia latifolia that when the bushes bearing the flowers were 
beaten with a net, the odor could be perceived several feet away. 

The effectiveness of adult females in defending their nests is shown by a 
series of nests of E. variabilis set up for observation in Brisbane. Within a 
day those containing no adults were robbed of all immature stages by ants. 
Those containing adults mostly survived for many days. 

Eges of most species are laid in a mass (usually criss-crossed) at the bot- 
tom of the burrow, not attached to the burrow walls. However, in some 
species the eggs are attached by their posterior ends to the wall of the burrow 
and project into the lumen of the nest (Rayment, 1948, 1951; Erickson and 
Rayment, 1951; and E. aterrima discussed below). In some cases such at- 
tached eggs are placed in a row; in other cases they are scattered about on 
the inner nest wall. The duration of the egg stage was not determined by me 
but is obviously long enough that numerous eggs (10 or more) may accumu- 
yate in a single nest, probably laid by a nels female, before the first ones 
hatch. The ovaries mature only one egg at a time, as in most bees, although 
three or four may be nearly mature (Figs. 77, 78). Rayment (1948) showed 
that at the temperatures of late winter in Sydney the egg stage lasted between 
two and three weeks. 


Larvae hatching from loose eggs remain in a cluster at the bottom of the | 


nest until they grow large enough to maintain a position above the bottom. 
Larvae hatching from attached eggs may retain their positions on the nest 
walls for a time Gna may be fed pollen in this position. 

The larvae are fed progressively with pollen mixed with a liquid (nectar?) 


LirE CyYcLE AND SOCIAL ORGANIZATION OF BEEs 319 


to form a rather sticky but firm mass. Sometimes more than one larva feeds 
from a single mass and this may be the usual way of feeding small larvae 
lying in groups. Larger and sometimes even the smallest larvae (e.g., in E. 
aterrima, see below) are fed by individual pollen masses placed on their 
venters. Even within a species the manner of feeding varies, as was noted for 
E. variabilis (see below) and for E. roddiana by Rayment (1948), who 
observed both communal and individual pollen stores, depending on the 
arrangement of the larvae in the nest. 

At any one time the number of larvae having pollen supplies on which to 
feed is small. For example, in one group of nests of E. variabilis (see below) 
only eight larvae out of 74 had food available when the nests were opened. 
Rayment (1951) showed that in 17 nests of E. richardsoni containing 209 
larvae, only six nests contained food and these had a total of only 23 pollen 
masses. 

Rayment has repeatedly (1946c, 1951; Erickson and Rayment, 1951) de- 
scribed the feeding of larvae, especially young ones, by a liquid from the 
mouthparts of females which he stated was “pap” from the pharyngeal glands. 
The scarcity of larvae feeding on pollen supports this idea. However, the 
long larval life suggests sporadic feeding. Moreover, the finding of very 
young larvae feeding on pollen indicates the inaccuracy of the idea that 
young larvae must be fed only liquid. (For E. aterrima, see below; for E. 
angophorae, asimillima, richardsont, etc., see Rayment, 1951, in which publi- 
cation pollen is reported among the eggs and very small larvae lying in a 
cluster at the bottom of the nest.) Disturbed adults often move the larvae 
about in the nest with their jaws and mouth the larvae; such movements 
might simulate feeding and cause erroneous accounts of feeding. Feeding 
of larvae by glandular secretion (or by other liquid from the mouth of the 
adult) requires verification, although it seems probable in view of the detailed 
accounts of Rayment. 

The young larvae lack ventrolateral lobes but have strong subspiracular 
ridges (Figs. 4,53, and 57). Ventrolateral lobes appear on the mesothorax in 
the penultimate larval stage and exist on most segments in the ultimate larval 
stage. The mesothoracic arms are much the largest of such lobes (Syed, 1963). 
As described below for E. variabilis, these lobes (arms, or pseudopodia) are 
partially retractile in the last larval stage but shrivel in the prepupal part of 
the last stadium. Rayment (1948) appears to have been correct in regarding 
these lobes as non-secretory protuberances. His subsequent certainty (e.g., 
1951) that they have secretory importance is not supported by my observa- 
tions of their external and internal structure. As Rayment said, the meso- 
thoracic arms are often mouthed and chewed by larvae but so are any other 
small soft objects which the larvae can reach. This activity therefore does 
not indicate any exudation. In contrast to Allodapula (Michener, 1962), 


320 Tue UNiversiry ScIENCE BULLETIN 


larvae of Exoneura are constantly active, bending about as though reaching 
for something. Immature stages generally lie roughly in order of youngest 
at the bottom to oldest near the entrance of a nest. Food, therefore, has to 
be carried past a series of pupae and prepupae to the larvae which are being 
fed. Feces are produced while larvae are still feeding (Rayment, 1951) and 
must be carried out of the nest by adult bees. 

The duration of the immature stages is not well known but is long, as 
Rayment clearly showed. Rayment (1951) thought that development from 
egg to adult of E. rufitarsis and richardsoni required from early July (mid- 
winter) to early December (early summer). He also showed (1948) that 
both larvae and adults could survive up to 90 days closed in a container with- 
out food, and Erickson and Rayment (1951) showed survival of adults for 
over 100 days at 10°C. Thus survival of both larvae and adults during periods 
of drought when flowers are scarce and during the winter is not surprising. 
Overwintering of larvae certainly occurs (Rayment, 1946b, 1951). It is prob- 
able in some of the species discussed in detail below. However, the principal 
way of overwintering is by means of fertilized adult females. Often several 
females overwinter together. In the spring new colonies are commonly 
established by single females (EF. variabilis, see below; E. concinnula, see 
Rayment, 1951). Sometimes, however, two or more adult females are present 
in nests in spring (E. dlastris, see Erickson and Rayment, 1951; E. richard- 
som, see Rayment, 1951). Whether or not supernumerary females later 
disperse from such nests, leaving monogynous colonies, is not known, but 
in E. variabilis such dispersal is suggested by the polygynous October (early 
spring) nests and the largely monogynous November nests. Some species 
maintain more nearly continuous reproductive activity than does E. vari- 
abilis, as is quite clearly the case for E. roddiana, concinnula, and rufitarsis 
which had many eggs in their nests in July (midwinter) (Rayment, 1951). 

At least in summer many nests contain two ore more adult females. Large 
groups usually consist in part of young individuals which have not yet dis- 
persed, but as many as three or four females may be long term inhabitants of 
a nest. Such groups, as will be shown in detail later, commonly consist of 
one or sometimes two egg layers® and one or more workers. The important 
role of workers in pollen collecting has been shown by Michener (1963). 
Further information on the division of labor and social organization is 
presented in the body of this paper and in brief form, in the summary. 

Rayment (1951) described males as guarding nests, leaving and returning 
to them, and in short constituting a significant part of the society. His obser- 
vations were probably made in winter and early spring when low tempera- 
tures presumably prolong all activities. Data obtained by me show that males 


3 c . . 
[ have not used the word queens, because so many of the nests at least in the species I 
studied contained only one female, which was like the egg layer in nests containing more than 
one female. 


No 
ast 


Lire CycLE AND SOCIAL ORGANIZATION OF BEES 3 


are produced in rather large numbers but after reaching maturity stay in 
nests for much shorter periods of time than females, so that adult nest popu- 
lations are predominantly female. Presumably under warm conditions males 
leave the nests and soon mate and die. 

A swarm of males of E. hamulata was found about noon of April 5, 1959, 
on Noosa Hill in Noosa National Park, Queensland. A few hundred males 
were flying about in the sun between large bushes, sometimes only two or 
three in a place, sometimes 25 forming a loose flying swarm. The bees never 
seemed to alight but remained in flight continually. No females were seen, 
but I have little doubt about this being a mating swarm. Probably only those 
species with large eyed, hairy males indulge in such swarming. 


METHODS 


The principal method used was to collect nests, plug the entrances in the 
field, bring them to the laboratory, and preserve for study the entire contents 
of each nest. Nests were mostly taken in early morning, late afternoon, or 
on cool or rainy days when all the bees would presumably have been in their 
nests. 

Each female bee was examined using the methods described by Michener, 
Cross, Daly, Rettenmeyer and Wille (1955). The wings of E. variabilis do 
not often become much worn. The number of nicks on the forewing margins 
was recorded in the belief that it provides some rough index of previous flight 
activity. For purposes of the present paper, these data are given as unworn 
(no nicks), slightly worn (1 to 5 nicks in total on both wing margins), well 
worn (6 to 10 nicks in total on both wing margins), and much worn (11 or 
more nicks in total). The last category was very rare. Mandibular wear was 
rather slight and recording of it was abandoned. 

Spermathecae were never found partially empty and it seems probable that 
once mated, a female retains a supply of sperm for life. 

For the ovaries, the maximum width of each was measured as well as 
the length of the longest developing egg. The average width of the two 
ovaries is correlated with the length of the longest developing egg (see figure 
1), and the latter has been used in the following sections of this paper on the 
theory that it better represents the reproductive activity of the female con- 
cerned (see also figures 65 to 82). 

For convenience in expression, ovaries with the longest oocyte 0.50 mm. 
long or longer are considered to be enlarged in the discussions which follow. 
As can be seen from figure 1, this means enlargement beyond the initial stages 
common among slender ovaries. 

The whitish areas noted in the section on ovaries of E. variabilis were 
used as an aid in judging age and egg laying by bees in the latter part of 
the study, after their meaning was realized. 


322 Tue University ScrENCE BULLETIN 


In the following discussions the words “adult” or “female” always mean 
adult female except when otherwise specified. Callow adults are those show- 
ing paler coloration than mature adults. The youngest callows have milky 
wings but the wings harden and become transparent long before full colora- 


$50 
600 
550 
500 
450 
-400 
.350 
.300 
.250 
.200 
150 


100 


Average Maximat Width of Ovaries in Millimeters 


.050 


-29 v5 1.25 v5 


Length of Longest Oocyte in Millimeters 


Fic. 1. Scatter diagram showing relations between ovarian width (maximum width of each / 
ovary measured and averaged for each individual) and length of the longest oocyte in Exoneura — 
variabilis. Small dots represent one individual; numbers of individuals represented by larger 
dots may be estimated by their diameters. 


(oN) 
Nh 
Us 


Lire Cycle AND SOCIAL ORGANIZATION OF BEEs 


tion of the body is attained. Callows are unfertilized, with unworn wings, 
and with regularly formed, slender ovaries (longest oocyte less than 0.22 mm.) 
having no whitish areas in the bases or stalks. Fully colored but otherwise 
similar females are also considered as young but are not recorded as callows, 
being listed as mature adults. 


EXONEURA VARIABILIS RAYMENT 


Habitat: Observations were all made in southeastern Queensland. Nests 
were found in large numbers at the margins of the rain forest areas which 
occur on the higher mountains of this region. None was found in Eucalyptus 
savanna or in the depths of the rain forest, but breaks in the forest caused by 
roads as well as natural forest margins provided suitable habitats. 

The localities where series of nests were taken for study are as follows: 
(1) roadside at summit of Cunningham’s Gap, near Mount Edwards (Octo- 
ber 3, November 9, December 27, February 25); (2) forest margins in Lam- 
ington Naional Park, near Binna Burra, in the McPherson Range (Decem- 
ber 8, January 11-13); and (3) roadside and forest margins on Tamborine 
Mountain, on the side nearest to the town of Tamborine (December 18, 
February 15, March 22). 

Nests: From the localities and dates listed above, 233 nests or at least 
burrows occupied by one or more Exoneura variabilis were opened for study. 
A few of these were merely burrows in which bees were resting, or were 
probably nests in the process of excavation, but 221 were nests in later stages 
of development. 

The nests are simple burrows in broken or cut, standing, dry, pithy stems 
of bushes, brambles, or vines. Rarely, they are in stems that are soft because 
of rotting rather than because of pithyness. The holes are obviously made by 
the bees themselves, as in most other Exonewra and in contrast to species of 
Allodapula (see Michener, 1962). They enter the stems at the broken or cut 
ends. In slender stems the burrows are nearly straight while in thick stems 
with much pith the burrows are often sinuous. The stems are typically erect 
and the nests more or less vertical, but they often also occur in slanting, 
horizontal, or even hanging stems. In the last case, the entrance is the lowest 
part of the nest. Such orientation is especially common in roadside bushes of 
Lantana whose dead stems have been cut or broken. 

Plants utilized for nesting sites were the following: Rubus rosaefolius 
(most nests at Cunningham’s Gap, many at Binna Burra, few at Tamborine 
Mountain), Lantana camara (most nests at Tamborine Mountain, few at 
Binna Burra), Eupatorium adenophorum and Plectranthus graveolens (many 
from each at Binna Burra), Homalanthus populifolius (some from Tam- 
borine Mountain), Erigeron (few from Cunningham’s Gap), Deeringia bac- 
cata (few from Binna Burra), cultivated Hydrangia (one from Binna Burra), 


324 THe UNiversiry SciENCE BULLETIN 


dead stems of rain forest vines (few from Binna Burra), and rotten stems of 
various plants (few from Binna Burra). Rayment (1949b) recorded a nest 
from a dead stem of a “reedy grass or sedge.” 

A few nests in Lantana and Deeringia were excavated through the dry 
portions of the stems and terminated at the beginning of moist living ma- 
terial. The same phenomenon was common in nests of Exoneurella lawsont 
(Rayment) (see Michener, 1964) and was shown to influence the length of 
the burrows. These few nests of Exoneura variabilis in such stems were 
excluded from consideration in the following paragraphs on nest size and 
growth although all of them are within the extremes noted for nests in dry 
stems. No significant correlation was noted between nest measurements and 
the kind of stem in which the nest was located. 

The nests are burrows 2.0 to 3.0 mm. in diameter. The mean of 140 
measurements of burrow diameters is 2.5 mm. At the entrances the burrows 
are narrowed to 1.5 to 1.6 mm. in diameter. The burrows are excavated with 
narrow entrances. However, if an entrance is artificially enlarged or if the 
apex of a stem, including the constricted part of a burrow, is broken off, the 
Exoneura narrows the entrance by means of small pieces of pith stuck to 
each other and to the wall of the burrow near the entrance. Often such pieces 
of pith can be noted at nest entrances not altered in any way by an experi- 
menter. 

Nests in dead, dry twigs or stems and containing immature stages were 
24-295 mm. in depth. The nests are seemingly deepened during the time 
that the young are being reared and the nest reused, as shown by the follow- 
ing data: Nests containing a single adult female and eggs or small larvae but 
no older immature stages ranged from 24-190 mm. in depth, mean 70.68.65 
mm. (n=32). Older nests, containing a single adult female and immature 
stages up to pupae, but no young adults, were 25-260 mm. in depth, mean 83.2 
£5.95 mm. (n=71). Still older nests containing more than one adult or con- 
taining a young egg layer and also older immatures apparently her sisters (1.e., 
nests whose use is continuing beyond the brood of the founding female) 
ranged from 38-290 mm. in depth, mean 118.411.41 mm. (n=33). Differ- 
ences in depth between the latter group of nests and each of the others 
are significant (p<.01) but the differences between the first two is not 
(a0 <p 10): 

A few nests were found which contained larvae or pupae and at the same 
time had loose bits of pith outside the entrances; this observation corroborates 
the statistical evidence that pith is sometimes and perhaps regularly removed 
from the bottom of a nest and pushed past the brood and out the nest entrance 
by adult bees with resulting lengthening of the burrow. 

Seasonal variation in nest depth was suggested, but the differences be- 
tween seasons were not significant. Nests in October and November were 


YI 


Life CyYcLE AND SOCIAL ORGANIZATION OF BEEs 32 


24-143 mm. in depth, mean 79.59.84 mm. (n=13). In February and March 
depths were about the same, 25-295 mm., mean 78.0£9.58 mm. (n=28). In 
summer, however, nests seemed deeper, 25-260 mm. in December and Jan- 
wary, mean 95.36.12 mm. (n=95). 

Life History Data: Group 1. This species was first encountered by me at 
Cunningham’s Gap in the spring, on October 3, 1958. Nine nests were found; 
in them were 17 adult females and no adult males. Three of the nests con- 
tained only one adult; all of these lacked immature stages and were pre- 
sumably excavated for establishment of new nests. The remaining six nests 
contained two or three adults each and five of them contained larvae, four to 
six each, mostly small; middle sized and large larvae were present in only 
one nest. Since there were no eggs nor pupae, I thought at first that these 
larvae had passed the winter in the nests and did not hatch from eggs early 
in the spring. 

At least 65 percent of the females taken from the nests in October, in- 
cluding all that were alone in their nests, contained no sperm cells in the 
spermathecae. In all but two the largest developing egg was less than 0.45 
mm. long; one was 1.00 and another 1.38 mm. in length. Most of the females, 
therefore, were unfertilized with scarcely enlarged ovaries, and most did not 
show any wing or mandibular wear. These data are in various ways con- 
fusing. If they represent an overwintered population, there should be adult 
males to fertilize at least the isolated unfertilized females; also they do not 
correspond with data from the overwintering fall populations (groups 8 and 
9) which were studied in much larger numbers and which included no 
young larvae, very few mature ones, and few unfertilized females. It is my 
tentative supposition that the high percentage of unfertilized females and 
the lack of eggs represent sampling errors (the sample size was small) and 
that the young larvae found in five of the nests hatched from eggs laid in 
September. 

Group 2 consists of 11 nests from the same locality taken on November 
9. Each contained only a single adult female, except for one with two. Im- 
mature stages were absent in three of the nests. In one nest, the only im- 
mature individual present was a single prepupa. In one nest there were 
seven small larvae and five eggs. In the remaining six nests there were two 
to 11 (average 5.5) eggs and no other young. The nest with two adult females 
was the one with 11 eggs. 

All of the adults that were successfully dissected (9) were fertilized. Those 
with eggs in the nest had enlarged ovaries with the longest oocyte 1.00 mm. 
or longer. Those without eggs had more slender ovaries with the longest 
oocyte .035-0.50 mm. in length. Of the 12 females, eight had one or more 
nicks in the wings and two were much worn. 

Clearly conditions had changed since the preceding month, most of the 
females being alone, fertilized, and in egg laying condition. 


326 Tue University ScrENcE BULLETIN 


Taste 1. Exoneura variabilis. Numbers of Young and Adults in Nests of Group 

3, taken on December 8. (Each horizontal row represents a nest. Nests without 

immature stages or with only eggs were omitted. Under adults C=callows. “I” 
indicates the parasitic bee, /nquilina.) 


Nest Small Medium Large Total Adult Adult 


No. Eggs Larvae Larvae Larvae Prepupae Pupae Young ee ad 
DN ie eee 3 1 4 1 

500; = 2 2 - 4 1 

Fo | ol NS 2 ] 1 2 6 1 

Di) (ee ee a 7 7 1+11 

XS) ee 2 2 1 

Di] Gum ene 2 4 =a 6 1 

PAY eee II 21° 2 5 1 

Oa ae ee *: ee b. p 2 1 

20 AR apes = = D, = 2 1 

ZO TBE eeceare = a z 1 3 a. 1 1 

303 3 = = 2 - II 111 12 2 

3043) == : z 2 iv = 2 1 

SOB pee == f 3 3 1 7 1 

SION) cosets eZ 1 3 1 

STA ia a 2 1 3 1 

SB ere tee 2 2 4 2 

321B 1 1 2+1C 1C 
3272BNE eS 2 2 D 

318B a 2 2 i 

Depo Pts 4 4 2 

De Digs Er et Es 7, 3 2 10 D, 

7) 13 - : 7 20 1 

2 gee 1] 3 3 17 2 

281 = | 3 1 

DS 6iF gi eee 5 re 3 3 ] 

288 pee de l(0) be = 2 1 3 16 3 

SU) Aaeeaiiae  2e = 1 a0 1 e ” 2 1 

29 6B ieee 3 - fe S a 1 4 1+2C 

2:9 3s eee mat) 4 i 2 cS = 12 3 

SOS eee, 4 & 4 p) 12 2+2C 
3020 2 - s = 4] 16] 22 3+11 

306B ee lel 8 z. ] is 4 24 KC, AC. 
SOR: ee l i a 7 1 14 DA aN XC 
SAB a 20 = = - = 10 30 2+4C 3C 
S153) ee 1] 6 6 8 3 6 40 3+10C 10C 
SIG: pees | re = s I 5 12 2 

SHIGE ee see I 2 2 5 1 


Group > consists of 44 nests taken at Binna Burra on December 8. Of 
these, four were only burrows without immature stages, three of them only 
10 to 18 mm. deep, apparently being excavated. Each contained a single 
fertilized adult female, two with slender and two with enlarged ovaries. 


WW 
i) 
“SI 


Lire Cycle AND SocIAL ORGANIZATION OF BEEs 


Three of these four bees showed some wear of the wing margins. Ap- 
parently new nests were being established at this season. 

Three other nests had progressed somewhat farther and contained one 
to four eggs each but no other young. A single adult female was in each 
of these nests; all three were fertilized, one had enlarged ovaries while two 
had slender ovaries, one showed some wing wear. 

The contents of the remaining 38 nests in Group 3 are indicated in Table 
1. Nests 274, 300B, and 287 were like those discussed in the preceding para- 
graph except that some of the eggs had hatched. The majority of the nests 
were older and their interpretations more complex, as indicated below. 

Several nests (269-322B) contained only older larvae or pupae. Except for 
303B, 317B, 321B, and 322B, these contained only one adult, which had been 
fertilized. About half of these females were unworn; the remainder showed 
wing wear. The ovaries were slender (longest oocyte 0.18-0.28 mm.) except 
for 276 and 310B which had swollen ovaries (longest oocyte 1.23 and 1.00 
mm.). Presumably at least some of these adult females were of the same 
generation as the immature stages; others may have been mothers of the im- 
matures in their nests. Those with enlarged ovaries would presumably soon 
have laid eggs, so that their nests would have been similar to those discussed 
below with young as well as old immature stages. Nest 322B contained two 
adults, both fertilized and with worn wings, one with slender ovaries, one 
with rather enlarged ovaries (longest oocyte 0.78 mm). Nest 303B con- 
tained two adults, both unfertilized, with very slender ovaries, and worn 
wings. Nest 317B also contained two adults, one fertilized and unworn, one 
unfertilized with worn wing margins (worker). Both had slender ovaries. 
Nest 275 contained two adults; their spermathecae were lost in dissection; 
both had worn wings; one had slender, the other enlarged ovaries. 

The remaining nests contained immatures of various ages, often young 
ones (eggs or small larvae) and older ones (prepupae, pupae, or callow or 
young adults) with few intermediates. For example, nests 272 and 273 con- 
tained eggs and young larvae as well as unworn adults with very slender 
ovaries without whitish areas; in addition 273 contained a fertilized, well 
worn adult with more robust ovaries (longest oocyte 0.55 mm.) with whitish 
areas in the stalks. The latter individual must have laid the eggs and could 
have been also the mother of the young adult. As shown in Table 1, most 
of the nests (numbers 272-319B) had a larger number of immature stages 
than did 273. One of these nests (306B) contained only callow adults and 
another (272) only young adults with slender ovaries. In spite of the eggs, 
no old adults were in these nests. Presumably they had died by the time I 
opened the nests. Nests 273, 281, 286, 288, 289, 296B, 315B, and 319B con- 
tained only one mature adult each. (In nests 288 and 315B, two of the three 
adults shown in Table 1 were probably young, perhaps only slightly older 


328 Tue UNiversiry SciENCE BULLETIN 


than those recognized as callows. Perhaps they were workers.) These mature 
adults were fertilized; those in nests 273, 286, 288, and 315B had worn wing 
margins and swollen ovaries (longest oocyte 0.55-1.42 mm.) while those in 
nests 281, 289, 296B, and 319B had unworn wings and slender or slightly 
enlarged ovaries (longest oocyte 0.17-0.87 mm.). Those having the longest 
oocytes, 0.87 mm. or longer, had eggs in the nests; the others (except for 
273 in which the longest oocyte was 0.55 mm.) did not. Probably only the 
worn individuals, and perhaps not all of them, were mothers of the older 
immatures in the nests. The other nests (298B, 301B, 302B, 308B, 314B, and 
316B) contained two or three seemingly mature adults each. In 298B and 
301B, the mature females were all fertilized; one in each nest had somewhat 
enlarged ovaries (longest oocyte 0.55 and 1.08 mm.) and worn wing margins, 
the others had slender ovaries and unworn wings. Nests 302B, 314B and 
316B had ovaries of all mature females enlarged (longest oocytes 1.03 to 
1.70 mm.). All were fertilized except one in 302B. Most showed wing wear 
but one each in 314B and 316B did not. The five mature females in 308B 
included two with swollen ovaries (longest oocytes 0.70 and 1.35 mm.) and 
worn wing margins which were probably fertilized. The other three had 
slender ovaries. One was unfertilized with worn wings, one fertilized with 
worn wings, and one fertilized with unworn wings. 

Group 4 consists of 35 nests taken at Tamborine Mountain on December 
18. Of these, one was only 11 mm. deep and contained only an unfertilized, 
unworn adult with slender ovaries. One lacked immature stages but con- 
tained an unworn adult with large ovaries, probably ready to start laying. An- 
other contained a similar but somewhat worn adult. Three nests contained 
two, four, and five eggs respectively but no other young, and a single fer- 
tilized adult with large ovaries (longest oocytes 1.13 and 1.3 mm.) was in 
each; one had unworn wings, another slightly worn wings, and the third had 
much worn wings as though she had had some other activity before laying 
the few eggs observed. Such a bee might come from a nest that had been 
destroyed, or might be a worker that later became a reproductive. 

Two nests lacked immature stages entirely but contained two and four 
adults respectively. Each contained one fertilized adult with somewhat en- 
larged ovaries (longest oocyte 0.75-1.20 mm.); the others were unfertilized 
with slender or slightly enlarged ovaries (longest oocytes 0.23-0.63 mm.). The 
wings were unworn or nearly so. Presumably these were adults of the same 
brood, reared in these nests. In each nest the fertilized bee would presumably 
soon have started laying; the others might have dispersed or remained as 
workers. 

The contents of the remaining nests in Group 4 are indicated in Table 2. 
Nest 311T seems to have passed just beyond the stage of those discussed in 
the preceding paragraph in that eggs had been laid. One adult, the egg 


Lire CycieE AND SocIAL ORGANIZATION OF BEEs 329 


layer, was fertilized with enlarged ovaries (longest oocyte 1.23 mm.) while 
the other was unfertilized with slender ovaries and must have been a 
worker. Both had worn wings. 

The nests listed as 297T to 326 in Table 2 contained only older larvae 
and pupae. In those containing only one female (297T, 308T, 324, 326), she 
was fertilized, and except for 324, had much enlarged ovaries (longest oocyte 
1.13-1.50 mm.), suggesting that egg laying was about to begin. Nests with 
two adult females (302T, 318T) contained one adult each that was fertilized 
with enlarged ovaries (longest oocyte 0.63 mm.) and one each that was un- 
fertilized with slender ovaries. The first would presumably have soon started 
egg laying while the second, probably a sister in each case, would have re- 
mained as a worker or perhaps left to start a new nest. 


Taste 2. Exoneura variabilis. Numbers of Young and Adults in Nests of Group 

3, taken on December 18. (Each horizontal row represents a nest. Nests without 

immature stages or with only eggs were omitted. Under adult females, C=callows. 
“T” indicates the parasitic bee [nquilina.) 


Nest Small Medium — Large Total Adult Adult 
No. Eggs Larvae Larvae Larvae Prepupae Pupae Young 2 @ 644 
ha 6 6 2 

225) ae 3 3 1 
ee 4 4 2 

0 ———n 1 6 7 1 Bs 
3 ili hn 3 3 2 1 
300 Gee I 1 2 1 

326° 3 3 I 

0S 2 1 2 3 ] 

31 5) 2 4 11 1 

375 rr 5 2 7 1 

327) 4 l 5 I 

323) ees 3} 2 5 1 

2 OG Deane 2, 7 1 10 6 

25 eee 1 3 3 7 3 

BORN oo: 8 3 3 i 21 4 

BU .2.-.--c----- 2 | 6 9 4 

3034) eee 8 2 3 13 2+11 
304i 3 1 . 4 8 2 

310) 6 2 1 6 ib) 5+1C 1 
3p ers 8 l l 1 11 2 

317A ee 2, 1 1 4 2 
2c 3 2 3 8 2 

320 6 3 2 11 5 

305). eee 8 + 2: | = 15 3 

31161 rr 10 2 2 2 5 21 2 

Balti ns a Pe l es 81 1-+-8I 3+11 

200 5 ee 9 2 2 l 5 19 2, 


WwW 
1S) 
=n) 


Tue University ScrENCE BULLETIN 


Nests 305T, 313T, 321T, 322T, and 323 contained only one adult each. All 
were fertilized except the one in 313T which was unfertilized, unworn, with 
slender ovaries, and could not have laid the eggs in the nest. The adult in 
305T had slender ovaries and worn wings; the adults in 321T, 322T, and 323 
had enlarged ovaries and those in the first two had unworn wings. It is very 
likely that such unworn individuals and perhaps the others as well are not 
mothers of the older immature stages in their nests; probably they are their 
sisters. They are, however, mothers of the eggs, and presumably care for and 
protect the older larvae and pupae present as well as their own offspring. 

Nests 293T to 325 on Table 2 all contained more than one adult. In each 
case one was fertilized, with large ovaries, and in a few cases with worn 
wings, while the other individuals were unfertilized, with slender ovaries and 
usually unworn wings. The unfertilized individuals may function as work- 
ers; presumably they are sisters of the egg layer in each nest and also of the 
pupae and prepupae when present. Nest 316T was similar except that the 
individual with slender ovaries was fertilized. Nests 304T and 328 contained 
only unfertilized individuals with slender ovaries which could not have pro- 
duced the eggs found in those nests. 

Group 5 consists of 23 nests. Some were taken from Tamborine Mountain 
on December 18, established in Brisbane for observation, and opened on De- 
cember 26; the others were taken at Cunningham’s Gap on December 27. 
Two lacked immature stages and contained one adult each. One of these 
adults was fertilized, unworn, with enlarged ovaries (longest oocyte 1.18 
mm.). The other had slender ovaries but its other characteristics were not 
ascertained. 

Three nests contained one, two, and seven eggs each and no other im- 
mature stages. The first two contained one adult each, fertilized, one with 
worn wing margins, the other unworn, both with enlarged ovaries (longest 
oocyte 0.90 and 1.35 mm.). The third nest contained two females, both 
fertilized with well worn wing margins, one with enlarged ovaries (longest 
oocyte 1.48 mm.), the other with long slender ovaries except for one large 
oocyte 1.45 mm. in length. The latter, although fertilized, may have been 
more or less workerlike in function. 

The contents of the remaining 18 nests of Group 5 are indicated in Table 
3. Nests 331 to 379 in Table 3 contained only older larvae and pupae. Of 
the two adults in 331, one was fertilized with enlarged ovaries (longest 
oocyte 1.18 mm.) and the other was unfertilized, unworn, with slender 
ovaries (longest oocyte 0.15 mm.), probably a worker. Nests 376, 377, and 
379 each contained a fertilized, unworn adult with somewhat enlarged 
ovaries. Probably eggs would soon have been laid; also it seems very probable 
that these adults were sisters rather than mothers of the pupae and prepupae 
in their nests, 


WwW 
— 


Lire CycLE AND SOCIAL ORGANIZATION OF BEEs Ss 


Taste 3. Exoneura variabilis. Numbers of Young and Adults in Nests of Group 

3, taken on December 26. (Each horizontal row represents a nest. Nests without 

immature stages or with only eggs were omitted. Under adults, C—callows. “I” 
indicates the parasitic bee /nquilina.) 


Nest Small Medium Large Total Adult = Adult 
No. Eggs Larvae Larvae Larvae Prepupae Pupae Young 29 fg 
331 eae = = & 61 = 61 2 

16 3 re 3 | 

VE 2 Zs 2 l 

3 ea 2 2 | 

374) eee ] 2 3 | . 
302: ae 9 | 2 10 ee 2 24 1+4C 4C 
BS Ome erie 10 5 1 5 21 1 

2) 2 2 = - Z 2, 6 1 
Ds 2 é ! 4 3 10 2-2 2C 

3 4 3 2 5} | 4 19 2+11 a 
Re === 5 ] 4 Zi | 13 4 1 
3 Z 2 D. 6 3 

37/3) (ee 8 > l 5 £ 19 6+2C 

Soil ee 7 l 4 12 3+11 

i). ee 1] 5 4 20 2+1C 

DD 8 4 12 2 

32) =e 8 2 10 2 

oo) = 6 3 1 10 2 


The remaining nests contained a variety of immature stages. Nests 374, 
382, 386, and 388 contained only one adult each (if callows are ignored); in 
every case this adult was fertilized, with worn wings, and with much en- 
larged ovaries (longest oocyte 1.50 to 1.68 mm.). Nests 329, 331, 332, 373, 
375, and 381 had one fertilized individual with swollen ovaries (longest 
oocyte 1.18-1.58 mm.) in each. Two had worn wing margins, three did not. 
The other adults in each of these nests were unfertilized, unworn, with 
slender ovaries, and in some cases were known to function as workers. Nest 
390 was similar except that the probable worker was fertilized, although with 
slender ovaries. Nests 385, 389, and 391 had two mature adults each, both 
fertilized; most were worn and all had enlarged ovaries (longest oocyte 0.85- 
152 mm.). Nest 330 contained two unfertilized adults with slender ovaries, 
one with worn wing margins. Probably both functioned as workers. 

Group 6 consists of seven nests taken at Binna Burra, January 11-13. They 
show no differences from those of Group 5 except that the small number of 
eggs may be significant. Five of the nests contained eggs, but three contained 
only one each, the others two each. Only two of the nests contained small 
larvae, one five, the other one. Correlated with the small number of eggs 
was the few adults with enlarged ovaries; there were only two. One was in 


WwW 
Oo 
is) 


Tue University SciENCE BULLETIN 


a nest without other adults and with one pupa; presumably the adult would 
soon have laid eggs. The other was in a nest containing young from the egg 
to pupal stages and two other adults, unfertilized, with slender ovaries, prob- 
ably functioning as workers. 

Group 7 consists of 31 nests taken at Tamborine Mountain on February 
15. The approach of autumn was already obvious. Fourteen of the nests 
contained no immature stages. Of the 17 remaining nests, only three con- 
tained eggs. Two others contained small larvae, so that five or about one 
sixth of the nests contained eggs or small larvae. Ten contained prepupae 
or pupae. The number of immature stages, even in nests containing some, 
was small (1-4) with the exception of three nests. These three contained 
eggs, small larvae, medium sized larvae, large larvae, prepupae, and pupae 
respectively, in the following numbers: (nest 421) 0, 0, 0, 3, 4, 15; (mest 443) 
Dy 2 oh, Sh JER aiinvel (Gnegesas)) Panam any lle 

The 31 nests of group 7 contained a total of 50 adults, one to seven per 
nest. Twenty-two, or nearly half, were unfertilized, the rest fertilized; 14 
showed some damage to the wing margins, the others showed no wear. Only 
five had enlarged ovaries; all five were fertilized but their distribution among 
the nests was not intelligible. Of the three nests containing eggs only one 
also contained a bee with enlarged ovaries. Probably the egg layers in the 
other two had died. The other four adults with enlarged ovaries were in 
nests with no young or only older larvae or pupae. Two were the only adults 
in their nests and two were in nests with another adult each. 

Apparently fertilized and unfertilized adults with slender ovaries were 
functioning as workers to feed larvae in most of the nests containing larvae. 

Group 8 consists of 13 nests taken at Cunningham’s Gap on February 25. 
One large larva, two dead prepupae, and one pupa were the only immature 
stages; these were distributed among three nests. Callow adults were present 
in another nest. Of the 17 adults, about 50 percent were fertilized, six showed 
worn wings, and none had enlarged ovaries (longest oocyte .25 mm.). 

Group 9 consists of 48 nests from Tamborine Mountain taken on March 
22. Immature stages were present in only five of these and callow adults in a 
sixth. Dead parasitized* prepupae were present in three others. The youngest 
immatures were large larvae, of which there was one each in three different 
nests. The total number of living immatures per nest was one in each of three 
nests and seven in each of two other nests. 

The 70 adults in the 48 nests were from one to six per nest. Twenty two 
were unfertilized; the rest fertilized; 28 showed worn wing margins, eight 
of them well worn or much worn; all had slender ovaries (longest oocyte 
0.13-.35 mm.). 

Summary of Discussion of Life History: The data for groups 8 and 9 


‘Parasitization was by chalcidoid wasps similar to those recorded by Rayment (1949b). 


oS) 
Ov 
Oo 


Lire Cycie AND SocIAL ORGANIZATION OF BEES 


above show that this species goes into winter primarily as fertilized adult 
females with slender ovaries. Young immature stages had disappeared by 
February 25 (late summer) and only a few older larvae and pupae remained. 
As shown in Table 4, about half of the adult females were worn, showing 
that many females that had been active earlier were now preparing for over- 
wintering. Most of the nests contained one or two adult females. Those with 
more than two often contained callows, showing that the larger number was 
usually due to a recently matured brood; probably such individuals would 
not remain together throughout the winter. 

In the spring, ovarian enlargement of some individuals occurs and eggs 
are laid in some nests in September and October. The presence of a few 
large immature stages suggests that a small number pass the winter in such 
stages rather than as adults. Reproduction goes on in September and October 
while several individuals are still in some nests, but group 2 (early Novem- 
ber) indicates that by that time nearly all females have dispersed from over- 
wintered groups and each has her own nest. Presumably each overwintered 
female ultimately develops enlarged ovaries and rears her offspring. 

Production of eggs and of young adults goes on throughout the summer 
as shown in the preceding section and in Table 4. New nests seem to be 
established throughout the summer by single females. Old nests often contain 
several females but the large groups are usually mostly young individuals, 
often partly callows. 

As shown in Table 4, production of immature stages is reduced by Feb- 
ruary 15 and many nests already lack immatures completely (see paragraphs 
above on group 7). Only 16 percent of the nests contained a female with 
enlarged ovaries (nearly all fertilized). However, the two nests containing 
several eggs and young larvae (443, 445) lacked egg layers; care of the young 
was by bees with slender ovaries, mostly unfertilized. Thus well before the 
end of summer the bees are progressing toward the overwintering condition. 

The Reproductive Cycle Within Nests: In order to better understand the 
summer part of the life cycle, the social interactions and reproductive cycle 
must be examined. Many nests, especially as summer advances, are difficult 
to interpret, but others, with greater or lesser certainty, can be placed in one 
or another of the groups discussed in the following paragraphs. Those that 
cannot be so placed are often difficult to interpret because of lack of knowl- 
edge of the age of the egg-laying female; one may not be able to decide 
whether she is a sister or the mother of her associates. Examples of nests 
of various types are given below; further details on the same nests can be 
obtained from Tables 1-3 or the commentary in preceding sections. 

Nest establishment and the egg layer with her first brood are exemplified 
as follows: Burrows without immature stages, usually containing only one 
adult and presumed to be young nests, were found throughout the summer. 


Tue Universiry SciENcE BULLETIN 


(PG Qa UM) (FD) (6) 7 
(1¢) + (9b)S (cQ)L (fp) L (0¢)9" 
= = OMe (60)¢ (16) UT 
($S)9° (9b) (¢g)¢° (62) (19)0'T 
(€8)0T  (79)9° (18)6° (LS)9° (16)7T 
(bp) (GS) a (Ss (67) F* (LI) 
(OODST (OOLET (OOL)9T (OOT)6! (COOT)EZ 
(OL) F (S1)Z (cg)7E (OOT)O'9 (82)0°6 
(9) 7 (@) Ve (0Z) ZL (fb)LZT (ZS) b'T 
Gye ae (SC) 9) (F1)T (OE)L 
(9) 1’ (is GOL; (€b)1'Z (OE)ST 
= a (OC bs (FI) I" (Og)¢° 
a eam) i (67) 6° (So)OT 
ae ae (ORS; (IZ)0'T  (0Z)0'F 
8b €1 Ig l €Z 
77-Ill C7-Il CI-II et L7-I1X 
6 8 l 9 ¢ 


(9) T CUR)S: Bei pe ae iat A ae in a) ae hele ae eS 2 2 py 
(IS)T'T (07) €" ae (LO) CEM gages pyro Sas e ey on ncaa IATIIYIO A\ 
(€8)6" (79) L (b9)L CONE ae a SE es ae SOUBAO passejpua yA, 
(0S)L° (£5) 8° (€L)L (CAS) |) Sie Mn sac: a ae cake cages in eg UIOA\ 
(98)6° CES) ical COON Oss COS) Pee” egies ae eee pezt eee 
(¢) 0 (91)S° ‘7 pe ee aioe. gered eas SMOT[LO 
(COOI)T'Z (OOT)O'T COOT)TT (OOT)6T 7 8) ore) amply 
(COUT MESS) VAI CSA) nS (ONS) CST EG Saaremaa maa SPINJBUILUT [PIO], 
(eh)rt (7E)07 rie Bay Wig ae ag A ce a ae aedng 
(F177 (€Z)¢° Ge ae avdndaig 
(LETT «(1 F) 01 ri a OI) eal Gan mG aC et ties cee Were avAIey ade] 
(€Z)E (02)S° Bre eid Cline aga ane er te arto ieee a  oeare] UINIpayy 
(TE) 8" (7£)6° (6) 9° (COGN Clr Vict an aa ail ers = ieere ayer eats IVAIR] [[LLUS 
(99)7E (OS)6O7T (H9)CE hae Wager Ral eee a eae Ge re ae s33q 

ce bh II 6 SISON JO ‘ON 
8I-IIx 8-HXx 6-IX €-X req 

b £ (l I ‘ON dnoiy 


*SOIeC] WsIIPAIC uo sasely [equowidoyaaaq SNOTIE A UT ISIN Jod syeNprAtpu] jo JToquun NY aISvIIAV “STPIQVIADA DANIUOXTT *f ATAV TL 


(aseis Yyoeo pouleqyuoy yey} ep yoed uO Ud YP} SJSoU JO 


sasequaoiad are sasayiuared uy) 


WwW 
oS) 
YI 


Lire CycLte AND SocIAL ORGANIZATION OF BEES 


Similar beginning nests, but with eggs already laid, were also found in each 
sample of nests studied. Other nests contained only eggs and young larvae; 
examples are nests 274 and 319B. Such nests usually contain a single fertilized 
adult female each; if she is continuing to lay eggs, her ovaries are enlarged. 
Usually several eggs are laid over a relatively short period, after which egg- 
laying by the mother ceases at least for a time, as shown by nests such as 276, 
278, 310B, and 297T in which the several young are large larvae or pupae but 
younger immatures are lacking. The mothers in such nests are fertilized, 
often with some wing wear as one would expect, and the ovaries usually 
somewhat enlarged. In other cases, however, the egg production seems to 
have been more protracted. Nest 287, for example, had young throughout 
the range from eggs to large larvae. The mother, as would be expected, was 
fertilized, with some wing wear, and with enlarged ovaries. The number of 
young in nests such as those described in this paragraph does not usually 
exceed seven or eight. 

Most egg layers probable survive to lay a second batch of eggs. Nests 
are rather common in which a group of older, immature individuals is pres- 
ent, and in which, in addition, there is a group of eggs or small larvae, 
young of intermediate ages being absent. When the egg-laying adult in such 
a nest shows considerable wing wear, I have interpreted the evidence as indi- 
cating resumption of egg laying by the same mother that produced the older 
offspring. An example is nest 270, in which the adult bee was much worn, 
fertilized, with much enlarged ovaries. The fact that, in addition to the seven 
pupae in the nest, there were 13 eggs may suggest that in addition to the 
old mother, another individual, possibly a sister of the pupal individuals, had 
contributed to the egg cluster. This seems unlikely because of the lack of 
callows or other young adults; this nest was taken at 5:00 a.m. when almost 
certainly all individuals living in it would have been present. Nests 286 and 
323 are other examples showing probable resumption of egg laying by a 
worn, fertilized female with large ovaries. 

Such nests can sometimes also be recognized at a later stage when some 
or all of the individuals of the first brood have reached adulthood. For ex- 
ample, in nest 273, in addition to eggs, there was an old, fertilized adult with 
well worn wings and enlarged ovaries and an unfertilized, unworn female 
with slender ovaries. The latter presumably represented the first brood of 
young; males and any other females of that brood must already have dis- 
persed. The eggs represented the second brood of young. The unworn 
female would probably have remained and functioned as a worker or might 
have mated and become reproductive. Nest 311T was the same except that 
the unmated individual with slender ovaries already had three nicks in one 
wing margin. Nest 288 presents a similar picture, six large immatures and 
two presumably young adults’ (unworn, unfertilized, with slender ovaries) 


*One of the young adults was teritological, lacking one front leg and one side of the 
prothorax. The mesonotum was so shaped that the bee probably could not fly. 


336 Tue University ScrENCE BULLETIN 


apparently representing the first brood and ten eggs representing the second 
brood of a worn, fertilized female with enlarged ovaries. 

Nests 314B and 316B are similar to those described above in having only 
young and old immature stages, without intermediate ages. Each also had 
4 worn, fertilized female with enlarged ovaries. In addition, each contained 
an unworn, fertilized female with enlarged ovaries. No doubt both indi- 
viduals contributed to the clusters of eggs, as indicated by the large number 
of eggs (20) in nest 314B. The 17 pupae and callow adults in the same nest 
suggest a double source for the earlier brood also. The sources of the unworn 
egg layers cannot be established but probably they are older sisters of the 
pupae and callow adults in these nests. 

Nest 315B also may contain one of its founders; at least there was a well- 
worn, fertilized female with large ovaries in the nest. The other adults in 
the nest were unworn, unfertilized, with slender ovaries, and no doubt were 
sisters of the callow adults and pupae. The large number of immature stages 
suggests that more than one egg layer was involved in their production. 

Nests such as 270 and 288 suggest that the second brood of a female may 
be larger than the first. This may be reasonable since there is often more than 
one female to care for young of the second brood, while the first brood ordi- 
narily must be cared for by the mother alone. 

There is evidence that the founder of a nest does not generally survive 
long after laying the eggs for her second bood. Only 24 well and much worn 
females were found in the course of the study; they become no more common 
as the season advances and are no more common in large colonies than in 
small ones. These facts argue against long survival of a colony-founding 
mother such as occurs in Lasioglossum malachurum (Noll, 1931) or & 
inconspicuum (Michener and Wille, 1961). Further support of this interpre- 
tation is provided by Skaife’s (1953) observations on Allodape and by the 
many nests containing immature stages, obviously being fed, yet lacking any 
individual that could have been their mother. Obviously care of young is 
taken over by other individuals, probably usually older sisters of the young 
concerned. The nests discussed below illustrate such phenomena. 

Nest 272, containing eggs and young larvae, was inhabited by two adult 
females, both unworn, one fertilized, one unknown because the spermatheca 
was lost in dissection, both with slender ovaries lacking white areas. Such 
ovaries could not have produced the eggs found in the nest. Presumably 
these adults were from the first brood of a female; her second brood con- 
sisted of the eggs and young larvae being cared for by the individuals of the 


: 
7 


first brood. This nest was taken at about 5:00 a.m., long before the bees be- | 


come active. Hence it is unlikely that any surviving inhabitants were away | 
from the nest. Nest 313T was similar in that it contained young immatures. | 


It also contained pupae and one unworn, unfertilized adult with slender — 


Lire Cycie AND SoctaL ORGANIZATION OF BEES 337 


ovaries; these must represent the first brood of a female no longer present, 
whose second brood was the eggs and young larvae. The young adult was 
obviously caring for the eggs and young in the nest. 

Nest 421 contained large larvae as well as prepupae and pupae. Of the 
seven adult females in the nest, only one was fertilized, three showed slight 
wing wear, and all had slender ovaries with no white areas. Obviously these 
adults were caring for the larvae although no one of them could have been 
their mother. 

Nests 443 and 445 both contained numerous immature individuals and 
several adults. One adult in each nest was fertilized, that in 443 slightly 
worn, but none of the adults had enlarged ovaries. Obviously one or more 
older bees, presumably dead by the date when the nest was taken, must have 
been the mothers of the immatures in each nest. Young bees were clearly 
caring for the eggs and larvae. 

Not only are young commonly cared for by their older, workerlike sisters, 
but some of the sisters also replace their mothers as egg layers, as indicated 
in the following paragraphs. Nest 295T is interesting in that it contained no 
immatures. The original egg layer must have failed to produce a second 
brood. The four bees in the nest were all unworn, only one was fertilized 
and she had enlarged ovaries. Presumably she would soon have laid a group 
of eggs. One of the unfertilized females had slightly enlarged ovaries (longest 
oocyte 0.55 mm.). 

It is my unverified suspicion that nests such as 275 (both adults with 
slightly worn wings, only one fertilized with enlarged ovaries) and 301B 
(one of the noncallow adults unworn, unfertilized, with slender ovaries; the 
other only slightly worn, fertilized, with enlarged ovaries) had lost their 
original founders and that each contained another egg layer which was pre- 
sumably one of the offspring of the founder. The principal reason for this 
surmise is the slight amount of wear of the wings of the individuals with 
enlarged ovaries. 

In other cases it seems clearer that the founder of a nest has died and been 
replaced by another egg layer. Nest 296B contained eggs, a pupa, callow 
adults, and one mature, unworn, fertilized adult with enlarged ovaries. It 
seems most likely that this individual is a sister of the pupa and callows but 
the mother of the eggs. Nest 293T was similar in that it contained young 
immature stages and young adults. Five of the adults were unworn, un- 
fertilized, with slender ovaries. The remaining adult differed in being 
fertilized, with enlarged ovaries. Presumably she was a sister of the other 
five adults and mother of the eggs and young larvae. Nest 463 can be in- 
terpreted along similar lines; the three adults were unworn, only one was 
fertilized with enlarged ovaries. 

Division of labor: From the above paragraphs it is evident that the typical 


338 Tue UNiversiry SCIENCE BULLETIN 


nest population contains only one fertilized adult female with enlarged 
ovaries. During the main reproductive period in summer (December), over 
half of the nests contain only one mature (i.e., not callow) female and at 
other seasons the percentage of monogynous nests is even higher. However, 
‘n some nests, additional mature females are present. Thus of 102 nests taken 
during December (groups 3-5), 48 contained more than one mature female. 
As shown in Table 5, 30 of these 48 nests contained two mature females and 


Taste 5. Exoneura variabilis. Frequencies of December Nests with Two or 
More Mature Females (out of total of 102 nests) and, among these, Frequencies 
of Nests with Varying Numbers of Reproductive Females. 


No. of mature @ 2 0 l 2 3 4 5 6 


Frequencies of nests with 
De OGMNOLEOITIACUTEs eG iy See ee z x 30 8 5 3 


bo 


Frequencies of such nests having 


Lo} 


varying nos. of 9 Q with enlarged ovaries. .. 4 29 12 


Frequencies of such nests having 


(oS) 


varying nos. of fertilized 9 Q. 72 2 28 15 


only small numbers had three or more. Table 5 also shows, as was suggested 
from less specific data at the beginning of this paragraph, that in most nests 
(28 or 29 out of 48) containing two or more mature females, only one 1s 
fertilized and only one has enlarged ovaries. In the great majority of cases 
the same individual shows both of these features and is obviously the only 
reproductive individual in the nest. However, among the 131 mature females 
in the 48 nests, 13 were fertilized but had slender ovaries. The 54 mono- 
gynous December nests contained 16 fertilized females with slender ovaries. 
This 27 to 30 percent fertilized individuals with slender ovaries consists of 
(1) individuals, probably relatively young, in their parental nests, sometimes 
associated with fully reproductive individuals, (2) individuals starting new 
nests (among the monogynous nests only), and (3) individuals that have laid 
their first brood of eggs and whose ovaries have regressed. There may be 
other conditions under which such individuals appear; from available data 
it is often difficult to place specific individuals in one or another of these 
classes. My impression is that classes (1) and (3) are about equally numerous 
and (2) less so. 

In addition to fertilized individuals with slender ovaries, there were 
among the 48 nests containing more than one female, three females with en- 
larged ovaries but no sperm cells in the spermathecae. One of these actually 
had only one oocyte enlarged (0.55 mm. long), so that its ovaries barely 
qualified as “enlarged”; it was in a nest (295T) containing two other young 
mature adults (unfertilized with slender ovaries) in addition to a fully re- | 


SN) 
Ne) 


Lire CycieE AND SOCIAL ORGANIZATION OF BEES 3 


productive female, and was undoubtedly only a young female which for some 
reason had an unusually enlarged oocyte. The other two unfertilized females 
with enlarged ovaries both had worn wings and the longest oocytes of each 
were 1.03 and 1.10 mm. long. Both had ovaries smaller than average for 
such large oocytes (0.25-0.30 mm. wide), as shown in Fig. 1, a fact which 
suggests that the ovarian enlargement may have been incomplete. One was 
in a nest (385) with a fully reproductive female, a callow female, and brood; 
the other was in a nest (302B) with two fully reproductive females and a 
female of the parasite Imguilina. Apparently such individuals are more or 
less in the nature of abnormalities. 

Among the December nests there were eight in which two of the females 
were fully reproductive. In several nests these were the only adults present 
while in other cases other mature adults were also present. 

Among the 131 mature females in the 48 December nests containing more 
than one mature female, 53 had enlarged ovaries and were fertilized. Such 
individuals are the principal reproductives and, as shown above, are usually 
distributed one per nest. At least many of the remaining individuals function 
as workers. This was established by capturing pollen collectors returning 
to their nests and dissecting them along with the other females in the nests. 
The pollen collectors were, in all cases studied, unfertilized bees with slender 
ovaries, and can be called workers. As examples, nest 332 contained a well 
worn reproductive, a worker, and two other individuals which seemed work- 
erlike and probably also served as workers. Probably the reproductive was the 
mother of the workers which were therefore functioning in the presence of 
a queen mother just as in many halictids and other social bees. Nest 329 
contained a reproductive, a worker, and two callows. Since all were unworn, 
they were probably all sisters. Nest 331 contained a reproductive and a 
worker. Because the reproductive’s wings were damaged in capture, it 1s 
not known whether she was an old bee or not. All the larvae in the nest were 
Inquilina; the reproductive evidently had not yet started to lay and was most 
likely a sister of the worker. Finally, in nest 330 there was no reproductive, 
the two females being unfertilized, with slender ovaries. At least one of 
them, however, was bringing pollen into the nest, presumably to feed larvae 
which probably were her sisters. Nests such as 330 substantiate the view 
already presented that young adults take over the work of caring for any 
young in the nest when the old adults die. 

The importance of workers in foraging in this and other species of 
Exoneura has been shown elsewhere (Michener, 1963) by dissections of 
pollen collecting individuals taken on flowers. At least at some seasons more 
than 80 percent of the pollen collectors were shown to be unfertilized and 
an even higher percentage have slender ovaries. The high percentage of 
workers collecting pollen indicates the inactivity of egg layers in such work, 
for less than 50 percent of the mature females from nearby nests were workers. 


340 Tue Universiry SciENcE BULLETIN 


ite) 

oo 

oO 

o 

Ww 

Te) 

(e) 

Te] 

ee 
@® 
—_— 
@® 
= 

o S 

i 
fe 
= 
5 

oO --= 

+ & 
wo 
oO 
@® 
pa FS 

Ww 

™ 

oO 

m 


© © o 


S|ONPIAIPU] yo saqunNn 


Fic. 2. Frequency distribution of head widths of females of Exoneura variabilis, based on 
70 egg layers taken in December, 70 workers and presumed workers taken in December, and 
70 females taken in March. 


Lire CycLeE AND SOCIAL ORGANIZATION OF BEES 341 


Since most workers are unworn, the problem arises as to whether workers 
die off as such or ultimately mate and disperse to establish new nests. The 
abundance of immature stages compared to the small number of adults 
(Table 1 to 3) suggests that the workers are short lived in this capacity—they 
either die after a short adult life or they disperse and cease to be workers. 
The number of new nests found in each group during the summer was 
small, as indicated in the discussion of nest groups 3 to 5. It is therefore my 
belief that most workers are relatively short lived, as in other social bees, and 
that they die before their wings become very worn. On the other hand the 
seemingly new nests sometimes found being established by worn bees may 
suggest that some workers do become reproductives later (see first paragraph 
in discussion of Group 4 above). The idea that the workers, or at least many 
of them, represent a caste rather than a stage in the adult life of females is 
further supported by the average size difference between workers and egg 
layers. Seventy fertilized individuals with large ovaries taken in December 
were compared with 70 unfertilized mature adults with slender ovaries taken 
in the same month and presumed to be workers (including 19 known to be 
workers from behavioral observations). Head width was used as a measure 
of size. The mean head width of the 70 egg layers was 60.9 .257 micrometer 
divisions, that of the 70 presumed workers was 58.1.101. These means are 
significantly different as shown by a ttest (p<.01). These data are shown 
in millimeters in Fig. 2. There was no significant or even suggestive size 
difference between lone egg layers and those in company with workers or 
other females. In the same figure is presented comparable data for 70 females 
taken on March 22. There was no significant difference between these 
females, prepared to overwinter, and the summer egg laying females. 

The difference in size between egg layers and workers is more clearly 
shown by comparing sizes in various nests separately. Among 33 nests taken 
in December and January in which egg layers and workers could be dis- 
tinguished with some degree of assurance, 24 had the egg layer larger than 
any worker in the same nest (Table 6). Moreover, 61 of the 71 workerlike 
individuals from these nests were smaller than the accompanying egg layers. 

Adults produced in December, as judged by callows, included a few 
individuals of maximum size but most were the size of workers; 21 of the 
24 callows taken in December with accompanying egg layers were smaller 
than those egg layers. In February, however, of six callows in three nests 
that contained egg layers, two were larger, two smaller, and two equal in 
size to those egg layers. These data, although few, suggest that at that time 
callows become overwintering egg layers rather than workers. 

Sex Ratio: The sex ratio may support the idea of a worker caste some- 
what similar to that of social halictines for there is an excess of females pro- 
duced. December and January pupae, callow adults, plus a few prepupae 


342 Tue University SciENcE BULLETIN 


Taste 6. Relation of Size of Workers (W) to Egg Layers in Various Nests of 
Exoneura vartabilis taken in December and January. 


Relation No. of nests No. of workers 


W <_— egg layer 24 54 
We =" ego layer 4 4 
Wo > ese layer 2 2 
W <>* egg layer l 2, 
We <<=* egg: layer ] 64 
W >=<*egg layer ] 3 


LoS) 
Qu 
NI 
— 


Totals 


* Multiple symbols indicate that some workers are smaller than, others larger than or equal to 
the egg layer. 
+ These six consisted of 5 smaller than and one equal to the egg layer. 


reared to the pupal stage for determination of sex, numbered 207. Of these 
only 40.1 percent were males. 

Males are generally short lived or leave the nests when young, for the 
number of noncallow males found in the nests is small (see Tables 1 to 3). 

Ovarian Development: The various stages of ovarian development among 
adults, from young callows to mature egg layers and workers and over- 
wintering individuals, are shown in Figures 65 to 82. The explanations of 
the illustrations are given in some detail and make an extended account here 
unnecessary. Drawings are based on specimens fixed in Kahle’s (Dietrich’s) 
solution. There are four ovarioles per ovary, and in each several oocytes 
enlarge simultaneously, thus permitting the laying of several eggs at about 
the same time. This is in sharp contrast to Adllodapula and Exoneurella (see 
Michener, 1962 and 1965). 

The white areas found in the posterior parts of some ovaries presumably 
result from resorption of eggs or from the laying of eggs or both and are 
therefore of value in giving some idea of previous ovarian activity. 

Immature Stages: The occurrence of immature stages in nests is given in 
detail in the section on Life History Data, including Tables 1 to 3. The de- 
tails of larval form and structure are given by Syed (1963), who showed 
that there are probably four larval stages. His drawings were from fixed 
specimens; Figures 3-13 were made from living material. 

The mesothoracic arms of the last stage larvae vary more than Syed in- 
dicated, as shown by Figures 7-37. These arms are sticky and soft in life. 
Branches may stick together or even to the body wall and then practically 
disappear. There is thus a good probability of observational errors in studies 
based on living material. Moreover, the arms are partially retractile; in in- 
active individuals they are not fully exerted. Tactile stimulation, for ex- 
ample with a needle, especially in the head region, results in full extension 
of the arms, as though by blood pressure; after a short time they retract. 


Lire Cycie aNp SoctaAL OrGANIZATION OF BEES 343 


Fics. 3-37. Exoneura variabilis. 3, egg (length 1.75 mm.); 4, small, probably second stage, 
larva (maximum diam. 0.58 mm.); 5, half grown (third stage) larva (maximum diam. 0.75 
mm.); 6, mature larva. Figures 7-37 represent maximally exerted mesothoracic arms of mature 
larvae before the prepupal stage. Figures 7-13 were drawn from live or freshly dead specimens. 
Figures 14-37, which tend to have the branches recurved, were drawn from specimens that had 
been fixed for weeks in Kahle’s (Dietrich’s) solution. Broken lines in some locations represent 
projections present on one side of the body, absent on other side of the same larva. Figures 
based on larvae from the same nests are 7 and 8, 9 and 10, 11 and 12, 17 to 19, 20 to 22, 
23 and 24, 25 and 26, 27 to 31, and 32 to 3/. 


Adult bees passing through the nest must cause this reaction which must be 
related to providing the pollen mass by the adults and its handling by the 
larger larvae. 

In the prepupal stage the larval appendages shrivel. 

The small projections on the mesothorax of third (?) stage larvae are 
not retractile. 

Specimens dropped alive into Kahle’s (Dietrich’s) solution have fully 
extended appendages which after a time tend to have recurved branches as 
shown in Figures 14-37. 

In the nests the immature stages are usually arranged from oldest above 


344 Tue University ScrENcE BULLETIN 


to youngest below. Perhaps because there are larger numbers of young of 
about the same age than in Allodapula (see Michener, 1962), the sequence is 
not so precisely kept as in that genus. For example, pupae of various degrees 
of pigmentation are usually mixed. The following lists illustrate selected 
cases in which marked irregularities in the usual sequence occurred. (In each 
list the nest number is given first and then the immature stages from top to 
bottom. Groups of immatures in clumps rather than in linear order are 
indicated in parentheses.) 


288: 2 large larvae, | prepupa, | white pupa, 2 black pupae (10 eggs). 

306B: 1 large larva, 4 pupae (11 eggs, 8 very small larvae). 

308B: 6 large larvae, 1 prepupa, 14 pupae, | large larva (1 egg, 1 small larva). 

310B: 3 large larvae, 3 prepupae, 1 white pupa. 

289T: 2 black pupae, 5 white pupae, 3 prepupae (1 small larva, 2 very small 
larvae, 8 eggs). 

316T: 2 large larvae, 2 prepupae, 3 pupae, | small larva, | pupa, 1 small larva, 
1 pupa (10 eggs). 

382: 10 large larvae (8 eggs, 1 small larva), 1 dead pupa, 1 pupa; 2 medium 
sized larvae, | egg. 


Feeding of the larvae is progressive, as in other members of the genus. At 
any one time relatively few larvae have food. For example, on December 18 
a series of nests was collected during the cloudy afternoon after a fine morn- 
ing during which the Exoneura were common on flowers. The nests were 
opened in the evening. At that time only eight of the 74 larvae in these nests 
had pollen masses on them. These eight included larvae of all sizes. 

I have two records of two small larvae curled around a single pollen mass 
but in general there is one mass per larva. The mass is on the ventral sur- 
face of the larva; the larva lies on its dorsal side and curls around the pollen 
as in the sketches reproduced in Figures 38-41. 


bye 
‘ 40s 


38 


Fics. 38-41. Exoneura variabilis \arvae feeding on pollen masses. 38, very small larvae feed- 
ing on pollen mass which has been largely consumed (maximum dimension of pollen mass | 
mm.); 39, third stage (?) larva; 40 and 41, last stage larvae. 


Eggs and small larvae are somewhat sticky and cling together in irregular 
clumps. The eggs are not attached in any way to the walls of the nest burrows. 

Larvae removed from nests produce feces from middle size onward. Feces 
are never found in the nests, however; they must be removed by the adults. 

Variation among adults: Rayment’s name for this species is appropriate 


Yt 


Lire Cycie ANpD SoctaL OrGANIZATION OF BEEs 34 


for there is more variability in coloration in this species than in any other 
Exoneura known to me. The face of the female is rarely wholly black. Com- 
monly there is a small cream colored mark on the lower part of each paraocu- 
lar area and sometimes these marks are rather large, nearly attaining the level 
of the summit of the clypeus. Commonly there is a longitudinal median 
clypeal mark, sometimes broadened at the summit to the full upper width 
of the clypeus. The clypeal mark may be present without the paraocular 
marks, or vice versa, or both may be present. 

The females in any one nest tend to resemble one another in the markings 
but there is enough variation within nests to show that all the variation 
described is intraspecific. 

There is also geographical variation. The abdomen, in specimens from 
Cunningham’s Gap, is orange. In specimens from the other localities it is 
considerably darker because of broad basal blackish bands on the segments, 
and sometimes it is almost wholly black. 


INQUILINA EXCAVATA (COCKERELL) 


This bee is similar in size and appearance to Exoneura variabilis and in- 
habits the nests of that Exoneura. As was pointed out by Michener (1961), 
the females of Inquilina do not possess a fully formed scopa. They have not 
been seen to collect pollen and appear to be social parasites in the nests of 
E. variabilis. This habit explains some of the “heterospecific companion- 
ship” discussed by Sakagami (1960). The details of the larval structure of 
I. excavata have been described and figured by Syed (1963). Variation in 
the mesothoracic arms of mature larvae is illustrated in Figures 42-49. 

No specimens of Inquilina were found with nest groups 1 and 2 of E. 
variabilis. Occurrence in December (groups 3 to 5) is shown in Tables 1 to 
3. In January (group 6) three nests (not included in previously presented 
statistics on E. variabilis) were found to contain only [nguilina. The contents 
were as follows (nest number at first of each line) : 


460: seven adult females. 

461: two adult females, six large larvae. 

464: five adult females, two adult males, one black pupa, one prepupa, seven 
large larvae. 


In another nest one female of Inguilina was associated with a female, two 
small larvae, and seven large larvae of E. variabilis. 

In February and March (groups 7 and 8) a total of eight /nquilina, all 
adult females, was taken, one in each of eight nests, and in every case in 
the company of one or more adult E. variabilis and sometimes with variabilis 
larvae. 

The relation of Inguilina to the Exoneura cannot be learned in detail from 


Tue University ScrENCE BULLETIN 


346 


"PAIv] JO JaquaA UO Apeare sseuT udT[od YM pue ‘MOrING Jo [[PM 0} payseNe UOLTOYD ur [As Apoq 

jo yred so1rajsod YW ‘easel payey AjMau ‘[T¢ ‘s3s39 Jo suontsod Sutmoys “sau JO pUd IOMOT ‘YG “MUuLdaID DANIUOXT “[C-)G ‘SOA 
‘JsoU IOYJOUL WOIF IAI] UO ‘s19YIO dy} “sau JUO UWIOIF AVAIL] UO Poseq 

aie ¢p pure Zp soinsty ‘suaumdeds paxy uo paseq ‘vvavrxd puyimbuy JO BAIR] IINeUT JO SLUR DIvIOYIOSIU JO SAaYNOYS “Gp-Zh “SA 


Lire Cycie AND SocIAL OrGANIZATION OF BEEs 347 


available data. It is obvious that the parasite must invade nests of Exoneura 
and lay eggs there. From the fact that whole groups of larvae and pupae 
in certain nests (302B, 303B, 304T, 331) were Inquilina, and from the fre- 
quency of adults in the nests, it seems clear that the Imqzulina take up resi- 
dence in association with Exoneura, rather than merely entering to lay an 
egg and then leaving. 

The larvae of Inguilina must obviously be fed by the Exoneura, since the 
parasite cannot collect pollen. The /ngzailina probably prevents the produc- 
tion of brood of the host in some way or destroys Exoneura eggs, in view of 
the seven nests (302B, 303B, 304T, 331, 460, 461, and 464) in which the whole 
broods seem to have been Inquilina. In some cases, e.g., 460, 461, and 464, the 
Exoneura appear to have reared their parasites and died, leaving the nest 
entirely to the parasite. In other cases, it seems that the Exoneura survive. 
For example, nest 331 was inhabited by two adult Exoneura, one of which 
was apparently about to start laying eggs. Perhaps such Exoneura resulted 
from larvae somewhat older than those of the parasites reared in the same 
nest; perhaps in the absence of Exoneura larvae of that age, the destruction 
of the Exoneura colony by the parasite is complete. 

The individual females of Imguilina isolated in nests of Exoneura were 
nearly all fertilized, not or slightly worn, with slender ovaries. Only two 
with enlarged ovaries were found; these were the adult females in nests 
302B and 381. Both were in nests with females of Exoneara which were also 
fertilized and with enlarged ovaries. From this it seems likely that egg-laying 
Inquilina live in nests where female Exoneura are actively laying eggs. The 
species of the eggs and young larvae in nests 302B and 381 is unknown, but 
it is reasonable to suppose that they were those of the [nquilina which proba- 
bly destroyed host eggs. 


EXONEURA HAMULATA COCKERELL 


Habitat: This species was studied at the nearby localities of Beerwah and 
Tibrogargen, in southeastern Queensland. The area is Eucalyptus savanna. 

Nests: Nests of this species have been recorded previously by Rayment 
(1946c, 1948). Twenty five nests of E. hamulata were found by me. All were 
burrows in standing dead, dry, flowering stems of Xanthorrhoea. Usually 
there was only one burrow in such a stem, but in one case two and in another 
case four burrows were parallel to one another in the pith, so close together 
as to be separated by only one half to one fourth of a millimeter of pith for 
considerable distances, yet never connecting. E. hamulata, like E. variabilis, 
makes its own burrows; it enters stems cut off by an insect which leaves a 
smoothly cut end. Stems broken or cut by man have rough ends and were 
not utilized. 

Nest burrows sufficiently advanced to contain young are 30-310 mm. deep 


348 Tue UNiversity SciENCE BULLETIN 


(mean of 21 measurements, 164 mm.). The entrances of the burrows are 
narrowed to 2.0-2.5 mm. (mean of six measurements, 2.4 mm.). 

The entrance may be narrowed by means of bits of pith forming a neat 
collar or the burrow may be excavated with a narrow entrance; perhaps all 
are first constructed in the latter way and a collar built only if the original 
entrance is damaged. The diameter of the burrow below the entrance is 
25-4 mm., usually being 3 mm. or more (mean of 39 measurements, 3.4 mm.). 

Life History and Reproductive Data: Information on the nest contents of 
the 21 nests that contained young is given in Table 7. These data suggest 


Taste 7. Numbers of Young and Adults in Nests of Exoneura hamulata. (Each 
horizontal row represents a nest. Nests without immature stages or callows and 
containing only a single adult were omitted from this table. Under adults, 


C=callows.) 

Nest Small Medium Large Total Adult Adult 
No. Date Eggs Larvae Larvae Larvae Prepupae Pupae Young je) 66 
88 X-17 a a 2 2 = - 4 2 

89 X-17 3 12 15 2 

90) X-17 7 x8 7 10 24 4 

92 X-17 a as 4 4 : 8 3 

245 XI-24 4 - 2. a = 1 5 2+5C 
253 XI-24 6 = 7 2 x 1 7 3+17C 6C 
400 1-6 3 2 = a n = 5 1+1C 
401 1-6 1 = 7 zs mi l 2 2C 

402 1-6 9 4 3 16 3 

403 1-6 10 2 a = = _ 12 1+3C 
404 1-6 9 - = ce 10 19 1+8C 2C 
405 1-6 4 1 2 6 2 15 5 

406 1-6 1 = A 4 2 

407 1-6 4 ie e s as 2 

408 1-6 = Es 4 | 2 3 be 
409 1-6 15 x. - s # 4 19 3-66 IC 
410 1-6 2 is es a 0 2+1C 
416 1-6 a a = 4 a mS 4 2 

419 1-6 es ae - x - 1 1 1 

420 1-6 I : 1 1+1C 
465 II-19 4 } | 7 2 


that nests are active more nearly continuously than in E. variabilis for nests 
full of immature stages were found both in October and February. October 
and February nests of variabilis do not contain such a variety of immature 
stages. A longer season for hamulata is not surprising in view of its warmer 
habitat, not far above sea level. Perhaps immature stages are numerous 
throughout the winter in these nests but adults were not taken on flowers 
in spring until mid-September nor in fall after April. Two lone females were 


Lire Cycie AND SoctIAL OrGANIZATION OF BEEs 349 


found making new nest burrows on November 24, two others on January 6. 
It is likely that new nests may be established at various seasons, as in E 
variabilis. None of the nests containing immature stages was recently estab- 
lished by a lone female, however; all contained older young and most con- 
tained more than one mature female. 

Obviously, as in E. variabilis, females lay considerable numbers of eggs 
over a short period and then rest, probably laying a second group of eggs 
later. The result is the normal occurrence of young of two very different ages 
(e.g., eggs and pupae, as in nest 404) in the nests. Occasionally (nest 405) 
a more or less continuous sequence of young exists in a nest, perhaps because 
of the presence of more than one egg laying female. 

The rather numerous callow adults in the nests (see Table 7) no doubt 
originated from the same groups of eggs as the pupae usually found in the 
same nests. 

Division of Labor: Consideration of mature adults in the nests suggests 
the existence of the same sort of colony organization at least from October 
through January or February as exists during December for E. vartabilis. 
Nests with a single mature female exist (e.g., 400, 403, 404). All show a major 
gap between young and older immature stages. The adult females in nests 
400, 403, and 404 were well worn, fertilized, with enlarged ovaries, as might 
be expected of females laying their second groups of eggs. 

Nests with several mature females are more common than in E. variabilis, 
as shown in Table 7; 15 of the nests contained two to five mature females 
each. The features of the mature females in these 15 nests are shown in 
Table 8. One or more fertilized bees were present in each nest. One fully 
reproductive individual (fertilized and with large ovaries) was found in 
each of six polygynous nests (90, 406, 408, 409, 416, and 465) and two in five 
nests (245, 253, 402, 407, and 410). Most of the remaining bees were (1) un- 
worn, fertilized individuals, not yet laying eggs (no white areas in ovaries), 
but caring for larvae which could not have been their progeny (e.g., all 
females in nests 88 and 92); or (2) unfertilized individuals, either worn or 
not, almost all of which had slender ovaries. These are presumably workers, 
except perhaps for a few which would later mate and become egg layers. The 
presence of worn workers (e.g., the last bee listed in nest 90, Table 8) sup- 
ports the view of workers as a distinct caste rather than as a phase in the life 
cycle, as elaborated in the discussion of E. variabilis. Wing wear is much 
more evident in larger bees than in smaller; it is therefore not surprising to 
find it more useful in studies of the large species, hamudlata, than in variabilis. 
Workers are relatively abundant in £. hamulata. 

The size of workers in E. hamulata probably averages smaller than that 
of egg layers. Workers in four nests were smaller than the associated egg 
layers, in a fifth, the reverse was true. Most of the callows were also smaller 
than associated egg layers. 


350 Tue University ScteENcE BULLETIN 


Taste 8. Data on Mature Females in Nests of Exoneura hamulata Containing 
Two or More Such Females. (Degrees of wing wear are from unworn, —, to 


very much worn, +-++-++-++.) 


Nest Worn Enlarged Nest Worn Enlarged 
No. Fertilized Wings Ovaries No. Fertilized Wings Ovaries 
88 tr = _- 405 = = we 
+ = — + fee 
89 — — — -— _— — 
+ — — ? — = 
90 — -- _ — —_ — 
ek 406 + = = 
= = = + = + 
a 407 + + : 
92 + = = “: — 4 
4 — — 408 — — — 
+ = = a 
245 of —- -} — — — 
=F SP4ES= =F 409 = ar = 
253 Be = av = a a 
- tet + + + 
“ a = 410 a Fier a 
402 at = — =F = =F 
Le — 4. 416 — — — 
+ — + + = + 
465 ? -- — 
+ — = 


Sex Ratio: Of 93 pupae and callow adults, only 42 percent were males. 
This is in almost exact agreement with variabilis. 

Ovarian Development: No differences in ovarian development were 
noted between E. hamulata and E. variabilis. 

Immature Stages: The details of larval structure were given by Syed 
(1963). The figures presented in the present paper (Figures 52-56) were 
based on living specimens. The arrangement of the stages in the nests is 
similar to that of EF. variabilis, perhaps less regular. 


EXONEURA OBSCURIPES MICHENER 


Six nests of this species were found, all in Lamington National Park near 
Binna Burra in the McPherson Range, Southeastern Queensland. They were 
in rotting, broken stems of vines in the rainforest, three of the six in a single 
broken stem. Nests containing immature stages ranged from 55 to 125 mm. 
in depth; diameters were 3 to 3.5 mm., with entrances narrowed to 2 mm. 
Eggs, small larvae, and large larvae were found both on December 8 and 
January 11. A new nest was being excavated by a lone, fertilized, unworn 
female on December 8. 


ten “MOYS JOU SIOP soptoqny jeuruopqe fo MoT [jeuus Jouut 94} 6¢ pue $S SOINGIT ul 


‘payasxa lou uoyM sasepuadde jo syisua_ MoYs saul] psNop ssurmesp ssoyy Ul “JapjO 6G ‘wMIpeIs Ise] UT SAIL] “GG PUL QC 
‘(Burp ays oF no paypnd sea W asojoq [[eys 389 ut [pHs sem Apoq jo uonsod sear) ware] [peuis */¢ “sadisnIsqo vANaUOXY “6C-LG ‘SOI 


“PAIR] DINILUT JO WIE DOeIOYIOSauT ‘gg ‘paytuo siey ‘edndaid ‘cg feArey sIMyeuI “pC + (Suryooys oF yno poijnd 
SPM UW a10Jaq [Jays 88a ur [jas sem Apoq Jo qed Ivar) ware] [peus ‘Eg f(t CT*Z YIsua_) 839 ‘7C ‘wIYJNWDY vINIUOXY “9C-TG ‘SOI 


ES 


9S 


Lire CycLE AND SOCIAL ORGANIZATION OF BEEs 


352 Tue University ScIENCE BULLETIN 


Nests with one, two, and three mature adults were found. The one with 
three contained an egg layer and two unfertilized, unworn probable workers. 
In each of the two nests with two adults, one was an egg layer, the other a 
fertilized, worn bee with slender ovaries. 

Larval structure was described by Syed (1963). The illustrations in- 
cluded herein (Figures 57-59) are based on sketches made from living 
specimens. 


EXONEURA ANGOPHORAE COCKERELL 


Only two nests of this species were found, both at Tibrogargen, in dead 
flowering stems of Xanthorrhoea, like those occupied by FE. hamulata. One 
was taken on November 24, the other on February 19. These nests were 62 
and 94 mm. deep with diameters varying from 3.5 to 5 mm., the entrance of 
one (for other not recorded) narrowed to 1.5 mm. Each nest contained eggs 
and small larvae but no other immature stages. Each contained more than 
one adult, however, suggesting that a previous brood of young had been 
reared in the nest. 

The nest taken on November 24 contained (1) a fertilized, worn female 
with enlarged ovaries, obviously the egg layer; (2) two unworn bees with 
slender ovaries, one fertilized, the other not, probably both acting as workers; 
and (3) three mature males. The nest taken on February 19 contained an 
unworn egg layer (fertilized, enlarged ovaries) and a worn individual with 
slender ovaries probably acting as a worker although fertilized. 


EXONEURA HACKERI COCKERELL 


Two nests of this species (which is apparently the same as imsularis 
Cockerell) were found, both in stems in the rainforest margin in Lamington 
National Park near Binna Burra, southeastern Queensland. One entered the 
stem through a beetle hole in the side of the stem instead of through a broken 
end. Lengths of the nests were 63 and 210 mm., diameters were 2.5 to 3 mm. 

One of the nests contained 12 eggs, one pupa, and two worn, fertilized 
females, neither with greatly enlarged ovaries. The other contained ten large 
larvae and six mature females, four fertilized, one unfertilized, and one doubt- 
ful. Two of the fertilized individuals were worn, the other adults were un- 
worn. One worn and one unworn fertilized female had enlarged ovaries. 


EXONEURA ATERRIMA COCKERELL 


A single nest of this species was found at Tibrogargen on November 24, 
1958 in a dead flowering stem of Xanthorrhoea. The entrance was through a 
hole (made by a beetle?) in the side of the stem rather than through a 
broken end. The burrow was only 19 mm. long, 3 mm. in diameter, narrowed 


oN) 
Wt 
(OS) 


Lire CycLrE AND SOCIAL ORGANIZATION OF BEEs 


to 1.5 mm. at the entrance. It contained four eggs and one very young larva 
with the rear half still in the egg shell but with a pollen mass on its under 
surface (Figure 51). 

An unusual feature, differing from all other Exoneura studied by me but 
resembling certain species studied by Rayment (1948, 1951) and Erickson and 
Rayment (1951), was the attachment of the eggs, usually by their posterior 
ends, to the burrow wall near the end of the burrow (Figure 50). The eggs 
therefore do not lie loose and form a clump as in other species dealt with in 
this paper. The chorion of the one egg that had hatched kept this position 
and held the young larva in the position shown in Figure 51. 


EXONEURA SUBBACULIFERA RAYMENT 


Five nests were found in dead, cut stubs of Rabus stems at Cunningham’s 
Gap on November 9. The nests ranged from 60-148 mm. in depth and from 
2.5-3 mm. in diameter. Some of them were being deepened (as shown by pith 
fragments thrown out of the entrances) in spite of the presence of brood. 

Like nests of E. variabilis found on the same date, those of sabbaculifera 
contained eggs, one of them also with small and medium sized larvae. The 
nests contained two to four females each, all but one fertilized, mostly worn, 
and one or two in each nest with enlarged oocytes. The one unfertilized 
individual is of interest because it was well worn with slender ovaries; it 
was probably a worker or workerlike individual that had survived the winter. 

The larvae were described and figured by Syed (1963) and are note- 
worthy for the large frontal projection, also shown in Figures 62 and 63 
based upon fresh material. Like the ventrolateral projections of this and 
other Exoneura, the frontal projection can be exserted; all are shown in the 
exserted position in the figures but when withdrawn the apical part of the 
frontal projection is reduced in size and less turgid and the abdominal ap- 
pendages are all mere convexities, like those shown on the eighth and ninth 
abdominal segments in Figure 62. 


EXONEURA BACULIFERA COCKERELL 


Two nests were found in broken stems at the edge of the rainforest in 
Lamington National Park near Binna Burra. 

The nests were 30 and 260 mm. deep, 2.75 to 4 mm. in diameter, the con- 
striction at the entrance of one of them had a diameter of 1.75 mm. 

The shallow nest was apparently new, containing a single fertilized 
female with enlarged ovaries and four eggs. The deep nest contained six 
mature females, five callow females, 13 males, six pupae, and three eggs. 
Two of the mature females were fertilized, worn, with enlarged ovaries. The 
other four were unfertilized, unworn (except for one nick in one wing), three 


“Cu (ZT Jojowerp winuitxew) p4af7novg vinauoxg yo edndaig ‘$9 ‘914 
“ures JO XeJOY JO IUOIJ PUL Pedy JO MATA [esIOp ‘Eg f("WIW YQ'Q YIPIM pray) eaAre| 
aimeul “7g ("Ww 99°) YIP peasy) eAie] (2) a3vys psryy Jo peay ‘Tg $ (WU 8L1 ysus]) 399 ‘99 ‘wuafynovggns vinauoxy *¢9-1)9 “sDIq 


UD» aa ‘ 


v9 9 19 09 


Tue University Science BuLLETIN 


354 


V1 


Lire Cycle AND SoOcIAL ORGANIZATION OF BEEs 35 


with slender ovaries, the fourth, surprisingly, with enlarged ovaries. 
The prepupa of this species is shown in Figure 64. The frontal projection 
is similar to that of EF. subbacultfera. 


SUMMARY 


This paper consists mainly of an account of the life cycle and social or- 
ganization of Exoneura variabilis, with less complete data on several other 
species of Exoneura and on the socially parasitic bee, Inguilina excavata. The 
bulk of the information was derived from statistical examinations of nest 
populations, spermathecal content, wing wear, number and stages of the im- 
mature forms, and the like, rather than from direct observation of behavior 
in the nests. 

E. vartabilis nests in simple burrows, not divided into cells, in dry pithy 
stems in rainforest boarder regions. It overwinters in such burrows, primarily 
as fertilized adult females of varying ages (or amounts of wing wear). One 
to three females inhabit a single nest. A few larvae apparently also over- 
winter. 

In spring the bees disperse so that most nests contain one female. Egg 
laying occurs throughout spring and summer and adults of both sexes are 
produced. New nests are established throughout the active season by lone 
females. However, apparently the majority of the young adult females do not 
leave the parental nest but remain there. Some replace the original egg laying 
female who dies soon after laying two, sometimes incompletely separated, 
batches of eggs. Since the larvae are fed progressively, those hatching from 
the second batch are usually reared by one or more of their adult sisters from 
the first batch of eggs. Such an individual may coexist for a time with the 
worn mother, but in many nests larvae are cared for in the absence of any 
individual that could be their mother. Commonly such sisters are workerlike, 
having slender ovaries and being unmated, but others mate and replace their 
mothers as egg layers. 

At all seasons over half the nests opened contained only one mature female 
(recently emerged adult females are not counted). However, especially in 
summer, some nests (e.g., 48 out of 102 in December, the first summer 
month) contained more than one mature female, usually two but rarely as 
many as five or six. Usually only one but sometimes more are egg layers; the 
rest have slender ovaries and are often unfertilized workers; they were seen 
to be active in collecting pollen for nests containing egg layers which did not 
leave during the period of observation. (Rarely unfertilized individuals have 
enlarged ovaries.) Workers average smaller in size than egg layers and are 
probably short lived compared to egg layers. 

A worker can be either daughter or (usually) sister of an egg layer in 
her nest. Female bees care for their younger sisters, their own progeny, and 


356 THe University SciENCE BULLETIN 


progeny of their sisters. Both workers and egg layers care for whatever young 
are in the nest, although egg layers in nests that also contain workers do 
little or no foraging. 

About 40 percent of the young produced are males. Adult males do not 
remain long in the nests, and are probably short lived. 

Ovarian development was studied and illustrated. In contrast to halictids, 
conspicuous white masses are formed, presumably indicating both resorption 
of eggs and prior laying. 


Fic. 65. Ovaries of soft, pale, young callow. Note that oocytes at the lower ends of the 
ovarioles are scarcely enlarged. 

Fics. 66 ano 67. Ovaries of callows. Note enlarged oocytes and well formed groups of 
nurse cells; in the right ovary of figure 66 each of the ovarioles has a second enlarging oocyte. 
There are no white bodies indicating oosorption or remains of nurse cells. 

Fics. 68-72. Ovaries of overwintering bees taken from nests and fixed on March 22, 1959. 
Figures 68 and. 69 were from unfertilized bees with unworn wings, 70 and 71 from fertilized 
bees with unworn wings, and 72 from a fertilized bee with wo-n wing margins. Figures 68, 71, 
and 72 show white areas (dark in figures) in one or both of the lateral oviducts indicating 
oosorption. Figures 69 and 70 seem to show an anterior retreat of the ovarioles as compared 
to the callows. This is common in overwintering bees. There is commonly only one distinct 
enlarged oocyte per ovariole but there may be none (some ovarioles, figs. 67 and 68) or as 
many as three (right side, fig. 68). The bee that had worn wings and was therefore presumably 
older than the others had larger ovaries, but none of the oocytes was very large (Fig. 72). 
Apparently at this season oocytes of only moderate size are resorbed. 


Fics. 73-78. Ovaries of some adults taken on December 26, 1959. Figures 73 (fertilized) 
and 74 (unfertilized) are based on bees with worn wings. They might have been either workers, 
or egg layers for some reason not at the moment laying eggs, and therefore with rather slender 
ovaries. The left hand ovary in figure 74 evidently produced a large oocyte, now being resorbed 
(lightly stippled). White areas (dark in figures) below the ovarioles indicate oosorption or 
remains of nurse cells. Figures 75 to 78 represent bees apparently in or approaching egg laying 
condition. Figure 75 shows no white areas below the ovarioles; this bee probably never laid 
an egg. Figures 77 and 78 (left side of each) show the break down of groups of nurse cells 
above fully formed eggs. Similar white masses (dark in figures) below the eggs probably are 
of this origin and indicate prior egg laying. 

Fics. 79-82. Ovaries of unfertilized workers taken on December 26, 1959. Often such 
ovaries are flattened against the wall of the crop so that all four ovarioles are in a single plane 
and visible from one view, as shown on one or both sides of figures 80 to 82. The lack of white 
areas suggests that these bees never laid eggs. Narrow apical oocytes such as are shown in 
figure 82 are very rare. 


358 Tue University SciENCE BULLETIN 


Immature stages are kept more or less in age groups in the nest. Food 
masses are placed on the ventral surfaces of the larvae. Occasionally two small 
larvae may curl around a single food mass. 

Life cycles and social organization of other Exoneura species seem similar 
to those of E. variabilis. In E. hamulata the colonies average larger and the 
active season longer. In E. aterrima the eggs, instead of being placed to- 
gether in a batch at the bottom of the nest, are stuck by their posterior ends 
to the walls of the nest and the young larvae retain this position and are fed 
there. 

Inquilina excavata is a social parasite in the nests of Exoneura variabilis 
and is closely related to Exoneura. Inquilina cannot collect pollen. Its 
females invade nests of E. variabilis and take up residence there in association 
with the Exoneura. Apparently the Inquilina prevent the laying or survival 
of Exoneura eggs in some way, for production seems to go wholly to Jn- 
quilina, but adult Exoneura (workers) must be present to bring food for the 
parasite. The only two Inquilina found with enlarged ovaries, ready to lay 
eggs, were in Exoneura nests containing a female Exoneura having similar 
ovaries. Perhaps the Inquilina synchronize their laying with that of the 
Exoneura in order to fit into the social system of the latter. Ultimately the 
Exoneura die off, leaving the nest to the by now maturing [nquilina, or 
some Exoneura survive and probably may reestablish the Exoneura colony. 


LITERATURE sei») 


Erickson, R., anpD T. RayMent. 1951. Simple social bees of Western Australia. Western 
Australian Nat. 3:54-59. 

MicHeNner, C. D. 1961. A new parasitic genus of Ceratinini from Australia (Hymenoptera, 

Apoides). Jour. Kansas Ent. Soc. 34:178-180. 

. 1962. Biological observations on primitively social bees of the genus ‘“‘Allodapula” in 

the Australian Region (Hymenoptera, Xylocopinae). Insectes Sociaux 9:355-373. 

. 1963. Division of labor among primitively social bees. Science 141:434-435. 

———. 1965. The bionomics of Exoneurella, a solitary relative of Exoneura (Hymenoptera, 
Apoidea, Ceratinini), Pacific Insects, 6:411-426. 

MIcHENER, C. D., E. A. Cross, H. V. Daty, C. W. RETTENMEYER, AND A. WILLE. 1955. Addi- 
itional techniques for studying the behavior of wild bees. Insectes Sociaux 2:237-246. 

MicHener, C. D., ano A. Witte. 1961. The bionomics of a primitively social bee, Lasioglos- 
sum inconspicuum. Univ. Kansas Sci. Bull. 42:1123-1202. 

Nott, J. 1931. Untersuchungen uber die Zeugung und Staatenbildung des Halictus malachurus 
Kirby. Zeit. f. Morph. u. Okologie d. Tiere 23:285-367, pls. I-III. 

RayMentT, T. 1935. A cluster of bees. 752 pp., Endeavor Press, Sydney. 

. 1946a. New bees and wasps—Part I. Victorian Nat. 62:178-184. 

—. 1946b. New bees and wasps—Part II. Victorian Nat. 62:230-236. 

— . 1946c. New bees and wasps—Part III. Victorian Nat. 63:63-68. 

—. 1948. New bees and wasps—Part VII. Victorian Nat. 65:85-91. 

—. 1949a. New bees and wasps—Part VIII. Victorian Nat. 65:208-212. 

——. 1949b. New bees and wasps—Part IX. Victorian Nat. 65:247-254. 

——. 1951. Biology of the reed-bees. Australian Zoologist 11:285-313, pls. 27-32. 

——. 1954. New bees and wasps—Part XXII. Victorian Nat. 71:13-16. 

. 1956. New bees and wasps—Part XXV. Victorian Nat. 72:173-174. 

SakacAMI, S. F. 1960. Ethological peculiarities of the primitive social bees, Allodape Lepeltier 
[sic] and allied genera. Insectes Sociaux 7:231-249. 

SkxalFE, S. H. 1953. Subsocial bees of the genus Allodape Lep. & Serv. Jour. Ent. Soc. S. Africa 
16:3-16. 

Syep, I. H. 1963. Comparative studies of larvae of Australian ceratinine bees (Hymenoptera, 
Apoidea). Univ. Kansas Sci. Bull. 44:263-280. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


THE LIFE CYCLE AND BEHAVIOR 
OF THE PRIMITIVELY SOCIAL BEE, 
LASIOGLOSSUM ZEPHYRUM 
(HALICTIDAE) 


By 
Suzanne W. 'T. Batra 


VoL. XLVI Paces 359-423 Fepruary 1, 1966 No. 10 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Voi. XLVI Paces 359-423 Fesruary |, 1966 No. 10 


The Life Cycle and Behavior of the Primitively Social Bee, 
Lasioglossum Zephyrum (Halictidae)* 


By 
SUZANNE W. T. Batra 


A 


ABSTRACT 


Lasioglossum zephyrum is a small halictine bee, widely distributed in the 
United States. In Douglas County, Kansas, aggregations of nests were found in 
exposed, south-facing banks of streams or in horizontal ground near such banks. 
Males and females from a population on Tauy Creek had relatively larger heads 
than those a few miles away along the Wakarusa River. Factors influencing nest 
distribution in suitable habitats may include visual and chemical attraction to 
areas where nests have been established. The high rate of mortality of colonies 
throughout the season apparently was not due to any of the obvious parasites 
common at the nesting sites. 

Males and females produced in fall were larger than those emerging in sum- 
mer, perhaps because the oviposition rate declines in late summer, permitting each 
cell to be amply provisioned before egg laying. Castes are not distinct. The over- 
wintered nest-founding egg-layers are often replaced by their daughters during 
summer. There is a continuum from queenlike to workerlike females in each 
nest and there are no discontinuous morphological or behavioral differences be- 
tween egg-layers and females that lay few or no eggs. Unfertilized females with 
enlarged ovaries are common during summer and probably lay male-producing 
eggs. Foragers usually have somewhat enlarged ovaries and are polylectic. They 
were induced to collect pollen from a petri dish. Guards defend the nests against 
some parasites. Burrows are often tortuous with many branches. The subhori- 
zontal cells are arranged along main burrow and branches, each cell being pro- 
vided with one pollen ball and one egg before being plugged with soil. Insem1- 
nated females are the only bees that overwinter. They hibernate in short branch 
burrows of the previous summer’s nest. In April most of them begin new nests. 
Each female makes, provisions, and oviposits in about 4 cells, then rests in the 
nest until her progeny emerge. The young females stay in the nest and deepen it 


1. Contribution number 1245 from the Department of Entomology, The University of 
Kansas, Lawrence. 


360 Tue UNiversity SclIENCE BULLETIN 


as reproduction continues throughout the summer. By August, nests may reach 
depths as great as 57 cm, with as many as 38 branch burrows, 85 occupied cells 
and 45 females. The number of males gradually increases during summer. They 
leave the nests and swarm over the nesting areas, where they mate with females 
of various ages. 


INTRODUCTION 


Halictine bees or sweat bees are of interest to many biologists because 
various levels of social organization occur within the subfamily. Even with- 
in Dialictus, a subgenus of Lasioglossum (to which L. zephyrum belongs), 
are found bees with solitary life cycles, such as L. herbstiellum (Claude- 
Joseph, 1926) and L. opacum (Michener and Lange, 1958); social bees with 
poorly defined castes, such as L. rhytidophorum (Michener and Lange, 
1958); and bees with fairly distinct castes, such as L. imitatum” (Michener 
and Wille, 1961). A comprehensive review of the literature on life cycles of 
halictine bees was given by Plateaux-Quénu (1959), and nest structures and 
some aspects of the social behavior of many halictines were discussed in de- 
tail by Sakagami and Michener (1962), with an extensive bibliography. 

In this paper, the life cycle and behavior of a primitively social, small (5-7 
mm), greenish-black sweat bee are discussed. This study is part of an in- 
vestigation by C. D. Michener and his associates on comparative halictine 
behavior, intended to elucidate the origin and evolution of social behavior. 

Lasioglossum zephyrum (Smith) is widely distributed in the United 
States (Michener, 72 Muesebeck, et al., 1951). A description of this species, 
as Dialictus zephyrus, is in Mitchell’s (1960) monograph on bees of the 
eastern United States. 

Rau (1922, 1926), Robertson (1926), Krombetn (1938), Sakagami and 
Michener (1962), Knerer and Atwood (1962), LaBerge and Isakson (1963), 
and Lin (1964) all noted nests of this species, mostly in aggregations in banks 
of streams or lakes; but sometimes in flat ground adjacent to such banks. 
The nests recorded by Krombein and by Lin were seemingly in flat ground 
not associated with any bank. The authors listed above recorded the general 
structure of nests and also the swarming of males (“sun-dances” of Rau) 
near the nesting places. They also recorded several females in some nests; 
Rau (1926) found 14 in one nest. 

The flight period has been shown to extend from March 21 to November 
in southern Illinois (Robertson, 1926) and from April to September in 
southern Ontario (Knerer and Atwood, 1962). 

Pearson (1933) found that L. zephyrum was a widespread and common 
bee, relatively abundant in moist dune groups and on flood plains near Chi- 
cago. Lasioglossum zephyrum was relatively more sensitive to dryness than 


2. Referred to as L. inconspicuum by Michener and Wille (1961). The synonymy was 
established by Knerer and Atwood (1962). 


2 


Tue Lire CycLteE AND BEHAVIOR OF THE PRIMITIVELY SoctAL BEE 361 


most bees experimentally tested. He thought that a function of the nest 
guard may be to keep moisture in the nest on warm, dry days. 

Robertson (1928) recorded L. zephyrum on the following plants (m= 
male, f=female, c=female collecting pollen) : 


Anacardiaceae 

Rhus canadensis 

Rhus copallina (£) 

Rhus glabra (m, f, c) 
Apocynaceae 

Apocynum androsaenufolium (f) 
Asclepiadaceae 

Gonolobus laevis 
Bignoniaceae 

Tecoma radicans (f, c; also f, m, on 

extra-floral nectaries) 

Caprifoliaceae 

Sambucus canadensis (c) 

Symphoricarpus orbiculatus (m, f) 
Celastraceae 

Euonymus atropurpureus 
Compositae 

Antennaria plantaginifolia (£) 

Aster ericoides (m, f) 

Aster paniculatus (£) 

Cacalia reniformis (£) 

Erigeron ramosus (m) 

Eupatorium altissimum (m) 

Eupatorium serotinum (m) 

Helianthus tuberosus (c) 

Rudbeckia subtomentosa (£) 

Rudbeckia triloba (£, m) 
Cornaceae 

Cornus florida 
Cruciferae 

Arabis dentata 

Arabis virginica 

Capsella bursa-pastoris (f, c) 

Radicula obtusa 

Radicula palustris 
Ebenaceae 

Diospyros virginiana (Ef, c) 
Hydrophyllaceae 

Ellisia nyctelea 
Labiateae 

Blephilia ciliata 

Pycnanthmum pilosum (f£) 
Lauraceae 

Sassafras varufolium (f, c) 
Leguminosae 

Cercis canadensis 


Liliaceae 

Asparagus officinalis (f, c) 

Smilacina racemosa 

Smilacina hispida (c) 
Malvaceae 

Malva rotundifolia (m) 
Onagraceae 

Ludvigia polycarpa (£) 
Papaveraceae 

Sanguinaria canadensis 
Phytolaccaceae 

Phytolacca decandra (f) 
Portulacaceae 

Claytonia virginica 
Ranunculaceae 

Actea alba (c) 

Clematis virginiana (m, f, c) 

Isopyrum biternatum 

Myosurus minimus (£) 

Ranunculus abortivus 

Ranunculus septentrionalis 
Rhamnaceae 

Rhamnus lanceolata (f, c) 
Rosaceae 

Amelanchier canadensis 

Crataegus mollis (£, c) 

Prunus americana (f, c) 

Prunus serotina (Ef, c) 
Rutacaceae 

Zanthoxylum americanum 
Salicaceae 

Salix amygdaloides (f, c) 

Salix cordata (£) 

Salix humilis (£) 

Salix longifolia (m, f, c) 

Salix nigra (f, c) 
Polygonaceae 

Polygonum persicaria (£) 

Polygonum scandens (m) 
Saxifragaceae 

Philadelphus grandifloris cult. 

Ribes gracile 
Scrophulariaceae 

Scrophularia marilandica (m, f) 

Veronica perigrina 
Thymeleaceae 


362 Tue Universiry ScrENcCE BULLETIN 


Dirca palustris Taenidia integerrima 

Tilia americana (Ef, c) Zizia aurea 
Umbelliferae Verbenaceae 

Cryptotaenia canadensis (m, f) Verbena hastata 


Sium cicutaefolium (m, f) 


Mitchell (1960) and Knerer and Atwood (1962) recorded L. zephyrum 


from flowers of the following additional genera: 


Caprifoliaceae Scilla 
Virburnum Malvaceae 
Compositae Althaea 
Baccharis Nymphaeaceae 
Cirsium Castalia 
Helenium Polygonaceae 
Leucanthemum Fagopyrum 
Solidago Pontederiaceae 
Taraxacum Pontederia 
Cruciferae Rhamnaceae 
Barbarea Ceanothus 
Cucurbitaceae Rosaceae 
Cucurbita Aronia 
Leguminosae Malus 
Lotus Rubus 
Medicago Saxifragaceae 
Melilotus Hydrangea 
Vicia Symplocaceae 
Liliaceae Sym plocos 


Obviously this is a polylectic bee, as are most other species of Lasioglos- 
sum. 

MATERIALS AND METHODS 

Observations were made during 1957, 1959, 1961, 1962, and 1963 in Doug- 
las County, eastern Kansas. Data obtained in 1957 and 1959 were kindly 
supplied by Dr. C. D. Michener. The techniques described by Linsley, 
MacSwain and Smith (1952) and by Michener, et al. (1955) were used in 
the field, with a few modifications to be described in appropriate places 
below. Methods for rearing L. zephyrum in insectary rooms were described 
by Batra (1964). 

A total of 255 nests was excavated; 1,075 females were measured, dissected 
to determine reproductive condition, and examined for wear of wings and 
mandibles, and 741 males were examined for wear and measured. Many 
observations on living bees in the field and in four insectary rooms were 
made. A total of 152 days were spent in the field (107 of these days were 
during 1962 and 1963). 


GENERAL LIFE HISTORY 
A brief account is given here to provide background for subsequent sec- 


tions of the paper. 


> 


Tue Lire Cycie AND BEHAVIOR OF THE PRIMITIVELY SociAL BEE 363 


Relatively unworn, inseminated females with slender ovaries are the only 
bees that overwinter. They stay at the ends of short branch burrows di- 
verging from the main burrows of the previous summer. In early spring, at 
about the time Ulmus americana is in bloom (March 24, 25 and April 1),” 
females appear at the entrances of the nests, and begin flying during warm 
hours of the day. Some of these females stay in the old nests, but most begin 
to excavate new nests. They feed on nectar and pollen as their ovaries en- 
large. New nests were first seen on April 2, 13, and 14, and the bees were 
first seen carrying pollen to nests on April 17 and 21. Three or four cells per 
bee are made and mass-provisioned with a mixture of pollen and honey. An 
egg is laid on the provisions in each cell, which is then plugged with a little 
soil. The females then cease activity and wait in the nests while their brood 
matures. Bees had ceased collecting pollen on May 7, 12, and 13, although a 
few were still flying about after these dates. 

The new generation of males and females emerged from the nests on 
May 22 and 25. A few young pollen collectors were seen on these dates. 
These young females stay in the nests started by the overwintered females, 
enlarge them and make and provision cells. Some mate and many of them, 
whether inseminated or not, develop large oocytes. They probably lay eggs 
and apparently replace the overwintered females (no marked overwintered 
females were found after June 14). The nests remain continuously open and 
the bees active through the rest of the summer. Castes are not clearly de- 
fined in this species and all intergradations from queen to worker are found 
in nearly every nest. During summer, the relative proportion of males pro- 
duced gradually increases, and by August they form large swarms over the 
nesting areas. 

The last females collecting pollen were seen on September 13, 20, 27, and 
October 2. The last males were seen October 2, 12, and 10. Overwintering 
females were guarding nests as late as October 19, 23, and 29. 

Nests are usually excavated in banks of streams at about a 90° angle to 
the soil surface. The circular entrance is about 2 mm in diameter; the bur- 
row, usually branched, is 4 mm in diameter and lined with a thin layer of 
compressed soil. Cells are 9 to 11 mm in total length, 4 mm in maximum 
width, and except for the neck, which is about 2 mm long, are lined inside 
with a shiny waterproof film. Constructed serially along the main burrows, 
they are bilaterally symmetrical (sense Malyshev, 1935) and subhorizontal. 
As in other halictines, each cell contains one ball of moistened pollen and 
one egg. After oviposition, the neck of each cell is filled with loose soil. 

3. Dates in series such as these refer to different years in which meaningful data pertaining 


to initiation or cessation of specific activities were available. In some years, observations con- 
cerning these activities were not frequent enough and are therefore omitted. 


364 Tue Universiry SciENCE BULLETIN 


Usually cells deepest in the nest contain the youngest stages of the brood. 
Nests are in groups known as aggregations and the several females in each 
nest constitute a colony. 


SEPT. Oct 


APR. |MAY ! JUNE! JULY | AUG. 


MAR. 


Oe ey, lp yee ae 


— 


Degrees Centigrade 


Fic. 1. Mean air temperatures (measured at a height of 1 m, in shade, upper line), and mean 
soil temperatures [at depths of 10 cm (dotted line) and 21 cm (lower line)] taken at 
the County Line and Wakarusa nesting areas between 10 a.m. and 4 p.m. over a period 
of 3 years. The number of measurements for each mean is indicated at each point. 
Below the temperature graph is a chart indicating the first dates when bees were seen 
emerging from hibernacula (E), excavating nests (N), collecting pollen (P), resting in 
closed nests (R), and when the first brood emerged (B). The last dates for pollen col- 
lectors (P), males (M) and guards (G) are also indicated. 


Tue Lire CycLe AND BEHAVIOR OF THE PRIMITIVELY SoctAL BEE 365 


NESTING AREAS 


Tue ENVIRONMENT AND THE AGGREGATIONS 


Douglas County, where my observations were made, is in central eastern 
Kansas. Altitudes are 236 to 366 m above sea level. The climate is continen- 
tal. Nearly three-fourths of the annual precipitation falls during the growing 
season (196 days). The average date of the last killing frost in spring is 
April 10 and the average date of the first killing frost in fall is October 23 
(Fig. 1). The mean annual precipitation at Lawrence is 88 cm and the mean 
annual temperature is 14°C (O’Connor, 1960). Non-cultivated areas are now 
covered with trees and brush, although before the arrival of European man, 
most of the area was prairie. 

Lasioglossum zephyrum, in this area, usually nests in the bare, steep, 
periodically eroded silty-clay banks of permanent or semipermanent streams. 
Compact soil, exposed to the sun much of the day (at least in spring, when 
nests are founded), seems to be preferred. 

Most observations were made at two locations where nests were numer- 
ous, described below: 


1. County Line—This area, on the farm of Mr. Roy Stecher, at the Douglas-Franklin 
County line, is near Tauy Creek, a semipermanent tributary of the Marais des Cygnes River. 
When this location was discovered in 1957, the bees were nesting in a south-facing, bare bank 
of clayey sand about 15 m from the creek. A few nests were found in sloping or horizontal 
ground (a cowpath) at the base of the bank and some nests were in the temporarily dry creek 
bed. In 1959, nests were confined to the almost vertical bank (1.5 x 2 m), with very few in 
the horizontal ground nearby. By 1961, bees had deserted the bank, which was crumbling, over 
90 per cent of the nests being located on nearby flat ground in the cowpath; maximum density 
was 30 nests per square meter (maximum internest distance 44 cm; minimum internest distance 
2.5 cm). This flat area of 16 sq. m in front of the bank was partly covered with clumps of 
short grass and weeds. Nests were in bare patches of ground as well as in the vegetation and 
Lasioglossum imitatum was also nesting here. The bank and flat area, partly shaded by trees, 
were in sun during most of the afternoon and were subject to occasional flooding. In March, 
1962, the owner widened the path with a bulldozer, cutting into the compact clay soil of a 
southeast-facing slope about 20 m away, and clearing all vegetation from the cowpath. In 
1962, the bees colonized this new vertical bank (Fig. 2), digging 49 nests in its surface of 
roughly 1.4 x 4 meters. An additional 23 nests were made in a horizontal area of 3 x 4 
meters in front of this second bank (maximum internest distance 2 m; minimum internest 
distance 3 cm). These areas were in sun all day during spring but the bank was shaded by a 
small tree most of the day in summer. Twenty-five nests were started that year in the original 
bank, and 5 nests in the cowpath in front of it, but all of them had died out by the end of the 
summer. In 1963, the horizontal areas in front of both banks had no nests, and were partly 
covered with grass and weeds. No bees were nesting in the crumbling original bank, which was 
riddled with old abandoned burrows of past years. The new bank had only seven nests. 

2. Wakarusa River—in June, 1962, a large aggregation of L. zephyrum was found in a 
steep alluvial bank of this tributary of the Kansas River, 7.5 miles southwest of Lawrence. The 
nearly vertical bank, facing southwest, extended from the water level to a height of about six 
meters (Fig. 3). It was composed of compact silty light brown clay previously deposited by 
the river. No vegetation was on its surface, but brush and a black walnut tree (Juglans nigra) 


366 Tue University SciENcE BULLETIN 


Fic. 2. The nesting area in a bank at County Line (1962). Nests were in the bank and the flat 
ground near the equipment. 


Tue Lire Cycte aNp BEHAVIOR OF THE PRIMITIVELY SoctAL BEE 367 


grew above it. The overhanging tree shaded the bank until 4:30 p.m. during the summer. 
Bees’ nests were in the topmost 1.5 meters of the bank, under loose hanging roots of the tree, 
reaching a maximum density of 320 nests per square meter (maximum internest distance 16 
cm; minimum internest distance 6 mm in this square meter). No nests were in a zone of 
heavy dark gray clay that was in the nesting area. The entire bank was occasionally eroded 
by flood water. Few bees remained in this bank in 1963 after a flood had eroded much of 
the bank (maximum density was two nests per square meter) and by August, no nests were 
there. A new nesting site was found in April, 1963, about 50 meters upstream. This bank, 
composed of similar soil, was about three meters high and faced south. It was exposed to sun 
all day and had no vegetation on its face, although grass and weeds grew above it. The nests 
were in the upper one to two meters of the bank among hanging roots and stems of Vitzs. 
There were 5 to 10 nests per 0.09 sq. m over most of the nesting area (maximum density 36 
nests per 0.09 sq. m; maximum internest distance 14 cm; minimum internest distance 6 mm). 
Nests in this bank all appeared to be new; there was no evidence of a previous year’s burrows 
or cells. 

The disappearance of aggregations at the County Line and Wakarusa nesting sites may 
have been in part due to my removal of colonies for study. Many nests were destroyed by floods 
but most of the colony mortality was due to unknown causes. 

Small aggregations of L. zephyrum were found also (1957, 1959, 1961, 1962) in a south- 
facing sandy bank of the Kansas River, in North Lawrence, and in a southwest-facing vertical 
clay bank near a small stream at Hole-in-the-rock (1959), 12 miles south of Lawrence. The 
latter aggregation reached a density of 20 nests per square meter; Anthophora abrupta was also 
nesting there. Twelve nests were found on the gently sloping bare clay banks of a small semi- 
permanent tributary of the Wakarusa River, but the bees died out during the summer (1962). 
Lasioglossum zephyrum may occasionally make isolated, scattered nests but no search was made 
for them. Two females were collected on the campus of The University of Kansas in April, 
1963, at a footpath on a north-facing partly shaded slope. They apparently were looking for 
nesting sites although far from water or known aggregations. No nests were established there. 


Factors INFLUENCING Nest DistriBUTION 


Although in Douglas County, L. zephyrum usually nests in bare, south- 
facing banks of compact silty alluvium and in horizontal surfaces of similar 
soil near inhabited banks, many apparently favorable habitats did not contain 
nests. The aggregations that were found were limited in extent and did not 
fully occupy apparently suitable areas, but nests were somewhat clustered 
together. 

When about to start nests, females often appear to be attracted to sites 
where others already have begun nesting. They may be attracted by the odor 
of existing nests or bees, as seems to be the case in Amegilla salteri (Miche- 
ner, 1961). When vials were opened in which living females of L. zephyrum 
had been kept for several weeks, a sweet pungent odor could be often de- 
tected. The pollen balls and wet, newly-made waterproof linings of empty 
new cells had a similar, rather stronger odor. 

Females are attracted by visual stimuli such as entrances of nests and 
other irregularities in the soil surface. In the insectary rooms, nests and 
shallow “sleeping places” were frequently started where a clod of soil rested 
on the soil surface, or around the edges of small stones. Females readily 


368 ‘Tue University ScreNcE BULLETIN 


Fic. 3. The nesting area on the Wakarusa River (1962). Nests were in the bank below tree 
marked ‘X.’ 


| 


Tue Lire Cycle AND BEHAVIOR OF THE PRIMITIVELY SoctaAL BEE 369 


entered holes 4 mm in diameter punched in the soil and some began nests 
in these. 

In the field, females about to begin nest construction in spring fly back 
and forth in front of banks, often alighting to investigate nest entrances, 
crevices, holes, or protuberances. Ten holes, 4 mm in diameter and about 
6 cm deep, were made by me in a bank, and one week later five of them 
were occupied (April 21, 1962). Females also appear to be attracted by con- 
cavities in the bank surface and by overhangs, in this respect resembling 
Thygater and Lonchopria (Michener, et al., 1958). Nests were often started 
during spring in cavities where I had previously excavated, and the greatest 


concentrations of nests were found below the overhanging tops of banks. 


SuRVIVAL OF COLONIES AND AGGREGATIONS 


In studying survivorship, numerous nests were marked by pushing a 


wire bearing a numbered aluminum tag into the soil 2 cm to the right of 


each nest. The number of such nests having a guard was counted at inter- 
vals. Because new nests are not normally founded during summer, counts of 
the number of active nests in designated areas, made at intervals, indicated 
colony mortality rates in large aggregations. When a sample of apparently 
vacant nests was excavated, no bees of any age were found. 

Most nests founded in spring were abandoned after a few months. The 
reasons for this mortality of colonies were not clear. Survivorship curves for 
colonies of L. zephyrum (Fig. 4) resembled those for L. imitatum (Michener 
and Wille, 1961). The survivorship curve for eight colonies in an insectary 
room was similar to curves for 510 colonies in seven aggregations at three 
locations in the field. 

This seasonal mortality does not seem to be due to any of the natural 
enemies discussed by Batra (1965) but a few observations of sudden extinc- 
tion of colonies in the insectary rooms suggested that disease may be a factor. 
In these instances, the adults ceased activity and died within a few days either 
staying in the nest or crawling out. Larvae and pupae (visible through the 
glass walls of the nest boxes) apparently were unaffected. When they 
emerged from cells as adults they probably died or left, for the nests were 
not repopulated. Dissections of dead and dying bees revealed no obvious 
parasites or lesions. It is possible that death of the egg-layer, and failure to 
replace her, may have been a factor in colony mortality. However, in this 
species there are usually some “workers” with enlarged ovaries in each nest 
and males were available all summer for insemination; it is probable that 
egg-layers are normally replaced. 

In 1957 a heavy infestation by the mermithid nematode, Hexamermis 
sp., apparently was responsible for the mortality of many colonies in an ag- 
gregation. However, this nematode is usually not very common. 


370 Tue UNIversiry ScIENCE BULLETIN 


ale 


Saige Gi, INOW 


AUG. 


JUNE JULY 


MAY 


APR. 


@ ©) @ O © 
78) Ke) T N 
Percentage of nests surviving 


Q 
©) 


Fic. 4. Survivorship curves for colonies in 8 aggregations. A, 34 nests at County Line, 1962; 
B, 8 nests in the insectary room, 1963; C, 28 nests at County Line, 1962; D, 36 nests at 
Wakarusa, 1963 (observations terminated (X) when a flood destroyed the aggregation) ; 
E, 320 nests at Wakarusa 1962 to 1963; F, 49 new nests begun in summer after a flood 
at Wakarusa, 1963; G, 7 nests at Wakarusa, 1963; H, 36 nests at Wakarusa, 1962 


to 1963. 


Tue Lire CycLteE AND BEHAVIOR OF THE PRIMITIVELY SociAL BEE 371 


Lasioglossum zephyrum, as mentioned above, usually nests in locations 
subject to flooding. All aggregations regularly observed were covered by 
water at least once annually. In one nest, excavated soon after it had been 
covered for several hours by about 1.5 m of water (July, 1963), living pupae 
were found in cells 4 cm or more below the surface of the bank. All pupae 
less than 4 cm from the surface of the flooded bank were dead. Several living 
adults of various ages were in this nest, as well as in others excavated at the 
same time. Twenty-four hours after water one meter deep had covered a 
nest site for six hours (September, 1961), guards were at the repaired en- 
trances of three of seven marked nests (the tags of many more were washed 
away). Apparently the nest entrances are plugged by mud, which traps 
sufficient air in nests for survival. Nests excavated after flooding did not 
appear to have had water flowing into them. 

Flooding causes great mortality when parts of banks containing nests are 
washed away. This happened to portions of two aggregations (July, 1963, 
September, 1962). Two days after the flood of July, 1963, in which hundreds 
of nests were completely washed away, many inseminated, somewhat worn 
females with enlarged ovaries excavated at least 50 new nests in the freshly 
eroded area. These nests, similar in structure (Table 5) to the early spring 
nests made by overwintered bees, rapidly died out (Fig. 4), perhaps due to 


the dryness that followed the flood. Apparently none of the brood reached 
adulthood. 


LOCAL AND SEASONAL VARIATIONS IN SIZE 


Loca S1zE DIFFERENCE 


Females and males from the County Line aggregations had relatively 
larger heads (mean head widths) and mandibles than individuals from the 
Wakarusa aggregations. Mean wing lengths of large samples of each sex 
from both locations were similar and will therefore be used in the discussion 
of seasonal sizes and castes. Although there was a conspicuous difference in 
the relative proportion of head size to body size (measured as wing length) 
between the County Line and Wakarusa populations, no significant differ- 
ences (except those rather obviously due to environmental factors) were 
found in the behavior or life cycles of bees in these populations. Mr. George 
Eickwort (personal communication) has found similar morphological differ- 
ences between populations of L. zephyram in Oregon and size variation in 
Ontario (Knerer and Atwood, 1962) may be of the same sort. Possibly popu- 
lations of this bee are genetically more isolated than those of other halictines 
because it nests along streams and mates at the nesting site. The County 
Line population on the Marais des Cygnes watershed is about 16 miles south 
of the Wakarusa population, on the Kansas River watershed. An extensive 


S72 Tue Universiry ScrENCE BULLETIN 


sampling of L. zephyrum from many areas should be made to clarify this 
situation. 

Head capsule widths of 22 preserved last instar larvae from the Wakarusa 
population averaged 1.07 mm (S.E.—.007). Head capsule widths of 17 last — 
instar larvae collected in June at the County Line nest site averaged 1.12 mm | 
(S.E.=.010) and 27 last instar larvae from that population in August aver- | 
aged 1.14 mm (S.E.=.007) in head capsule width. Larvae, as well as adults, 
from the Wakarusa population had relatively smaller head widths than those 
from County Line. 


SEASONAL DIFFERENCES IN SIZE 

Seasonal size differences in both males and females have been recorded 
for L. imitatum (Michener and Wille, 1961), and seasonal size differences 
in females are known for L. duplex (Sakagami and Hayashida, 1958), L. 


Tasce 1. Seasonal variations in mean wing lengths of females in Groups A, C and | 
E. Standard errors (S.E.) and sample sizes (N) are indicated. Young females — 
(Group E, wear < 3) significantly differ in mean size at the 5°/ level when May | 
and June, June and July-Aug., July-Aug. and Sept.Oct. samples are compared. | 
Mean sizes of Group A females did not differ significantly during the year, and 
little;worn Group A females are not significantly smaller than Group A females. 


66,99 


in general. (All comparisons made with “Student’s “t” test.) 


Group A Group A Group E Group E 
total wear <3 Group C wear <3 wear >5 
“queens” new “queens” total young bees | old “workers” 


WN 
Zo ve 


4.565 
April iE: .087 
11 0 
x 4.626 4.616 4.371 
May SE. 054 114 03 
N 31 3 27 0 | 
x ; 
June S.E. 4 : 
N 
x 4.555 4.200 4.289 
July SE; 051 019 
N 17 2 64 4.149 4.164 
4.270 4.184 018 044 
Aug. 022 188 22 
be 4.469 l 48 
Sept. S.E. 072 5.310 4.526 4.513 
and N ot 046 369 
Oct. 1 96 14 


Tue Lire Cycite aNpD BEHAVIOR OF THE PRIMITIVELY SocrAL BEE 373 


rhytidophorum (Michener and Lange, 1958), Halictus ligatus (Knerer and 
Atwood, 1962) and L. malachurum (Stockhert, 1923; Bonelli, 1948), all of 
which show social tendencies and production of young in both summer and 
fall. In these species, females emerging in summer are smaller than those 
which emerge in the fall, overwinter, and start new nests in spring (queens). 
Lasioglossum zephyrum also shows this pattern of seasonal size variation in 
males and females. This trend is most clearly seen if only the youngest of 
the females are considered (Table 1, unworn group E females), probably 
because older bees are of variable and unknown ages, so that they confuse 
the relation between size and season. 

Observations made in the insectary room (Batra, 1964) suggested that 
seasonal size variation may be caused by fluctuations in the relative number 
of pollen collectors and egg-layers in each nest. Mean numbers of eggs and 
pollen balls without eggs per nest in the field (Fig. 21) show peaks in egg 
production in early June and early August. In contrast to the situation in 
preceding months, the mean number of eggs in late July and August was 
clearly greater than the mean number of pollen balls without eggs, possibly 
indicating some competition for available provisions among egg-layers. If, 
under such circumstances, pollen balls are completed and eggs laid on them 
before they reach maximum size, production of relatively small bees in late 
summer could occur (Table 1, unworn group E females). The larger, over- 
wintering bees seen in October and spring may arise from the few eggs laid 
in late August and early September in more amply provisioned cells, when 
pollen collectors are more active than egg-layers according to Fig. 21. Too 
few pollen balls were measured, due to difficulty in handling them, to deter- 
mine whether there are significant seasonal differences in their sizes. 


FEMALES 
CastE DIFFERENCES 


In studying castes, methods recommended by Michener, et al. (1955) were 
used, with minor alterations. Before nests were excavated, their entire popu- 
lations were aspirated and immediately preserved in Carnoy’s fixative. (The 
material was less brittle than when preserved in Dietrich’s solution and 
could be used for later histological work.) Often the activities of marked 
individuals were observed for a day or longer before their nests were opened. 
The wing lengths and head widths of preserved bees were measured, and 
the number of nicks in both forewings was counted. Mandibles were 
grouped in seven classes of wear (Michener and Wille, 1961), ranging from 
unworn (class 1), to very worn, with the subapical tooth and notch gone 
(class 7). Index of total wear was a sum of number of wing nicks and 
mandibular class. This index ranges from 1 (unworn) to a maximum of 30. 
Abdominal terga of the preserved females were removed, the spermathecae 
examined for sperm cells, contents of the gut noted, amount of fat estimated, 


374 Tue University ScIENCE BULLETIN 


degree of enlargement of the poison glands of the sting apparatus noted, and 
development of oocytes in the ovaries recorded. 

Lasioglossum zephyrum resembles L. rhytidophorum (Michener and 
Lange, 1958) in the weak differentiation of castes. There are no discontinu- 
ous morphological or behavioral differences between probable egg-layers and 
females that lay few or no eggs. Rather, there is a continuum from queen- 
like to worker-like individuals. Young (unworn) inseminated or non- 
inseminated females commonly develop enlarged ovaries as they become 
older (more worn), although one or more bees with enlarged ovaries may 
be living in the nest at the same time. Most workers are moderately worn 
females with somewhat enlarged ovaries and they probably occasionally lay 
eggs. Relatively few much-worn females with undeveloped ovaries, com- 
parable to the females that constitute the worker caste in social bees, includ- 
ing L. imitatum, L. malachurum, and others, are found. 

For this study, five groups of females were recognized, based on degree of 
ovarian development and presence or absence of spermatozoa in the sperma- 
theca. As in L. imitatum (Michener and Wille, 1961), “group A” females 
were fertilized, with several large oocytes, the anterior portions (terminal 
filaments) of the ovaries sometimes being pressed into folds and containing 
somewhat swollen oocytes. Usually such bees had one or more eggs large 
enough to lay (1 mm or more in length). 

Group B females were fertilized, with somewhat enlarged oocytes, and 
may or may not have had an egg ready to lay (largest oocyte 0.25 mm or 
more in length). 

Group D females were fertilized but oocytes were not enlarged (largest 
oocyte less than 0.25 mm long). 

Group C females were unfertilized, their degree of ovarian development 
like that of group B. They probably laid some male-producing (haploid) 
eggs. 

Group E females were unfertilized with small oocytes less than 0.25 mm 
long, or rarely without oocytes visible with the dissecting microscope. 

The percentage of females belonging to each of these groups varies dur- 
ing the year (Table 2). In spring, all bees are large, fertilized, overwintered 
females (groups A, B, D). A given female could have been classified in any 
of these three groups, depending on the amount of ovarian enlargement 
after she emerged from hibernation (as a group D female) and whether she 
had recently oviposited (possibly reducing her from group A to group B). 

In May, the first young, on the average smaller, unfertilized females ap- 
pear (group E). Many of them develop enlarged ovaries within about 15 
days after emergence, thus joining group C. 

Group C females (Fig. 5) comprise about 35 per cent of the population 
during midsummer. They may lay many haploid eggs, resulting in the 


Tue Lire Cycte AND BEHAVIOR OF THE PRIMITIVELY SocrAL BEE 375 


NOW DEC. JAN. 


SEFT. OC 


JUNE JULY AUG. 


> 
< 
> 
ag 
Oo 
< 
© O © O O © O 
© X@) wv 60) aN) Se 


Percent in each group 


Fic. 5. Seasonal fluctuations in proportions of females belonging to groups A and C from nest 
populations. See Table 2 for information concerning other groups. 


August and September peaks in production of males (Fig. 22). Unfertilized 
bees probably produce many of the males in L. malachurum also (Noll, 
1931). In early August, as the number of males increases, many group C 
females probably become inseminated, thus joining group B, which increases 
at this time. 

During summer, the proportion of females in group A decreases as the 
total number of females per nest (in other groups) increases. The number 


376 THe University SCIENCE BULLETIN 


Tase 2. Seasonal fluctuations in the percentages of 831 females from nest popu- 
lations that belong to each ovarian group. 


Inseminated Not Inseminated Number 

A B D Cc E of bees 
April eee 30.6 18.3 51.0 0.0 0.0 49 
May palclis ieee eee 60.0 6.6 33.3 0.0 0.0 30 
May 16-30 ...........- 32.7 20.0 Seat 4 36.3 55 
june lolb 36.7 45 45 31.0 229 87 
June 16-30 .......... IPB ed] 2 3.3 27.1 30.5 59 
Julyalels 12.6 S// 3.4 333}05) 44.8 87 
July 16-30) 8.2 12.4 DD 35.1 38.6 145 
Augustamlcilp eee 7.8 IPAS) 4.7 Sei 37.0 127 
August 16-30 ...... pall 20.8 18.0 WS) Si 72 
September 1-15 ... 3.5 26.7 44.6 ile 23.2 56 
@ctobers jess = 0.0 8.5 57.4 Da S19 47 
INovember” == 0.0 9.0 91.0 0.0 0.0 11 
Januatyaee 0.0 0.0 100.0 0.0 0.0 6 


in group B remains fairly constant, perhaps because inseminated group C fe- 
males and group D females whose ovaries had enlarged join this group. 

In August and September, the total percentage of females with enlarged 
ovaries (groups A, B, and C) declined from about 58 to about 13 per cent, as 
the total percentage of females with slender ovaries increased from about 
42 to about $7 per cent (Table 2). It is not clear why the ovaries of females 
do not enlarge in late summer, but perhaps environmental influences such 
as declining average temperatures (Fig. 1) or shortened day length have 
some effect. By October most bees with slender ovaries are inseminated 
(group D), and only these and a few little-worn group B females with slight- 
ly developed ovaries enter hibernation. 


“QUEENS” 


Because there is, during summer, a continuum between queenlike and 
workerlike individuals, only females belonging to group A will be classified 
as queens in the discussion below. Females in groups B and C probably lay 
some eggs, but many of them act as foragers (workers). All females in 
spring before the summer bees emerge are, of course, “queens,” although they 
first appear as group D and later pass through group B to group A. 

An effort was made to determine whether the large, overwintered egg- 
layers (group A) live longer than “workers” (groups B, D, E, and C in sum- 
mer). Marking bees with “dope” and enamels was unsuccessful, for the spots 
of paint did not adhere well and often peeled away within a few days. Many _ 
overwintered females were marked in April and May by clipping off the | 


tarsus of one middle leg, and replacing the bees in their marked nests. Five _ 
| 


Tue Lire Cycie anp BenHavior oF THE PrimitivELy SoctaL Bee 377 
of them lived until the second week of June, but no marked spring egg-layer 
was recovered after June 14. These old egg-layers had an index of wear of 
five to eight, no pollen in the crop, and belonged to groups A and B. 

The index of total wear may be used to indicate the relative age of bees, 
assuming that all bees are active or work equally and fairly constantly. The 
index of wear of group A bees showed a steady increase until the end of May 
after which it declined (Table 3). In June, an increase in wear in all groups 
was seen, probably related to a sharp increase in production of brood at that 
time (Fig. 21). During the rest of the summer, the mean index of wear of 
group A females fluctuated but showed no overall increase, probably indi- 
cating that as egg-layers became older (more worn) some of them were re- 


Taste 3. Indices of total wear of females belonging to groups A to E during the 
year. All specimens were taken from nests. Mean (X), range (R) and number 
of bees examined (N) are indicated. CL and W refer to females from County 
Line and Wakarusa, respectively. The usually lower indices of wear of County 
Line females may be due to the relatively lesser wear of their larger mandibles. 


Inseminated Not Inseminated 


A B Eo TC a: 
X,(R),N X,(R).N X,(R),N X,(R),N X,(R),N 
April Cr, 0 2.29(1-4)7 —-2.58(1-6)24 0 0 
W. —_5.40(3-8)15 —-2.50(2-3)2.-——-3.00(3)1 0 0 
May 1-15 CL.  4.50(4-5)4 0 0 0 0 
W. _8.07(4-25) 14 4.50(4-5)2. ——- 5.80(3-9) 10 0 0 
May 16-30 CL. 6.00(5-8)8 —3.50(3-4)2 0 0 3.25(1-8)4 
W. —-8.23(4-25) 13. 6.50(3-11)6 —4.33(1-10)3_—-2.33(2-3)3 «(1.56 (1-3) 16 
June 1-15. CL. 4.29(1-12)18 4.00(4)1 233(1-4)3° ~— 3:89(1-7)19"  S77eioyis 
W. —-7.07(3-12)15_2.00(2)2 1.00(1)1 3.00(1-5)8  3.14(2-4)7 
June 16-30 C.L.  6.92(3-14)12  4.00(4)4 9.00(9) 1 4.73(1-13)15  2.40(1-9) 15 
W. — 10.33(8-13)6 10.00(10)1 6.00(6) 1 6.00(6) 1 3.30(3-4)3 
Miel-15 CL. 4.88(2-11)8  2.50(1-4)2 , 1.00(1)2 1.55(1-3)18  2.42(1-16)26 
W. 6.60(4-11)5— 3.00(3)1 1.00(1)1 3271-8) 11) 2etdes 1s 
July 16-30 C.L. 5.38(1-13)13. 6.00(1-9)5 —-.2.00(1-4)7 — 4.34(1-15) 32 1.97(1-6)30 
W. —_-9.55(2-19)11 3.00(1-8)8 —«:1.00(1)1 2.68(1-4)19  2.00(1-6)26 
August 1-15 C.L. 5.38(3-10)13  3.86(2-6)7 —-2.50(1-6)6 —-3.48(1-8)42~—-2.47(1-10) 45 
W. — 4.00(4)2 3.50(3-4)4 0 4.33(3-6)6  4.00(4)2 
August 16-30 C.L. 7.05(1-30)21 9.00(3-15)2  4.62(1-10)13  4.89(2-12)9 — 3.37(1-18)27 
W. 0 0 0 0 0 
September C.L. 11.33(7-17)3  7.71(3-19) 14 4.80(1-12)251.00(1)1 2.00(1-12) 13 
W. 0 0 0 0 0 
October CL. 5.50(2-9)2 —9.00(6-12)2.-1.48(1-4)27—‘:1.00(1)1 2.53(1-7) 15 
Ww. 0 0 0 0 0 
November CLL. 0 6.00(1)1 3.70(1-10) 10 0 0 
Ww. 0 0 0 0 0 
January GE. 0 0 0 0 0 
W. 0 0  2.30(1-5)6 0 0 


378 Tue University SCIENCE BULLETIN 


placed by relatively unworn younger bees. However, observations in the 
insectary room (Batra, 1964) suggested that the egg-layers do little work in 
or outside the nest and it is possible that the index of wear of such bees does 
not significantly increase as they become older. But some group A females 
(indices of wear 1 to 3) which were not worn enough to have begun the 
nests in spring were found from early June to October, which strongly sug- 
gests that replacement of worn egg-layers by at least some unworn (young) 
ones is likely (Table 3). 

The large mean size of group A females even in late summer may indi- 
cate either that few were replaced, or that only the largest summer bees 
joined group A. A sample of 6 little-worn group A females in June was not 
significantly smaller than 30 group A females in June, suggesting that large 
summer bees join group A (Table 1), but it is also possible that these little- 
worn group A females had overwintered and for some reason had done little 
work. Replacement of egg-layers may be common, since males are present 
all summer and ovaries of many females appear to enlarge spontaneously 
(i.c., without mating), as in L. rhytidophorum (Michener and Lange, 1958). 

Females in group A are not always the largest individuals in each nest. 
Thirty-five nests containing at least one group A female each were examined 
from May to August, before production of the large overwintering bees be- 
gan. In 21 of these nests, a group A female was the largest, but in 14 of 
them, a bee belonging to another group was largest (group E in 6 nests, 
group D in 2 nests, group C in 3 nests and group B in 3 nests). This also may 
indicate that the original large overwintered group A female is sometimes 
replaced by one of her smaller daughters. 

Nests having no inseminated female with enlarged ovaries are common 
(Fig. 6). In April, many of the overwintered bees were just beginning nests 
and their ovaries had not yet developed mature oocytes. Before the first 
young females emerged from their cells at the end of May, many of the 
overwintered egg-layers that were resting in closed nests again had slender 
or only moderately enlarged ovaries. During mid-summer, death of egg- 
layers with no immediate replacement probably occurred. After August, 
development of eggs appeared to be inhibited. 


“WORKERS” 


Females with various degrees of ovarian development act as workers dur- 
ing summer. Dissections of 113 females that were collecting pollen in July, 
1962, when nests were not being provisioned by solitary egg-layers, showed 
that 63 per cent of them belonged in group C, 20 per cent in group E, 1 
per cent in group B, 4 per cent in group A, and 2 per cent in group D. Ag 
that time of year (Table 2) about 33 per cent of females in nests belonged in 


Tue LirE Cycie AND BEHAVIOR OF THE PRIMITIVELY SoctAL BEE 379 


SEPT. ©GCie sNOV “vA: 


AUG. 


JUNE = JULY 


MAY 


APR. 


O @ O 
© tT A 


‘e) 
6) 


100 


Percent 


Fic. 6. Percentages of nests having no female (‘‘queen’’) belonging to groups A or B. Numbers 
across bottom refer to numbers of nests examined each month. 


380 Tue UNiversiry SCIENCE BULLETIN 


group C, 45 per cent in group E, 6 per cent in group B, 13 per cent in A and 
3 per cent in group D. Relatively more group C and fewer group E females 
acted as foragers than would be expected, if foraging were randomly dis- 
tributed among the groups. Fewer females belonging to groups A, E, and D 
were collecting pollen than were present in the general population, probably 
because the group A females (“queens”) mainly laid eggs and group D and 
E females were immature. 

From the end of May to July, most foragers belong to group C, but in 
August most of them were in group E, perhaps reflecting the inhibition of 
oocyte development seen in late summer (Table 2): 

The mean mandibular wear of pollen collectors during most of the sum- 
mer was about three but wing wear fluctuated, maximum wing wear oc- 
curring in June and August, when peaks in production of brood were seen 
Figs. 7, 21). The increase in mean mandibular wear of foragers in fall 
probably occurred because younger, less worn bees emerging at that time did 
not also collect pollen. 

The moderately worn mandibles and enlarged ovaries of most pollen- 
collectors suggest that, in this species, foragers are bees that have done or are 
doing some excavation in the nest while their ovaries enlarge. Observations 
in the insectary rooms showed that the youngest bees (probably group E) 
excavate the burrows, and only after a week or more do they usually begin 
to collect pollen and make cells (Batra, 196+). The relatively few group E 
females active as pollen collectors in the field also suggested this, although 
such bees leave their nests for other purposes. 

Group C females appeared in nests 15 to 20 days after the emergence of 
group E females in May, indicating that females belonging to group C (most 
foragers) were individuals that had previously belonged to group E, as was 
expected. 

Unlike L. imitatum (Michener and Wille, 1961), group E females with 
an index of total wear of five or more (typical “workers”), were rare, com- 
prising only 3.4 per cent of summer females. Lasioglossum zephyrum does 
not have a large distinct worker caste, to which the older (more worn) fe- 
males can be assigned. Worn group E females were of approximately the 
same sizes as unworn (total wear 3 or less) group E females (Table 1D) 

Apparently L. zephyrum is a bee in which castes are indistinct. Develop- 
ment of large oocytes occurs in most fertilized or unfertilized females during 
summer as they become older. Insemination is probably not related to age 
or to the degree of enlargement of the ovaries and occurs all summer, per- 
mitting replacement of group B females and of at least some females in 
group A. Many workers in groups C and B probably lay eggs and oviposi- 
tion by a forager was actually observed (Batra, 1964). 


Tue Lire Cycie AND BEHAVIOR OF THE PRIMITIVELY SociAL BEE 381 


RPOLEEINMEOEREECTORS 


%: 


MAI" JUNE* JUL AUGS SBPIr OCT 


Fic. 7. Mean wing lengths in mm (WL), mean mandibular wear (MW) and mean number of 
wing nicks (WN) of pollen collectors and guards during various months. Standard 


errors are indicated for each point, and across the bottom of each graph, sample sizes 
are indicated. 


Tue University SCIENCE BULLETIN 


ios) 
CO 
bo 


ACTIVITIES VISIBLE OUTSIDE THE NEST 


ConpiITIONS FOR ACTIVITY 


Most aspects of the seasonal cycle are probably regulated by climate and 
are not obligatory. Colonies in the insectary rooms maintained continuous 
activity during fall and winter if suitable foods were available and other 
environmental conditions adequate. The resting period that follows provi- 
sioning by nest-founding females is, however, not dependent on environ- 
mental conditions and females that had started nests at any time of year in 
the insectary rooms always ceased activity until their progeny emerged. 

Temperature was an important factor that influenced behavior in spring 
and fall. In March, guards appeared at the nest entrances when the soil at a 
depth of 21 cm was at least 7° C. New nests were excavated when the 
temperature of the soil (at a depth of 21 cm) was about 11° C (Fig. 1). 
Females removed from their nests at the end of April, when air and soil 
temperatures were 14° C, could not fly, although some of them had already 
been excavating. At air temperatures of 15° to 18° C (as measured in shade, 
1 m above the ground on sunny days), females flew slowly, often alighting 
to “sun” themselves on the soil surface. At 16° C, bees were able to fly when 
the sun was out but when it became cloudy they apparently could not fly and 
crawled about or clung to vegetation. If it was windy at this temperature, 
they did not leave their nests but did fly on warmer windy days. During 
summer, flights occurred at air temperatures as high as 40° C, activity of 
this species not being inhibited by high temperatures. In fall, at air tempera- 
tures of 14° to 17° C, both males and females were active. Males were some- 
what more active than females, this difference in activity perhaps facilitating 
copulation. The hibernating females are resistant to freezing. When nests in 
frozen soil were excavated, the bees (at 0° C) moved their legs and antennae 
feebly and when brought to 20° C, they began walking about. Males were 
not resistant, none being found after the first killing frost of autumn. 

Changes in light intensity may influence activity. Normal flights occur 
on clear and slightly overcast days. One morning in July at the height of 
pollen collecting activity, as dark storm clouds rapidly approached, the maxi- 
mum light intensity quickly dropped to 350 ft-c. Females no longer left their 
nests, and many, none carrying pollen, entered nests. Honeybees also return 
to the hive if dense clouds appear (Percival, 1947). Males continued their 
usual swarming activity, until I was forced to leave by sudden rain and 
tornadic winds. Both males and females maintained normal levels of activity 
during a partial solar eclipse in July, 1963. 

Females continued to collect pollen during light rain (drizzle), but if 
rain became heavier (shower), they ceased collecting and returned to their 


Tur Lire Cycie AND BEHAVIOR OF THE PrimMiTIVELY SocraL BEE — 383 


nests. Males seemed to be somewhat more active in rain than females, but 
rested in vegetation or entered abandoned burrows during heavy rain. 

Females are most active outside their nests during the morning hours 
(Fig. 8). Lasioglossum imitatum and other halictines show the same pat- 
tern of behavior (Michener and Wille, 1961). Unlike L. imitatum, females 
of Lasioglossum zephyrum return from most trips without pollen through- 
out the day. The morning peak in foraging activity may be related to other 
activities inside the nests. Usually, cells are made during the night and are 
ready for provisioning in the morning. By afternoon many of them have 
been provisioned and the foragers are engaged in making pollen balls or the 
cells have been closed after oviposition (Batra, 1964). 

Few males or females were flying as early as 8:00 a.m., even on warm 
days, and in the insectary rooms they did not begin activity until the lights 
had been on for at least one half hour. By sunset (about 7:30 p.m. in sum- 
mer), most females had returned to their nests and few males were still 
flying. Nests were guarded as late as 9:30 p.m. (with full moon) although 
the guards were timid. Some nests were guarded on dark nights, but others 
were not. 


Foracinc BEHAVIOR 


Foragers are usually females that have done some work in the nest before 
beginning to collect pollen. Marked individuals lived as long as 31 days in 
the insectary room and some collected pollen at intervals for as long as 19 
days. Observations in the insectary rooms showed that pollen collectors also 
excavate the burrow, guard, make cells, make pollen-balls, provide nectar for 
the provisions, and occasionally oviposit. For example, one bee collected pol- 
len and nectar, made a pollen ball, and made a cell in one day; another col- 
lected pollen, made a ball and worked in the burrow in one day (Batra, 
1964). 

Increases in mean wing wear occurred at the same times that increased 
brood production was seen (Figs. 7, 21). This suggests that individual 
foragers make more flights when more cells are available for provisioning, 
and spend more time in the nests when there are fewer cells to provision. 
Seventeen marked foragers made one to eight pollen collecting trips per day 
(mean=3 trips) in the field. There is considerable variation in the amount 
of pollen collecting activity of different foragers in a nest and of the same 
forager on different days. This is probably related to the number of cells 
available for provisioning and to the physiological condition of the foragers. 

Lasioglossum zephyrum visits many species of angiosperms for provi- 
sions. Other authors (see Introduction) have listed a large number of 
flowering plants visited by this bee; for this reason, little time was spent col- 


384 


Fic. 


Number of nests observed at each hour 


8. 


Tue University SCIENCE BULLETIN 


1 


12M 


11 


10 


SAM 9 


an 
O 


2) 
N 


US) 
14 
12 


@r Ose 
sn @) » © 
Entries per nest per hour 


OA 


ip 


Average number of times per hour that females entered their nests with and without 
pollen loads. The number of nests observed for each hour is indicated at the top of the 
graph. Observations of 67 nests were made in June and July during fair weather. Sev- 
eral adjacent colonies, having various degrees of activity, were observed on each day. 
Some nests were watched for as many as 3 days, others for only half a day. 


Tue Lire Cycite ann BEHAVIOR OF THE PRIMITIVELY SoctAL BEE 385 


lecting L. zephyrum on flowers. A list of plants on which it was observed 
during this study is given in Table 4. Most observations on foraging be- 
havior were made in the insectary rooms, where the bees appeared to behave 
normally. 

Foragers do not always exhibit flower constancy. Although females 
usually returned to the same species of plant, sometimes a female collected 
pollen from as many as three genera of plants in a single collecting trip. 
Bees frequently go back and forth between two different genera, finally 


Tasie 4. List of Mowers visited by L. zephyrum. P indicates that females were 
collecting pollen, N that they were taking nectar. M indicates that males were 
taking nectar. -BR indicates that observations were made in the bee-room. 


Liliaceae Draba reptans N 
Allium mutabile N-BR Onagraceae 
Iridaceae Oenothera speciosa N-BR 
Iris cult. var. N-BR Scrophulariaceae 
Commelinaceae Verbascum blattaria N-BR, P-BR 


Tradescantia canaliculata P-BR 
Ranunculaceae 

Thalictrum dasycarpum P-BR 
Leguminosae 

Melilotus alba M, N, P, M-BR, N-BR, 

P-BR 

Amorpha canescens P-BR 

Petalostemon candidum P-BR 

Petalostemon purpureum P-BR 

Petalostemon multiflorum P-BR 
Asclepiadaceae 

Asclepias sp. N-BR 

Asclepias sullivanti N, M, N-BR, M-BR 

Asclep:as syriaca N-BR 
Rosaceae 

Rosa setigera P, P-BR 

Rosa suffulta P, P-BR 

Rubus ostryifolius P-BR 

Fragaria virginiana N 

Spiraea cult. var. P-BR 
Solanaceae 

Solanum rostratum P, P-BR 
Oxalidaceae 

Oxalts stricta N, M, M-BR 
Verbenaceae 

Verbena hastata N 
Cruciferae 

Rorippa sinuata N 

Brassica campestris P-BR, N-BR 

Capsella bursa-pastoris P-BR 


Campanulaceae 
Campanula americana P-BR 

Labiatae 
Monarda fistulosa N-BR 
Salvia azurea M-BR 

Compositae 
Cacalia tuberosa N-BR 
Cirsium lanceolatum P-BR 
Chrysanthemum leucanthemum P-BR 
Chrysanthemum cult. var. P-BR, N-BR 
Amphiachyris dracunculoides P-BR 
Echinacea pallida P-BR 
Aster sp. P-BR, N-BR, M 
Aster oblongifolius M-BR 
Aster ericoides M-BR, P-BR 
Erigeron sp. N, P, N-BR, P-BR 
Erigeron ramosus N-BR 
Rudbeckia hirta M-BR, N-BR 
Rudbeckia sp. M-BR 
Ratibida pinnata N-BR 
Helianthus annuus M-BR, P-BR 
Helianthus grosseserratus M-BR, N-BR 
Silphium integrifolium N-BR, P-BR 
Silphium laciniatum N-BR, P-BR 
Taraxacum vulgare N 
Vernonia fasciculata N-BR, P-BR 
Solidago sp. M-BR, N-BR, P-BR 
Solidago altissima M 
Solidago glaberrima M-BR, N-BR, P-BR 
Solidago rigida M-BR, N-BR, P-BR 


386 Tue University SciENcE BULLETIN 


carrying a mixed pollen load to the nest. They sometimes obtained nectar 
from the pollen source but more often visited another kind of plant for this, 
or took the honey and water mixture provided on feeding stands and on 
various artificial flowers. Although this species is polylectic, foragers show 
strong preferences for certain flowers, and the flowers of many species of 
plants provided for them in the insectary rooms were completely ignored or 
visited much less frequently than those of other plants. Pollen collecting trips 
to flowers were 3 to 32 minutes in duration, most of them lasting about 
10 minutes. 


| 


Foragers collecting pollen usually first loosen the pollen grains from the — 


anthers by biting them. They rapidly brush back the pollen with the front 
tarsi, transferring it briefly to the middle tibiae. The middle legs immediately 
press it into the scopa on the hind legs and abdomen. As in L. imitatum 
(Michener and Wille, 1961), there is no regular alternation of movements, 


the forelegs and midlegs on one side of the body sometimes transferring — 


pollen two or three times in succession before those on the other side are used. 
As the load of pollen becomes larger, foragers begin packing it into the scopa 
by pressing the hind legs together and against the abdomen. About one- 
third of the pollen load is carried on the abdominal scopa. Females that were 
collecting pollen of Vernonia sometimes used the mandibles to hold onto the 
anthers while the legs brushed up pollen. 

The apices of the antennae of pollen-collectors are almost constantly in 
contact with the pollen. Bees, attracted to flowers of Helianthus having few 
pollen-bearing anthers, wandered across the disc and did not begin pollen- 
collecting motions until their antennae contacted the pollen. Foragers drink 
nectar before, after, or during the collection of pollen and sometimes eat 
pollen also. Sometimes females take no nectar, but only pollen, to the nest; 
others take both or only nectar. An individual often collected pollen on 
some trips and nectar on others. The kind of provisions collected depends on 
the consistency of the provisions deposited previously in the cells (Batra, 
1964). Bees collecting nectar sometimes pause for five or ten minutes, while 
they extend and retract the glossa with a drop of nectar, about twice per 
second. Several other halictines show similar behavior (Plateaux-Quénu, 
1959), probably evaporating water from the nectar. 

Foragers and males were collected on flowers about 230 meters from the 
nearest known nesting area, and marked females readily returned to their 
nests when released about 170 meters away. Only one of nine marked 
females returned to her nest from a distance of 0.5 mile (809 meters). Few 
flowers are available near the Wakarusa nesting site, and foragers probably 
visit flowers in pastures that are at least 150 meters away. 

Foragers apparently are able to learn what are suitable pollen or nectar 


Tue Lir—E Cyct—E AND BEHAVIOR OF THE PRIMITIVELY SocIAL BEE 387 


sources, and ability to collect pollen seemed to improve with time. Two days 
after the first young bees were seen in insectary nests (May 20, 1963), three 
foragers on flowers of Rosa suffulta did not fill their scopas although they 
made the usual motions and pollen was abundant. Probably these were 
young, inexperienced foragers, since most bees on Rosa rapidly filled their 
scopas. Females in the field returning to nests with incomplete pollen loads 
(July, 1963) were like those returning with full loads with respect to man- 
dibular and wing wear and ovarian development. Factors other than im- 
maturity apparently were responsible for the partial pollen loads in this 
instance. 

Females in the insectary room were induced to collect pollen of Alnus 
sp. (which L. zephyrum is not known to visit in nature) from a petri dish. 
Flowers of Aster sp., from which bees were collecting pollen, were dusted 
with the pollen of Alnus, which they then picked up. The Aster plants were 
placed near some yellow plastic imitation sunflowers, each of which had a 
small pile of Alnus pollen on its disc. Bees began collecting this pollen, and 
the asters were removed. A petri dish, having a circular piece of yellow 
paper beneath it, was placed near the sunflowers. A black “disc” and black 
“rays” had been drawn on the paper. The following day, foragers were col- 
lecting Alnus pollen from a small heap placed over the “disc” area. Because 
these females had not been marked, it is not known which individuals 
learned to collect this pollen. They were probably later replaced by young 
untrained bees as they died off. A colony of bees was maintained through 
the winter, using this pollen alone. Foragers usually hovered over the dish 
for a few seconds, alighting at the edge of the pile of pollen. When bees were 
collecting the loose Alnus pollen, foraging trips were brief, lasting only one 
or two minutes, unless they also visited the honey-feeding stations. This sug- 
gests that the duration of a foraging trip depends in part on the ease with 
which pollen is gathered. 

When crude paintings of flowers were made on one wall of an insectary 
room (1962), several females were attracted, alighting on the “stems” and 
“leaves.” They also alighted at the centers of asterisk-shaped blue “flowers.” 
After two days, bees were no longer attracted to the paintings, apparently 
having learned to ignore them. 

Artificial flowers were effective competitors with real flowers for attention 
outdoors as well as in the insectary rooms. Females in the field fed on honey 
at plastic flowers placed near the nesting site. 

Halictine bees are well known for their annoying tendency to lick up 
human perspiration. Males and females of L. sephyrum do this occasionally. 
Both sexes also licked the stems, tendrils, and both sides of the leaves of 
Vitis sp. growing near the nesting site. Females were observed licking stems 


388 Tue UNIversity SCIENCE BULLETIN 


and buds of Aster sp. in the insectary rooms, and they licked dry sugar at a 
feeding stand. In the field, they licked the honeydew on stems and leaves of 
suckers of Populus deltoides that were infested with aphids. When licking 
skin or plants, the bees were unusually persistent and not easily frightened 
away. One female that was licking a pencil was caught and squeezed tightly 
so that she buzzed and struggled. She immediately flew back to the pencil 
and resumed licking when released. This treatment and response were re- 
peated three times. As the bee walked slowly over the pencil, the antennae 
moved about, lightly contacting it. A trace of liquid could sometimes be 
seen at the rapidly moving apex of the glossa, which was extended back 
under the thorax. Moisture was readily available at streams near the nesting 
sites, but bees were never seen drinking water. They may have been licking 
up sugars (of plants) or salts (in perspiration). 


GUARDS 


Overwintered solitary females that are actively provisioning their nests in 
spring usually guard them intermittently. In the insectary rooms, such 
females often plugged the nest entrance with a little soil when they had 
finished provisioning for the day, and were about to oviposit (usually in the 
afternoon), and the nests remained plugged for several hours or overnight. 
When the overwintered females had ceased activity in May and were waiting 
for their brood to mature, the nest entrances were plugged with soil or were 
left open (Fig. 9) and were usually without guards. When several overwin- 
tered females stay together in a nest in spring, it is guarded. After the young 
females emerge at the end of May, most nests are always guarded. Guarding 
by solitary females does not occur in L. imitatum and other halictines (Mich- 
ener and Wille, 1961). 

Observations of colonies in the insectary rooms have indicated that the 
youngest females guard their nests and excavate for several days before leav- 
ing on foraging trips. Figure 7 shows that many large, overwintered females 
were probably replaced as guards in summer by their smaller daughters. 
Many of the overwintered females (group A) as well as smaller group E 
bees were guarding in June, perhaps accounting for the relatively large 
standard error of mean size of guards in that month. Throughout the sum- 
mer, the mean mandibular wear of guards remains at about three, and wing 
wear did not show increases associated with brood production as did wing 
wear of pollen collectors (Figs. 7,21). There were differences between wing 
wear of guards and pollen-collectors in spite of the fact that many foragers 
guard before leaving nests on foraging trips, and some of them were probably 
included in the samples of guards. Females parasitized by gregarines, nema- 
todes, female Strepsiptera and conopid larvae were seen guarding nests. 


Tue Lire Cycre anp BEHAVIOR OF THE PrimitivELy SociaL BEE 389 


IG. 9. A. Excavated soil forming a heap at the base of a bank that contains many active nests. 


B. Entrances of two new nests in spring. C. Entrance of a nest in summer, with a guard. 


390 Tue UNIversiry ScIENCE BULLETIN 


Percentages of guards belonging to each ovarian group in June and 
August are indicated below: 


Group A B (G D E 
jJune*(NE= 29) == eee ey eee 34% UNG, 24% 4°, SIs 
AN ost Ne 0S 9) ees Bio 19% 61% 3%5 14% 


Guards of L. sephyrum in general resemble those of L. imitatum (Miche- 
ner and Wille, 1961) in behavior. Aggressiveness of guards in defending the 
nest entrance seemed to depend on time of day and temperature. When the 
soil temperature was less than 16° C and few or no bees were flying, guards 
were timid. After about 7:00 p.m. (in summer) when all bees had ceased 
flying, guards were timid and retreated into their nests when disturbed. 
Solitary overwintered females that had finished provisioning were also timid 
and often retreated to the bottoms of their nests, leaving them unguarded. 
Possibly guards are females “intending” to fly out of the nest and the stronger 
the urge to leave, the more aggressive the guard (Legewie, 1925). In the 
insectary room, females that had emerged from their cells 24 hours earlier 
were aggressively guarding nests and flying. 

Guards defend the nests against small organisms such as phorid flies and 
ants simply by blocking the constricted nest entrance with the head (Fig. 9), 
or by striking out with open mandibles. More aggressive organisms, such as 
mutillids, are kept out of the nest by plugging the entrance with the posterior 
abdominal terga. Guards frequently leave their nests to attack nearby fe- 
males of Pseudomethoca frigida (Batra, 1965). When heavy rain falls on the 
nest entrances, guards sometimes block the burrows with their abdomens or 
plug them with a little soil to protect the nests from flooding. Guards en- 
gaged in attacking an intruder would, at the same time, briefly withdraw to 
permit nest mates to enter, but when guards block the nest entrances with the 
abdomen, other females can neither enter nor leave. 

When nests are well populated, usually several females stay in the burrow 
just below the guard. Marked guards in such nests were replaced as guards 
by other individuals every 15 to 30 minutes. 

Females occasionally guard shallow burrows (1 or 2 cm deep) excavated 
as overnight sleeping places, and bees have been found resting in detached 
turrets at nesting areas as if guarding them. One female, collected as she was 
guarding a nest in the field, was put into a nest in the insectary room. With- 
in one minute, she began to repair the entrance of this nest, apparently not 
disturbed by the transfer. (See Batra, 1964, for additional information on 
guards.) 


Lost BrEEs 


As in L. imitatum (Michener and Wille, 1961), young females make 
more elaborate orientation flights when leaving the nest than experienced 


Tue Lire Cycie AND BEHAVIOR OF THE PRIMITIVELY SociAL BEE 391 


females do, and they appear to have difficulty locating their nests on their 
return. In the insectary rooms, newly emerged females and overwintered 
females often flew up toward the lights when first leaving their nests, but 
after a few days in the rooms, were no longer attracted to the lights. Over- 
wintered females, in the field and in the insectary rooms, at first had difh- 
culty finding their newly established nests. By the time they began collecting 
pollen, however, they usually entered and departed rapidly. One forager, 
that had been searching for her nest for about ten minutes, alighted and 
brushed off her pollen load before continuing. Females, like some other 
Hymenoptera, apparently relate the positions of their nests to landmarks. 
One disoriented bee was attracted to the southwest side of several clumps of 
grass before she found her nest, which was at the southwest side of a grass 
clump. Another female flew to the wrong nest-box in the insectary room and 
spent about five minutes flying and walking over the soil at its right side, 
minutely examining the surface with her antennae and occasionally digging. 
Finally she flew to the neighboring nest box, where she immediately entered 
her nest in the soil at the right side of the box. Many lost bees were seen 
when floods or trampling by cattle had apparently destroyed landmarks. On 
the other hand, bees readily located nest entrances covered with dirt or 
debris by the observer, if the surrounding area was undisturbed. 

Disoriented bees sometimes enter the wrong nests and are usually not 
repulsed by the guards. Most of these females leave the nests immediately 
but rarely a bee, marked as a member of the population of one nest, was 
later found behaving normally in another nest. In early spring, when over- 
wintered females are establishing nests, they frequently enter nests of other 
females when inspecting holes for nest sites. One overwintered female that 
was excavating a burrow only about 2 cm deep, was joined by a second bee. 
Both females tried to stay in the nest, first one, then the other being pushed 
to the surface, each briefly taking a guarding position before trying to crowd 
deeper into the nest; one of them finally left. In two instances, a previously 
marked overwintered female left her own nest and was found living in the 
nest of another bee, along with the original inhabitant. Probably the newly 
established nests containing two females (relatively common in L. zephy- 
rum) arise in this way. 

During summer, disoriented bees were sometimes much worn or para- 
sitized individuals. In early June, several “lost” females that were dissected 
appeared to be old, overwintered bees of groups A or B; but others were 
newly emerged. In the insectary room, one senile marked queen was appar- 
ently stung to death by the guard of a nest (not her own) that she had per- 
sisted in attempting to enter during the morning. That afternoon, the guard, 
still in the entrance, was biting one of the front legs of the weakly quivering 
bee. Occasionally dying bees were found near nest entrances in the field. 


392 Tue Universiry SciENcE BULLETIN 


These females of various ages, may have been also injured by guards. Guards 
sometimes refused to let certain individuals enter their nests, although at the 
same time other bees were permitted to enter. The females that were at- 
tempting to enter frequently stood near the nest entrances, facing the guards, 
as if waiting. They made orientation flights around the nests at intervals and 
occasionally tried to push into the nests past the guards, which bit at the 
intruders, sometimes grasping their mandibles. Females that had been re- 
pulsed by guards repeatedly cleaned their heads and antennae while waiting 
near the nests. Some of these bees were apparently lost, for they later entered 
other nests, but some belonged in the nests they were trying to enter. Bees 
that had left their nests to attack mutillids were often kept from re-entering 
immediately. Probably bees that approached the guards rapidly were more 
readily allowed to enter than those that hesitated, such as lost, inexperienced, 
senile or parasitized females. In 1957 many disoriented females were seen in 
July at an area in which most bees were infected by the nematode, Hexamer- 
mis. Six of a sample of eight such lost females contained the nematode. Bees 
containing puparia of male Halictoxenos (Strepsiptera) were found walking 
or weakly flying about the nesting site as if lost. 


NEST STRUCTURE 


INTRODUCTION 


The structure of nests changes somewhat during the year as the popula- 
tion of bees in each nest varies in number and in composition. The generali- 
zations made below are based in part on statistics given in Table 5 and in 
Fig. 10. Nests characteristic of various phases of the life cycle are illustrated 
in Figs. 11 to 17. 


HIBERNACULA 


Inseminated females overwinter in nests constructed and occupied during 
the summer. One exception may have been a female in a new unbranched 
burrow 4 cm deep, collected in November. 

The entrances of overwintering nests in banks are about 2 mm in diame- 
ter, and not plugged with soil, but those in horizontal ground are usually 
closed by rain or other disturbances. Each hibernating female usually oc- 
cupies a branch burrow (hibernaculum) 1 to 1.5 cm long, of the same diame- 
ter as the main burrow (about 4 mm). These branch burrows may be 
homologous to the resting places seen in nests during summer and also to 
“sleeping places” excavated by females that have no nests. Overwintering 
nests were 24 to 40 cm deep, the bees in branches at depths of 12 to 37 cm 
(Fig. 11). The overwintering females apparently make the side branches 


Tue Lire Cyctt AND BEHAVIOR OF THE PRIMITIVELY SociAL BEE 


593 


Taste 5. Characteristics of nests throughout the year. Means (X), standard 
errors (S.E.), range (R) and number of nests examined (N) are indicated. 
“New” July nests are those newly excavated after a flood in July, 1963. 


Maximum Minimum Maximum Number Bees Brood Cells 
Nest Depth Depth of Per Rex, 
Depth Eggs Eggs Branches Nest Nest 
April x 9.42 5.00 8.00 1.14 2.69 
(New Nests) SE. 0.66 0.51 0.42 0.08 0.54 
(R)N (2-15)26  (3-9)18 ~~ (3.5-14) 18 0 (1-2)21  (0-9)26 
April x 20.63 4.63 
(Old Nests) SiE; 33355) 0.89 
(R)N  (5-39)8 0 0 0 (1-8)8 0 
May 1-15 x 12.06 5.43 8.14 0.29 1.42 B12 
S.E. 0.91 0.84 Vg 0.19 0.09 0.93 
(RIN (7-18)17  -(@2-15)14 G-19)14 ~—s(03) 17 (13) 36— “(0-16)17 
May 16-30 x 14.79 6.62 10.19 0.48 2.43 3.86 
S.E. 0.96 0.62 0.80 0.16 0.28 0.78 
(ROIN| § (9-28)19) (2214)211 (4-16)21 (03) 21 (1-6) 21 O20 
June 1-15 me 21.45 14.85 17-55 1.65 5.30 9.80 
S.E. 2.09 1.69 1.68 0.67 1.56 178 
(R)N (7-50)20  (5-40)20 = (5-40)20 = (0-13)20  (1-34)20 (1-37)20 
June 16-30 x 25.42 17.74 22.91 2.00 3.86 19.58 
S.E. 1.62 1.43 1.84 0.20 0.96 3.03 
(R)N (11-42)26 (5-33)23  (9-41.5)22_ (0-16)26 (1-22)21 (0-57)26 
July 1-15 oe 31.00 20.60 22.60 2.43 12.71 14.43 
SE. 3.46 1.36 2.02 0.72 278 5.39 
(R)N (20-44)7  (17-24)5.— (17-28)5.— (0-5)7._—s (8-26)7 ~—- (0-36) 7 
July 16-30 i 26.83 18.18 22.18 5.42 11.18 24.92 
S.E. 3.07 3.50 2.93 1.55 3.18 6.22 
(R)N (14-48)12 (9.5-36)11 (10.5-38)11 (1-17)12 (4-38)11 (0-65) 12 
July 15-30 4 6.67 4.40 5.60 0.50 1.08 1.08 
(New) S.E. 0.91 121 1.50 0.26 0.08 0.47 
(RIN (2211) 12-9 (1E8'5)5.  Gie10)5 (0-3)12 (1-2)12  (0:4)12 
August 1-15 xX 39.67 29.13 35.50 10.30 14.30 45.30 
S.E. 7.85 3.36 3.45 72 4.08 10.81 
(R)N (32-57)9  (16-45)8 — (24-55)8 = (3-29)10 (5-45) 10 (0-85) 10 
September 1-15 X 31.56 5.60 8.30 18.40 
S.E. 3.81 2.37 3.81 7.82 
(R)N (15-49)9 0 0  (0-26)10 (1-41)10 (0-81)10 
Winter i 32.36 4.88 4.44 
S.E. 1.50 1.88 1.39 
(R)N (24-40) 11 0 0 (0-16)8 (1-14)9 0 


which are sometimes plugged with a 


main burrow. 


little soil, but do not deepen the 


394 Tue University ScreNcE BULLETIN 


PotycyNnous Sprrnc Nests 


In spring, some females remain together in the nests in which they hiber- 
nated; others leave these nests to begin new ones. 


APR, MAY JUNE JULY AUG. SEPT. 'OCT. “NOY. 


Characteristics of nests during the year. ND, mean nest depth in cm; LC, mean depths 
of lowest cells containing eggs in cm; UC, mean depths of uppermost cells with eggs 
in cm; C, mean number of cells per nest; BN, mean number of females per nest; 
B, mean number of branches per nest. (0), old spring nest; (n), new spring nest. 


Fic. 10. 


Fic. 11. Hibernacula. A, Overwintering females in short branch burrows. Nest opened October 
3, 1957; B and C. Females in characteristic hibernation attitude, January 3, 1964. In 
diagrams of nests, the following symbols are used: N, new empty cell; B, pollen ball 
or loose pollen in cell; E, egg; S, small larva; M, medium larva; L, large larva or 
prepupa; 6, male pupa; @, female pupa; X, dead or parasitized brood. Stippled 
areas indicate cells or burrows that have been filled with soil by the bees. Small letters 
indicate connecting points, when diagrams have been broken to conserve space. 


396 Tue UNiversiry ScIENCE BULLETIN 


When several overwintered females were found using the same nest 
during early spring, it seemed probable that they had hibernated together if 
old earth filled cells were found near the burrow or if the nest had been 
marked the previous year. Females emerging from hibernation fill the side 
branches and lower part of the burrow with soil before beginning to make 
new cells (Fig. 12). New spring cells in old nests were excavated at about 
the same depths as those of new nests in spite of the availability of deep 
burrows. Nests containing several bees were constantly guarded and had 
well-maintained, constricted entrances. 


New Sprinc NEsts 


Shallow nests, with no traces of old cells or burrows, containing one or 
sometimes two females, were often known to be new. They often have a 
branch burrow about 1 cm long near the entrance, but most have no other 
branches. Burrows made in banks are fairly straight but those in horizontal 
ground are often more crooked. New nests in spring frequently have un- 
repaired entrances 3 mm or more in diameter and are guarded only inter- 
mittently. Such nests usually are not guarded after the females finish pro- 
visioning the cells (Fig. 14). New nests were about three times as numerous 
as old polygynous nests in spring. Nests resembling new spring nests in 
structure were excavated in July by females whose original nests had been 
destroyed by a flood. 


SUMMER NEs ts 


After the young females emerge in late May, nests are gradually extended 
deeper into the earth. The number of branches and cells is directly propor- 
tional to the number of bees per nest (Fig. 10). The level of occupied cells 
descends with the deepening of the burrows. The deepest nest reached 57 cm 
(August), the maximum number of branch burrows was 38 (July), the 
largest number of occupied cells was 85 (August), and the maximum num- 
ber of females per nest was 45. Other data are given in Table 5 and Fig. 10. 
In late summer, the number of bees and inhabited cells rapidly decreases, 
until by October only the overwintering females are left in the nests where 
they stay to hibernate. Summer nests in horizontal ground are in general 
vertically oriented but often tortuous, with many branches ramifying in vari- 
ous directions. Those in banks are somewhat straighter, the branches being 
usually fairly straight and running into the bank (Figs. 15, 16, 17). Perhaps 
nests in horizontal ground are tortuous because L. zephyrum usually nests in 
banks; such nests often have subhorizontal sections like nests in banks. The 
actual amount of tortuousness is dificult to draw in two dimensions. 


397 


Tue Lire CycteE aNp BEHAVIOR OF THE PRIMITIVELY SocrtaL BEE 


‘syueyd Jo sjapoor yorya Aq papunosins st yey? [[29 peuopueqe woz [euaeU [e92J pur ]Ios Jo sseur 
pedwuog “Fy ‘[los Jo ss9Xe] QUaUO0D Jo dn aYINq ya4IM) JOYS ‘4 ‘SauUeNUS IsaU aAOqE sNNLUMY UT UOITII ayl-fouuny “yf fooeyzans 
su uo sinjstow Jo sjajdorp puke Suruy peystjod SutAyrspun soacois Surmoys ‘[[99 speur A[MEN “| !196] “ET AL Ysou MoU Ul a]/euw 
“aq “CQ :punoj sem suo AyUO pur [los yWM pal]y useq Apuosedde savy passquIMiszao Koy) YTYM UT sayoueI IWOYs 2YL “[96] ‘T Judy 
uoneusoqry Worf Surss9ua saaq ‘OQ S196] “E] APJ 240Jaq [los YM p2|[y useaq sey (KX) MoLING Plo 24, ‘paqeusaqry pey Asyd yoy 
UT ysou B UT sopetUay ‘q [96] “T [Udy ‘Iscu ke ajeAvoXa oy SuluuIseq aewag ‘SY ‘aiMjoNYs sou Jo sjtelap pue Sutids Ayres UL S]SON, 


ZO 


398 Tue Universiry Scrence BULLETIN 


5¢m. 


Fic. 13. Spring nests in a bank, as viewed from above. A, Polygynous nest with old cells, 
May 7, 1963; B, New nest with two bees, April 23, 1963. 


Tue Lire Cycie anp BEHAVIOR OF THE PrimitivELy SociaL BEE 399 


Brera 


Fic. 14. New spring nests and early summer nests in banks as viewed from above. A, April 2a. 
1963; B, May 2, 1963; C, Resting female in closed nest with maturing brood, May 10, 
1963; D, Resting female in open nest. She dug a narrow escape tunnel when the nest 
was dug up for study, May 10, 1963; E, and F, Nests with young summer females and 
new cells, June 8, 1963. 


400 Tue UNIversiry ScIENCE BULLETIN 


GENERAL CONSIDERATIONS 


Nest entrances are usually constricted to about 2 mm in diameter at the 
County Line site, where guards have relatively large heads. At the Wakarusa 
site, entrance diameters average about 1.8 mm. Nests in horizontal ground 
often have radial tumuli of granular excavated soil particles surrounding the 
entrance. Depending on the activity of the bees and the weather, tumuli 
reach diameters of 4.5 cm and heights of 3 cm. The inner wall of the tumulus 
is sometimes made of compressed soil, leaving a turret when loose soil blows 
away. Soil particles pushed from nests excavated in banks form piles up to 
8 cm deep on ledges and at the bases of the banks (Fig. 9). Nests in banks 
sometimes have short turrets up to 7 mm in outside diameter and 6 mm in 
length, apparently made of concentric rings of mud pushed out of the nest 
when the soil is wet. Funnel-like depressions are rarely seen around the 
entrances of nests in horizontal ground (Fig. 12). No subsurface enlarge- 
ments of the burrows (“sentry boxes”) are made by the populations studied 
although L. Chandler (personal communication) reports such structures in 
nests of L. zephyrum in Indiana. 

The main burrow and branch burrows are usually about 4 mm in diame- 
ter (range 3.2-455 mm). Burrows maintain very roughly a 90° angle to the 
surface of the soil, so that those in horizontal ground are more or less vertical 
and those in steep banks are approximately horizontal in orientation although 
sections of nests are often oriented differently. Intermediates were seen, de- 
pending on the steepness of the slopes. Branch burrows were of various 
lengths, some of them nearly equalling the main burrow, others only 1 cm 
long. Usually a short blind burrow is present as a continuation of the main 
burrow or branch burrows beyond the deepest cells. The burrow walls are 
smooth and sometimes a lining of compressed soil, about 0.5 mm thick, can 
be detected. This lining is conspicuous when burrows pass through different- 
ly colored layers of soil or through rotten wood buried in the soil. 

In arrangement of cells, nests of L. zephyrum belong to subtype IIIb, 
OCH'"B or to subtype IIIc, OCH’"B of the classification of Sakagami and 
Michener (1962). Vertical sections of burrows often have cells arranged 
radially but in burrows or sections running horizontally the cells are ar- 
ranged serially along each side of the tunnel (Figs. 15, 16), with few excep- 
tions. The degree of concentration of the cells is variable, but in general, 
spring or summer nests containing several females have cells more crowded 
together than nests occupied by only one or two bees. Cells are sometimes 
separated by only 0.5 mm to 1 mm of soil. Excavating females probably de- 
tect the secretion that soaks into the soil around cells, thus being able to 
avoid cells already completed when making new cells or burrows. 


Tur Lire Cycite aNp BEHAVIOR OF THE PRIMITIVELY SOCIAL Bee 401 


Fic. 15. Early summer nests and polygynous spring nest in horizontal ground. A, June 8, 1962; 


B, June 11, 1962; C, May 13, 1961. 


402 Tue University ScrENcE BULLETIN 


Fic. 16. Summer nest in horizontal ground, July 17, 1957. The diagram has been broken to 
conserve space. Letters indicate connecting points. 


Tue Lire Cycre anp Benavior oF THE PrimitiveLy SoctaL BEE 403 


aan 


Fic. 17. Two summer nests in banks, viewed from above. A, July 18, 1962; B, September 13, 
1962. The diagrams have been broken to conserve space and letters indicate connect- 
ing points. 


404 Tue UNiversiry ScIENCE BULLETIN 


Cells are subhorizontal and bilaterally symmetrical, 9 to 11 mm in total 
length including the neck (extremes 8 to 15 mm), 4 mm in maximum verti- 
cal width (3.5 to 4.5 mm), and slightly more in maximum horizontal width. 
The necks of the cells are usually about 2 mm wide (1.7 to 2.5 mm) and 
2 to 4 mm long. Cells are closed after oviposition with a plug of loose soil 
pellets which extends into the neck for a distance of about 2 mm. The por- 
tion of the plug forming part of the burrow wall is of smooth compressed 
soil, so that there are no visible differences in the burrow surfaces between 
regions where there are plugs and those without plugs. An area about 7 mm 
long, occupying the widest end of the cell, is lined with a thin, shiny, trans- 
parent, waterproof film that can be flaked off when cells are soaked in water. 
Observations in the insectary room have shown how this film is applied over 
a layer of compressed soil. Completed cells in the insectary room nest-boxes 
had distinct halos where a liquid apparently had soaked into the soil for a 
distance of about 1 mm around each cell. In the field, newly-made cells with 
wet, sweet-smelling walls were occasionally found, the secretion penetrating 
the surrounding soil for about 1 mm. It appears unlikely that this amount of 
secretion could be produced by the salivary glands alone, although the glossa 
was used by bees lining cells in the insectary rooms (Batra, 1964). Bordas 
(1894) has shown that the Dufour’s glands of the sting apparatus of young 
females of L. leacozonium (a solitary species) are greatly enlarged in spring. 
Semichon (1906) noticed that the Dufour’s glands of freshly killed Antho- 
phora personata were greatly enlarged and filled with an oily liquid, which 
when dried, became a white waxy solid that resembled the lining of the cells 
of this bee in solubility and melting point. These glands were most enlarged 
at the time when the bees were making cells. He believed that the secretion 
of Dufour’s gland was used by Anthophora in making cells. In L. zephyrum, 
11 of 14 solitary females taken from nests with newly made, empty cells in 
April had greatly enlarged Dufour’s glands, but only two of six females 
taken from closed nests during the resting period in May had enlarged 
Dufour’s glands. During summer, enlarged Dufour’s glands were found 
mainly in females belonging to groups C, B, and A in this order of fre- 
quency. Observations in the insectary room have shown that foragers (prob- 
ably most in groups C and B) made cells. Possibly the secretion from 
Dufour’s glands soaks into the soil and is spread about by the glossa or aug- 
mented by a salivary secretion. One cell, excavated in the field, had a polished 
lining with minute droplets on its surface, which suggested that two secre- 
tions may have been used (Fig. 12). The waterproof layer becomes brown 
in old abandoned cells. Small grooves, probably made by the pygidial plate 
in tamping and smoothing the earthen walls, are often seen underlying the 
shiny secreted linings of the cells (Fig. 19). 


Tue Lire Cycie AND BEHAVIOR OF THE PRIMITIVELY SoctAL BEE 405 


IMMATURE STAGES AND PROVISIONS 
PROVISIONS 

In the field as well as in the insectary rooms, more than one female usually 
provisions each cell during summer. As in L. imitatum (Michener and 
Wille, 1961) and L. calceatum (Plateaux-Quénu, 1963), there are often more 
pollen-collectors per nest than cells being provisioned at that time, indicating 
cooperative activity. For instance, five marked individuals were carrying 
pollen into a nest that yielded only three newly provisioned or empty cells 
when excavated that day. Of course, the number of foragers that were sup- 
plying nectar alone could not be counted. 

Cells opened in the field while being provisioned contained one or another 
of the following sorts of pollen masses: (1) a small amount of loose, dry 
pollen, (2) a small, moist but crumbly pollen ball, (3) a similar small pollen 
ball on top of a mass of dry pollen, (4) a large amount of dry pollen (Fig. 18), 
sometimes in differently colored layers, (5) a mass of loose, somewhat moist 
pollen that may be molded by the fingers like dough, (6) a small pollen ball 
with a mass of loose pollen on top of it, (7) a pollen ball with a few flakes of 
agglutinated pollen grains on top of it, and (8) a full-sized, smooth pollen 
ball. The consistency of completed pollen balls varies from that of butter at 
room temperature to that of dough. The different types of provisions listed 
above may be related to the stage at which provisioning was interrupted; the 
process of provisioning is discussed by Batra (1964). 

Pollen balls collected in the field are frequently more moist outside than 
in the center. When females from several nests were fed a honey and water 
mixture to which blue food color had been added, some of the pollen balls 
that were removed from the nests a few days later were stained blue. One of 
them, bearing an egg, was blue only in its upper half. This suggested that 
honey was added to the pollen ball after it had been shaped, and either that 
the working over of the ball by the egg-layer just before oviposition did not 
disturb its surface, or that the honey may have been added after oviposition 
(progressive provisioning). Usually pollen balls appear dull in texture, but 
glossy nectar oozes out when their surfaces are scratched. 

The flattened, subspherical pollen balls of L. zephyrum measure 3 to 4.1 
mm in diameter and 2.0 to 2.5 mm in height; there is no depression for the 
egg. Completed pollen balls, each bearing one egg, are usually somewhat 
tilted in the cells (Fig. 18), unlike those of other halictines. No loose pollen 
or nectar is visible on the walls of cells that contain completed pollen balls. 
Pollen balls have a strong, pungent, sweet odor. 


Eces 


As in other halictines, the eggs are white, smooth and curved. Eggs, in- 
cluding those laid by a single female, vary in curvature. Some contact the 


406 Tue Universiry SciENcE BULLETIN 


Fic. 18. Development of the brood. A, Loose pollen in a partly provisioned cell; B, Egg on a 
characteristically tilted pollen ball. The upper end of the egg will develop into the 
head of the larva. C, Young (probably first instar) larva, with head raised. 


Tue Lire Cycte aNnp BEHAVIOR OF THE PRIMITIVELY SociAL BEE 407 


= ES RR 


Fic. 19. Development of the brood. A, Second or third instar larva. A cross-section of the 
burrow is at the right; B, Larva (probably third instar), with head raised in alarm 
pose. The grooves in the shiny wall of the cell may have been made by the pygidial 
plate of the female that constructed it; C, Fourth or fifth instar larva eating a pollen 


ball. 


408 Tue UNIversiry ScIENCE BULLETIN 


pollen ball only at each end, but others contact it throughout their entire 
lengths. The widest, blunt end of the egg is directed toward the neck of the 
cell, and develops into the head of the larva. The tapering, narrower end 
terminates as a transparent area. A few eggs had a small watery droplet on 
their upper surfaces (some such eggs hatched). Eggs were 1.39 to 153 mm 
long and 0.44 to 0.48 mm across the widest end (Fig. 18, 20). 


LarvaE 

Larvae were reared in cavities made in a layer of embedding paraffin in 
large, covered Stender dishes kept in darkness at 16° C in an incubator. Each 
ege or larva, with its pollen ball, was placed in a separate cavity as soon as 
its cell was opened in the field. Larvae survived best if kept cool and not 
exposed to the sun or allowed to dry out. Each Stender dish contained a 
piece of moist cotton to maintain high humidity (usually so that water con- 
densed on the lid). A honey and water mixture was added to the pollen balls 
periodically, for the larvae did not eat provisions that became dry (in spite 
of the apparently high humidity). Larvae were examined and measured at 
intervals of two or three days. Dead ones were removed to discourage growth 
of fungi. About one-third to one-half of the eggs and larvae died within two 
days of collection. Many of these had probably been injured by handling 
with forceps and showed black areas at the sites of injury, which spread over 
the entire body within an hour or two. Twenty-five eggs and young larvae 
lived for considerable periods in the Stender dishes, although only one de- 
veloped completely from egg to pupa. Prepupae and pupae were relatively 
easy to raise and required little attention. 

Before hatching, the segments and pulsating head of the young larvae 
could be seen. As in L. duplex (Sakagami and Hayashida, 1960), the chorion 
was first removed from the anterior part of the larva in hatching. Larvae 
began to eat and raise their heads if disturbed within 24 hours after hatching, 
although the rear half was still encased in the chorion. Eggs in the laboratory 
hatched after two to ten days. 

Five larval instars (Fig. 20) could be distinguished by observing discrete 
increases in widths of the head capsules of growing individuals. The dura- 
tion of each instar was highly variable, perhaps because of adverse conditions 
in the rearing dishes. Head capsules of first instar larvae were 0.41 to 0.48 
mm‘ in width and the first instar lasted two to ten days (N=4). Second 
instar head capsules were 0.54 to 0.58 mm in width, this stage lasting two to 
ten days (N=4). Third instar head capsules were 0.65 to 0.75 mm wide, this 
stage lasting two to four days (N=8). Fourth instar head capsules were 
0.78 to 0.88 mm wide, this stage lasting one to three days (N=6). Fifth 


4. Some larvae were larger than others throughout all instars. This may account for the near 
overlap of head capsule widths of different instars. 


Tue Lire Cycie AND BEHAVIOR OF THE PRIMITIVELY SocIAL BEE 409 


Fic. 20. Immature stages, drawn from fixed specimens all to same scale. A, Egg, with hyaline 
region at narrow end; B, First instar larva; C, Second instar larva; D, Third instar 
larva; E, Fourth instar larva; F, Fifth instar larva about to defecate; G, Prepupa; 
H, Pupa of male. 


instar larvae defecated and became prepupae without an additional molt 
(N=9). Their head capsules were 1.03 to 1.16 mm wide and they lived three 
to thirteen days before defecating. Although the size of the head capsule in- 
creased only at each molt, the rest of the body grew gradually larger. A 
sample of preserved larvae from the field had the same distribution of head 
capsule widths, which suggested that five larval instars are normal for this 
species. During molting, the old larval skin was gradually pushed back from 
the head (the delicate transparent skins had been dusted with pollen grains 
to make them visible). 


410 Tue UNiversity ScIENCE BULLETIN 


Behavior of larvae is in general like that of L. duplex (Sakagami and 
Hayashida, 1960) and L. imitatwm (Michener and Wille, 1961). The young- 
est larvae (first to fourth instars) eat the portions of the pollen balls near 
their mouths and are relatively inactive. Last instar larvae curl around their 
pollen balls (Figs. 18, 19). They consume about two-thirds of the provisions, 
moving the head from side to side as they chew. These larger larvae are less 
sensitive to disturbances and desiccation of pollen balls than are earlier in- 
stars. When the pollen balls have been consumed, such larvae continue to 
make chewing motions until ready to defecate. In the insectary room, a large 
larva that had eaten all of its provisions continued to chew and actively 
moved about in the cell as if looking for more food, completing a 360° revo- 
lution within 35 minutes. 

In the insectary rooms, at 28° to 33° C, the egg and larval stages com- 
bined lasted eight to 11 days (one egg developed to about second instar in 
4.5 days from time of oviposition). Development from egg-laying to adult, 
regardless of sex, required 21 to 24 days at these temperatures. In the field, at 
a mean soil temperature of about 18.5° C, development from egg to adult re- 
quired about 38 days. 


PupPAE 


Defecation and pupal development are essentially as outlined for L. du- 
plex (Sakagami and Hayashida, 1960). Pupae in the insectary rooms occa- 
sionally rotate on their long axes in the cells by twisting their abdomens. Al- 
most all pupae and prepupae excavated in the field had their heads directed 
toward the plugs of their cells, although fifth instar larvae were oriented in 
various ways. The only exceptions were one prepupa (parasitized by two 
small mutillid larvae) and one normal black-eyed male pupa, excavated in 
July and August, 1963, after a flood. Jay (1951) found that honeybee larvae 
depend on differences in texture of cell and cap for orientation. Possibly, the 
flood disoriented these L. zephyrum larvae by somehow changing the texture 
of the plugs or walls of their cells. 

Teneral adults of both sexes usually stand with their heads pressed against 
the plugs of the cells (in rearing dishes, they pressed against the sides of the 
Wax Cavities). Females attempt to sting when their cells are opened but do 
not try to escape. 

After bees have emerged, their cells are filled with soil brought from 
other parts of the burrow. The fecal material, decomposed by fungi, small 
oligochaetes and mites, is gradually converted into a yellow or brown pow- 
dery material. Roots of plants sometimes penetrate abandoned cells and 
surround the masses of soil and the probably nitrogenous fecal material in 


them (Fig. 12). 


Tue Lire CycLte AND BEHAVIOR OF THE PRIMITIVELY SoctAL BEE 411 


GENERAL CONSIDERATIONS 


Peaks in production of young were seen in June and early August (Fig. 
21 and Table 6), but the broods are not as distinct as in some species of 
Lasioglossum (such as the Palearctic malachurum, duplex). Probably the 
relatively short lives of overwintered females and oviposition by successive 
groups of young females are responsible for the fluctuations in brood produc- 
tion that are not seen in L. imitatum (Michener and Wille, 1961), a species 
with a long-lived queen and few egg-laying workers. The condition in L. 
zephyrum seems to be intermediate between that of solitary species with 
distinct broods and that of the more social Nearctic species with fairly con- 
stant production of young. 

In general, cells nearest the soil surface are constructed and provisioned 
first, so that there is a sequence in ages of the immature stages from oldest 
above to youngest below. Cells along a branch burrow usually contain the 
older stages of brood nearest the origin of the branch. However, many irreg- 
ularities occur, especially when there are several females in the nest. 

As in other halictines (Sakagami and Michener, 1962), solitary females 
as well as females nesting together often make and provision two or more 
cells simultaneously. 


MALES 


SEASONAL CYCLE 


Males first appear at the nesting areas during the third week of May, at 
the same time that the first female progeny of the overwintered females 
appear, but they are scarce. They gradually become more abundant until the 
second half of September when their numbers begin to decrease somewhat. 
The last males were seen during the second week of October, none being 
found after the first frost. The relative abundance of males observed flying 
around the nesting sites corresponds with fluctuations in the percentage of 
pupae that were males (Fig. 22). Males leave their nests permanently after 
emerging from their cells; Marshall and Musgrave (1937) found them sleep- 
ing in large aggregations in curled leaves. The few non-teneral males found 
in nests were not used to estimate abundance. The probable relationship 
between number of group C females and number of males produced has 
been discussed. 

Fluctuations in the mean number of wing nicks of males flying around 
the nest sites during the season agrees well with the fluctuations in produc- 
tion of males. When many young males are being produced during July to 
September, males on the average, are less worn than in June. 


Taste 6. Composition of the brood during the year. For each month and each 

immature stage, the percentage (°/,), mean number of cells per nest (X), standard 

error (S.E.) and range (R) of cells containing that stage are given. “New” July 
nests are those newly excavated after a flood in July, 1963. 


Empty Small Larva Large 
Cell Pollen Egg Larva Med. Larva Prepupa 
April 1-30 of, 20.8 18.0 41.6 6.9 12.5 0.0 0.0 
XC 0555 Mel til “Ol85" “014728 
S.E. 0.154 0.154 0.294 0.093 0.506 
R 0-3 0-3 0-5 0-2 0-3 
May 1-15 % Dey, 11.5 4.3 4.3 1.4 18.8 23.1 
x 0.750 0.400 0.150 0.150 0.050 0.650 0.800 
S.E. G22¢— 0133 ONl09" 01150) = 0.418 0.451 
R 0-3 0-2 0-2 0-1 0-1 0-8 0-8 
May 16-30 of, 7.3 21.9 15.8 2.4 0.0 12 0.0 
ne 0.285 0.857 0.619 0.095 a 0.047 
S.E. 0.122 0.474 0.335 0.095 ae a 
R 0-2 0-9 0-5 0-2 ae 0-1 = 
June 1-15 of, 9.6 17.2 19.2 ies 15.7 13.1 6.5 
x 0.950 1.700 1.900 1.100 1.550 1.300 0.650 
S.E. 0.303 0.391 0.397 0.176 0.426 0.616 0.335 
R 0-5 0-6 0-5 0-2 0-6 0-12 0-6 
June 16-30 of 31 3.3 3.9 6.2 5.5 115 18.5 
x 0.615  0:653- 0.769 “.1.230 1.076 ~ “2-269 0NiiS-c53 
S.E. 0.205 0.166 0.217 0.310 0.293 O77 "0686 
R 0-4 0-4 0-3 0-4 0-17 0-5 0-12 
July 1-15 of, 7.9 1.0 0.0 2.0 2.0 5.0 12.9 
Ne 1.142 0.142 .: 0.285 01285" | 10x74 esoy 
S.E. 0.705 a a 0.286 0.286 0/421) seita20 
R 0-5 0-1 Be 0-2 0-1 0-3 0-9 
July 16-30 of, 5.3 5 9.9 10.6 4.0 11.9 8.6 
xX 1.333 1.333 2.500 2.666 1.000 3.000 2.166 
S.E. 0.541 0.284 1.005 0.964 0.407 1.700 0.860 
R 0-5 0-3 0-10 0-20 0-4 0-20 0-10 
July 16-30 of, 84.6 0.0 15.4 0.0 0.0 0.0 0.0 
(New) x 0.916... 0.166 
S.E. 0.313 a 0.112 
R 0-3 fs) 0-1 
August 1-15 of 4.4 1.5 9.5 11.0 75 9.1 8.6 
axe 7.000 0.700 4.300 5.000 3.400 4.100 3.900 
SE. 1.265 1.158 2.128 3.066 1.905 581 9225 
R 0-13 0-3 Qe) = n0=31 0-18 0218 0-10 
September 1-15 % 4.4 0.5 0.0 1.6 Ant 19.7 Dep) 
Se 0.800 0.100 0.300 0.800 3.600 0.400 
S.E. 0.327 0.213 0.416 1.694 0.267 
R 0-2 0-1 0-2 0-3 0-17 0-2 
October 1-15  % 0.0 16.7 0.0 0.0 0.0 0.0 0.0 
x 0.150 
SE 0.433 


Tue Lir—E CycLtE AND BEHAVIOR OF THE PRIMITIVELY SoctaL BEE 413 


Tase 6. (Continued ) 


Female Female Male Male Total Total 
Pupa Imago Pupa Imago Cells Nests 
April 1-30 of 0.0 0.0 0.0 0.0 
x 72 27 
S.E. 
May 1-15 of 14.0 0.0 0.0 0.0 
x 0.500 69 20 
S.E. 0.211 
R 0-3 
May 16-30 of, 41.4 47 0.0 0.0 
x 1.619 0.380 82 21 
S.E. 0.428 0.128 
R 0-7 0-2 
June 1-15 Wh 6.0 0.0 1.0 0.0 
xe 0.600 0.100 197 20 
S.E. 0.550 0.100 
0-11 0-2 
June 16-30 oy, 35.3 4.9 6.2 1.6 
x 6.961 0.961 1.230 0.307 513 26 
S.E. 1.060 0.306 0.547 0.198 
R 0-29 0-5 0-9 0-5 
July 1-15 of 40.6 18.8 6.9 3.0 
x 5.857 2.714 1.000 0.428 101 y 
S.E. 27 14 1.208 0.655 0.297 
R 0-16 0-8 0-4 0-2 
July 16-30 A 78 7.0 7.9 iby 
Se 7.000 1.750 2.000 0.416 302 12 
S.E. 2.243 0.780 1.095 0.260 
R 0-22 0-8 0-11 0-3 
July 16-30 of 0.0 0.0 0.0 0.0 
(New) xX 13 12 
S.E. 
R 
August 1-15 Wis 30.9 2.9 We 7; 2.9 
a 14.000 1.300 5.300 1.300 453 10 
S.E. 4.119 0.578 2.724 0.989 
R 0-47 0-5 0-26 0-10 
September 1-15 % 45.9 2.7 1931 0.0 
x 8.400 0.500 3.500 184 10 
S.E. 3.949 0.307 2.007 
R 0-37 0-3 0-20 
October 1-15 So 77.8 0.0 5.6 0.0 
xe 3.500 0.250 18 4 
S.E. L8i7 


R 0-12 0-1 


414 


BiG: 


Average number per nest 


Tue University ScIENCE BULLETIN 


ay ey ey ef sh 
ey 2 IS eS 


mb wo A OF OFN WD © 


—- 


/ 


APR. 'MAY "JUNE JULY -AUG. SEPLP OCT. 


21. Fluctuations in the production of brood. PB, pollen balls without eggs; E, eggs; ML, 
medium sized larvae; P, female pupae. The probable relationships between the num- 
ber of pollen balls without eggs and the number of eggs are discussed in the text and 
additional figures are given in Table 6. 


Tue Lire Cycite AND BEHAVIOR OF THE PRIMITIVELY SocitaL Bee 415 


Fright Bewavior oF Mates 


Males of L. zephyrum usually fly around the nesting sites, although they 
visit flowers for nectar and sometimes mate there. In the field, males appar- 
ently are attracted by conspicuous objects. At the County Line nesting area, 
swarms of males were seen flying around large plants or bushes on which 
some of them rested at intervals. They also flew back and forth, 7 to 30 cm 
above the surfaces of banks or horizontal areas that contained nests. At the 


30 


PERCENT MALES 
De) 
O 


O 


Fic. 22. Percentages of pupae and teneral adults that are males. The total number of pupae and 
teneral adults examined is shown for each half-month. A few adult males are found 
at nesting areas in May although none were in the small sample of pupae examined 
during that month. 


416 THe UNiversiry ScIENCE BULLETIN 


Wakarusa nesting area, the greatest concentrations of males were found 
along the upper edges of the banks and among the hanging roots of plants 
that had been exposed by erosion. Males flew along the upper edges of con- 
cavities in the bank where I had removed nests. After a flood in 1962 that 
destroyed many nests in the left side of a bank, males no longer swarmed 
there (where formerly they were most abundant), but flew over the remain- 
ing nests in the right side of the bank, perhaps attracted by the heads of 
guarding females. In the insectary rooms, males established flight patterns 
around high, conspicuous objects, such as strips of corrugated cardboard on 
the wall (where they also rested), beams, the door frame, tall flowers, and 
the feeding stations that were suspended from the ceiling. They did not 
aggregate around the nesting boxes which were relatively low and incon- 
spicuous. Males generally zig-zagged upward along a cardboard strip or 
other object. On reaching the top, they usually flew to another landmark 


C 


Fic. 23. Flight pattern of a male in the insectary room, semi-diagrammatic. A, Viewed from 
above; B, Side view. Cardboard strips are indicated (C). 


Tue Lire CycteE AND BEHAVIOR OF THE PRIMITIVELY SocIAL BEE 417 


and then flew upward along it. Objects were not visited in any definite se- 
quence by individual males (Fig. 23). When lights above one side of an 
insectary room were turned off, males confined their flights to the lighted 
side. Males apparently flew independently, but when they were numerous, 
a swarm (the “sun-dance” of Rau, 1926) was seen around the landmarks. 
They usually faced toward the bank or other vertical object as they rapidly 
flew back and forth, and upward along its surface. Males were continuous- 
ly active from 8:00 a.m. until 7:00 p.m. in summer. In the insectary rooms, 
they usually did not begin activity until one and one-half hours after the 
lights were turned on, although some females were active within one-half 
hour. 


Matinc BEHAVIOR 


Mating occurs when females are standing or walking at one of the land- 
mark areas visited by males. In the field, mating usually occurred at the 
banks where males were flying, but in the insectary rooms, females were 
attacked when they alighted on feeding stations or other objects that attracted 
males. Males do not attack females in flight, although groups of males or 
single males often followed females that were about to enter their nests, and 
pounced on them as soon as they alighted. 

Attempted matings, lasting a few seconds, are more common than true 
matings, during which the bees remain together for 20 to 35 seconds (as 
determined by stopwatch). In June, six attempts were seen for every success- 
ful copulation observed. 

Visual stimuli apparently initiate attack by males, but actual copulation is 
probably controlled by chemical stimuli. Males pounce on many objects but 
only attempt to mate with females. When offered stunned or headless fe- 
males, they often persistently tried to copulate but were unsuccessful, perhaps 
because some cooperation by the female is necessary. 

Males, like those of other halictines (Bohart, 1950; Michener and Wille, 
1961), are not discriminating, and attack various objects of about the same 
size as females of L. zephyrum. They pounced at small pebbles, nail heads, 
female mutillids, millipeds, rhipiphorids and stationary males, but immedi- 
ately flew away. When a female was contacted, the male grasped her with 
his legs and immediately attempted to insert his genitalia. Whether mating 
occurs apparently depends on the amount of resistance by the female. Fre- 
quently, females let go of the banks or other surfaces on which they are walk- 
ing so that the pair falls or rolls a meter or more, and the males are often 
dislodged at this time. If the male continues to hold on, most females effec- 
tively resist him by curling away the abdomen, pushing him off with the 
hind legs or sometimes biting at his abdomen. Males usually let go after five 


418 Tue University Science BULLETIN 


seconds or less in such cases. Some females do not struggle or merely fall 
down. These are usually successfully mated. During copulation, females are 
passive, standing still or walking a few millimeters (they are able to fly if 
provoked). The male usually maintains a horizontal position above the 
female, with the end of his abdomen curled around the tip of the female’s 
abdomen. The male’s head is above the abdomen of the female, with the 
antennae slanting backwards, the legs lightly contacting the female or the 
nearby soil (Fig. 25). The male’s abdomen expands and contracts about once 
per second during copulation. If the female begins to move, the male some- 
times assumes a perpendicular position. The male terminates the copulation 
by abruptly detaching and immediately flying away. The female usually flies 
up as soon as the male leaves her but sometimes first cleans herself for a few 
seconds. Copulating pairs can be gently lifted up and sometimes aspirated 
without disturbing them. 

By placing laboratory-reared virgin females with clipped wings on a bank 
were males were numerous and allowing them to mate, it was determined 
that spermatozoa had not entered the spermathecae of females killed immedi- 
ately after mating. Spermatozoa were found in spermathecae of such females 
killed 36 hours after true mating, but not in those that resisted and were con- 
tacted by males for ten seconds or less. 

Samples of females were collected in the field in July as they were copulat- 
ing or struggling to avoid copulation, in order to determine if willingness to 
mate was related to age, condition of ovaries or presence of spermatozoa al- 
ready in the spermatheca. These females were killed immediately, before the 
newly received spermatozoa could be transferred to the spermatheca. Fe- 
males passively accepting the male for 15 seconds or more included: two in 
group E, one in group C (unfertilized), four in group D, and three in 
group B (fertilized). Indices of total wear of these females ranged from one 
to four. Females violently struggling to free themselves for ten seconds or 
more included: two in group E (unfertilized), four in group D, one in group 
E (fertilized) and one parasitized by a female Halictoxenos. Index of wear 
ranged from one to nine. No clear difference was seen between females ac- 
cepting copulation and those resisting it. 

When males were numerous, two or three were often seen simultaneously 
attempting to mate with a single female, which sometimes succeeded in 
avoiding all of them. Males were attracted to the heads of guards at nest 
entrances but were struck at by them. They unsuccessfully tried to copulate 
with guards that had blocked nest entrances with their abdomens. In 
autumn, males appeared to be more aggressive than during the summer. 
They frequently darted into unguarded nests. When entrances of occupied 
nests were artificially widened, males entered and could be seen attempting 


Tue Lir—E Cycte anp BEHAVIOR OF THE PRIMITIVELY SociAL BEE 419 


to copulate with the inhabitants. None were seen entering guarded nests 
with normal constricted entrances, however. Copulation within the nest is 
little known among bees but is usual in L. marginatum (Plateaux-Quénu, 
1959). Worn males of L. zephyrum sometimes taken in nests containing 
females probably had entered them and were not newly emerged bees. 

Males usually attacked females that hesitated before entering their nests. 
Such females included many young, inexperienced bees. Probably females 
entering hibernation are all inseminated because (a) males are numerous in 
fall; (b) males enter the nests; (c) the many young inexperienced females 
about to overwinter are attacked outside the nests; and (d) males are some- 
what more active than females at low fall temperatures. 

Males probably mate repeatedly. Individuals lived in vials for one week 
after mating. In the insectary room, marked individuals (with one antenna 


clipped) lived for 32 days. 


VARIATIONS IN COLOR 


Claude-Joseph (1926) observed considerable variation in color of the ab- 
domens of males, and to a lesser extent, of females, of Ruizantheda mutabilis. 
Males, but not females, of L. zephyrum as well as of various European 
species of Lasioglossum show great variation in color of the abdomen. Ab- 
domens of males range from almost entirely orange to uniform dark brown 
or black. Intermediates show different patterns and degrees of banding (Fig. 
24), apparently similar to patterns seen in R. mutabilis. 

The color of the abdomen is not associated with the age or size of the 


Fic. 24. Range of variation in color of the abdomens of males, semidiagrammatic. Black areas 
are represented by stippling. 


420) Tue University ScreNcE BULLETIN 


Fic. 25. Activities at the nesting site. A, Copulating pair; B and C, Guards watching a nearby 
mutillid (P. frigida). 


male. No overall seasonal differences in abdominal color were noted and 
males reared at 16° C were no darker than usual; temperature probably is 
not a factor in determining color. Males reared from pupae of the same sum- 
mer nest showed a wide range of color variation, and males of both the 
County Line and Wakarusa populations showed similar variations in color. 
These variations are probably genetically determined. 


Tue Lire Cycie anp BEHAVIOR OF THE PRIMITIVELY SoctAL BEE 421 


CONCLUSION 


Lasioglossum zephyrum is of special interest due to its weakly social 
behavior. The nest founding female (queen) lives usually for several weeks 
with her daughters which cooperatively work in their natal nests. As the 
daughters grow older, they gradually equal or replace the queen as egg- 
layers, many simultaneously acting as foragers. Apparently these daughters 
tend to behave like solitary bees because they provision cells when their 
ovaries contain large eggs; however, they do not individually establish new 
nests as do solitary species. 

In other social Hymenoptera, workers have small or atrophied ovaries 
and normally do not fully replace the queen as egg-layers. Probably in L. 
zephyrum there is little or no inhibition of ovarian development in workers 
by the queen. Physiological and biochemical studies to elucidate the mechan- 
isms of colony maintenance and integration in this species need to be made. 
Lasioglossum zephyrum may be a species taking early steps in the evolution 
of social behavior. 


ACKNOWLEDGMENTS 


I wish to thank Dr. C. D. Michener for guidance and the use of facilities 
during this study and for reviewing this manuscript. Drs. K. A. Stockham- 
mer and R. L. McGregor also kindly reviewed the manuscript. 

Mr. Roy Stecher generously permitted me to study a nesting area on his 
farm and Mr. William Kerfoot located another nesting area. 

I wish also to thank my husband, Lekh, for his self-sacrifice, patience and 
many helpful suggestions. 

This study was made possible by National Science Foundation grant 


G 11967 to Dr. C. D. Michener. 


LITERATURE CITED 


Batra, S. W. T. 1964. Behavior of the social bee, Lasioglossum zephyrum, within the nest 

(Hymenoptera: Halictidae). Insectes Sociaux 11:159-185. 

. 1965. Organisms associated with Lastoglossum zephyrum (Hymenoptera: Halictidae). 

Jour. Kansas Entomol. Soc. (in press). 

Bounart, G. E. 1950. Observations on the mating habits of halictid bees. Pan-Pacific Entomol. 
26:34-55). 

BonELui, B. 1948. Osservazioni biologiche sull’ “Halictus malachurus’ K. Boll. Inst. Entomol. 
Univ. Bologna 17:22-42. 

Borpas, L. 1894. Apparéil glandulaire des Hyménoptéres. Thése, Paris. Ann. Sci. Nat. 1-362. 


CiaupE-JosrpH, F. 1926. Recherches biologiques sur les Hyménoptéres du Chili (Melliféres), 
Ann. Sci. Nat., Zool., 9:113-268. 

Jay, S.C. 1961. The longitudinal orientation of larval honey bees (Apis mellifera L.) within 
cells. Anim. Behav. 9:230-231. 

Knerer, G. ano C. E. Arwoop. 1962. An annotated check list of the non-parasitic Halictidae 
(Hymenoptera) of Ontario. Proc. Entomol. Soc. Ontario 92:161-176. 

Krompien, K. V. 1938. Notes on the biology of Pseudomethoca frigida (Smith) (Hymenop- 
tera, Mutillidae). Bull. Brooklyn Entomol. Soc. 33:14-15. 


422 Tue University SciENCE BULLETIN 


LaBercr, W. E. anp O. W. Isaxson. 1963. The nest of Lasioglossum (Chloralictus) zephyrus 
(Smith) in Wyoming (Hymenoptera, Halictidae). Entomol. News 74:113-116. 
Lecewie, H. 1925. Zur Theorie der Staatenbildung. I Tiel: Die Biologie der Furchenbiene, 

Halictus malachurus K. Zeitschr. f. Morph. u. Okol. d. Tiere 3:619-684. 


Lin, Norman. 1964. Increased parasitic pressure as a major factor in the evolution of social 
behavior in halictine bees. Insectes Sociaux 11:187-192. 

Linstey, E. G., J. W. MacSwain ano R. F. SmirH. 1952. Outline for ecological life histories 
of solitary and semi-social bees. Ecology 33:558-567. 

MatysHey, S. I. 1935. The nesting habits of solitary bees. Eos 11:201-309. 


MarsHatt, G. E. anp L. I. Muscrave. 1937. A progress report on the Microlepidoptera of 
southern Indiana, and their parasites. Canadian Ent. 69:100-106. 


MEIwELL, O. 1958. A terrarium to investigate solitary and parasitic bees (Hymenoptera-Apidae) 
in an artificial climate. Univ. Bergen Arbok Naturvitenskapelig rekke Nr. 5:1-16. 


Micuener, C. D. 1961. Observations on the behavior of a burrowing bee (Amegilla) near 
Brisbane, Queensland (Hymenoptera, Anthophorinae). The Queensland Naturalist 
16:63-69. 

MicuHEneR, C. D., E. A. Cross, H. V. Day, C. W. RETTENMEYER AND A, Witte. 1955. Addi- 
tional techniques for studying the behavior of wild bees. Insectes Sociaux 2:237-246. 


Micuener, C. D. anp R. B. Lance. 1958. Observations on the behavior of Brazilian halictid 
bees (Hym., Apoidea) V. Chloralictus, Insectes Sociaux 5:379-407. 


MicHENER, C. D., R. B. LANGE, J. J. BIGARELLA AND R. SaLamuni. 1958. Factors influencing 
the distribution of bees’ nests in earth banks. Ecology 39:207-217. 


Micuener, C. D. ano A. Witte. 1961. The bionomics of a primitively social bee, Lasioglossum 
inconspicuum. Univ. Kansas Sci. Bull. 42:1123-1202. 


MircHet, T. B. 1960. Bees of the Eastern United States. Vol. I. North Carolina Agric. Expt. 
Station Tech. Bull. No. 141, 538 pp. 


Mueseseck, C. F. W., K. V. Krompein anp H. K. Townes. 1951. Hymenoptera of America 
north of Mexico. U.S.D.A. Agriculture Monograph No. 2, Washington, D.C., 1420 pp. 

Nott, J. 1931. Untersuchungen iiber die Zeugung und Staatenbildung des Halictus malachurus 
Kirby. Zeitschr. f. Morph. a. Okol. d. Tiere 23:285-368. 

O’Connor, H. G. 1960. Geology and ground-water resources of Douglas County, Kansas. State 
Geological Survey of Kansas Bulletin 148. Lawrence, Kansas. 200 pp. 


PEarson, J. F. W. 1933. Studies on the ecological relations of bees in the Chicago region. Ecol. 
Monogr. 3:377-441. 

Percivat, M. 1947. Pollen collection by Apis mellifera. The New Phytologist 46:142-173. 

PLATEAUX-QuéNu, C. 1959. Un nouveau type de société d’insectes: Halictus marginatus Brulle 
(Hym., Apoidea). Année Biologique 35:326-444, plates i-x. 

. 1963. Sur les femelles d’été de Halictus calceatus Scopoli. C. R. Acad. Sci. [Paris] 256: 
2247-2248. 

Rav, P. 1922. Ecological and behavior notes on Missouri insects. Trans. Acad. Sci. St. Louis 
24:1-71. 

. 1926. The ecology of a sheltered clay bank; a study in insect sociology. Trans. Acad. 
Sci. St. Louis 25:157-277. 


Robertson, C. 1926. Phenology of inquiline and nest-making bees. Psyche 33:116-120. 


. 1928. Flowers and Insects. Carlinville, Illinois, 221 pp. 


SaKacamt, S. F. AND K. Hayasuwa. 1958. Biology of the primitive social bee, Halictus duplex 
Dalla Torre I. Preliminary report on the general life history. Annot. Zool. Japonenses 
S151 155. 

—. 1960. Biology of the primitive social bee, Halictus duplex Dalla Torre II. Nest strue- 
ture and immature stages. Insectes Sociaux 7:57-98. 


SaKAGAMI, S. F. AND C. D. MicHENER. 1962. The nest architecture of the sweat bees. Univ. 
Kansas Press, Lawrence, 135 pp. 

Semicon, M. L. 1906. Recherches morphologiques et biologiques sur quelques melliféres soli- 
taires. Bull. Sci. Fr. Belg. [Paris] 40:281-442. 


SrocKHERT, E. 1923. Uber Entwicklung und Lebensweise der Bienengattung Halictus Latr. und 
ihrer Schmarotzer. Konowia 2:48-64, 146-165, 216-247. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


A REPORT ON THE OCCURRENCE AND 
DISTRIBUTION OF CLADOCERA AND 
COPEPODA IN LEWIS AND CLARK 
LAKE, SOUTH DAKOTA 


By 
Jerry C. Tash, George A. Swanson, and Richard E. Siefert 


eran peeecsteseresnencutnerees 


VoL. XLVI Paces 425-432 Fesruary 1, 1966 No. 11 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 425-432 Fesruary 1, 1966 No. 11 


A Report On The Occurrence And Distribution Of Cladocera 
And Copepoda In Lewis And Clark Lake, South Dakota 
By 


Jerry C. Tasu,! Greorce A. Swanson,” AnD RicHarp E, SreFERT? 


INTRODUCTION 


North Central Reservoir Investigations, Bureau of Sport Fisheries and 
Wildlife, Yankton, South Dakota, is conducting biological studies on Lewis 
and Clark Lake, a main-stem reservoir on the Missouri River separating 
South Dakota from Nebraska. This is the smallest reservoir on the Missouri 
River and the lowest point of control in the reservoir system. 

In South Dakota, most limnological investigations have emphasized 
vegetational, ichthyological, or physicochemical aspects (Over and Churchill, 
1927; Gastler and Moxon, 1948; Shields, 1957). A few studies include re- 
cords of microcrustacea (Galtsoff, 1924; Berner, 1951; Nash, 1959; Nelson, 
1960). No study has considered the distribution of microcrustacea within a 
single aquatic locality such as a reservoir. The distribution of microcrustacea 
in a reservoir correlated with stomach analyses of fishes collected in that 
habitat may provide information on trophic dynamics, feeding areas, and 
movements of some fishes. This report provides information on the occur- 
rence and distribution of Cladocera and Copepoda in Lewis and Clark Lake. 


METHODS 


One-hundred and four samples, collected during 1962-1963, were analysed 
for Cladocera and Copepoda. Eighty-six of these samples were collected 
‘Department of Zoology, Kansas University, Lawrence, Kansas. 


*North Central Reservoir Investigations, Bureau of Sport Fisheries and Wildlife, Yankton, 
South Dakota. 


426 THe University SciENcE BULLETIN 


from a boat using a Miller plankton sampler with a conical net of #10 mesh, 
and were taken either by towing the sampler or by hanging the sampler over 
the side of the boat while anchored in a swift current. The remaining 18 
samples were collected in a stationary sampler installed inside Gavins Point 
Dam, using a conical net of #10 mesh. 

Conductivity, expressed as micromhos at 25°C., was measured with an 
Industrial Instruments conductivity bridge, turbidity (ppm) with a Hach 
“Direct Reading” colorimeter, and temperature with an electronic thermome- 
ter. Zooplankton were identified with the aid of keys by Brooks (1957, 1959), 
Pennak (1953), Wilson (1959), and Yeatman (1944, 1959). 


DESCRIPTION OF COLLECTING STATIONS 
Shields (1957) and Schmullbach (1962) described Lewis and Clark Lake. 


The collecting stations were grouped according to river, littoral, or limnetic 
characteristic (Fig. 1). 
River AREAS 


Fifteen samples were analysed from collections taken during June-August, 
1963, in swift currents near the junction of the Missouri and Niobrara rivers, 
and from collections taken during October, 1962, and June-August, 1963, in 
downstream areas at Running Water and Springfield. The river level varies 
due to fluctuating discharges from Ft. Randall Reservoir upstream. 

At station 1, samples were collected along the south edge of the Missouri 
River about 100 meters upstream from the Niobrara River junction. Tur- 
bidity was 12-37 ppm., conductivity 785-820 micromhos/cm, and water tem- 
perature 15.8-24.2°C. The river in this area remains relatively clear year- 
around because streamflow is stabilized by Ft. Randall Reservoir about 35 
miles upstream. 

At station 20, samples were collected in water 1-2 meters deep at mid- 
channel of the Niobrara River, approximately 100 meters from the mouth. 
Turbidity was 190-1,250 ppm., conductivity 260-445 micromhos/cm, and tem- 
perature 21.1-26.4°C. 

At stations 2 and 4, samples were collected along the north edge of the 
Missouri River channel. Turbidity was 52-83 ppm., conductivity 735-800 
micromhos/cm, and temperature 16.9-24.9°C. 


LirroraL AREAS 


Thirty-five samples were analysed from collections taken during June- 
August, 1963, in creeks, bays, basins, and the western end of the lake. In- 
creased currents caused by rainfall influence all of these areas except the boat 
basins. 


OccuRRENCE AND DiIsTRIBUTION OF CLADOCERA AND CopEpPopaA 427 


Stations 7, 8, and 9 are creeks of 15-60 meters wide at their mouths, 1.5-2.4 
meters deep, have silt bottoms, lack aquatic vascular plants, and are protected 
from most wind action. Snatch Creek is land-locked during dry periods. 

Stations 12, 16, and 17 are open bays of 1.2-4.0 hectares, 5.4-6.1 meters deep, 
have silt or sand bottoms, and are influenced by wind. Only station 17 has 
aquatic vascular plants (two species of Potamogeton). 

Stations 3, 5, 14, and 18 are boat basins of 0.4-12.0 hectares, 1-5 meters deep, 
have silt bottoms, lack aquatic vegetation, and are protected from wind action. 
Boats entering and leaving create waves and mixing. 

Station 5, in the western end of the lake, is 1.5-3.1 meters deep except in 
the river channel, has a silt bottom, and contains scattered aquatic vascular 
plants of Scirpus sp., Typha sp., and Polygonum sp. This area is protected 
from wind, but it is influenced by currents coming from the Missouri River. 


Liwnetic AREAS 


Fifty-four samples were analysed from collections taken in five limnetic 
zones of the lake. All collections were taken during July-October, 1962, and 
during January-August, 1963, in water 6-20 meters deep. 

Stations 9, 10, 11, and 13 are in the old river channel that winds the length 
of the lake. Turbidity was 12-65 ppm., conductivity 740-805 micromhos/cm, 
and temperature 19.4-25.8°C. These areas are influenced by wind action. 

Station 15 is adjacent to the southwest side of the dam and water passes 
from it through the turbines. The area is protected from the wind. 


Species ACCOUNT 


Collections from the 20 stations yielded 37 species of Cladocera and 
Copepoda (Table 1). Eighteen were present in river samples, 13 from the 
Missouri River and 15 from the Niobrara River. Few individuals of any 
species were in Niobrara River samples, but many individuals per species 
were common in Missouri River samples. Most of the species found in the 
Missouri River samples probably originated in Ft. Randall Reservoir. Of the 
37 species distinguished, 6 were found in only littoral samples, 5 in only 
limnetic samples, 9 in both littoral and limnetic samples, and 7 in samples 
from all areas. 

Table 2 summarizes relative abundance of Cladocera and Copepoda from 
all collecting stations. Diaptomus forbest, D. clavipes, D. ashlandi, Cyclops 
bicuspidatus thomasit, Daphnia galeata mendotae, D. pulex, and Bosmina 
longirostris showed the greatest relative abundance throughout the year. 
Mesocyclops edax, Leptodora kindtu, Diaphanosoma brachyurum, Daphnia 
retrocurva, and Moina spp. were relatively abundant from May to October. 
All of the above species except L. kindtii were found in river samples, sug- 
gesting the possibility that river-introduced species may influence the popu- 
lation dynamics of the same species occurring in the lake. 


428 THe UNIversiry ScIENCE BULLETIN 


DISCUSSION 


Pennak (1957) reported that, at any one time, most limnetic habitats in- 
clude 2-4 species of Cladocera and 1-3 species of Copepoda. The limnetic 
zone of Lewis and Clark Lake contained as many as 13 species of Cladocera 
and 11 species of Copepoda during July, 1963. Collections taken during the 
winter of 1962-1963 contained 3 species of Cladocera and 4-6 species of Cope- 
poda in the limnetic samples, and the number of species during this period of 
time more readily fit Pennak’s proposal of a typical limnetic species composi- 
tion. Pennak’s proposal was based on studies conducted on lakes more stable 
than those through which large rivers flow, as Lewis and Clark. The com- 
plex species composition of Cladocera and Copepoda in Lewis and Clark Lake 
can be accounted for if the varieties of habitats and modes of species introduc- 
tion are considered. The large drainage area of the tributaries entering the 
lake introduce different species from a variety of habitats; these species may 
establish themselves in the niches provided by the diverse littoral and limnetic 
areas. 

Despite the rapid movement of water through the lake (it takes approxi- 
mately eight days for a volume of water equal to the holding capacity of the 
lake to pass through), the introduction of plankton by several large tributaries, 
and the extreme wind action, populations of Cladocera and Copepoda species 
have developed and maintained themselves. Past records (Galtsoff, 1924; 
Siebrass, 1960; Nelson, 1960) of species composition in the Missouri River 
and in Lewis and Clark Lake compared to the species listed in this paper 
indicate a high degree of stability in species composition of Cladocera and 
Copepoda of the river and lake. Only four of the 15 species of Cladocera 
and Copepoda recorded by Galtsoff (1924) for the upper Missouri River and 
only one of the 11 species of Cladocera and Copepoda recorded by Siebrass 
(1960) and Nelson (1960) for Lewis and Clark Lake were absent from the 
species listed in this report. 

Further discussion on the ecology of the Cladocera and Copepoda in 
Lewis and Clark Lake is beyond the scope of this report. A program is now 
in progress to collect data on the biological and environmental parameters 
affecting the entire biota. 


SUMMARY 


Distribution and relative monthly abundance of Cladocera and Copepoda 
during July-October, 1962, and January-August, 1963, have been studied in 
Lewis and Clark Lake, Gavins Point, South Dakota. Thirty-seven species of 
Cladocera and Copepoda were identified from 104 samples collected from 20 
different stations on Lewis and Clark Lake and on the Missouri and Niobrara 
rivers. Distribution of Cladocera and Copepoda were related to river, littoral, 
and limnetic areas. Only seven species occurred year-around while all others 
showed seasonal periodicity. 


OccuRRENCE AND DIsTRIBUTION OF CLADOCERA AND CopEPoDA 429 


PITERALURE CIleD 


Berner, L. M. 1951. Limnology of the lower Missouri River. Ecology 32:1-12. 

Brooks, J. L. 1957. The systematics of North American Daphnia. Memoirs of the Conn. Acad. 

of Arts and Science, Yale University Press, New Haven, vol. 13. 180 pp. 

. 1959. Cladocera, 587-656. In W. T. Edmondson (ed.), Freshwater Biology, 2nd ed., 

John Wiley and Sons., Incorp., New York. 

GarsorF, P. S. 1924. Limnological observations in the upper Missouri, 1921. U.S. Dept. Int., 
Fish and Wildl. Ser. Fish Bull. 39:347-438. 

GasTLeER, G. F. anp A. L. Moxon. 1948. Composition of Missouri River water samples taken at 
monthly intervals from May 15 to September 15, 1947. Proc. South Dakota Acad. Sci., 
27:32-35. 

Nasu, J. C. 1959. A taxonomic study of microcrustacea: the Branchiopoda, Ostracoda, and 
Copepoda of three gravel pit ponds in Clay County, South Dakota. Unpublished M.S., 
State University, South Dakota. 

Netson, W. R. 1960. Report of fisheries investigations during the sixth year of impoundment 
of Gavins Point Reservoir, South Dakota. South Dakota Dept. of Game, Fish and Parks, 
Dingell-Johnson Project F-1-R-10, Job No’s. 2, 3, 4. Unpublished mimeo. report. 56 pp. 

Pennak, R. W. 1953. Freshwater invertebrates of the United States. The Ronald Press Co. 

N. Y. 769 pp. 

. 1957. Species composition of limnetic zooplankton communities. Limnology and Ocea- 

nography 2(3):222-232. 

Over, W. H. anv E. P. Cuurcuitt. 1927. A preliminary report of a biological survey of the 
lakes of South Dakota. South Dakota Geol. Nat. Hist. Surv., Circular No. 29:18 pp. 

SCHMULBACH, J. C. AND H. A. SANDHOLM. 1962. Littoral bottom fauna of Lewis and Clark 
Reservoir. Proc. South Dak. Acad. Sci. XLI:101-112. 

SuieLbs, J. T. 1957. Report of fisheries investigations during the second year of impoundment 
of Gavins Point Reservoir. South Dakota Dept. of Game, Fish and Parks, Dingell-Johnson 
Project F-1-R-6. Unpublished mimeo. report. 34 pp. 

Steprass, H. H. 1961. The plankton of two large artificial impoundments in South Dakota. Un- 
published M.S., State University of South Dakota. 

Witson, M. S. 1959. Calanoida, 738-794. In W. T. Edmondson (ed.), Freshwater Biology, 
2nd ed., John Wiley and Sons, Incorp., New York. 

YeatmaNn, H.C. 1944. American cyclopoid copepods of the viridis-vernalis group, (Incl. a Desc. 

of Cyclops carolinianus n. sp.). Amer. Mid]. Nat. 32(1):1-90. 

. 1959. Cyclopoida, 795-815. In W. T. Edmondson (ed.), Freshwater Biology, 2nd ed., 

John Wiley and Sons, Incorp., New York. 


TOATY BIVIQOINT 


‘07 


Ya2xID pursiA ‘6 
ulseg Jeog puesio\, 


Avg a10yg ynos 
aIOYS YINOoG jo ysveq Avg 


"SaNS BuIdI]JOI ay Jo UONKICT IY} Surmoys sotueynqiy sy pue yr] YAe[O pure simoy] Jo Yoys mvuumviseiq 


81 
7A 
Ol 


We] JIOg SUIAPH “C] 
eULeW “pI 

Paly dauUNTT jauULYD “¢] 
Avg I0g SUIARD ‘Z| 

poly WQeuuN'yT pauueyyD “[] 


=—_ oo 


ayia 


voly snout] JeuueYD “OT 
BOTY IOUUITT [VUUPYD) *6 


yaoxig AsjreyO °8 
HREMO) MONARY 
9 


ulseg jog danoiddy ° 


ted WOISTAIp | 


ale oe ie ed 
£9 Soe ec: 


yuariny ¢— 


nN— —— 


ayer] yo pug 19189 AA. “¢ 
swonog ppysutdg fp 
ulseg iog yoedoig *¢ 


AK SS Se Se e008 


Jae Ay Suruuny 
JEUNIT MEMOS A 


C 
‘|! 


:puasaT 


Ha 
‘T “SI i 


OccuRRENCE AND DistRIBUTION OF CLADOCERA AND CoPEPODA 431 


Tasre 1. Distribution of Cladocera and Copepoda in study areas of Lewis and 
Clark Lake and in the Missouri and Niobrara rivers during July-October, 1962, 
and January-August, 1963. The numbers in the habitat columns indicate the 
numbers of separate collections from each type of environment in which the 
species was encountered. 


Habitat 

Species Composition River Littoral Limnetic 
Cladocera 
WeepiodoralRinaii (Rocke) O44) <2---2<-e-2 cen ecenceeecteeaneoeee —— 11 18 
Sida crystallina (O. F. Miller) 1785. -.....-.-.-2-----.22----2----- — 1 1 
Diaphanosoma brachyurum (Liéven) 1848)... 9 23 25 
Mipnmioeparvula Nordyce) VOOM eee ce een ec ene cance anes 3 1 14 
D. galeata Sars 1864 mendotae Birge 1918 ..............-----.- 10 3 29 
RDM CLOCUTUG ROLES: OSL) cee: cearecc oo cene ete eee ccs cee sean ecenaee sees 5 8 20 
D. pulex Leydid 1860 amend. Richard 1896 _.....-......- 1 11 38 
!D, GOMES EN Eo Mo OV 7 eee ee — — 1 
PV OLCUCTI SHRED IIA SATS M9 () 5) eee eeren seen carers eee — 4 — 
Ceriodaphnia reticulata (Jurine) 1820 ...............-------------- — _- 1 
MeepulchellanSars (SOL eco soso cee esvcesecsececevestewscwaenee== 1 4 6 
WANGEEEGE “S5O) Os} Ze ee ES See eee | 21 22 
Bosmina longirostris (O. F. Miller) 1785 -..... ieee ae 4 21 33 
Macrothrix laticornis (Jurine) 1820 -......-.....-.------.--------- _- Zz 1 
Camptocercus rectirostris Schgdler 1862 -.................-.-..---- 1 a — 
Leydigia quadrangularis (Leydid) 1860 -.................-..-...... -- 1 3 
AIDED CDGHTTB SENT Cl OPA ee ee ee ec -— ] — 
Ay RAGA GDIBAS ENS tel eee eee eee — 3 — 
Pleuroxus denticulatus Birge 1878 ...........----------------------- — 2 —_ 
miydorius elobosus: Baird V85 0) oe eee ee ce -ceeceecec cen aee es eeee eee _ 1 — 
(C., Swipe pens (Oy les INNS ey 7's) Se eee 1 6 15 
BRLOTEC LIES ARO ATS! Mls OD meeeesreree sees ier ee eee eee -- 1 9 
Copepoda 
Wrapiomus jorbest Wight W938) -2-22.-.-cc--2e.ccee--c8- egeceeeeeeeese=- 8 12 37 
Wmclavipes Schacht: 1897 <.22-cccscccc2-0-2ccceeeseesee sce La 10 4 339) 
RDMESICZIPS ROMA OLD ES MUO BD yoo: cece oes feee races, ee eee ee — ~- 2 
MUMS andiaNarshiiliG 95% eerste ee 8 16 31 
mmerciloides Willyeboroy [889) --.--..-c2- 1 ceeceece-eoeeeeee-eeeresee= 5 3} 16 
(Dla fRERIRTORTIS TS Were (ela RGA) ee ee ee — — 1 
Bay clopsiagiits: (Koch) 18382 ee os 5 12 
E. agilis montanus (Brady) ................-- ie eee eh oe — 8 7 
Trophocyclops prasinus (Fisher) 1860 ................-.-.-. oer: ] 3 5 
tayclops vernais Fischer W853) <-2.2-22.-<-0-22e0soee eee eee eee on ee 3 19 18 
C. bicuspidatus thomasit S, A. Forbes 1882 .....................- 12 15 42 
(erevaricans rubellus, Lilljeborg 1901 —_...---...-.--------.--- —- 1 3 
WMiesocyclops edax (S. A. Forbes) 1891] ..................-.....--- 10 2 DD 
Macrocyclops albidus (Jurine) 1820 _....00-.200----- AOL — — 2 
Ergasilus chautauquensis Fellows .................2..2-..-0---------- — 10 9 

Total number of Samples takem ...............................- 15 35 44 


432 Tue UNiversiry ScrENCE BULLETIN 


Taste 2. The monthly relative abundance of Cladocera and Copepoda in study 
areas of Lewis and Clark Lake and in the Missouri and Niobrara rivers during 
July-October, 1962, and January-August, 1963. Code to letters in the columns 
under the dates: A—more than 50 individuals per sample; C—10 to 50 individuals 
per sample; R—1 to 9 individuals per sample. 


Species 


Leptodora kindtit 
Sida crystallina 


Diaphanosoma brachyurum 
Daphnia galeata mendotae 
DS PORUU Ge eee ee 


Dy CLL OGULU Lire eee 
DVS Dulas een ane eee on cee eee 
DD SSCRO GI Crip ere es 


Scapholeberis kingi 


Ceriodaphnia reticulata 


Gapulchella 22 ieee ened eae 
IMO AE. S05 0s eee eee 
Bosmina longirostris ........---..- 
Macrothrix laticornis ..........-------- 
Cam ptocercus rectirOstris .....2..---- 
Leydigia quadrangularis ___........ 
VALOMGRGO SL GLA eee ne eee 
VARS EGLAIE (L1G) ee ee ee 
Pleuroxus denticulatus ...........---- 
Chydorus globosus __....--.---------- 
CHSPRGCTICU Se 
Aloncllgks pee es 
Digpromus: fOnbest 222. 22-2 2eeee 
WCLEULD CS: 07 ns ose ee ee 
WESTCI LES geen rete he Eee ee 
NGShla2d ee ee 
BESICtO1G Coen eee 
MG PLIGG US perce eee 
UGH CLO pss a oilis pee eee 


SissS 795 


ESR CEINE OTE LITA Speen eee ees re 
Trophyocyclops prasinus —......-.. 
Gyelopsvernalis =. 
C. bicuspidatus thomas. _............ 
C. varicans rubellus 
Mesocyclops €dax. ........:.--0-2--.--- 
Macrocyclops albidus ..............-..-- 
Ergasilus chautauquensis 


©) @) yes | 


Qi! 


e229) (OW eeat: 


a= @y NW ee) eel pos) Ml eel ee ey D 


Q 


Poh 


pola 


eH) Over! 


il ey oy 0 bel A 


am! 


AAD Pp! 


P(e) Eclit@u@y bel elites oles) teed teo) too} deel poate (ay ll Ce) tell p> @) @) Per | 


@) es) 2 p> eel be) es) O 


eo) 


ed @ el) el ll eel 1 er @) 1 ©) @y ee) ee) 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


MOSSES OF THE GREAT PLAINS AND 
ARKANSAS RIVER LOWLANDS OF KANSAS 


By 
Harold L. Smith 


VoL. XLVI Paces 433-474 Fepruary 1, 1966 No. 12 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Voi. XLVI Paces 433-474 Fepruary 1, 1966 No: 12 


Mosses Of The Great Plains And Arkansas River 
Lowlands Of Kansas 


By 


Harotp L. SmitTH 


Axstract. The purpose of this investigation is to increase the knowledge and 
understanding of the mosses of the Great Plains and Arkansas River Lowlands of 
Kansas. Climatic, edaphic, lithologic, and topographic factors have a very im- 
portant influence upon the growth and distribution of mosses in Kansas. From 
east to west in Kansas there is a decrease in mean temperature, in the length of the 
growing season, in precipitation, and in humidity, and an increase in the number 
of sunny days and wind velocity. These environmental factors operate together to 
produce a progressively less favorable habitat from east to west for the growth 
of mosses. Sandstone and the soil produced from it have a coarse texture and hence 
good moisture-retention properties; shale, limestone, chalk, and gypsum and the 
soils derived from them are fine-grained and hence contain a relatively small 
amount of moisture. Mosses are most common where there is enough topographic 
relief to provide north-facing slopes and ravines with favorable moisture condi- 
tions. Mosses are able to exist on the dry plains because of a protoplasm with a 
high resistance to desiccation and morphological features such as awns, involute 
and revolute margins, photosynthetic filaments, lamellae, and papillae which hold 
a capillary film of water onto the surface of the leaf. The pygmy mosses are 
mosses which are less than 5 mm in height and have a seta less than 2 mm long 
and a capsule which is round, less than 1 mm in diameter, with or without decidu- 
ous operculum, but always without peristome teeth. The ten genera of pygmy 
mosses in Kansas have arisen by specialized evolution in six different families. 
Four genera reveal step-wise evolutionary reduction from larger and morphologi- 
cally more complex relatives. Three genera, all with morphological adaptations of 
the leaves to hold a film of capillary water, are very important in the moss floras of 
the Great Plains. The exceptional ability to complete a life cycle on the dry plains 
is evidence that the pygmy mosses are highly adapted for existence on the prairies 
of the Great Plains. The mosses of the Great Plains and Arkansas River Lowlands 
are discussed as a unit for each of the following physiographic areas: the Smoky 
Hills, the Blue Hills, the Red Hills, and the High Plains (all in the Great Plains), 


434 Tue Universiry SciENcCE BULLETIN 


and the Wellington Lowland, the Great Bend Lowland, and the Finney Lowland 
in the Arkansas River Lowlands. Each physiographic area provides unique micro- 
habitats which select a characteristic assemblage of mosses. From east to west in 
Kansas there is a decrease in the importance of mosses of both eastern and western 
geographical affinities in the United States and an increase in importance of 
mosses of western affinities. A total of 117 species and varieties in the Great Plains 
and Arkansas River Lowlands with 23 new records for the state of Kansas are 
reported. Although knowledge of the mosses is still incomplete, this study should 
establish without doubt that the Great Plains and Arkansas River Lowlands of 
Kansas are areas of great bryological interest. 


INTRODUCTION 


The purpose of this investigation is to increase the knowledge and under- 
standing of the mosses of the Great Plains and Arkansas River Lowlands of 
Kansas. Previous collections of mosses from this area have been reported 
by McGregor (1950) and McGregor and Hartman (1956). In the present 
investigation, the western two-thirds of Kansas is approached from the 
physiographic standpoint with special emphasis on the microhabitats of the 
different rock outcrops of the various physiographic divisions. The discussion 
begins with a section on general physiography and is followed by one on 
ecology. Following, there is a morphological and an evolutionary considera- 


tion of the pygmy mosses, uniquely adapted to existence on the dry plains. | 


In the section on mosses of the Great Plains and Arkansas River Lowlands 
the moss floras of the following areas are discussed and compared to each 
other: the Smoky Hills, the Blue Hills, the Red Hills, and the High Plains 
(all in the Great Plains), and the Wellington Lowland, the Great Bend Low- 
land, and the Finney Lowland in the Arkansas River Lowlands. A list of 
species is provided. 


ACKNOWLEDGMENTS 


The writer wishes to express sincere appreciation to Dr. Ronald L. Me- 
Gregor, who suggested this problem, provided information of great assistance 
in carrying out field work, and directed the research in an atmosphere of 
academic freedom. The writer wishes to thank Dr. R. C. Jackson for reading 
the manuscript. The astute criticism by Dr. J. M. Jewett of the sections per- 
taining to geology is greatly appreciated. 


GENERAL PHYSIOGRAPHY 


Physiographically, Kansas is a plain (Schoewe, 1949). The surface of the 
plain slopes gradually eastward at a rate of 10 to 15 ft. per mile. The highest 
spot, located in Wallace County, is 4,135 feet above sea level; the lowest 
spot, located in Montgomery County, is slightly less than 700 feet above sea 
level. The surface of this plain is not everywhere flat and featureless. On the 


5 


MosskEs oF THE GREAT PLAINS AND ARKANSAS River LowLanps 435 


contrary, hills and valleys interrupt the flatness of the plain with slopes that 
are sometimes steep and precipitous. Local relief attains a maximum of 300 
ft. in a mile in areas along the major streams. Elsewhere it is seldom more 
than 100 ft. 

The classification of major and minor physiographic divisions of Kansas 
is given in Table 1 (Schoewe, 1949). The area of Kansas under consideration 
includes parts of two major physiographic provinces of the United States: 
the Great Plains Physiographic Province and the Central Lowlands Physio- 
graphic Province (Plate 1). The Great Plains Physiographic Province is 
represented in Kansas by two sections: the Dissected High Plains and the 
High Plains. Of the Central Lowlands the only section occurring in the study 
area is the Arkansas River Lowlands. 


Taste 1. Classification of major and minor physiographic divisions of Kansas 
(Schoewe, 1949). Used by permission of the author. Mosses have been collected 
from areas starred. 


Major Division Province Section Minor Division 
Interior Highlands Ozark Plateaus Spring field- 
Salem Plateaus 
Flint Hills 
Osage Cuestas 
Osage Plains Chautauqua Hills 


Cherokee Lowlands 


Kansas Drift Border 
*Central Lowland Dissected Till Attenuated Drift 
Plains Border 


*TInterior Plains 


*Wellington Lowland 

*Arkansas River McPherson Lowland 

Lowlands *Great Bend Lowland 
*Finney Lowland 


*Smoky Hills 
*Dissected High *Blue Hills 
*Great Plains Plains *Red Hills 


*High Plains 


THE GREAT PLAINS 


The Great Plains Physiographic Province in Kansas lies west of a line 
drawn from Washington County along the northern border of the state, 
southwestward to Harper County on the southern border. According to 
Fenneman (1931), the Great Plains is a broad belt of highland which slopes 
gradually eastward from the Rocky Mountains to the Central Lowland. The 
Great Plains surface is not continuous but is separated into a northern area 
and a southern area by the Arkansas River Lowland extension of the Central 


436 THe UNIversiry SCIENCE BULLETIN 


Lowland Province. The eastern boundary of the Great Plains north of the | 
Arkansas River is delimited by the eastern margin of the outcropping of | 
Cretaceous rocks (Plate 2). South of the Arkansas River the eastern boundary 
of the Great Plains is formed by the outcropping of resistant sandstones inter- 
bedded in soft shales of Permian age. The Great Plains Province is divided 
into two physiographic sections: the Dissected High Plains section and the 
High Plains section. 

The flatness of the Great Plains is due to the low angle of dip of the 
bedrock. According to King (1951) the structure of the rocks beneath the 
Great Plains is a homocline known as the Prairie Plains homocline, the 
eastern limb of the very asymmetric Great Plains syncline. The rocks of the | 
homocline dip westward at a low angle. In the western part of the Great | 
Plains the homocline is buried beneath Pliocene fluviatile deposits forming the — 
High Plains. In the eastern part of the Great Plains, erosion has removed the 
fluviatile mantle and has exposed sedimentary rocks of Cretaceous and Per- | 
mian ages. Because the different westward-dipping, sedimentary formations 
vary in resistance to erosion, differential erosion has here produced east-facing 
escarpments upheld by the more resistant layers of rock. This strip of eroded 

| 


cuestas along the eastern front of the High Plains is known as the Dissected 
High Plains. 


Tue Disstcrep Hicu Puiains 


The Dissected High Plains form the eastern part of the Great Plains. In 
this area erosion has removed the Pliocene fluviatile mantle and has exposed 
Cretaceous and Permian rocks. There are three minor divisions included in 
the Dissected High Plains: the Smoky Hills, the Blue Hills, and the Red 
Hills. The Smoky Hills and the Blue Hills divisions are located north of the 
Arkansas River and are underlain by Cretaceous rocks; the Red Hills division 
is located south of the Arkansas River and is underlain by Permian rocks. 

Smoky Hruis.—The Smoky Hills, a belt trending from Washington 
County southwestward to Rush County, form the eastern part of the Dis- 
sected High Plains north of the Arkansas River. These maturely dissected 
hills are carved in the Cretaceous Dakota Formation, the dark color of which 
gives the area its name. Numerous outlying hills and mounds occur in the 
Smoky Hills. 

Biuz Hitrs—The Blue Hills, so called because of the bluish haze which 
frequently invests them, form the western part of the Dissected High Plains 
north of the Arkansas River. This area extends from Republic and Jewell 
counties southwestward to Hodgeman and Finney counties. The rocks con- 
sist of inter-stratified limestones and shales of Cretaceous age. Two dissected 
cuestas occur in the Blue Hills. Other topographic features present are flat- 
topped buttes and small mesas. 


Rep Hitts.—The Red Hills represent the Dissected High Plains south of 


Mossts oF THE GREAT PLAINS AND ARKANSAS River LowLanps 437 


the Arkansas River. The name Red Hills comes from the red soil and rocks 
present in the area. Included in the Red Hills are parts of Kingman, Harper, 
Barber, Comanche, Clark and Meade counties. Underlying rocks are red 
shales, siltstones, and sandstones, along with gypsum and anhydrite, of Per- 
mian age. Topographic features include scarps, bluffs, mesas, and steep-sided 
buttes. 

‘Tue Hicu Prains 


The High Plains in Kansas is a locally featureless area lying between 2,000 
and 4,000 feet above sea level. According to Fenneman (1931), they are rem- 
nants of a former great fluviatile plain which stretched from the mountains 
on the west to the Central Lowland. Rivers have cut broad valleys into this 
plain, forming a dissected plateau with extensive flat uplands. The High 
Plains are generally underlain by unconsolidated or poorly cemented de- 
posits of Pliocene (Ogallala) and Pleistocene ages. The lime-cemented part 
of the Ogallala Formation is known as mortar beds. These mortar beds up- 
hold cliffs which are an important topographic feature of the High Plains. 
Cretaceous chalk and shale formations are exposed along the river valleys. 
Schoewe (1949) includes the area of outcrop of these Cretaceous rocks in the 
High Plains; Fenneman (1931) places this area, along with the Smoky Hills, 
the Blue Hills, the Red Hills, and the Great Bend Lowland, in the Plains 
Border. Along the river valleys there occur Pleistocene alluvium and terraces 
and low sand dunes. 


ARKANSAS RIVER LOWLANDS 


The Arkansas River Lowlands are low-lying areas extending east and west 
along the Arkansas River. The surface rocks are of two major kinds: un- 
consolidated gravels, silts, and clays of Pleistocene age and red silty shales of 
Permian age. The Arkansas River Lowlands are divided into the Welling- 
ton Lowland, the McPherson Lowland, the Great Bend Lowland, and the 
Finney Lowland. All but the McPherson Lowland are considered. 

WELLINGTON Lowtanp.—The Wellington Lowland lies mostly in Sumner 
and Harper counties with small areas in Barber, Sedgwick, and Reno counties. 
Unlike the flat topography of the rest of the Arkansas River Lowlands, the 
Wellington Lowland topography is decidedly rolling. The surface is under- 
lain by Permian shales. 

Great Benp Lowtanp.—The Great Bend Lowland is an undulating plain 
of little relief, extending from Dodge City in Ford County eastward to Hut- 
chinson in Reno County and southeastward to Wichita in Sedgwick County 
and Arkansas City in Cowley County. The surface of the lowland is covered 
by Pleistocene dune sand, terraces, and alluvium. 

Finney Lowianp.—The Finney Lowland comprises all of the Arkansas 
River Valley in Kansas west of Dodge City. The topography of the inner 


438 THe UNIversiry ScIENCE BULLETIN 


valley or valley bottom is essentially flat. The outer valley includes the 
terraces and land sloping toward the river but lying below the level of the 
High Plains. An extensive area of sand dunes occurs in the outer valley. 


ECOLOGY 


As an area for the growth of mosses, the Great Plains has long been mis- 
understood. Dr. Henry S. Conard, a well-known bryologist who has done 
extensive moss collecting in a state as close to Kansas as Iowa, has written 
(1956): “There is no moss among the grasses of the Tallgrass Prairies or the 
Shortgrass Plains. Moist spots or bare spots or wooded spots among these 
Plant Associations have their mosses.” Yet, in many areas of the Great Plains 
mosses abound in spots that are neither moist, bare, nor wooded. Many mosses 
grow on soil on slopes of hills and ravines where the grass cover does not pro- 
duce too much shade. Other mosses grow on rocks forming cliffs or out- 
cropping along the sides of hills or in ravines. A few mosses grow on the 
bark of trees; a few others grow on sand dunes with grass or sagebrush cover. 

Climatic factors have a very important influence upon the growth and 
distribution of mosses in Kansas. The authority for the climatic data of Kan- 
sas is S. D. Flora (1948). The annual mean temperature for the state is 
55.0°F. The normal annual mean Fahrenheit temperatures for Cheyenne, 
Morton, Washington, and Sumner counties, located respectively at the north- 
west, southwest, northeast, and southeast corners of the study area are as 
follows: 


Washington 54.3 


Cheyenne all 
9 Sumner 58.0 


Morton : 


Wr 
Wb 


The average length of the growing season in days for the same counties is: 


Cheyenne 160 Washington 171 
Morton 179 Sumner 185 
The average number of clear days per year increases from 162 at Wichita 
at the eastern border of the study area to 177 at Dodge City in the High 
Plains. The average annual precipitation decreases from 30 inches at Wichita 
to 17 inches at the western border of the state. The mean relative humidity 
in the spring for Wichita and Dodge City is as follows: 


March April May 
Wichita yess ee 60% 59%, 64% 
Dodge City .......... Dis 56% 60% 


The average hourly wind velocity increases from 12.1 at Wichita to 12.9 at 
Dodge City. Thus, from east to west there is a decrease in mean tempera- 
ture, in the length of the growing season, in precipitation, and an increase 
in the number of sunny days and wind velocity. These environmental factors 
operate together to produce a progressively less favorable habitat from east 
to west for the growth of mosses. 


MosskEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 439 


Edaphic and lithologic factors, as well as climatic factors, have an im- 
portant effect upon the growth and distribution of mosses in Kansas. Because 
soils differ in texture and moisture-retention properties, they differ in their 
capacity to support the growth of mosses. Soils derived from sandstone have 
a coarse texture due to the presence of sand grains and moisture retention is 
good. Soils derived from shale and limestone have a fine texture with poor 
moisture-rention properties. However, the moisture retention of soils de- 
rived from shale and limestone is often improved by the presence of thin 
layers of fine-grained sands interbedded in the shale and limestone. Like 
soils, the rocks differ in their ability to support growth of mosses. Because 
shale, limestone, gypsum, and chalk are fine-grained and relatively non- 
porous, they contain relatively small amounts of water. Sandstone and lime- 
cemented sands are coarser grained, more porous, often friable, and contain 
a better supply of water. Except where they are situated in ravines where 
there is seepage of water to moisten the rocks, shale, limestone, gypsum, and 
chalk are devoid of mosses. In contrast, sandstone and the lime-cemented 
sands support the growth of mosses, not only in ravines, but even where the 
rocks are exposed on wind-swept ledges and cliffs. 

Because of the heat and dryness of the Great Plains in the summer, mosses 
growing there must have the ability to withstand extreme desiccation. An 
experiment was conducted to see what would result if mosses that had been 
stored in herbarium packets three to 20 months were placed on moist peat 
or sand. Of 32 species tested, the following 17 showed renewed growth: 
Amblystegium juratzkanum, Astomum muhlenbergianum, Bryum argen- 
teum, B. bicolor, B. caespiticium, Ceratodon purpureus, Desmatodon plin- 
thobius, Grimmia apocarpa, G. laevigata, G. pulvinata, G. wright, Hedwigia 
ciliata, Leskea gracilescens, Phascum cuspidatum var. americanum, Physco- 
mitrium pyriforme, Pterigoneurum subsessile, and Ptychomitrium incurvum. 
All species except Desmatodon plinthobius and Grimmia wright produced 
new rhizoids and axillary shoots from the pieces of shoots that had been 
dried. Desmatodon plinthobius and Grimmia wright started new growth 
from the apex of the old stem. In some cases renewed growth started within 
two weeks after the dried mosses were placed upon a moist substrate. In 
other cases it was a month before growth was renewed. The presence of 
mosses on the Great Plains year after year and the results of the above ex- 
periment indicate that they have protoplasm with a very high resistance to 
desiccation. 

In addition to a very resistant protoplasm, mosses have various morpho- 
logical adaptations of the leaves which tend to hold a capillary film of water 
on their surfaces (Goebel, 1905). These morphological adaptations are awns 
or hairpoints, involute and revolute margins, photosynthetic filaments, lamel- 
lae, and papillae. The most common mosses with smooth to slightly-toothed 


440, Tue UNIversiry ScIENCE BULLETIN 


awns are Bryum argenteum, B. caespiticium, Desmatodon plinthobius, 
Phascum cuspidatum var. americanum, Grimmia rauei, and G. wright. 
Mosses with toothed awns, the irregular surfaces of which can hold consider- 
able capillary water, are Grimmia laevigata, Pterigoneurum subsessile, and 
Tortula ruralis. The strongly involute leaves of the genera Astomum and 
Weissia form small receptacles which can serve as a reservoir for water. A 
film of capillary water can be held by revolute leaf margins, a morphological 
adaptation common in the Pottiaceae. The genus dloina has chains of 
photosynthetic cells attached to the upper surface of the leaves. These fila- 
ments serve both to increase the photosynthetic area and to increase the 
amount of capillary water held by the leaf. Lamellae, thin sheets of cells 
standing perpendicular to the upper surface of the leaf, perform a similar 
function in the genus Pterigoneurum. Papillae, small knobs on the outer 
cell wall, are an effective capillary device. Papillae are present on the leaves 
of many species of mosses in the Great Plains. Two species, Hedwigia ciliata 
and Weissia glauca, have branched papillae which are especially effective in 
holding a film of capillary water. These morphological adaptations enable 
the moss leaf to obtain moisture from dew, mist, and light rain. Since pre- 
cipitation on the Great Plains is often in the form of mist or light rain, these 
morphological adaptations play a vital role in the existence of mosses in the 
area. 

All of the most common mosses of the Great Plains have one or more of 
the morphological adaptations discussed in the paragraph above. Bryuwm 
caespiticium and B. argenteum, both common in the Great Plains and found 
even growing under sagebrush on sand dunes, have awned leaves. In B. 
argenteum the upper leaf cells are dead and form a protective mantle for the 
living cells of the lower part of the leaf (Goebel, 1905). Grimmua laevigata, 
the characteristic moss of dry wind-swept rocks in the Smoky Hills, has 
leaves with toothed awns. Grimmia rauei and G. wright with somewhat 
spinulose leaf awns occur on the dry wind-swept Ogallala rocks in the High 
Plains. Barbula unguiculata, common on dry limestone outcrops in the Blue 
Hills and on the Ogallala Formation in the High Plains, has papillose leaf 
cells and revolute leaf margins. The morphological adaptations found in 
Phascum cuspidatum var. americanum, a moss distributed from the Smoky 
Hills and the Red Hills west to the High Plains, are awns, papillose leaf 
cells, and revolute leaf margins. Tortzla ruralis has leaves with papillose leaf 
cells, revolute margins, and toothed awns, features of importance in adapting 
this moss to its habitat on dry ravine slopes on the Ogallala Formation in the 
High Plains. Prerigoneurum subsessile, a moss which grows most abundantly 
under sagebrush plants on sand dunes, has leaves with lamellae and toothed 
awns. In both Astomum muhlenbergianum and Weissia controversa papillose 
leaf cells and strongly involute leaf margins are combined in a leaf with an 


MosseEs OF THE GREAT PLAINS AND ARKANSAS River Low Lanps 44] 


especially good morphology to hold capillary water. These two mosses are 
of great importance on the Great Plains. From the foregoing discussion it 
can readily be seen that mosses of the Great Plains show considerable diver- 
sity in their adaptations to hold capillary water. 


Pycmy Mosses 


The term “pygmy mosses” is here used to denote a group of unrelated 
mosses with the same growth form. The plants, gametophytes and sporo- 
phytes, together are less than 5 mm in height; the setae are less than 2 mm 
long; the capsule is round, less than 1 mm in diameter, cleistocarpous or with 
deciduous operculum, but always without peristome teeth. Thus, the pygmies 
are minute mosses with the leaves of the gametophyte surrounding the im- 
mersed capsule. On the Great Plains of Kansas ten genera have been found 
which can be considered pygmy mosses: Archidium, Ephemerum, Aschisma, 
Pterigoneurum, Pleuridium, Bruchia, Pyramidula, Phascum, Acaulon, and 
Astomum. The evolutionary relationships of the genus Archidium are 
obscure; Conard (1956) places this genus in the monogeneric Archidiaceae. 
Similarly, the genus Ephemerum is placed in the monogeneric Ephemeraceae. 
The relationships of the genera Aschisma and Pterigoneurum in the Pottia- 
ceae are unclear. However, within the genus Pterigoneurum there is an 
evolutionary trend from P. ovatum with an exserted, cylindrical capsule to 
P. subsessile with immersed spherical capsule. The genera Pleuridium and 
Bruchia belong to the Ditrichaceae and the Dicranaceae, respectively (Bryan, 
1956). The remaining four genera of pygmy mosses reveal a step-wise 
evolutionary reduction from larger and morphologically more complex rela- 
tives. Three such specialized evolutionary trends are illustrated by mosses 
growing on the Great Plains: one in the Funariaceae, one in the subfamily 
Pottioideae of the Pottiaceae, and one in the subfamily Trichostomoideae of 
the Pottiaceae. Each trend begins with a moss with a large gametophyte, 
long seta, and a large, cylindrical, operculate capsule with single or double 
peristome. From this stage different genera and species have undergone 
diminution in size and loss of structures until the pygmy condition is attained. 
Since the particulars of each trend differ in detail, each will be discussed 
separately. 

The most morphologically complex species in the Funariaceae is Funaria 
hygrometrica with nodding annulate capsules with both inner and outer peri- 
stomes well developed. Funaria flavicans is similar, but the inner peristome 
teeth are short and truncate. Funaria americana has a peristome like that of 
FP. hygrometrica, but the capsule is smaller, almost erect, and lacks an annulus. 
In Entosthodon drummondii the capsule is erect and has only a single peri- 
stome. Further reduction is illustrated by Physcomitrium pyriforme in which 
the setae are shorter and the capsule lacks peristome teeth. Physcomitrium 
pyriforme var. serratum has setae about 2 mm long and thus approaches the 


442 Tue UNIversiry ScIENCE BULLETIN 


pygmy habit. The pygmy condition in the Funariaceae is represented by 
Pyramidula tetragona, a moss with a short seta and operculate capsule covered 
by the 4-angled calyptra. 

The trend in the subfamily Pottioideae of the Pottiaceae begins with 
Tortula ruralis which has long seta, and a cylindrical, operculate capsule with 
long, spirally twisted peristome teeth. Desmatodon obtusifolius is a smaller 
moss with shorter, untwisted peristome. In some species of the genus Pottia 
the peristome is lacking, although the capsule is operculate. The pygmy 
mosses represented in this trend are Phascum cuspidatum var. americanum 
and Acaulon rufescens. Both are perfect examples of the pygmy habit with 
small, inoperculate, spherical capsules immersed in a bud-like cluster of 
gametophytic leaves. 

The forms present in the subfamily Trichostomoideae of the Pottiaceae 
illustrate in a striking manner the specialized evolution of normally developed 
mosses to the pygmy condition. The trend begins with Barbula unguiculata, 
a species with seta 5 mm long, a cylindrical, operculate capsule with twisted 
single peristome teeth. Didymodon tophaceus is similar, but in this species 
the peristome is shorter and not twisted. The genus Wezssia has species 
showing two stages in reduction from the condition represented by Barbula 
and Didymodon. Weissia controversa has seta 4 mm long and an operculate 
capsule with shorter peristome. Wezssia microstoma is still more reduced; 
this species has seta 1 mm long and an operculate capsule with a closing 
membrane in place of the peristome teeth. Although the operculum sepa- 
rates as a unit when the capsule is moist, spores are released from capsules 
maturing in the dry prairies in June by irregular breaking of the capsule wall. 
This loss of function of the operculum leads in a natural way to the next 
genus. The genus Astomum represents the pygmy habit in the Trichosto- 
moideae. In all three species the capsule is inoperculate; the spores are not 
released until the wall of the capsule is broken. However, in Astomum 
ludovicianum the apiculus, the tip end of the capsule, is delimited from the 
rest of the capsule by a distinct ridge. Thus, the apiculus represents a vesti- 
gial operculum and the ridge represents a vestigial annulus. In A. phascoides 
and A. muhlenbergianum the apiculus is merely the pointed tip end of the 
capsule. In A. /udovicianum the short-cylindrical capsule is up to 1 mm long, 
the apiculus up to 0.3 mm long, and the seta is about 0.5 mm long. In A. 
phascoides the spherical capsule is 0.7 mm in diameter, the apiculus 0.4 mm 
long, and the seta is 1.5 mm long. Thus, compared to A. ludovicianum, A. 
phascoides has a longer seta and apiculus, but a smaller capsule with no 
dividing line separating the apiculus from the rest of the capsule. A. muhlen- 
bergianum represents the most reduced condition in this trend. The seta, less 
than 0.2 mm long, supports a spherical capsule 0.4 mm in diameter with an 
apiculus 0.1 mm long. 


Mossts OF THE GREAT PLAINS AND ARKANSAS River LowLaAnps 443 


Although seven genera of the pygmy mosses are rare, three genera, all 
in the Pottiaceae, are of great importance in the moss flora of the Great Plains. 
Astomum muhlenbergianum is a common moss in the Smoky Hills, the 
Red Hills, the Blue Hills, the Wellington Lowland, and the Great Bend 
Lowland, and is present in the High Plains. Only in the Finney Lowland is 
this species absent. Phascum cuspidatum var. americanum is present in all 
seven physiographic regions in the area studied except for the Blue Hills 
and is an important moss in the Smoky Hills and the Wellington Lowland. 
Pterigoneurum subsessile is an important moss in the Finney Lowland, is 
fairly common in the High Plains, and is present in the Red Hills, the Blue 
Hills, and the Wellington Lowland. The leaves of these pygmy mosses have 
morphological adaptations to enable the leaf to hold a film of capillary water. 
Astomum muhlenbergianum has leaves with papillose cells and involute 
margins. The leaves of Phascum cuspidatum var. americanum have papillose 
cells, revolute margins, and awns. Pterigoneurum subsessile has leaves with 
lamellae and toothed awns. With the exception of Grimmia rauer and G. 
wright, which could be considered pygmy mosses except for the peristome 
teeth, these three pygmy mosses are the only ones that regularly fruit through- 
out the Great Plains of Kansas. This exceptional ability to complete a life 
cycle on the dry plains is evidence that these pygmy mosses are highly 
adapted for existence on the prairies of the Great Plains. 

Bryan (1956, 1956a) has reported chromosome numbers of four genera 
of pygmy mosses. Species of Bruchia have chromosome numbers of n=—14, 
n=15, and n=28. Because of the similarities of the chromosomes of Bruchia 
and Trematodon (n=28), the genus Bruchia is removed from the Ditri- 
chaceae and placed in the Dicranaceae. The similarities of the chromosomes 
of Ditrichum pallidum (n=26) and the pygmy moss Plearidium subulatum 
(n=13) substantiate their classification in the Ditrichaceae. The genera 
Weissia and Astomum are closely related taxonomically and cytologically. 
Weissia controversa has 13 bivalents, including one M-chromosome. The 
chromosomes of Astomum ludovicianum (n=13) are of similar size al- 
though two are somewhat smaller. Astomum muhlenbergianum is a sporo- 
phytic tetraploid with 26 bivalents, including one large elongated bivalent 
resembling the M-chromosome of Weisssia controversa. Acaulon rufescens 
(n=26) and Phascum cuspidatum var. americanum (n=26) are also sporo- 
phytic tetraploids. 


MOSSES OF THE GREAT PLAINS AND ARKANSAS RIVER 
LOWLANDS 


The mosses of the Great Plains and Arkansas River Lowlands are dis- 
cussed as a unit for each of the following physiographic areas: the Smoky 


Hills, the Blue Hills, the Red Hills, the High Plains of the Great Plains, the 


444 Tue University SCIENCE BULLETIN 


Wellington Lowland, the Great Bend Lowland, and the Finney Lowland of 
the Arkansas River Lowlands. The discussion of the mosses of each area 
follows the same pattern. First, there is a treatment of geology and physiogra- 
phy in greater detail than in the section on General Physiography. Then, 
there is an ecological comparison of each area to the Smoky Hills where 
conditions for the growth of mosses are the most favorable of all regions 
under study. Next, the flora is analyzed from the standpoint of how many 
species are both eastern and western in their distribution in the United States, 
how many are eastern, and how many are western. Then, the most common 
species are listed and their distributional affinities stated, after which the rest 
of the mosses are listed. Following, there is a discussion of the various mosses 
and their microhabitats. Finally, the mosses on the microhabitats not char- 
acteristic of the area in general are considered. At the end of the section 
there is a comparison of the floras of the different physiographic areas. Plate 
1 shows the localities from which mosses have been collected by the author. 


MOSSES OF THE SMOKY HILLS 


The Smoky Hills form the eastern part of the Dissected High Plains 
north of the Arkansas River. Outcropping rocks belong to the Dakota 
Formation of Cretaceous age. The Dakota Formation is divided into the 
lower Terra Cotta Clay Member and the upper Janssen Clay Member (Plum- 
mer and Romary, 1947). The Terra Cotta Clay Member consists of massive 
clay, silt, and sandstone and comprises approximately the lower two-thirds 
of the Dakota Formation. The Janssen Clay Member includes beds of lig- 
nite, massive clay, silt, and some shale. Approximately three-fourths of the 
thickness of the Dakota Formation is clay. The case-hardened, iron oxide- 
cemented silt beds and sandstone lenses are more resistant to erosion than 
the clays and form the cap rock of steep-sided hills of irregular shape. The 
Dakota Formation is an important aquifer. Springs and seeps are numerous 
due to the abundance of saturated sandstone outcropping. An area of special 
interest is Rock City, a group of concretions located 2, miles southwest of 
Minneapolis in Ottawa County. According to Schoewe (1949), there are 
over 200 concretions, spherical to irregular, and 8 to 27 ft. in diameter. The 
sandstones forming the concretions are distinctly cross-bedded. The con- 
cretions are areas in the sandstone where the quartz sand grains have been 
cemented together by calcite. Because of the resistant nature of the concre- 
tions, weathering and erosion have removed the uncemented, softer sand- 
stone, thus exposing the concretions. Pleistocene alluvium and alluvial ter- 
races and dune sand occur in the river valleys. 

More species of mosses grow in the Smoky Hills than in any other physio- 
graphic area in the region under study. The reasons for the comparative 
richness of the moss flora are the following: 


MosskEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 445 


1. Due to their eastern position, the Smoky Hills receive a relatively large 


amount of precipitation. 


2. The dissected topography has many ravines and north-facing hill slopes 


favorable to the growth of mosses. 


3. The sandstone is porous and contains a good supply of water. As an 
important aquifer, the Dakota Formation provides water for many 
springs and seeps on hill slopes and in ravines. 

4. The soils formed from the Dakota Formation contain fine to medium 
sand which provides a comparatively good soil texture as far as moisture 


retention is concerned. 


All of the above factors operate together to make the Smoky Hills the most 
favorable area in the Great Plains of Kansas for the growth of mosses. 

Of the 81 species and varieties collected in the Smoky Hills, 77°% are dis- 
tributed in both the eastern and the western United States. Seventeen per 
cent are eastern species extending as far west as Kansas. Six per cent are 


of western affinity. 


Ten of the 81 species and varieties are especially important in the moss 


flora: 


Astomum muhlenbergianum 
Bryum argenteum 

B. caespiticium 

Campylium chrysophyllum 
Chamberlainia salebrosa 


Desmatodon obtusifolius 

Grimmua laevigata 

Phascum cuspidatum var. americanum 
Physcomitrium pyriforme 

Weissia controversa 


All of these mosses are distributed in both the eastern and the western 


United States. 


Thirteen other species are common to fairly common in the Smoky Hills: 


Astomum ludovicianum 
Atrichum angustatum 
Barbula unguiculata 
Campylium hispidulum 
Ceratodon purpureus 
Ditrichum pallidum 
Entodon seductrix 


Grimmia apocarpa 
Hedwigia ciliata 
Homomallium adnatum 
Mnium cuspidatum 
Pohlia nutans 
Ptychomitrium incurvum 


Fifty-eight species and varieties are rare: 


Acaulon rufescens 
Amblystegium juratzkanum 
A. serpens 

A. varium 

A. varium var. ovatum 
Atrichum undulatum 

A. undulatum var. minus 
Barbula fallax 

Bryum bicolor 

B. cuspidatum 

B. pendulum 

B. pseudotriquetrum 
Campylium chrysophyllum var. brevifolium 
Chamberlainia acuminata 

C. oxyclada 


Desmatodon plinthobius 
Dichodontium pellucidum 
Dicranella varia 
Didymodon rigidulus 
Drepanocladus aduncus var. polycarpus 
Entodon compressus 
Ephemerum sp:nulosum 
Eurhynchium hians 

E. serrulatum 

Fabronia imperfecta 
Fissidens bryoides 

F. obtusifolius 

F. osmundioides 

F. sublimbatus 

Funaria americana 


446 Tue University SCIENCE BULLETIN 


F, flavicans Physcomitrium hookert 
F, hygrometrica P. pyriforme var. serratum 
Grimmia plagiopodia Plagiothecium geophilum 
G. pulvinata Platygyrium repens 

G. wright Pleuridium subulatum 
Hygroamblystegium iriguum Pohlia wahlenbergu 
Leptobryum pyriforme Polytrichum juniperinum 
Leptodictyum riparium P. piliferum 

L. trichopodium Pylaisia selwynu 

Leskea gracilescens Pyramidula tetragona 

L. obscura Thuidium virginianum 
Orthotrichum pumilum Timmia megapolitana 
O. strangulatum Tortula pagorum 
Philonotis longiseta T. ruralis 


Mosses are commonly found in ravines and on northfacing hill slopes. 
Most of the 81 species and varieties grow on rocks and soil in such situations. 
Some, however, grow in distinctive habitats. One striking example is the 
common occurrence of little hummocks of Grimmia laevigata on dry, wind- 
swept rocks on the sides and tops of hills. Also growing on dry rocks are 
Weissia controversa, Desmatodon obtusifolius, and Hedwigia ciliata. Plants 
of the genus Fissidens are usually found on rock ledges or soil on vertical 
ravine banks where grass leaves hang over and form a protective cover. 
The various pygmy mosses most often grow on small level areas on hill slopes. 
At the opposite ecological extreme from the dry wind-swept rocks, there are 
shaded spring areas characterized by the presence of the following corticolous 
species: Amblystegium juratzkanum, A. serpens, A. varium, Hygroambly- 
stegium irriguum, Leptodictyum riparium, L. trichopodium, Leskea graciles- 
cens, and Orthotrichum pumilum. 

Some interesting mosses have been found growing on the sandstone con- 
cretions at Rock City. The cross-bedded sandstones of the concretions have 
been weathered to produce many ledges and crevices, ideal places for the 
growth of mosses. In a paper on the flora of Rock City, Horr (1937) men- 
tioned two mosses but did not specify them by name. The following mosses 
have been found growing on the sandstone concretions: Bryum argenteum, 
Campylium hispidulum, Desmatodon plinthobius, Fissidens obtustfolius, 
Grimmia apocarpa, G. wrightit, and Weissia controversa. Of special interest 
is the presence at Rock City of Desmatodon plinthobius and Grimmua 
wrightii, two species of mosses, unknown elsewhere in the Smoky Hills, 
normally growing in Kansas only on the lime-cemented rocks of the Ogallala 
Formation in the High Plains. The occurrence of these mosses at Rock City 
is to be explained by the calcite cement of the concretions. 

Dune sand occurs in small areas along the river valleys in the Smoky Hills. 
Mosses growing on these dunes are Bryum argenteum, B. caespiticium, Phas- 
cum cuspidatum var. americanum, and Weissia controversa. On a north- 
facing vertical roadbank covered by grass and shaded by a box elder tree the 
following species were found: Amblystegium serpens, Astomum muhlen- 


MossEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 447 


bergianum, Barbula unguiculata, Bryum argenteum, B. caespiticium, Cam- 
pylium hispidulum, Chamberlainia salebrosa, Eurhynchium  serrulatum, 
Funaria hygrometrica, and Mnium cuspidatum. 


MOSSES, OF THE BLEUE HILLS 


The Blue Hills form the western belt of the Dissected High Plains north 
of the Arkansas River. The underlying rocks are inter-stratified limestones 
and shales belonging to the Gulfian Series of Cretaceous rocks. In ascending 
order the formations are the Graneros Shale, the Greenhorn Limestone, the 
Carlile Shale, and the Fort Hays Limestone. According to Fishel and 
Leonard (1955), the Graneros Shale is a fissile, black non-calcareous, 
marine shale with some sandstone, siltstone, and limestone beds. The 
Greenhorn Limestone consists of a blue-gray calcareous shale alternating 
with thin beds of dark crystalline or white chalky limestone. The Carlile 
Shale is represented by a lower member of fissile, light gray to buff, cal- 
careous, marine shale and an upper member of fissile, non-calcareous, black, 
marine shale with a thin sandy zone. The Fort Hays Limestone of the 
Niobrara Formation is a massive, white, chalky limestone. The two lime- 
stone formations uphold two dissected cuestas in the Blue Hills, the Green- 
horn Limestone cuesta in the east and the Fort Hays Limestone cuesta in 
the west. Areas underlain by the nonresistant shales have a rolling topogra- 
phy. Along the foot of the Fort Hays Limestone escarpment the Carlile 
Shale is intricately dissected into local small badlands. The more resistant 
limestones form the cap rock of flat-topped buttes and small mesas bounded 
by steep rocky escarpments up to 100 ft. high. 

Several factors are responsible for the fact that the Blue Hills are de- 
cidedly less favorable than the Smoky Hills as an area for the growth of 
mosses. Because the Blue Hills lie to the west of the Smoky Hills, the an- 
nual precipitation is one to five inches less. The topography of the Blue Hills 
is not as favorable as that of the Smoky Hills. Much of the Blue Hills is 
underlain by nonresistant shale which produces a rolling topography with 
insufhcient relief for exposures suitable for moss growth. However, where 
the shale is overlain by resistant limestone, erosion has produced hills and 
ravines where mosses can be found. Unlike the porous sandstone of the 
Smoky Hills, the shale and limestone of the Blue Hills are non-porous and 
do not support the growth of mosses. The fine-textured soil in the Blue Hills 
does not provide as much moisture for the growth of mosses as does the 
sandy soil in the Smoky Hills. Thus, the precipitation, topography, lithol- 
ogy, and edaphic conditions in the Blue Hills compare unfavorably with 
those of the Smoky Hills. 

Of the 33 species and varieties of mosses growing in the Blue Hills, 679%, 
are distributed in both the eastern and the western United States. Twenty- 


448 Tue Universiry ScIENCE BULLETIN 


four per cent are eastern species extending as far west as Kansas. Nine per 
cent are western in their distribution. 

The most common mosses of the Blue Hills are Astomum muhlenbergi- 
anum, Barbula unguiculata, Bryum caespiticium, Didymodon rigidulus, 
and Weissia microstoma. Their distributional affinities are varied. Bryum 
caespiticium, Astomum muhlenbergianum, and Barbula unguiculata are dis- 
tributed in both the eastern and the western United States. Weissia micro- 
stoma is an eastern species that does not grow west of Kansas. Didymodon 
rigidulus is a western species that reaches its eastern limit in Iowa and 
Michigan. Thus, the Blue Hills have selected an unusual assemblage of 
mosses as the characteristic elements of the flora. 

The other 28 species and varieties are rare: 


Amblystegium juratzkanum Ditrichum pallidum 

A. serpens Drepanocladus aduncus 
Astomum phascoides Ephemerum cohaerens 
Barbula cruegeri Eurhynchium serrulatum 

B. fallax Fissidens minutulus 

Bryum argenteum Grimmia apocarpa 

B. bicolor Leptodictyum trichopodium 
B. cuspidatum Leskea gracilescens 
Campylium chrysophyllum Orthotrichum diaphanum 
C. hispidulum O. pumilum 

Ceratodon purpureus Physcomitrium pyriforme var. serratum 
Chamberlainia acuminata Pterigoneurum subsessile 

C. oxyclada Pylaisia selwynit 

C. salebrosa Weissia controversa 


Mosses in the Blue Hills are found growing on north-facing slopes de- 
veloped along the contact between nonresistant shales and overlying resistant 
limestones. Shale and limestone rocks are too dry to support the growth of 
mosses, but a number of interesting species grow on soil and bark. The 
most commonly encountered mosses on soil in limestone areas are Astomum 
muhlenbergianum, Barbula unguiculata, and Bryum caespiticium. Didy- 
modon rigidulus, a moss characteristic of the Ogallala rock in the High 
Plains, is frequent on soil in the Blue Hills. Of special interest is Werssia 
microstoma. This moss is abundant on soil at certain horizons of the 
Greenhorn and Fort Hays limestone outcrops where there are numerous 
flat pieces of limestone rock scattered upon the surface of the soil. The as- 
sociation of Weissia microstoma with such outcrops is striking, inasmuch 
as this moss is rare or altogether lacking in other physiographic areas of 
Kansas. The prominent corticolous species are Campylium hispidulum, 
Leskea gracilescens, Orthotrichum diaphanum, and O. pumilum. 

Shale outcrops vary greatly in their capacity to support the growth of 
mosses. In one area there are within one-quarter mile of each other three 
microhabitats, each with its own moss flora. Astomum muhlenbergianum 
and Barbula unguiculata (not fruiting) were found on a dry hill slope. At 
the bottom of a moist ravine there were Bryum cuspidatum, Leptodictyum 


MosseEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 449 


trichopodium, and Drepanocladus aduncus. On a hill slope at a sandy 
horizon in the shale the following mosses were found: Astomum muhlen- 
bergianum, Barbula unguiculata (fruiting), Bryum cuspidatum, Physcomt- 
trium pyriforme var. serratum, Campylium hispidulum, Chamberlainia 
salebrosa, Eurhynchium serrulatum, and Ceratodon purpureus. 

The following species have been found growing on sandy Pleistocene 
alluvium: Bryum caespiticium, B. argenteum, Barbula unguiculata, Phys- 
comitrium pyriforme var. serratum, and Funaria hygrometrica. 


MOSSES OF THE RED’ HILLS 


The Red Hills form the Dissected High Plains south of the Arkansas 
River. According to Swineford (1955), the Red Hills are developed on pre- 
dominantly red Permian rocks consisting of very fine-grained sandstones 
and siltstones, silty shales and shales, and extensive thin beds of gypsum and 
anhydrite. The red color is due to a hematite stain on the mineral grains. 
Topographic features present in the Red Hills include jagged scarps, mesas, 
steep-sided buttes, and stream valleys with steep bluffs. Fenneman (1931) 
states that it is gypsum, a resistant rock in dry climates, that upholds the 
topographic features of the Red Hills. The prominent scarp forming the 
eastern boundary of the Red Hills is upheld by resistant sandstones (Bayne, 
1960). Thus, the topographic relief, which locally reaches a maximum of 
300 ft., is to be explained by the differential erosion of nonresistant shales 
and resistant sandstones and gypsum. 

Swineford (1955) has classified the Permian rocks in the Red Hills in 
the following formations, listed in ascending order: the Salt Plain Formation, 
the Cedar Hills Sandstone, the Flowerpot Shale, the Blaine Formation, and 
the Dog Creek Shale of the Nippewalla Group of the Leonardian Series, and 
the Whitehorse Sandstone, the Day Creek Dolomite, and the Taloga Forma- 
tion of uncertain classification. These formations outcrop from east to west 
in the same order. The Salt Plain Formation consists of reddish-brown, 
flaky siltstones, thin, sandy siltstones, and very fine-grained sandstones. A 
plain of low relief has developed upon most of the outcrop area. The Cedar 
Hills Sandstone includes brownish-red, massive, very fine-grained sandstones 
and sandy siltstones with beds of argillaceous siltstone and silty shale. Box 
canyons have developed in the thick sandstones. The Flowerpot Shale con- 
sists of reddish-brown gypsiferous shale and silty shale with a few thin beds 
of sandstone and siltstone; the Blaine Formation above contains massive 
gypsum, thin dolomite, and brownish-red shale. The resistant Blaine gyp- 
sum forms the cap rock of mesas and buttes, the slopes of which are de- 
veloped upon the nonresistant Flowerpot Shale lying below the gypsum. 
Where the protective gypsum cover has been removed by solution, deep 
ravines have been eroded into the soft shales. Prominent benches formed 


450 Tue University ScrteNce BULLETIN 


by resistant sandstones outcrop on the shale slopes. The Dog Creek Shale 
consists of thin beds of dark red, silty shale, brownish-red and greenish-gray 
siltstone, and very fine-grained sandstone, dolomite, dolomitic and gypsifer- 
ous sandstone, and gypsum. The Whitehorse Sandstone includes red, 
friable sandstone, siltstone, and shale, with minor quantities of white to 
buff sandstone and dolomite. The Day Creek Dolomite is a single bed of 
pale gray to pink, -n-, fine-grained dolomite with some chert. The Taloga 
Formation consists G1 montmorillonitic shale, silty shale, siltstones, and some 
very fine-grained sandstones. A rolling topography has developed in this 
formation. 

Like the Smoky Hills, the Red Hills are located toward the eastern part 
of the study area and receive a relatively large amount of precipitation. The 
topography of the Red Hills, like that of the Smoky Hills, is dissected with 
many ravines and north-facing hill slopes favorable for the growth of 
mosses. However, the shale, siltstone, and gypsum of the Red Hills are 
less porous and contain less water than the more porous sandstone of the 
Smoky Hills. Although the texture of the soii in the Red Hills is in places 
improved by the addition of fine sand, the soil of the Red Hills does not 
compare favorably with the predominantly sandy soil of the Smoky Hills 
in regard to the retention of moisture. As a result of the comparatively un- 
favorable lithologic and edaphic conditions in the Red Hills, there are only 
half as many species of mosses growing in the Red Hills as in the Smoky 
Hills. 

Most of the mosses growing in the Red Hills also grow throughout all 
or nearly all of the United States. Of the 31 species and varieties of mosses 
found in the Red Hills 68% are widely distributed throughout the United 
States. Ten per cent are eastern species extending as far west as Kansas. 
Nineteen per cent are essentially western species. Aschisma kansanum is 
endemic to Kansas. 

The most common mosses in the Red Hills are Astomum muhlenbergi- 
anum, Barbula ungutculata, Campylium chrysophyllum, Weissia contro- 
versa, and Didymodon rigidulus. Except for Didymodon rigidulus, which 
is of western affinity, the mosses characteristic of the Red Hills are species 
widely distributed throughout all or most of the United States. 

The remaining 26 species and varieties are rare: 


Acaulon rufescens Didymodon tophaceus 
Aloina rigida Ditrichum pallidum 
Amblystegium juratzkanum Fabroma ciliaris 

A. serpens Fabronia imperfecta 
Aschisma kansanum Fissidens minutulus 
Barbula fallax F. sublimbatus 

Bryum argenteum Homomallium mexicanum 
B. caespiticium Leskea gracilescens 

B. cuspidatum Orthotrichum diaphanum 


Campylium hispidulum O. pumilum 


MossEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 451 


Phascum cuspidatum var. americanum Tortula pagorum 
Pterigoneurum subsessile T. ruralis 
Pylaista selwynit Weissia microstoma 


The most characteristic mosses on slopes of hills and ravines are Asto- 
mum muhlenbergianum, Barbula unguiculata, Campylium chrysophyllum, 
and Weissia controversa. Mosses are not abundant on gypsum, but the 
notable ones are Amblystegium juratzkanum, A. serpens, Didymodon 
rigidulus, Fissidens minutulus, and F. sublimbatus. The important corti- 
colous species include Fabronia imperfecta, Leskea gracilescens, Orthotri- 
chum diaphanum, O. pumilum, and Tortula pagorum. Aschisma kansanum 
grows beneath and around the edges of small, round, white quartz pebbles. 

In the Red Hills there are several habitats quite unlike those found on the 
Permian rocks. The Cheyenne Sandstone of the Dakota Formation of 
Cretaceous age outcrops in local areas. Desmatodon obtusifolius, Grimmua 
laevigata, and G. rauei grow on the sandstone rocks, while Bryum ar- 
genteum, B. caespiticium, Ceratodon purpureus, Ditrichum pallidum, and 
Weissia controversa grow on the sandy soil. Except for Grimmua rauet, 
which normally grows on the Ogallala Formation in the High Plains, these 
mosses are most commonly found on the Dakota Formation in the Smoky 
Hills. Rocks of the Ogallala Formation outcrop at Big Basin, a sink located 
in the western part of the Red Hills. Didymodon rigidulus and Grimmuia 
wrightii grow on rocks; Aloina rigida, Bryum argenteum, B. caespiticium, 
Pterigoneurum subsessile, Phascum cuspidatum var. americanum, and 
Weissia controversa grow on the soil. All of these species are also found in 
the High Plains. The mosses common on sandy Pleistocene alluvium are 
Bryum argenteum, B. caespiticium, Barbula unguiculata, and Phascum 
cuspidatum var. americanum. Other mosses found on alluvium are Ar- 
chidium ohtoense, Aloina rigida, Astomum muhlenbergianum, Bryum bi- 
color, Ephemerum cohaerens, Physcomitrium pyriforme var. serratum, 
Pterigoneurum subsessile, and Pottia arizonica. Except for Pottia arizonica, 
which is known in Kansas from only one area, these mosses are found on 
sandy substrates in other physiographic areas in Kansas. 


MOSSES: OF THE HIGH PLAINS 


The High Plains include approximately the western one-third of Kansas. 
The surface materials are of Pleistocene and Pliocene (Ogallala) ages. 
According to Frye, Leonard, and Swineford (1956), the Ogallala Formation 
consists of feldspathic sand, silt, and gravel with irregular calcareous cementa- 
tion. The cemented deposits of the Ogallala Formation, known as mortar 
beds, are resistant and form cliffs along the sides of hills and ravines. Where 
the Ogallala Formation is covered by Pleistocene deposits, the topography 
is flat. Cretaceous chalk and shale outcrops are exposed along the river 


452 Tue University SCIENCE BULLETIN 


valleys. Also in the river valleys are Pleistocene alluvium and alluvial ter- 
races and low sand dunes. 

Insofar as topographic, lithologic, and edaphic conditions are concerned, 
the High Plains compare favorably with the Smoky Hills as an area suitable 
for the growth of mosses. Away from the flat inter-stream divides there are 
prominent cliffs and ravines. The Ogallala Formation, porous and often 
friable, is the only rock formation other than the Dakota Formation in the 
Smoky Hills to support the growth of mosses. Likewise, the soil texture is 
such as to contain moisture for moss growth. However, precipitation in the 
High Plains is 3 to 10 inches less than in the Smoky Hills. Also, the humidity 
is lower and the wind velocity is higher in the High Plains than in the 
Smoky Hills. As a result, considerably fewer species of mosses grow in the 
High Plains than in the Smoky Hills. 

Of the 44 species and varieties of mosses growing in the High Plains, 
61°% are distributed in both the eastern and the western United States. Six- 
teen per cent are eastern species extending as far west as Kansas, and 20% 
are western. Aschisma kansanum is endemic to Kansas. 

The most common mosses of the High Plains are Grimmuia wrighti, G. 
rauet, Didymodon rigidulus, Barbula unguiculata, and Bryum caespiticium. 
Grimmia wrighti, G. rauet, and Didymodon rigidulus are mosses of 
western affinity, while Barbula unguiculata and Bryum caespiticium are dis- 
tributed in both the eastern and the western United States. 

Four other species of mosses fairly common in the High Plains are 
Bryum argenteum, Desmatodon plinthobius, Pterigoneurum subsessile, and 
Tortula ruralis. 

The remaining 35 species and varieties are rare: 


Aloina brevirostris Didymodon tophaceus 

A. rigida Eurhynchium serrulatum 

Amblystegiella subtilis Fissidens bryoides 

Amblystegium americanum F. minutulus : 
A. juratzkanum F. sublimbatus | 
A. serpens Grimmia teretinervis 

Aphanorhegma serratum Hygroamblystegium irriguum 

Aschisma kansanum Hypnum cupressiforme 

Astomum muhlenbergianum Leptodictyum riparium 

Barbula fallax L. trichopodium 

Bryum bicolor Oreoweisia serrulata 

B. gemmiparum Orthotrichum diaphanum 

Campylium chrysophyllum Phascum cuspidatum var. americanum 

C. hispidulum Physcomitrium hookert 

Chamberlainia acuminata Weissia controversa 

C. salebrosa W. glauca 

C. velutina W. microstoma 


Desmatodon obtusifolius 


The Smoky Hill Chalk Member of the Niobrara Formation is exposed 
in areas bordering the major streams. Mosses growing on north-facing slopes 
of ravines are Astomum muhlenbergianum, Barbula unguiculata, Didy- 


MossEs oF THE GREAT PLAINS AND ARKANSAS River LowLanps 453 


modon rigidulus, Bryum caespiticium, Campylium hispidulum, Wetssta con- 
troversa, and W. microstoma. All of these are known also from Ogallala 
outcrop areas in the High Plains. 

Pleistocene alluvium and alluvial terraces and sand dunes occur in the 
river valleys. Mosses collected on sandy alluvium in the western part of the 
High Plaines include Barbula unguiculata, Bryum argenteum, B. caespiti- 
cium, Phascum cuspidatum var. americanum, and Pterigoneurum subsessile. 
On sandy terraces in Kingman and Reno counties of the eastern part of 
the High Plains the following mosses are abundant: Astomum muhlen- 
bergianum, Bryum argenteum, B. caespiticium, Phascum cuspidatum var. 
americanum, and Weissia controversa. Also present are Acaulon rufescens, 
A. schimperianum, A. triquetrum, Archidium ohioense, Astomum ludo- 
vicianum, Barbula unguiculata, Bryum cuspidatum, Ephemerum cohaerens, 
E. spinuolsum, Funaria flavicans, and Pleuridium subulatum. Bryum 
argenteum, B. caespiticitum, and Pterigoneurum subsessile grow abundantly 
on sand dunes, but Barbula unguiculata, Phascum cuspidatum var. amert- 
canum, and Physcomitrium pyriforme vat. serratum are rare. 

The Cockrum Sandstone of the Dakota Formation of Cretaceous age is 
exposed in local outcrops along streams in Stanton and Morton counties. 
Grimmia rauei, Weissia glauca, and Didymodon rigidulus grow on rock, 
while Aloina rigida, Barbula unguiculata, Bryum argenteum, B. caespiticium, 
Phascum cuspidatum var. americanum, Pterigoneurum ovatum, and P. 
subsessile grow on sandy soil. At present Pterigoneurum ovatum is known 
in Kansas only at Cockrum Sandstone outcrops in Stanton County. 


MOSSES OF THE WELLINGTON LOWLAND 


The Wellington Lowland is situated mostly in Sumner and Harper 
counties. The rocks that outcrop in the Wellington Lowland are of Permian 
and Pleistocene ages (Bayne, 1960; Walters, 1961). Although bedrocks are 
Permian, they are in many areas mantled by Pleistocene deposits. The 
Permian rocks are represented by the Sumner Group below and the Nip- 
pewalla Group above, both of the Leonardian Stage. The Sumner Group 
underlies the eastern part of the Wellington Lowland and consists of two 
formations: the Wellington Formation below and the Ninnescah Shale 
above. The Nippewalla Group underlies the western part of the Welling- 
ton Lowland and includes the Harper Siltstone below and the Salt Plain 
Siltstone above. The Wellington Formation consists of mostly shale and 
silty shale, usually of gray and green color, but sometimes red. The Nin- 
nescah Shale is a reddish-brown, silty shale with beds of thin calcareous silt- 
stone, blocky, reddish-brown shale, and a very fine-grained sandstone. The 
Harper Siltstone consists of reddish-brown, argillaceous siltstone and silty 
shale and a few thin beds of silty sandstone. The Salt Plain Siltstone con- 


454 Tue Universiry SciENcE BULLETIN 


sists of red silty shale with thin, silty sandstone and siltstone beds. Since the 
Permian shales and siltstones are easily eroded, and topography of the Well- 
ington Lowland is characterized by gently rolling hills. 

Although the average annual precipitation for the counties in the Well- 
ington Lowland is three inches more than those in the Smoky Hills, the 
Smoky Hills is a much better area for the growth of mosses. The topographic 
expression of the nonresistant shales and siltstones in the Wellington Low- 
land is a rolling area with only rare development of hills and ravines with 
protected exposures favorable for the existence of mosses. The shales and 
siltstones and the soils produced upon them are poor substrates as far as a 
moisture supply for mosses is concerned. The best growth of mosses occurs 
at sandy horizons of the Ninnescah Shale where a fine sand improves the 
texture of the soil, and the shale is eroded to produce north-facing slopes. 
Thus, despite a more favorable precipitation, the less favorable topographic, 
lithologic, and edaphic conditions in the Wellington Lowland make 
moisture conditions for the growth of mosses there less suitable than in the 
Smoky Hills. 

Of the 14 species of mosses known presently from the Wellington Low- 
land, 799% are distributed in both the eastern and the western United States. 
Fourteen per cent are eastern species and 7°% are western species. 

The characteristic mosses of the Wellington Lowland are Astomum 
muhlenbergianum, Barbula unguiculata, Bryum argenteum, and Phascum 
cuspidatum var. americanum. All of these are known from both the eastern 
and the western United States. 

The other 10 mosses are rare: 


Astomum ludovicianum Physcomitrium pyriforme 
A. phascoides P. pyriforme var. serratum 
Bryum caespit:c1um Pterigoneurum subsessile 
Chamberlainia salebrosa Pyramidula tetragona 
Leskea gracilescens Weissia controversa 


The most common mosses growing on sandy Pleistocene alluvium are 
Barbula unguculata, Bryum caespitictum, and Phascum cuspidatum var. 
americanum. Others include Astomum ludovicianum, A. muhlenbergi- 
anum, Barbula fallax, Bryum argenteum, Chamberlainia oxyclada, and 
Ephemerum cohaerens. Those growing on bark are Leskea gracilescens, 
Orthotrichum diaphanum, O. pumilum, and Tortula pagorum. 


MOSSES OF THE GREAT BEND LOWLAND 


The Great Bend Lowland extends along the Arkansas River from Ar- 
kansas City in the east to Dodge City in the west. Much of the area is 
underlain by Pleistocene dune sand. The dune sand consists of fine- to 
medium-grained quartzose sand with minor amounts of clay, silt, and 
coarse sand (Latta, 1950). Erosion has reduced the older dunes to a low, 


MosstEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 455 


hummocky topography. The younger dunes, subjected to erosion for a 
shorter time, form moderately steep, irregular, grass-covered hills surround- 
ing shallow, undrained basins (Bayne, 1956). Small ponds form in some 
of these undrained basins. Most of the Great Bend Lowland is covered by 
the older subdued dunes which are often in cultivation. However, younger, 
irregular dunes occur in various places, notably between the Arkansas and 
the Little Arkansas rivers. Pleistocene terraces and alluvium also occur in 
the Great Bend Lowland. 

Precipitation in the Great Bend Lowland compares favorably to that in 
the Smoky Hills. However, the topographic diversity present in the Smoky 
Hills is lacking in the former. Furthermore, dune sand does not have the 
good moisture retention properties characteristic of the sandy soil of the 
Smoky Hills. As a result of these factors, the Great Bend Lowland does 
not have as rich a moss flora as the Smoky Hills. 

Of the 30 species and varieties of mosses growing in the Great Bend 
Lowland, 70% are species found in both the eastern and the western United 
States. Twenty-seven per cent are eastern species extending as far west as 
Kansas, and three per cent are western species. 

The characteristic mosses of the Great Bend Lowland are Astomum 
muhlenbergianum, Bryum argenteum, B. caespiticium, and Physcomitrium 
pyriforme. All of these are distributed in both the eastern and the western 
United States. 

The remaining 26 species and varieties are rare: 


Amblystegium serpens E. spinulosum 

Archidium ohioense Funaria flavicans 

Atrichum angustatum F. hygrometrica 

Barbula unguiculata Leskea gracilescens 

Bruchia sullivanti Orthotrichum diaphanum 

Bryum bicolor O. pumilum 

B. pseudotriquetrum Phascum cuspidatum var. americanum 
Campylium chrysophyllum Physcomitrium pyriforme var. serratum 
Ceratodon purpureus Pleuridium subulatum 

Chamberlainia acuminata Pohlia nutans 

Ditrichum pallidum Pylaisia selwynit 

Ephemerum cohaerens Tortella humilis 

E. crassinervium Wetssia controversa 


Mosses are most commonly found on young sand dunes which provide 
enough relief to form shallow depressions. The only species commonly 
found where the dunes are above the water table are Astomum muhlen- 
bergianum, Bryum argenteum, B. caespiticium, and Physcomitrium pyrt- 
forme. The remaining species are normally found only in areas where the 
water table intersects the surface of the depressions and moist spots and 
small ponds are formed. Species growing on bark include Amblystegium 
serpens, Chamberlainia acuminata, Leskea gracilescens, Orthotrichum dia- 
phanum, O. pumilum, Pylaisia selwynit, and Tortella humilis. 

At Pawnee Rock in southwestern Barton County there is an isolated out- 


456 Tue UNiversiry ScIENCE BULLETIN 


crop of the Dakota Formation. Mosses growing there are Barbula unguicu- 
lata, Bryum argenteum, B. caespiticium, Desmatodon obtusifolius, and D. 
plinthobius. Desmatodon plinthobius is a moss occurring elsewhere in Kan- 
sas on the lime-cemented sands of the Ogallala Formation in the High 
Plains and on the lime-cemented sandstone concretions in the Smoky Hills. 


MOSSES OF THE FINNEY LOWLAND 


The Finney Lowland includes the Arkansas River Valley west of Dodge 
City. Extensive sand dunes occur in the outer valley of the river. Mosses 
have been collected from several areas in the sand dunes along the Arkansas 
River west of Syracuse in Hamilton County. The altitude of the dunes is 
only a few tens of feet higher than the Arkansas Valley and is more than 
150 feet lower than the upland area of the High Plains (McLaughlin, 1943). 
The maximum relief in the dunes is 70 feet. The dunes are bare or are 
covered by grass and sagebrush. 

Because the precipitation is only 17 inches per year and the dunes lie 
above the water table, only a few drought-resistant mosses grow in the area. 
However, the species present occur abundantly, mostly under the protective 
cover of sagebrush plants. The sagebrush bushes serve to collect dew and 
light rain and direct it toward the mosses growing underneath. As a result, 
even light rains can be effective in providing moisture. 

Only six species of mosses have been found growing on the sand dunes. 
Pterigoneurum subsessile, Bryum argenteum, and B. caespitictum are abun- 
dant, while Phascum cuspidatum var. americanum, Barbula unguiculata, 
and Funaria hygrometrica are rare. These six mosses represent four different 
geographical patterns in the United States. Brywm argenteum, B. caespiti- 
cium, and Funaria hygrometrica are distributed throughout the United 
States. Barbula unguiculata is known throughout the United States but is 
rare in the West Coast States. Phascum cuspidatum var. americanum is 
known from the eastern United States and Arizona. In contrast, Ptert- 
goneurum subsessile is a western species occurring mostly from Arizona to 
Texas and Kansas, but also reported from North ‘Dakota and Illinem, ie 
moss flora of the sand dunes of the Finney Lowland is, then, a unique 
assemblage. 

All mosses growing on the sand dunes have morphological adaptations 
of the leaves to hold capillary water. Bryum argenteum and B. caespiticium 
have leaves with slightly toothed awns. In addition to awns, the leaves of 
Phascum cuspidatum var. americanum have papillose leaf cells and revolute 
margins. Pterigoneurum subsessile, growing in irregularly shaped hum- 
mocks up to 5 cm across and fruiting profusely, has even more specialized 
adaptations. Both the erect lamellae attached to the upper surface of the 
leaves and the toothed awns projecting from the surface of the hummocks 
serve as very effective devices to hold capillary water. 


MosstEs oF THE GREAT PLAINS AND ARKANSAS RiveER LowLaNps 457 


COMPARISON OF THE FLORAS OF THE SEVEN 
PHYSIOGRAPHIC AREAS 


Table 2 shows the geographical affinities of the entire moss floras of the 
seven physiographic areas in the region under study. 


Tasre 2. Percentage of the species of the total moss floras of the seven physio- 
graphic areas with geographical distribution in both the eastern and the western 
United States, only in the eastern United States, and only in the Western United 
States. 


Eastern and 


Western Eastern Western 
SEGRE TUNGS BS ee ee eee eee 77 17 6 
Bitiewetills. -..! 2.22. PAE Si ee a ede Nee 67 24 9 
“heal THM Mey Cages SRR eee h owe ae EE aoe eee eee 68 10 19 
‘Hhyaln [DIGI eee Oe Ree eee eee ene eee eer 61 16 20 
Wellington Lowland ..................... ae are Ears 79 14 7 
Rerectambena Wowilamd) occ -seesse see cee eee ceeee 70 27 3 
faromevamleow land): 28. 22 ee eee 83 0 17 


From east to west in the Great Plains there is a decrease in the percentage of 
mosses of both eastern and western affinities and an increase in the per- 
centage of mosses of western afhnities. 

Table 3 shows the geographical affinities of the characteristic mosses of 
each of the seven physiographic areas. As in the comparison of the entire 
moss floras, the comparison of the most characteristic elements of each of 


Tas_e 3. Percentage of the characteristic moss species of the seven physiographic 
areas with geographical distribution in both the eastern and the western United 
States, only in the eastern United States, and only in the Western United States. 


Eastern and 


Western Eastern Western 
Bes iam 19 |S eee eee ee ee ee 100 0 0 
Echree TERUG) 2 ee cen Ra ti see rc 60 20 20 
"cmeGL TRUDI Aspe aoe each ne eee pha ep ae ee 80 0 20 
Hevemeblains) sts ee ee ees aid 40 0 60 
Pfellimeaton: eowlamd, 922.3 .2 <2. ciccscoceesecesecetenceceesee 100 0 0 
reat Bend Lowland .............-....-..-----.- Ce ee 100 0 0 
|Pampomeny! “Ufone aH ee ne ct 67 0 33 


the floras of the seven physiographic areas shows from east to west a de- 
crease in the percentage of mosses of both eastern and western distribution 
in the United States and an increase in the percentage of mosses of western 
distribution. However, in the comparison of the most common mosses there 
is a greater difference in the percentages from one area to the next. Werssia 
microstoma accounts for the 20°% of eastern mosses reported from the Blue 
Hills. The presence of this moss in the Blue Hills and its rarity in the rest 
of Kansas are very intriguing. Especially noteworthy are the high percentage 


458 THe UNIversity SCIENCE BULLETIN 


of mosses of both eastern and western affinity in the Smoky Hills, the 
Wellington Lowland, and the Great Bend Lowland and the high percentage 
of western species in the High Plains. 

Table 4 is a compilation of the most common or characteristic mosses 
of each of the seven physiographic areas. Although there are common ele- 
ments from area to area, each assemblage of mosses is unique to its physio- 
graphic area. The Smoky Hills is notable for the relatively large number of 
mosses commonly found. The other areas have comparatively few common 
mosses, but with the exception of the Great Bend Lowland these floras can- 
not be considered an impoverished Smoky Hills flora because of the presence 
of mosses not characteristic of the Smoky Hills. Wezssia microstoma 1s 
characteristic of only the Blue Hills; Grimmia rauei and G. wrighti1, of only 
the High Plains, and Prerigoneurum subsessile, of only the Finney Lowland. 
Three of the characteristic species of the Red Hills are common species in 
the Smoky Hills, while the other two species are common in the High Plains. 
The Wellington Lowland shares three species in common with the Smoky 
Hills and one species with the High Plains. The Smoky Hills and the High 
Plains have only one characteristic species in common, Brywm caespiticium. 
Thus, it can be seen that each physiographic area provides unique micro- 
habitats which select a characteristic assemblage of mosses. 


Tase 4. The characeristic mosses of the seven physiographic areas. 


Astomum muhlenbergianum Desmatodon obtusifolius 
Bryum argenteum Grimmia laevigata 
Smoky B. caespiticium Phascum cuspidatum var. americanum 
Hills Campylium chrysophyllum Physcomitrium pyriforme 
Chamberlainia salebrosa W eiss:a controversa 


Didymodon rigidulus 
Blue Astomum muhlenbergianum Bryum caespitici1um 
Hills Barbula unguiculata Wetssia microstoma 


Campylium chrysophyllum 


Red Astomum muhlenbergianum Weissia controversa 
Hills Barbula unguiculata Didymodon rigidulus 
Didymodon rigidulus 
High Grimmia rauei Barbula unguiculata 
Plains G. wrightii Bryum caespiticium 
Wellington Astomum muhlenbergianum Bryum argenteum 
Lowland Barbula unguiculata Phascum cuspidatum var. americanum 
Great Bend Astomum muhlenbergianum Bryum caespiticiuum 
Lowland Bryum argenteum Physcomitrium pyriforme 
Finney Bryum caespiticium 
Lowland Bryum argenteum Prerigoneurum subsessile 


It should be stated here that the mosses of the Great Plains and Arkansas 
River Lowlands are still imperfectly known. This area, of great geographical 
extent and geologic diversity, requires much more investigation before the 


Mossts OF THE GREAT PLAINs AND ARKANSAS River LowLaNps 459 


knowledge of the mosses is complete. The mosses of the McPherson and 
Wellington Lowlands are only poorly known. The Cedar Hills Sandstone 
and the Whitehorse Sandstone of the Red Hills are promising areas that 
have not been collected from. Further investigation should not overlook the 
Laverne and Rexroad formations exposed in Seward County in the High 
Plains. The mosses of the Smoky Hill Chalk Member of the Niobrara 
Formation may need to be treated as a separate unit, rather than as a sub- 
ordinate unit in the High Plains. However, this study should establish 
without doubt that the Great Plains and Arkansas River Lowlands of Kansas 
are areas of great bryological interest. 


DISCUSSION OF INDIVIDUAL SPECIES 


Each taxon is listed and its geographical range in the United States and 
in the Great Plains and Arkansas River Lowlands is stated. Geographical 
ranges are based on information in Grout (1928-1940) and on reports in 
volumes 18 to 60 The Bryologist, McGregor (1950), and McGregor and 
Hartman (1956). Morphological features of special interest are also dis- 
cussed. Starred taxa were previously unknown from Kansas or were with- 
out voucher specimen. A total of 117 species and varieties are reported with 
23 new records. Unless otherwise noted, all taxa are on file in the herbarium 
at the University of Kansas, Lawrence, Kansas. 

Acaulon rufescens Jaeg. Known from the eastern United States and also 
reported from Arizona, this species is rare and was found on soil in Cloud, 
Clay, Ottawa, and Ellsworth counties in the Smoky Hills, in Barber County 
in the Red Hills, in Reno County in the Great Bend Lowland, and on 
Pleistocene terraces in Kingman County in the High Plains. 

*Acaulon schimperianum Sull. Formerly known only from Arizona, 
Texas, and Iowa this species has been collected once in Kansas. The plant 
was growing beneath a round, white, quartz pebble on a Pleistocene terrace 
in Reno County in the Great Bend Lowland. 

Acaulon triquetrum (Spruce) C. M. Known from the eastern United 
States, this moss has been reported by Cridland (1960) on a Pleistocene 
terrace in Reno County in the High Plains. Specimens are not on file in 
the herbarium at the University of Kansas. 

Aloina brevirostris (Hook. & Grev.) Kindb. Distributed in Montana 
and the Rocky Mountains, this moss has been reported by Cridland (1960) 
on silty calcareous soil in Meade County in the High Plains. The specimen, 
sterile and tentatively assigned to this taxon, is not on file in the herbarium 
at the University of Kansas. 

*Aloina rigida (Hedw. ex Schultz) Kindb. This moss has been reported 
from California, the Rocky Mountains, Arizona, Iowa, and Illinois. In Kan- 
sas this rare species grows on soil in Cheyenne, Stanton, and Clark counties 


460 Tue University ScIENCE BULLETIN 


in the High Plains, and in Clark and Barber counties in the Red Hills. The 
Kansas material is sterile and is referred to this taxon because the leaves are 
curled-in when dry and spreading when moist. 

*Amblystegiella subtilis (Hedw.) Loeske. This species, distributed in 
the eastern United States, is known in Kansas from one locality in Cheyenne 
County in the High Plains. The habitat was bark and rocks. 

Amblystegium americanum Grout. This moss has been reported by 
McGregor (1950) on moist soil near a spring in Scott County in the High 
Plains. Known from Wisconsin, Iowa, and North and South Dakota, this 
taxon is considered by Conard (1959) to be a variety or form of A. com- 
pactum, which is distributed across the United States. 

Amblystegium juratzkanum Schimp. Widely distributed throughout the 
United States, this is a rare moss, growing on bark, soil, and rocks in the 
Smoky Hills, the Red Hills, the Blue Hills, and the High Plains. 

Amblystegium serpens (Hedw.) Bry. Eur. This moss, widespread in the 
United States except for the West Coast States, is a rare species growing on 
bark, soil, and rocks in the Smoky Hills, the Blue Hills, the Red Hills, the 
Great Bend Lowland, and the High Plains. 

Amblystegium varium (Hedw.) Lindb. This widely distributed species 
of the United States, is rare in Kansas and is found on rocks, bark, and soil 
in the Smoky Hills. 

Amblystegium varium (Hedw.) Lindb. var. ovatum Grout. Known 
from Pennsylvania, Missouri, and Iowa, this variety has been reported by 
McGregor (1950) on moist soil in Ottawa County in the Smoky Hills. 

Aphanorhegma serratum (Hook. & Wils.) Sull. This moss, known from 
the eastern one-half of the United States, has been reported by McGregor 
(1950) from Decatur County in the High Plains. 

*Archidium ohioense Schimp. This moss has been reported from New 
York, North Carolina, Florida, Texas, and Minnesota. In Kansas this rare 
moss has been found on Pleistocene alluvium in Harper County in the Red 
Hills, on sand dunes in Harvey County in the Great Bend Lowland, and on 
Pleistocene terraces in Kingman County in the High Plains. 

Aschisma kansanum Andrews. This moss is presently known only from 
Kansas. It was described by Andrews (1915) from material collected by E. 
Hall on the prairies of western Kansas. Type material is in the herbarium 
of the New York Botanical Garden and the Sullivant and James herbarium. 
Cridland (1959, 1960) reported this moss from Barber, Reno, and Pratt 
counties. A persistent protonema develops beneath rounded, relatively trans- 
lucent, quartz pebbles with diameters of one to five cm. The pebbles, with 
protonema growing beneath them, are a milky-white color; dark pebbles 
are apparently too opaque to support growth of the protonema. Gameto- 
phores are produced around the edge of the pebble near the surface of the 


MosskEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 461 


soil and occasionally even beneath the pebble itself. This moss was found 
fruiting profusely on April 28, 1962, in section 28, R. 32 S., T. 14 W., in 
Barber County in the Red Hills. Pebbles with Aschisma kansanum have 
also been found in Kiowa County in the Red Hills and in Clark County in 
the High Plains. 

Cridland (1959) reported Acaulon sp. from Reno County and Weissia 
sp. from Pratt County in association with quartz pebbles. To these may be 
added Acaulon schimperianum from Reno County; Astomum sp. (sterile) 
from Cheyenne, Kingman, Phillips, Scott, and Wallace counties; Barbula 
unguiculata from Cheyene County; Brywm argenteum from Kingman and 
Scott counties; B. caespiticium from Kingman County; Grimmia wrightit 
from Scott County; Phascum cuspidatum var. americanum from Kingman 
County; and Prerigoneurum subsessile from Wallace County. As Cridland 
points out, no other taxa utilizes the habitat under the quartz pebble as 
successfully as Aschisma kansanum. Occasionally Oscillatoria sp. and 
Nostoc sp. grow with the protonema of the mosses. 

*Astomum ludovicianum Sull. This moss was formerly known only 
from West Virginia to Florida and Texas. The Kansas plants extend the 
range of this taxon much farther inland. The species is common in the 
Smoky Hills and especially abundant on soil at Rock City near Minneapolis 
in Ottawa County. It has been found also in Sumner County in the Well- 
ington Lowland and in Kingman County in the High Plains. 

Astomum muhlenbergianum (Sw.) Grout. This moss is distributed 
throughout the United States except for the West Coast States. It is an 
important species on soil in the Smoky Hills, the Red Hills, the Blue Hills, 
the Wellington Lowland, and the Great Bend Lowland. It is rare in the 
High Plains and presently unknown in the Finney Lowland. 

*Astomum phascoides (Hook.) Grout. This moss was formerly known 
only from North Carolina, Ohio, Minnesota, Texas, and Arizona. It is 
rare in Kansas, known on soil only in Smith County in the Blue Hills and 
Sumner County in the Wellington Lowland. 

Atrichum angustatum (Brid.) Bry. Eur. This moss, widely distributed 
in the eastern United States, is common on soil and rocks in the Smoky Hills 
and is rare in the Great Bend Lowland. 

Atrichum undulatum (Hedw.) Beauv. Widely distributed in the United 
States, this moss has been reported by McGregor (1950) on soil in Wash- 
ington County in the Smoky Hills. 

Atrichum undulatum (Hedw.) Beauv. var. minus (Lam. & De Cand.) 
Web. & Mohr. Known from Pennsylvania, Iowa, Kansas, Texas, and Wash- 
ington, this variety has been reported by McGregor (1950) on soil in Wash- 

ington County in the Smoky Hills. 
 Barbula cruegeri Sond. Known from North Carolina to Kansas and 


462 Tue University ScIENCE BULLETIN 


southward, this moss has been reported by McGregor (1950) on soil in 
Republic County in the Blue Hills. No gemmae were found but the stems 
and midribs of leaves were red. 

Barbula fallax Hedw. This moss, reported from the United States, ex- 
cept from the West Coast States, is a rare species, growing on rocks and 
soil in the High Plains, the Red Hills, the Blue Hills, and the Smoky Hills. 
McGregor (1950) has reported it on Pleistocene terraces in the Wellington 
Lowland. 

Barbula unguiculata Hedw. Known throughout the United States but 
rare in the West Coast States, this is an important species on soil and rocks 
in the High Plains, the Red Hills, the Blue Hills, and the Wellington 
Lowland. It is common in the Smoky Hills but rare in the Great Bend Low- 
land. McGregor (1950) has reported it from the Finney Lowland. 

Bruchia sullivanti Aust. This moss of the eastern United States has been 
found in the study area at one locality in sand dunes of Harvey County in 
the Great Bend Lowland. 

Bryum argenteum Hedw. This moss, widely distributed throughout 
the United States, is an important moss on soil and rocks in the Smoky 
Hills, the Wellington Lowland, the Great Bend Lowland, and the Finney 
Lowland. It is fairly common in the High Plains but is rare in the Red 
Hills and the Blue Hills except on sandy substrates. 

Bryum bicolor Dicks. Distributed throughout the United States, this is 
a rare moss in Kansas, growing on soil in Ottawa County in the Smoky 
Hills and on sand dunes in Harvey County in the Great Bend Lowland. 
It has been reported by McGregor (1950) from the Blue Hills, the Red 
Hills, and the High Plains. 

Bryum caespiticium Hedw. This moss is widely distributed in the 
United States and is an important species on soil in the Smoky Hills, the 
Blue Hills, the High Plains, the Great Bend Lowland, and the Finney 
Lowland. It is rare in the Wellington Lowland and the Red Hills except on 
sandy substrates. 

Bryum cuspidatum (Bry. Eur.) Schimp. This species, known from the 
United States except for the West Coast States, is rarely found on soil in 
the Smoky Hills, the Blue Hills, the Red Hills, and the High Plains. 

Bryum gemmiparum DeNot. This species is distributed throughout the 
United States and has been reported by McGregor (1950) on moist soil and 
rocks at a spring in Scott County in the High Plains. 

Bryum pendulum (Hornsch.) Schimp. This moss is found throughout 
the United States except for the West Coast States. It is a rare moss on 
soil and rocks in the Smoky Hills. 

Bryum pseudotriquetrum (Hedw.) Schw. This species is rare in Kansas, 
growing on rocks in the Smoky Hills and on sand dunes in Reno County in 
the Great Bend Lowland. It is distributed throughout the United States. 


MossEs oF THE GREAT PLAINS AND ARKANSAS River LowLanps 463 


Campylium chrysophyllum (Brid.) Bryhn. This species, known from 
the United States except from the West Coast States, is common on soil 
and rocks in the Smoky Hills and important but less common in the Red 
Hills. It is rare in the High Plains, the Blue Hills, and the Great Bend Low- 
land. 

Campylium chrysophyllum (Brid.) Bryhn var. brevifolium (R. & C.) 
Grout. This moss, known from the eastern United States, is rare on soil and 
rocks in Ottawa and Ellsworth counties in the Smoky Hills. 

Campylium hispidulum (Brid.) Mitt. This moss is fairly common on 
soil, bark, and rocks in the Smoky Hills and on soil at local areas of Ogallala 
and Smoky Hill Chalk outcrops in the High Plains. It is rare in the Blue 
Hills. The species is unknown from the West Coast. 

Ceratodon purpureus (Hedw.) Brid. This moss, widely distributed in 
the United States, is fairly common on soil in the Smoky Hills but is rare in 
the Great Bend Lowland, the Red Hills, and the Blue Hills. The specimens 
reported by McGregor (1950) for Clark and Rawlins counties in the High 
Plains are Didymodon rigidulus. 

Chamberlainia acuminata (Hedw.) Grout. Known from the eastern 
United States and Arizona, this rare moss grows on soil, rocks, and bark in 
the Smoky Hills, the Blue Hills, the Great Bend Lowland, and the High 
Plains. 

Chamberlainia oxyclada (Brid.) Robins. Brachythecium oxycladon 
(Brid.) J. & S. Restricted to the eastern United States, this moss has been re- 
ported by McGregor (1950) on soil in the Smoky Hills, the Blue Hills, and 
the Wellington Lowland. Because the Kansas material is sterile and the 
vegetative characters of C. oxyclada and C. salebrosa are not completely 
distinct, the assignment of specimens to these two species is difficult. Brachy- 
thecium oxycladon var. dentatum (Lesq. & James) Grout has been reported 
by McGregor (1950) for Washington County. Robinson (1962) does not 
mention this variety. It is possible the variety should be suppressed. 

Chamberlainia salebrosa (Web. & Mohr) Robins. Brachythecium sale- 
brosum (Web. & Mohr) Bry. Eur. This widely distributed species is com- 
mon on rocks and soil in the Smoky Hills but rare in the Blue Hills, the 
Wellington Lowland, and the High Plains. Brachythecium flexicaule R. & 
C., reported by McGregor (1950) for Cloud County, is referred to this taxon. 
According to Robinson (1962), Grout’s concept of B. flexicaule was appar- 
ently based on material that was C. salebrosa. The Cloud County plants 
identified as B. flexicaule cannot be referred to C. calcarea (Kindb.) Robins., 
B. calcareum Kindb. and B. flexicaule R. & C. (correctly conceived) as 
synonyms, because the dry leaves are not secund. 

*Chamberlainia velutina (Hedw.) Robins. This moss, distributed 
throughout the United States, is presently known in Kansas from one locality 


464 Tue Universiry SCIENCE BULLETIN 


in Rawlins County in the High Plains. It was found in a recess beneath an 
overhanging ledge of mortar beds in a ravine. Dried tumbleweeds had 
piled up in front of the recess. 

Desmatodon obtusifolius (Schw.) Jur. This moss, known throughout 
the United States, is common on rocks in the Smoky Hills but rare in the 
High Plains and the Red Hills. It is also known from an isolated outcrop of 
the Dakota Formation in Barton County in the Great Bend Lowland 
(McGregor, 1950). 

Desmatodon plinthobius Sull. & Lesq. This moss is fairly common on 
rocks in the High Plains and is present on sandstone concretions at Rock 
City in the Smoky Hills. It is also known from an isolated outcrop of the 
Dakota Formation in Barton County in the Great Bend Lowland. It is 
unknown from the West Coast but has been reported for most other parts 
of the United States. The specimen reported by McGregor (1950) for 
Sumner County is Phascum cuspidatum var. americanum. 

*Dichodontium pellucidum (Hedw.) Schimp. This moss, distributed in 
the United States except from the Southeast, is known in Kansas from one 
locality on sandy soil in McPherson County in the Smoky Hills. 

Dicranella varia (Hedw.) Schimp. Widespread throughout the United 
States, this species has been reported by McGregor (1950) for Washington 
County in the Smoky Hills. 

Didymodon rigidulus Hedw. This western species is known eastward 
to Texas, Kansas, South Dakota, Iowa, and Michigan. It grows on rocks 
and soil and is the most common moss in the High Plains. It is less common 
but still important in the floras of the Blue Hills and the Red Hills and is 
occasionally found in the Smoky Hills. Some plants produce smooth, 
spherical, 2 to 6-celled gemmae in the axils of the leaves. The cell contents 
of the gemmae are green and the cell walls are a reddish-brown color. 

Didymodon tophaceus (Brid.) Jur. This moss, widely distributed in the 
United States, is rare on rocks and soil in Barber County in the Red Hills 
and in Cheyenne and Clark counties in the High Plains. 

Ditrichum pallidum (Hedw.) Hampe. This moss is distributed in the 
eastern United States and is fairly common on soil and rocks in the Smoky 
Hills. It is rare in the Red Hills, the Blue Hills, and the Great Bend Low- 
land. 

*Drepanocladus aduncus (Hedw.) Warnst. Known across the United 
States, this moss is a rare species found growing in a moist ravine in Jewell 
County in the Blue Hills. 

Drepanocladus aduncus (Hedw.) Warnst. var. polycarpus (Bland.) 
Warnst. This widely distributed species of the United States has been re- 
ported by McGregor (1950) on very moist sandy soil in Mitchell County 


MosstEs OF THE GREAT PLAINS AND ARKANSAS River LOWLANDS 465 


in the Smoky Hills and in Trego County (specimen in the herbarium at 
Kansas State University, Manhattan, Kansas). 

Entodon compressus (Hedw.) C. M. This moss of the eastern United 
States has been reported by McGregor (1950) on bark and soil in Ottawa 
and Washington counties in the Smoky Hills. 

Entodon seductrix (Hedw.) C. M. Known from the eastern United 
States and Arizona, this species is fairly common on rocks, bark, and soil in 
the Smoky Hills. 

Ephemerum cohaerens (Hedw.) Hampe. This species, distributed in the 
eastern one-half of the United States, is a rare moss, growing on sandy soil 
in Reno and Harvey counties in the Great Bend Lowland, in Sumner County 
in the Wellington Lowland, in Harper County in the Red Hills, and in 
Republic County in the Blue Hills. 

Ephemerum crassinervium (Schw.) C. M. Known from the eastern one- 
half of the United States, this species is rare on dune sand in Reno County 
in the Great Bend Lowland. 

*Ephemerum spinulosum Schimp. This moss is distributed in the east- 
ern one-half of the United States and is a rare moss, growing on soil in 
Ottawa and Ellsworth counties in the Smoky Hills and on sand dunes in 
Reno and Harvey counties in the Great Bend Lowland. 

Eurhynchium hians (Hedw.) J. & S. This rare moss is known for the 
eastern United States and Arizona and grows on soil in the Smoky Hills. 

Eurhynchium serrulatum (Hedw.) Kindb. This moss, known from the 
eastern United States, is a rare species, growing on soil, bark, and rocks in 
the Smoky Hills, the Blue Hills, and the High Plains. 

Fabronia ciliaris (Brid.) Brid. Reported for the eastern United States, 
New Mexico, and Arizona, this species is known from one locality in the 
Great Plains of Kansas. It was found growing on oak bark in Barber 
County in the Red Hills. 

*Fabronia imperfecta Sharp. This moss has been previously reported 
only from Tennessee by Sharp (1933). The peristome is imperfect or lack- 
ing. The spore size of the Kansas plants ranges up to 20 microns, 5 microns 
larger than those of the Tennessee plants. This variation is not surprising 
in a genus which Conard (1956) has described as: “A whole series of inter- 
grading forms as to shape of leaf, size of marginal teeth, size and shape of 
cells, size of spores.” Conard adds: “Some combinations are named, others 
fortunately, are not.” This taxon has been found growing on elm bark 
in Ellsworth County in the Smoky Hills and in Barber County in the Red 
Hills. 

*Fissidens bryoides Hedw. This moss, distributed widely in the United 
States, has been found in Kansas on soil in Washington County in the 
“Smoky Hills and on rock in Rawlins County in the High Plains. 
| 
| 


466 Tue University SciENCE BULLETIN 


Fissidens minutulus Sull. Known from the eastern United States and 


Arizona, this is a rare moss which grows on soil in Jewell, Rooks, and — 


Trego counties in the Blue Hills, in Barber County in the Red Hills, and in 
Rawlins County in the High Plains. 

Fissidens obtusifolius Wils. This moss, known from the eastern United 
States, Colorado, and Arizona, is a rare species which grows on rock in the 
Smoky Hills. 

Fissidens cbtusifolius Wils. var. Ransanus R. & C. This variety was de- 
scribed from material collected by Henry in Saline County (Grout, 1936). 
The type specimen is not in the herbarium at the University of Kansas. This 
variety has since been reported from Arizona. 

*Fissidens osmundioides Hedw. Reported from the eastern United 
States, Wyoming, and Idaho, this species has been collected once in Kansas 
on rock in Ellsworth County in the Smoky Hills. 

Fissidens sublimbatus Grout. This moss has been reported from Arizona, 
New Mexico, Texas, and Missouri. It is a rare moss, growing on rocks and 
soil in Washington and Ellsworth counties in the Smoky Hills, in Barber 


County in the Red Hills, and in Decatur and Scott counties in the High — 


Plains. 
*Funaria americana Lindb. This moss, known from the eastern United 


| 


| 


| 


States and California, is a rare species on rocks and soil in Jewell and Ells- 


worth counties in the Smoky Hills. 

Funaria flavicans Mx. Known from the eastern United States and 
Wyoming, this is a rare moss which grows on soil in McPherson County in 
the Smoky Hills and on sand dunes in Reno and Harvey counties in the 
Great Bend Lowland. 

Funaria hygrometrica Hedw. This moss, widely distributed throughout 


the United States, is rare on soil in the Smoky Hills, the Great Bend Low- | 


land, and the Blue Hills. McGregor (1950) has reported it from the Finney 
Lowland. 


Grimmia apocarpa Hedw. Widely distributed throughout the United | 


States, this moss is fairly common on rocks in the Smoky Hills and rare in 
the Blue Hills. It is excluded from the High Plains as reported by Mc- 
Gregor (1950). 

Grimmua laevigata (Brid.) Brid. This moss, widely distributed in the 
United States, is common on dry wind-swept rocks in the Smoky Hills. 
It is also known on the Cheyenne Sandstone in Kiowa County in the Red 
Hills. 

Grimmia plagiopodia Hedw. Known from the western United States 
except the West Coast States eastward to Iowa, this species has been reported 
from Kansas by McGregor (1950) on dry rocks in Rice County in the 
Smoky Hills. 


MossEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 467 


Grimmuia pulvinata (Hedw.) Sm. This moss is known from the western 
United States and eastward to Iowa and has been found on rocks in Ells- 
worth County in the Smoky Hills. 

Grimmia raue: Aust. As treated in Grout (1940), G. rauer has leaves 
with entire apical margins and G. wrighti has leaves with denticulate apical 
margins. Because of the variability of denticulation, Sayre (1952) referred 
plants with ovate leaves to G. rauer and plants with obovate or oblong leaves 
to G. wright. The Kansas plants are classified according to Sayre’s concept 
of the two species. Since the publication in 1952 of Sayre’s key to the genus 
Grimmia, G. rauei has been reported in “The Bryologist” from South 
Dakota, Texas, and Arizona. Conard (1956) reports it from Minnesota to 
Texas and Arizona. G. rauei is common on the mortar beds in the High 
Plains and rare on the Cheyenne Sandstone in Comanche and Kiowa 
counties in the Red Hills. 

Grimmua teretinervis Limpr. This moss has been reported by Kucyniak 
(1952) for North America from Quebec, Wisconsin, Minnesota, Missouri, 
and Kansas. McGregor and Hartman (1956) reported this species on sand- 
stone rocks in Russell County in the Smoky Hills and on mortar beds in 
Norton and Scott counties in the High Plains. This moss has again been 
found on the mortar beds in Scott County. It is well known in the Alps 
of Europe (Kucyniak, 1952). The dry wind-swept cliffs in Kansas appear to 
resemble the alpine habitat with respect to the growth of this moss. Mc- 
Gregor (1956) states: “It (Grimmua teretinervis) appears to be an alpine, 
non-arctic species which can occur at low altitudes, even sea level, where 
conditions of wind and other climatic factors somewhat duplicate alpine 
conditions.” 

*Grimmia wrightu Aust. Since the publication in 1952 of Sayre’s key to 
the genus Grimmuia, G. wrightii has been reported in “The Bryologist” from 
South Dakota and Texas. In Kansas it is an important species on dry wind- 
swept rocks in the High Plains. This taxon is also present on the sandstone 
concretions at Rock City in Ottawa County in the Smoky Hills. 

Hedwigia ciliata Hedw. This widely distributed moss is fairly common 
on rocks in the Smoky Hills. 

Homomallium adnatum (Hedw.) Broth. This moss, known from the 
eastern United States and Arizona, is fairly common on rocks and bark in 
the Smoky Hills. 

— *Homomallium mexicanum Cardot. This species has previously been re- 

ported from Arizona, New Mexico, and Texas. The Kansas plants extend 
the range considerably northward. Presently this taxon is known in Kansas 
from only one locality on soil and gypsum rock in Barber County in the 
Red Hills. 

_ Hygroamblystegium irriguum (Wils.) Loeske. Distributed throughout 


468 Tue UNrversity ScIENCE BULLETIN 


the United States, this is a rare moss which grows on rocks, soil, and bark in 
the Smoky Hills and the High Plains. 

*Hypnum cupressiforme Hedw. This moss, distributed in the United 
States except for the West Coast States, is known from one Kansas locality in 
Scott County in the High Plains where it was growing on soil and rocks. 

Leptobryum pyriforme (Hedw.) Schimp. This widespread moss was 
found on soil in McPherson County in the Smoky Hills. 

Leptodictyum riparium (Hedw.) Warnst. This widespread species was 
found on bark at a spring in Ellsworth County in the Smoky Hills. Me- 
Gregor (1950) has reported it on moist soil in Meade County in the High 
Plains. 

Leptodictyum trichopodium (Schultz) Warnst. This eastern species is 
rare on soil and bark in the Smoky Hills, the Blue Hills, and the High 
Plains as reported by McGregor (1950). 

Leskea gracilescens Hedw. This moss, distributed in the eastern United 
States and westward to the Rocky Mountains, grows on bark in the Smoky 
Hills, the Blue Hills, the Red Hills, the Wellington Lowland, and the Great 
Bend Lowland. 

Leskea obscura Hedw. Widely distributed east of the Rocky Mountains, 
this moss has been reported by McGregor (1950) on bark in Mitchell and 
Lincoln counties in the Smoky Hills. 

Lindbergia brachyptera (Mitt.) Kindb. This moss is distributed in the 
eastern United States and Arizona (Crum, 1956). Crum has examined and 
verified the specimen collected by Henry from Saline County but did not 
specify the particular herbarium where the specimen is located. 

Mnium cuspidatum Hedw. This widely distributed species is fairly com- 
mon on rocks and soil in the Smoky Hills. 

*Oreoweisia serrulata (Funck) DeNot. This widespread moss has been 
found on rocks in Decatur, Rawlins, Scott, and Clark counties in the High 
Plains. The upper part of the leaves is partially monostratose and partially 
bistratose. 

Orthotrichum diaphanum Brid. This moss has been reported from 
Arizona, New Mexico, Colorado, Texas, Oklahoma, and Kansas. Pre 
viously this taxon was known in Kansas only from Reno County as re- 
ported by McGregor (1950). This bark-inhabiting species is now known 
in Barton, Stafford, and Reno counties in the Great Bend Lowland, in 
Sumner County in the Wellington Lowland, in Barber County in the Red 
Hills, in Rooks County in the Blue Hills, and in Decatur County in the 
High Plains. 

Orthotrichum pumilum Dicks. Distributed in the United States except 
from the West Coast States, this species grows on bark in the Smoky Hills, 
the Wellington Lowland, the Great Bend Lowland, the Red Hills, and the 
Blue Hills. 


MossEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 469 


Orthotrichum pusillum Mitt. Distributed in the eastern United States, 
this moss has been reported from Reno County by McGregor (1950). The 
specimen, collected by Voth, is located in the herbarium at the Chicago 
Natural History Museum. 

Orthotrichum strangulatum Schw. This moss, known from the eastern 
United States and Utah, has been found on rocks in the Smoky Hills in 
Ellsworth County and in Mitchell County as reported by McGregor (1950). 

Phascum cuspidatum Hedw. var. americanum R. & C. Known in the 
eastern United States and Arizona, this is an important species on soil in 
the Smoky Hills and the Wellington Lowland. It is rare in the Great Bend 
Lowland, the Red Hills, and the High Plains. 

Philonotis longiseta (Rich.) E. G. B. This moss, known from the eastern 
one-half of the United States, has been collected on a sandy creek bank in 
Washington County in the Smoky Hills. 

Physcomitrium hookeri Hampe. This species is known from the central 
United States, New York, and Utah, and has been collected on soil in Mc- 
Pherson County in the Smoky Hills. McGregor (1950) has reported it on 
soil at the edge of a swamp in Meade County in the High Plains. 

Physcomitrium pyriforme Brid. This widely distributed moss is common 
on soil in the Smoky Hills and rare on sand dunes in the Great Bend Low- 
land. 

Physcomitrium pyriforme Brid. var. serratum (R. & C.) Crum and An- 
derson. (P. kellermani Britt. and P. drummondii Britt. as synonyms). This 
variety is limited to an area roughly corresponding to the Great Plains from 
Louisiana and Texas to North Dakota (Crum and Anderson, 1955). It is 
a rare moss on soil in Washington and Clay counties in the Smoky Hills, in 
Sumner County in the Wellington Lowland, in Harper County in the Red 
Hills, in Jewell and Ellis counties in the Blue Hills, and on sand dunes in 
Reno and Stafford counties in the Great Bend Lowland and in Phillips 
County in the High Plains. The spores of the Kansas plants showed no 
‘tendency to adhere in diads, triads, and tetrads as reported for most speci- 
‘mens by Crum and Anderson. 

Plagiothecium geophilum (Aust.) Grout. This moss, distributed in the 
eastern United States, New Mexico, and Arizona, is a rare species which 
grows on rocks in Ellsworth County in the Smoky Hills. 

Platygyrium repens (Brid.) Byr. Eur. This moss, distributed east of the 
Rocky Mountains, has been reported by McGregor (1950) on bark in 
Mitchell County in the Smoky Hills. 

Pleuridium subulatum (Hedw.) Lindb. This rare species is known from 
the eastern United States and California and grows on soil in Washington 
and Ellsworth counties in the Smoky Hills, on sand dunes in Harvey County 


47() Tue UNiversiry ScIENCE BULLETIN 


in the Great Bend Lowland, and on Pleistocene terraces in Kingman County 
in the High Plains. 

Pohlia nutans (Hedw.) Lindb. This widely distributed moss is fairly 
common on soil and rocks in the Smoky Hills and rare on sand dunes in 
the Great Bend Lowland. 

Pohlia wahlenbergii (W. & M.) Andr. Widely distributed in the United 
States, this species has been reported by McGregor (1950) on rocks and 
soil in the Smoky Hills. 

Polytrichum juniperinum Hedw. This widely distributed species is rare 
in the Great Plains of Kansas. It is presently known on rocks only in Ells- 
worth County in the Smoky Hills. 

Polytrichum piliferum Hedw. This widely distributed moss has been 
reported by McGregor and Hartman (1956) from a sandy hillside in Ells- 
worth County in the Smoky Hills. 

*Pottia arizonica Wareham. This species has been reported from Arizona 
and southern California. The Kansas plants were found growing on sandy 
Pleistocene alluvium in Clark County in the Red Hills. 

*Pterigoneurum ovatum (Hedw.) Dixon. This moss has previously 
been reported from Washington, Idaho, Utah, Wyoming, North Dakota, 
Arizona, New Mexico, and Texas. The Kansas plants are known from one | 
locality on soil where the Cockrum Sandstone outcrops in Stanton County — 
in the High Plains. | 

Pterigoneurum subsessile (Brid.) Jur. This moss has been reported 
from Arizona, Texas, North Dakota, Kansas, and Illinois. This taxon 1s - 
important on sand dunes in the Finney Lowland and the High Plains. It 
is fairly common in the remainder of the High Plains and rare in the Red — 
Hills, the Blue Hills, and the Wellington Lowland. 

Pterigoneurum subsessile (Brid.) Jur. var. henricu (Rau) Wareham. 
The type specimen, collected by Henry in Saline County, is in the herbarium 
at the New York Botanical Garden (Grout, 1939). Grout considered it 
possible that the variety should be suppressed. Plants resembling this variety 
have been reported from Texas. 

Ptychomitrium incurvum (Mubhl.) Sull. Known from the eastern one- 
half of the United States, this species is fairly common on rocks in the 
Smoky Hills. 

Pylaisia selwynii Kindb. This moss, known from the eastern United 
States and Arizona, is a rare species which grows on bark in the Great Bend 
Lowland, the Blue Hills, the Red Hills, and the Smoky Hills as reported by 
McGregor (1950). 

Pyramidula tetragona (Brid.) Brid. This moss has been reported from 
Colorado, Texas, Oklahoma, Kansas, Nebraska, Iowa, Minnesota, and 
Indiana. It is rare on soil in the Smoky Hills and the Wellington Lowland. 


| 


MossEs OF THE GREAT PLAINS AND ARKANSAS River LowLanps 471 


Thuidium virginianum (Brid.) Lindb. This moss, known from the 
eastern United States and Arizona, is a rare moss on soil in the Smoky 
Hills. 

Timmia megapolitana Hedw. This widely distributed species has been 
reported by McGregor (1950) on sandy soil in Mitchell County in the Smoky 
Hills. 

Tortella humilis (Hedw.) Jenn. Known from the eastern United States 
and Arizona, this species has been found on bark at base of a honey locust 
tree in Reno County in the Great Bend Lowland. 

Tortula pagorum (Milde) DeNot. Distributed throughout the United 
States, this moss grows on bark in the Smoky Hills, the Red Hills, and the 
Wellington Lowland. 

*Tortula ruralis (Hedw.) Smith. This widely distributed species 1s 
fairly common on rocks and soil in the High Plains. It is rare in the Red 
Hills and the Smoky Hills. 

Weissia controversa Hedw. Widely distributed, this common moss grows 
on soil and rocks in the Smoky Hills. It is a rare moss in the Blue Hills, the 
Red Hills, the High Plains, the Wellington Lowland, and the Great Bend 
Lowland. 

*Weissia glauca Bartr. This moss has been previously reported from 
Arizona and Texas. It is rare in the Great Plains on rocks and soil on 
Ogallala outcrops in Cheyenne, Rawlins, Wallace, and Scott counties and 
on Cockrum Sandstone outcrops in Stanton County. 

*Weissia microstoma (Hedw.) C. M. This moss has been reported from 
Massachusetts, New Jersey, Ohio, and Texas. It is most abundant on lime- 
stone outcrops in the Blue Hills. It has also been found on Smoky Hill 
Chalk outcrops in Phillips County in the High Plains, on Ogallala outcrops 
in Clark County in the High Plains, and on Flowerport Shale outcrops in 
Barber County in the Red Hills. 


BIBLIOGRAPHY 


Anprews, A. L. 1915. Bryological notes I. Aschisma kansanum, a new species with remarks 
upon the genus. Torreya 15:63-67. 


Baynr, C. K. 1956. Geology and ground-water resources of Reno County, Kansas, State 


Geological Survey of Kansas, Bulletin 120. 


/ ——. 1960. Geology and ground-water resources of Harper County, Kansas. State Geological 


Survey of Kansas, Bulletin 143. 


Bryan, V. S. 1956. Chromosomes and systematic position of the inoperculate mosses, Pleuridium 
and Bruchia. Amer. Jour. Bot. 43:460-468. 


. 1956a. Cytological and taxonomic studies of some species of Astomum, Acaulon and 
Phascum. The Bryologist 59:118-129. 


Conarp, H. S. 1956. How to know the mosses and liverworts. Revised edition, Wm. C. Brown 
Company, Publishers, Dubuque, Iowa. 

. 1959. Amblystegium. The Bryologist 62:96-104. 

Criptanp, A. A. 1959. The habitat of Aschisma kansanum. The Bryologist 62:132-135. 


472 Tue Universiry ScIENCE BULLETIN 


. 1960. Notes on Kansas mosses. The Bryologist 63:51-53. 

Crum, H. A. 1956. Lindbergia brachyptera in North America. The Bryologist 59:203-212. 

AND ANpERSON, L. E. 1955. Taxonomic studies in the Funariaceae. The Bryologist 

58:1-15. 

FeNNEMAN, N. M. 1931. Physiography of western United States. McGraw-Hill Book Company, 
Inc., New York. 

FisHeL, V. C. anp A. R. Leonarp. 1955. Geology and ground-water resources of Jewell County, 
Kansas. State Geological Survey of Kansas, Bulletin 115. 

Fora, S. D. 1948. Climate of Kansas, report of the Kansas State Board of Agriculture. Vol. 
LXVII, June, 1948, no. 285. 

Frye, J. C., A. B. Leonarp, AND A. SwineForp. 1956, Stratigraphy of the Ogallala Formation 
(Neogene) of northern Kansas, State Geological Survey of Kansas, Bulletin 118. 

GorseL, K. 1905. Organography of plants, part II. (Translated by Isaac B. Balfour.) The 
Clarendon Press, Oxford. 

Grout, A. J. 1928-1940. Moss flora of North America, north of Mexico. 3 volumes, published 
by the author. Newfane, Vermont. 

Horr, W. H. 1937. The flora of “Rock City.”” Trans. Kans. Acad. of Sci. 40:193-194. 

Kino, P. B. 1951. The tectonics of middle North America, middle North America east of 
the Cordilleran System. 2nd printing 1954, Princeton University Press, Princeton, New 
Jersey. 

Kucyniak, J. 1952. The occurrence of Grimmia teretinervis Limpr. in North America. The 
Bryologist 55:35-47. 

Larra, B. F. 1950. Geology and ground-water resources of Barton and Stafford counties, 
Kansas. State Geological Survey of Kansas, Bulletin 88. 

McGrecor, R. L. 1950. Studies on Kansas mosses I. Univ. Kans. Sci. Bull. 33:291-311. 

anp E. L. Hartman. 1956. Studies on Kansas mosses II. Univ. Kans. Sci. Bull. 38: 

B3)1=333.. 

McLaucuun, T. G. 1943. Geology and ground-water resources of Hamilton and Kearny 
counties, Kansas. State Geological Survey of Kansas, Bulletin 49. 

Prummer, N. anv J. F. Romary. 1947. Kansas clay, Dakota Formation. State Geological 
Survey of Kansas, Bulletin 67. 

Ropinson, H. 1962. Generic revisions of North American Brachytheciaceae. The Bryologist 
65:73-146. 

Sayre, G. 1952. Key to the species of Grimmia in North America. The Bryologist 55:251-259. 

ScHorwr, W. H. 1949. The geography of Kansas, part II, physical geography. Trans. Kans. 
Acad. of Science 52:261-333. 

SHarp, A. J. 1933. Three new mosses from Tennessee. The Bryologist 36:20-23. 

SwineForp, A. 1955. Petrography of upper Permian rocks in south-central Kansas. State Geo- 
logical Survey of Kansas, Bulletin 111. 

Totstep, L. L. anp A. SwineForp. 1959. Kansas rocks and minerals. State Geological Survey 
of Kansas. 

Watters, K. L. 1961. Geology and ground-water resources of Sumner County, Kansas. State 

Geological Survey of Kansas, Bulletin 151. 


473 


MossEs OF THE GREAT PLAINS AND ARKANSAS River LowLaNps 


“royne ayy Aq paqda][o9 Udsaq JAY SassOL a1Z4M 
A} [RIOT B suasaidas Ystayse YRY (“GPHT ‘AMIOYIS WOT) “AMI0YIS *H{ JaIEA\ “IC Jo uorsstursad Aq pas—, ‘sesuey Jo depy s1ydesSosAyg *] ‘Of 


NVALVId XYVZO 
. Ses lah A eee Pees SNIVId tV3y9 


yee me ee SuRE TS erie eh ean een 
a i ino ueeuilley: a as es } 
*% 


le a f SNiva | 
sass qua: joe: ae \ S i j a — ete Stl o Ns a | ae H ; 
(Ea an one a | il PR YO Os” A pn see 
ae =a) » af} | \ SS] | =P " Sreinitns SS a | i * + 
ORTHO oh i ee! j i j 
Seema | ; oles iH \ he poe ere nO 7 ‘ 
a @8403849 a ee T ee Ren j 
| | ' | 4 [ Ste B 4 nO \ : me 
i cee = EL aoonnazv$ | feral . at fi d, SQNV, rk sae ! 
| Ga aCe ssn XO ex | 
SNIVId- "3 9¥S0 TP puamoW 2] owoums: 559) | 
| watt | roneow | 34509 ' ; mae ! pibg wi | i i “s 
| sojeono (eto80! ) aw Hy coe Ne eee era i i ei i i | 
i | wt $ ue ES i a ee 1S usnu | Ae ss as | : 
Send dees eevee. enna % Cees Pa ieee 
. W390 | { Siweon J } | et N — : os a ae 
| ne “ecco 4 ; “Jaton a aes =] ay - | ‘ 
pothee spree SAL i [ 92 SNIVId HOIH | 


| | * , 


| 

| | 
xa 2409 | Nv900 j 
5 ! ees See ea nt. eine 


. oa) } 
$ PN: & Timms 1A apie ey ee (7: 

5 “ro4y Lbiac FEY p Ne me ES reo a ui v) | 
LMM NOREASIIE | Q *P8p, \ Ox ! iS = FOE s (Op | l 
SE ean oy ep ow the to | 

| NOBNONE | aWOLwm re vt a0 | Wx ~ n * s youre sw00u | ww I ‘ 
\\ i sesssiuill Peete 3 dese) (Cy tore | aa I eens aad et SY ROHAHiT venues | 
jnowmeuy jf ani ee ar | | | Se Wiles seers 
Deas I 4ISNIW Id: i ENING egies i P 
Emons mL OILOASSIO! ~y ret | i owe | ! ! a! 
i} ta ike | i i 22 near 
“iam wren peer | ibngye sy _ovenan | Eee | mame | saving sown nue! ganna gu | 


aNyTMOT SS ae SNIGNeimlgauBG ce Lt meio 


THe UNIversiry ScIENCE BULLETIN 


474 


yctiaan ~ CIO PRL CBE 
Vacs Corer Ses | A Lip e- GOUOAKMA SI at ie 
HAG \ ynonwinvHo alates 
f L 5 Yj PY sun Go =i | aN 
US seal SORE Fa i tee 
’ | . Lo Lod WT SET 2, "2 oa 
: ) : } 
| = VB ase ; j — 
ORE eae 4 ows | nostm | A veo > >> oe Garon ae eh 
i RG Ae ry 0 : Glas BM Sowiase =, = mae Trt. 
| | t By202 | 45imd03 g40j Sees Yr) = OAD) 
{ -(\snoauby | ca tt Shoo pts ound | 
| Noaunoal: NatnWe _NOSGOOM | : 
‘ eel eee DAS: —| — a00mN33u9 _ 
=e | jae 
| ae a an 


San : «! ° 1 
ar ees = ABNNIZ{* ANUWSN * note 

ae 33NMVd a Ts 1 see ae epee bn cere a ate 
ae ON: ieee eats 4 5 


eee TS 
o* 1 ; 
nets Wr i oun 
rs os i 7) a 
[aera saniavetl : | 
: — . 
ve Oe =eaeree | AVSi|| nen 
1. ,S¥79N00 j ale p 
Den. ope 
Se 
Bf of 
— ff -—- VAVLLO | 
\ O Wo} oD 5. Of ff f- Yf SNOOUD/(xX-/ AL \--—--—- Dale 
HLYOM | Qo = 


“Ngawa7 | NOSH3II30 | 
ee S65 ee 
nme o!l 

»\ ell ye 
‘ NOSIMD1V o 
Sees 


eS H 
¢ NvxdiNOO | oS 


aes oveiline 
et 2 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


SOME INTERSECTIONAL HYBRIDS AND 
RELATIONSHIPS IN HAPLOPAPPUS 


By 
R. C. Jackson 


Vot. XLVI Paces 475-485 Fepruary 1, 1966 No. 13 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Vou. XLVI Paces 475-485 Fepruary |, 1966 No. 13 


Some Intersectional Hybrids and 
Relationships in Haplopappus’ 


By 


R. C. JAcKson 


INTRODUCTION 


Over the past several years, morphological, cytological and hybridization 
data have been accumulated for several distantly related species of the genus 
Haplopappus as circumscribed by Hall (1928). These data support parts of 
Hall’s treatment but point up some basic errors in others, particularly in his 
scheme of phylogenctic relationships based on certain “primitive” types in 
section Blepharodon. Section Prionopsis, certain disruptive elements of sec- 
tion Blepharodon, and one proposed new section are dealt with here. Most 
of the species discussed have been described in detail by Hall and only perti- 
nent or general morphological data are presented. 

The cytological and hybridization techniques used in this study have 
been described in detail elsewhere (Jackson, 1962, 1965). Voucher specimens 
of the intersectional hybrids and of the species with chromosome counts 
reported for the first time are in the University of Kansas Herbarium. Col- 
lection data for the chromosome counts are as follows: Haplopappus brick- 
ellioides Blake, Clark Co., Nevada, east of Mercury in the southern part of 
the Spotted Range, Oct. 1963, Johnson 2117; Haplopappus phyllocephalus 
subsp. primitivus Hall (=Machaeranthera), State of Mexico, about 15 miles 
southeast of Queretaro border along hyw. 57, Aug. 9, 1964, Jackson 5157; 
Haplopappus aureus Gray, Harris Co., Texas, jctn. of hyw. 290 and West 
39th St., Houston, Oct. 7, 1964, R. C. Jackson 5242. 


"This study has been supported by NSF Grant GB-3071. 


476 Tue UNIversiry SCIENCE BULLETIN 


GENERAL RELATIONSHIPS OF THE SECTIONS 


Blepharodon, as delimited by Hall, is perhaps taxonomically the most 
difficult of the North American sections of Haplopappus. Field and morpho- 
logical studies and cytogenetic analyses of Fi hybrids have served to clarify 
and define certain diagnostic characters which delimit the section. The basic 
chromosome number is X==4 (Jackson, 1962). The basic karyotype appears 
to be stable in both the diploids and tetraploids studied so far, and it is quite 
distinct from those of several taxa previously included in the section (Fig. 
1-4). The section has one chromosomally polymorphic species, H. gracilis 


x y 


a 


Fics. 1-4. Root tip chromosomes. Fic. 1. Haplopappus spinulosus (Pursh) DC. (sect. 
Blepharodon). The arrow at upper left designates a chromosome ca. 5u in length. Figures 3 
and 4 are at a somewhat larger scale. Fic. 2. H. havardii (sect. Havardia). Fig. 3. H. annuus 
(sect. Isocoma). Fic. 4. H. ciliatus (sect. Prionopsis). 


Some INTERSECTIONAL Hysrips AND RELATIONSHIPS IN HapLopappus 477 


(Nutt.) Gray, which has races with n=2 and n=3 (Jackson, 1965). Super- 
numerary chromosomes have been found in most of the diploid and aneu- 
ploid taxa (Jackson, 1960; Raven et al., 1960; Jackson & Li, 1961; Jackson, 
unpubl.). 

In contrast to Hall’s (1928) treatment, Cronquist and Keck (1957) agreed 
with Shinners’ (1950) transfer of Haplopappus gymnocephalus DC. and 
H. nuttallu Torr. & Gray from section Blepharodon to the genus Machae- 
ranthera. My own study of the type of Haplopappus phyllocephalus subsp. 
primitivus Hall (n=4) and living material at the type locality leave no doubt 
that it too is a Machaeranthera, probably M. gymnocephala. As the name 
primitivus indicates, Hall considered this subspecies as the ancestral type of 
several annual taxa of section Blepharodon. 

The other subspecies of H. phyllocephallus (sensu Hall) are being re- 
tained in Haplopappus and have been studied in some detail. Haplopappus 
phyllocephalus DC. subsp. phyllocephalus and H. p. subsp. megacephalus 
(Nash) Hall are merged under H. phyllocephalus. H. p. subsp. annuus is 
considered as a distinct species, H. annuus (Rydb.) Cory. Crosses between 
H. annuus and H. phyllocephalus have demonstrated a sterility barrier in 
the F; hybrid, and I have not found that H. pAyllocephalus passes insensibly 
into H. annuus as stated by Hall. These two species and the shorter, endemic 
H. aureus Gray are annuals with a chromosome number of »=6 (Jackson, 
1960). In addition to chromosome number, there are several morphological 
differences that set these three species apart from all other taxa that Hall in- 
cluded in section Blepharodon. All of these differences need not be dealt 
with here, but the one of disc corolla shape is most striking and suggestive of 
certain relationships (Fig. 5-10). In the three species listed above the disc 
corolla is a slender tube that is abruptly dilated above the middle (Fig. 8). 
All other species of section Blepharodon have a disc corolla that gradually 
tapers and is funnelform (Fig. 5). This character difference is most striking 
in living plants and is easily discernible in dried material. Shinners (1950, 
p. 39) apparently did not notice this character when he merged section 
Blepharodon (sensu Hall) with Machaeranthera, stating that the two differ 
in no important technical feature of florets, achenes, or pappus, but only in 
the trivial one of ray color. 

The disc corolla shape in H. phyllocephalus, H. annuus, and H. aureus 
is indistinguishable from that found among various taxa of section [socoma 
(Fig. 7). Moreover, this section has a basic chromosome number of X=6, 
and the somatic and meiotic chromosomes show staining reactions similar to 
those of these three species. As generally recognized, section Isocoma is char- 
acterized by discoid heads with turbinate to narrowly campanulate involucres 
compared with radiate heads and broadly campanulate involucres of the 
three species listed above. However, vestigial rays have appeared in at least 


478 THe University SciENCE BULLETIN 


one species of section Isocoma that I have in cultivation so this qualitative 


difference is no longer absolute for separating the two groups. 
6 ZL 10 
ta 13 


9 


| 
v4 


11 


Fics. 5-13. Disc florets (9) and median leaves (X1) of Haplopappus species and hybrids. 
Fic. 5. H. spinulosus (Pursh) DC. (sect. Blepharodon). Fic. 6. H. havardii (sect. Havardia). 
Fic. 7. H. venetus (sect. Isocoma). Fic. 8. H. annuus (sect. Isocoma). Fic. 9. H. ciliatus X 
H. annuus. Fic. 10. H. ciliatus (sect. Prionopsis). Fic. 11. H. ciliatus. Fic. 12. H. ciliatus 


< H. annuus. Fic. 13. H. annuus. 


SoME INTERSECTIONAL Hysrips AND RELATIONSHIPS IN HapLopaprus 479 


In view of the data presented, H. phyllocephalus, H. annuus, and H. 
aureus are tentatively assigned to section Isocoma. The origin of these an- 
nuals is almost certainly centered in this primarily perennial section, and they 
may have derived from certain broad-leaved, campanulate involucral forms 
of the Central Mexican Highlands comprising H. venetus (H.B.K.) Blake 
subsp. venetus and perhaps some undescribed taxa. 

Another disruptive element in section Blepharodon is H. brickellioides 
Blake. This is a rarely collected species of Death Valley and adjacent regions 
of California and Nevada with a chromosome number of n=6. In leaf, in- 
volucre, and flower shape, it is closely related to species of section Hazardia 
and probably should be included in this group. Contrary to Hall’s descrip- 
tion, the species does have ray flowers although they are small. 

Haplopappus havardu was described by Waterfall (1943) as allied to the 
H. phyllocephalus complex. In annual habit, leaf shape, and overall size, it is 
similar to H. phyllocephalus and related species. However, in disc flower 
shape (Fig. 6), chromosome number (n=4; Jackson, 1960), and karyotype 
(Fig. 2) it is quite distinct from these taxa. Furthermore, its fusiform achenes 
are different from any of the species of section Blepharodon (as delimited 
here) and [socoma. Although a genetic relationship has been demonstrated 
with H. annuus, it is sufficiently distant so that taken together with its mor- 
phological and chromosomal divergence from other species, H. havardit must 
be considered as representing a separate section, and such is proposed here. 


Haplopappus section Havardia R. C. Jackson, sect. nov. 

Planta erecta rigida foliosa glandulose annua. H. annuo affinis, a quo 
differt glandulis confertioribus; foliis numerosis minoribus obtusis breviter 
dentatis; capitulis minoribus ramulos conspicue foliosos terminantibus; 
acheniis fusiformis. 

Type and only known species: Haplopappus havardu Waterfall 


Haplopappus section Prionopsis has but one species, H. ciliatus (Nutt.) 
DC. Originally described as a distinct genus, the section was doubtfully 
placed in Haplopappus by Hall (1928) who remarked on its similarity to 
Grindelia. The chromosome number of H. ciliatus is n=6 (Jackson, 1959) 
as is that of several species of Grindelia, but attempts to cross the two genera 
have thus far been unsuccessful. The general similarity in leaves, pappus, 
and resinous-vicid involucres of the two taxa may be the result of paralleled 
evolution or it may be indicative of a real but distant genetic affinity. 

During mitosis and meiosis, the chromosomes of H. ciliatus are similar in 
size and staining reaction to those of H. phyllocephalus and H. annuus. The 
karyotype of H. ciliatus is, however, quite different from these species (Fig. 
1-4). Although a genetic relationship can be demonstrated with H. annuus, 
A. ciliatus is morphologically and cytologically distinct from any other species 


480 Tue UNiversiry SciENCE BULLETIN 


of Haplopappus. In doubtfully including the species in Haplopappus, Hall 
suggested that even if a closer relationship of H. (Prionopsts) ctliatus to 
Grindelia could be demonstrated, this would scarcely bring the two genera 
together. In my opinion, H. ciliatus should be retained in Haplopappus as a 
monotypic section because of the demonstrated genetic relationship with 
H. annuus and indirectly with other taxa by way of this species. 


INTERSECTIONAL HYBRIDIZATION 
Haplopappus (Prionopsis ) ciliatus (n=6) * H. (Isocoma) annuus (n=6). 


Hybrids of this combination were produced in 1961 and again in 1964. 
They were generally intermediate for leaf and flower characters of the par- 
ents (Fig. 8-13). However, their growth and maturation were somewhat 
slower than H. annuus and more closely resembled H. ciliatus. Several heads 
were fixed for cytological analyses and others were used for backcrossing to 
H. ciliatus. 

Meiotic prophase stages beginning with early pachytene were studied in 
as much detail as possible. Details of synapsis at pachytene were difficult to 
analyze because most of the chromosome ends seemed to be involved in more 
complex arrangements than bivalents. However, deletion or duplication con- 
figurations were noted in several cells (Fig. 14), and what appeared to be a 
combination of a translocation and inversion configuration involving at least 
four chromosomes (Fig. 15) was observed in one cell. The latter configura- 
tion may result in a plant heterozygous for an inversion with a translocation 
distal to it. 

In a few cells at an early diplotene, all 12 chromosomes appeared to be 
connected in a branched, complex chain (Fig. 16). At later diplotene (Fig. 
17), ten chromosomes were involved in a branched chain in at least five cells. 
By diakinesis, terminalization had broken down the long chains to smaller 
ones (Fig. 18). Table 1 gives the kinds of configurations and their frequen- 
cies at late diplotene and diakinesis. 

Metaphase I was usually characterized by two short chains, bivalents, and 
univalents but this and later stages were not analyzed in detail. The second 
meiotic division showed lagging chromosomes and occasional bridges at 
anaphase II. At telophase II, there were several to many micronuclei which 
yielded many abnormally small microspores. 

In contrast to the 96-99°%, pollen fertility of the parental species, the Fi 
had one stainable pollen grain out of the 1012 examined (0.09°%), and it 
apparently resulted from an unreduced microsporocyte. 

Repeated backcrossing of the Fi to H. ciliatus produced no viable seed 
from either parent. 


Some INTERSECTIONAL Hysrips aNp RELaTIoNnsuips IN Haptopappus 481 


Haplopappus (Isocoma) annuus (n=6)  H. (Havardia) havardu GS 


Numerous attempts have been made to cross these species. A number of 
seed had small, aborted embryos, and one Fi seedling was obtained several 
years ago. This plant grew very slowly to a height of about three cm. Leaves 


Fics. 14-18. Meiotic stages in H. ciliatus X H. annuus. Fic. 14. Segments of pachytene 
chromosomes showing loops (arrows) which are characteristic of either duplications or dele- 
tions in paired chromosomes. Fic. 15. Arrow indicates an inversion configuration involving 
three or more chromosomes. Fic. 16. Early diplotene showing a branched chain of 12 chromo- 
somes. Fic. 17. Mid-to late diplotene with a complex chain of ten chromosomes. Possibly the 
chain was torn away from an association with the nucleolar organizing bivalent. Fic. 18. A 
chain of seven chromosomes, two bivalents, and an univalent at diakinesis. 


482 THe University SciENcCE BULLETIN 


Taste 1. Pairing configurations and minimum chiasmata at late diplotene and 
diakinesis in the F; hybrid of Haplopappus ciliatus (n=6) < H. annuus (n=6). 


No. cells Chain(s) Pair(s) Univalent(s) Minimum chiasmata 
5) 10 1 = 10 
l 10 Z 9 
] 9,3 10 
] 9 ] 1 9 
] 8, 4 10 
1 8, 3 = 1 9 
2 8 2 9 
2 8 1 2 8 
3 Tad 2 8 
2 7 2 l 8 
3 7 1 3 6 
1 6,3 1 2 7 
1 5,4 1 2 7 
1 yes) = 1 8 
1 Da3 2 8 
1 5 3 7 
2 5 2 6 
2 4,4 2 _ 8 
3 4.3 2 1 7 
] 4h 8) 1 3 6 
3 33,5) 2 2 6 
1 3 4 1 6 

Total 39 7 a 305 
X=7.8 


and stems showed irregular patches of tissue that closely resembled “sector- 
ing” in irradiated plants. Toward the end of its growth, the Fi produced 
many small heads two to three mm in diameter, but mature flowers were not 
formed. Since meiotic material was not available for analyses, somatic 
chromosomes and nuclei of some young leaf tissue were studied. Ten 
chromosomes (Fig. 19, 20) were found in one cell and these corresponded 
in size and structure to those of the parental species. Several anaphase stages 
had a chromosome bridge, and in some cells it appeared that fragmentation 
of chromosomes had occurred. Several late telophase nuclei were connected 


by a chromatin bridge (Fig. 21). 


DISCUSSION 


The morphological hiatus between Haplopappus (Prionopsis) ciliatus 
and H. (Isocoma) annuus is accompanied by an equally great change in 
karyotype. Haplopappus ciliatus has four sets of metacentric and two sets of 


SoME INTERSECTIONAL Hyprips AND RELATIONSHIPS IN HapLtopappus 483 


acrocentric chromosomes (Fig. 3), whereas H. annuus has four sets of acro- 
centric and two sets of metacentrics (Fig. 4). 

The difference between the two species in relative position of centromeres 
may have been caused by pericentric inversions, unequal reciprocal trans- 
locations, or a combination of both. However, analyses of pairing in the F; 
hybrid of the two species indicate that the primary changes have been trans- 
locations. The complex, branched chains of chromosomes observed at early 
diplotene suggest that all of the chromosomes were involved in the configura- 
tion. Cells at late diplotene and diakinesis seemed to bear this out when 
chains of up to ten chromosomes were observed. Branching of the chromo- 
some chains is probably the result of interstitial and/or multiple transloca- 
tions in which one chromosome may be made up of segments of several 
different chromosomes in comparison to the ancestral condition. 

Six chiasmata is the minimum number needed to hold the six parental 
bivalents together until metaphase I. The mean minimum chiasma fre- 
quency per cell was 7.5 in H. annuus and 8.2 in H. ciliatus. The means were 
derived from a study of 20 cells for each species. The Fi had a mean of 7.8 
chiasmata per cell. Despite the large number of translocations and the prob- 
able reduction in chiasmata caused by them, these data indicate that a high 
degree of homology still exists among the chromosome segments of the 
parental species. 

The arrangement of the Fi chromosomes at metaphase I indicated that 
anaphase I segregation involved short chains, some bivalents, and univalents. 
This would probably yield a lower pollen fertility than the expected 1.56%, 
from a translocation heterozygote with six independent translocations, as- 


19 20°55 ‘ey 


Che 


Fic. 19-21. Leaf squashes of Haplopappus annuus X H. havardii. Fics. 19, 20. Photograph 
and tracing of same showing ten chromosomes at late prophase of mitosis. Note difference in 
size of chromosomes. Fic. 21. Chromatin bridge connecting leaf cell nuclei at late telophase. 


484 Tue Universiry SciENCE BULLETIN 


suming that a single translocation causes 50° sterility. The relatively high 
chiasma frequency of the Fi: may have contributed to the sterility if the de- 
letion or duplication loops noted at pachytene were the result of inversions 
and re-inversions. Estimations of sterility based on the translocations and 
other aberrations indicate that the sterility barrier separating H. annuus and 
H. ciliatus is largely due to structural repatterning of the chromosomes of 
one or both parental species. 

In gross morphology H. havardiu (n=+4+) closely resembles H. phyllo- 
cephalus and H. annuus (n=6) more than any other species of the genus, 
even though it differs from these species in chromosome number and disc 
corolla shape. Furthermore, genetic relationship, as demonstrated by the 
production of an Fi hybrid, is only with H. annuus and indirectly with H. 
phyllocephalus because the latter crosses readily with H. annuus. Attempts 
at crossing H. havardi to species of other sections have failed repeatedly. 

Since the hybrid between H. annuus and H. havardii is the first in the 
genus between a four- and six-paired species, it is unfortunate that meiotic 
material was not available for analysis. The chromosomes of H. havardu 
are larger than those of H. annuus in parental plants and in the Fi. In view 
of the morphological similarity and demonstrated genetic link of the two 
species, and the relationship of H. annuus to section Isocoma (X=6), I sug- 
gest that »=6 is more primitive in this case than n=4. The origin of n=4 
from n=6 probably resulted from unequal reciprocal translocations and loss 
of centromeres as previously shown for another species of Haplopappus 
(Jackson, 1962). This is not to say, however, that one species is older than 
the other; the inference is simply that both species had their origin in an 
ancestral group with X=6. 

The hybridizations among the three species discussed here verify a genetic 
relationship which was previously assumed because of morphological simi- 
larity. Other than this, the lineage or true phylogenetic relationships among 
them are unknown and will probably remain so because connecting morpho- 
logical forms are no longer extant. They are well isolated by internal barriers 
(Stebbins, 1950) in the hybrids. In the case of H. annuus X H. ciliatus there 
is chromosomal sterility. Haplopappus annuus and H. havardu are separated 
by hybrid weakness and failure of flowering, and, if meiotic stages had been 
produced, there would surely have been chromosomal sterility. 


LITERATURE Cli bp 


Cronouist, A. AND D. D. Kecx. 1957. A reconstitution of the genus Machaeranthera. Brit- 
tonia 9:231-239, 

Hari, H. M. 1928. The genus Haplopappus: a phylogenetic study in the Compositae. Carnegie 
Inst. Wash. Publ. No. 389. 1928. 

Jackson, R. C. 1959. In Documented Chromosome Numbers of Plants. Madroftio 15:52. 

. 1960. Supernumerary chromosomes in Haplopappus gracilis. Evolution 14:135. 

. 1960. In Documented Chromosome Numbers in Plants. Madrofio 15:222. 


SoME INTERSECTIONAL Hysrips AND RELATIONSHIPS IN HapLopappus 485 


. 1962. Interspecific hybridization in Haplopappus and its bearing on chromosome evolu- 

tion in the Blepharodon section. Amer. Jour. Bot. 49:119-132. 

. 1965. A cytogenetic study of a three-paired race of Haplopappus gracilis. Amer. Jour. 

Bot. (in press). 

Li, N. ano R. C. Jackson. 1961. Supernumerary chromosomes in Haplopappus spinulosus ssp. 
cotula. Amer. Jour. Bot. 48:419-426. 

Raven, P. H., O. T. Sorsric, D. W. KyHos, anp P. Snow. 1960. Chromosome numbers in 
Compositae. I. Astereae. Amer. Jour. Bot. 47:124-132. 

SHINNERS, L. H. 1950. Notes on Texas Compositae. V. Field & Lab. 18:32-42. 

SresBins, G. L. 1950. Variation and evolution in plants. Columbia Univ. Press, N.Y. 

WarerFALL, U. T. 1943. A new species of Haplopappus from southwestern Texas. Rho !ora 


45730-3953. 


mi 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


EMBRYOLOGY OF HAPLOPAPPUS GRACILIS 
(NUTT.) GRAY 


By 
S. Venugopalan 


VoL. XLVI Paces 487-493 June 17, 1966 No. 14 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 487-493 June 17, 1966 No. 14 


Embryology of Haplopappus Gracilis (Nutt.) Gray’ 
By 


S. VENUGOPALAN~ 


INTRODUCTION 
The genus Haplopappus, monographed by Hall (1928), is of late receiv- 


ing much attention by experimental taxonomists. Interspecific hybridization 
among various species of the section Blepharodon and related cytogenetic 
studies have been done in this genus (Jackson, 1959, 1962, 1963, 1964; Li and 
Jackson, 1961). With the report of the diploid chromosome number as four 
for Haplopappus gracilis by Jackson (1957), this species has become a handy 
tool for radiation cytologists (Kamra & Kamra, 1962), plant physiologists 
(Blakely & Steward, 1964), and studies on growth and development by tissue 
culture methods. Harling (1951) described briefly the embryology of two 
species of this genus, H. croceus and H. lyallii, in a comparative study of the 
tribe Astereae. Except for this, little is known of the embryology of the genus. 
Since embryological characters are known to be of value in the elucidation 
of taxonomic problems at various levels and of phylogeny as well (Mahesh- 
wari, 1950, 1963), the present work was undertaken as a preliminary to a 
comparative embryological study of the different sections of Haplopappus 
(sensu Hall, 1928). 


MATERIALS AND METHODS 


The materials were collected from plants growing in the greenhouse. 
Carnoy’s fluid and formalin-acetic-alcohol were used as fixatives. Dehydration 


* This study has been made possible by a Danforth Indian Fellowship Grant to the author 
and by National Science Foundation Grant 20864 to Dr. R. C. Jackson. 


~ Present address: Vivekananda College, Madras-4, India. 


488 THe UNIversiry ScIENCE BULLETIN 


was performed with ethanol and tissue mat was used for embedding. Sections 
were cut at 10-18» and stained with Heidenhain’s Iron-alum haematoxylin 
and counterstained with Fast green or Orange G. 


OBSERVATIONS 


Ovule. The primordium of the ovule develops as a ventral protuberance 
from the floor of the ovary and eventually assumes an anatropous position. 
The ovule is monotegmic. Even as the archesporium is being differentiated, 
the primordium of the integument stands out on the side away from the 
raphe. In a well developed ovule the integument is a massive structure of 
several cells in thickness. The micropyle appears to be a uniformly narrow 
canal leading to the nucellus which can be hardly distinguished. From its 
inception the nucellus is a poorly developed tissue represented by a layer of 
cells surrounding the megaspore mother cell. By the time the functional 
megaspore begins to elongate, the nucellar cells disorganize and appear as 
darkly stained patches. The integumentary cells adjoining the nucellus be- 
come differentiated into a distinct endothelium. Its cells remain uninucleate 
throughout, although they divide to cope with the elongating embryo sac 
and the ovule. 

Megasporogenesis. The archesporium consists of a hypodermal cell both 
in H. gracilis and H. spinulosus, and it functions directly as the megaspore 
mother cell (Fig. 1, 2). The two successive divisions of the mother cell 
result in a linear tetrad of megaspores (Fig. 3). The chalazal megaspore 
functions while the remaining ones degenerate (Fig. 4). The functional 
megaspore elongates toward the micropyle and forms the embryo sac. In H, 
croceus, Harling (1951) reported that the chalazal as well as the micropylar 
megaspores appear to enlarge. 

Embryo Sac. Some of the representative stages in the development of the 
embryo sac are illustrated in Figs. 5 to 8. The degeneration of the nucellar 
cells brings the embryo sac close to the integument. The embryo sac shows 
eight nuclei, resulting from three mitotic divisions of the nucleus of the 
functional megaspore (Fig. 8). Three of these organize into the egg apparatus 
with two pyriform synergids and an egg cell, all of which lie in the same 
plane. The polar nuclei frequently lie near the egg (Fig. 8). 

The three antipodal nuclei in the eight-nucleate embryo sac do not re- 
main together as is the case in other angiosperms. One of them lies in the 
middle of the embryo sac while the other two are farther towards the chalazal 
end. The sister nucleus of the lower polar nucleus divides while near the 
middle of the embryo sac. Each set of antipodal nuclei is embedded in a 
separate mass of cytoplasm, but there is no evidence of wall formation before 
fertilization. The mature embryo sac thus consists of nine nuclei at the time 


EmpryoLocy oF Haplopappus gracilis (Nutr.) Gray 489 


of fertilization (Fig. 8). However, following fertilization the embryo sac is 
partitioned into three chambers (Fig. 9): the micropylar chamber with the 
zygote and the primary endosperm nucleus, the middle chamber with a 
binucleate antipodal cell, and the chalazal chamber, rather hypertrophied and 
uninvested by endothelial cells. The nuclei in the middle chamber unite 
to form a larger nucleus with many nucleoli. The nuclei at the chalazal pocket 
also tend to fuse. Similar nuclear fusions in antipodal cells have been reported 
for Tridax procumbens (Maheshwari & Roy, 1952) and T. trilobata (Kapil 
& Bala Sethi, 1962). But in the above two cases the antipodal cells persist as 
elongated tubular processes whereas they are vesicular and ephemeral in 
Haplopappus gracilis. There is no division of the primary antipodal cells as 
has been reported by Harling (1951) for H. /yallit. 

Endosperm and Embryo. The division of the zygote is transverse and 
it precedes that of the endosperm nucleus (Fig. 10). Subsequent divisions of 


Ficures 1, 3-15. Haplopappus gracilis. Ficure 2. H. spinulosus. 
Ficures 1-7. Fic. 1. Megaspore mother cell, X 353. Fic. 2. Hypodermal archesporium with 
the primordium of the integument, X 193. Fic. 3. Linear tetrad, X 353. Fic. 4. Same with 
three degenerate megaspores. Fic. 5. Two nucleate embryo sac, X 193. Fic. 6. Four nucleate 
embryo sac, X 193. Fic. 7. Eight nucleate embryo sac, X 353. 


490) Tue Universiry ScreNcE BULLETIN 


the proembryonal cells could not be followed for want of adequate prepara- 
tions. The primary endosperm nucleus first divides vertically but there is no 
wall formation. Further divisions of these two free nuclei result in the forma- 
tion of cellular endosperm. The development of endosperm is strictly con- 
fined to the micropylar chamber during the early stages. But soon, with the 
haustorial action of the antipodal cells and enlargement of the embryo sac, 
formation of endosperm extends to the chalazal region. Digestion of the 
endosperm cells does not begin until the heart-shaped stage of the embryo 
(Fig. 14). In the mature seed, however, all but the outermost layer of endo- 
sperm disappears. The endothelial cells degenerate as does also the integu- 
mentary tissue adjoining them. The early heart-shaped stage of the embryo 


Ficures 8-13. Fic. 8. Mature embryo sac; note the four antipodal nuclei, K 193. Fic. 9. 
Fertilized embryo sac; The primary endosperm nucleus lies close to the zygote and the antipodal 
cells are binculeate, X 353. Fic .10. Bicelled proembryo and endosperm nucleus, X 353. 
Fic. 11. L. s. ovule showing the two binucleate antipodal cells and their haustorial action on the 
chalazal ovular tissue: ant = antipodal cells; ¢ = chalazal tissue affected by haustorial action of 
antipodals; e = endothelium; t = testa; X 36. Fic. 12. Early heart-shaped embryo, X 193. 
Fic. 13. Early dicotyledonous embryo, & 193. 


EmpryoLocy oF Haplopappus gracilis (Nutr.) Gray ap 


(Fig. 12), the mature embryo with cotyledons, and a three celled suspensor 
are shown in Figs. 12 and 13. 

Seed Coat. Of the fourteen or more layers of cells making up the integu- 
ment in a mature ovule, all but four or five layers degenerate under the 
haustorial action of the enlarging chalazal antipodal chamber. The integu- 
ment in a mature seed consists of rectangular and uninucleate flattened cells 
(Fig. 15). The pericarp external to the testa consists of four layers of 
flattened cells and some remnants of vascular tissue. Long epidermal hairs 
are seen on the surface of the pericarp. No special thickenings were observed 
on the testa or the pericarp. 


Ficures 14-15. Fic. 14. L. s. developing ovule showing the various parts: c = tissue of the 
integument showing haustorial action of antipodals; em = embryo; end = endosperm; e = 
endothelial cells; t = testa; X 36. Fic. 15. L. s. portion of fruit: t = testa; Dp = pericarp; 


€ = endothelial cells; c = disorganized integumentary cells, & 139. 


492 Tue Universiry SCIENCE BULLETIN 


SUMMARY AND CONCLUSIONS 


The present study on Haplopappus gracilis reveals that the archesporium 
of the ovule is unicellular and hypodermal and that the development of the 
female gametophyte is monosporic and of the polygonum type. This is in 
agreement with the findings of Harling (1951) for H. /yalli1. However, these 
two species seem to differ in some details regarding the character of anti- 
podal cells. In H. gracilis, the female gametophyte has nine nuclei, constituted 
into three uninucleate cells of the egg apparatus, two polar nuclei, and four 
antipodal nuclei. The increase in number of antipodal nuclei from three to 
four is due to secondary division of one of the nuclei of the chalazal quartet. 
Further, the formation of cell walls surrounding these nuclei takes place 
only after fertilization, and two binucleate primary antipodal cells are 
formed as a result. Of these, the chalazal one enlarges considerably and the 
nuclei in this cell as well as the other antipodal cells may fuse to form a 
large multinucleolate nucleus. In some, one nucleus divides and thus a 
multinucleate cell may result. Such endomitotic, heteroploid, primary anti- 
podal cells are reminiscent of the microsporangial tapetum. Primary anti- 
podal cells of this nature are known to occur in many of the embryologically 
studied species in the tribe Heliantheae (cf. Schnarf, 1931). Two other genera, 
Grindelia and Aphanostephus, both belonging to the tribe Astereae, are 
reported by Harling to exhibit enlarged primary antipodal cells. 

In contrast with this is a secondary antipodal complex reported by Har- 
ling (1951) for H. lyalli7. In this species the two primary antipodal cells keep 
on dividing to form 12 to 15 cells. The nuclear divisions are not always fol- 
lowed by cell divisions, thus resulting in small multinucleate secondary anti- 
podal cells. These regions persist, restricting the endosperm formation to the 
micropylar region. Such a secondary antipodal complex has been held by 
Harling to be prevalent in the great majority of species showing monosporic 
development of the female gametophyte in the tribe Astereae, with the ex- 
ception of Erigeron, Vittidinia, and Aster section of Macranthera. Grangea 
maderaspatana (Mitra, 1947) and Solidago altissima (Beaudry, 1958) also 
form secondary antipodal cells. 

The prevalence of enlarged antipodal cells in Grindelia and Aphanostephus 
has been taken by Harling as indicative of the connection of these two genera 
with Heliantheae. Further, he has stated “it is perhaps no mere chance that 
Grindelia squarrosa exhibits a connection to the said tribe also in respect of 
chromosome number (n=6). The number six and multiples of this are in 
fact common in the Heliantheae but they have not been met with in the 
Astereae apart from Grindelia.” 

Two points deserve mention here: (1) Heliantheae has haploid numbers 
ranging from four to 19 and six is not the common number of all these. (2) 
In Astereae there are some species of Haplopappus with n=6, besides 


EmpryoLocy oF Haplopappus gracilis (Nurt.) Gray 493 


Grindelia and other genera, so it will be but proper to consider the correla- 
tion between embryological characters and chromosome number in Grindelia 
as fortuitous. Additional information on the embryology of the species of 
Haplopappus with a haploid number of six will be necessary before any 
valid conclusions can be drawn in this regard. In any case, secondary anti- 
podal tissue in 1. /yallii and species of Solidago suggest the latent ability of 
the female gametophyte to organize a nutritive mechanism independent of 
double fertilization. 


LITERATURE CITED 


Beaupry, J. R. 1958. Studies on Solidago L. Ill. Megasporogenesis, development of the mega- 
gametophyte and mode of reproduction in Solidago altissima L. Proc. Genet. Soc. 
Canada 3:7-14. 

BrakELy, L. M. anp Srewarp, F. C. 1964. Growth and development in cultured cells. VII. 
Cellular variation. Amer. J. Bot. 51:809-820. 

Hart, H. M. 1928. The genus Haplopappus: A phylogenetic study of the Compositae. Carnegie 
Institution of Washington. Publ. No. 389. 1-391. 

Hariine, G. 1951. Embryological studies in the Compositae. III. Astereae. Act. Hort. Berg. 
16:73-120. 

Jackson, R. C. 1957. New low chromosome number for plants. Science 126:1115-1116. 

. 1959. A study of meiosis in Haplopappus gracilis (Compositae). Amer. J. Bot. 46: 

550-554. 

——.. 1962. Interspecific Lybridization in Haplopappus and its bearing on the chromosome 
evolution in the Blepharodon section. Amer. J. Bot. 49:119-132. 

——.. 1963. Variation in the short arm of chromosome B of Haplopappus gracilis, Canad. J. 
Genet. & Cytol. 5:421-426. 

———.. 1964. Preferential segregation of chromosomes from a trivalent in Haplopappus gracilis. 
Science. 145:511-513. 

Kamra, O. P. ano Kamra, J. K. 1962. Factors influencing X-ray tolerance of Haplopappus 
gracilis seeds. Canada J. Genet. & Cytol. 4:255-262. 

Kapit, R. N. anp Bara Serut, S. 1962. Development of seed in Tridax trilobata. Phyto- 
morphology 12:235-239, 

Li, N. anp Jackson, R. C. 1961. Supernumerary chromosomes in Haplopappus spinulosus ssp. 
cotula. Amer. J. Bot. 48:419-426. 

Manesuwart, P. 1950. An Introduction to the Embryology of Angiosperms. New York. 

. 1963. Embryology in relation to taxonomy. In Vistas in Botany. Vol. III. Oxford. 

AND Roy, S. K. 1952. The embryo sac and embryo of Tridax procumbens L. Phyto- 
morphology 2:245-252. 

Mirra, J. 1947. A contribution to the embryology of some Compositate. J. Indian Bot. Soc. 
26:105-123. 

Scunarr, K. 1931. Vergleichende Embryologie der Angiospermen. Berlin. 


< an 
ie 
a [ao 
Tae @ 


«yy 8! 


of 


a 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


BIORBIA (BORAGINACEAE) IN THE CENTRAL 
U. S. PLIOCENE 


By 
Ronald H. Segal 


VoL. XLVI Paces 495-508 June 17, 1966 No. 15 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 495-508 June 17, 1966 No. 15 


Biorbia (Boraginaceae) in the Central U. S. Pliocene 
By 


Rona.p H. SEGAL 


INTRODUCTION 


During middle and late Tertiary time extensive sedimentary deposits 
were laid down over the mid-continental High Plains region, an area ex- 
tending 400 miles eastward from the Rocky Mountain erosional front, stretch- 
ing from Texas to South Dakota. These fluviatile sediments are preserved 
as erosion remnants of the originally broad layers of sediments. The most 
extensive of these broad sheets, known as the Ogallala Formation of late 
Miocene and Pliocene (Neogene) age forms “. . . relatively thin sheet of 
deposits, in part fine-grained, but containing lenses of coarser gravels carried 
eastward from the Rocky Mountains in the channels of withering streams. 
In many places its layers are peculiarly cemented by caliche as a result of 
soil-forming processes in an arid climate, so that they are resistant to erosion 
and create the cap rock of the plain.” (King, 1959). 

The Ogallala Formation has been famous for the numerous mammals 
contained in its sediments. Fossil plants, chiefly representatives of the family 
Gramineae, are also quite well-known. It is the purpose of this work to 
present a taxonomic revision of the genus Biorbia (Boraginaceae) which is 
the most abundant fossil found in the Ogallala Formation of the Great Plains 
region. 

Though the Ogallala Formation is widely distributed, much of it is un- 
available for study due to a covering of Quaternary deposits. In many areas, 
extensive searching must be done over almost featureless terrain in order to 
locate an exposed outcrop. The Ogallala Formation in western Kansas, in 
many cases of quite limited exposure, readily yielded thousands of specimens, 
While several extensive exposures in western Oklahoma and Texas were 
non-productive, the exception being a locality in Garza Co., Texas. Several 


496 Tue University ScIENCE BULLETIN 


other fossiliferous sites have been reported from northwest Texas by Frye 
and Leonard (1957), who established the rather extensive distribution of | 
Biorbia papillosa Leonard in Texas. 

The author has collected thousands of specimens of Biorbia fruits from 
various outcrops of the Ogallala Formation over the past two years (see the 
following list of collection sites and distributional map, Fig. 1). Fossils from | 
all of these sites are permanently housed in the Paleobotanical Collection of 
the Department of Botany, University of Kansas, Lawrence. 

The following is a list of collection sites of Biorbia from outcrops of the 
Ogallala Formation. The number of nutlets of Brorbia fossilis and B. glabrum 
collected are given in parentheses for at least one representative locality of 
each county in which collections were made. 


KANSAS. Cheyenne Co.: about 2 mi. north of St. Francis along Hwy. 27, 
Segal 121, 26 Aug. 1964, Biorbia fossilis (1002), Biorbia glabrum (412). Clark 
Co.: SEY Sec. 36, T. 30 S., R. 25 W., McGregor 17214, 24 June 1962, Biorbia | 
fossilis (654), Biorbia glabrum (309); Clark Co. State Lake, Segal 112, 12 Oct. 
1963, Biorbia fossilis (580), Biorbia glabrum (166). Hodgeman Co.: SW'14 SE'4 
See330; 1223 S.eR:.2> W., Pee 17203, 23 June 1962, Biorbia fossilis (333), 
Biorbia abas (535); Loren Co.: NY SW% Sec. 24, T. 12.S.7ReS2aWe Segal 
105, 6 Sept. 1963, Biorbia fossilis (13), Biorbia glabrum (16). Meade Co.: 4 mi. 
south, 2 mi. west of Meade, Beamer, 16 Aug. 1945, Biorbia fossilis (186), Biorbia 
glabrum (37). Phillips Co.: NW% NW% Sec. 12, T. 2 S., R. 20 W., Segal 90 
4 Sept. 1963, Biorbia fossilis ( (272), Biorbia glabrum (38). Rawlins Co: NW% 
SW, Sec. 12, T. 3 S., R. 33 W., McGregor 17418, 29 July 1962, Biorbia fossilis 
(256), Biorbia glabrum (668). Scott Co.: SE4 SEY Sec. 2, T. 16 S., R. 33 Wa 
McGregor 17229, 25 June 1962, Biorbia fossilis (102), Biorbia glabrum (161). 
Sheridan Co.: Sheridan Co. State Lake, Segal 120, 24 Aug. 1964, Biorbia fossilis 
var. parvulus (1637), Biorbia glabrum var. minimum (287). Smith Co.: center 
west line of SW Sec. 32, T. 1 S., R. 14 W., Segal 89, 4 Sept. 1963, Biorbia 
fossilis (39), Biorbia glabrum (19). Wallace Co.. NWY% SW% Sec. 16, T. 14 S., 
R. 39 W., Segal 99, 6 Sept. 1963, Biorbia fossilis (113), Biorbia glabrum (1975 
center west line of Sec. 32, T. 11 S., R. 42 W., Segal 117, 11 July 1964, Brorbia 
fossilis (3), Biorbia glabrum (129). 

SOUTH DAKOTA. Bennett Co.: 5 mi. east of Vital, McGregor 18803, 11 
Aug. 1964, Biorbia fossilis (151). 

NEBRASKA. Keith Co.: Lake Ogallala State Recreation Area, Segal 122, 25 
Aug. 1964, Biorbia fossilis (76), Biorbia glabrum (22). 

TEXAS.! Garza Co.: 3.0 mi. west of the intersection of U. S. Hwys. 380 
and 84 near Post, Segal, July, 1963. Briscoe Co.: 104% mi. west of Briscoe-Hall 
Co. line and 6 mi. north of Briscoe-Floyd line. Floyd Co.: 1 mi. west of Floyd- 
Motley Co. line and 7 mi. north of U. S. Hwy. 70. Howard Co.: south wall of 
Wildhorse Creek valley, 5.2 mi. north of Coahoma. Randall Co.: 200 yards west 
of south end of Buffalo Lake dam. 


‘The Texas collections, other than by Segal, were reported by Frye and Leonard (1957) 
and are included in the distribution map (Fig. 1), but specimens from these localities are not 
in the Paleobotanical Collection of the Department of Botany, University of Kansas. 


Biorbtia (BoracinackakE) IN THE CENTRAL U. S. PLioceNr 17, 


SYSTEMATIC TREATMENT 
Biorbia Elias, 1932, Kans. Univ. Sci. Bull. 20:350; pl. 29, figs. 2a-2e. 


Type species: Lithospermum fossilium var. rugosum Berry 1928 (—Biorbia 
fossils (Berry) Cockerell), Proc. U. S. Nat. Mus. 73:1-3, pl. 1, figs. 1-10. 

Generic Diagnosis (emended): 

Only nutlets known. Nutlets keeled, asymmetrically inflated, being more 
convex on the dorsal side. Surface rugose, covered with ridges arranged in 
honeycomblike net, or smooth and without network of ridges, or densely 
papillose and without prominent reticulate surface sculpture. Dorsal keel 
turns to right or left, spoiling the apparent bilateral symmetry of the nutlets. 
Probably four nutlets, two right and two left ones, constituted a complete 
fruit. Scar of attachment round, comparatively small and elevated, consists 
of a double ring and an elaiosome in the center. Distinct canal in the ventral 
part of the scar (Fig. 8). The slight emendment of Elias’s original description 
of the genus Brorbia is necessary in order to accommodate the new species 
B. levis and the previously described species B. papillosa Leonard. 


KEY TO THE SPECIES AND VARIETIES OF BIORBIA 


Bee Mcnciae SMMOOE | ees OME sito. oe A eee B 
iNutletstusose or papillose, not smooth 20 oeeecceeceeeeeee ee Cc 
IpemINGilet 2°. 3-3..4: mamar ome: 2.cicee ce cee a esecec ceases ee oe 4. B. levis 

INutleevl77-2.3) muni lone 2.2 5. B. levis var. minimus 
C. Nutlet 1.7 mm long, sculptured by 
papilloscvexchescences 92-8 ee ee 1. B. papillosa 
INutlet rugose, covered by reticulate sculpture 2 3 D 
DraiNutlet 273-304 mma long. 2s ae 2. B. fossilis 
INutletdl. 7-222 mm: long’ 2 a. 3. B. fossilis var. parvulus 


1. Biorbia papillosa Leonard, 1958, Univ. Kans. Sci. Bull. 38:1396, pl. 1, 
figs. 1-3. 


Holotype: Museum of Natural History, University of Kansas; Catalogue No. 
9917; collected by A. B. Leonard and John C. Frye, June 30, 1955. 


Type locality: Partly cemented sand and gravel at the rim of the escarp- 
ment produced by resistant Pliocene deposits, 3.8 miles west of the intersection 
of U. S. Hwys. 380 and 84 near the center of Post, Garza Co., Texas. 

Nutlet small, total length 1.7 mm, transverse diameter 1.3 mm, vertical 
diameter 1.3 mm; ovoid in shape, swollen near base, tapering toward cuspidate 
apex, but also compressed; ventral, bladelike ridge arising from base of 
style at anterior rim of scar of attachment, extending forward in midline over 
apex, thence backward along mid-dorsal axis of nutlet for a short distance be- 
fore extending diagonally to the left, over side of the nutlet to reach scar of 
attachment; surface sculpture of closely spaced, conical, papillose excrescences, 


498 THe UNiversity ScIENCE BULLETIN 


0.2-0.3 mm high; in a few places along side of the nutlet, the papillae fused 
to form short ridges; scar of attachment circular in outline, depressed strongly 
in center, within outer rim a parallel groove, followed medially by a parallel 
row of six low, rounded papillae or bosses; a small cylindrical elevation 
(elaiosome) in center of scar of attachment, rising slightly above outer rim 
of scar; the style, situated on anterior rim of scar at point of origin of keel, 
elevated well above rim of scar, spatulate, possessing a large median vein 
and two smaller, parallel veins laterally (Fig. 5). 

I found Biorbia papillosa at the type locality, but no new details can be 
added to the above account by Leonard. B. papillosa was also found by 
Leonard (1957) in several other exposures of the Ogallala Formation in 
northwest Texas. 

Biorbia papillosa is a distinct species known only from the Pliocene of 
Texas; it has not been found associated with Brorbia fossilis. Since prolific 
quantities of B. fossilis were excavated from southwest Kansas (for instance, 
Clark Co.) an attempt was made to find fossil remains from the panhandle 
region of Oklahoma with the idea that B. fossils or the new species B. glab- 
rum might be found with B. papillosa. Such a discovery would be significant 
from both a geological and biologic view. However, no such fossiliferous 
exposures of the Ogallala Formation were discovered. 


2. Biorbia fossilis (Berry) Cockerell, 1933, Torreya 33:15. 


Lithospermum fossilium var. rugosum Berry, 1928, U. S. Nat. Mus., Proc. 73:1-3, 
pl. 1, figs. 1-10. 
Celtis microendocarpica Watt, 1928, Penn. Acad. Sci., Proc. 2:54. 
Celtis microendocarpica Brooks, 1928, Carnegie Mus. Ann. 18:299, pl. 17, fig. 1. 
Celtis microendocarpica Brooks, 1929, Carnegie Mus., Ann. 19:135-137, 1 pl. 
Biorbia rugosa (Berry) Elias, 1931, State Geol. Surv. Kansas, Bull. 18, vol. 32, no. 
7p: 150: 
eee rugosa (Berry) Elias, 1932, Univ. Kans. Sci. Bull. 20:350, pl. 29, figs. 2a-2e. 
Biorbia fossilia (Berry) Elias, 1942, Geol. Soc. Amer., Spec. Paper 41, p. 106, pl. 15, 
fig. 12. 
Isotypes: United States National Museum. 
Type localities: Phillips Co., Kansas, and Kit Carson Co., Colorado. 
Fossilized nutlets 2.2-3.4 mm long, 1.8-25 mm wide, hollow, walls ex- 
clusive of ridges ca. 0.08 mm thick, asymmetrically inflated, the dorsal side 
being more convex than the ventral. On the ventral side a keel extends from 
a cuspidate apex nearly or quite to the scar of attachment, unless lost on the 
way; surface rugose, covered with narrow and, when perfectly preserved, 
elevated ridges forming an irregular honeycomblike structure; scar of attach- 
ment round and elevated, about as high above the surface as the ridges of 
the exterior; scar consists of double ring, inside of which is prominent central 
elevation apparently corresponding to the elaiosome of living Boraginoideae- 
Anchuseae; a short and ventrally curved delicate pipe is ordinarily observed 


Biorbia (BoraGINacEAE) IN THE CENTRAL U. S. PLIocENE 499 


ventrally between the outer and inner rings of the scar, but when not pre- 
served, a corresponding thin canal leading inside the nutlet can be seen 
(Fig. 4). 

Frye, Leonard, and Swineford (1956) recognized that many seeds of 
Biorbia fossilis (Berry) Cockerell were smooth and lacked the reticulate 
sculpture of this, the rugose form, but they did not attempt to judge whether 
the smooth nutlets represented another species. Evidence will be presented 
to show that two species are probably represented, and the creation of the 
new species Brorbia levis is believed to truly delineate the biological diversity 
of the genus. 


3. Biorbia fossilis var. parvulus, var. nov. 


Biorbia fossilia var., Ellias, 1942, Geol. Soc. Amer., Spec. Paper 41, 106, pl. 15, fig. 
10a. 


Holotype: No. 120 (1a) from the type locality; Segal 120, Aug. 24, 1964. Iso- 
types: 1636 isotypic examples from the type locality; Segal 120, Aug. 24, 1964. 
Type locality: partly cemented sand and gravel deposit, Sheridan Co. State Lake, 
Sheridan Co., Kansas. 

Nutlet small, asymmetrically inflated, 1.7-2.3 mm long, 1.3-1.8 mm_ wide; 
ovoid with swollen base; tapers to cuspidate apex; ventral keel arises from anterior 
rim of round and elevated scar of attachment which extends over cuspidate apex, 
turns to either left or right on the dorsal side and connects diagonally with the 
scar of attachment. Surface of the nutlet rugose with prominent reticulate surface 
sculpture, small papillae present between the reticulate ridges; scar of attachment 
a double ring enclosing central elaiosome, inner ring dentate; remnant of style at 
anterior rim of scar (Fig. 7). 


This variety is easily distinguished from collections of Biorbia fossilis 
(Berry) Cockerell by its consistently smaller size as illustrated in Fig. 9. 

Apparently this new variety is identical to that noted by Elias (1942) as 
a “small variety of Brorbia” from a locality near Sharon Springs, Wallace Co., 
Kansas, but he did not formally describe it. 

Though we failed to find this variety at the locality specified by Elias, 
over 1600 specimens were located at the Sheridan Co. site. The difference in 
size in populations of B. fossilis and B. fossilis var. parvulus is quite striking 
and there is no difficulty in distinguishing these entities when populations 
are compared. Occasionally a small specimen or two is found in collections 
of B. fossilis, and in such instances one must assume this is an immature 
nutlet, or represents an abortive one. In living species of the Boraginaceae 
one or several nutlets frequently abort. 


4. Biorbia levis, sp. nov. 


Holotype: No. 17214(1) from the type locality; McGregor 17214, June 24, 
1962. Isotypes: 308 isotypic examples from the type locality; McGregor 17214, 
June 28, 1962. Type locality: Consolidated coarse sand, SE% Sec. 36, T. 30 S., R. 
25 W., Clark Co., Kansas. 


500 Tue Universiry SciENCE BULLETIN 


Similar to Biorbia fossilis (Berry) Cockerell except that the nutlets are 
smooth with the keel representing the only surface ornamentation (Figs. 
TW) 3 

I believe that the recognition of these fossils as a distinct species is 
taxonomically correct and more adequately represents the biologic diversity 
within this genus. 

At first I thought that the smooth and rugose nutlets were simply forms 
of the same species. It is known that various species of the Boraginaceae 
have heteromorphous fruits in which one of the four nutlets of each fruit 
may be distinguished by size or sculpture. Such a plant is Cryptantha crasst- 
sepala in which the fruit consists of one smooth and three granulate nutlets. 
This ratio is constant as determined from herbarium specimens, with the 
smooth being slightly larger than the granulate forms. 

While the smooth fossil nutlets are generally also slightly larger than the 
rugose, the ratio of smooth to rugose nutlets is extremely variable. At the 
Bennett Co., South Dakota, exposure only rugose forms were found. At the 
Wallace Co., Kansas, localities the smooth forms were dominant, and at one 
site (center west line of Sec. 12, T. 11 S., R. 42 W.) rugose nutlets were rare. 
On the basis of the opposed ratios, I feel that the establishment of a new 
species 1s necessary. 


5. Biorbia levis var. minimus, var. nov. 


Holotype: No. 120(1) from the type locality; Segal 120, Aug. 24, 1964. Isotypes: 
287 isotypic examples from the type locality; Segal 120, Aug. 24, 1964. Type 


locality: partly cemented sand and gravel deposit, Sheridan Co. State Lake, Sherr — 


dan Co., Kansas. 


Similar to Biorbia fossilis var. parvulus except that the nutlets are smooth 
with the keel representing the only ornamentation (Fig. 6). 
Again, as for B. fossilis var. parvulus, there is a sharp distinction in the 


size of the species and its variety. It is possible that if more information on — 


the plants as a whole were available, the varietal forms would be elevated 
to species. In extant plants of the family, such differences in the form of the 
fruit are diagnostic of species (Fig. 10). 


GENERAL DISCUSSION 


The Pliocene strata of the Ogallala Formation of the High Plains is known 
to contain the following plant fruits: 


Family GRAMINEAE S. breve Elias 
Tribe Stipeae S. coloradoense Elias 
Genus Stipidium S.commume Elias 
Stipidium aristatum (Berry) Elias S. elongatum Elias 


S. asymmetricum Elias S. grande Elias 


Biorbia (Boractnackak) IN THE Cenrraw U. S. PLioceNe S01 


S. intermedium Elias 
S. kansasense Elias 
. minimum Elias 
. nebraskense Elias 
.novum Elias 
. tubus Elias 
. tubus var. leave Elias 
. variegatum Elias 
S. hirsutum Elias 
S. schereri Elias 
S. ventricosum Elias 
Genus Berriochloa Elias 
Berriochloa am phoralis Elias 
. conica Elias 
. glabra (Berry) Elias 
. inflata Elias 
. maxima Elias 
. minuta Elias 
. primaeva Elias 
. tuberculata Elias 


ANNHNHnHN 


BDonannns 


P. eliasi Leonard 

Genus Setaria Beauvois 
Setaria chasea Elias 

Dribei(@?)} 

Genus Clementsiella Elias 

Clementsiella laminarum (Cocker- 
ell) Elias 
Family BORAGINACEAE 

Genus Biorbia Elias 
Biorbia fossilis (Berry) Cockerell 
B. fossilis var. parvulus Segal 
B. levis Segal 
B. levis var. minimus Segal 
B. papillosa Leonard 

Genus Cryptantha Lehman 
Cryptantha auriculata (Elias) Segal 
C. chaneyi (Elias) Segal 
C. coroniformis (Elias) Segal 

Genus Prolithospermum Elias 


Genus Paleoeriocoma Elias 
Paleoeriocoma hitchcock Elias 
Genus Nassella Desvaus 
Nassella amphora Elias 
Tribe Paniceae 
Genus Panicum Linne 
Panicum elegans Elias 


Prolithospermum johnstoni Elias 

Family COMPOSITAE 
Genus Achaenites Braun 

Achaenites kansanum Segal 

Family ULMACEAE 
Genus Celtis Linne 

Celtis willistonu (Cockerell) Berry 

With the exception of Celtis willistonii, all of these fossils represent the re- 
mains of grass and herbaceous fruits. Unlike the vast majority of late Cre- 
taccous and Tertiary floras, the plant parts preserved in the Ogallala Forma- 
tion represent the reproductive parts of the plant rather than the vegetative, 
a fact which facilitates their identification. Through all of the extensive col- 
lecting accomplished at the sites described above, only some rather sinuous 
fragments which resemble root or stem axes were found; no anatomical de- 
tails were preserved. 

The borages, represented prominently by the genus Brorbia, are without 
doubt the most abundant fossil in the Ogallala Formation. Though the large 
number of specimens collected attest to the fact that these remains are very 
prolific, sometimes rather diligent searching is required to find these fruits 
in the massive strata. The fossils usually occur in isolated clusters or pockets. 
I have found several thousand specimens of Biorbia in just a few cubic inches 
of sediment. 

The fossil fruits appear to be buried approximately at the place of their 
growth. It seems likely that the fossils accumulated into clusters when small 
ephemeral pools of water dried up; these fruits could have survived very 
little transportation and still retain their detailed sculpture in water carrying 
coarse sediments in which the fossils are preserved. 


502 Tue UNiversity ScIENCE BULLETIN 


The endocarps referred to as Celtis willistoni are the most wide-ranging 
petrifactions in the Ogallala Formation. They are of little stratigraphic sig- 
nificance in comparison to the various species of the grass and borage families 
which have been utilized in stratigraphic studies (Frye, Leonard, & Swine- 
ford, 1956). 

It is an interesting facet of preservation that only fruits are so prominently 
found in the Ogallala Formation; and these, as shown for Celtis willistoni, 
are silicifications (Swineford & Franks, 1959). This is probably due to the 
fact that the fossil fruits, as some living borages and grasses, were so con- 
stituted that their naturally hard and indurated fruits were not readily sus- 
ceptible to decay, and were gradually replaced by silica. Organs such as 
leaves, even of the borages, would not normally be expected to be preserved, 
particularly in the coarse sediments which compose the mass of the Ogallala 
Formation. Significant discoveries of leaves and other vegetative structures 
might be expected if fine-grained lake sediments could be located. 

The absence of extensive arboreous vegetation is puzzling. In support of 
at least the presence of isolated elements of a woody flora, one can mention 
the Beaver County Flora (Chaney & Elias, 1936) and the Kilgore Flora of 
late Miocene or possibly early Pliocene age (MacGinitie, 1962). A study 
of these isolated floras reveals that arboreous vegetation, as Populus, Salix, 
Ulmus, and Fraxinus was present in the High Plains coexistent with the rich 
herbaceous vegetation emphasized by Elias (1942). This is especially sig- 
nificant since Wells (1965) has suggested that the development of the broad 
grasslands and the treeless landscape currently occupying much of the High 
Plains region is relatively recent. 

Acknowledgments: I sincerely thank Dr. Ronald L. McGregor for his 
generous help and advice during this study, and for accompanying me in my 
field work. Miss Elizabeth A. Gibson prepared the line drawings. 


DTPRATURE ChbED 


Cuaney, R. W., anv M. K. Extras. 1936. Late Tertiary Floras from the High Plains. -Carnegie 
Inst. Wash. Publ. 476. 

Evias, M. K. 1932. Grasses and other plants from the Tertiary Rocks of Kansas and Colorado. 

Univ. Kans. Sci. Bull. 20:333-367. 

. 1942. Tertiary prairie grasses and other herbs from the High Plains. Geol. Soc. Amer. 

Spec. Pap., No. 41, 176 p. 

Frye, J. C., anv A. B. Leonarp. 1957, Studies of Cenozoic geology along eastern margin of 
Texas High Plains, Armstrong to Howard Counties. Bureau of Economic Geology, 
Report of Investigations No. 32. 62 p. 

Frye, J. C., A. B. LEonarp, AND A. SwINEForD. 1956. Stratigraphy of the Ogallala Formation 
(Neogene) of Northern Kansas. State Geol. Surv. Kans. Bull. 118. 92. p. 

Kino, P. B. 1959. The Evolution of North America. Princeton Univ. Press, New Jersey. 189 p. 

Lronarp, A. B. 1958. Two new fossil plants from the Pliocene of Northwestern Texas. Univ. 
Kans. Sci. Bull. 38:1393-1400. 

MacGinitiz, H. D. 1962. The Kilgore Flora, a Late Miocene Flora from Northern Nebraska. 
Univ. Calif. Pub. Geol. Sci. 35:67-158. 


Biorbia (BoraciNackak£) IN THE CENTRAL U. S. PLioceNr 


503 


SwinErorp, A., AND P. C. Franks. 1959. Opal in the Ogallala Formation in Kansas, p. 111- 
120. In Andrew H. Ireland (ed.), Silica in Sediments. Soc. Econ. Paleontol. Mineral. 
Spec. Publ. No. 7. Tulsa, Oklahoma. 


Wetts, P. V. 1965. Scarp woodlands, transported grassland soils, and concept of grassland 
climate in the Great Plains region. Science: 148:246-249. 


cmon { au ~ see = “4 
woe > = | scoot J 
¢ au nano neowooo | MICOKLET mck wana 
Cc = \ 
mere VAs stamccy womasaury | BROOKINGS 7 
Op, Tk aut *e, joooat| owaTEO 
—— . 2 umm | 
Povminaton oe OMS [eal samen | ane |woooy a 
weston on | malic 
All waanTin | PARIBAULT 
4c coon | MIMMEMAA ere 
= A \¢ a nowne 
SOUTH DAKOTA als oa Pe de 
seunnon ae | von osecoul “a ue 
Terr Cale Turns) & nossUTH “hey conn = "ey 
rao | no 
+ oe "8 uy Ie oemen| car | iro =) 
moenans wou 
comvemse (spleens 1 a 
EVA Pane soro sume He, \wmant | “ry, | eurcen 
pos mee rmwou™ %e wee oy, ‘Oy 
os pounwresten, %), 
Lr wooosuey | 1a] 
c 
more eu, - «| ‘stom 
= + 4 *¥rony wonona |cRawrono |canmot.| GREE ‘ Boon! 
2 s gf 
| o | ost usosson| f° |oumina 
. ror 
ih | Re eaenon erst, aura DALLAS 
vaccer |ameeuey amon |uanasea! 
anne \ | 
[ | ig |" pain MADSON| WAR! 
Larue — ae a = 
Ot WAP! 
we womano) wu [oom = union [evannc | Lucas |MON! 
ee pale 
Mag, \pecatun) ie 
Sas pao TAYLOR 3 
Lame % = porn 
n 2 “ug, pence 
\ J | *, 
—— G 1 @% 
| lee Pena) FIN 
bd 
sources 
f i =e \anomento, | oantss | 
==}—| Sale i a 
+ | comets wasHinaTOM vuua the q 
", 
anal ——_—_—— CHCrONNE —nawuns | OfcATUR u “On, 
somes ARAPAHOE ro, af 
ie ——|— eas 
| vovaLas | cur Lo 
J ‘ ane | wrench AW Me : 
[ a ae ee leet | 
f(.—— corarerre : 
] IAcKSOn = 
a a umcour | s 
J } umcoun 
if E Paso —_ | sonnson | PUTT 
— revue | cnevenne 
| — “14 
Fmtwont — = el 
| | nome 
———__, cxomecy | sates 
\ cusren ruce.o ie ij ] 7 sl 
i. | J 
—_ ment | fener vennon 
Ba oreo | [iesow cease | sas} 
nutmrano | { oun 
eS | | i 
a f a Barton Ara = ne 
wa S s | | | | 3 
} < | {—} anne | 
Costin OS am enc = ieee é | 
} se Tom | STEVENS meaoe | cuane | eanoce | nanera | suunen | comer | gh | fll =tmatacatia 
/ wonton Sowaro COMANCHE Ra x c | 
| o newton ttt 
F | 
erm |S | vance 
le juec COMAL O| | 
cour I 
os | 
| emo =a 
= | | 
) L r— DALAM | SHERMAN HawsroRD lcmurae Meacos af ool 
Som \ — | fo) a f° | noi") anon 
= } | — | o | 
a { manoina | | | | 
ime aS _— atic moore | | moscars | wears 
“ 
SAN miGuE, T T + custen 
T= Cron | porten cmon | ontr | ween = 
an eas | pecnnane 
= ‘Quay =F T +— + 
: | i Sis s 
cunoniure | r sam | amon | | comer |” erty 
tI! XZ [= F 4 
— Lr | = 
Tim iconey | ¢ 
L | PARCA | CASTRO | Swisnem | BmSCOE | HALL ca 
— ie | Cease =| L_ = i | rz 
Is AL | 
rs mooseverr Jamey! we | na | novo | Money | come 
| | 
COUN | Le =i | i ails 
(ee | | | 
| ees MOCKLEY LUBBOCK | CROseY ONCRENS Td 
ouves = | 
| ? 
J 
in voanum | rome | Lymn Berri erect ihe 
———— l 
| | | | 
| w ome sallow 
—_ i 
| oor 
| sooarws 
| 
“\ Lovina | smacen tcroa | 
Wosermn 
= CULERSON oe) + 
mane / me 
as im * | | Ne ance ) 
men LA urtow | P | Tom aattn | concno UN CONEY 
7 \ | 


Fic. 1. Map showing distribution of the genus Biorbia. O=both Biorbia fossilis and Biorbia 
levis present, with B. fossilis dominant; Be Biorbia papillosa only; += eee fossilis only; 
[J=both Biorbia fossilis and Biorbia levis present with B. levis dominant; © =both Biorbia 
fossilis var. parvulus and Biorbia levis var. minimus with B. fossilis var. parvulus dominant. 


Tue Universiry SciIENCE BULLETIN 


Yj 
YAY) 


ir 


Fic. 2. Drawing of Biorbia levis, new species, ca. 30X. 


Fic. 3. Drawing of Biorbia levis, new species, ca. 30X. 


Fic. 


ate 


Drawing of Biorbia fossilis (Berry) Cockerell, ca. 30X. 


Fic. 5. Drawing of Biorbia papillosa Leonard, ca. 38X. 
gs 


Biorbia (BoraGINACcEAE) IN THE CENTRAL U. S. PLIOCENE 


Fic. 6. Comparison of ‘bia levis (two large specimens), new species, and Biorbia levis var. 


Ne 


minimus (two small specimens), new variety, ca. 13X. 


Fic. 7. Comparison of Biorbia fossilis (Berry) Cockerell (two large specimens), and Biorbia 
fossilis var. parvulus (two small specimens), new variety, ca. 13X. 


THe University ScrENCE BULLETIN 


Fic. 8. Details of basal scar of attachme 


507 


Biorbia (BoractINacEaE) IN THE CENTRAL U. S. PLIOCENE 


34 
° 
323) fo) ° 
° 
00 
(eo) 
»~ ° re) 
° 
° ° 
29 
ro) ers) 
O° 
273 
Cw 
E ier 
E 4 6 
© 257 ° 
D %, 
(Ss 
o 


240) 


207 


19 


2.15 Za) 249 265 


199 


width (mm) 


Fic. 9. Graph showing relative sizes of Biorbia fossilis and Biorbia fossilis var. parvulus. Data 
are based on the following collection sites: SE% Sec. 36, T. 30 S., R. 25 W., Clark Co., Kansas, 
for Biorbia fossilis; McGregor 17214, June 24, 1962. Sheridan Co., State Lake for Bzorbia 
fossilis var. parvulus; Segal 120, Aug. 24, 1964. O=B. fossilis; A= B. fossilis var. parvulus. 


1.32 149 165 1B2 


UW 
S 
e} 


Tue Universiry ScreNcE BULLETIN 


34 
° 
° o 6 ° 
° 
° ° OG 
323 ° 
° 
o 
oO 
30 euate ° 
0° 
oa 
29 
27 Ones 


ny 
u 


I@¢ngthgmnyp 
° 


240 


2.07| 


190 


1.66 


32 149 165 182 199 215 232 249 2.65 


Width (mm) 


Fic. 10. Graph showing relative sizes of Biorbia levis and Biorbia levis var. minimus. Data 

are based on-the following collection sites: SE% Sec. 36, T. 30 S., R. 25 W., Clark Co., Kan- 

sas, for Biorbia levis; McGregor 17214, June 24, 1962. Sheridan Co. State Lake for B. levis 
var. minimus. O=B. levis; AA=B. levis var. minimus. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


SYSTEMATICS OF THE GENUS 
AUGOCHLORELLA (Hymenoptera, Halictidae) 
NORTH OF MEXICO 


By 
Ellen Ordway 


Vor. XLVI Paces 509-624 June 17, 1966 No. 16 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Wot. XLVI Paces 509-624 JUNE 17, 1966 No. 16 


Systematics of the Genus Augochlorella (Hymenoptera, 
Halictidae) North of Mexico’ 
By 


ELLEN Orpway? 


ABSTRACT 


In a revision of the bee genus Augochlorella, descriptions and keys are given 
for seven species occurring north of Mexico and one species and a subspecies from 
Mexico. One species from Texas, 4. bracteata and the subspecies A. neglectula 
maritima from Mexico are new, and the male of A. edentata is described for the 
first time; 4. neglectula is raised from synonymy, and A. aurata of recent authors 
has been divided so that most of the specimens from north of the Gulf Coast states 
are placed in A. persimilis. Other species treated are A. pomoniella, A. gratiosa 
and A. striata. Regional as well as individual variations are treated in detail since 
there is considerable morphological intergradation among females of some species. 


INTRODUCTION 


Although bees of the genus Augochlorella (Halictidae) are common over 
much of the United States, their interspecific relationships and biologies are 
little understood. This paper is a result of a study of interspecific relation- 
ships of the two species of Augochlorella occurring in the vicinity of Law- 
rence, Kansas. Females of the two species, persimilis and striata, could not 
be satisfactorially distinguished and preliminary biological observations indi- 
cated little or no difference between them. In order to understand the nature 
of the variations and intergradations occurring between these two, all species 
occurring within the United States were examined. They are redefined and 
illustrated, keys for their identification are provided, and variations within 
species analyzed. 


"Contribution number 1280 from the Department of Entomology, The University of Kansas, 
Lawrence, 


2 5 orc is 5 5 5 “y a : ~ 
Present address, Division of Science & Math., University of Minnesota, Morris, Minnesota. 


510 ‘Tue Universiry SCIENCE BULLETIN 


For those investigators who are interested chiefly in the species occurring 
in a certain area, primary use should be made of the Regional Account of the 
Species. Other sections to be consulted include the discussion of the genus, 
the section on terminology, the keys to the species and under the Species 
Accounts, parts on comparisons with other species and seasonal activity. 

For those interested primarily in studies of variation, special attention 
should be given to sections on Specific and Other Characters, Terminology, 
Species Groups and under the Species Accounts, parts on comparisons and 
variation. The regional account should also be of value. 

Augochlorella is a genus of bright metallic blue or green bees. As in most 
halictines, the species are morphologically variable and are in many cases 
difficult to distinguish. Therefore, special attention is given to the nature and 
extent of variations within and among species. Since other species, not being 
considered in this work, occur in Mexico, Central and South America, a com- 
plete analysis cannot be attempted until all species have been studied, prefer- 
ably both biologically and morphologically. Eight species are treated here; 
seven occur north of Mexico, the other is from Mexico and Central America. 

The more than 14,000 specimens examined in this study were obtained 
from a total of 50 private and institutional collections (Table 1). In addition 
to the pinned material, 414 specimens were examined from 134 nests exca- 
vated near Lawrence, Kansas, and the variation compared with that of popu- 
lations throughout the country. 


Taste I. Collections studied. 


Code No. Collection 
] University of Arizona, Tucson. 
2 University of Arkansas, Fayetteville. 
3 California Academy of Sciences, San Francisco. 
ay University of California, Berkeley. 
5 Los Angeles County Museum, Los Angeles, California. 
6 P. H. Timberlake Collection, Riverside, California. 
7 University of Colorado Museum, Boulder, 
8 Colorado State University, Ft. Collins. 
9 United States National Museum, Washington, D.C. 
10 State Department of Agriculture, Gainesville, Florida. 
11 University of Georgia, Athens. 
12 Fattig Collection, University of Georgia, Athens. 
13 Chicago Natural History Museum, Chicago, Illinois. 
14 Illinois Natural History Survey, Urbana. 
15 Robertson Collection, Illinois Natural History Survey, Urbana. 
16 Purdue University, Lafayette, Indiana. 
17 Indiana University, Bloomington. 
18 Iowa State University, Ames. 
19 Kansas State University, Manhattan. 


20 University of Kansas, Lawrence. 


SYSTEMATICS OF THE GENUS Augochlorella Nort oF Mexico — 511 


Code No. Collection 
72 Carl W. Rettenmeyer Collection, Kansas State University, Manhattan. 
22 Museum of Comparative Zoology, Cambridge, Massachusetts. 
13} University of Massachusetts, Amherst. 
24 R. R. Dreisbach Collection, Michigan State University, East Lansing. 
25 Michigan State University, East Lansing. 
26 University of Minnesota, St. Paul. 
27 University of Missouri, Columbia. 
28 University of Nebraska, Lincoln. 
PKS) David W. Ribble Collection, University of Nebraska, Lincoln. 
30 Rutgers, The State University, New Brunswick, New Jersey. 
33 Cornell University, Ithaca, New York. 
32 American Museum of Natural History, New York City, New York. 
33 North Carolina State, Raleigh. 
34 North Dakota Agricultural College, Fargo. 
35 University of North Dakota, Grand Forks. 
36 Ohio State University, Columbus. 
Sy Oklahoma State University, Sullwater. 
38 Academy of Natural Sciences of Philadelphia, Philadelphia, Pennsylvania. 
39 Carnegie Museum, Pittsburgh, Pennsylvania. 
40 South Dakota State College, Brookings. 
4] University of Tennessee, Knoxville. 
42 Alvin F. Shinn Collection, Oak Ridge, Tennessee. 
43 Utah State University, Logan. 
44 George E. Bohart Collection, Utah State University, Logan. 
a5) Brigham Young University, Provo, Utah. 
46 Milwaukee Public Museum, Milwaukee, Wisconsin. 
47 University of Wisconsin, Madison. 
48 British Museum (Natural History), London. 
49 Canadian National Collection, Ottawa. 
50 Naturhistorisches Museum, Wien, Austria. 


The genus Augochlorella was proposed (with the type species Augochlora 
gratiosa Smith) by Sandhouse (1937) in a revision of an assemblage of forms 
that had previously been included in the genus Auwgochlora. Prior to the re- 
vision by Sandhouse, all green halictines except Agapostemon were com- 
monly referred to Augochlora, including forms now placed in Augochlora, 
Augochloropsis and Augochlorella. |For synonymies of these genera, see 
Sandhouse (op. cit.).| I am not certain of the status of Pereirapis Moure, 
which was synonymized with Augochlorella by Michener (1954). Since the 
species included in Pereirapis, except for edentata (see Species Groups), are 
distinctly different from those of the north and are separated from each other 
by entirely different characters, I feel that they form a legitimate group 
which should perhaps be recognized at the subgeneric level. 

Moure (personal communication, 1961) has included the following 
species in the Pereirapis group: A. bidentata Michener, A. cerasis (Vachal), 
A. chrysaspis (Vachal), A. edentata Michener, A. phoenicis (Vachal), A. 


a2 Tue Universiry SciENcE BULLETIN 


semiauratus (Spinola) |=A. tatania (Smith), A. hypixts (Vachal) |, A. semi- 
nigra (Cockerell) (=P. rhysophila Moure) and A. simotes (Vachal). He 
has also indicated that the following Central and South American species are 
to be placed in Augochlorella proper: A. cladopyga (Cockerell), A. comis 
(Vachal), A. ephyra (Schrottky) [=A. traumatias (Vachal), A. ictis 
(Vachal) and Oxystoglossidia uraniella Moure|, A. iopoecila Moure, A. 
iphigema (Holmberg), A. michaelis (Vachal) (=Oxystoglossidia uranioides 
Moure), A. tredecim (Vachal), A. urania (Smith). 


SoME D1acNostic CHARACTERS 


The genus can be distinguished from the other genera of green halictines 
occurring north of Mexico by the key below. A detailed description was 
given by Sandhouse (1937) and modified by Michener (1954). The follow- 
ing account, therefore, is brief and presents only the most diagnostic and 
readily visible characters by which Augochlorella can be distinguished from 
the other nearctic halictines of similar appearance. 

HEAD: The epistomal suture (Fig. 2, es.) in Augochlorella and Augo- 
chloropsts extends ventrolaterally on each side of the clypeus to the level of 
the mandibular base where it angles abruptly laterally. In Awgochlora and 
most Agapostemon it loops below the level of the mandibular base into the 
clypeal area, thus forming a lobate extension of the paraocular area (Fig. 1). 

The clypeus is flattened in Augochlorella but with the apical half beveled 
or at a slight angle to the upper half. Females have the beveled portion 
brown or black with large, coarse, well separated punctures. The upper half 
is metallic green or blue with smaller punctures of varying sizes. In Augo- 
chlora the clypeus is flat, or the beveling is not noticeable since the entire 
surface is green except for a narrowly black apical edge. In Augochloropsis 
the clypeus is protuberant and rounded, sometimes slightly darkened medial- 
ly at the apex. The clypeus of Agapostemon is similar to that of Augochlor- 
ella except for the ventrolateral or lobate extensions of the epistomal suture. 

The vertex in Augochlorella is short, about equal to the oculo-ocellar 
distance, and abruptly angled or declivous between the posterior ocelli and 
the occipital carina (Fig. 5). In Augochlora this area is broadly rounded and 
longer than the oculo-ocellar distance (Fig. 3). In Augochloropsis the vertex 
is long and sharply angled so that the occipital carina is usually hidden from 


Fics. 1-2. Front view of head. Fig. 1, Augochlora; Fig. 2, Augochlorella; es: epistomal 
suture. 

Fics. 3-5. Side view of head. Fig. 3, Augochlora; Fig. 4, Augochloropsis; Fig. 5, Augo- 
chlorella. 

Fics. 6-9. Forewings. Fig. 6, Augochlorella; Fig. 7, Augochlora; Fig. 8, Augochloropsis; 
Fig. 9, Agapostemon. 

Fics. 10-11. Body measurements. Fig. 10, a, width of head; b, length of clypeus; c, width 
of clypeus; Fig. 11, a, length of metanotum; b, length of propodeum. 


SYSTEMATICS OF THE Genus Augochlorella Nortu or Mexico 


514 Tue UnNiversiry ScIENCE BULLETIN 


above except when the head is strongly depressed (Fig. +). In Agapostemon 
the vertex is similar to that of Azgochlorella. 

THORAX: The posterior vertical surface of the propodeum of Aga- 
postemon may be distinguished from that of all the other genera by the key 
character given below. 

The marginal cell of the forewing in Augochlorella is pointed at the 
wing margin (Fig. 6), whereas in the other three genera it 1s narrowly trun- 
cate at the tip or bent away from the margin of the wing. In Augochlora 
and some Agupostemon and Augochloropsis the lower vein of the marginal 
cell (Rs, Fig. 7) extends beyond the tip of the cell. 


Key to GENERA OF GREEN Ha ticTINEs OF THE UNITED STATES 


1. Body surface strongly sclerotized with large, deep punctures, similar 


to that of chrysidids; female without scopa ........--..----+------1-------- Temnosoma 
— Body surface not as above; females with scopa on hind legs ...............-- 2 | 
2. Posterior vertical surface of propodeum enclosed by a strong carina; first 
recurrent vein (Ist m-cu) basad of Ist r-m (Fig. 8) -..---------------- A gapostemon 
— Posterior vertical face of propodeum not enclosed by a carina; first 
recurrent vein interstitial with Ist r-m (Fig. 6) --:.-.—-.-.2---22e 3 
3. Epistomal suture forming deep loop into clypeal area, extending below 
levelrof base of mamdable: (Big 1) ooo cess ress cee cree eee Augochlora 
— Epistomal suture forming right angle, bending laterally at level of 
mandibular base (Fig. 2) isc. nose seer een 4 


4. Marginal cell pointed on margin of wing (Fig. 6); inner hind tibial 

spur of female serrate; metasomal terga without fringe of long apical 

fain ie ren a Bn Meee Seg re eee Augochlorella 
— Marginal cell either truncate and appendiculate apically or pointed 

below margin of wing (Fig. 9); inner hind tibial spur of female 

PCC EMTS cesar nace nea Gree 5 
5. Tegula with inner posterior angle lobate; pronotum with dorsal edge 

carinate or lamellate; apical margins of metasomal terga green, usually 

with fringe of sumplejapicale lacs sess. ere te eos Augochloropsis 
— Tegula oblong, not lobate posteriorly; pronotum with dorsal edge 

rounded or angulate; apical margins of metasomal terga black, with- 

OUit frimee Ot rapical Manns je eee ce cee esac nese eee rere eee Pseudaugochloropsis 


TERMINOLOGY 


Terminology for all morphological structures follows that of Michener 
(1944) unless otherwise stated. In the interests of simplicity and conservation 
of space, certain terms have been used in the descriptions and discussions of 
variation which may seem vague; they have been used, however, in precise 
ways as explained below: 


SYSTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 515 


The body sarface: the integument of a sclerite in general or the areas be- 
tween punctures, striations or other specifically mentioned conformations 
when present; its degree of shininess and roughening are often significant 
characters. 

Shininess: polished, absolutely smooth at magnifications used and highly 
reflective; shiny, reflective but not brilliant and not necessarily smooth; dull, 
not shining although sometimes appearing smooth at magnifications used. 

Roughness: smooth, without obvious irregularities or unevenness; granu- 
lar, with very small, round pits, usually regularly distributed, giving the illu- 
sion of raised bumps as on fine sand paper; rough or roughened, with a 
slightly irregular or uneven surface but without any obvious pattern to the 
unevenness; ragose, deeply roughened or wrinkled, often forming a regular 
pattern with the elevations and ridges occupying an area as large as or larger 
than the depressions; areolate, deeply rugose, forming a network of ridges 
with depressions occupying greater areas than the elevations or ridges. 

The degree of roughening or punctation is described by the following ad- 
verbs in sequence starting with the least amount of roughening: minutely, 
weakly, finely and coarsely. An adjective when used by itself (i.e., “rough”) 
describes a condition between fine and coarse. 

Punctures distinct: This expression is used if each puncture is separated 
from every other puncture, with the outlines easily discernible. Punctures 
may be very close together but nevertheless easily recognizable as individual 
punctures. 

Punctures indistinct: This expression is used if the punctures are not 
easily recognizable as distinct entities and may be either shallow or minute 
and vaguely defined or may merge together. 

Disc refers to the dorsal area of the propodeum only. The edge of the disc 
refers to the angle formed between the dorsal and posterior surfaces of the 
propodeum. It grades from sharp or weakly carinate to gradually rounded. 
The shape or outline of the disc is the pattern formed by the edge of the disc 
when viewed from above. 

Form refers only to particular variants in the species striata. 


METHODS 


Equipment. All specimens were examined under 40% magnification of a 
dissecting microscope. A 100 watt incandescent bulb was used when record- 
ing body sculpturing. A microscope light with blue filter was used to view 
the color of body and pubescence. 

All photographs were taken at the same magnification and those appear- 
ing in any one plate are reproduced at the same magnification. 

_ Measurements were made by means of an ocular micrometer at 40 


516 Tue UNiversity SciENCE BULLETIN 


magnification. In all cases, the specimen was aligned so that both extremities 
of the structure being measured were in focus and a maximum measurement 
obtained of the distance between them. 

Synonymies. In the synonymy of each species all known references are 
given, with annotations indicating the content as follows: descriptions or 
descriptive comparisons with other species (descr.), distributional records 
(distr.), flower records (fl.), keys (key), annotated or unannotated lists of 
species such as catalogues or regional compendia (list), and taxonomic treat- 
ments usually including keys, distributions, descriptions, etc. (tax.). Sec- 
ondary references to synonymies in catalogues or taxonomic treatments are 
not included. Certain authors (Rau, 1922) have confused species of Awgo- 
chlorella with Augochlora pura. References published under specific names 
of Augochlorella (e.g., striata) but known to refer to Augochlora pura are of 
course omitted from the synonymies. 

Types. The type (holotype or lectotype) has been seen for each species 
unless otherwise noted. 

Descriptions. With the exception of persimilis and striata, all specimens 
of a species were examined for all characters. During the course of the study 
all but those characters finally used in the descriptions were eliminated, usual- 
ly because they were found to have little or no diagnostic value. The genitalia 
of about 509% of the males of each species, from localities throughout the 
range, were examined. 

Because of the large number of specimens of striata and persimilis, the de- 
scriptions and detailed examinations of these species were made from a sam- 
ple of about 50 specimens, mostly from one state. Samples from all other 
states were then compared with the description, and corrections and varia- 
tions noted. Every specimen in both species, both male and female, was 
critically examined with regard to all propodeal characters and metasomal 
punctures and for the basitarsal hair of males. Other characters received at- 
tention commensurate with their diagnostic value. For variable characters in 
persimilis, enough specimens were examined to determine the extent of the 
variation and the distribution of the variants. For striata, variation was 
studied by recognizing several lettered “forms.” The details are indicated in 
the discussion of that species. 

Variations. An attempt has been made to indicate as closely as possible 
the range of variation of each character in each species. Whenever practical, 
particular specimens have been cited to illustrate certain variational features, 
or percentages are given if large numbers of individuals are involved. Par- 
ticular specimens are identified by their label data followed by the code num- 
bers, in parentheses, of the collections in which they are located (see Table 1). 

In the section on regional variation all species occurring within each geo- 
graphical area are compared. 


SysTEMATICS OF THE GENUs Augochlorella Norru or Mexico — 517 


Although much attention has been paid to the problem of variation in 
this group, this is not primarily a study of variation and at least in one species, 
striata, a considerable amount of work, both analytical and statistical, still 
needs to be done before a complete understanding can be achieved. 

Distributions. Although a complete list of all label data has been made, 
and may be obtained from the author, localities are listed in this paper only 
by counties and are indicated on maps. Specific localities are given where the 
counties are large with widely varying habitats. Only those specimens that 
have been examined are recorded (unless otherwise specified). Records from 
the literature are omitted due to the unreliability of specific determinations. 

For economy of space, listing of localities is omitted in whole or in part 
for certain common eastern species. Localities are shown approximately by 
the maps, and counties are listed by Ordway (1965). 

Seasonal Activity. The data on this topic were obtained from pinned ma- 
terial unless otherwise noted. Dates given for seasonal activity are those on 
which collections were made and do not necessarily represent the entire 
season of the bees’ activity. 

Flower Records. Flower records were taken from the literature and from 
labels on pinned specimens but not from laboratory observations since host 
plant preferences in the laboratory are shown to have little correlation with 
those in the field under natural conditions (see Ordway, in press). The 
records listed under each species show the flowers on which bees were found 
without regard to the sex of the bee or to whether the bee was collecting 
pollen or nectar, since this information is usually not available. All flower 
names have been checked and the appropriate synonymies made according 
to the following references: Fernald (1950), Gould and Thomas (1962), 
Kearney and Peebles (1951), Munz (1959), Smith (1933) and Index Kewen- 
sis (1895-1955). 


SPECIFIC CHARACTERS 


Nearly all of the specific characters vary in a continuous manner, so that 
it is difficult to categorize the differences within and among species. Many of 
the characters are self explanatory but others require explanation and are dis- 
cussed below. Some of these characters are of no value in distinguishing the 
North American species from one another but are of value in distinguishing 
certain neotropical species. 

Body color. Coloration throughout the range of the species is extremely 
variable. The usual color is a bright green, but specimens may range from 
bronzy or yellowish green to a deep violet-blue. Blue specimens are found 
only in Florida. The part of the body with the greatest color variation is the 
metasoma where coloration is often incomplete, allowing various amounts of 


518 Tue UNiversiry SciENcCE BULLETIN 


brown to show through, so that in some specimens this area looks brown 
with metallic reflections. Of the species considered in this paper, only the 
males of edentata have the metasoma consistently and naturally brown. It 
was found that in dried specimens, normally testaceous color (pale yellow- 
white) structures may in some specimens turn orange due to ageing or other 
factors. This condition has not been found in live or freshly killed bees. 

The metallic color is structural and can be altered artificially by various 
environmental conditions or chemicals. Limited tests have shown that dried 
bright green bees turn blue-green to blue within five minutes in ethyl acetate 
(liquid) and that on drying, the bees stay blue-green. If they are then put 
into water the bright green color slowly returns. Bees preserved in alcohol or 
in Dietrich’s or Carnoy’s solutions remain green, but dry, bright green or 
yellow-green bees turn coppery-green to reddish in an atmosphere containing 
phenol (as in a relaxing chamber). Depending on the concentration and 
exposure time, the altered color may remain after the specimen has redried. 

Chemicals may affect the coloration of insects found in collections. The 
color of live bees may be determined or influenced during the pupal stage by 
atmospheric or soil moisture. Specimens of Awgochlorella found in Florida 
are uniformly darker green than they are to the north but it is not known 
whether the bright blue individuals, frequently found in collections from 
Florida, actually are this color in nature. Many of the blue specimens were 
collected by Graenicher. It is possible that he and some other collectors in 
Florida used acetate killing jars, thus changing the color of the bees. Some 
persimilis were reared in the laboratory in wet soil. Emerging females were 
usually a dark green but one blue individual was produced. The males pro- 
duced were mostly the yellow-green color typical of both sexes of this species 
in Kansas. The blue reflections found on the frontal area of the head in 
some species appear to be a natural phenomenon with variation only in the 
intensity. 

Size. Total body length, although inexact, gives a rough indication of the 
overall length and is given for comparison with species described by earlier 
workers. When a wide variation of lengths occurs the extremes are given. 
All measurements are made on individuals with bodies in extended positions. 
Width of the head is a more standard and reproducable measurement and 
has been shown (Michener and Lange, 1958), at least for Lasioglossum 
rhytidophorum, to be highly correlated with other body measurements such 
as wing length and thoracic size. Both in A. striata and persimilis a high 
correlation (r values significant at < .01) was also found between the width 
of the head and length of the wing with r values of .65 and .55 in striata and 
persimilis respectively (see Ordway, in press, a, for full discussion of varia- 
tion). Head width was measured across the widest part of the face, at about 
the level of the ocular emargination (Fig. 10a). When available, at least 50 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 519 


males and 100 females of each species were measured from throughout its 
geographic range, representing as wide a span of collecting dates as possible. 
An effort was made to include all size extremes. In striata, all “forms” were 
lumped together. Wing length was not used, as it is a difficult and unreliable 
measurement on dried material. The width-to-length relationship of the 
head is given as an indication of the shape of the face. Bees in which the 
length of the head is greater than the width have a face that appears long 
(Fig. 52); when the length is equal to the width (+ 0.4 mm), it appears 
round (Fig. 53), and when the width is greater than the length, the face 
appears wide (Fig. 54). In some species the shape of the face is relatively 
constant; in others it is variable. The length of the face was measured from 
the apex of the clypeus to the vertex, with the head positioned so as to give a 
maximum measurement and with the distance being determined when both 
extremities were in focus. It is a poor measurement to use by itself, as it is 
frequently inexact or cannot be reproduced exactly on the same specimen. 
However, as a comparison with the width (greater than, equal to or less 
than), it is quite usable, and the same proportions may be obtained after 
multiple readings. 

Clypeus. The clypeus in all species studied is largely green or similar in 
color to the rest of the head. In males, the apex is narrowly testaceous medial- 
ly, slightly more extensively so at the lateral corners. In females the apex of 
the clypeus is nonmetallic brown or black, the extent of the nonmetallic area 
being variable within and among species. The length (Fig. 10b) is usually 
equal to or slightly greater than the width (Fig. 10c) in both sexes. The size 
and spacing of the punctures on the clypeus usually varies within species al- 
though some species have a more uniform pattern than others. 

Supraclypeal area. This area is slightly convex and looks protuberant in 
all species except edentata. This protuberance is due more to depression of 
the epistomal suture and antennal sockets than to an elevation of the area 
itself. In edentata and some other neotropical species, there are scarcely any 
depressed areas so that the entire face looks broadly and evenly convex. 

Tegula. The length to width ratio of the tegula differs among species. In 
all species unless otherwise noted, this structure is oval, shiny and smooth, 
transparent to pale yellowish anteriorly, becoming darker posteriorly and 
dark brown, usually with metallic reflections, along the proximal edge. The 
anterior part is usually minutely pilose. 

Propodeum. The characters of the propodeum are highly variable. All 
variation is more or less continuous so that there are rarely distinct gaps sep- 
rating species, and yet these characters are still the most diagnostic for dis- 
tinguishing species, especially in the females. The length of the disc (the 
dorsal area) is measured along the median line from the anterior carina 
Separating the metanotum and propodeum, to the posterior margin of the 


520 Tue UNiversiry ScrENCE BULLETIN 


disc. When the posterior margin forms a sharp V medially, the posterior 
edge is assumed to be evenly rounded and the tip of the V is omitted from 
the measurement (Fig. 11b). 

The outline or shape of the disc is classified into four categories, although 
all gradations occur among these. The “bracket” shape (“\) is perhaps the 
most distinctive (Figs. 12-14, 57, 58) and is characteristic of gratiosa, aurata 
and some striata. The posterior edge comes to a point or V medially, slopes 
laterally to the posterolateral corners of the propodeum, then angles abruptly 
anterolaterally. The V-shape is similar but without the distinctive angula- 
tion at the posterolateral corners. The V may be deep, so that the length of 
the disc is considerably longer medially than at any other point, as in some 
striata B (Figs. 15, 63), or it may be shallow or “obtuse” with the length of 
the disc essentially equal throughout (Fig. 59). In addition, the posterior 
edge may be sharply pointed medially (Fig. 12) or blunt and rounded (Fig. 


14), or may extend posteriorly in the plane of the rest of the disc or be de | 


pressed onto the posterior vertical propodeal surface (Fig. 63). The U-shape 
(Figs. 16, 62) lacks the medial V and is evenly rounded posteriorly. The 
length of the disc is greater medially than laterally. The semicircular shape 
is similar to the U-shape but is shorter in relation to its width, with shorter, 
more rounded posterolateral corners (Figs. 18, 64). Gradations among all 
shapes occur (Figs. 17, 65). Relationships of the different shapes are dia- 
grammed in Figure 19. 

The sharpness of the edge and the flatness of the disc is indicated by the 
profile type. In type 1 (Figs. 20, 21), the edge of the disc is weakly or sharply 
angulate to carinate and slightly elevated, giving the disc a concave appear- 
ance in profile. In type 2 (Fig. 22), the edge is abruptly rounded, not ele- 
vated, and distinct when viewed from the top. The surface of the disc is flat. 
In type 3 (Fig. 23), the edge is rounded but slightly prominent or thickened, 
so that although the edge is definite when viewed from above it is not as 
distinct as in type 2. Type 4 has a gradually and smoothly rounded edge 
(Fig. 24) without any demarcation between the dorsal and vertical surfaces. 
Type 5 (Fig. 25) includes only pomoniella and combines characteristics of 
types 3 and 4. The edge of the disc, although gradually and smoothly 
rounded, is at the same time prominent and somewhat elevated. The dorsal 
surface of the disc is therefore slightly concave, a feature that is evident both 
from the top and from the side. 


Fics. 12-18. Diagrams showing shapes of propodeal disc. Figs. 12-14, bracket-shaped; Fig. 
12, A. gratiosa; Fig. 13, striata a; Fig. 14, aurata. Fig. 15, V-shaped, striata p. Fig. 16, U-shaped, 
pomoniella; Fig. 17, U-shaped, edentata. Fig. 18, semicircular, bracteata. 

Fic. 19. Relationship of disc shapes: v, V-shaped; s, semicircular; u, U-shaped. 

Fics. 20-25. Diagrams showing profiles of thorax. Fig. 20, A. striata a, type 1; Fig. 21, 
gratiosa, type 1; Fig. 22, aurata, persimilis, bracteata, type 2; Fig. 23, striata 8, pv, neglectula, 
type 3; Fig. 24, striata c, edentata, type 4; Fig. 25, pomoniella, type 5. 


SYsTEMATICS OF THE GeNUs Augochlorella Nortu or Mexico 


521 


1 
i) 
Nm 


Tue University ScrENcE BULLETIN 


When the striae do not reach the posterior edge of the disc or the edge is 
unclear as in profile type 4, the surface of the disc beyond the striae may be 
variously marked with granulations (Fig. 62), fine ridges (Fig. 66) or minute 
transverse lines or reticulations (Figs. 67, 69, 70); each type of surface is 
characteristic of certain species. 

The posterior vertical surface of the propodeum is also variously sculp- 
tured in the different species, grading from smooth and shiny with minute 
punctures (Fig. 73) to smoothly granular (Fig. 74) to roughened and rugose 
(Fig. 75). There is some variation within species, particularly among the 
forms of striata, but in general each species is characterized by a certain type 
of sculpturing. 

Metasoma. Because the first abdominal segment is incorporated into the 
thorax, segments of the apparent abdomen are numbered from one on and 
are called metasomal segments. Numbered terga and sterna always refer to 
these metasomal segments. 

The size and density of punctures on the first and second tergum are 
variable within and among species in both males and females, sometimes 
varying geographically. The punctures may be distinct and regular or they 
may be indistinct, irregular or variable in size and spacing. The third and 
following terga in all species studied are densely and minutely punctured, 
with the punctures inconspicuous and blending together giving the surfaces 
of the terga a minutely reticulate appearance. The apices of the terga are 
narrowly margined with brown in all species. The sterna of both males and 
females are brown, with long hair on at least the apical halves. In females 
this hair is longer than in males and is frequently used by the bee as part of 
the scopa. In addition to color variations noted in a previous paragraph, the 
first and second terga may show dull, discolored areas on the dorsal median 
surface. This is due to a waxlike secretion from this area (Fig. 56) that leaves 
the otherwise shiny surface dull. The nature and function of this secretion 1s 
not known, but it occurs only in females of all species of Augochlora and 
Augochlorella examined. It is not found on fresh, young (entirely unworn) 
specimens but the discoloration, if not the waxlike substance itself, may be 
found on most older specimens. 

Pubescence. Over most of the dorsal part of the body there are two types 
of pubescence, the long, simple or branched hairs usually referred to as the 
pubescence or hairs and exceedingly short, fine, highly plumose pubescence 


Fics. 26-28. Male genital capsule of Amgochlorella striata. Fig. 26, dorsal view; Fig. 27, 
ventral view (hi: inner lobe of gonostylus; lz: outer lobe of gonostylus; gn: gonostylus); Fig. 28, 
side view of genital capsule. 

Fics. 29-30. Male genital capsule of Augochlorella seminigra. Fig. 29, dorsal view; Fig. 30, 
ventral view. 

Fics. 31-36. Lobes of the gonostyli. Fig. 31, 4. bracteata; Fig. 32, striata; Fig. 33, persimilts; 
Fig. 34, gratiosa; Fig. 35, pomoniella; Fig. 36, edentata. 


SysTEMATICS OF THE GENUS Augochlorella Nortu of Mexico — 523 


524 Tue University SclIENcCE BULLETIN 


not visible except when the longer hairs have been worn away and the sur- 
face is seen in profile. This latter pubescence is white in Awgochlorella and 
may be rather dense in unworn specimens, especially on the head, posterior 
surface of the propodeum and metasoma. The longer hairs are white to 
golden-white, depending on the region of the body and on the individual 
specimen, 

Male genitalia. The male genital capsule (Figs. 26-28) is basically similar 
for all Nearctic species but differs significantly in structure from that of A. 
seminigra of the Pereirapis group (Figs. 29-30). The only specifically vari- 
able structure on the capsule is the inner lobe of the gonostylus (1, Fig. 27). 
This lobe is essentially similar in all the eastern species (aurata, striata, 
gratiosa, persimilis and bracteata) (type 1, Figs. 31-34) but is distinctive for 
each of the remaining western and Mexican species (types 2-4, Figs. 35-39). 
In type 1, considerable variation occurs within species in the length of the 
attentuated, finger-like projection, the number of long setae, and the round- 
ness and slope of the apical portion of the lobe. There are average differences 
in shape among the species, as seen in Figs. 31-36, but these differences are 
not constant and cannot be used as diagnostic characters. The number of 
setae is not constant but averages about 10 on the rounded portion of the lobe 
and 2 to 3 on the attenuation. The outer lobe (lz, Fig. 27) is similar for all 
type 1 species and varies little except for the number of long, unbranched 
setae. The genitalia of the remaining (western and Mexican) species are 
described under those species. 

Hidden sterna. Only three structural types are recognized among the 
eight species studied (Figs. 40-42), with all eastern species and neglectula 
belonging to type 1. Slight individual variations occur in type 1 (Fig. 40) 
that involve the degree of sclerotization, the shape of the apex, and the num- 
ber and position of setae on the seventh sternum. The central thickening 
appears to be absent in neglectula. In pomoniella (Fig. 41), the sterna differs 
by the presence of a variably shaped knoblike median projection on the 
eighth sternum, by the minute setose projections on the distal arms, by the 
apparent lack of central thickening and the consistent lack of apical setae on 
the seventh sternum. In edentata (Fig. 42) the structure of the eighth ster- 
num is similar to that of the eastern species, but the seventh sternum is 
elongate and truncated apically. 

Eighth tergum. This tergum is hidden, mostly membranous, internal, 
usually closely adherent to the seventh tergum and attached by weakly sclero- 
tized arms to the eighth sternum. There is a row of spiculate, finger-like pro- 


Fics. 37-39. Lobes of the gonostylus of A. neglectula. Figs. 37-38, n. neglectula; Fig. 39, 
n. maritima. 

Fics. 40-42. Hidden metasomal sterna, a: sternum 7, b: sternum 8. Fig. 40, type 1, A. 
striata; Fig. 41, type 2, pomoniella; Fig. 42, type 3, edentata. 


~~ 


SYSTEMATICS OF THE GeNUs Augochlorella Nortu or Mexico 525 


526 Tue University SCIENCE BULLETIN 


jections along the anterior (inner) edge when the sclerite is in its normal 
inverted position (Fig. 43). However, this whole structure may be everted, 
in which case the projections extend posteriorly to the outside (Figs. 44, 45). 
There are differences among species in the number, shape and spacing of the 
finger-like projections. All the eastern species as well as pomoniella belong 
to type 1 (Fig. 43) characterized by 10 to 14 closely arranged, densely setose 
projections with the entire tergum minutely setose. The projections are 
sometimes branched at the tips in gratiosa. In neglectula (Fig. 46) there are 
8 to 9 short, thin, widely separated, and sparsely, minutely and inconspicuous- 
ly pubescent projections. The membrane appears finely pubescent laterally 
but bare medially. In edentata (Fig. 47) the projections number 12 to 14 are 
long, thin and very sparsely setose. The arms are minutely pubescent, but 
the transverse part appears bare. The differences noted here are easily ob- 
servable under 200% magnification of a compound microscope. Differences 
in the shape of the tergum, length and shape of the lateral arms, presence 
and location of punctures and arrangement and pattern of the pubescence 
are not considered of diagnostic value because of the fragile nature and 
eversibility of the entire structure. The eighth tergum is most easily removed 


with the genital capsule and seventh and eighth sterna to which it is attached. | 


OTHER CHARACTERS 


Other characters were studied and rejected because the characters were 
identical in all species, because they were too variable intraspecifically to 
have any meaning, or because the measurements or definitions were not 
precise or reproduceable. A total of 20 characters, most with multiple charac- 
ter states, were used in a factor analysis of differences between females of 
persimilis and striata. Although none of the characters was rejected on the 
basis of this study, it was shown that those characters deemed most useful in 
species recognition were indeed most highly correlated with the species (used 
in the analysis as one of the characters). These most useful characters were 
located in the propodeal area. 

Characters studied but seldom or not found useful among the nearctic 
species are: 

Head: Ratio of clypeal length to width; protuberance of clypeus and su- 
praclypeal area, color and shape of labrum; degree of rugosity of frontal area; 
roughening of vertex; relations between upper interocular, the interocellar 


Fics. 43-47. Hidden eighth metasomal tergum. Fig. 43, A. striata (normal inverted posi- 
tion); Fig. 44, gratiosa (semi-everted); Fig. 45, gratiosa (everted position); Fig. 46, neglectula; 


Fig. 47, edentata. 


Fics. 48-51. Male hind basitarsus, inner view. Fig. 48, A. striata; Fig. 49, neglectula; 


Fig. 50, persimilis; Fig. 51, gratiosa. 


SYSTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 527 


AG GEE 
ey) 


Ce 
Ho 


yy 


: Ye j Lt 


a7 
Y 


WI 
NO 
(oe) 


Tue University SciIENcE BULLETIN 


and the ocellocular distances; smoothness, color and length of pubescence of 
genal area; antennal contrasts in color of different segments. 

Thorax: The angle between the dorsum of the scutum and the slope of 
the propodeal disc; wing length; color, size, shape of the tegula; precise 
measurements and comparisons of size and spacing of punctures anteriorly, 
centrally, laterally and at anterolateral angles of the scutum; sculpture, color 
and smoothness of the scutellum; sculpture, color and smoothness of each 
part of the pleuron; size and spacing of the propodeal striations; ratio of dark 
to light areas on the femora and tibiae of the male; length to width ratios of 
the hind tibia and tarsal segments; color of the basitibial plate. 

Metasoma: Detailed color variations of the terga; descriptions, measure- 
ments and comparisons of size and spacing of punctures and surface mark- 
ings on all terga; punctures and lineations or other markings on the sterna; 
depth of the emargination of fourth sternum in males; comparisons of cer- 
tain structures of the genital capsule of males—length of the attenuation of — 
the inner lobe; roundness and height of the apex or crown of the inner lobe; 
spacing and number of setae on the inner lobe; shape, location and area of | 
fusion of other lobes and processes on the gonostylus; comparison of profiles 
(side view) of the capsule. 


SPECIES GROUPS 

The Nearctic group of species (1.e., Augochlorella s. str., not Pereirapts) 
includes both Neotropical and Nearctic species although only the Nearctic 
species are considered at this time. This group is divisible into eastern, west- 
ern and southern subgroups, each showing certain distributional and mor- 
phological affinities. The only species included in this work from the 
southern subgroup is edentata, but there are undoubtedly other species to 
the south. The western or “pomoniella subgroup” contains pomoniella and 
neglectula. The eastern or “striata subgroup” consists of aurata, striata, per- 
similis, gratiosa and bracteata. 

The pomoniella subgroup ranges from northern California southeast into 
Arizona and New Mexico in the United States and at least as far south as 
Costa Rica and Panama. The striata subgroup is found from south central 
and southeastern Canada through the eastern, midwestern and southern 
United States and as far west as Colorado and New Mexico. Except for 
bracteata, none of the included species ranges into Mexico. 

Morphological characters separating the striata and pomoniella subgroups 
involve the shape of the fourth metasomal sternum of males (margin straight 
in pomoniella subgroup, emarginate in striata subgroup) and the shape of 
the inner lobe of the gonostylus of the male genital capsule. The degree of 
morphological variation occurring within the two groups is also significant. 


SYSTEMATICS OF THE GENUs Augochlorella Nortu or Mexico 529 


These characters and their relationships among the species are discussed 
under each taxon. It is quite probable that significant biological differences 
may be found between the subgroups, but as yet, only the biologies of 
persimilis and striata are known (see Ordway, 1965a; in press). 

The species of the eastern subgroup seem more closely related to each 
other than are the two species of the western subgroup. In the striata sub- 
group there is greater morphological variation and intergradation among the 
species and the male genitalia are all essentially alike. In the pomoniella sub- 
group the male genitalia differ, and there is comparatively little variation 
and virtually no intergradation between the two species. 

The relationship of edentata to the northern Augochlorella and to the 
Peretrapis group is not clearly understood. The female appears to belong to 
Peretrapis and has been placed in that group by Moure. The male, however 
(previously undescribed), has clypeal and genitalic characteristics of the 
northern species but in other respects looks similar to Pereirapis. 


SPECIES ACCOUNT 
KEY TO THE sPECIES OF Augochlorella 


This key must be used in conjunction with the regional keys and the sec- 
tion entitled Regional Discussion because of the wide variability within 
species and the continuous or intergrading nature of all characters. The key 
for females will not work for all specimens since intermediates are found 
among many of the species and forms. The term “disc” in all cases refers to 
the dorsal area of the propodeum. 


FEMALES 


1. Striae of disc reaching edge posteriorly, i.c., with little or no unstriated 


area between ends of striae and edge of disc (Figs. 57-63) .........c0ecc000.0--- 2 
— Striae of disc not reaching edge posteriorly, with distinct unstriated 

region between ends of striae and edge of disc (Figs. 64, 67, 69) ............ 11 
2. Scutum coarsely punctate and strongly rugose anteriorly (Fig. TM) a 3 
— Scutum with small, distinctly separated punctures, finely roughened to 

weakly rugose anteriorly (Fig. 78) ......... , Sade Uae aaa see ater se aires Aer 2. ha 9 


3. Striae of disc fine and close together; disc at least weakly bracket-shaped 
(Figs. 57, 58), length less than or only slightly longer than 


MME tAMO LUI 2.222 occ ec deecttec veces aurata, gratiosa, striata a (see regional keys) 
— Striae of disc coarse, close together to widely separated; disc variably 
shaped, longer than metanotum a... ase eee (striata) 4 


4. Posterior edge of disc sharply angulate (Fig. 21) or abruptly rounded 
(Fig. 22); disc sharply pointed medially; length of disc less than 1.5 
fimmesithatok metanotum (Fig.:60)\ -2.2.0.c ee eae striata A 


10. 


a 


IZ. 


1S 


Tue University SCIENCE BULLETIN 


Posterior edge of disc gradually rounded or at most abruptly rounded, 
not sharply pointed medially, or if so, length more than 1.5 times 


as lomg as metamOtusny en. --ececsseneen cae cteeteeccee eects roe 5 
Disc at least weakly bracket-shaped, length equal to or only slightly 
longer than metanotum (Fig. 71) ~~ striata D 
Disc V-shaped, U-shaped or semicircular (Figs. 16-19) = 6 
Disc obtusely and bluntly V-shaped (Figs. 15, 63), length 1.5 to 2 
times that of metamotumm -2..:------_----2---c-cecenenneeenecec cess eeee eee censor striata B 
Disc U-shaped or semicircular, or if V-shaped then less than 1.5 times 
lemgth Of TmetamOtuna ec c-ce--aeteeeccnn eect 7 


Shape of disc semicircular or U-shaped; edge smoothly and gradually 
rounded posteriorly (Figs. 24, 62) without distinct difference in sculp- 
ture between dorsal and vertical surfaces ~.........-.--------------- striata © 
Shape of disc variable, posterior edge slightly raughened, abruptly 
angulate (Figs. 22, 23) or with distinct contrast in sculpture between 


dorsal and vertical surfaces. 2-.c.cec..-ncqececc-seceenenecseecectseesee sooner a 8 
Disc roundly V-shaped with medial striae ending abruptly at posterior 

Ps oO POE ance eer err striata D 
Disc semicircular with striae ending gradually near posterior 

COSC ce. asetcsecies ee eepete eter tee corpse striata © 


Shape of disc semicircular (Fig. 66); scutum with small, distinctly 
separated punctures, space between punctures smooth and_ shiny 
(Fig. 78) (Texas and Mexico only) .....-----------1---1 bracteata 
Shape of disc at least weakly bracket-shaped (Figs. 57, 58); scutum 
with punctures irregular in size and shape, very close or contiguous 
with little or no smooth shiny space between (Fig. 81), giving scutum 
a slightly roughened appearance .......-------------s---esssessercrerrr ttn 10 
Posterior vertical surface of propodeum finely and evenly granular 
or smooth (Fig. 74), length of disc equal to or slightly greater than 
length of mietamotunny 2. ecceccccan cece tte rece rec aurata 
Posterior vertical surface of propodeum irregularly or coarsely granular 
with minute irregular ridges (Fig. 76), length of disc equal to or less 
than“ metanotuny —l0es)4 2 ee ee gratiosa 
Scutum with small, distinct punctures, surface between punctures 
smooth, or minutely roughened and without distinct punctures (Figs. 


78, 19). cvcestc pean octet A gc a sate cacao 1g 
Scutum with surface rough, coarsely punctate or rugose, at least an- 
(Gsinlohd hf eee persimilis, striata c, neglectula, pomoniella (see regional keys) 
Antenna with yellow tip; scutum without distinct punctures (Fig. 
79): propodeal disc more than 1.5 times as long as metanotum ........ edentata 
Antenna with dark tip; scutum with close, distinct punctures (Figs. 
78, 80); propodeal disc 1.5 times as long as metanotum or less ....2.20 eee 13 
Striae occupying three-quarters the length of the disc or less as in 
Figure 69, with posterior edge of disc smooth and shiny ......---------- pomoniella 


Striae occupying more than three-quarters the length of the disc, 
pying q 8 
posterior edge roughened and dull Aloe ee ee 14 


SYSTEMATICS OF THE GENUs Auwgochlorella Nortu or Mexico 531 


14. Posterior vertical surface of propodeum finely and evenly granular or 
smooth; tegula shiny without distinct punctures bracteata 
— Posterior vertical surface of propodeum roughened with horizontal 
rugae extending across posterolateral corners; tegula dull, with small 
buiiycIstmnct qumetUTeS: cases ee ey neglectula maritima 


MALes 


No attempt is made to separate males of striata into forms a to p, since 
only a few can be so classified. Differences among these forms, when present, 
are explained at the end of the description of striata. 


I. Fourth metasomal sternum with apical margin straight 0... 2 
— Fourth metasomal sternum with apical emargination a 
2. Metasomal terga green or blue; first sternum with metallic reflections ... 3 


— Metasomal terga brown; first sternum without metallic reflections .. edentata 
3. Posterolateral corner and lateral vertical surface of propodeum smooth, 


with punctures; frons without bluish reflections pomoniella 
— Posterolateral corner and lateral vertical surface of propodeum rough 
or rugose, not punctate; frons with bluish reflections neglectula 


4. Hind basitarsus with erect hairs of two distinctly different lengths 

(exclusive of basal tuft), longest hairs at least twice as long as width 

of segment, usually curved at tips (Figs. S025) Set aie ae 5) 
— Hind basitarsus with erect hairs of similar lengths (exclusive of basal 

tuft), 1.5 times as long as width of segment or less; all hairs straight 

GiresM 48-40 Tee aurata, striata, bracteata (see regional keys) 
5. Basal third of hind basitarsus without long erect hairs; longest hairs on 

apical two-thirds about twice as long as width of segment (Fig. 50); 

last antennal segment similar in color to preceding segment, not en- 

tirely dark brown; fifth metasomal sternum without greenish reflec- 

TCI ee Da LO ee oth Seis WAL ee MRE AS, CT persimilis 
— Basal third of hind basitarsus bearing long curved hairs, four times 

as long as width of segment (Fig. 51); least antennal segment entirely 

dark brown; fifth metasomal sternum with greenish reflections ........ gratiosa 


REGIONAL Keys 


If the user of the keys is directed to the Regional Keys by the main keys, 
he should select the region to which his specimens belong and continue 
keying at the appropriate couplet. 


NortHeEAst (Canada, Maine, New Hampshire, Vermont, Massachusetts, 
Connecticut, Rhode Island, Pennsylvania, New York, New Jersey.) 
FEMALES 
From couplet 3: all specimens are sfriata A. 
From couplet 11: all specimens are striata c. 


532 Tue University ScIENCE BULLETIN 


MALEs 
From couplet 4: all specimens are s¢riata. 
Nortu Crentrat (Ohio, Indiana, Michigan, Illinois, Kentucky, Kansas, 
Colorado, Wyoming, Nebraska, Iowa, Wisconsin, Minnesota, North 
Dakota, South Dakota, Montana. ) 


FEMALES 

From couplet 3: all specimens are striata A. 

From couplet 11: intergrades occur throughout the region but par- 
ticularly in Illinois, Indiana, Missouri, Nebraska and Kansas. 
Check regional discussion for description of variations. 

Surface of disc beyond striae smooth; striae extending almost 
to posterior edge of disc (Eig 103) eee ere striata © 
Surface of disc beyond striae usually linearly roughened; striae 
extending about four-fifths or less the length of disc (Fig. 
(6 geen eo Cnty a ai) Pau rendre at Pee teat Nee bctoic., persimilis 
Mates 
From couplet 4: all specimens are striata. 
SoutHEast (Florida, Georgia, South Carolina, North Carolina, Virginia, 
West Virginia, Maryland, Delaware.) 


FEMALES 
From couplet 3: 
A. Posterior vertical surface of propodeum finely and evenly 
granular or smooth (Fig. 74); length of disc equal to or 
slightly greater than length of metanotum ...............--.--.-----------— B 
— Posterior vertical surface of propodeum irregularly or coarsely 
granular (Fig. 76); length of disc equal to or shorter than 


FAME CANO CUI setae reste eee gratiosa 
B. Striae of disc thick, regular, straight or slightly wavy (Fig. 

20) aera ales nea Sey distant ue ieee erat Fae recite certcecncnccns striata & 
— Striae of disc fine, close, irregular (Fig.57) 2222 eee aurata 


From couplet 11: nearly all specimens are persimilis from this area 
as striata c is not common and is usually large (more than 6 mm 
long) with characters not usually confused with those of persimi- 
lis. Neither one has been seen from Florida. 
Striae of disc almost reaching posterior edge; surface beyond 
striae smooth and shiny or at most minutely reticulated (Fig. 
(eee reese ep ee rect cereeecrtotc foo striata C 
Striae extending about three-fourths the length of disc medially, 
area beyond striae usually finely and linearly roughened 
parallel to edge of disc: (Fig. 64) ) 2 2-2 eee persimilis 
MALEs 
From couplet 4: 
Striae of disc fine and close together (Fig. 57); scutum with small, 
distinct but crowded punctures, weakly rugose anteriorly .... aurata 
Striae of disc coarse, widely or closely spaced (Figs. 59, 60, 63); 


SYSTEMATICS OF THE GeNus Augochlorella Norra or Mexico 533 


scutum smooth with large, widely spaced punctures or rough 
and coarsely punctate, coarsely rugose anteriorly 00... striata 
SoutH Centrat (Texas, Louisiana, Mississippi, Alabama) 


FEMALES 
From couplet 3: 

Posterior vertical surface of propodeum uniformly and finely 
granular (Fig. 74); length of disc equal to or longer than 
metanotum medially, striae straight or irregular and indistinct: 
apex of clypeus less than one-third brown: in Texas, first 
metasomal tergum strongly punctate (Fig. 82) 0.00 aurata 

Posterior vertical surface of propodeum roughly or irregularly 
granular (Fig. 76); length of propodeal disc equal to or 
shorter than metanotum, striae straight and regular (Fig. 

58); clypeal apex more than or less than one-third brown; 
in Texas, first metasomal tergum finely punctate (Fig. 83) _.. gratiosa 
From couplet 11: 

A. Posterior vertical surface of propodeum and posterolateral 
corners finely and transversely rugose; south-western 
IX CES Ese eds ae ee me neglectula 
— Posterior vertical surface of propodeum and _ posterolateral 

corners smooth and shiny or finely granular B 
B. Striae extending almost to posterior edge of disc; surface 
beyond striae smooth or slightly irregular or roughened .... striata p 
— Striae extending about three-fourths the length of disc; sur- 
face beyond striae usually linearly roughened ................... persimilis 
MAtes 
From couplet 4: 
Scutum finely roughened anteriorly; posterolateral corners and 
lateral vertical face of propodeum punctate to weakly 
DUM CtORUS OSE cee ee ee bracteata (Texas only) 
Scutum with anterior margin coarsely rugose to areolate; pos- 
terolateral corners and lateral vertical surface of propodeum 
FOWP I MONS OSG ae arate we Ge ke ce striata (throughout the area) 
West (California, Nevada, Utah, Arizona, New Mexico) and Mexico. 


| FEMALEs 
From couplet 3: none of these species range into this region. 
From couplet 11: 

A. Scutum with large, distinct, but close punctures; surface be- 
tween punctures smooth; posterolateral corners of propodeum 
prominent, smooth and shiny (Fig. 69); posterior vertical 
surface of propodeum smooth, finely punctate (Fig. 73); striae 
of disc short, rarely extending more than two-thirds the length 
of the disc medially, surface beyond striae smooth and 
minutelyerchiculated (Figs 69)... eee pomoniella 


Tue Universiry SCIENCE BULLETIN 


Ut 
WwW 
_ 


— Scutum coarsely punctate; posterolateral corners of propodeum 
rough, dull, not prominent (Fig. 70); posterior vertical surface 
of propodeum finely roughened to weakly rugose (Fig. (by): 
striae of disc reaching about three-fourths the length of disc 
medially, surface beyond striae dull and granular -........- B 
B. Tegula shiny, oval, smooth, without distinct 
PUULCEUINGS eee neglectula neglectula 
— Tegula dull, oblong, distinctly and roughly punctate (Mexico 
Of Ip) Neieten apes MEP Sra at toca Beh eR ee AB EA cons neglectula maritima 
Mates 
From couplet 4: all Mexican specimens are bracteata; all western 
specimens are striata. 


Augochlorella pomoniella (Cockerell) 


Augochlora pomoniella Cockerell, 1915, Pomona Jour. Ent. Zool. 7:232 (descr.); 
Cockerell, 1916, Pomona Jour. Ent. Zool. 8:51 (descr., distr.); Bray, 1917, 
Pomona Jour. Ent. Zool. 9:99 (list); Sandhouse and Cockerell, 1924, Proc. 
California Acad. Sci. (4)13:339 (distr., key); Cockerell, 1926, Ann. Mag. Nat. 
Hist. (9)18:624 (distr.); Cockerell, 1927, Pan-Pacific Ent. 3:162 (distr., 
descr.); Michener, 1936, Pan-Pacific Ent. 12:172 (distr.); Cockerell, 1937, 


Amer. Mus. Novitates 948:12 (distr.); Michener, 1937, Ann. Mag. Nat. Hist.) 


(10)19:314 (descr.); Cockerell, 1939, Proc. California Acad. Sci. (4)23:42% 
431 (distr., f.); Cockerell, 1939, Bull. So. California Acad. Sci. 38:139 (distr.); 


Cockerell, 1941, Proc. 6th Pacific Sci. Congr. 4:289 (distr.); Linsley, Mac-¥ 


Swain, Raven, 1963, Univ. California Pub. Ent. 33:44 (f1.). 

Augochlora (Augochlorella) pomoniella pomoniella: Michener, 1951, in Muese- 
beck et al. U.S. Dept. Agr., Agr. Monogr. 2, p. 1125 (ist). 

Augochlorella pomoniella: Sandhouse, 1937, Jour. Washington Acad. Sci. 27:69, 
71 (key, tax.); Michener, 1954, Bull. Amer. Mus. Nat. Hist. 104:55 (descr.); 
Linsley, 1962, Proc. Ist Internat. Sympos. on Pollination, Copenhagen 1960, 
p. 194 Ci.). 

Augochlorella pomoniella pomoniella: Krombein, 1958, U.S. Dept. Agr., Agr. 
Monogr. 2, Ist suppl. p. 232 (list). 


Augochlora utahensis Michener, 1937, Ann. Mag. Nat. Hist. (10)19:314 (descr.). 


Augochlora (Augochlorella) pomoniella utahensis: Michener, 1951, 1 Muese- 
beck et al., U.S. Dept. Agr., Agr. Monogr. 2, p. 1125 (list). 


Types. Augochlora pomoniella, holotype female, from Aliso Canyon (2 
miles from Laguna Beach) [Orange County |, California (R. LaFollette) is 
in the collection of the U.S. National Museum. Augochlora utahensis, holo- 
type female, from Rockville | Washington County], Utah, May 7, 1951 (1. 


Fics. 52-54. Shapes of heads of Augochlorella. Fig. 52, neglectula, \onger than wide; Fig. 
53, persimilis, as long as wide; Fig. 54, pomoniella, wider than long. 

Fic. 55. Front view of head of A. edentata. 

Fic. 56. First and second metasomal terga with waxlike exudate. 


5 


6 


Tue University SCIENCE BULLETIN 


Fic. 57. Propodeal disc of A. aurata. 
Fic. 58. Propodeal disc of A. gratiosa. 


SYSTEMATICS OF THE GENUS Augochlorella NorrtH or Mexico —_537 


Fic. 59. Propodeal disc of A. striata a from Florida. 
Fic. 60. Propodeal disc of A. striata a from east coast (standard). 


Tue University SciENCE BULLETIN 


Fic. 61. Propodeal disc of A. striata c. 
Fic. 62. Propodeal disc of A. striata c (standard). 


SYSTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 


Fic. 63. Propodeal disc of A. striata B (standard). 
Fic. 64. Propodeal disc of A. persimilis. 
Fic. 65. Propodeal disc of A. striata B-c intermediate. 


540 Tur Universiry ScIENCE BULLETIN 


Fic. 66. Propodeal disc of A. bracteata. 
Fic. 67. Propodeal disc of A. edentata. 
Fic. 68. Propodeal disc of A. striata c-persimilis intermediate. 


SYSTEMATICS OF THE GENUs Augochlorella Nortu or Mexico 541 


Fic. 69, Propodeal disc of A. pomoniella. 
Fic. 70. Propodeal disc of A. neglectula. 


at 


Tue University SCIENCE BULLETIN 


Fic. 71. Propodeal disc of A. striata p (standard). 
Fic. 72. Propodeal disc of A. striata B-D intermediate. 


SYSTEMATICS OF THE GENUs Augochlorella Nortu or Mexico 543 


Fic. 73. Posterior vertical surface of propodeum, A. pomoniella. 
Fic. 74. Posterior vertical surface of propodeum, A. persimilts. 

Fic. 75. Posterior vertical surface of propodeum, A. neglectula. 

Fic. 76. Posterior vertical surface of propodeum, A. gratiosa. 


5) 


44 


Tue University ScIENCE BULLETIN 


Fic. 77. Mesoscutum, A. striata. 
Fic. 78. Mesoscutum, A. bracteata. 
Fic. 79. Mesoscutum, A. edentata. 


SysTEMATICS OF THE GeNUs Augochlorella Nortu or Mexico = 545 


Fic. 80. Mesoscutum, 4. pomoniella. 
Fic. 81. Mesoscutum, A. gratiosa. 


VERSITY SCIENCE BULLETIN 


546 


Fic. 82. Metasomal punctures, A. aurata. 
Fic. 83. Metasomal punctures, A. gratiosa. 


SYSTEMATICS OF THE GENUs Augochlorella Nortu or Mexico — 547 


Wilson), on Datura, and one paratype female are in the Snow Entomological 
Museum, The University of Kansas. 

Description. Female: (1) Length 7 to 9 mm; head width 1.8 to 2.6 mm, 
averaging 1.90 mm, usually wider than long (rarely as wide as long). (2) 
Color bright green to blue-green; frons often with slight bluish reflections; 
metasoma similar in color to head and thorax or darker or browner in Mexi- 
can specimens. (3) Mandible usually with basal half dark brown, lighter 
brown becoming ferruginous apically, with or without metallic reflections 
basally. (4) Clypeus about twice as wide as long; basal half green with large, 
widely spaced punctures, closer basally; apical half dark brown or black and 
slightly beveled, with large elongate punctures; surface between punctures 
usually smooth and shiny, sometimes minutely roughened. (5) Supraclypeal 
area shiny and irregularly punctured; surface smooth or sometimes minutely 
roughened. (6) Paraocular area punctorugose to finely rugose below level 
of antennae, more coarsely rugose above. (7) Antenna entirely dark brown; 
flagellum often slightly lighter below than above; pedicel as long as broad; 
first flagellar segment almost twice as wide as long. (8) Scutum regularly 
and densely punctate; punctures small, deep and close; anterior margin 
roughened medially, becoming weakly rugose at anterolateral angles. (9) 
Tegula less than twice as long as wide. (10) Scutellum with small, deep, 
close punctures. (11) Pleuron rugose, more coarsely so anteriorly. (12) Pro- 
podeum with disc 1.5 times as long as metanotum; outline of disc U-shaped, 
profile type 5, posterior edge abruptly rounded, shiny and smooth; disc ex- 
tending slightly onto posterior surface of propodeum and particularly onto 
posterolateral corners; striae wavy, irregular, moderately coarse, extending 
about two-thirds the length of disc medially, usually attaining edge at ex- 
treme lateral corner; surface beyond striae minutely reticulate; posterior 
vertical surface smooth and shiny, sparsely and minutely punctured; pos- 
terolateral corners prominent, shiny, smooth, with few widely spaced punc- 
tures; lateral surface finely and linearly rugose anteriorly, weakly punctate 
toward posterior angles. (13) Legs brown, fore and hind coxae and trochan- 
ters with strong metallic reflections, femora usually with weak metallic re- 
flections. (14) First metasomal tergum polished and brilliant, punctures fine 
and widely spaced anteriorly, smooth and shiny dorsally with punctures 
small, distinct, numerous, closely and regularly spaced; other terga with 
minute, often indistinct punctures close together; first sternum without me- 
tallic reflections. (15) Pubescence white on head, pale, golden to white on 
vertex, thorax, metasoma and legs; pubescence short and sparse on genal 
area, denser and coarser in Mexican specimens. 

Male: (1) Length 9 mm; head with 1.75 to 2.25 mm, averaging 2.18 mm, 
width equal to length. (2) Color bright green to blue-green; frons without 
blue reflections on green specimens; metasoma dark green, usually darker 


548 Tue UNiversiry SciENCE BULLETIN 


than rest of body. (3) Mandible usually with metallic reflections basally. 
(4) Clypeus with punctures large, widely and irregularly spaced, surface be- 
tween punctures usually smooth and shiny. (5) Supraclypeal area punctate 
to punctorugose below antennae, usually smooth, shiny and sparsely punctate 
basally. (6) Paraocular area with small distinct punctures below level of an- 
tennae, finely ruguose above. (7) Flagellum dark brown above, yellow- 
brown below; scape, pedicel and usually first flagellar segment dark brown 
in the United States and some Mexican specimens, last one or two flagellar 
segments often slightly darker than preceding segments; pedicel and first 
flagellar segment about equal in size, each about 1.5 times as wide as long. 
(8) Scutum shiny and smooth, uniformly punctured; punctures distinct, sep- 
arated by less than a puncture width medially, more crowded at periphery 
than at center in specimens from areas outside of California; anterior margin 
roughened or finely rugose, becoming slightly more coarsely so at lateral 
angles. (9) Tegula twice as long as wide, entirely pale yellow in some Mexi- 
can specimens. (10) Scutellum shiny, roughened and punctate; punctations: 
generally irregular in size and spacing. (11) Pleuron rugose to punctorugose, 
becoming areolate anteriorly. (12) Propodeum with disc 1.5 times as long as 
metanotum; outline of disc truncately to obtusely U-shaped; disc nearly 
horizontal and slightly concave, slightly lower at posterolateral corners than 
posteromedially; posterior edge abruptly rounded, smooth and shiny; striae 
rather coarse, distinct, wavy, extending about two-thirds length of disc medi- 
ally or nearly to margin when median striae depressed, reaching edge lateral- 
ly, often extending onto lateral surfaces of propodeum; surface of disc beyond 
striae shiny, smooth, minutely reticulate, smooth area extending onto pos- 
terior surface; posterior surface of propodeum shiny, usually slightly rough- 
ened with shallow, widely spaced punctures of variable size and density, 
distinct to indistinct; posterolateral corners closely and shallowly punctate; 
lateral surface regularly and distinctly punctate with surface between punc- 
tures smooth and shiny. (13) Legs brown, coxae, trochanters, femora and 
tibiae with metallic reflections; hind basitarsi with erect hairs of uniform 
length and density from base to apex, about as long as width of segment; 
basal tuft absent. (14) Metasomal terga dark green with apical margins 
slightly depressed, narrowly brown. First tergum polished anteriorly with 
numerous widely spaced punctures, smooth, usually dull dorsally, punctures 
small and close; second and third terga dull, punctures small and close; sterna 
light brown, minutely pubescent, hairs often longer and denser at apical 
margins than elsewhere, less so medially than laterally; first sternum usually 
with feeble metallic reflections; apical margins of all sterna straight or slight- 
ly convex. (15) Pubescence white. (16) Genital capsule as in Figures 26, 27, 
35, type 2; inner lobe of gonostylus “fan-shaped,” bearing variable number 
(usually about 8) of large heavy setae, inner portion divided to form finger- 


SYSTEMATICS OF THE GENUS Augochlorella Norra or Mexico 549 


like section with two to five setae (usually 2 or 3); separation deep or shal- 
low (so that “finger” may appear long or short, but always evident) ; sterna 7 
and 8 without setae, variable in shape, 8 with broadly rounded basal edge 
medially, not truncate (Fig. 41b); 7 produced into knob-shaped lump medial- 
ly, each arm with minute setose protuberance on distal inner angle (Fig. 
41a); tergum 8 of type 1 (Fig. 43). 

Comparisons. A. pomontella is, perhaps, the most distinctive of all the 
North American Augochlorella. It is the largest (Fig. 86), smoothest and 
most brilliant of the species and the least variable. It seems most closely re- 
lated to neglectula, although the male genitalia show closer affinities with 
those of the eastern species. Both male and female can be distinguished from 
neglectula by the more widely separated and larger scutal punctures and the 
generally smooth body surface, the greener legs. the polished posterior part 
of the propodeal disc and posterolateral corners. the smooth posterior propo- 
deal surface, and in the males by the genitalia (distincuishable from all other 
North American species) and the sternal punctures and setae. 

No pomoniella were found that would be confused with any other species. 

Variation. Californian specimens, together with those from Nevada. both 
male and female, are the most morphologically stable of any species. Vari- 
ability in size and color increases in Arizona and Mexico, where specimens 
tend to be smaller with the brown areas paler. 

Specimens are larger in California (mean head width=2.33 mm) than in 
Mexico or Arizona (mean head width=2.07 mm) (Fig. 84) while females 
from Utah are about the same size as in California. Differences noted by 
Michener (1937) represent normal variations found not only among individ- 
uals from Utah but also in Californian material. 

The metasoma of both males and females is noticeably darker in Mexican 
specimens than in specimens from the United States, with a greater tendency 
for the brown to dominate the green in males or brown to become black in 
females. Other brown structures, such as the mandibles, tarsi and tegulae, 
are paler in males from Mexico than in those from the United States, so that 
on the mandible, the dark brown area is restricted to the basal portion, and 
the tarsi of most specimens are lighter in color than other parts of the legs, 
contrasting with the uniformly dark brown legs of specimens from the 
United States. One series of nine specimens from Yucatan (13*) had both 
tibiae and tarsi pale although the tibiae were normal in the two other Yuca- 
tan males examined. All male specimens from Yucatan and Sonora had pale 
testaceous tegulae, although specimens from other parts of Mexico were 
normal. None of the females showed these variations in color. 


* See Table 1. 


550 Tue UNiversity ScIENCE BULLETIN 


Thoracic punctation is uniform both in size and spacing in males from 
California, but in Arizona and Mexico the punctures are closer together and 
irregular in size at the periphery of the scutum. Females vary little in this 
respect. 

The surfaces of the head and thorax are dull and finely roughened by 
minute reticulations in many female specimens from Mexico. This roughen- | 
ing is most apparent on the supraclypeal area, clypeus, scutum and scutellum. 
In addition, the clypeus is entirely brown or black or the brown area extends 
medially to the base. Such variants were not found in Baja California; only 
1 of 28 specimens from Sonora showed such characters, but all 27 specimens 
from other Mexican states showed them in varying degrees. 

As with other species of Awgochlorella, the propodeal area shows the 
greatest amount of variation but even this is less than in other species and 
there is no geographical trend or seasonal pattern in the variation. The striae 
of the disc are distinct in all males but in females there are occasional in- | 
dividuals in which the striae are fine, weak, or barely recognizable. The 
striae normally extend half the length of the disc, although in about one-third 
of the specimens they are longer medially and almost attain the edge of the 
disc. Such length is more noticeable in the males where the median area 1s 
slightly depressed when this condition occurs. 


E A 


X= 2.30 


frequency 


millimeters 


Fic. 84. Measurements of head widths of females of A. pomoniella. A. From California and 
Utah. B. From Arizona and Mexico. 


~ 


SYSTEMATICS OF THE GEeNUs Augochlorella Nortu or Mexico ‘551 


Distribution. From northern California just north of San Francisco and 
western Nevada, southward through central and southern California; Wash- 
ington Co., Utah, throughout the western half and southern part of Arizona; 
Baja, California, the west coast of Mexico to Chiapas and eastward to Yuca- 
tan, southward into Guatamala and Costa Rica (Map: Fig. 85). 

A total of 118 males and 631 females have been seen. ARIZONA: Co- 
chise Co..5 & (Huachuca Mts.: Mouth of Carr Canyon; 7 mi. SW. Wilcox; 
Benson) (July, September); Coconino Co.,2 4,26 2 (Grand Canyon: In- 
dian Garden ca. 3800 ft., Phantom Ranch ca. 2500 ft., South Rim 6800 ft.; 
Supai; Havaisu Canyon 3500 ft.) (June-August); Maricopa Co.,1 8,7 2 
(Granite Reef Dam; 25 mi. E. Gila Bend; 20.7 mi. S. Gila Bend; Tempe) 
(March, July, September); Mohave Co. 1 2 (Grand Canyon, mi. 179.2 at 
Lava Falls) (June); Pima Co., 22 6, 78 2 (Tucson; Sta. Catalina Mts.: 
Pepper Sauce Canyon, Sabino Canyon, Sabino Basin 3800 ft., Ventana Can- 
yon, Cape Canyon, Hitchcock Highway mile posts No. 5, 6, 8, 9, 5500 ft.; 
Saguaro Nat. Mon.; 18 mi. W. Sells; 5 mi. N. Tucson; 20 mi. E. Tucson; 
Lowell Ranger Sta. 2700 ft.; Sahuarita; Baboquivari Mts., Kits Peak Rincon; 
Sierritas 31°51’ N. 111° 16’ W.; Black Dike Prspct. ca. 3750 ft.; Tanque 
Verde; Continental; Quitobaquito, Organ Pipe Nat. Pk.) (March-Novem- 
ber); Pinal Co.,4 8,8 2 (Superior; 20 mi. W. Casa Grande; Florence Jct.; 
Coolidge; Rio Aravaipa 2500 ft.) (February-March, June-July); Santa Cruz 
Co., 3 % (Coyote Mts. 31° 58’ N. 111° 29’ W., ca. 3500 ft.; Sonoita) (July- 
August); Yavapai Co.,2 2 (Seligman; 3 mi. N. Rock Spr.) (July). CALI- 
FORNIA: Alameda Co., 5 2 (Tesla) (October); Calaveras Co.,2 2 
(Murphys 2500 ft.); Contra Costa Co. 2 % (Mt. Diablo) (July); Fresno 
Co. 2 2 (Coalinga; Orange Cove) (April-May); Inyo Co. 3 8, 63 
(Lone Pine; Darwin Falls; Mazourka Canyon; Panamint Mts., Surprise 
Canyon; 7 mi. W. Westgard Pass; Death Valley; 5 mi. W. Lone Pine; Big 
Pine; Independence; Payson, 39° 19’ N. 118° 08’ W.; Inyo Mts. 7000-9000 ft.; 
Antelope Spr., 8 mi. SW. Deep Spr.) (April-August); Kern Co.,5 @ (Ar- 
vin; Democrat Spr.; 6 mi. W. Inyokern, Short Canyon; Caliente) (March, 
June); Kings Co. 2 @ (12 mi. SW. Avenal) (August); Los Angeles Co., 
13 3,43 2 (Sta. Catalina Is.: Avalon, Cape Canyon, Pebbly Beach, Rancho 
Escondito; Newton; Clairmont; 5 mi. S. Pearblossum; Acton: Eagle Rock; 

Altadena; Whittier; 6 mi. W. Palmdale; Pasadina) (February- September, 
Beember): Mariposa Co., 1 2 (1500 ft.) (May); Mendocino Co., 1 @ 
(Ryan Cr.) (July); Mono Co., 2 2 (Oasis) (May); Monterey Co. 8 6, 

1 & (Paraiso Spr.; Jamesburg; Sta. Lucia Mts., Hastings Nat. Hist. Res. 
1900-2700 ft.) (April-May, August-September, November); Napa Co., 14 2 
(Pope Valley; Mt. St. Helena; Conn Lake; Chiles) (March, May, Septem- 
ber); Orange Co. 3 2 (Newport Bay; Serra; Aliso Canyon nr. Laguna 
fies Laguna Beach) (July-August); Riverside Co. 4 6, 32 2 (Palm 


Dp2 Tue University ScreNce BULLETIN | 
Spr.; Riverside; San Jacinto; San Jacinto Mts.: Idyllwild Keen Camp; 10 mi. : 
W. Perris; Henshaw; The Gavilan; Whitewater; Corona; Palm Springs; 
Cathedral City; 2 mi. E. Anza; Elsinor; Murrieta; Andreas Canyon, Palm 
Spr.) (March-May, July, October); San Bernardino Co. 1 6,13 2 (Mill 
Cr.; Crestline; 5 mi. SE. Hesperia; 12 mi. SE. Ivanpah; E. Highlands; Mo- 
rongo; Chino Canyon; Argus Range, Indian Joe Spr. 2600 ft.; Colton) 
(March-May, August-September); San Diego Co., 3 &, 32 2 (Jacumba; 
Poway; Vista; Borego; 2 mi. N. Warner Spr.; 3 mi. S. Oak Grove; Barrett 
Spr.; El Cajon; San Diego; Campo; Descanso; Torry Pme ize Warren) 
(March-April, July-September); San Joaquin Co. 1 2 (Tracy) (July); 
San Luis Obispo Co., 2 ? (Paso Robles; 2.5 mi. S. Creston) (April, Septem- 
ber); San Mateo Co.,1 2 (Jasper Ridge) (September); Santa Barbara Co., 
1 @ (Sta. Cruz Is.) (May); Santa Clara Co, 2 8,9 2 (Alum Rock Came 
yon; Stanford, Palo Alto; San Jose; Mt. Hamilton; Uvas Cr.) (July-August, 
October); Sonoma Co.,5 2 (Guerneville; Preston) (May, July); Szani- 
slaus Co..1 2 (del Puerto Canyon) (April); Tulare Co,1 6,8 2 (Linge 
say; Lemon Cove 500 ft.; 3-Rivers 600-800 ft.; Porterville; Kaweah) (June- 
July, November); Tuolumne Co.,5 @ (Jamestown; 18 mi. SW. Sonora; 
13 mi. SW. Sonora) (April, July); Ventura Co., 54 ? (Quantal Canyon; 
5 mi. S. Gorman, Hungary Valley) (May). NEVADA: Douglas Co. 1 4, 
I 2 (3 mi. S. Genoa) (August); Washoe Co. 2 2 (Pyramid; Sutcliff) 
(June-July). NEW MEXICO: Rio Arriba Co. 1 2 (Willow Creek) 
(August). UTAH: Washington Co. 2 6,66 2 (Zion Nat. Pk.; Toquer- 
ville; Leeds; St. George; Sta. Clara; La Verkin; Washington; Hurricane) 
(May-September). 

BAJA CALIFORNIA: 16 2 (Cedros Is.: Bernstein Spr.; Isla Espiritu 
Santo; Coyote Cove, Conception Bay; Los Frailes; La Paz; 19 mi. E. Rosario; 
Canipol; 40 mi. S. El Arco Mine) (March, June, August, October). CHIA- 
PAS: 1 2 (2mi.N. Suchiapa) July). GUERRERO: 19 6,45 2? (7am 
N. Chilpancingo 2250 ft.; Chilpancingo 3700 ft.; 5.2 mi. E. Chilpancingo 
5700 ft.; 2 mi. S. Chilpancingo; 42 mi. N. Acapulco 1550 ft.; 9 mi. W. Aca- 
pulco; Acapulco; 42 mi. N. Acapulco 1550 ft.; Xalitla 1500 ft.) (March, 
August, December). MICHOACAN: 1 6,3 2 (11 mi. E. Apatzingan; 
4 mi. E. Apatzingan; Apatzingan; 10 mi. N. Morelia 5900 ft.) (July-August). 
MORELOS: 1 ¢ (11 mi. S. Tlaltizapan) (August). NAYARIT: 238 
(km. 78 Rte. 15) (September). OAXACA: 4 6,6 2 (23 mi. S. Matias 
Romero 200 ft.; 4 mi. NW. Tehuantepec 700 ft.; 14 mi. NW. Tehuantepec 
700 ft.; 6 mi. S. Tehuantepec 200 ft.; 10 mi. NE. Juchitan; 50 mi. N. La Ven- 
tosa) (June-July). SINALOA: 7 2 (6 mi. NE. Villa Unidén 350 ft.; 3 mig 
NW. Concha 50 ft.; 14 mi. SE. Elota; Los Moschis) (May, July). SONORA: 
7 6,18 2 (37 mi. N. Guaymas; 70 mi. N. Hermosillo; 4 mi. N. Guaymas; 


wi 
wii 
oS) 


SYSTEMATICS OF THE GENUs Augochlorella Nortu or Mexico 


Rio Mayo; 20 mi. SE. Empalme) (April, September). YUCATAN: 11 4, 
1 @ (8km.N. Muna; Mérida; Chichen Itza) (July). “San Carlos Bay” 1 ? 
(September); “San José de Guaymos” 4 6,1 2 (April). 

GUATEMALA: Retalhuleu, 2 ? (Champerico) (April). 

COSTA RICA: Guanacaste, 3 ¢ (El Coco) (August). 

A. pomoniella appears to be most abundant in California with smaller 
populations occurring from Baja California down the west coast of Mexico 
into Central America. Small and possibly isolated populations of pomoniella 
exist in Utah and possibly in New Mexico. The one specimen from New 
Mexico is without question pomoniella but does seem to be out of place. 
Whether this is due to mislabelling or to a lack of collecting cannot be de- 
termined from the available information. Specimens from the most south- 
western county of Utah, Washington Co., are little different from those in 
California and seem to be a part of the main Californian stock. Specimens 
have been taken from near sea level to 7000 feet elevation in California and 
Arizona and from 200 to 6000 feet in Mexico. Specimens were found near 
sea level in Costa Rica. 

The only species of Augochlorella with which pomoniella is in contact in 
the United States is neglectula, a species found chiefly on the Mexican Pla- 
teau and in the mountains of southern New Mexico and Arizona. Although 
the ranges of the two species broadly overlap in southern Arizona, they are 
apparently somewhat different ecologically, judging by random collections 
made in the Santa Catalina Mountains near Tucson, Arizona. These collec- 
tions show pomoniella occurring from the Sonoran desert near Tucson, a 
saguaro-mesquite habitat at about 2800 feet near the base of the mountains, 
to about 5200 feet in oak-juniper association. A. neglectula has been col- 
lected at about 3800 feet and ranges well into the pines near the tops of the 
mountain at about 8100 feet elevation. A. neglectula has not been collected 
in the desert area around Tucson. This comparison is based only on the 
label data of 73 pomoniella and 64 neglectula from this one area. 

Seasonal Activity. Females have been collected between February 28 and 
December 27 and males between May 3 and December 17, with little differ- 
ence in dates throughout the range. Females were collected with pollen in 
their scopes from early March until the beginning of October in California, 
with similar dates being recorded from other areas as well. The wide varia- 
tion in size (Fig. 84) is due to geographical differences rather than to seasonal 
or caste differences. 

Flower Records. Acacia, Arctostaphylos, Argemone, Asclepias, Aster, 
Baileya, Bebbia, Brassica, Carnegia, Centaurea, Cercidum, Chrysothamnus, 
Cirsium, Cissus, Cleome, Cryptantha, Dalea, Encelia, Eriogonum, Eucnide, 
Fendlerella, Gossypium, Gutierrezia, Haplopappus, Helianthus, Heliotropi- 


554 Tue University SciENCE BULLETIN 


Fic. 85. Distribution of A. pomoniella (solid dots) and bracteata (rings). 


um, Heterotheca, Hymenothrix, Isomerts, Kallstroemia, Melilotus, Oeno- 
thera, Opuntia, Penstemon, Peucephyllum, Rhus, Salix, Salvia, Senecio, 
Sisymbrium, Sphaeralcea, Tamarix. 


Augochlorella neglectula neglectula (Cockerell) 


Augochlora neglectula Cockerell, 1897, Bull. New Mexico Coll. Agr. Exper. Sta. 
24:43 (descr.); Cockerell, 1898, Bull. Sci. Lab. Denison Univ. 11:47 (descr.); 
Cockerell, 1898, Zool. 2:80 (fl., distr.); Cockerell, 1899, Catalogo de las Abejas 
de Mexico p. 6 (list); Cockerell, 1899, Canad. Ent. 31:256 (fl. distr.); Cockerell 
1900, Amer. Nat. 34:488 (fl., distr.); Cockerell, 1901, Ent. News 12:39 (i; 
Cockerell, 1902, Amer. Nat. 36:811 (f.); Cockerell, 1903, Ann. Mag. Nat. 
Hist. (7)12:442 (descr.); Cockerell, 1906, Trans. Amer. Ent. Soc. 32-283 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 555 


bracteata 


aurata 


neglectula 


pomoniella 


‘ n 
2 | 3 ; sericea X- 1.66 


frequency 


eee n 
persimilis 
X 


gratiosa 


millimeters 


Fic. 86. Histograms showing head widths of field caught females throughout the range of 
each species, samples from throughout the season, 


556 Tue University ScIENCE BULLETIN 


(distr.); Cockerell, 1915, Pomona Jour. Ent. Zool. 7:232 (descr.); Cockerell, 
1927, Pan-Pacific Ent. 3:162 (descr.). 

Augochlora dimissa Cockerell, 1923, Proc. U.S. Nat. Mus. 63:5 (descr.) (new 
synonymy ). 

Augochlora confusa: Cockerell, 1897, Bull. New Mexico Coll. Agr. Exp) Sta 
24:23, 25 (misidentification ). 

Augochlorella aurata: Sandhouse, 1937, Jour. Washington Acad. Sci. 27:71 (in 
part). (Since 1937, when Sandhouse erroneously synonymized neglectula with 
aurata, various authors have repeated this synonymy, but additional data per- 
taining to neglectula does not appear to have been given.) 


Types. Augochlora neglectula, holotype male, from Filmore Canyon 
[Dona Ana Co.|, New Mexico, August 29 (Townsend), in the collection of 
P. H. Timberlake, University of California at Riverside, California. Cotype 
(paratype) female, Filmore Canyon, New Mexico, August 24 (Townsend), 
No. 4345, in the collection of the U.S. National Museum. In 1906, Cockerell 
listed the type locality as Organ Mountains, New Mexico; probably Filmore 
Canyon is merely a more specific statement of the same locality. The species 
was recognized as new on the basis of male characters, and the holotype is 
well marked with “type” labels, although not in Cockerell’s usual hand. 
Augochlora dimissa, holotype female, from Victoria [Tamaulipas], Mexico, 
March 16, is in the collection of the U.S. National Museum, No. 25582. 

Description. Female: (1) Length 6 to 7 mm; head width 1.68 to 2.21 
mm, averaging 1.97 mm; head usually wider than long. (2) Color bright 
green to dark blue; frons usually with blue reflections; metasoma often 
darker than head and thorax and suffused with brown. (3) Mandible with 
basal third dark brown, reddish brown centrally, rufous apically, without 
metallic reflections basally. (4) Clypeus slightly wider than long; basal halt 
green with rather large punctures about a puncture width apart; apical half 
dark brown and slightly beveled, with about three to five large, often elongate 
punctures; surface between punctures smooth and shiny or finely reticulated 
at base and laterally. (5) Supraclypeal area irregularly punctate, sparsely so 
medially; surface between punctures usually shiny and smooth, sometimes 
minutely roughened and dull. (6) Paraocular area closely punctorugose be- 
low level of antennae, coarsely rugose above. (7) Antenna dark brown, fla- 
gellum slightly lighter below than above; pedicel as broad as long; first 
flagellar segment less than twice as wide as long. (8) Scutum with punctures 
variable in size and spacing, grading from distinctly and closely punctate to 
punctorugose, usually closer together laterally than medially; anterior margin 
finely roughened medially, becoming finely to coarsely rugose at anterolateral 
angles. (9) Tegula with length slightly greater than width, shiny, without 
conspicuous punctures. (10) Scutellum with small, close, irregular sized 
punctures, becoming indistinct in Mexican specimens. (11) Pleuron coarsely 


SYSTEMATICS OF THE GENUs Augochlorella Nortu of Mexico — 557 


rugose, areolate anteriorly. (12) Propodeum with disc usually less than 1.5 
times as long as metanotum; outline of disc broadly semicircular, profile type 
3, posterior edge abruptly rounded medially, becoming gradually rounded 
laterally; striae variable, usually regular, fine and radiating from medial area, 
medially extending about three-fourths length of disc and ending gradually, 
laterally nearly reaching edge or extending onto vertical surface; surface be- 
yond striae dull and granular to edge; posterior and lateral vertical surfaces 
finely roughened, usually with fine horizontal rugae extending from lateral 
to posterior surface across rounded posterolateral corners. (13) Legs brown; 
fore and hind coxae with strong metallic reflections; fore femur sometimes 
weakly metallic. (14) First metasomal tergum with anterior surface polished, 
sparsely and finely punctate, dull to shiny dorsally, with fine, distinct, close 
punctures; second tergum with fine, close punctures, surface between punc- 
tures smooth, shiny to dull; first sternum with or without metallic reflections. 
(15) Pubescence white on head and ventrally on thorax and metasoma, white 
to pale golden or dorsal parts of thorax and metasoma and on legs, golden 
on thorax of most Mexican specimens. 
Male: (1) Length 7 to 8 mm; head width 1.66 to 2.04 mm, averaging 1.85 
mm, greater than, equal to or less than length. (2) Color bright green to 
blue, frons with bluish reflections in all specimens; usually variably blue- 
green over entire body. (3) Mandible with or without metallic reflections 
basally. (4) Clypeus with punctures large, widely spaced, surface between 
punctures minutely roughened to smooth and shiny. (5) Supraclypeal area 
variably punctate, surface between punctures smooth and shiny or sometimes 
minutely reticulate and dull. (6) Paraocular area closely punctate to finely 
rugose. (7) Flagellum dark brown above, yellowish brown below; scape and 
pedicel entirely dark brown or black; last one or two flagellar segments 
usually slightly darker below than preceding segments; pedical as broad as 
long; first flagellar segment about twice as wide as long. (8) Scutum shiny 
with punctures distinct, variably spaced medially to parapsidal lines, closer 
laterally; becoming weakly rugose at lateral anterior margin. (9) Tegula less 
than 1.5 times as long as wide, shiny, with punctures inconspicuous or lack- 
ing. (10) Scutellum shiny, distinctly punctate, more densely so posteromedi- 
ally than elsewhere. (11) Pleuron areolate anteriorly and laterally. (12) Pro- 
podeum with disc slightly longer than, to 1.5 times as long as metanotum; 
outline of disc semicircular, posterior edge abruptly rounded; striae coarse, 
irregular, wavy, not reaching edge medially, extending onto vertical surface 
laterally; surface of disc beyond striae narrowly smooth and shiny or slightly 
toughened; posterior vertical surface shiny, finely roughened or rugose; pos- 
terolateral corners with lineate, horizontal rugae extending from lateral to 
posterior surfaces; lateral vertical surface finely rugose with horizontal lineate 
-rugae anteriorly. (13) Legs brown, fore and hind coxae and trochanters with 


558 Tue University SciENCE BULLETIN 


strong metallic reflections; femora and fore tibia weakly metallic; middle and 
hind tibiae and all tarsi brown; hind basitarsus with straight, erect hairs of 
uniform length along entire segment; these hairs almost 1.5 times as long as 
width of segment; basal tuft reduced to inconspicuous. (14) Metasomal 
terga green with apical margins brown; first tergum polished with fine, 
widely scattered punctures anteriorly, smooth but less shiny dorsally with 
punctures small and close together; second tergum minutely to indistinctly 
punctate; sterna brown, occasionally suffused with black, pubescence short 
and fine, evenly distributed; first sternum with weak metallic reflections; 
second through sixth sterna with apical margins straight. (15) Pubescence 
white over entire body to golden in some Mexican specimens. (16) Genital 
capsule of type 4 (Figs. 37-38); inner lobe of gonostylus long and thin with 
blunt apex, bearing 3 to 5 stout setae; posterior edge of lobe with setae vari- 
able in size and number; setae on outer lobe (lz Fig. 27) usually branched; 
seventh tergum type 2 (Fig. 46); seventh and eighth sterna type 1 as figured 
(Fig. 40). 

Comparisons. Most neglectula can be distinguished from all other North 
American species of Augochlorella by the fine radiating striae that extend 
only about three-quarters of the way across the propodeal disc, by the rugose 
nature of the posterior propodeal surface, by the blue areas on the frons, and 
by the shape of the inner lobe of the male gonostylus. In addition to these 
characters, it differs from pomoniella by the usually close, deep, often con- 
tiguous scutal punctures giving the scutum a rough or even rugose appear- 
ance. In pomoniella the scutum is smooth with distinct, widely spaced 
punctures. The legs, especially the trochanters and femora, are more uni- 
formly brown in neglectula than in pomoniella. In most specimens there are 
no distinct posterolateral corners to the propodeum (Fig. 70) or if there are, 
they are weak. The corners are rarely polished and shiny as in pomontella 
but usually are traversed by horizontal rugae extending from the lateral to 
posterior surfaces. This character will also distinguish neglectula from 
striata when the striae of the disc of neglectula become coarser and less 
strongly radiating than usual. Males can be easily distinguished from striata 
on the basis of the fourth sternum and genitalia. 

A truer picture of the relationships of neglectula to the other North 
American species of Augochlorella must necessarily wait until a study is 
made of the Mexican, Central and South American species of the genus 
with which it is possibly more closely related. It has no close affinities with 
Pereirapis and among the species from the United States is most similar to 
pomoniella. A. neglectula and pomoniella may look similar in areas where 
their ranges overlap but there is no evidence of genic exchange since the 
features of each species are maintained. There is a similar resemblance in a 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 559 


few individuals to striata but in all such cases also, neglectula maintains its 
identity. 

Variation. Populations of neglectula in Arizona show the greatest amount 
of variation, particularly in the degree of thoracic roughness. Females are 
more variable than males. 

There is not much difference in size among individuals from different 
areas of the range although the mean head width of females is largest in 
Mexico (1.99 mm) and smallest in Texas (1.85 mm). The width of the head 
is less variable in males, but no trends can be described due to the meager 
samples available from most areas. 

The supraclypeal area is shiny in most males and females from the United 
States but is rarely polished or brilliant. It may be entirely punctate or, more 
frequently, sparsely punctate medially, more densely so laterally. In many of 
the Mexican females, as in some Mexican pomoniella, the supraclypeal area 
is dull due to minute reticulations on the integumental surface. This dullness 
extends onto the basal area of the clypeus and over the scutum, and the 
brown color on the apex of the clypeus extends in a narrow medial line to 
the base of the clypeus. In no case is the clypeus entirely brown. But unlike 
pomoniella, other less dull or even shiny neglectula sometimes show this 
same variation of clypeal coloration. The dull specimens may be found 
throughout the range of Mexican neglectula from March to September. The 
supraclypeal area of the males is rarely dull even though it may be coarsely 
punctured. The minute fine reticulations were found on only three out of 
42 males from Arizona (from Yuma, Cochise and Pima Counties) and on 
four out of 18 Mexican individuals (from Chihuahua, San Luis Potosi, Du- 
rango and Guerrero). Although the roughening may extend onto the cly- 
peus, it does not appear on the scutum and the dullness is considerably less 
striking than on the females. The female holotype of Augochlora dimissa 
from Victoria, Mexico, shows this dull condition although in every other 
respect appears to be a normal Mexican neglectula. Since there is no appar- 
ent morphological difference associated with this condition and no geo- 
graphical pattern or even distinct population of dull individuals, there is no 
basis for recognizing this variant either as a species or subspecies. 

Although the blue patch on the frons is characteristic of this species, in 
many Mexican females the blue area is obscure and can only be found with 
difficulty. It is not found at all in the Mexican subspecies maritima. 

The punctures of the scutum are distinct with their diameters about 
equal to the spaces between them in about half the males (40 specimens) 
from all parts of the range, but are close, deep, and contiguous or form a 
rugose surface in the other half (37 specimens). The scutum is coarsely and 
closely punctured in females (except for six females from Arizona which 
have the punctures more widely spaced). 


560 THe UNiversiry SciIENcE BULLETIN 


There is considerable variation in the nature of the propodeal area, 


especially in specimens from Arizona. In the females, the Mexican speci- — 


mens show the typical neglectula pattern with fine, straight, radiating striae 
on the disc, often reaching the posterior medial edge of the disc. The hori- 
zontal rugae are rarely present on the posterior and lateral surfaces of the 
propodeum but these surfaces are rough and dull. When the striae are larger 
or less fine the rugae are present posterolaterally. “Typical” neglectula are 
also found in Arizona but individuals occasionally may resemble either 
pomoniella or striata. Those similar to pomoniella {10 out of 57 specimens 
from the Santa Catalina Mts. (1*), and two out of five specimens from 
Globe (9) | have short striae that are finer than found on the usual pomoniel- 
la. The posterolateral corners are weak and the edge of the disc is shinier 
than usual, but the polished surface does not extend onto the vertical sides. 
All these specimens have horizontal rugae on the posterior surface. Those 
that resemble striata have coarser striae than normal, often reaching almost 


to the posterior margin of the disc, but in all cases the vertical surfaces are | 


rough with well defined horizontal rugae. Such specimens were found com- 
monly in New Mexico and Texas although the rugae are often less distinct 
in Texan specimens. None of the specimens from New Mexico or Texas 
resembles pomoniella. 

In males, the striae of the disc are variable in thickness but generally 
rather coarse. In 10 of the 42 males from Arizona and 3 of the 18 Mexican 
males (Zacatecas, Chihuahua and “Guadalupe”), the propodeum resembled 
that of pomoniella, having shiny, smooth posterolateral corners and smoother 
vertical surfaces than is normal for neglectula, with the horizontal rugae 
indistinct or absent. Punctures on the propodeum were distinct although 
close and coarse so that the roughened character of neglectula is maintained 
in these specimens. 

The inner lobe of the gonostylus of the male genitalia is rather constant 
in shape but variable in the number and character of the setae it bears on its 
outer edge. These setae may number two or three and be short, very thin 
and flaccid (Fig. 37), but may vary to long, thick, heavily sclerotized and up 
to 14 in number. When the larger number are present the series is con- 
tinuous with the setae at the apex of the lobe, with one or two weaker setae 
between those on the posterior and apical margins (Fig. 39). If the setae are 
weak, there is usually a space between those of the two series (Fig. 38). As 
the setae become longer and thicker the lobe itself becomes shorter and 
broader. All intermediate conditions occur between the two extremes, and 
all forms apparently occur throughout the range. No correlation has been 
found between these genitalic differences and external morphological varia- 


* See Table 1. 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico — 561 


tion. In no case does the inner lobe look similar to that of any other species 
here considered. 

The outer lobe of the gonostylus bears long, fine, branched or unbranched 
setae. Branched hairs cannot be correlated with either locality or external 
morphological variation. 

The Guatemalan and Panamian specimens look like those from Mexico. 

Distribution. From the southern half of Arizona and southwestern New 
Mexico into the Big Bend area of Texas, southward through all of central 
Mexico, with specimens also from Guatemala and Panama (Map: Fig. 87). 

A total of 79 males and 279 females have been seen: ARIZONA: Apache 
Co.,1 2 (White Mts.) (June); Cochise Co.,7 6,23 2 (Huachuca Mts., 
Ramsey Canyon; Chiricahua Mts.: SW. Res. Sta. 5400 ft., Rustlers Park 
8784 ft., Portal 5000 ft.; Douglas; 6 mi. NE. Douglas) (March-August) ; 
Coconino Co.,1 ? (Oak Creek Canyon 3500 ft.) (August); Gila Co.,7 9 
(Bot Fly Canyon, Pinal Mts. 3500 ft.; Globe; Payson) (May, July, Septem- 
ber); Graham Co..3 6,2 2 (Graham Mts.: 6000-7000 ft., Wet Canyon) 
(july); Maricopa Co.1 2? (Reef); Navajo Co..1 2 (Carrizo Cr.) (June); 
Pima Co., 24 6,72 2 (Santa Catalina Mts.: mile posts Nos. 9, 10, 23 Hitch- 
cock Highway, Pepper & Sauce Canyon, Sabino Canyon, Sabino Basin ca. 
3800 ft., Ventana Canyon, Catalina Springs, Molino Basin; Santa Rita Mts. 
4000-8000 ft.; Baboquivari Mts.: Kits Peak Rincon ca. 1050 ft., Brown’s Can- 
yon) (February, April-November); Pinal Co..1 8,1 2 (Superior) (Feb- 
ruary, July); Santa Cruz Co. 2 6,5 2 (Ruby, Sycamore Canyon; Pata- 
gonia; 17 mi. W. Nogales) (March-May, August, November); Yavapai 
Ge! 2 G mi. S. Jerome) (July); Yuma Co.,5 $,1 2 (Parker Creek; 
Sierra Ancha Exper. Sta.) (May, August). NEW MEXICO: Catron Co., 
1 2 (Mogollon Mts.) (August); Dona Ana Co.,2 6,28 2 (Las Cruces; 
Organ Mts.: La Cueva ca. 5300 ft., 5100 ft., Riley’s Ranch, Filmore Canyon 
5700 ft.. Dripping Springs) (April, June, August-September); Grant Co., 
26,1 2 (6 mi.N. Silver City; 14 mi. N. Silver City; Pinos Altos) (June- 
July); Otero Co.,1 2 (Alamogordo) (May). TEXAS: Brewster Co.1 é, 
6 2 (65 mi. S. Marathon; Basin 5000 ft., Big Bend Nat. Pk.; Chisos Mts.) 
(June-July); Jeff Davis Co.,11 6,7 2 (Davis Mts.; Ft. Davis; 23 mi. W. 
Ft. Davis) (April-July); Val Verde Co.1 8,6 2 (Devil’s River) (May). 

CHIHUAHUA: 5 ¢,4 2 (92 km. N. Chihuahua; 80 km. N. Chihua- 
hua; Terrero 5500 ft.; Valle de Olivos 5500 ft.) (May-July). DURANGO: 
1¢6,5 @ (12 mi. N. Alamillo; Nombre de Dids; El Tascate 6400 ft.; 69 mi. 
N. Durango, Hwy. 31; 12 mi. W. Durango) (February, June-July). GUER- 
RERO: 7 2 (2 mi. S. Chilpancingo; 5.2 mi. E. Chilpancingo 5700 ft.; 5 mi. 
S., 2 mi. E. Chilpancingo 3800 ft.; Chilpancingo 4000 ft., 3700 ft.) (August). 
EWDALGO: 1 6, 21 ¢ (23 mi. NE. Jacala 5075 ft.; 38 mi. NE. Jacala 
3100 ft.; Zimapan; 6 mi. E. Tulancingo; 45 mi. W. Pachuca) (June-Septem- 


562 Tue University ScrENCE BULLETIN 


ber). JALISCO: 5 9 (22 mi. NW. La Piedad; 15.5 mi. NE. Lagos de 
Moreno 6200 ft.; 6 mi. SE. Lagos de Moreno 5900 ft.) (July). MEXICO: 3 


@ (Teotihuacan Pyramid; Valle de Bravo 6500 ft.) (June, August). MI- 
CHOACAN: 1 28, 3 2 (Quiroga 6300 ft.; Morelia) (February, July). 
MORELOS: 1 ¢,5 2 (Cuernavaca 5500 ft.; 5 mi. S. Cuernavaca 4000 ft.; 
3 mi. W. Cuernavaca 6500 ft.; Yautepec 4000 ft.) (March-May). NUEVO 
LEON: 1 6,7 2 (4 mi. W. El Cercado 2100 ft.; 12 mi. S. Linares; China) 
(June, August, December). OAXACA: 3 8,6 2 (7 mi. SE. El Camerén; 
Tehuantepec; 5 mi. E. Oaxaca; 10 mi. NE. Oaxaca; Oaxaca; 23 mi. Nie 
Nochixtlan 7000 ft.; 12 mi. SE. Nochixtlan 7100 ft.) (April, June, July, De- 
cember). PUEBLA: 3 2 (2 mi. NW. Petlalcingo 4600 ft.; 8 mi. SE. Te- 
huitzingo 4100 ft.; 13 mi. E. Villa Juarez 1300 ft.) June). QUERETARO: 
3 2 (Queretaro) (June). SAN LUIS POTOSI: 2 ¢, 30 2 (EI Salto 1500- 


1800 ft.; 5 mi. E. Ciudad Maiz 4700 ft.; 3.4 mi. NE. El Naranjo 800 ft.; 5 mi. 


W. Xilitla; 14 mi. W. Xilitla 4200 ft.; 8 mi. W. Xilitla 3500 ft.; 4.3 mi. NW. 


Nuevo Morelos; El Huizache; 10 mi. NE. San Luis Potosi 6200 ft.; 29 mi. 


SW. San Luis Potosi 6800 ft.) (June-September). TAMAULIPAS: 2 9 
(Llera, Victoria) (March, June). TLAXCALA: 1 2 (8 mi. W. Apizaco 
8500 ft.) (June). VERACRUZ: 3 2 (4 mi. NW. Rinconada Antigua 1350 
ft.) (June). ZACATECAS: 1 6, 4 9<(5 km. E. Sombrerete; 2 mig 
Fresnillo; 9 mi. S. Fresnillo; Fresnillo 7000 ft.) (July-August). 
(GUATEMALA. Alta Verapaz; 12 (Txece Aguas): 

PANAMA. Canal Zone, 1 é (Fort Clayton) (May). 

This species is widespread in the central plateau area of Mexico. The 
range extends northward through the Chihuahua desert to New Mexico and 
Arizona. It also extends southward at least as far as Panama. Augochlorella 
n. neglectula is not known from along the coasts of Mexico, although it ap- 
proaches them in Oaxaca and Veracruz. It has been taken at elevations from 
800 to 8500 feet in Mexico and from 1050 to 8100 feet in Arizona and New 
Mexico. (See also discussion under pomoniella.) 

Titus (1901) reported two female specimens of neglectula from Ft. Col- 
lins and Greeley, Colorado. One specimen from each locality (8) with dates 
matching those given by Titus has been examined but both specimens were 
normal Colorado striata p. All other specimens examined from Colorado 
have also been A. striata and therefore it is assumed that Titus was mistaken 
in his identification and that neglectula extends only as far north as Catron 
Co., New Mexico, and Yavapai Co., Arizona. 

Dreisbach (1945) reports neglectula from Michigan, but this is clearly a 
case of confusion in synonymy. (See distribution of persimilis.) 

Seasonal Activity. Females have been collected from mid-February to 
the end of November in Arizona and to the end of December in Mexico. 
Males have been taken from March through November in the United States 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 563 


and all through the year in Mexico. Females with pollen in their scopas were 
collected from April through August in the United States and from Febru- 
ary through December in Mexico. 

It seems probable that there is little or no activity during December and 
January in the United States and that nesting takes place from the middle of 
April into September or October, depending on the elevation. In Mexico, the 
bees are apparently active and nesting throughout the year, at least at the 
lower elevations. 

Flower Records. Acacia, Aesculus, Baccharis, Ceanothus, Chilopsis, 
Chrysanthemum, Chrysopsis, Dalea, Descurainia, Echinocactus, Erigeron, 
Eschscholtzia, Fendlera, Gaillardia, Gossypium, Gutierrezia, Helenium, 
Heterotheca, Lepidium, Manzanita, Melilotus, Opuntia, Penstemon, Prunus, 
Pyrus, Rosa, Senecio, Sida, Sisymbrium, Sphaeralcea, Ungnadia. 


564 THe University SciENcCE BULLETIN 


Augochlorella neglectula maritima new subspecies 


Types. 22 males, 30 females (holotype male, allotype female and para-_ 
types), 20 mi. E. Acapulco, Guerrero, Mexico, August 12, 1962 (Univ. Kansas 
Mexican Exped.). Additional paratypes as follows: MEXICO. Guerrero: 3 
females, 5 males, 20 mi. E. Acapulco, 11 August 1962 (Univ. Kansas Mexican 
Exped.); 1 female, 20 mi. E. Acapulco, 11 August 1962 on Microspermum 
EEDA, (E. Ordway); 6 females, Acapulco, 6 August 1954 
(Univ. Kansas Mexican Exped.); Nayarit: 1 male, San Blas, 13 September 
1957 (R. & K. Dreisbach). 

Holotype, allotype and 31 female and 19 male paratypes are in the Snow 
Entomological Museum, The University of Kansas. Two female and two_ 
male paratypes are in each of the following collections: Michigan State Uni- 
versity, American Museum of Natural History, U.S. National Museum and 
the California Academy of Sciences. 

Description. his subspecies, known only from the Pacific Coast of 
Mexico, differs from the true neglectula only as follows: 

Female: (1) Length of head greater than width (width 1.66 to 1.96 mm, 
averaging 1.81 mm). (2) Frons without bluish reflections. (3) Mandible 
with basal half brown. (4) Clypeus as long as broad with apical third brown. 
(8) Scutum finely and very closely punctured, punctures becoming indistinct — 
on extreme anterolateral angles. (9) Tegula oblong, twice as long as wide; 
surface coarsely punctate and dull on dark brown area. (12) Edge of propo- 
deal disc more angulate than in typical neglectula. (13) Legs dark brown 
with weak greenish reflections usually on fore and hind coxae, fore tro- 
chanter, fore and middle femora and fore tibia. 

Male: (1) Length of head greater than width (width 1.52 to 1.84 mm, 
averaging 1.70 mm). (2) Color bright green, metasoma suffused with black; 
frons without bluish reflections. (7) Antenna with first flagellar segment 
entirely dark brown. (9) Tegula oblong, more than twice as long as wide, 
with distinct, close punctures on dark brown area; surface dull. (11) Pleuron 
coarsely punctured, becoming rugose laterally. (16) Genital capsule with 
inner lobe of gonostylus heavily sclerotized with straight outer edge and 
thick, coarse, long bristles to apex, without gap or finer bristles separating 
outer and apical series of setae (Fig. 39). 

Comparisons. A.n. maritima looks similar to neglectula but can be readi- 
ly distinguished by the size, shape, punctations and dullness of the tegulae. 
In addition, it lacks the blue areas on the frons and is less rugose on the 
face, thorax and posterior surface of the propodeum. The general appearance 
of the propodeum is smoother than in n. neglectula but otherwise little dif 
ferent. The first flagellar segment of the male is entirely dark in maritima 
and the inner lobe of the gonostylus of the male genitalia is more heavily 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico — 569 


sclerotized and bears stronger setae than most n. neglectula, although inter- 
gradations exist. The oculo-ocellar ratios show that the top of the head is 
slightly narrower than that of neglectula and the head is longer in relation 
to the width. The overall impression is of a longer head with the eyes nar- 
rower and less emarginate, and the clypeus longer than in n. neglectula. 
The habitat of maritima appears to be distinct from that of n. neglectula; 
the former has so far only been found on the sand dune areas along beaches 
of the west coast of Mexico (circles, Fig. 87). 


Augochlorella edentata Michener 


Augochlorella edentata Michener, 1954, Bull. Amer. Mus. Nat. Hist. 104:58-59 
(descr.). 


Types. Holotype female, from Panama, Coclé Province, El Valle de 
Anton, April 1, 1945 (Michener), is in the American Museum of Natural 
History. One female paratype each: type locality, January, 1947 (Krauss), 
at The University of Kansas; Canal Zone, Summit, November, 1946 
(Krauss); Balboa, May 25, 1914, in shady jungle (Hallinan). 

Description. Female: (1) Length 6 to 7 mm; head width 1.55 to 1.87 
mm, averaging 1.66 mm, width greater than length. (2) Color yellow-green 
to dark green with silky sheen or luster; frons without blue reflections; meta- 
soma usually similar in color to head and thorax. (3) Mandible dark brown, 
becoming yellow-brown just before rufous tip, without greenish reflections at 
base. (4) Clypeus about as long as wide, almost flat; basal half green with 
distinct but shallow punctures, widely but uniformly spaced, usually finely 
and closely punctured or roughened along basal suture; apical half brown, 
very slightly beveled, with elongate, shallow, groove-like punctures separated 
by about their own diameters; surface between punctures shiny and smooth. 
(5) Face broadly convex from apex of clypeus to vertex; supraclypeal area 
flat to weakly rounded, smooth and shiny to minutely roughened with or 
without punctures. (6) Paraocular area finely and uniformly roughened 
throughout by widely scattered minute granules. (7) Antenna dark brown, 
becoming yellow-brown at tip; last flagellar segment entirely yellow-brown; 
flagellum lighter below than above; pedicel width subequal to length, first 
flagellar segment almost twice as wide as long; pedicel longer than and 
about as wide as first flagellar segment. (8) Scutum without distinct punc- 
tures, surface finely and irregularly roughened throughout (Fig. 79); an- 
terior margin smooth medially to anterolateral angles, except for minute 
reticulations on surface. (9) Tegula about 1.33 times as long as wide. (10) 
Scutellum finely and irregularly roughened, without punctures. (11) Pleu- 
ron finely rugose, weakly areolate anteriorly. (12) Propodeum with disc al- 
most twice as long as metanotum; outline of disc semicircular to broadly 


566 Tue UNiversiry ScIENCE BULLETIN 


U-shaped, profile type 4; posterior edge of disc smoothly and gradually | 
rounded medially and laterally; striae fine and close, irregular and branch- 
ing medially, straight and distinct laterally, occupying basal half to two- | 
thirds of disc medially, sometimes attaining edge laterally; area beyond striae 
minutely reticulated; horizontal area of propodeum only partially occupied 
by disc laterally and at posterolateral corners; posterior vertical surface ex-_ 
ceedingly finely and evenly granular in larger specimens, smooth and shiny 
with widely spaced minute punctures in smaller specimens; posterolateral - 
corners finely roughened to nearly smooth; lateral vertical surface weakly 
roughened. (13) Legs brown with slight metallic reflections on hind coxa | 
only. (14) Metasomal terga dark green to golden-green, suffused with 
brownish in some specimens; first tergum with anterior surface polished 
with widely scattered punctures laterally; dorsal surface with minute punc-_ 
tures evenly spaced; second tergum with minute punctures, more crowded | 
than on first; first sternum without metallic reflections. (15) Pubescence 
white to golden-white dorsally on head and thorax, dorsally and ventrally on 
metasoma; white ventrally on head, thorax and basal parts of legs; golden — 
on tibiae and tarsi. | 
Male: (1) Length 6 mm; head width 1.48 to 1.63 mm, averaging 1.56 
mm, width slightly greater than length. (2) Color olive-green with yellow- 
green sheen, varying to yellow-green with golden sheen in some specimens; | 
frons without bluish reflections; metasoma dark brown. (3) Mandible yel- 
low-brown, only slightly rufous at tip, basal condyles usually dark brown; 
basal metallic reflections absent. (4) Clypeus with punctures shallow and 
widely spaced, sometimes closer basally; surface between punctures brightly 
polished. (5) Face broadly convex from apex of clypeus to vertex; supra- 
clypeal area scarcely protuberant, impunctate but variably roughened. (6) 
Paraocular area smooth and polished below level of antenna with a few, 
widely spaced, minute punctures, becoming rougher toward frons; frons 
with dense mat of short white pubescence extending from antennal sockets 
to vertex. (7) Flagellum slightly darker above than below; scape and pedicel 
dark brown; last one and a half to two flagellar segments uniformly dark 


brown; pedicel and first segment each less than twice as wide as long. (8) 
Scutum shiny with satiny luster; punctures very weak, sparse and widely 
spaced (wider than own diameters), minute centrally, larger and closer 
laterally, becoming slightly deeper and contiguous to weakly and finely 
punctorugose anterolaterally; anterior margin smoothest medially, becoming 
rougher laterally; sculpturing exceedingly fine, shallow and indistinct (Fig. 
79). (9) Tegula twice as long as wide. (10) Scutellum shiny, slightly rough- 
ened and punctate, punctures weak and widely spaced. (11) Pleuron shal-| 
lowly and irregularly punctorugose, shallowly areolate anteriorly; surface 
between impression minutely roughened, dull. (12) Propodeum with disc 


SYSTEMATICS OF THE GENUs Augochlorella NortuH of Mexico — 567 


about 1.75 times as long as metanotum; outline of disc U-shaped, posterior 
edge indistinct, smoothly and gradually rounded and shiny; striae distinct, 
fine, close, irregular, often branched, extending about three-fourths length of 
disc medially, reaching edge laterally; surface of disc beyond striae smooth, 
minutely reticulated, irregularly and minutely wrinkled; posterior vertical 
surface of propodeum and posterolateral corners shiny and smooth with 
widely spaced, minute punctures; lateral vertical surface of propodeum fine- 
ly but regularly roughened becoming distinctly but shallowly punctate dor- 
sally. (13) Legs dark brown; fore and hind coxae and femora with slight 
greenish reflections; tibiae and tarsi lighter brown at bases; hind basitarsus 
with erect hairs short, uniform in length, dense, slightly longer than width 
of segment; basal tuft indistinct. (14) Metasomal terga dark brown with 
slight bluish highlights and broad, reddish brown apical margins; first ter- 
gum polished anteriorly, with slight olive-green metallic reflections, smooth 
but less shiny dorsally, sparsely and minutely punctate; second and following 
terga smooth, dull, essentially impunctate but with widely scattered minute 
punctures; all sterna brown, finely and uniformly pubescent, without metal- 
lic reflections, with apical margins straight; surfaces smooth without reticu- 
lations or punctures. (15) Pubescence golden dorsally, white ventrally; face 
with mat of short, dense, white hair from antennae to vertex and longer, 
fine, golden hair dispersed over entire face; pubescence golden dorsally on 
thorax, entire metasoma and legs, white ventrally on head and thorax. (16) 
Genital capsule, seventh and eighth sterna and eighth tergum type 3 (Figs. 
ap) 42, 47). 

Comparisons. This species has certain affinities with both the species of 
the north and those of the Pereirapis group. It is small, has the convex face 
and the similar fine, smooth sculpturing and discal shape, and has the 
straight margin of the fourth sternum in the male, all as in Perezrapis. But 
the genital capsule and the white apex of the clypeus of the male are more 
similar to those of the North American species than to Pereirapis. A. eden- 
tata is characterized by its round, convex face (Fig. 55), comparatively 
smooth, fine and very shallow body sculpturing, and usually by the yellow 
tipped antennae in the female. The clypeus is also flatter than in other North 
American species but comparative material is necessary to recognize this 
character. 

Variation. The body color in both males and females varies from dark 
green to yellow-green. Although the bee is shiny, the color is dull, with 
olive-green tones. In females, the metasoma is green or only slightly suffused 
with brown in some of the darker specimens, or yellower green in lighter 
colored individuals. The metasoma is consistently brown in males. 

There is greater variation of the head width to length ratio in females than 


568 THe University SciENCE BULLETIN 


in males. (However, the available males are nearly all from the same popula- 
tion.) Both the length and width are variable in females. | 

The clypeal punctures are generally rather uniform in males but variable 
in size, number and spacing in females. The basal punctures range from 
absent to crowded in both sexes. 

The tips of the antennae of females are brownish yellow in all specimens 
examined from Mexico and Costa Rica, the intensity of the yellow varying | 
only slightly. However, in the holotype the antennal tips are brown, only 
inconspicuously paler than the rest of the flagellum. The one paratype seen, 
also from the type locality, has the antennal tips yellower than the holotype 
but darker than those from Mexico and Costa Rica. More specimens should 
be examined to see if this is individual or geographical variation. 

There is slight variation in the quality of the roughening of the scutum 
among females. On about half the specimens the smoothness of the antero- 
medial edge extends down each side of the median suture, creating a trans- 
verse gradient in degree of roughness from the center line to the parapsidal 
lines. However, the anterolateral angles are no more roughened than other 
parts of the scutum in either sex. 

The punctures on the scutellum are variable in size, number, spacing and 
depth among males. 

There is little variation in the propodeal area of males, but in females the 
length of the striae is variable, extending more than two-thirds the length of 
the disc in only one specimen |“Rin Antonio” (9*) |. In this specimen, only 
a small area is unstriated medially, and the lateral striae are strong, ending 
abruptly at the edge. Although lateral striae usually reach the edge, they 
my be shorter in the larger specimens. The sloping horizontal area of the 
propodeum is usually only partially occupied by the disc laterally, the rest of 
the horizontal area being characterized by sculpturing similar to that of the 
vertical surface (Fig. 67). The posterior vertical surface in females varies 
from very smooth, broken only by widely scattered minute punctures, to 
weakly uneven or slightly roughened. There is also some variation in 
amount of roughening of posterolateral angles. There is little variation in 
the lateral vertical surface. 

The metasomal punctures are similar in all specimens; however, in fe- 
males they may vary somewhat in depth and are barely detectable in some 
individuals. 

The color of the pubescence is difficult to determine in most specimens of 
both sexes, usually because of wear or dirt. Some appear lighter than others 
so that the dorsal pubescence appears to range from all white to golden. The 
short, dense facial pubescence is matted and gummy in most specimens, giv- 


* See Table 1. 


SysTEMATICS OF THE GENUS Augochlorella Nortu oF Mexico — 569 


ing the face a yellowish cast. This facial hair is worn away on some speci- 
mens. 

Distribution. This species has been taken only as far north as southern 
San Luis Potosi in Mexico. It ranges at least as far south as Panama and has 
been taken at elevations from 200 to 4500 feet (Map: Fig. 88). 

The following specimens have been seen: MEXICO. MORELOS: 1 4, 
22 (3 mi. N. Alpuyeca 3400 ft.; S. end Cuernavaca 4500 ft.) (March-April) ; 
OAXACA: 1 2 (Rin Antonio); SAN LUIS POTOSI: 12 6,4 ¢ (Hui- 
chihuayan; Xilitla 1450 ft.; 5 mi. E. Xilitla) (July, September); VWERA- 
RUZ: 1 3,6 2 (17 mi. NW. San Andrés Tuxtla 900 ft.; 6 mi. N. Jesus 
Carranza, Isthmus of Tehuantepec 200 ft.; 1.4 mi. N. Santiago Tuxtla 1150 
ft.; Cordoba) (January, June). 

GUATEMALA. ALTA VERAPAZ, 2 2 (Trece Aguas) (June). 


Fic. 88. Distribution of A. aurata (solid dots) and edentata (rings). 


570 Tue University Science BULLETIN | 

i 
EL SALVADOR. LA LIBERTAD, 6 2 (5 mi. W. Quezaltepeque) 
(June-August). 

COSTA RICA. CARTAGO: 3 6,1 2 (Turrialba) (August); f 
MON: 2 6, 12 2 (Pandora) (August); PUNTARENAS: 1 ¢, 1474m 
(6 mi. NE. Esparta; Gromaco 34 km. SE. Potrero Grande, Rio Coto Brus — 
1000 ft.; Playén, 8 mi. N. Parrita 30 m) (June-July, December); SAN / 
JOSE: 2 6,5 2 (Pozo Azul, junc. Rios Parrita and Candelaria 85 m) | 
(August). | 

PANAMA. See localities for type material above. 

Seasonal Activity. Females of edentata are apparently active throughout 
the year; males have been collected from March through August but are — 
probably active in other months as well. Pollen collectors have been found © 
during June and July in Mexico and in August and December (1 specimen) 
in Costa Rica. 


7 
; 


Augochlorella bracteata new species 


Augochlorella aurata: Sandhouse, 1937, Jour. Washington Acad. Sci. 27:70 (in 
part) (taxon.). 


Types. Holotype female, Southmost, Cameron Co., Texas, March 27, 
1951 (R. H. Beamer) taken on Prosopis; allotype male, same locality, April 
13, 1950 (Beamer, Stephen, Michener, Rozen); 45 female paratypes, same 
locality, March 27, 1951 (Beamer); 26 female paratypes, same locality and 
date (Michener); 4 male paratypes, Brownsville, Texas, June; 2 male para- 
types, Cameron Co., Texas, August 3, 1928 (J. G. Shaw); 1 male paratype, 
Brownsville, Texas, April 17, 1952 (Michener, Beamer, Wille, LaBerge). 

Holotype, allotype, 55 female and 3 male paratypes are in the Snow Ento- 
mological Museum of The University of Kansas; one male and four female 
paratypes are in each of the following: the American Museum of Natural 
History, the U.S. National Museum, the University of Nebraska and the 
California Academy of Sciences. I have seen 14 males and 137 females from 
other localities in Texas and Mexico, not included in the type series. The 
holotype and female paratypes represent a series of large and uniformly 
similar bees, probably spring queens. Not included in this series are smaller 
and more morphologically diverse specimens, including probable worker- 
like individuals. 

Description. Female: (1) Length 5 to 6 mm; head with 1.47 to 1.79 mm 
(holotype=1.79 mm), averaging 1.63 m, width equal to, or slightly greater 
than length. (2) Color yellow-green to dark green (bright green in holo- 
type); frons without bluish reflection; metasoma often slightly browner than 
rest of body. (3) Mandible with basal third dark brown, yellowish brown 
centrally, rufous at tip, without green basal reflections. (4) Clypeal width 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico — 571 


equal to or subequal to length (length slightly greater than width in holo- 
type), basal two-thirds green, punctures variable, irregularly spaced, smaller 
basally than apically; apical third brown and slightly beveled with large, 
deep punctures; surface between punctures shiny and smooth. (5) Supra- 
clypeal area sparsely punctured medially, becoming densely punctured pe- 
ripherally; surface smooth to minutely reticulated. (6) Paraocular area with 
large contiguous punctures below antenna, finely rugose above antenna. 
(7) Antenna dark brown; flagellum usually lighter below than above; pedi- 
cel longer than wide; first flagellar segment slightly wider than long; pedicel 
longer and narrower than first segment. (8) Scutum smooth, finely and 
uniformly punctured throughout, punctures small, close, distinct, extending 
almost to anterior edge; anterior edge with surface slightly roughened medi- 
ally as on vertex becoming slightly rougher laterally to very finely rugose at 
anterolateral angles. (9) Tegula 1.5 to 2.0 times as long as wide. (10) Scu- 
tellum rough with fine, shallow, irregular to indistinct punctures. (11) Pleu- 
ron punctorugose to shallowly and finely rugose (finely punctorugose in 
holotype), areolate anteriorly. (12) Propodeum with disc equal to or slightly 
longer or shorter than metanotum (slightly longer in holotype); outline of 
disc semicircular, profile type 2, posterior edge of disc abruptly rounded, in- 
distinct, more gradually rounded laterally; striae irregular, vermiform or 
straight (rather straight in holotype), fine and close together, usually ending 
almost at edge of disc medially, straight and reaching edge but not crossing 
it laterally; surface at ends of striae minutely roughened, narrowly shiny; 
posterior vertical surface and posterolateral corners shiny but surface un- 
even, without punctures or coarse roughening, often weakly granular posteri- 
orly and finely granular at posterolateral corners; lateral surface moderately 
roughened to weakly rugose (weakly rugose in holotype), without basal 
subhorizontal rugae. (13) Legs brown, fore and hind coxae with strong 
green reflections, femora and hind trochanter with weak metallic reflections. 
(14) Metasomal terga green, suffused with brownish; apical margins nar- 
rowly pale brown; first tergum shiny, polished, with numerous, fine, distinct 
punctures anteriorly, almost impunctate along narrow median longitudinal 
area, smooth and shiny dorsally with numerous, small, close, distinct, regu- 
larly spaced punctures; second tergum with punctures more numerous and 
smaller; first sternum without metallic reflections. (15) Pubescence golden- 
white dorsally, white ventrally on head, thorax and basal parts of legs, golden 
on leg extremities and ventral part of metasoma. 

Male: (1) Length 5 to 6 mm; head width 1.53 to 1.76 mm, averaging 1.63 
mm, width less than, equal to, or greater than length (width equal to length 
in allotype). (2) Color yellowish green to dark green; frons without bluish 
reflections; metasoma variably suffused with brownish. (3) Mandible with- 
out metallic reflections basally. (4) Clypeus with punctures large medially, 


) 
small and close basally and laterally; surface between punctures usually — 
smooth and shiny. (5) Supraclypeal area finely punctate laterally; surface | 


shiny basally, usually minutely roughened above. (6) Paraocular area finely 
and weakly rugosopunctate. (7) Antenna dark brown above, dark yellow 


SY? Tue Universiry SciENcE BULLETIN 


below; scape usually narrowly yellow below; last flagellar segment entirely — 
pale brown to dark brown, pedicel and first flagellar segment each less than 
15 times as wide as long. (8) Scutum shiny and smooth with punctures | 
small, distinct, usually separated by at least the width of a puncture medially, 
closer laterally and anteromedially; anterior margin weakly roughened as on 
vertex, becoming weakly rugose at extreme anterolateral angles. (9) Tegula 
slightly more than 1.5 times as long as wide. (10) Scutellum shiny and 
punctate; punctures distinct, closer posteriorly than elsewhere. (11) Pleuron 
rugosely punctate, areolate anteriorly. (12) Propodeum with disc slightly 
longer than metanotum; outline of disc obtusely V- or U-shaped to semi- 
circular (semicircular in allotype), posterior edge sharply angulate to gradu- 
ally rounded (abruptly rounded in allotype); striae fine to coarse, regular to | 


irregular (fine and irregular in allotype), reaching edge medially, crossing 
edge laterally; marginal surface narrowly shiny and slightly roughened pos- 
teriorly, shiny posterolaterally; posterior vertical surface either with widely 
separated shallow punctures and surface between punctures shiny, or weakly 
roughened and impunctate; posterolateral corners evenly punctate; lateral — 
vertical surface more closely punctate with punctures usually distinct, regu- 
lar, separated by about the width of a puncture; surface between punctures 
finely roughened or entire surface weakly rugosopunctate (punctate in type). 
(13) Legs light brown; fore and hind coxae, trochanters and femora with 
metallic reflections; tibiae light brown, yellow-brown at apices; tarsi uni- 
formly pale yellow-brown; hind basitarsus with erect hairs of uniform 
length, about as long as width of basitarsus, basal tuft slightly shorter. (14) 
Metasomal terga dark green, usually suffused with brownish, yellowish or 
reddish with apical margins brownish, scarcely contrasting in color with rest 
of tergum; first tergum polished anteriorly with small, widely spaced punc- 
tures throughout; smooth but less shiny dorsally, with punctures variable, 
usually large, distinct and irregularly spaced; second tergum with punctures 
of same size but very closely spaced; sterna dark brown, smooth, with short 
inconspicuous white pubescence; first sternum with weak metallic reflec- 
tions; fourth sternum broadly and shallowly emarginate. (15) Pubescence 
short and white over entire body. (16) Genital capsule, seventh and eighth 
sterna and eighth tergum of type 1 (Figs. 31, 40, 43). 

Comparisons. The females of this species look most like persimilis or the 
striata-persimilis integrades. The size is small, at most the size of persimilis. 
The female can be distinguished from the other species found in the United 
States by its small size, the small but distinctly separated punctures on the 


SysTEMATIcs OF THE GENUs Augochlorella Nortu oF Mexico — 573 


scutum, and the lack of coarse roughening or rugosity anteriorly on the 
scutum (Fig. 78). It is further separated from gratiosa and neglectula by 
the very smooth posterior surface of the propodeum. Although this is also 
consistently smoother than aurata, persimilts or striata, the differences among 
these species in the posterior surface of propodeum are usually too slight to 
be recognized without comparative material. 

The male is about the size and color of the male of persimilis but has the 
characteristics of a small striata with an emarginate fourth sternum, short 
basitarsal hair, and with the last segment of the antenna dark. The male 
genitalia are not perceptibly different from those of other members of the 
eastern species group, although there is a tendency for the process of the 
inner lobe of the gonostylus to be shorter and blunter than in other species 
(Fig. 31). Like the female, the male can be distinguished from all forms of 
striata by both its small size and its smooth but distinctly punctured scutum. 

Variation. The greatest variation in bracteata is found in the characters 
of the propodeum. Males are in general more variable than females, the 
females varying chiefly in connection with differences in size. There is little 
geographical variation, probably due to the limited range and to the few 
specimens available outside of Texas. 

As with most of the eastern species of Augochlorella, there is considerably 
more size variation in females than in males. This may be due to caste differ- 
ences. In females the width of the head is usually greater than the length, 
but sometimes the width and length are equal. No such general pattern can 
be established for the males since the length-width ratio is highly variable, 
even though the overall variation in size is not great. 

Color varies from dark green to yellow-green in both males and females, 
with no apparent correlation between size, date or location, There seems to 
be an unusual amount of discoloration, fading or bleaching in many of the 
specimens, especially on the metasoma. Among males, many of the speci- 
mens are coppery or reddish (see section on Specific Characters). In spite of 
the large proportion of specimens thus discolored, freshly caught specimens 
are probably normally green. 

The number and size of punctures on the clypeus of the female is vari- 
able. When the punctures are sparse, the clypeus looks smooth, shiny and 
gently rounded, with the apex also smooth. This condition is apparent 
especially in the smaller (worker?) individuals. There is little variation in 
the size and density of punctures in males. 

The supraclypeal area in both males and females is usually punctate, with 
the surface between punctures variably roughened. A female with a sparsely 
punctate clypeus will usually have a sparsely punctate supraclypeal area, with 
the central portion smooth and shiny. In the male the supraclypeal area may 


574 Tue UNiversity ScIENcCE BULLETIN 


be shiny with few punctures (149% of the specimens) or with the upper half 
rough and lower half shiny (67°94) or entirely rough (19%). 


There is little variation in the punctation of the thorax except as related — 


to the size of the individual bee. On particularly small females the punc- 


tures are exceeding small and close and may give the scutum the appearance | 


of being granular rather than punctate. 

The rugosity of the mesepisternum also becomes very fine on small fe- 
males so that the surface may look similar to that of the metepisternum. 

The disc is the most variable structure in this species. Although the edge 
is rounded in both males and females, it may be smooth and shiny, minutely 
or weakly roughened, or smooth with minute reticulations. In males the 
punctures of the posterior vertical surface of the propodeum may reach the 
dorsal part of the edge on some shiny specimens, and the edge may be more 
or less sharply defined (but never carinate) with shininess often associated 


with a rounded edge and dullness with a more distinct edge. A V-shaped © 
depression is rarely apparent but may be indicated on those specimens with — 


a relatively defined edge. 

The shape of the disc shows little variation in females. In all cases it is 
semicircular, like that found in persimilis or in the persimilis-striata inter- 
mediates. There is somewhat more variation in males, with the outline vary- 
ing from semicircular to roundly V-shaped, but not bracket-shaped as in 
striata form a. The striae may be of any thickness from very fine to coarse, 
especially in males, and may vary from straight to vermiform or, as in some 
small females, may be so irregular as to be unrecognizable as striae. Striae 
usually reach the edge of the disc, at least medially, where they are usually 
branched or irregular. In both sexes they are frequently slightly shorter on 
each side of the median line, where the edge then becomes thicker (see Fig. 
66). The striae are straighter and more regular in the larger females than in 
smaller ones. There is as much variation in the characters of the disc as in 
persimilis and its intermediates, with some specimens resembling the small 
striata form c and others resembling persimilts. 

The posterior surface of the propodeum is smoother in some females than 
others but never equals the smoothness of the anterior part of the first ter- 
gum. The roughening takes the form of fine granulations or irregularities 
on an otherwise smooth surface. 

The lateral vertical surface of the propodeum ranges from rugose to 
finely roughened in the females and distinctly and evenly punctured to weak- 
ly roughened, largely punctorugose or finely reticulate, in the males. 

The metasomal punctures are variable in both sexes, although the type 
series of 72 females, all collected at the same time and place, are similar in 
this feature. The punctures on the first tergum in females may be absent in 
some small specimens or small and irregularly spaced to large, distinct, and 


SYSTEMATICS OF THE GENUs Augochlorella Nortu of Mexico — 575 


closely spaced. There is little indication of regional variation, although most 
specimens with large, close punctures are from southern Texas. The four 
specimens from Mexico have small to minute punctures. In males the punc- 
tures of the first tergum vary from very small to large, close to widely spaced, 
with no regional pattern evident. The second tergum of females usually has 
very small punctures regardless of the size of those of the first tergum. In 
males both the size and the spacing of punctures on the second tergum are 
variable, but usually the punctures are smaller than on the first. The third 
tergum in males sometimes has small but distinct punctures. 

There are a few females with all white pubescence over the body. These 
are usually discolored individuals with brown metasoma (Victoria, Kings- 
mille, etc.). 

Distribution. From northeastern to southern Texas, southward through 
eastern Mexico to Hidalgo (Map: Fig. 85). 

In addition to the type series, 14 males and 137 females were seen: TEX- 
AS: Bexar Co..5 2 (March, July-August); Blanco Co.,2 23; Cameron 
Co. 6 6, 33 2 (January-August, September); Comal Co. 1 2 (May); 
Dallas Co. 8 2 (March-May); Hidalgo Co..1 4,4 2 (April-June); Jack- 
son Co.,2 2 (March); Karnes Co.,6 2 (March, September); Kleberg Co., 
1 2 (June); Lee Co. 2 6,20 2 (February-June, September); Mavarick 
mors °. (April); Nacogdoches Co, 1 2 (June); rNueces Co. 1 9 
(April); Refugio Co, 1 2 (April); Robertson Co. 15 2 (April); San 
Szricio Co. 3 2 (June-July); Pravis Co..1 9°; Val Verde Co,1 46,1 ° 
(May-June); Victoria Co.,1 6, 20 2 (March-April, September, Novem- 
ber); Webb Co..6 2 (December); Wilson Co. 1 2 (October). 

MIDALGO: 1 ¢,3 2 (18 mi. NE. Jacala 4750 ft.) Qune). NUEVO 
LEON: 1 @ (General Terdn) (July). 

Seasonal Activity. The females are apparently active throughout the 
year, and the males are present at least from March through November. 
None of the 30 females taken between October and February was collecting 
pollen, and most of these specimens were clean and unworn, although both 
larger and smaller individuals were represented. Of the 90 specimens taken 
in March, the only 5 collecting pollen were captured on or after the 24th of 
the month. From April through September many specimens have pollen in 
the scopa; one was taken with a pollen load as late at September 29. 

From these data it would seem that although the bees are active through- 
out the year, nesting takes place only from early spring into September. 
From variations in size of females it also seems that this species probably does 
have a worker caste, as do striata and persimilis (see Ordway, 1965a). 

Flower Records. Agastache, Callirhoe, Chamaesaracha, Coreopsis, Dalea, 
Englemannia, Haplopappus, Helenium, Helianthus, Monarda, Oenothera, 
Opuntia, Phacelia, Prosopis, Pyrrhopappus, Ratibida, Rubus. 


576 Tue UNIversity SCIENCE BULLETIN 


Augochlorella gratiosa (Smith) 


Augochlora gratiosa Smith, 1853, Catalogue of the Hymenoptera in the British 
Museum 1:80 (descr.); Dalla Torre, 1896, Catalogus Hymenoptorum 10:95 
(list); Bingham, 1897, Trans. Amer. Ent. Soc. 24:162 (list); Cockerell, 1905, 
Trans. Amer) Ent. Soc. 312363) (list): 


Augochlora (Augochlorella) gratiosa: Michener, 1951, in Muesebeck et al. U.S. © 


Dept. Agr., Agr. Monogr. 2:1125 (list); Montgomery, 1957, Proc. Indiana 
Acad. Sci. 66:132 (list, fl.). 

Augochlora festiva: Graenicher (not Smith, 1853), 1930, Ann. Ent. Soc. Amer. 
3271574 list, fl:) 

Augochlorella gratiosa: Sandhouse, 1937, Jour. Washington Acad. Sci. 27:69 
(tax.); Lovell, 1942, Kentucky Acad. Sci. Trans. 10:20, 21, 23 (key, descr.), 
Michener, 1954, Amer. Mus. Nat. Hist. Bull. 104:55 (descr.); Mitchell, 1960, 
Bees of the Eastern United States 1:460 (tax.). 


Types. Augochlora gratiosa, holotype female, from Georgia, is in the | 
British Museum (Natural History). Although labels on the holotype agree | 
with those indicated by Smith’s published description of gratiosa, the de-— 
scription agrees better with the holotype of aurata. Conversely, his descrip- — 


tion of aurata fits best the type of gratiosa. All recent descriptions and most 
determinations of gratiosa agree with the gratiosa type and not with Smith’s 
description. Since Smith himself probably mixed labels and descriptions, 
since his descriptions are scarcely decisive, and since utmost confusion would 
result from reversing the application of the names in a group already so 
dificult taxonomically, it seems best to follow usage and the labeled types, 
which bear the proper locality data, and ignore the inconsistencies in the 
descriptions. I have not seen the types, but Dr. C. D. Michener took detailed 
notes on them and compared them with submitted specimens. 

Description. Female: (1) Length 6 to 7 mm; head width 1.68 to 2.13 
mm, averaging 1.92 mm; head length to width ratio variable. (2) Color 
yellowish green to blue; frons often with slight bluish reflections; metasoma 
usually more yellow or brownish than head and thorax. (3) Mandible with- 
out metallic coloration basally. (4) Clypeal width about equal to length, 
basal two-thirds shiny green with punctures irregular in size and shape, 
smaller and closer at margins of clypeus than centrally; apical third brown 
and slightly beveled with punctures large, elongate; surface between punc- 
tures smooth and shiny. (5) Supraclypeal area usually impunctate medially, 
lateral punctures variable in size, surface between punctures minutely rough- 
ened laterally. (6) Paraocular area punctorugose below level of antennae, 
coarsely rugose above. (7) Antenna dark brown; flagellum slightly lighter 
below than above; first segment of flagellum less than 1.5 times as wide as 
long, pedicel slightly longer and narrower than first flagellar segment with 
ratio of length to width variable. (8) Scutum roughly punctate to rugoso- 


SYSTEMATICS OF THE GENUS Augochlorella Nortu of Mexico — 577 


punctate; punctures small and contiguous; anterior margin rugose, becom- 
ing areolate at anterolateral corners. (9) Tegula almost 1.5 times as long as 
wide. (10) Scutellum coarsely roughened and granular, punctures indistinct 
or absent. (11) Pleuron coarsely rugose, areolate anteriorly. (12) Propodeum 
with disc equal to or shorter than metanotum; outline of disc bracket-shaped, 
profile type 1; posterior edge of disc sharply angulate posteriorly, rounded 
laterally; striae fine, straight, close together, extending full length of disc; 
posterior vertical surface finely, unevenly roughened or granular, usually 
dull, occasionally with very fine, indistinct, irregular, subhorizontal rugae; 
posterolateral corners coarsely and often linearly roughened; lateral vertical 
surface weakly rugose, rugae usually lineate along anterior and ventral edge, 
reticulate centrally. (13) Legs brown; fore and hind coxae and outer surface 
of fore femur with metallic reflections. (14) First metasomal tergum with 
anterior surface minutely and sparsely punctured, polished and finely pubes- 
cent; surface less shiny dorsally with close, minute punctures; second tergum 
with surface minutely reticulated, similar to following terga; sternum with- 
out metallic reflections. (15) Pubescence white ventrally on head and thorax, 
golden elsewhere; short and dense on genal area. 

Male: (1) Length 6 to 8 mm; head with 1.72 to 2.00 mm, averaging 1.86 
mm, greater than, equal to or less than length. (2) Color yellow-green to 
royal blue, usually bright, shiny green; frons without blue reflections on 
green specimens; metasoma usually slightly redder than rest of body or color 
uniform over entire body. (3) Mandible usually with metallic reflections 
basally. (4) Clypeus with large, irregularly shaped punctures separated by 
less than their diameters; surface between punctures smooth and shiny. 
(5) Supraclypeal area with punctures large, shallow and close; surface be- 
tween punctures minutely roughened. (6) Paraocular area with punctures 
small, deep and crowded. (7) Flagellum brown above, yellow-brown below; 
scape dark brown except for narrow light area below; pedicel usually all 
yellow; last flagellar segment entirely dark; pedicel and first flagellar seg- 
ment each less than 1.5 times as wide as long. (8) Scutum rough, with 
punctures deep and distinct medially, deep and contiguous at parapsidal 
lines, becoming rugose anteriorly; anterior margin variably roughened medi- 
ally, rugose to areolate laterally. (9) Tegula 1.5 times as long as wide. (10) 
Scutellum shiny, coarsely punctate, punctures irregular in size and spacing. 
(11) Pleuron rugose, more coarsely so anteriorly. (12) Propodeum with disc 
equal or subequal in length to metanotum; outline of disc weakly bracket- 
shaped, dorsal surface slightly concave, posterior edge sharply to weakly 
angulate, well defined, rounded laterally; striae usually straight, regular, 
fine, widely to narrowly spaced, reaching edge posteriorly and laterally; 
posterior vertical surface coarsely rugose, usually without punctures; postero- 
lateral corners rounded with posterior rugosity extending onto lateral sur- 


578 Tue University SciENCE BULLETIN 


face; lateral vertical surface less coarsely rugose than posterior surface, rugae 
lineate, perpendicular to anterior and ventral edges. (13) Legs brown, fore 


} 


and hind coxae brightly metallic anteriorly; trochanters and femora with at — 


least slight metallic reflections; tibiae dark yellow, usually brownish central- 
ly; basitarsi pale yellow with following segments slightly darker; posterior 


basitarsus with erect hairs very long on basal third of segment, four times as 
long as width of basitarsus, curved at tips, becoming shorter on apical two- — 


thirds of segment; basal tuft reduced, often inconspicuous. (14) Metasomal 
terga green, sometimes lighter or browner than thorax; first tergum polished 
anteriorly with punctures separated by slightly more than their diameters 
medially, closer laterally, surface smooth but less shiny dorsally, with punc- 
tures slightly larger, more distinct and denser; second tergum with punctures 
similar to first; first sternum strongly metallic, fifth and sometimes fourth 
sterna with feeble metallic reflections; fourth sternum deeply and broadly 
emarginate apically. (15) Pubescence white to golden-white over entire 
body, yellowish on tarsi. (16) Genital capsule, seventh and eighth sterna 
and eighth tergum of type 1 (Figs. 34, 40, 44). 

Comparisons. A. gratiosa is not a very common species. It has been col- 
lected during all months of the year and most frequently from Georgia and 
Florida. It comes in contact with persimilis, aurata, striata and bracteata, the 
females intergrading morphologically with both azrata and striata. Although 
the males of gratiosa are distinct and can usually be identified by the key 
characters (see exception below), females are more difficult to distinguish 
and the subtle differences can be difficult to recognize unless samples of each 
are available. 

A. gratiosa is characterized chiefly by the propodeal characters in the 
females. The disc is as short as or shorter than the metanotum (Fig. 58). 
Its posterior border is usually well delineated, sometimes by a weak carina 
but more usually by its abruptly declivitous edge which has a weak, medial, 
V-shaped depression (Figs. 12, 20). The striae are fine, straight, and distinct, 
extending the full length of the disc; the posterior vertical surface is dull and 
granular, usually with short subhorizontal lineate irregularities (Fig. 76). 
The second metasomal tergum is granulose and similar to the following 
terga rather than minutely punctulate as is the first tergum. 

Females of gratiosa differ from aurata by the distinct, clear-cut features of 
the propodeal disc, by the complete, straight striae and the nature of the 
posterior face and second metasomal tergum. Although aurata may also 
have striae as fine and close together as those of gratiosa, they are rarely as 
straight or regular, nor do they extend onto the posterior margin of the disc 
(Fig. 57). Also, in aurata, the edge of the disc is less sharply posteriorly, 
usually without the medial V-shaped depression, and the posterior surface 
is smoother, shinier and lacks the lineate irregularities. In Texas, the two 


SysTEMATICS OF THE GENUS Augochlorella Nortu of Mexico 579 


species are easily distinguished by the above characters and in addition, the 
first metasomal tergum is strongly punctate in aurata (Fig. 82) but finely 
punctate as usual in gratiosa (Fig. 83). In striata the striae usually extend 
the full length of the disc, the disc in form a is frequently as well delineated 
posteriorly as in gratiosa, possesses the V-shaped depression and may approxi- 
mate the bracket-shaped outline of gratiosa. However, if the striae are as 
straight and regular as those of gratiosa they usually are considerably coarser 
(fewer in number with greater space between them). If they are as fine as 
those of gratiosa they rarely are as straight, regular or well defined. In addi- 
tion, the posterior vertical surface of the propodeum in striata is smoother, 
shinier, with minute punctures and without the lineate granular irregulari- 
ties found in gratiosa, and the second tergum is similar to the first rather 
than the third or fourth terga. 

A. gratiosa is apparenly more closely related to striata a or aurata than to 
any other group. There is comparatively little variation and variation at- 
tributed to it in the past is seemingly partly due to misidentification based 
chiefly, if not exclusively, on the nature of the propodeal disc. When other 
characters are also used, identification becomes easier. The strict definition of 
gratiosa is based chiefly on the distinct and unvarying characters in the male, 
as opposed to the high degree of variability found in males as well as females 
of striata. 

Variation. The few variations that exist are associated chiefly with size 
and color. There is a wide variation in head size (Fig. 86) with the largest 
individuals being found in Louisiana and Alabama. A sample of 22 female 
specimens was measured from Florida, and all females from other states were 
measured. There is some indication that there may be caste differences in 
size although there is no correlation between width of head and season (i.e., 
large, small or average individuals may be found at any time of the year). 
Males show similar variations in size. The width of the head, with only a 
few exceptions is regularly greater than the length. 

Body color is rather uniformly yellowish green except in Florida where it 
varies from dark green to deep violet-blue. Most males (14 out of 18) and 
about 40°% of the females are blue in Florida. In these specimens there are 
weak metallic reflections at the bases of the mandibles in females and on the 
hind tibiae of males, variations rarely found in green specimens. Only 1 out 
of 20 females from Georgia | Tifton, Ga., 6-13-96, Lot 209 (38*) | showed 
such reflections on the mandibles and two green Floridian males had slight 
reflections on the hind tibiae | Jacksonville 9-3-11 (9); Levy Co., Fla., 9-10- 
55 (10) |. 

The extent of brown on the clypeus of females is variable from specimen 
to specimen with no apparent regional trend, but it does not exceed one-third 
of the total length. 


580 Tue Universiry ScieENcE BULLETIN | 


The degree of roughness and amount of punctation on the supraclypeal - 
area is also variable throughout the range. The frons shows very weak bluish 
reflections in most green specimens when the light is properly reflected from > 
the surface but usually there are no readily visible spots of blue such as are 
found in neglectula. All blue-green specimens from Florida showed differen- | 
tial coloring on the various parts of the body with the head usually darker 
(bluer) than the thorax, and the metasoma lightest (greenest) in color. 

In females from Louisiana the antennae are lighter below than above as | 
usual, but the apical third of the flagellum is lighter above than the preced- | 
ing segments. 

The scutum in females is roughly and closely punctured throughout the 
range but is less so than that of striata. Punctures are usually distinct but are _ 
very close or become indeterminate or rugose anteriorly and laterally. This © 
rugosity is not correlated with size, color or distribution, although in Florida — 
rugosity occurs with higher frequency in blue specimens than in green ones 
(in 45°% of the blue and 9°% of the green). 

The characters of the propodeum are remarkably stable in gratiosa, com- | 
pared with the variability in the other eastern species. The sharply delineated | 
disc is usually bracket-shaped and narrow in females, somewhat less distinct- | 
ly so in males. Three of the seven females from Louisiana [2 from 8.5 a-I.ms. 
New Roads, 6-22-60 (20); Olivier, 5-04 (9)]; 2 of the 146 from Florida’ 
| Homestead 4-18-23 (32); Homestead 8-31-27 (9)] and one from Summer- 
ville, S.C. [5-10- (20) | out of seven seen, were found with a more rounded 
outline although it is dificult to draw the line between one type and the 
other. When the posterior margin is subbracket-shaped, the medial V-shaped 
depression is not evident and the sharply angulate edge becomes abruptly 
rounded. The size of the striae varies little and only in the males do the 
spaces between striae vary. On the lateral sides of the propodeum of females 
there may or may not be very fine lineate rugae perpendicular to the anterior 
and ventral margins. They are present in all males seen with the exceptions 
noted below in Texas. Again, this seems to be a variation within populations 
and not correlated with season or distribution. 

The first sternum of the metasoma is variously metallic in females, 
strongly blue or blue-green in many darker specimens to brown with faint 


metallic reflections in others, especially the paler specimens. 

The amount of metallic coloration on the legs of females is relatively 
constant although any or all trochanters and femora may be slightly colored 
in addition to the always colored fore and hind coxae. 

All Floridian females have deep golden pubescence over all the body in- 
cluding the ventral parts of head and thorax, whereas in Texas, pubescence 


* See Table 1. 


SysTEMATICs OF THE GeNUs Augochlorella Nort of Mexico — 581 


is paler with hairs whitish on the upper parts of the head and basal parts of 
the legs, in addition to the venter. The males do not vary in this character, 

Male genitalia have the inner lobe of the gonostylus as shown in Figure 
34. The fingerlike projection averages slightly longer than in the other 
eastern species but enough variation occurs in each of the eastern species that 
distinctions cannot usually be made. The outer lobe has long, unbranched 
hair as in other eastern species. 

The two males from Texas | Nacogdoches, X-3-60 (42) and Victoria, VI- 
10-07 (9) | are divergent individuals falling between gratiosa and persimilts, 
not fitting either group well but appearing to be most like gratiosa. On the 
basis of characters 1, 3 and 7, the specimen from Nacogdoches is most like 
gratiosa (the specimen from Victoria is without a head). The thoracic and 
metasomal characters of both specimens resemble those of either persimilis 
or gratiosa, and the hind basitarsus of each specimen is intermediate between 
the two species although more similar to gratiosa than to persimults. 

Distribution. From New Jersey and Washington, D.C., southward to the 
keys of Florida, along the Gulf Coast states into southeastern Texas, extend- 
ing inland as far as northern Georgia and eastern Tennessee (Map: Fig. 87). 

A total of 29 males and 199 females have been seen: ALABAMA: Mo- 
bile Co.,1 2; Washington Co..1 2 (June). FLORIDA: Alachua Co., 1 
6,8 @ (February-May, August, October, December); Bradford Co.,1 ¢, 
1 2 (April); Brevard Co.,1 4,1 2 (April, November); Broward Co., 
m2 (February, October); Collier Co.1 8,5 2% (April); Dade Co., 13 ¢, 
49 2 (January-April, June-December); Duval Co.,1 6,2 2 (April, Au- 
gust, November); Gadsden Co. 1 2 (April); Hendry Co, 2 2 (July, 
October); Hernando Co.,1 2 (December); Highlands Co.,6 2 (April- 
June, August); Hillsborough Co.,3 2 (April, August); Lake Co., 8 2 
(January-February, April); Lee Co..1 @ (March); Levy Co, 1 6,19 2 
(February, April, June, September); Manatee Co., 2 2 (February, April); 
Marion Co., 4 2 (February, April); Martin Co, 1 2 (March); Monroe 
Co.,4 2 (January, May); Nassau Co. 4 2 (July-August); Orange Co., 
4 9 (February-April, December); Palm Beach Co., 2 2 (March, Septem- 
ber); Pasco Co.,3 2? (January, August); Polk Co.,2 2 (May, September) ; 
Putnam Co.,6 2 (May); Seminole Co..1 9; St. Lucie Co.,1 2 (April); 
Holusia Co.,1 2: DISTRICT OF COLUMBIA: 1 ¢ (May). GEORGIA: 
Haralson Co.,1 2 (June); Lowndes Co.,2 2 (July); Thomas Co. 2 2 
(April); Tift Co..5 6,14 2 (June); Townes Co.1 é (August); Walker 
Co.,.1 2 (June); Ware Co.,1 6,2 2 (July). LOUISIANA: East Baton 
Rouge Parish, 2 2 (July); Iberia Parish, 1 2% (May); Pointe Coupée 
Parish, 3 2 (June); Tangipahoa Parish, 1 @ (June). MISSISSIPPI: For- 
test Co., 3 2 (August-September). NEW JERSEY: 1 @. NORTH 
CAROLINA: Columbus Co. 1 2 (August); Moore Co. 1 2 (Novem- 


582 Tue UNiversiry ScIENCE BULLETIN 


ber); Tyrrell Co. 1 2 (July); Wake Co. 3 2 (April-May). SOUTH 
CAROLINA: Dillon Co.,3 2 (April); Dorchester Co..1 2 (May). TEN2 


NESSEE: 1 2. TEXAS: Colorado Co.,1 2 (March); Jackson Cosa 


(March); Nacogdoches Co.,1 6 (October); Victoria Co..1 6 (June). 

A single male (9) from Washington, D.C., and one female labeled “N. J. 
1786” (9) were examined, although no specimens of this species have other- 
wise been taken north of North Carolina. One female |Chickamauga, Ga. 
V1-24-98 (38) | taken in the northwest corner of Georgia in Walker County 
and one female labeled “E. Tenn.” (38) represent the most inland records. 
Although specimens are scarce and widely scattered in this region, they do 
not differ from others except that the individual from Walker County is 
somewhat paler with more brown on the metasoma than usual. 

Lovell (1942) records this species from Jefferson County, Kentucky, in 
the north central part of the state. I have not seen any of his specimens but 
it seems unlikely that gratiosa ranges that far inland. 

Seasonal Activity. Females of gratiosa are active throughout the year in 
Florida and have been taken from April through September in other states, 


Fic. 89. Distribution of A. gratiosa. 


SYSTEMATICS OF THE GENUS Augochlorella NortH or Mexico — 583 


Males have been collected from March through December in Florida and 
June through October elsewhere. Pollen collectors are found from mid-April 
at least through July and probably well into the fall throughout the range 
and from the end of February to September in Florida. 

Flower Records. Aster, Ampelopsis, Asclepias, Berteroa, Bidens, Calli- 
carpa, Cassia, Chrysobalanus, Chrysopsis, Cirsium, Citrus, Clethra, Cratae- 
gus, Crotonopsis, Cunila, Erigeron, Eryngium, Galactia, Gerardia, Gossypt- 
um, Helianthus, Hypericum, Ilex, Ixora, Jacquemontia, Lepidium, Malva, 
Melilotus, Aenothera, Opuntia, Piriqueta, Polygala, Polygonum, Pterocaulon, 
Pycnothymus, Rhus, Rubus, Sabal, Senecio, Solidago, Taraxacum, Teucrium, 
Vaccinium, Verbena, Warea. 


Augochlorella aurata (Smith) 


Augochlora aurata Smith, 1853, Catalogue of the Hymenoptera in the British 
Museum 1:82 (descr.); Cresson, 1887, Trans. Amer. Ent. Soc., Suppl. 14s:293 
(list); Dalla Torre, 1896, Catalogus Hymenoptorum 10:94 (list); Bingham, 
1897, Trans. Amer. Ent. Soc. 24:162 (list); Brimley, 1938, Insects of North 
Carolina p. 454 (list, fl.). 

Augochlora austrina Robertson, 1893, Trans. Amer. Ent. Soc. 20:147 (descr.); 
Dalla Torre, 1896, Catalogus Hymenoptorum 10:94 (list); Cockerell, 1922, 
Proc. U.S. Nat. Mus. 60:16 (descr.). 

Augochlora (Augochlorella) aurata: Michener, 1951, in Muesebeck ef al., U.S. 
Dept. Agr., Agr. Monogr. 2:1125 (in part) (list). 

Augochlorella aurata: Sandhouse, 1937, Jour. Washington Acad. Sci. 27:71 (in 
part) (tax.); Mitchell, 1960, Bees of the Eastern United States 1:459 (in part). 


Types. Augochlora aurata, holotype female, from St. John’s Bluff [? St. 
Johns Co.], Eastern Florida, is in the British Museum (Natural History). 
Although labels on the holotype agree with those in Smith’s published de- 
scription of aurata, the description agrees best with the holotype of gratiosa. 
Conversely, his description of gratiosa better fits the type of aurata. For 
further discussion of this problem, see Augochlorella gratiosa. 1 have not 
seen the type but Dr. C. D. Michener took detailed notes on it and compared 
it with submitted specimens. Augochlora austrina, holotype female, No. 
12859 [Robertson’s number], from Inverness [Citrus Co.], Florida, 1892, is 
in the Robertson Collection at the Illinois Natural History Survey. For the 
most part, the azrata of Sandhouse (1937) and Mitchell (1960) is persimilis. 

Description. Female: (1) Body length 7 mm; head width 1.58 to 1.98 
mm, averaging 1.84 mm, width to length ratio variable. (2) Color yellow- 
green to blue-green; frons without bluish reflections on green specimens, 
metasoma similar in color to other body regions. (3) Mandible with basal 
third dark brown, yellow-brown centrally, rufous at tip, rarely with metallic 


584 Tue University ScieENcE BULLETIN 


reflection at base. (4) Clypeal width subequal to length; basal part green | 
with large, irregularly spaced punctures; apical fifth or less, brown, brown | 
area not exceeding one-third, slightly beveled; punctures in brown area _ 
round, or slightly elongate when brown area exceeds diameter of puncture; — 


surface between punctures usually shiny and smooth. (5) Supraclypeal area 


variably punctate, surface between punctures shiny and smooth or finely — 


roughened. (6) Paraocular area punctorugose to rugose below antenna, 
more coarsely rugose above. (7) Antenna dark brown, often slightly lighter 


below than above; pedicel with length subequal to width, first flagellar seg-_ 


ment wider than long. (8) Scutum coarsely punctate; punctures close to 
contiguous over entire dorsum, similar to frons; anterior margin and antero- 
lateral corners finely rugose to finely areolate. (9) Tegula almost twice as 
long as wide. (10) Scutellum rere or shallowly “ind irregularly punc- 


tate. (11) Pleuron finely rugose, becoming areolate anteriorly. (12) Pro- | 


podeum with disc equal to or slightly longer than metanotum; outline of 
disc bracket-shaped to semicircular, usually weakly bracket-shaped, forming 
blunt point medially, profile type 2; posterior edge of disc distinct, abruptly 
rounded to sharp, gradually rounded laterally; striae fine, close, wavy, usual- 


ly irregular, ending just before edge or at edge medially, often leaving edge | 


slightly raised, roughened and dull, reaching or crossing edge laterally; pos- 
terior vertical surface evenly and finely granular, not rough; posterolateral 


corners not prominent, usually slightly more roughened than posterior sur- | 


’ 
| 


face; lateral vertical surface weakly rugose. (13) Legs brown, fore and hind — 
coxae, mid and hind trochanters and femora with metallic reflections. (14) 


First metasomal tergum with anterior surface shiny but not polished, with 


numerous widely spaced punctures; dorsal! punctures minute and close, or 


large, close and distinct (Texas); first sternum without metallic reflections. 
(15) Pubescence golden-white on dorsum and legs and ventrally on meta- 
soma; white ventrally on head and thorax; pubescence short and thick but 
not dense on genal area. 

Male: (1) Length 7 mm; head width 1.81 to 1.91 mm, averaging 1.86 
mm, width to ener ratio variable. (2) Color bright green; often with 
bluish reflections on frons; metasoma cotren slightly redder or browner above 
than on other parts of body. (3) Mandible with dark metallic reflections 
basally. (4) cre with punctures variable in size and spacing; surface 


peuvent punctures shiny and smooth. (5) Supraclypeal area variably punc- | 


tured with surface smooth and shiny or irregularly roughened. (6) Paraocu- 
lar area finely punctorugose. (7) Flagellum dark brown above, yellowish 
brown below; scape dark brown; pedicel partially light brown above, dark 
brown below; last flagellar segment entirely dark brown; pedicel and first 
flagellar segment each about 1.5 times wider than long. (8) Scutum with 
punctures distinct but crowded medially, separated by less than their di- 


SYSTEMATICS OF THE GENUS Augochlorella NortH oF Mexico — 585 


ameters, becoming contiguous at parapsidal lines; anterior margin rugose, 


becoming areolate laterally. (9) Tegula twice as long as wide. (10) Scutel- 
lum shiny, punctate to punctorugose. (11) Pleuron rugose, becoming areo- 
late anteriorly. (12) Propodeum with disc equal to or slightly longer than 


metanotum; outline of disc weakly bracket-shaped to obtusely V-shaped, 


posterior edge abruptly rounded; striae fine, wavy and irregular, reaching 


edge posteriorly, crossing edge laterally; posterior vertical surface and pos- 


-terolateral corners finely rugose or roughened; lateral vertical surface rugose. 


(13) Legs brown, with fore and hind coxae, trochanters and femora reflect- 
ing green; tibiae reflecting green medially, testaceous at extremities; tarsi 
testaceous; hind basitarsus with erect hairs uniform in length, up to 1.5 


times as long as width of segment, variable among individuals, pale yellow 


in color; basal tuft distinct. (14) Metasomal terga green; first tergum 


polished, with widely scattered fine punctures anteriorly, smooth but less 


shiny dorsally, punctures small and close; sterna brown, pubescence short, 
fine over entire sterna; first sternum with weak metallic reflections, fourth 
sternum shallowly emarginate. (15) Pubescence white on head, white to 
golden dorsally on thorax, golden on metasoma and legs. (16) Genital cap- 
sule, seventh and eighth sterna and eighth tergum all of type 1 (similar to 
Bies, 32, 40, 43). 


Comparisons. Very few specimens of aurata have been collected outside 


of Florida and Texas although nine specimens are available from Alabama, 


Georgia and North Carolina. The females are most similar in appearance to 


those of gratiosa and the males to those of striata, the range of variation in 


Florida overlapping those of both striata a and gratiosa. 

The females can usually be distinguished from gratiosa by the slightly 
longer propodeal disc, the flatter bracket-shape (Figs. 57, 58) and less acutely 
angulate posterior edge of the disc and the smoother posterior vertical sur- 
face of the propodeum. In Texas they can be additionally distinguished by 
the deep, crowded, distinct punctures on the first and second metasomal 
terga (Figs. 82, 83). They can be distinguished from striata a in North Caro- 
lina and Georgia by the flatter bracket-shaped disc with a less acutely angu- 


late posterior edge and the finer, more irregular striae, and in Florida tigre, 


_ by finer more irregular striae and the less rugose sculpturing on the thorax. 


There are no striata females in the south with which this species could be 
confused. 


The males are similar to striata with short basitarsal hairs of more or less 
uniform length, shallowly emarginate fourth metasomal sternum and dark 
tipped antennal flagellum. The range of variation is not known since only 
five males from Georgia have been positively identified and these were 
similar to one another. Ten males from Florida are also tentatively included. 


| 
| 
| These look very similar to variants of striata. The males from Georgia differ 


586 Tue Universiry SciENcCE BULLETIN 


from all striata males in the flatter, more finely striate, propodeal disc. The | 
difference in scutal punctures will also separate what are believed to be 
Floridian awrata from Floridian s¢riata. 

I believe that aurata is more closely related to striata than to gratiosa al- 
though it is more difficult to distinguish the females from those of gratiosa, 
The paucity of males may reflect limited collecting at the proper time of year | 
or their occurrence in locations other than where the females were taken or 
it may be that they have not been distinguished from striata males since the 
nature of the variation in striata has not been fully evaluated. 

Variation. There is comparatively little variation in size although par- 
ticularly large or small individuals may occasionally be found in Florida. | 
The color, usually a yellow-green to bright green, is often blue-green in 
Floridian specimens although yellow-green individuals may also be found. 

The clypeus is apically brown, usually for one-sixth to one-fourth its 
length in females, but is one-third brown in some specimens from Florida 
and one-half brown in some specimens from Georgia. The face looks long 
(Fig. 52) in all Texan and some Floridian specimens but round in the rest 
(Fig. 53). 

The propodeal disc of the female is usually slightly longer than the meta- 
notum, or it may be equal to the metanotum but is not shorter. The disc is_ 
nearly always at least weakly bracket-shaped and bluntly pointed medially 
(Fig. 57). The edge is usually abruptly rounded and distinct although it 
may be either sharply angulate or rounded and indistinct in some specimens — 
from Texas and Florida. There is more variability among Floridian speci-_ 
mens in this character than elsewhere in the range; the four specimens from | 
Georgia are all similar to Figure 57. The striae are rarely as straight and 
well defined as in gratiosa (except for a few from Florida) but are very fine, - 
irregular or vermiform, with no definite spaces between them. The posterior 
edge is usually minutely roughened when rounded and the striae end grad- 
ually in this roughened area. The most striking variation occurs in the 
metasomal punctures of specimens from Texas. Throughout the rest of the 
range the punctures are small, close, shallow, and almost inconspicuous as 
in gratiosa (Fig. 83). In Texas, the punctures, although also close, are 
slightly larger, much deeper, and more conspicuous (Fig. 82), giving the 


tergum a coarse or roughened appearance. 

Males have been seen only from Georgia and Florida. Those from Geor- 
gia have the disc resembling that of the females, widely bracket-shaped, with 
fine irregular striae ending in a minutely roughened area at the edge. None 
of the presumed aurata males from Florida have discs similar to this or to 
that of the females. The shape of the disc in Florida varies from weakly 
bracket-shaped and narrow to long and roundly V-shaped. The striae are 
usually rather thin and close, and the posterior edge is usually abruptly 


SYSTEMATICS OF THE GENUs Augochlorella Nortu of Mexico — 587 


rounded; the posterior surface may be weakly roughened as in the Georgian 
specimens or rugose as in many male striata. These specimens are all differ- 
ent from one another, resembling males of striata s but are unlike the 
striata a from Florida. There is considerable variation in the characters of 
the disc of females in Florida; perhaps the variability is as great in males. 
Due to the uncertainty in identification of the Floridian males, the above 
description of the male is based solely on the specimens of azrata from 
Georgia. 

Throughout Florida, occasional female specimens are found that are 
brilliantly shiny, very finely punctured, with a body surface finely roughened 
or at most, weakly rugose on the thorax. These individuals are always dark 
blue-green in color but in all other respects are similar to other Floridian 
aurata. Eleven such specimens have been seen. They do not form an iso- 
lated population nor have corresponding males been found and therefore it 
seems improbable that these few specimens represent a different species. 
[Such specimens are from: W. Palm Beach, [X-3-27 (Graenicher) (32*); 
Highlands Co., VI-6-60 (Weems) (10); Pasco Co., HI-2-57 (Weems) (10); 
Port Sewell, I-24-29-44 (Sanford) (32); Jacksonville Beach, VIII-5-36 
(Mitchell) (33); Archibald Res. Sta., Lake Placid, [V-6-61 (Dietrich) (32) |. 
_ Distribution. This species is found from Florida along the Gulf coast 
into Texas and northward along the east coast as far as North Carolina 
(Map: Fig. 88). 

_ A total of 15 males and 163 females have been seen: ALABAMA: Hous- 
lion Co, 1 2 (July). FLORIDA: Alachua Co..1 6, 8 2 (March-May, 
October, December); Brevard Co.,1 8,10 @ (March-April, July, Novem- 
ber); Broward Co.2 2 (February); Collier Co..1 2 (April); Dixie Co., 
1 6 (August); Duval Co. 1 6,16 2 (May, August-September); Flagler 
}Co., 2 2 (February, December); Glades Co..1 @ (March); Hendry Co., 
1 é (July); Highlands Co..5 @ (April, June); Hillsborough Co., 3 2 
(April): Indian River Co., 1 2 (March); Jackson Co. 1 2 (August); 
Levy Co.,1 2 (September); Marion Co.,1 2 (March); Martin Co.,1 2 
(February); Nassau Co. 2 2 (July); Okeechobee Co, 1 2 (April); 
Orange Conse 52) 2 (March- June); Palm Beach Co.,2 6,5 2 (Septem- 
ber); Pasco Co. 1 9 (March); Pinellas Co., 1 9 (April, IAW RACs 
s(March-April); Sz. Lucie Co, 1 2 (April); Sz. Johns Co. 2 2 (seul) 
Seminole Co. 4 2 (May, July); Volusia Co.,2 6,6 2 (June-September). 
GEORGIA: Brooks Co, 1 2 (March); Dougherty Co, 1 8 (June); 
iowndes Co. 4 8 (July); Pike Co, 1 2 (April); Thomas Co. 1 9 
(April). NORTH CAROLINA: “N.C.” 1 ¢@..TEXAS: Aransas Co., 
4 2 (July-August); Bastrop Co.,2 % (May); Bee Co.1 2 (May); Cal- 


*See Table 1. 


588 Tue University SciENcCE BULLETIN 


houn Co.,2 2 (April); Cameron Co. 4 2 (June); Colorado Co, 1 2 
(April); Fayette Co.,3 ? (March); Goliad Co.5 2 (May); Jackson Co., 
4 9° (March, July); Kenedy Co, 2 2 (April); Lee Co, 20°98 (@Maneum 
June); Nacogdoches Co. 3 2 (April, September); Refugio Co, 1 2 
(April); Vuctoria Co. 16 2 (February-May, August, October); Waller 
Com 2 (@Nprall), 

Seasonal Activity. Females of aurata are apparently active throughout 
the year and have been collected from early February through December. 
Males are active from early June through November. Pollen collectors have 
been found from early February to the beginning of September although 
nesting activities may continue later into the fall. 

Flower Records. Aster, Cirsium, Citrus, Crataegus, Lythrum, Opuntia, 
Polygonum, Rubus, Viburnum. 


Augochlorella persimilis (Viereck) (n. comb.) 
Halictus (Oxystoglossa) persimilis Viereck, 1910, in Smith. Ann. Rept. New 


ersey State Mus. for 1909, p. 688 (list) (new name for Augochlora similis | 
y p g 


Robertson, preoccupied in Halictus; Viereck’s identification was in error but 
his new name stands for similis Robertson). 

Augochlora persimilis: Rau, 1922, Trans. Acad. Sci. St. Louis 24:33 (list, f1.); 
Graenicher, 1935, Ann. Ent. Soc. Amer. 28:302 (list). 

Augochlorella persimilis: Ordway, 1964, Jour. Kansas Ent. Soc. 37:139-152 (biol.); 
Ordway, 1965, Insectes Sociaux 12:291-308 (biol.); Ordway, in press, Jour. 
Kansas Ent. Soc. (biol.). 


Augochlora similis Robertson, 1893, Trans. Amer. Ent. Soc. 20:146 (descr.); | 


Robertson, 1894, Trans. Acad. Sci. St. Louis 7:436-472 (fl.); Robertson, 1896, 
Trans. Acad. Sci. St. Louis 7:175 (fl.); Dalla Torre, 1896, Catalogus Hy- 
menoptorum 10:96 (list); Bridwell, 1899, Trans. Kansas Acad. Sci. 16:210 
(list); Cockerell, 1899, Ent. News 10:3 (list); Graenicher, 1911, Bull. Pub. 
Mus. Milwaukee 1:234 (list); Banks, 1912, Ent. News 23:107 (fl.); Brimley, 
1938, Insects of North Carolina, p. 455 (list, fl.). 

Augochlora (Oxystoglossa) similis: Robertson, 1902, Canad. Ent. 34:247 (key). 


Oxystoglossa similis: Robertson, 1928, Flowers and Insects, pp. 1-221 (fl.); Pear- | 


son, 1933, Ecol. Monogr. 3:386, 396 (biol.). 

Halictus (Oxystoglossa) xystris Vachal, 1911, Misc. Ent. 19:50 (key, descr.); 
Mitchell, 1960, Bees of the Eastern United States 1:460 (tax.). 

Augochlorella aurata: Sandhouse, 1937, Jour. Washington Acad. Sci. 27:71 
(in part) (tax.); Lovell, 1942, Kentucky Acad. Sci. Trans. 10:20, 21, 23 (keyg 
descr.) ; Mitchell, 1960, Bees of the Eastern United States 1:459 (in part) (tax.); 
Sakagami and Michener, 1962, Nest Architecture of Sweat Bees, pp. 1-135 
(biol.); Eickwort and Fischer, 1963, Ann. Ent. Soc. Amer. 56:350 (list) (mis- 
identifications ). 

Augochlora ( Augochlorella) aurata: Michener, 1951, in Muesebeck et al., U.S. 


| SYSTEMATICS OF THE GENUS Augochlorella Norvu or Mexico — 589 


Dept. Agr., Agr. Monogr. 2:1125 (in part) (list); Montgomery, 1957, Proc. 
Indiana Acad. Sci. 66:132 (ice fl.) (misidentifications ). 

Augochlorella neglectula: Dreisbach, 1945, Michigan Acad. Sci. Arts and Letters. 
Papers 30, p. 225 (misidentification ). 


Types. Augochlora similis, \ectotype female here designated, No. 1104, 
from Carlinville, Macoupin Co., Illinois, 1891, is in the Robertson Collection 
at the Illinois Natural History Survey. This specimen has been selected from 
among 19 syntype females. It agrees with the original description and may 
be considered “typical” of the species. There are eight male syntypes in the 
Robertson Collection, one (No. 474, Robertson’s number) being a large 
striata and 18 syntype females all from Carlinville, Macoupin Co., Illinois, 
with dates from 1885-1887 and 1890-1892. Halictus xystris Vachal, lectotype 
female and two syntypes each labeled “xystis [sic] @ Wach.” are located in 
the Vachal collection at the Museum National d’Histoire Naturelle, Paris. 
I have designated as lectotype one of the three syntypes that best fits the de- 
scription and the only one with a locality label. The handwritten label reads 
“Augochlora pura, from Potentilla, S. Ill. June. . .”. The label has been 
cropped closely, obliterating the rest of the date. It is probable that Vachal 
mistook the handwritten “S. Ill.” for “S.W.” and published the locality as 
“southwestern United States”. Two of the three specimens including the 
lectotype resemble persimilis from southern Illinois. The third is a persimilts- 
striata intermediate that might be a small striata form c. It does not agree as 
well with the original diagnosis as the propodeal striae are regular and almost 
reach the edge of the rounded disc so that there is scarcely any smooth shiny 
area beyond the striations. I was able to see these specimens thanks to the 
generosity and cooperation of Mademoiselle S. Kélner-Pillault of the Mu- 
seum National d’Histoire Naturelle, who lent them to me. 

Although Sandhouse (1937) synonymized this species with aarata Smith 
1853, the type of aurata is different from Robertson’s similis. 

Description. Female: (1) Length 5 to 6 mm; head width 1.44 to 1.9] 
mm, averaging 1.69 mm; head width greater than length. (2) Color bright 
green to yellow- or coppery-green; frons without bluish reflections; meta- 
soma usually more golden, coppery or brownish than head and thorax. 
(3) Mandible with basal third dark brown, yellow-brown centrally, rufous 
at tip, without green basal reflections. (4) Clypeal length equal to width or 
slightly longer; basal part green with punctures variable in size, smallest and 
closest near basal angles, becoming larger apically and separated by about 
twice their diameters or more; apical third to two-thirds of clypeus brown 
and slightly beveled, with elongate punctures becoming shallow and indis- 
tinct at apex, giving apex roughened appearance; surface between punctures 
smooth and shiny. (5) Supraclypeal area with surface weakly roughened, 


590 Tue University SciENCE BULLETIN 


punctures small just below antennae and along subantennal sutures, some- | 
times with few scattered punctures centrally. (6) Paraocular area usually | 
closely punctate below antenna, finely rugose above antenna. (7) Antenna — 
brown, flagellum lighter below than above; first flagellar segment wider — 
than long; pedicel slightly longer and narrower than first flagellar segment, 
ratio of fener to width variable. (8) Scutum coarsely and irregularly punc- 
tured, punctures close, usually with little or no space between them, rarely — 
separated by as much as a puncture width centrally, becoming closer and . 
coarser laterally; surface between punctures, when present, smooth and shiny | 
centrally; anterior margin roughened with surface finely lineolate and dull 
medially, becoming rugose laterally and at anterolateral angles. (9) Tegula 
twice as long as wide. (10) Scutellum shiny and roughened, irregularly — 
punctate or rugose. (11) Pleuron irregularly rugose, becoming coarsely 
areolate anteriorly. (12) Propodeum with disc slightly longer than to a little” 
more than 1.5 times as long as metanotum; outline of disc roundly semi-~ 
circular, profile type 2, posterior edge indistinct and gradually rounded; 
striae variable, usually irregular, branched, vermiform, occupying 60 to 80_ 
percent the length of disc medially, reaching edge laterally; surface beyond | 
striae smooth but minutely lineate or anely roughened; posterior vertical 
surface shiny and smooth with sparsely scattered minute punctures, or sur- 
face finely granular with granular texture extending across upper part of 
posterolateral corner to lateral surface; lateral vertical surface weakly rugose 
or coarsely roughened with widely separated or reticulated rugae. (13) Legs 
brown; fore and hind coxae strongly metallic; trochanters and femora with 
feeble metallic reflections. (14) First metasomal tergum with anterior sur- 
face brilliantly polished with a few widely spaced fine punctures; punctures 
more numerous and surface less brilliant dorsally; second tergum with 
numerous fine punctures separated by about twice their diameters; surface 
usually dull; first sternum sometimes darker than others, often greenish but 
not metallic. (15) Pubescence golden-white dorsally on head, thorax, apical 
segments of legs, and on dorsal and last two ventral metasomal segments; 
white ventrally on head and thorax; white or golden on basal segments of 
legs and ventral part of metasoma. 

Male: (1) Length 7 mm; head width 1.45 to 1.80 mm, averaging 1.66 
mm; head width equal to, greater than or less than length with no regional 


or seasonal pattern. (2) Color yellow-green to coppery-green; frons without 
blue reflections; metasoma usually more golden or reddish than rest of body. 
(3) Mandible with or without metallic reflections basally. (4) Clypeal sur- 
face shiny between rather large punctures; punctures irregular in size, shape 
and spacing, smallest along basal edge. (5) Supraclypeal area with small 
scattered punctures; surface between punctures minutely roughened, dull; 
rougher immediately below antennae than just above clypeus, or shiny and 


SysTEMATICS OF THE GENUS Augochlorella NoxtH oF Mexico — 591 


smooth. (6) Paraocular area finely and closely punctate to rugosopunctate. 
(7) Flagellum dark brown above, yellow below; scape entirely dark brown 
except for small apical yellow area on underside; pedicel dark brown and 
yellow; last flagellar segment rarely darker below than preceding segments, 
but if so, then only partially dark apically; pedicel and first flagellar seg- 
ment about 1.5 times as wide as long. (8) Scutum shiny with punctures 
distinct, separated by less than their own diameters medially, slightly closer 
laterally, smaller and closer posteriorly; anterior margin roughened to finely 
rugose medially to rugose laterally. (9) Tegula more than 1.5 times as long 
as wide. (10) Scutellum shiny, coarsely punctate, punctures distinct but ir- 
regular in size and spacing. (11) Pleuron rugose to rugosopunctate, coarsely 
areolate anteriorly. (12) Disc of propodeum longer than metanotum; outline 
of disc semicircular, posterior edge prominent, abruptly rounded medially, 
gradually rounded laterally; striae variable, usually moderately coarse, irreg- 
ular or wavy, not quite reaching edge medially, attaining edge laterally; 
surface of disc beyond median striae coarsely roughened to smooth and 
shiny with minute reticulations; posterior vertical surface usually smooth 
and brilliant, or only weakly and irregularly roughened, upper part of pos- 
terolateral corners minutely punctate to weakly roughened; lateral vertical 
surfaces roughened to rugose or finely areolate with weak horizontal rugae 
along anterior and ventral margins. (13) Legs with fore and hind coxae, 
trochanters and femora bright green, tibiae yellow-brown, usually darker 
centrally, hind tibia sometimes weakly reflecting green on inner surface; 
tarsi pale yellow; hind basitarsus with erect hairs along apical two-thirds of 
segment only, longer basally than apically; longest hairs about twice as long 
as width of segment, usually slightly curved at tips; basal third of segment 
without erect hairs; basal tuft short and sparse. (14) Metasomal terga 
green with apical margins usually narrowly brown; first tergum polished 
anteriorly with few, widely scattered fine punctures, smooth but less shiny 
dorsally, punctures denser, minute, separated by 1.5 times their diameters or 
less; sterna brown, first sternum with green reflections; fourth sternum 
emarginate apically. (15) Pubescence short, thick and white between an- 
tennae and on paraocular areas, white on cheeks, venter of thorax and basal 
segments of legs; white to golden-white on clypeus, frons and vertex, dor- 
sum of thorax, on tibiae and tarsi and metasoma. (16) Genital capsule, 
seventh and eighth sterna and eighth tergum of type 1 (Figs. 33, 40, 43). 

Comparisons. Although the range of persimilis overlaps that of four 
other species of Augochlorella, females intergrade only with form c of strt- 
ata. There is no sure way of separating females of the two species where 
intermediates occur although the key will distinguish a majority. Males of 
the two species remain distinct and are readily distinguishable by the key 
characters. 


592 Tue UNiversity SCIENCE BULLETIN 


Both males and females of persimilis may superficially resemble bracteata | 
in size, coloration and often in characters of the propodeal disc. There is 
however, only slight overlap in ranges and the consistantly rougher thorax | 
of persimilis effectively serves to distinguish the two species. r 

Apart from bracteata and striata c, persimilis can be distinguished from ~ 
the eastern species by its generally smaller size, smoother body and propodeal - 
disc with rounded posterior edge and short medial striae. The males are 
distinguishable by the long hind basitarsal hairs (Fig. 50) although in Texas | 
two male specimens of gratiosa were found that looked very similar to those | 
of persimilis (see variation under gratiosa). The fourth metasomal sternum © 
of the males is about as emarginate as that of gratiosa, so that any distinction — 


is difficult to make, especially when the segments are telescoped. 

Variation. Body color varies in persimilis from blue-green to golden or 
coppery-yellow. The metasoma is usually lighter in color than the head and | 
thorax (it may be browner or yellower). The males are predominantly 
yellowish green throughout the range except in Arkansas, Virginia and_ 
Georgia where all individuals seen are coppery in color. In Kansas, where a_ 
large sample was available, 85°/, of the females are bright green, 147%, yellow- 
green or coppery and 1% blue-green. More than half the specimens are 
yellow-green or coppery-green in Tennessee, North Carolina, Maryland, 
Texas, Arkansas, Indiana, Missouri, Minnesota and Virginia, but populations 
are predominantly bright green in Illinois, Kansas, Nebraska and Wiscon- 
sin. About half the specimens are bright green in Louisiana and Oklahoma. 

In both males and females, the width of the head varies widely within 
any one area. When measurements were pooled for each sex, however, nor- 
mal distributions were obtained with only a slightly skewed distribution in 
the case of males. The head width may be greater than, equal to or less 
than the length in the case of the males with only slight differences between 
width and length. In females the width is consistently greater than the 
length. Little difference in size was found between populations from differ- 
ent areas. In females, the average head width of field caught bees is greater 
during the spring (March, April, May) (Fig. 89) than during other months; 
at least throughout much of the range only queens are present in spring. 
During the rest of the season, both queens and workers are present. (See 
Ordway, 1965a, for discussion of size and caste data.) 

The length of the clypeus in females is about equal to the width giving 
the face a round appearance, especially in smaller individuals. The spacing 
and number of punctures on the clypeus are variable among individuals. 
Also, the extent of the brown color on the apical portion varies from about 
one-third to one-half the length. No regional trends were observed for 
either character. 


SYSTEMATICS OF THE GENUS Augochlorella Nortu of Mexico — 593 


The supraclypeal area is always roughened in females and at least par- 
tially so in males. The amount of punctation in this area is variable as is the 
degree of roughness. The paraocular areas are usually punctate near the 
lower ends of the eyes and finely rugose elsewhere. However, in about one 
percent of the females from throughout the range the roughness extends to 
the bases of the mandibles. 

The antennae of some males have the last flagellar segment slightly darker 
below than the preceding segments but the segment is only partially dark 
and the darkening is slight. 

The punctures of the scutum in females are rather constant in size but 
vary in respect to their spacing. The punctures are always distinct centrally 
but are closer toward the edges and the scutum may become rugose laterally 
and anteriorly. The amount of space between the central punctures varies; 
usually the punctures are close together with little space between, giving the 
surface a rather rough appearance. The surface looks smoother when the 
punctures are more widely spaced as in many of the specimens from Arkan- 
sas and a few from Virginia. The anteromedial surface is finely lineolate or 
roughened. This roughening extends for varying distances along the median 
suture but is always evident at least at the anterior end of the suture. There 
is little variability in scuta of males. 

The scutellum in males and many females has punctures of various sizes 
and unequal spacing. In the females the scutellum may have distinct punc- 
tures or the punctures may run together or the surface may be entirely 
rugose, with all conditions occurring in populations throughout the range. 
When the punctures are distinct in females, they are closer and smaller at 
the edges, becoming rugose along the posterior margin and along the medial 
line, usually being shallower and smaller than those on the scutum. In males 
the punctures may vary from widely spaced to crowded. 

There is little variation in the pleural region. Although the sculpturing 
of the mesepisternum and metepisternum is about equally coarse, the rugose 
patterns of the two areas are different. 

As in other eastern species the propodeal area is highly variable, yet in 
females, it remains the most diagnostic character available. The disc is longer 
than the metanotum in both sexes. Only 1 male out of about 50 measured 
was found with disc and metanotum equal in length [Illinois, 566, Hart 
Coll. (14*) |. In males the metanotum showed greater variation in length 
than the propodeum; in females, both structures varied in length. 

The shape of the margin of the disc and the lack of a V-shaped midapical 
depression is unvarying in all males. In 2 males out of about 300 examined, 
the usually thick posterior margin was thin, abruptly angulate and almost 


* See Table 1. 


594 Tue UNIversity ScIENCE BULLETIN 


carinate | Indiana, Warren Co., VII-25-50 (16); Missouri, Tecumseh, VI-9-60 
(25) |. In females the shape of the margin varies little and there is no V- 
shaped depression. Although the smooth, lineate, posterior part of the disc 
does extend onto the posterior surface in the shape of a V, there is no median 


depression as is found in gratiosa or some striata. The posterior edge in fe- 


males usually is thickened as in males, but may be very narrow, flat or in 


certain cases completely rounded so that there is no clear demarkation of the — 


margin [Illinois, 16966 (14); Nebraska, Nebraska City, VII-23-01 (28) ]. 


The striae are extremely variable in both sexes although more so in females. 


Variation in males is limited chiefly to the thickness of the striae and to the | 
amount of their separation. The striae in both males and females may be _ 
regular and straight or, more usually, at least partially wavy, branched or | 


irregular. All grades of irregularity occur in the striae of females, but rarely 
are the striae straight and distinct in the central area, and in no case was a 
specimen found in which the striae were both thick and straight, and widely 
spaced as in the large striata c or Floridian striata a. Although the striae 
rarely exceed 80°% of the length of the disc centrally, specimens may be 


f 


found where they reach into the lineate region medially | Texas (38); Kan- | 


sas, Douglas Co. (20) etc.].. The lateral striae are nearly always rather 
straight and distinct in both males and females. Various types of “extreme” 
conditions appear periodically in females of various populations. It is not 
feasible to cite them all, but they include forms without striae and with only 
fine roughening along the basal half of the disc | Kansas, Lawrence, VIII-3-58 
(20); Wisconsin, Oshkosh, VIII-7-16 (47) |, or with fine rugae running 
transversely and joining with lateral striae | Kansas, Douglas Co., [X-5-53 
(28); Missouri, Buffalo, VI-8-52 (20) | or with striae so irregular that there 
is no linear quality at all | Illinois, Algonquin, VI-4-09 (14); Missouri, Big 
Spring St. Pk. (20) etc.|. The posterior surface of the propodeum of females 
may be shiny and smooth or slightly less brilliant and granular in nature. 
No specimen was seen with rugae on this area. Specimens from the east 
(Maryland, Georgia, North Carolina, Virginia) are predominantly shiny 
and smooth; in other areas both conditions occur in about equal proportions. 
Three specimens were found in which the propodeal area was somewhat 
misshapen with the result that the posterior surface was “wrinkled looking,” 
shiny and without the usual minute punctures | Arkansas, Jonesboro, V1-29- 
52 (20); Wisconsin, Oshkosh, VIII-7-16 (47); Nebraska, Nebraska City, 
VII-23-01 (28) |. In males the posterior surface is usually very shiny and 
only slightly but variously roughened. This roughening may be in the form 
of shallow punctures which may or may not be distinct or may be merely 
unevenness of the surface. Two specimens, however, were found with very 
rough and somewhat duller posterior surfaces |TIlinois, 32408 (14); Illinois, 


Willow Spr., VIII-12-05 (14) ]. 


SysTEMATICS OF THE GENUs Augochlorella NortH or Mexico = 595 


There is little variation in coloration of the legs in either males or females 
except for the intensity of green. This coloration seems to be correlated with 
the darkness of body coloration, the paler (yellower) individuals having less 
strongly green legs. Such variation occurs throughout the range. The length 
of the hairs on the hind basitarsus of the male is rather constant. Only one 
male was found where the long hairs were as short as one-half the width of 
the basitarsal segment and in this case they originated close to the basal 
tuft with somewhat less space separating the tuft and the hairs than is nor- 
mal | Arkansas, Malvern, VI-15-58 (25) |. 

The metasoma shows the usual color variation of other body regions. 
The first sternum of the male is variously tinted with green. Some speci- 
mens have the metallic nature barely visible [ Louisiana, 2392 (9); several 
Minnesota specimens, etc.] whereas others are bright green or intermediate. 
In females the first sternum is not green although it may vary from light 
brown to dark brown and may be shiny and greenish but never metallic. 

One male was found in which the second tergum is granulose and simi- 
lar to the third rather than punctate as is the first | Missouri, Buffalo, VI-8-52 
(20) |. The third tergum is frequently punctate like the first in females. In 
females the second tergum may be similar either to the first or third or even 
occasionally intermediate [Indiana, Tippecanoe Co., VI-16-53 (16) |. Again, 
this variation appears to be individual rather than regional in nature. 

The color of pubescence in females varies regionally to a slight extent. 
In the eastern states (Georgia, Maryland, North Carolina, Virginia, Tennes- 
see, Louisiana) the ventral part of the metasoma and basal segments of the 
legs have golden rather than white hair. In the midwest (Oklahoma, Wis- 
consin, Minnesota, Nebraska, Kansas, Missouri, Texas, Arkansas) most 
specimens are paler below with the hairs on the basal leg segments white 
and on the venter white or golden-white, although. in Iowa, Indiana, Okla- 
homa and Arkansas, individuals are variable so that all combinations can 
be found. One male was found that had all white pubescence | Illinois, “Air- 
port Region” Peoria, VU-20-41 (14) |. 

On the male genital capsule, the inner lobe is variable and usually similar 
to that of striata. There is a tendency for the rounded portion to slope off 
sooner at each side of the apex, whereas in striata it is more broadly rounded. 
The finger-like process is variable in length although it is rarely as long as 
in gratiosa or striata. 

Distribution. From the eastern Appalachian Mountains, Maryland to 
Georgia, westward to about the 97th parallel, from southeastern Minnesota 
and Wisconsin southward to northeastern Texas and Arkansas (Map: Fig. 
90). Detailed data are omitted for areas where there are numerous localities 
(see Methods), but Figure 90 shows all localities. 

More than 300 males and 2,300 females were seen: ALABAMA: Cull- 


596 Tue University ScIENCE BULLETIN 


man Co.,1 & (July); Jefferson Co.,1 6 (August). ARKANSAS: (Fig. 90). 
GEORGIA: Clark Co,1 6,1 2 (June); Cobb Co. 1 2 (July); Fulton} 
Co.,3 é (June); McDuffie Co.,1 2; Meriwether Co. 1 2 (July); Polk 
Co. 1 2 (May); Rabun Co, 1 8,1 2 (June-July); “Head River” 1 9} 
July). ILLINOIS: (Fig. 90). INDIANA: Harrison Co. 1 2 (July}R) 
Lake Co. 1 2 (August); Spencer Co. 1 2 (September); Tippecanoe Co., 
6 6, 102 @ (April-September); Warren Co, 1 6, 22 (July-August). 
IOWA: Fremont Co.,1 2 (July); Louisa Co,1 2 (June); Story Co, 5m 
(May). KANSAS (see Fig. 90). KENTUCKY: Graves Co. 1 6 (June 
LOUISIANA: St. Landry Parish, 1 6,5 9. MARYLAND: Montgomery 
Co. 6 2 (July). MICHIGAN: Lenawee Co. 1 & (September). MIN-| 
NESOTA: Fillmore Co. 2 2 (May); Houston Co. 25 2 (May); Ee 
Sueur Co.,1 % (August); Olmsted Co, 1 8 (July); Ramsey Co. 1 9) 
(May). MISSOURI (see Fig. 90). NEBRASKA: Cass Co. 13 2 (May-! 
July); Douglas Co.1 2 (August); Lancaster Co.1 2 (July); Otoe Co., 
4 2 (May, August); Richardson Co, 1 2 (uly); Saunders Co. 3.2m 
(May); “Child’s Point”, 2 @ (July). NORTH CAROLINA: Haywood 
Co, 1 8,5 2 (May, July-August); Rutherford Co, 1 2 (June); Swain 
Co.,4 2 (April, June). OHIO: Lawrence Co.,1 6,1 2 (August); Wash- 
ington Co.,1 2 (June). OKLAHOMA (see Fig. 90). PENNSYLVANIA: 
Delaware Co..1 2 (June). SOUTH CAROLINA: Greenville Co., 18 
(August). TENNESSEE: Knox Co. 1 4,2 2 (May, August); Limcota 
Co.,1 2 (April); Montgomery Co.,1 2 (July); Sevier Co, 7 2 (ulyge 
Shelby Co, 2 6,7 2 (June). TEXAS: Bowie Co..3 9 (March); Fama 
Co., 2 2 (May); Hunt Co, 7 2 (March-June); Lamar Co.) cea 
(June); Nacogdoches Co.,2 6 (October); Red River Co. 1 2 (April); 
Tarrant Co, 1 6 (June). VIRGINIA: Botetourt Co, 4 $,3 2 (June); 
Fairfax Co, 1 6,14 9 (March, May-August); Fauquier Co, 1 6,29 
July); Fredrick Co.,2 2 (May); Prince William Co.,1 2 (July); “Bar- 
croft”, 11 2 (May-July, September). WISCONSIN: Dane Co.,1 6, 12 $ 
(May-August); Grant Co, 2 & (July); La Crosse Co, 1 2 (August); 
Pierce Co.,1 6,4 2 (July-August); Vernon Co.5 6,3 2 (July-August); 
Winnebago Co.,1 2? (August). 

This species seems to be most common in eastern Oklahoma and Kansas, 
and throughout Arkansas, Missouri and Illinois. Although it does range east 
of the Mississippi River as far as the eastern slopes of the Appalachian chain, 
the populations apparently decrease in numbers. This is not entirely due to 
lack of collecting since ample specimens of striata have been obtained from 
many of these areas, but rather seems to reflect an actual thinning out of the 
species. One male was taken on the southern border of Michigan but speci- 
mens have not been taken further north in this state in spite of intensive col- 
lecting. South of Michigan the apparent gap could be due to inadequate 


SYSTEMATICS OF THE GENUS Augochlorella Norru or Mexico — 597 


collecting in northeastern Indiana and northwestern Ohio. In the West, the 
abrupt line at about the 97th parallel reflects the decrease in rainfall and 
therefore corresponding changes in edaphic and vegetational conditions in 
this region. 

Sandhouse (1937) limited the range of the species (using the name 
aurata) to south of 42 degrees north latitude even though she saw speci- 
mens from Minnesota, north of this line. Specimens recorded by her and 
others from Colorado and New Mexico are now recognized to be striata 
and neglectula. The Floridian specimens recorded by Sandhouse as aurata 
are the true aurata of Smith, 1853. 

The distribution given by Mitchell (1960) was largely taken from the 
literature and reflects the complex errors in identification and synonymy. 
For example, J. B. Smith (1910) and Viereck (1916) record “aurata” and 
“persimilis (similis Robt.)” from New York, New Jersey and Connecticut. 
It is uncertain what they were regarding as auwrata, but their persimilis was 
undoubtedly the small striata form c that is occasionally found in these states 
or small individuals of other forms of striata. 

The report by Rau (1922) concerning the nests of A. similis in a log refers 
to Augochlora pura. But perhaps the most complex error was made by Dreis- 
bach (1945) who refers to “Augochlorella neglectula (=A. aurata Sm.)” as 
occurring in Michigan and gives as his reference Titus 1901, who referred to 
this species under the name A. similis. These specimens are not aurata, 
neglectula or similis (= persimilis) but are undoubtedly striata, essentially 
the only Augochlorella found in Michigan. 

Seasonal Activity. A. persimilts is active from early April to about mid- 
October although nesting takes place only from the end of April to about 
the middle of August. Males start appearing with the emergence of the 
first brood at the end of May and can be found on flowers until the first frost 
in the fall. Although there is division of labor in colonies of this species, 
morphological castes cannot be distinguished. For details of the biology of 
persimilis, see Ordway (in press, a, b). 

Flower Records. Achillea, Agastache, Ailanthus, Alisma, Althaea, Am- 
mannia, Amorpha, Antennaria, Anthemis, Aphanes, Apocynum, Arabis, 
Asclepias, Asparagus, Aster, Barbarea, Bidens, Blephilia, Borago, Brassica, 
Callirhoe, Camassia, Campanula, Capsella, Cardamine, Cassia, Ceanothus, 
Celastrus, Cerastium, Chrysanthemum, Chrysopsis, Cichorium, Cirsium, 
Citrullus, Claytonia, Convovulus, Coreopsis, Cornus, Cotoneaster, Crataegus, 
Cucurbita, Daucus, Descurainia, Diospyros, Echinacea, Erigeron, Erysimum, 
Eupatorium, Euphorbia, Fragaria, Geranium, Geum, Gnaphalium, Gutier- 
rezia, Hedeoma, Helenium, Helianthus, Heliopsis, Heterotheca, Heuchera, 
Houstonia, Hypoxis, Ipomoea, Justicia, Kolkwitzia, Krigia, Lepidum, Les- 
pedeza, Lippia, Lobelia, Lotus, Ludwigia, Lycopus, Malva, Medicago, Meli- 


598 Tue University ScIENCE BULLETIN 


lotus, Monarda, Nepeta, Nothoscordum, Oenothera, Oxalis, Paeonia, Paro- 
sela, Parthenium, Passiflora, Pastinaca, Petalostemum, Phacelia, Plantago, 
Polemonium, Polygonum, Polytaema, Potentilla, Prunus, Psoralea, Pycnan- 
themum, Ranunculus, Raphanus, Rhus, Rorippa, Rosa, Rubus, Rudbeckia, 
Sabatia, Sagittaria, Salix, Salvia, Senecio, Sida, Silphium, Stsyrinchium, 
Smilacina, Smilax, Solidago, Specularia, Stellaria, Symphoricarpos, Taenidta, 
Taraxacum, Thaspium, Tradescantia, Trifolium, Valerianella, Verbena, 


Verbesina, Vernonia, Veronica, Virburnum, Zigadenus, Ziza. 


Fic. 90. Distribution of A. persimilis. 


Lees 
a 


SysTEMATICS OF THE GENUS Augochlorella Nortu oF Mexico — 599 


Augochlorella striata (Provancher) 


Augochlora striata Provancher, 1888, Additions et Corrections au Volume II de la 
Faune Entomologique du Canada, traitant des Hyménopteres 2:317-318 
(descr.); Dalla Torre, 1896, Catalogus Hymenoptorum 10:96 (list); Procter, 
1938, Biological Survey of the Mount Desert Region, part VI, p. 443 (list); 
Procter, 1946, Biological Survey of the Mount Desert Region, part VII, p. 
506 (list); Evans & Lin, 1959, Wasmann Jour. Biol. 17:120, 123, 127, 131 
(biol.). 

Augochlorella striata: Sandhouse, 1937, Jour. Washington Acad. Sci. 27:70 (tax.); 
Procter, 1938, Biological Survey of the Mount Desert Region, part VI, p. 443 
(list); Lovell, 1942, Kentucky Acad. Sci. Trans. 10:20-22 (key, descr.); Dreis- 
bach, 1945, Michigan Acad. Sci. Arts and Letters, paper 30, p. 225 (distr.); 
Procter, 1946, Biological Survey of the Mount Desert Region, part VII, p. 506 
(list); Moure, 1950, Dusenia 1:310 (key); Stephens, 1951, North Dakota Agr. 
Exper. Sta. Bull. 14:63 (list); Mitchell, 1960, Bees of the Eastern United States 
1:461 (tax.); Michener & Wille, 1961, Univ. Kansas Sci. Bull. 42:1130 (biol.); 
Knerer & Atwood, 1962, Proc. Ent. Soc. Ontario 92:174 (dist., f1., biol.); 
Sakagami & Michener, 1962, Nest Architecture of Sweat Bees 1-135 pp. 
(biol.); Eickwort & Fisher, 1963, Ann. Ent. Soc. Amer. 56:350 (descr.); Ord- 
way, 1964, Jour. Kansas Ent. Soc. 37:139-152 (biol.); Judd, 1964, Canad. Ent. 
96:1475 (fl.); Evans, 1964, Psyche 71:142, 147 (biol.); Michener, 1964, Am. 
Zool. 4:233 (biol.); Ordway, 1965, Insectes Sociaux 12:291-308 (biol.); Ord- 
way, in press, Jour. Kansas Ent. Soc. (biol.). 

Augochlora (Augochlorella) striata: Michener, 1951, in Muesebeck, et al. U.S. 
Dept. Agr., Agr. Monogr. 2:1126 (list); Montgomery, 1957, Proc. Indiana 
Acad. Sci. 66:132 (list, fl.). 

Augochlora matilda Robertson, 1893, Trans. Amer. Ent. Soc. 20:147 (descr.); 
Dalla Torre, 1896, Catalogus Hymenoptorum 10:95 (list); Cockerell, 1922, 
Proc. U.S. Nat. Mus.'60:16 (list). 

Augochlora confusa Robertson, 1897, Trans. Acad. Sci. St. Louis 7:324 (descr.); 
Bridwell, 1899, Trans. Kansas. Acad. Sci. 16:210 (list); Cockerell, 1899, Ent. 
News 10:3 (list); Titus, 1901, Canad. Ent. 33:134 (descr.); Cockerell, 1902, 
Amer. Nat. 36:811, 816 (descr., biol.); Cockerell, 1906, Trans. Amer. Ent. Soc. 
32:295 (list); Lovell, 1908, Psyche 15:40 (list); Cockerell, 1911, Canad. Ent. 
43:391 (list); Graenicher, 1911, Bull. Public Mus. Milwaukee 1:234 (list); 
Crawford, 1913, Canad. Ent. 45:271 (list); Cockerell, 1915, Pomona Jour. Ent. 
Zool. 7:232 (descr.); Stephens, 1921, Canad. Ent. 53:68 (list); Rau, 1922, Trans. 
Acad. Sci. St. Louis 24:32 (biol.); Hendrickson, 1930, Iowa State Coll. Jour. 
Sci. 4:162 (list); Phillips, 1933, Jour. Agr. Res. 46:860 (list); Michener, 1937, 
Ann. Mag. Nat. Hist. (10)19:314 (descr.); Brimley, 1938, Insects of North 
Carolina p. 455 (list). 

Augochlora (Oxystoglossa) confusa: Robertson, 1902, Canad. Ent. 34:247 (key). 

Oxystoglossa confusa: Hart & Gleason, 1907, Bull. Illinois State Lab. Nat. Hist. 
7:256 (list); Robertson, 1928, Flowers and Insects pp. 1-221 (f1., list); Pearson, 


600 Tue University ScreNcE BULLETIN 


1933, Ecol. Monogr. 3:386, 396, 416, 417 (biol.); Procter, 1938, Biological 
Survey of the Mount Desert Region, part VI, p. 443 (list). 

Halictus (Oxystoglossa) confusus: Viereck, 1916, Connecticut Geol. Nat. Hist, 
Survey Bull. 22:701, 703, 705 (key, list); Britton, 1920, Connecticut Geol. Nat. 
Hist. Survey Bull. 31:342 (list); Leonard, 1926, Cornell Univ. Agr. Sta. Mem. 
101:1025 (list). 

Augochlora coloradensis Titus, 1901, Canad. Ent. 33:133 (descr.); Cockerell, 1911, 
Canad. Ent. 43:390 (list); Hicks, 1931, Canad. Ent. 63:176 (biol.); Cockerell, 
1934, Amer. Mus. Novitates 697:1 (list). 

Augochlora confusa coloradensis: Cockerell, 1906, Trans. Amer. Ent. Soc. 32:295 
(list); Cockerell, 1907, Univ. Colorado Studies 4:243 (list); Cockerell, 1915, 
Ann. Mag. Nat. Hist. (8)15:269 (descr.); Cockerell, 1928, Univ. Colorado 
Studies 16:101 (list). 

Augochlora pseudopurella Strand, 1914, Archiv. Naturg. 80:163 (list). 

Augochlora aurata: Evans & Lin, 1959, Wasmann Jour. Biol. 17:120, 123) tam 
(biol.). 

Augochlora pura: Robertson, 1893, Trans. Amer. Ent. Soc. 20:146; Robertson, 
1894, Trans. Acad. Sci. St. Louis 7:436-480 (in part) (f.) (misidentifications). 

Augochlora neglectula: Titus, 1901, Canad. Ent. 33:134; Cockerell, 1928, Univ. 
Colorado Studies 16:101; Dreisbach, 1945, Michigan Acad. Sci. Arts & Letters, 
paper 30, p. 225 (misidentifications ). 

Augochlora similis: Titus, 1901, Canad. Ent. 33-134 (misidentification). 


Oxystoglossa similis: Britton & Viereck, 1906, in 29th Ann. Rept. Connecticut — 


Agr. Exper. Sta., New Haven, 1905, part 4, p. 212 (misidentification). 

Hfalictus (Oxystoglossa) persimilis: Viereck, 1910, in Smith, Ann. Rept. New 
Jersey State Mus. 1909:688; Viereck, 1916, Connecticut Geol. Nat. Hist. Surv. 
3ull. No. 22, 5:701, 703, 705; Britton, 1920, Connecticut Geol. Nat. Hist. Surv. 
Bull. No. 31, p. 342 (misidentifications ). 


Halictus (Augochlora) auratus: Viereck, 1910, in Smith, Ann. Rept. New Jersey | 


State Mus. 1909, p. 688 (misidentification ). 
Halictus (Oxystoglossa) purus: Vachal, 1911, Misc. Ent. 19:50, 53, 111 (mis- 
identification ). 


Types. Augochlora striata, female lectotype, male lectoallotype, from 
Quebec, Canada, are in the collection of Laval University, Department of 


Biology, Ste. Foy, Quebec, Canada. These specimens have been carefully — 


compared by Dr. René Beiqué, Curator of Entomology at Laval University, 
with specimens I submitted. Dr. Beiqué’s careful examination and illustra- 
tion clearly show that these specimens are typical of striata of eastern Canada 
and typify form a of the discussion below. According to Dr. Beiqué (per- 


a = oe oe 


sonal communication), these types are the only two specimens of this species — 


in the Provancher collection, although the original series contained two 
females and four males. The lectotypes were labeled (but not published) by 
Mr. Noel Comeau, the former curator of the Provancher collection, in 1941, 
and are labeled as follows: female specimen No. 119: with a small yellow 


SysTEMATICs OF THE GEeNUs Augochlorella NortH or Mexico — 601 


Rebel marked 1475 (Provancher number), a white label with red border 
bearing the identification in Provancher’s handwriting, a red label marked 
lectotype with the identification, Comeau’s signature, dated 1941 and No. 
119; male specimen No. 120: with a small white label with é sign, a yellow 
label with Provancher’s number 1475 A, and a purple label marked Allotype, 
No. 120, with Comeau’s signature and dated 1941. The lectotype designa- 
tion is here published for the first time. The location of the rest of the syn- 
type series is not known. 

— Augochlora confusa, \ectotype female No. 927 (Robertson’s number) is 
from Carlinville, Macoupin Co., Illinois, 1886, and is in the Robertson Col- 
lection at the Illinois Natural History Survey. This specimen was selected 
from among 33 females of Augochlorella striata in the syntype series. Eigh- 


teen other females in the series are Augochlora pura. 1 have also seen seven 
male syntypes, all striata, and similar in appearance. I have not seen one 
female and four males of the original species. It seems certain that Robert- 
-son’s description was based upon the Awgochlorella striata and not the Augo- 


chlora pura. This lectotype designation is also published here for the first 


time. 
Augochlora coloradensis, \ectotype female, Ft. Collins [Larimer Co.}, 
Colorado, June 13, 1899, is at the U.S. National Museum. It has a red U.S. 


National Museum cotype label No. 19459, and an identification label by 
Titus. I have selected this specimen as the lectotype because it agrees as well 
as any with the description, is in good condition, and will be located at the 
‘same museum as other Titus types. Other known female syntypes are 


] 


| 


located at the Museum of Comparative Zoology (1 specimen), U.S. National 
Museum (1 other specimen), Purdue University (2 specimens), University 
of Kansas (1 specimen), and Colorado State University (6 specimens). This 
species was originally described from numerous females and two males. I 
have not seen the males or other females, if any. 

Augochlora matilda, lectotype female, No. 12247 (Robertson’s number), 
from Inverness, Citrus Co., Florida, 1892, is in the Robertson Collection at 


the Illinois Natural History Survey. I have seen only one of the two syntypes 


and here designate it as the lectotype. 

Augochlora pseudopurella Strand was proposed for Halictus purus Va- 
chal (not Say). It does not seem likely that Strand designated a holotype 
from among the “numerous” specimens which Vachal misidentified as H. 
purus Say. I have not seen these specimens which are in the Museum Na- 
tional d’Histoire Naturelle de Paris, but Pe. J. S. Moure (personal communi- 
cation) has verified that they are Augochlorella and not Augochlora. The 
specimens from Canada, the northeastern United States and possibly Louisi- 
ana would be striata; those from Orizaba and Oaxaca, Mexico, are probably 
neglectula. 


602 Tue Universtry SciENcE BULLETIN 


This species consists of highly variable, intergrading groups of individuals, 
In order to describe and discuss the variation, four forms (a-p) have been 
recognized. The variability is such over most of the range that no definite 
line can be drawn between the four groups and therefore it is not always 
possible to assign certain specimens to any of the groups. These unplaced 
individuals are called “s”. Although detailed studies were made of variations 
among the males, it seemed impractical to characterize the forms, so that 
many of the males are therefore assigned to group s. The following discus- 
sions of the forms concern only females unless specific reference is made to 
males. Biological information from Kansas (see Biosystematics) indicates 
that there may be at least two populations or species, but until further eco- 
logical and behavioral data are obtained, there is little justification for recog- 
nizing more than one species. However, when possible, I have kept the in- 
formation concerning the four groups separated in the following discussion 
and records, in the event that further biological work substantiates the hints 
that there may be sibling species involved. 

In discussing regional variation among the different forms, the specimens 
are compared with a standard.” This is a specimen of each form chosen 
from an area where the form is usually distinct, where there are few, if any, 
intergrades, and where the majority of specimens look alike. The term 
“standard” refers only to these specimens in the following discussions. The 
standard of form a is from Philadelphia, Pennsylvania, [X-24-14 (38); that 
of form B is from Alleghany Co., North Carolina, along Little River, nr. 
Eunice, VII-26-28-57 (R. Baileys & C. F. Walker) (25); that of form c¢ is 
from Lee Co., Iowa, VI-28-29- (Parks) (6); and that of form p is from 
Colorado Springs, Colorado, VI-6-52 (W. E. LaBerge) (20). 

The type of striata is form a. Robertson’s matilda is also form a but is 
too coarsely sculptured to be typical except in Florida. The type of Robert- 
son’s confusa is typical of form c. Form B is morphologically between a and 
c. In some areas it intergrades with form a, in other areas with form c. 
Occasionally there is a continuum from a to c but usually 8 is entirely sep- 
arable from both a and c. The type of coloradensis Titus is a representative 
of form p. In some areas this form appears to be a variable intergrade be- 
tween a, B and c but in other areas it is quite distinct. 

The following descriptions apply to all forms of striata except as noted. 

Description. Female: (1) Length 5 to 8 mm; head width 1.55 to 2.82 
mm, averaging 1.96 mm, width greater than length. (2) Color varying re 
gionally from blue-green to yellow-green; frons without bluish reflections 
in green specimens; metasoma similar in color to head and thorax. (3) Man- 
dible with basal third dark brown, reddish brown centrally, rufous at tip; 
without green reflections basally. (4) Clypeal length equal to, or slightly 
greater than width; basal part green with large, irregularly spaced punctures, 


SysTEMATICS OF THE GENUs Augochlorella Nortu or Mexico 603 


smaller and closer basally; apical one-fourth to one-half brown, slightly 
beveled, with elongate punctures or irregularities; surface between punctures 
smooth and shiny. (5) Supraclypeal area variously punctate with surface 
between punctures smooth or roughened. (6) Paraocular area punctorugose 
to rugose below antennae, coarsely rugose above antennae. (7) Antenna 
dark brown, flagellum slightly lighter below than above, pedicel as long as 
broad, first flagellar segment slightly wider than long; pedicel longer than 
but equal in width to first flagellar segment. (8) Scutum coarsely punctate 
to rugose (some form a only) medially, becoming more coarsely rugose be- 
tween parapsidal lines; anterior margin smoothly roughened at midline, 
becoming coarsely areolate laterally. (9) Tegula about 1.5 times as long as 
wide. (10) Scutellum finely and irregularly roughened, without distinct 
punctures. (11) Pleuron rugose, becoming areolate anteriorly. (12) Propo- 
deum with dise variable in size, shape and sculpturing; length equal to meta- 
notum (some form a) or more usually, longer than metanotum, rarely more 
than twice metanotal length; outline of disc sharply to roundly bracket- 
shaped (forms a, p) to deeply roundly or obtusely V-shaped (forms a, B) to 
broadly U-shaped or semicircular (form c), profile types 1-4; posterior edge 
sharp or weakly carinate (form a), abruptly rounded and thickened (forms 
B, D) to smoothly and gradually rounded and indistinct (form c); striae 
usually distinct, slightly irregular or straight, usually reaching edge pos- 
teriorly; posterior vertical surface coarsely and deeply roughened (some form 
A only) to smooth, shiny and granular; posterolateral corners finely granular 
to finely roughened (forms 8, c, p), to strongly roughened (forms a, B) or 
rugose (form a); lateral vertical surface finely to coarsely (some form A 
only) rugose or reticulated. (13) Legs brown; coxae bright green, tro- 
chanters and femora usually with weak metallic reflections. (14) Metasomal 
terga with apical margins narrowly, often inconspicuously margined with 
brown; first tergum with anterior portion polished, sparsely and finely punc- 
tate, dorsal surface variously punctate with punctures minute and inconspic- 
uous to large, and widely to closely, regularly to irregularly spaced; second 
tergum with punctures similar to first but with punctures closer; first meta- 
somal sternum with or without weak metallic reflections. (15) Pubescence 
golden on dorsum and legs and ventrally on metasoma; golden to white 
ventrally on thorax and head. 

Male: (1) Length 6 to 8 mm; head width 1.63 to 2.13 mm, averaging 1.85 
mm, usually equal to or less than length, rarely wider than long. (2) Color 
yellow-green to dark blue-green, usually bright shiny green, frons without 
blue reflections on green specimens, usually uniformly colored over entire 
body. (3) Mandible with or without metallic reflections basally. (4) Clypeus 
with punctures variables in size and number, separated by about their own 
diameters, surface between punctures usually smooth and shiny. (5) Supra- 


604 THe University SCIENCE BULLETIN 


clypeal area protuberant, variably punctate, with surface between punctures | 


|| 
i 


roughened or somtimes smooth and shiny at least basally. (6) Paraocular 
area with small close punctures below level of antennae, minutely but deeply — 
punctorugose above level of antennae. (7) Flagellum dark brown above, — 
yellow-brown below; scape and pedicel dark brown with yellow apical area 
below; width of pedicel and first flagellar segments variable, each averaging — 
1.5 tmes as wide as long. (8) Scutum shiny with punctures variable in size | 
and spacing; anterior margin and anterolateral angles areolate, smoother 
anteromedially. (9) Tegula about two times longer than wide. (10) Scutel- | 
lum with surtace irregular, punctate to rugose; punctures, when present, dis- 
tinct to indistinct, irregular in size and spacing. (11) Pleuron punctate to 
rugose, becoming areolate anteriorly. (12) Propodeum with disc equal to or | 
slightly longer than metanotum; outline of disc varying from distinctly 
bracket-shaped to obtusely U-shaped or semicircular, posterior edge varying | 
from sharply angulate and prominent to gradually rounded; striae fine to 
coarse, regular to irregular or branched, straight to wavy, widely separated to — 
close together, usually reaching edge posteriorly, or slightly before when edge 
of disc gradually rounded, reaching edge laterally; posterior vertical surface | 
minutely to finely rough; posterolateral corners with or without subhorizon- 
tal rugae extending from lateral to posterior faces; lateral vertical surface 
irregular, rugose with weak lineate rugae perpendicular to anterior and 
ventral edges. (13) Legs brown, fore and hind coxae, trochanters and femora 
green, tibiae dark brown, with greenish reflections at least on anterior side of 
hind leg, usually yellow-brown apically and basally; tarsi brown; hind basi- 
tarsus with erect hairs of uniform length, equal to or longer than basal hairs, 
length variable, not exceeding 1.5 times width of segment; basal tuft present. 
(14) Metasomal terga green with brown apical margins; first tergum polished 
with widely scattered, fine punctures anteriorly, smooth but less shiny dor- 
sally, punctures variable in size and spacing; second tergum with punctures 
variable in size, denser than those of first, indistinct on third and following 
terga, surface minutely reticulated in appearance, pubescence fine, short to 
long depending on wear; first sternum usually with metallic reflections vari- 
able in intensity; fourth sternum distinctly but weakly emarginate. (15) 
Pubescence white to golden, usually golden dorsally, white ventrally with 
long golden hairs and short white hairs on head and ventral part of abdo- 
men; golden on tibiae and tarsi, white on coxae, trochanters and femora. 
(16) Genital capsule, sterna 7 and 8 and tergum 8 of type 1 (Figs. 32, 40, 43). 


Forms 


The following accounts describe the “standard” individual of each form 
and a series from the same locality. They do not include total variation of 


SYsTEMATICS OF THE GENUS Augochlorella Nortu oF Mexico — 605 


the form or attempt to describe intergrades among the forms. The males 
were described only from specimens that could be definitely placed as to 
form and were usually from the same areas as the females. 


A. striata form A 


Female: Disc sharply bracket-shaped to obtusely V-shaped (Figs. 59, 60), 
shorter to slightly longer than metanotum, these sclerites usually about equal 
in length; length of disc at posterolateral corners as long as length postero- 
medially; edge of disc weakly carinate to sharply defined posteriorly (Fig. 
21), becoming rounded laterally; disc pointed or sharply V-shaped medially 
and depressed onto posterior vertical surface. Striae variable, straight, thick 
and well defined to irregular, branched and close, or thin, fine and very 
close; always reaching well defined edge. Posterior vertical surface of pro- 
podeum smoothly granular to coarsely roughened or rugose. 

Male: Disc usually as long as metanotum or only slightly longer, with 
distinct, usually sharp edge, often bracket-shaped; striae well defined, regu- 
lar but wavy, reaching edge; posterior surface of propodeum variable, usually 
uneven, may be rather smooth to rough. Hind basitarsus with hairs appear- 
ing short and sparse, only slightly longer than width of basitarsus, contrast- 
ing only slightly in length with basal tuft. 

All striata with sharp bracket-shaped discal areas belong in this group. As 
the bracket-shape and edge become rounded it is less easy to recognize this 
form. This form grades gradually into forms 8, c and p as the disc becomes 
more rounded in shape and rounded along its edge. The outline of the disc 
is similar to that of gratiosa and females may look similar to gratiosa and 
aurata when the striae are fine. 


A. striata form B 


Female: Disc longer than metanotum, up to twice as long, obtusely V- 
shaped, longer medially than laterally, with medial portion of V rounded and 
sometimes extending onto posterior vertical surface (Fig. 63), edge distinct, 
often thickened, rough and abrupt but rounded (Fig. 23), not ridged or 
carinate. Striae large, irregular and branched, reaching edge at all points; 
posterior vertical surface of propodeum finely and regularly granular (Fig. 
74), lateral vertical surface rugose. 

Male: Disc long medially, obtusely V-shaped, up to twice as long as 
metanotum, with edge thickened but rounded and often roughened; striae 
wavy but regular, reaching edge; posterior surface of propodeum usually 
shiny but rough to smooth, punctured or finely rugose, variable throughout 
range; hind basitarsal hair dense, usually of more or less uniform length, 


606 Tue University SCIENCE BULLETIN 


about twice as long as basal tuft; hind basitarsus usually appearing large, 
with long, dense hair. | 


These bees are usually large and light green. They grade into forms a, ¢ 
and p in certain areas but are predominant and most similar to the standard — 


in the southeastern part of the range. 


A. striata form c 


Female: Disc large, up to twice length of metanotum, broadly U-shaped 
(Fig. 62); edge of disc indistinct, smoothly and gradually rounded from 
vertical to horizontal plane; striae large and distinct, straight or irregular, 
wavy and branched, sometimes widely separated; striae ending gradually at 
indistinct edge of disc; surface between striae shiny and smooth or minutely 
reticulated or minutely roughened; posterior vertical surface of propodeum 
evenly granular (Fig. 74), lateral vertical surface rugose. 

Male: Disc broadly and deeply U-shaped, longer than metanotum; edge 
of disc smoothly rounded, shiny; striae straight, often widely separated with 
surface shiny between, usually reaching edge or ending gradually just before 
edge; posterior surface of the propodeum shiny but uneven or roughened 
and irregularly and minutely punctured, often rough; hind basitarsus with 
hairs long but sparse, distinctly longer than basal tuft. 

There is wide variation in size in this form but the most distinctive or 
characteristic bees of this group are large. Small individuals look similar to | 
A. persimilis, especially if the striae fade out before the edge of the disc leav- 
ing a shiny area between the striae and the edge. These bees are most preva- | 
lent in the northern part of the range and are most similar to the standard in 


; 


Iowa and Illinois. 
A. striata form bp 


Female: Propodeal disc equal to or usually slightly longer than meta-_ 
notum, obtusely U-shaped to weakly bracket-shaped (Fig. 71), edge distinct, 
often slightly thickened and rough or uneven, sometimes extending medial- 
ly onto posterior surface as indistinct or rounded V; striae large, irregular or 
vermiform, usually reaching edge posteriorly; posterior vertical surface of 
propodeum smooth and finely granular (Fig. 74); lateral vertical surface 


coarsely reticulate to rugose. 

Male: Disc similar to those of female but with smoother, rounded pos- 
terior edge and often straighter striae; edge slightly extended medially but 
rounded rather than V-shaped; posterior vertical surface of propodeum 
shiny, shallowly punctured to weakly rugose, finely and linearly rugose or 
punctorugose over posterolateral corners and on lateral vertical surface; 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico — 607 


hairs of hind basitarsus less than twice as long as width of segment but 
appearing long, rather dense and distinctly longer than basal tuft. 

These bees are usually bright green to dark green. They intergrade with 
forms a, B and occasionally with c. They are predominant and most similar 
to the standard in the western part of the range from Texas to the Dakotas. 

Comparisons. A. striata is the most widely distributed and morphological- 
ly diverse of all the North American species of Agochlorella. It overlaps 
the ranges of all the species north of Mexico except pomontella and inter- 
grades morphologically, at least in the females, with these species. It fre- 
quently is the largest and most coarsely sculptured of the eastern species but 
due to the wide variability in size, cannot always be distinguished by these 
features. 

In the southeastern region most specimens of striata are distinctive al- 
though gradations toward aurata and persimilis do exist. Males of striata a, 
in particular, are easily confused with those of awrata, and although the key 
separates the two, without biological information I am uncertain whether 
the separation represents a valid difference between populations or merely an 
artificial or arbitrary dichotomy. Females of striata a may be separated from 
aurata and gratiosa where ranges overlap by their generally coarser striae, 
the rougher sculpturing and other key characters. 

A few specimens in the south and southeast and many in the central re- 
gion that are small individuals of striata c or possibly p, intergrade complete- 
ly with persimilis, so that differentiation of females cannot always be certain. 
In these individuals the body size, the characters of the disc, and the body 
sculpturing all resemble those of persimilis. 

In Texas, some striata p may resemble bracteata in the characters of the 
disc but in this region most striata are larger and more rugose than bracteata, 
and the two species should not be confused. 

There seems to be little if any intergradation with neglectula where the 
ranges overlap in southeastern New Mexico. A. striata (mostly p) are usually 

less rough on the posterior vertical surface of the propodeum than neglectula 
and can be easily separated by the key characters. 
Variation (all forms). Body size and head width in both males and 
females varies considerably throughout the range (Fig. 86), with the largest 
‘specimens (8 mm) occurring among Floridian specimens of form a and the 
smallest (6 mm) among New Mexican specimens of form p. Small worker- 
like individuals were found in all forms usually during summer months, the 
small individuals of form c usually intergrading with, or becoming indistin- 
guishable from, persimilis. 

Body color varies regionally in both sexes with dark blue-green individuals 
found chiefly in Florida and New Mexico. Elsewhere, throughout the range, 
“most specimens are a bright green but may range from yellow-green to blue- 


608 Tue University SciENcE BULLETIN 

green. The different regions of the body of any one specimen are similar in 
coloration, so that striata is more uniform in color than other eastern species. 
The mandibles of females are usually brown as described; however some 
Floridian specimens do have weak greenish reflections at the bases. 

The supraclypeal area in both sexes is variously punctate. In females, the 
surface between punctures is usually shiny and smooth but may be minutely 
roughened or weakly rugose as in some Floridian specimens. In males, the 
supraclypeal area is roughened but may be smoother basally than just below 
the antennae. 

The punctation patterns on the scutum of males can be divided into four 
groups, all groups occurring throughout the range but with one pattern often 
regionally predominant. This character appears to be more geographically 
variable than it is variable among forms and cannot be correlated with any 
of the other major characters. The following puncture patterns can be recog- 
nized: a) Punctures distinct and widely spaced centrally, separated by dis- 
tances equal to or greater than one diameter; closer laterally, distinct to al- 
most contiguous at the parapsidal lines. b) Evenly and widely spaced (sep- 
arated by two times their diameters) over entire scutum. c) Evenly and 
closely spaced (separated by distances, equal to or less than their own diame- 
ters) over entire scutum. d) Distinct but unevenly spaced medially, becom- 
ing very close and rugose just medial of the parapsidal lines. No such varia- 
tion is found in females. 

Males are less easily separated into forms than females on the basis of 
propodeal characters and show a wider range of variation so that in certain 
areas each of a number of individuals has a different combination of charac- 
ters (Florida, Kansas, Nebraska, etc.). Deviations from the descriptions of 
the four forms occur among the intergrades between different forms and 
among intergrades of striata and other species but it is not feasible to de- 
scribe all the variations found in the continuum. 

The legs show little color variation; paler specimens show less metallic 
coloration than darker ones. In males, the extent of metallic coloration on 
the outer surface of the hind tibia is individually variable. The hairs on the 
hind basitarsus of males, although similar in length for all forms, look longer 
and denser in form 8 with the hairs of the basal tuft appearing proportional- 
ly shorter than in forms a, c and p. Although differences do exist in these 
characters, they form a continuum and cannot be correlated with other 
characters. | 

The metasomal punctures are variable in size and spacing in both sexes, | 
but like scutal punctures can be divided into groups. Any one group may be 
regionally predominant (Indiana, Illinois, Michigan, etc.) or the punctation” 
may vary among individuals within any one region (Nebraska, Minnesota, - 
Texas, Connecticut, etc.). In males the punctures are always distinct, al 


: 
| SYsTEMATICS OF THE GENUs Augochlorella Nortu oF Mexico — 609 
| 


though regularly or irregularly spaced and separated by more than to less 
than their own diameters. In the females the punctures are frequently very 
small or inconspicuous but may also become large. Spacing varies from very 
close to widely scattered and regularly or irregularly spaced. Not all speci- 
mens show the greenish reflections on the first sternum, and in those that do, 
the amount or intensity is variable. This coloration occurs more often in 
males than in females, and there seems to be no correlation between this 
character and the form or region. 

Distribution. Southern Canada to southern Florida, westward to the 
Rocky Mountains. More than 6700 females and 1400 males were seen. Due 
to the abundance of this species throughout its range, locality data are indi- 
cated only by Figure 91 (see Methods). 

Seasonal Activity. A. striata is active from early April to about the middle 
of October throughout most of its range. Pollen collectors are found from 
the end of April through middle or late August and males occur from late 
May to late October. In the North the season is slightly shorter, lasting from 
the end of April to the end of September, whereas in the South the bees are 
active throughout the year with pollen collectors being found from early 
April to early September. 


= \ 
See 


Fic. 91. Distribution of A. striata. 


610 Tue University SCIENCE BULLETIN 


There is considerable variation in size in this species (Fig. 86) within all 
forms. Usually the small individuals are similar in structure and appearance 
to the larger females of the same form although somewhat less coarsely 
sculptured. There is a queen and usually one or more workers (individuals 
that do not lay eggs) in each nest, but these cannot always be distinguished 
morphologically. Where both large and small individuals are present in one 
nest, one of the large females is always the queen and the remaining large 
and small females are workers. All gradations in size may occur within one 
colony, or all bees of a colony may be approximately the same size. Small 
individuals (probably of form c) do sometimes found nests in the spring, but 
then all offspring are as small as the queen or smaller. For further details 
on the biology of this species, see Ordway (1965a; in press). 

Flower Records. Achillea, Aesculus, Agastache, Agoseris, Althaea, Ame- 
lanchier, Amorpha, Anemone, Anemonella, Antennaria, Apocynum, Aqut- 


Fic. 92. Distributions of forms of A. striata. (Black—form a; dots—form 8; gray shading— 
form c; cross-hatched—form pb; white—group s.) : 

Each circle includes each state it overlaps except where the state is divided by broken line. 
Circle No. 117 includes West Virginia, Kentucky, Tennessee only. Circle No. 276 does not 
include New Mexico. The numbers beside each circle represent the total number of specimens 
examined from the area represented by the circle. 


SysTEMATICS OF THE GENUs Augochlorella Nortu or Mexico 611 


legia, Arabis, Aralia, Argemone, Aruncus, Asclepias, Aster, Astragalus, Bar- 
barea, Berteroa, Bidens, Brassica, Callirhoe, Calopogon, Camassia, Camelina, 
Campanula, Capsella, Cardamine, Carduus, Cassia, Caulophyllum, Ceano- 
thus, Celastrus, Centaurea, Cephalanthus, Cercis, Chaerophyllum, Chrysan- 
themum, Chrysopsis, Cichorium, Cicuta, Circaea, Cirsium, Citrullus, Clay- 
tonia, Clethra, Convovulus, Coreopsis, Cornus, Crataegus, Crypthantha, 
Cryptotaenia, Cubelium, Cucumis, Cucurbita, Cunila, Daucus, Dentaria, 
Diervilla, Dodecatheon, Echinacea, Echium, Ellisia, Erigenia, Erigeron, 
Erysimum, Euonymus, Eupatorium, Euphorbia, Fragaria, Gaillardia, 
Geranium, Gerardia, Glechoma, Gnaphalium, Gossypium, Grindelia, 
Gutierrezia, Haplopappus, Hedeoma, Hetracium, Helenium, Helianthus, 
Heliopsis, Heracleum, Heterotheca, Heuchera, Hieracium, Houstomia, Hy- 
banthus, Hydrangea, Hydrocotyle, Hydrolea, Hydrophyllum, Hypericum, 
Impatiens, Inula, Ipomoea, Iris, Isopyrum, Kolkwitzia, Krigia, Lactuca, 
Lathyrus, Lepidium, Lespedeza, Lesquerella, Linum, Lippia, Lobelia, Lo- 
matium, Lonicera, Lotus, Lycopersicum, Lycopus, Lythrum, Malus, Malva, 
Medicago, Melilotus, Mentha, Mertensia, Muikania, Monarda, Myo- 
soton, Nigella, Nothoscordum, Oenothera, Onopordum, Opuntia, Osmor- 
hiza, Oxalis, Paeonia, Parthenium, Parthenocissus, Paspalum, Pastinaca, 
Penstemon, Perideridia, Petalostemum, Phryma, Physalis, Polemonium, 
Polygonatum, Polygonum, Polymnia, Polytaenia, Pontederta, Potentilla, 
Prenanthes, Prunella, Prunus, Psoralea, Ptelea, Pteridium, Pycnanthemum, 
Pyrrhopappus, Pyrus, Ranunculus, Ratibida, Rhamnus, Rhus, Ribes, Rortp- 
pa, Rosa, Rubus, Rudbeckia, Sagittaria, Salix, Salvia, Sanicula, Sapindus, 
Satureja, Scrophularia, Scutellaria, Sedum, Senecio, Sericocarpus, Sida, 
Silphium, Sisymbrium, Sisyrinchium, Smilacina, Smilax, Solanum, Solidago, 
Sonchus, Specularia, Sphaeralcea, Spiraea, Stellaria, Stokesia, Strophostyles, 
Symphoricarpos, Syringa, Taenidia, Tanacetum, Taraxacum, Teucrium, 
Thaspium, Tradescantia, Tragopogon, Trifolium, Trillium, Triosteum, 
Vaccinium, Verbascum, Verbena, Vernonia, Viburnum, Vicia, Viola, Vitis, 
Waldsteima, Xanthoxylum, Zizia. 


BIOSYSTEMATICS OF 4. STRIATA AND PERSIMILIS 


The biology of persimilis and striata is discussed in detail by Ordway 
(1965a; in press). Some of the results obtained by excavating 133 nests near 
Lawrence, Kansas, are of systematic importance and are discussed here since 
populations from these nests shed some light on the complexities of the inter- 
specific and intraspecific variation. Except as otherwise indicated, the follow- 
ing discussion relates only to females. 

About 21° of the nests contained unquestionable persimilis; 54°, were 
clearly striata (sJ.) and 25°% contained apparent mixtures of persimilis and 


612 THe Universiry SciENCE BULLETIN 


striata or individuals of uncertain identity intermediate between the two 
species. 

The 72 nests of striata contained pure colonies of forms B, c, or p individ- 
uals, or mixed colonies of both 8 and p individuals, mixed colonies of B-c or 
B-D intermediates or more usually, mixtures of B and B-c, B and B-D or ¢ and 
cp individuals. One nest was found to contain a-p intermediates. There 
were six nests containing recognizable striata of forms B, c, p, or their inter- 
mediates, as well as small, persimilislike individuals, but in all cases the 
small individuals looked more like small striata c than like persimilis. In 
contrast, two nests were found with small but clearly striata-like females as 
well as males of persimilis. 

Of the 25 nests containing persimilis-striata mixtures or intermediates, 20 
contained individuals intermediate between persimilis and striata and five 
contained apparent mixtures of both persimilis and striata. None of these 
nests contained males. 

Since males of striata and persimilis are easily distinguished, 30 females 
intermediate between the two species were brought into the laboratory and 
allowed to establish nests. Male progeny from these females were exam- 
ined after they emerged from the nests or as they were leaving. Of the 116 
males recovered throughout the summer, 94° were persimilis, indicating 
that most of the original females belonged to this species. Judging by the 
frequency and periodic appearance of the striata males, it appeared probable 
that they were produced by a single female. 

Although these data are fragmentary, they do serve to crystallize some of 
the problems involved. Within the species striata, it seems that forms 8 and 
p are not different biologically since both can be regularly found within a 
single nest population. Form c may be biologically distinct since no nests 
were found in which both c and another form coexisted, although c-p inter- 
mediates occur in nests with form c. 

The small specimens, intermediate between striata and persimilis that 
were found within the striata nests, were probably striata, and those in nests 
of persimilis were probably persimilis although the possibility that F: hybrids 
exist cannot be ignored. Male (presumably haploid) offspring of both striata 
and persimilis are produced by the intermediate females but it has not been 
established if any one female can produce both. 

So far, biological, behavioral or ecological differences have not been 
found between the forms of striata as they occur in eastern Kansas and such 
differences between the two species are only slight if extant. If significant 
differences are discovered or if methods could be found to keep progeny of 
various females segregated in the laboratory, it would be worth the time and 
effort to examine in much greater detail the interspecific and intraspecific 
variation in an attempt to define or categorize those individuals now being 


613 


‘syued OM} OUI paplArp useq sey UOrIZaI sty} a]qQeI sIy Jo sasodind ay) I0J x 


eS  eeeeE———————————————————EEE————E———————— 


90b 8°sT “LE 8'bb 0 a 0 0 C0 ae gis a a OOIxa/Al 
606 0 799 O17 0 0 0 0 AS ane tate es eS aso NN 
Z07'9 0 0 0 0 0 eTb 0 OC Ma Sie Serie Tenw29 
+86 0 0 ts PL 8°€Z re |bicd 109 Weer oe cee qanogs 
pls 0 0 0 A 0 0 9°82 Cae wee iene rig ele) fl 
102 0 0 0 el 0 rTl V9 (GANS rc ce aco #1SEOTINOS 
688°7 0 0 0 0 0 0 a) OO cognn e aarn aee ysvay ION 

P]Jero;psosny Pjejuspa = By Jatuowod =e N\Da]3au reine ejeaVIgq = stjtuutsiad ~—sesorneas PyeII}s Poly 
T2301, Zo 7 7 7 Zo 7 7 7 


"eaIY YY WIZ vpjasopy20snp Jo saidadg Jo sduUsIINIIQO ws Jog ‘7 ATAV, 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 


614 Tue Universiry SciENcE BULLETIN 


Tasce 3. Per Cent Occurrence of Forms of A. striata from Each Area. 


% te We So %o Total 
Area form a form B form c form D group § specimens — 
Northeast, 2222222 40.1 6.8 19.1 0 34.0 2,885 
Southeast®* 225. 46.4 35.5 15S 0.2 2.8 569 
Blorida** 96.2 0.3 0.3 0 Bs 312 
South ‘Central 1521 20.5 1.0 al 18.3 590 
North Central .... 7.8 13.8 28.4 28.7 PA. 3} 3,640 
West-Mexico ...... 0 Se 0 61.0 35.6 118 


* These are aurata-striata intergrades (probably striata form a). 
** Kor this table the southeastern region has been divided into two parts. 


classed as intermediates between the forms or species. We have here an 
excellent case of very similar, sympatric populations with no apparent eco- 
logical segregation. 


REGIONAL ACCOUNT OF THE SPECIES 


The area under consideration in this paper, including Canada, the United 
States and Mexico, has been divided into regions, as indicated for the Re- 
gional Keys, in order to discuss and compare variation within and among 


species. 

The relative abundance of each species within each region is shown in | 
Table 2, and the relative abundance of each of the forms of striata is given in 
Table 3 and Figure 92. The regions in the latter are subdivided to give a _ 
more accurate indication of the relative frequency of the different forms. 
Unless otherwise stated, the discussions of striata below concern only females. | 


NortTHEAst 


Species represented: A. striata a, B, c. 

A. striata is the only species known to occur widely in this region, al- 
though other species have been reported in the literature, presumably erro- 
neously, and occasional specimens of gratiosa or persimilis may occur along: 


the southern borders of this region. The three forms of striata are sometimes — 
difficult to distinguish, especially in this region, and therefore the percentage 
of group s (unplaced specimens) is comparatively large (Fig. 92). 

In the southern part of the region, as far north as New Hampshire and 
Vermont, form a is distinct and easily separated from B and c, though ex- 
tremely variable. The disc is seldom bracket-shaped but is, characterized by 
a sharp to carinate posterior edge, at least medially, with the median portion 
coming to a distinct point. Its length is usually equal throughout and its 


SysTEMATICS OF THE GENUS Augochlorella Nortu or Mexico 615 


shape is a broad but well defined V (Fig. 60). The length of the disc and 
the size and regularity of the striae are variable as is the size of individual 
specimens. Nearly all specimens of form a are weakly rugose or finely 
roughened on the posterior vertical surface of the propodeum. Large, coarse- 
ly rugose specimens resembling form a from Florida (Fig. 59) are occasion- 
ally found as far north as Massachusetts, but they have the light green color 
of northern specimens. In Massachusetts [ Revere, VII-28-92 (F. A. Eddy 
Coll’n) (22), 2 specimens; Needham, V-18-20 (F.X.W.) (22), 1 specimen] 
these large individuals contrast strikingly with other individuals of form a, 
but to the south they tend to intergrade with other form a variants and are 
less noticeable. In Maine and Canada nearly all specimens differ from the 
standard of form a in that the posterior edge of the disc is less sharply angu- 
Jate, the shape of the disc is more variable, usually somewhat rounder, and 
the median V is often only weakly indicated. 

In Canada and Maine, group s consists chiefly of individuals that have 
characteristics of all forms rather than resembling variants of any one form. 
In the rest of the northeastern region specimens of group s are mostly varia- 
tions of form a that do not agree with the standard either because they lack 
the median V or the angulate edge of the disc or because the shave of the 
disc is unlike that of the standard or the usual variants of form a from this 
region. A few scattered unplaced specimens resemble form p and a few may 
be variations of B or c. 

Form c is generally more common than g in the Northeast, especially to 
the north. Only in Pennsylvania is 8 more abundant. Specimens similar to 
the standards for forms B (Fig. 63) and c (Fig. 62) are present throughout 
the region, but because of modifications of the shave and distinctness of the 
posterior edge of the disc. complete intergradation occurs making the separa- 
tion of the two forms difficult (Fig. 65). In Canada and Maine the two 
forms are more clearly separable, with fewer intergrades. 

If form p is present in this region it is not common and cannot be recog- 
nized as distinct. An occasional specimen resembles form p, especially some 
small individuals of form B, but these are few even in New York State from 
which more than 900 specimens were examined. 

The color of most specimens of all forms is a bright green, although 
yellow-green or dark green specimens are not uncommon and blue-green 
individuals are occasionally found. The scutal and metasomal punctation is 


variable in both sexes throughout the area. Many males can be segregated 


| 


into the lettered forms, especially form a, but individual variation is great 


and numerous and complex combinations of characters can occur, adding to 


the confusion rather than clarifying the nature of the variation. 


616 THe UNiversity ScIENCE BULLETIN 


SOUTHEAST 


Species present: A. striata a, B, C, aurata, gratiosa, persimilts. 

In Florida three species occur throughout the state, striata, gratiosa, and 
aurata. In Georgia, striata a, B, and c occur throughout the state, form 8 
being most common. A. gratiosa and aurata occur chiefly in the southern 
counties but are also found sporadically in the mountains of the north; per- 
similis is found in the northern part of the state. In the Carolinas, persimilis 
has been taken in the mountains to the west, gratiosa appears nearer the 
coast and the three forms of striata occur throughout. The one specimen of 
aurata seen from North Carolina lacked detailed locality data. In Virginia, 
striata a, B, and c occur throughout the state, and persimilis ranges along the 
western border as far as Maryland, apparently its northern limit in the east. 
A. gratiosa probably occurs in Virginia as it has been taken in nearby Mary- 
land, but this appears to be its northern limit also. Except for the few gratio- 
sa from Maryland, only striata has been found in that state. 

A. striata. There is comparatively little variation within the three forms 
of striata in this region. Only form a is present in Florida (except for one 
female of form 8 from Gainesville and a male of form c simply labeled 
“Florida”). About 90°% of the female specimens are large, dark green to 
blue-green and very coarsely sculptured, especially on the scutum and pro- 
podeum. The other 10°% are less rugose and resemble striata a from other 
regions. The disc in both sexes is usually bracket-shaped with a sharply — 
angulate or carinate posterior edge (Fig. 59), although the disc may become 
more V-shaped and develop a well defined but less sharply angulate pos- 
terior margin in some specimens. The posterior vertical propodeal surface 
varies from rugose to smooth in females (always rugose in males), and the 
punctures on the first metasomal tergum are variable but usually small and 
widely spaced. Small workerlike individuals are rare and always look simi- 
lar to the larger, rough specimens. All but four of the males from Florida 
belong to the large, robust and coarsely sculptured form a. One of the four 
is form c, the other three, similar to males from outside Florida, cannot be 
classified as to form. Each differs from the others in the shape and size of | 
the disc and in the type of scutal and metasomal punctures. 

Throughout the rest of the southeastern region, form p as a distinct group 
is not found although B-p intermediates (Fig. 72) do occur. A few of the 
coarsely rugose Floridian representatives of form a range into Georgia but 
most specimens from Georgia and elsewhere in this region are similar to 
the standard of form a. Most such form a specimens are bright green in color 
although some are yellow-green. They differ from forms B and c throughout 


—————————— EE —E— SS" — eae aero 


the region, without intergradation. 
Form B is usually distinctive although intergrades with p and c are found 


1 
| 


SyYsTEMATICS OF THE GeNus Augochlorella Nortu or Mexico — 617 


in both sexes. Most specimens of form B have a long, roundly V-shaped disc 
(Fig. 63) with the edge of the disc well defined. In Georgia, the disc may be 
slightly shorter and rounder than normal, resembling that of form p (Fig. 
72), or less V-shaped in appearance, thus resembling form c (Fig. 65). 

Form c is considerably more variable than either a or 8. The shape of the 
disc varies from semicircular to U-shaped and grades into the V-shape of 
form B. The edge is smoothly rounded and shiny, with little differentiation 
between the disc and the vertical surfaces. A few small females throughout 
the area resemble persimilis but have the striae reaching or almost reaching 
the posterior margin of the disc. Both 8 and c are usually yellow-green in 
the southern part of the region and a yellow-green to bright green in Vir- 
ginia, Maryland and the District of Columbia. The metasomal punctures 
in all forms are usually small, close, distinct and regularly spaced in both 
males and females although other punctation patterns may also occur. Most 
of the males can be placed as to form even though the size, shape and sculp- 
turing of the disc is variable within any one form. 

A. persimilis. There is little variation in persimilis in this region. Speci- 
mens are generally a light green to yellowish or coppery-green color and the 
disc (Fig. 64) and posterior vertical surface of the propodeum (Fig. 74) are 
similar to those of other specimens throughout the range. The pubescence, 
however, may be more golden in color than it is to the west. The species is 
not numerous and occurs chiefly in the western edge of the region. Inter- 
mediates between persimilis and small striata c may occasionally be found 
among the females. 

A. gratiosa. This species is particularly abundant throughout Florida 
and probably Georgia but becomes sparse to the north. Specimens are usual- 
ly dark green to deep blue in Florida but are generally a yellow-green to 
bright green in the rest of the region. This species shows little morpho- 
logical variation except in Florida where the propodeal disc is often exceed- 
ingly short (Fig. 58) with the posterior edge of the disc more sharply angu- 
late than elsewhere in the range. The characters of females in this species 
intergrade with those of awrata in Florida although most specimens can be 
distinguished by the key characters. 

A. aurata. In Florida, the range of variation of aarata overlaps that of 


both striata a and gratiosa. In Georgia, however, the species seems to be dis- 


‘tinct. Most female specimens look similar to gratiosa with a rather short, 
weakly bracket-shaped disc (Fig. 57), fine striae and a similar body size. 


The males are most similar to s¢riata with short hair on the hind basitarsus 


and a narrowly emarginate fourth metasomal sternum. The variation in 


“specimens from Florida mainly involves body color, and the size and sharp- 
ness of the posterior edge of the disc in both sexes. The few specimens seen 


from Georgia were all alike in coloring and characters of the propodeum. 


618 Tue Universiry SciENCE BULLETIN 


Nortru CENTRAL 


Species present: A. striata A, B, C, D, persimilis. 


In the northern part of this region (Michigan, northern Wisconsin and 
Minnesota) only striata is found, forms a, 8, and c occurring to the east and 
B, Cc, and p to the west. In the east form a is the most abundant, in the west 


form p is most abundant, form c being more abundant than B throughout 


the region (Fig. 92). Further south, persimilis is also found, its greatest 
abundance in the western part of its range, particularly in Kansas, Iowa and 


Ilinois. In the southern part of the region, striata c is more abundant than pg, 


with form a decreasing in abundance and form p increasing from east to 
west. A. persimilis, though present, is less abundant than it is farther north, 

Throughout this region where persimilis and striata c are common, inter- 
mediates between the females of the two species are found. In all areas, 
striata 1s more abundant than persimilis. 

A. striata. In the North Central region there is a shift in the proportional 
abundance of forms from east to west. Form a is distinct from other forms 
in Ohio, Indiana, Illinois, Tennessee and Kentucky (Fig. 56). There seem 
to be several variants of this form present, so that the form could be easily 
divided into a number of subtypes based chiefly on shape of the disc and 


robustness of the bee. To the north, in Michigan, Minnesota and Wiscon- | 


sin, the edge of the disc becomes less sharp than in the standard and eastern” 


specimens, and the shape of the disc is variable so that none of the subtypes 
are well defined or distinctive. One specimen similar to those from Florida 
(Fig. 59) was found in Iowa [ Ames, XI-1-59 (D. Easterman) (18) ]. It con- 


trasts strongly with the other specimens, the usual form a being smaller, 


yellower, less robust and less coarsely rugose. Form a becomes less abundant 


to the west and grades into form p, although a few rather distinct individuals 


of form a occur in Nebraska, Kansas and Colorado. They are apparently 
absent in Oklahoma, Arkansas and Missouri although the form may be rep- 


resented as intergrade types in these areas exhibiting an abrupt, but not | 


sharp, posterior discal edge, with or without a medial V. 

Form 8 usually is distinct, most specimens agreeing well with the stand- 
ard (Fig. 63). The proportional abundance, however, decreases sharply to 
the west and the form is entirely absent in Colorado. Variants from the 
standard grade into both c and p. Form s-c intergrades (Fig. 65) are found 


frequently to the north in Michigan and Minnesota, and Ohio to IIlinois— 


and possess an elongate disc grading from V- to U-shaped, usually with a 
distinct but rounded edge. Form B-p intergrades (Fig. 72) are found more 
frequently to the west in Illinois, lowa and Kansas where the disc becomes 
shorter and more semicircular. Form c-p intergrades are found throughout 
the Central region. In Kansas, four nests were found containing both s and 


SysTEMATICs OF THE GEeNUs Augochlorella NortH of Mexico — 619 


p individuals, but in no case did any of the specimens agree with the stand- 
ards. Also, there were four nests containing B-c intergrades, two of the nests 
also containing at least one specimen distinctly of form B. 

Form c is variable and poorly defined in the eastern part of the North 
Central region but numerous and distinctive to the north and west. There 
are a number of different subtypes represented, with specimens resembling 
the standard (Fig. 62) found chiefly in Illinois, although they also occur in 
varying proportions elsewhere throughout the region. In Ohio, Tennessee, 
and Indiana there are many individuals resembling b-c intermediates (Fig. 
65), but in Indiana and lower Wisconsin the majority of the individuals of 
this form are small and intergrade with persimilis (Fig. 28). In Kansas, Mis- 
souri, Arkansas and Oklahoma about 50° of the form c group consists of 
these small persimilis-like intermediates. These small specimens also occur 
sparsely in Minnesota, Iowa and Nebraska. In Michigan, South Dakota and 
Colorado there are some individuals with persimilistike discs but the speci- 
mens are large and would not be confused with persimilis, a smaller species. 
These larger subtypes also occur in Kansas, Nebraska and Missouri. Other 
variations from the standard also occur throughout the North Central region 
(Fig. 61); sometimes they represent intergrades with forms B or p. 

Form p is not abundant nor distinct as a form in the eastern and northern 
states, although p-like specimens are occasionally found. It intergrades chief- 
ly with form a in Michigan and Wisconsin, where the disc is weakly bracket- 
shaped and the edge of the disc may become sharper than normal, with a 
weakly developed medial V. This form intergrades with form B in South 
Dakota, where the disc becomes more V-shaped and longer than the standard. 
Form p individuals are proportionally more abundant than other forms in 
the Great Plains states and although variable, the majority are similar to 
the standard of form p (Fig. 71). Small workerlike individuals are present, 
some resembling the larger individuals, others being intermediate between 
the small form c and persimilis. Biological data are scarce but six nests of 
this form were found in Kansas, all containing large and rather standard 
individuals, without worker forms. Four other nests contained both distinct 
B and p individuals. 

A. persimilis. There is little variation in size or structure in persimuilis 
throughout this region, although specimens tend to be somewhat greener 
(less yellowish) and the pubescence generally whiter than in the east. 
Throughout the region the disc of the propodeum is as in Fig. 64. 

A. gratiosa. A. gratiosa is not known in this region, although it may be 
found along the southeastern borders. It will look similar to striata a but 
with finer, closer striae (Fig. 58) and a rough or granular posterior vertical 
propodeal surface (Fig. 76). One specimen has been seen from eastern 
Tennessee. 


620 Tue UNiversiry ScIENCE BULLETIN 


SoutH CENTRAL 


Species represented: A. striata a, B, v, persimilis, gratiosa, neglectula, 
aurata, bracteata. 


In the western part of this region, the southwestern portion of Texas 


along the Mexican border as far east as Val Verde Co. and north into the 
Davis Mountains, neglectula and occasionally striata have been found. In 
the southern part of Texas, east of the Edwards Plateau, both bracteata and 
aurata occur, bracteata occurring as far north as Dallas Co., and aurata into 
Nacogdoches Co. A. bracteata has also been taken along the Rio Grande 
west to Val Verde Co., where it meets but probably does not overlap the 
range of neglectula. 

A. striata, gratiosa and persimilis also occur to the east of the Edwards 
Plateau, gratiosa coming from the east and occurring from south of Galves- 
ton to Nacogdoches, persimilis going only as far south as Nacogdoches, and 
striata ranging south to near Corpus Christi. A. striata is the only species on 
the Edwards Plateau, and although there seems to be a gap between the 
eastern populations and those of New Mexico and southwestern Texas, all 
the bees of this species look similar to one another. In Louisiana, Mississippi 
and Alabama, gratiosa and probably aurata occur to the south, persimilis is 
rare, and striata occurs commonly throughout. 

A. striata. In the eastern part of this region, form a looks similar to its 
standard (Fig. 60) but becomes less distinct to the west and all but disappears 


as a distinct form, grading into form p. Thus, the posterior edge of the disc | 


becomes less sharply defined than in the standard although abruptly rounded, 
and the disc becomes less angular and the bracket-shape less well defined. 

In the eastern part of the region, most specimens of form B are large and 
agree well with the standard. Toward the west the characters of form 8 
grade into those of form p. The posterior edge of the disc remains abruptly 
rounded but the length of the disc decreases and the outline gradually 
changes from the broad V typical of form B (Fig. 63) to the semicircular 
shape of p (Fig. 71). Form B also decreases in proportional frequency al- 
though a few specimens similar to the standard are found in eastern Texas. 

In form p the length of the striae and angulation of the posterior margin 
of the disc as well as the shape of the disc are variable but in general resemble 
that in Figure 71. Form c is not distinct in this region, although occasional 
specimens may show some resemblances to it. 

In Louisiana and Texas, a few small pers¢milislike individuals are pres- 
ent that would belong to form c in other parts of the range, but are probably 
worker individuals of form p in this region. These are always distinct striata, 
however, and except for size do not intergrade with persimilis. To the west 
the percentage of form p increases. 


Systematics oF THE GENus Augochlorella Nortu or Mexico 621 


Throughout the region both color and punctation of females are variable. 
Males are individually variable with only a few specimens distinctive enough 
to be placed in one form or another; none of the males in this region seem to 
intergrade with any other species. 

A. persimilis. A. persimilis occurs very sparsely in this region and appears 
to differ little from those elsewhere in the range (Fig. 64). Most specimens 
are a yellow-green or coppery green in color. This species does not intergrade 
with striata c or p in this region, although there are small individuals of 
striata in Texas and Louisiana. 

A. gratiosa. Only a few specimens of gratiosa have been seen from this 
area. All are similar, usually bright green or yellow-green in color. The pro- 
podeal disc is usually about equal in length to the metanotum, sharply de- 
fined but rarely carinate. Some females may be confused with females of 
aurata in this region. 

A. aurata. Females of this species have been taken from throughout 
southeastern Texas and, although variable in disc characteristics, can be dis- 
tinguished from other species by the key characters. The females are most 
similar to gratiosa. No males have been seen, and no specimens of either sex 
have been taken between southeastern Alabama and Texas, probably due to 
lack of collecting in this area. 

A. neglectula. This species is a dark green, roughly sculptured species 
(Figs. 70, 75) with all specimens similar to one another in this region. Al- 
though striata and bracteata have also been taken in the same area, it is 
probable that the three species occur in different habitats. 

A. bracteata. This species is similar to the small striata c-perstmilis inter- 
mediate in size and in appearance of the propodeal disc (Fig. 66). Both of 
these characters show considerable variation, but the thoracic sculpturing 
(Fig. 78) is distinctive and unvarying and serves to separate this species 


from persimilis, aurata and striata in Texas. 


WEsT 


Species represented: pomoniella, neglectula, striata B, D. 
The only species occurring in California, Utah and Nevada is pomoniella. 


A. pomoniella enters Utah only in the extreme southwestern corner, in 
Washington Co. In Arizona, both neglectula and pomoniella are found, 


pomoniella to the west and neglectula to the east, with their ranges broadly 


overlapping in the center of the state. In New Mexico, striata comes in from 


the northeast and ranges southward to the Mexican border through the 


eastern half of the state. 4. neglectula ranges across the southern half of 
the state. 


622 Tue University ScIENCE BULLETIN 


A. striata. Most of the individuals of striata are small (6 mm), about the © 
size of large persimilis. The scutum is rather finely punctured or finely 


rugose, and the shape of the disc is somewhat variable, but the striae in all 
cases fill the discal area and reach the distinct posterior edge. The posterior 
surface of the propodeum is finely granular or weakly roughened, shiny or 


dull, and the color of the body varies from dark green to blue-green. Most | 


striata belong to form p in this region, although an occasional form B is 
found. Individuals weakly a-like and intergrades between B and p are also 
occasionally found. 


A. neglectula. There is little overall variation in this species except that 


the propodeal area varies in the degree of roughness of the posterior and 
lateral vertical surfaces. The smoother individuals superficially resemble 
pomoniella in Arizona or striata in New Mexico, although there is usually 
little difficulty in distinguishing the species, and the number of specimens 
showing this similarity is few. 

A. pomoniella. There is little variation in size, color or morphological 


characters in specimens from California, Nevada and Utah, but in Arizona, — 
specimens are smaller and more variable in the propodeal characters and in— 


color. A. pomoniella is, however, distinctive throughout the area and does 
not intergrade with neglectula which it overlaps in Arizona. 


Mexico 


Species present: A. neglectula, pomoniella, bracteata, edentata. Other 
species found only in the Neotropical region and belonging to Pereirapis 
are not included here. 

Throughout the central part of Mexico from the northern border into 
Central America, A. neglectula is the most common of the species consid- 
ered in this paper. It occurs throughout Mexico except along the coasts. 
Both edentata and bracteata are also found in the central area but occur only 
sparsely in the eastern part. 

In Baja California and along the western coastal area, pomoniella is found, 
and along some beaches of the west coast, maritima, a subspecies of neglec- 
tula is found. Apparently both pomoniella and neglectula occur together 
south of Morelos although they have not been collected simultaneously from 
any one area. In Chiapas and Yucatan, only pomoniella has been collected. 

A. neglectula. This species shows little variation in Mexico, except for 
Pacific Coast populations placed in the subspecies maritima. Occasional 
female specimens are dull rather than shiny, with the body surfaces minutely 
reticulated, especially on head and thorax. The blue patches on the frontal 
areas are rarely conspicuous on Mexican specimens, and occasional speci- 
mens with unusually smooth propodeal areas may look similar to pomoniella. 


SYSTEMATICS OF THE GENUs Augochlorella Norru or Mexico — 623 


Many of the specimens of the subspecies maritima are from the same popula- 
tion and therefore show little morphological variation among themselves. 

A. pomoniella. ‘This species is rather variable in size, color and sculptur- 
ing, especially in southern Mexico. Specimens from Baja California and 
Sonora are large and much like those from California, but those from further 
south are small, often with paler brown and darker green coloration. The 
range of pomoniella overlaps that of neglectula in Guerrero and Oaxaca, and 
females from these areas may intergrade morphologically. 

A. bracteata. Only a few specimens of this species have been taken in 
Mexico so that the nature of the variation has not been determined. The 
specimens seen are like those from Texas and are easily differentiated by the 
key characters from other species of this region discussed in this paper. 

A. edentata. This species is somewhat variable in both color and sculptur- 
ing, but until more specimens can be seen, the extent of the variation cannot 
be described. It is distinguished from other Mexican species discussed here 
by its smaller size, smooth sculpturing (Fig. 79), convex face and other key 
characters. 


ACKNOWLEDGMENTS 


I wish to thank all the individuals and institutions listed in Table 1 for 
their cooperation in sending me specimens of Awgochlorella for this study. 

I am particularly indebted to Dr. C. D. Michener, Dr. G. W. Byers and 
Dr. H. S. Fitch for reading this manuscript and for their helpful comments 
and criticisms. 

This study was aided by Grant No. GB-91 from the National Science 
Foundation (Dr. C. D. Michener, principal investigator). 


LITERATURE CLiLED 


DreispacH, R. R. 1945. The green Halictine bees of the genera Agapostemon, Augochlora, 
Augochlorella and Augochloropsis. Michigan Acad. Sci. Arts and Letters, paper 30, 
pp. 221-227. 

Fernatp, M. L. 1950. Gray’s manual of botany, 8th ed. American Book Co., illus. lxiv -+ 
1632 pp. 

Goutp, F. W. ano G. W. Tuomas. 1962. Texas Plants—a checklist and ecological summary. 
Texas Agr. Exper. Sta. Publ. MP-585, pp. 1-112. 
Kearney, T. H. anp R. H. Persies. 1951. Arizona Flora. Univ. California Press, viii-1032 pp. 
Lovett, H. B. 1942. The bright green bees of the genera Agapostemon, Augochlora, Augo- 
chloropsis and Augochlorella in Kentucky. Kentucky Acad. Sci. Trans. 10:19-23. 
Micuener, C. D. 1937. Records and descriptions of North American bees. Ann. Mag. Nat. 
Hist. (10) 19:313-329. 

——. 1944. Comparative external morphology, phylogeny, and a classification of the bees 
(Hymenoptera). Bull. Amer. Mus. Nat. Hist. 82:151-326. 

———. 1954. Bees of Panama. Bull. Amer. Mus. Nat. Hist. 104:1-176. 

.AND R. B. Lance. 1958. Observations on the behavior of Brasilian halictid bees. V. 

Chloralictus. Insectes Sociaux 5:379-407. 

Mircuett, T. B. 1960. Bees of the eastern United States, vol. 1. North Carolina Agr. Exper. 
Sta. Tech. Bull. No. 141, 583 pp. 


624 Tue University SciENcE BULLETIN 


Mounz, P. A. 1959. A California flora. Univ. California Press, 1681 pp. 
Orpway, E. 1965. The systematics of the genus Augochlorella (Hymenoptera, Halictidae) nor 
of Mexico and the bionomics of A. striata and A. persimilis in eastern Kansas. Ph. J 
Thesis. Univ. of Kansas, 347 pp.; also available in microfilm. 
.1965a. Caste differentiation in Augochlorella striata and A. persimilis (Hymenopter 
Halictidae). Insectes Sociaux. 12:291-308. 
. (In press.) The bionomics of Augochlorella striata and A. persimilis (Hymenopter 
Halictidae), in eastern Kansas. Jour. Kansas Ent. Soc. 
Rau, P. 1922. Ecological and behavior notes on Missouri insects. Trans. Acad. Sci. St. Lot 
24:30-71. 
SanpHousE, G. A. 1937. Bees of the genera Augochlora, Augochloropsis and Augochlorel 
(Hymenoptera, Apoidea) occurring in the United States. Jour. Washington Acad. § 
27:65-79 
Smirn, J. B. 1910. Annual report of the New Jersey State Museum including a report of th 
insects or New Jersey for 1909, 880 pp. 
SmitH, JoHN K. 1933. Manual of the southeastern flora. New York, published by the autho 
xxil + 1554 pp. 
Tirus, E. S. G. 1901. On some bees of the genus Augochlora. Canadian Ent. 33:133-137. 
Viereck, H. L. 1916. Guide to the insects of Connecticut. Part HI. The Hymenoptera | 
wasplike insects of Connecticut. Connecticut State Geological and Nat. Hist. Sur 
Bull. No. 22, 824 pp. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


THE BIONOMICS OF TENUIPALPOIDES 
DORYCHAETA PRITCHARD AND BAKER (1955) 
(ACARINA, TROMBIDIFORMES, 
TETRANYCHIDAE) 


By 


George Singer 


VoL. XLVI Paces 625-645 June 17, 1966 No. 17 


y AN if 


a 
WaTAR 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 625-645 June 17, 1966 No. 17 


The Bionomics Of Tenuipalpoides dorychaeta 
Pritchard and Baker (1955) 


(Acarina, Trombidiformes, Tetranychidae)" ~ * 


By 


GeEorGE SINGER* 


ABSTRACT 


The species T. dorychaeta is discussed in detail with regard to its biology and 
morphology. Rearing techniques involved the use of a live honey locust tree and 
isolation cells of plastic and glass afxed to the trunk of this tree. At 72°F the 
egg has an incubation period averaging 12 days. A low percentage of emergence 
at this stage is attributed to unfavorable humidity. The hexapod larva requires 2 
days each for a completion of the active and quiescent stages. The octopod proto- 
nymph stage has the same duration as the larva. The deutonymph may be identi- 
fied to sex. Five days are required for completion of this stage. Adult males are 
active, feed intermittently, and have an average longevity of 10 days. Adult 
females feed almost continuously and deposit an average of 14 eggs during their 
20-day life span. They are arrhenotokous. After oviposition the female lays 
down a network of fine silk strands over the egg. This was the only time that 
spinning was observed in this species. Adult females remain in a feeding position 
on the host throughout the winter in Kansas. 


‘Contribution No. 1214 of the Department of Entomology, University of Kansas, Lawrence, 
Kansas. 


* Portion of a thesis submitted in partial fulfillment of the requirements for the degree of 
Master of Arts in Entomology at the University of Kansas. 

* This study was supported in part by a grant from the General Research fund of the Uni- 
versity of Kansas. 

‘The author wishes to express his sincere appreciation to Dr. R. E. Beer for his assistance, 
interest, and encouragement during the course of this study. Present address: Zoology Depart- 
ment, University of Montana, Missoula, Montana. 


626 Tue UNIversiry ScrENcCE BULLETIN { 


INTRODUCTION 


The genus Tenuipalpoides Rekk and Bagdasarian (1948) has characters 
similar to both the subfamilies Bryobiinae and Tetranychinae. The char- | 
acter and placement of the setae of tarsus I, the complex distal enlargement — 
of the peritreme, the integumental structure, and the lateral position of the | 
inner sacral setae are all characteristic of the Bryobiinae. It is, however, in- 
cluded in the Tetranychinae by the following characters: no tenant hairs on | 
empodium; adult female, male, and all immature instars with two pairs of | 
anal setae. 

Tenuipalpoides dorychaeta is the only known North American representa-_ 
tive of the genus and is the second species recorded for the genus. Little has 
been published on its life history, other than comments on repeated collection | 
from the bark of honey locust (Gleditsia triacanthos L.) and black locust | 
(Robinia pseudo-acacia L.) in Louisiana, North Carolina, and Utah. Both | 
males and females are recorded from these hosts during the summer (Prit- 
chard & Baker 1955). 

Teniupalpoides zizyphus Rekk & Bagdasarian (1948) is the other species, 
recorded from Erevan, Armenia, on Zizyphus vulgaris Lam. (Rhamnaceae). | 
Descriptions of this species are from females; males are not recorded. 


REARING METHODS 


A survey for these mites was conducted in the spring of 1960. Good popu- 
lations were found on two widely separated honey locust trees on the Uni- 
versity of Kansas campus. Individuals collected from these trees were used 
for the life history studies and morphological descriptions. Collections were 
made during June and November, 1960, and February, 1961. 

Branches 4 inch in diameter and smaller were cut from the trees. These 
were then cut into sections about one foot in length to facilitate handling. 
In the laboratory the leaves were removed from the branches, beaten with a 
pencil over white paper, and then checked under a dissecting microscope. 
No T. dorychaeta individuals were ever collected from the leaves. The branch 
sections were also beaten, but this method proved inferior to using the dis- 
secting microscope for locating and dislodging the mites from the bark 
where they were normally found wedged into cracks with their mouthparts 
inserted. In such a position they were not readily dislodged by beating. 
Occasionally mites were observed moving over the surface of the bark. 

Attempts were made to rear the mites on freshly cut bark chips floating 
on tap water or sucrose solutions. This proved inadequate because the chips 
became waterlogged and overgrown with mold within 12 hours. Sections 
of branches with one end sealed with paraffin and the other immersed in 
water also proved inadequate. There was no means for insuring a relatively 


Tue Bionomics or Tenuipalpoides dorychaeta 627 


constant food source, nor could the mites be prevented from falling into the 
water. Therefore, a method of rearing was developed that employed a mini- 
mum of handling because the mites were easily injured and time involved in 
manipulations would prohibit the study of the large numbers needed for 
observations. 

The most satisfactory method for rearing the mites included the use of a 
live tree for a constant food source and cells of plastic and glass for isolating 
them. A three-year-old thornless variety of honey locust was obtained in 
July, 1960, and the branches and tree top were pruned so that the tree stood 
five feet tall. The roots were washed clean of soil and potted in a round 
plastic bucket 9x12 inches deep to facilitate handling. Cheesecloth was tied 
to the trunk and the rim of the bucket to prevent a loss of soil when the 
container was tilted. By this arrangement the tree could be laid horizontally 
on a table and rolled back and forth to expose areas of the trunk under the 
dissecting microscope. While the tree was in this position, cells were affixed 
and the mites were observed and handled. Between obsrvations of this sort 
the tree was returned to an upright position. 

Isolation cells were construced from washers with a 4 inch outer and 
5/16 inch inner diameter, drilled and cut from a sheet of 0.40 inch clear 
plastic. These washers were affixed to the bark with melted paraffin, using 
a hot needle. A No. 1, 12 mm round coverglass, secured by a small amount 
of parafin on one edge of each washer, provided a lid for the cell which 
could be easily removed, quickly and securely replaced with a hot needle, 
and through which observations could be made easily. The bark under each 
cell was first scraped to remove excess wax and then washed with alcohol 
and distilled water to remove any acaricidal chemical residues. The cells 
were oriented in rows parallel to the axis of the tree. Mites from the field or 
those isolated after emergence were introduced into separate cells with the 
use of a fine sable brush. 

A glass tube, inserted through the soil to the bottom of the bucket, pro- 
vided a means for watering and fertilizing the plant. This device assured that 
water reached the roots at the bottom of the bucket and reduced fertilizer 
odors. Of the three commercial fertilizers tried, Ortho-gro, Vigoro, Ra-pid- 
gro, the first appeared to give the best results in fostering rapid growth. 

From November, 1960, until the end of the study, in March, 1961, the 
photoperiod in the laboratory was kept above 13 hours so that some of the 
green foliage was retained by the tree. Normally, in greenhouses where the 
temperature never reaches freezing, the reduction of the photoperiod by 
early winter results in dormancy defoliation until March. It was hoped that 
a more constant host environment would be maintained by a retention of 
summer-type conditions. 

The mites and their eggs were found to be very susceptible to handling 


628 Tue University ScIENCE BULLETIN 


} 
| 
, 
f 


injury; therefore, handling was kept at a minimum. Egg deposition times | 
were recorded and their positions were plotted on a map diagram of each | 
cell. By this means the incubation period for each egg could be determined 
without disturbing it. 

Observations were conducted at approximately 12 hour intervals to reduce 
developmental-time error. The temperature and relative humidity during 
the study averaged 74°F and 309% respectively. 


LIEE HISTORY 


The life cycle of T. dorychaeta follows the generalized pattern found in 
the Tetranychidae: a normally developing egg, hexapod larva, eight-legged_ 
protonymph, deutonymph, and adult male or female. Each active stage, | 
other than the larva, is preceded by a quiescent or “pupal” stage during 
which the succeeding stage develops. Prior to each pupation the mites en- 
gorge with plant protoplasm and then become quiescent with their chelicerae_ 
inserted and their claws hooked to the substrate. After a period of six hours 
their appendages become white from autolysis of the tissues, and the gut con-_ 
tent is reduced in quantity and in coloring matter by digestion and assimila- 
tion. At emergence, the old exoskeleton splits transversally between the 
propodosoma and the hysterosoma and then posteriorly along the lateral | 
margins. The emergent instar backs out of this slit and begins feeding after 
a short rest period. 

In nature, the mites appear to feed on non-chlorophyll-bearing cells under 
the cork layer of the bark and remain a unicolorous orange-red throughout 
the life cycle. Under the conditions afforded during this study, however, 
the mites were exposed to chlorophyll-bearing trunk tissues and acquired 
dark greenish-black areas in the idiosoma. The red color was modified from 
a red to orange to yellow in the gnathosoma and legs. During feeding the 
mites flexed their legs close to the body and remained in this position for 
long periods of time, apparently until the food source had become depleted 
at that site. In this feeding position, only the dorsum of the legs and idiosoma 
were exposed. The tough nature of the dorsal idiosoma, coupled with the 
presence of large serrate setae, probably tends to repel or discourage predators. 
The mites were observed to move rapidly when disturbed. At no time did 
they attempt to feed on leaf tissues of the host, even when this was offered 
as the only food source. The preferred site for feeding was in cracks and 
crevices of the stem or branch bark, and it is in such locations that they are 
usually found in nature. When such cracks or bark irregularities were not 
available, the mites fed through the smooth bark but changed feeding sites 
more often. At no time did their presence seem to affect the vitality of the 
host. 


Tue Bronomics or Tenuipalpoides dorychaeta 629 


_ The Egg: The spherical egg has a smooth chorion and a deep carmine- 
‘red color. The diameter varies from 125 to 143m with a 25 egg average of 
137». Immediately after oviposition the female lays down a fine network of 
silk strands over the egg that tends to flatten and cause it to become some- 
what disc-shaped. During embryonic development, a dark brown area oc- 
‘cupying about one quarter of the sphere is formed at about the fifth day, 
while the rest of the egg becomes straw-colored. As the embryo develops 
further, the egg returns to a unicolorous red and, in a dorsal view, a flat 
area appears on one side occupying about 4 of the circumference. At this 
time the carmine eyes of the embryo can be seen. The egg then takes on a 
pearly sheen as the larva detaches itself from the chorion and prepares for 
emergence, which is accomplished through a circular slit in the dorsum of the 
egg shell and through the retaining silk strands. Total developmental time 
for 62 eggs (reared through to adults) was 9 to 14 days, with an average of 
12. From a total of slightly over 600 eggs, only 73 completed embryonic 
development and emerged as larvae. In many cases the embryos appeared to 
complete development but apparently could not extricate themselves from the 
egg shell or became trapped by the silk strands. The remainder either failed 
to differentiate at different levels of development or embryonic development 
was not initiated. The reason for this low percentage emergence is believed 
to be due to unfavorable relative humidity in the cells, as such a factor has 
been observed to influence emergence rates while not affecting emergence 
time: Kremer (1956) found that in Bryobia praetiosa Koch the greatest per- 
cent hatch of winter eggs occurred at 10 to 30°% relative humidity. Humidity 
in excess of 80°% induced the lowest percent emergence. Anderson (1948) 
found the opposite for Paratetranychus pilosus Can. & Franz. (=Panonychus 
ulmi), where the lower humidity reduced the percentage of emergence. The 
effect of humidity on emergence rate, but not on emergence time, is not re- 
stricted to the tetranychids but is also known to occur in species in other 
suborders of the Acarina (Camin, 1953). 

The Larva: The hexapod larva is orange-red on emergence and its sex is 
not identifiable. The larva does not begin to feed immediately on emergence 
but will wander about for an undetermined period of time. This may be of 
some selective advantage in aiding or increasing the dispersal of the species 
on a given host tree or, by air currents, to other trees. Once a feeding site 
is selected the larva will remain motionless, feeding constantly. Pupation 
usually takes place at the feeding site. The active larval instar has a duration 
of about two days while the quiescent period that follows requires somewhat 
less than two days before the emergence of the succeeding instar. 

The Protonymph: The octopod protonymph emerges and begins feed- 
ing almost immediately. No clear distinction was found that would separate 
the sexes in this instar. The active feeding period of two days is followed 
by a two day quiescent period. 


630 THe Universiry SctENcCE BULLETIN 


The Deutonymph: The sexes are distinguishable in this stage; males are 
smaller and have a relatively pointed abdomen as compared to females. This 
form begins feeding almost immediately after emergence. The completion 
of each of the active and quiescent stages required about 2.5 days. 

The Adult Male: After emergence, the males feed for a short time and 
then begin to wander apparently at random. The front legs are tapped up - 
and down ahead of the mite as it walks. The following is an account which - 
summarizes several observations: When a male blunders upon a female. 
deutonymph pupa, it becomes excited and circles the female stroking it with - 
his palps. After a short period of time the male becomes quiet and stradles_ 
the quiescent female. Several hours later the male moves to the side of the 
female and feeds for a period of time, then it resumes its former position. | 
As the adult female begins to emerge the male becomes highly excited, } 
orienting himself behind the emerging female, stroking her with his palps, — } 
and pulling on her dorsal setae with his palps. The emergent female char- 
acteristically moves to the side of the exuvium and attempts to feed. The | 
male’s advances, including pulling on her dorsal setae and stroking her with 
his front legs and palps, appears to cause the female to raise her opisthosoma _ 
and remain motionless. At this point the male moves forward and under — 
the female, at the same time arching the tip of his opisthosoma dorsally 
and anteriorly. The male continues moving forward under the female while 
probing the venter of the female with his palps until their genitalia contact — 
and copulation takes place. During this interlude the female remains 
motionless with her stylets inserted into the bark. In several observed in- 
stances copulation required from five to seven minutes, after which the male 
relaxed his opisthosoma into its normal position and backed out from 
under the female. Following copulation, both mites moved off several steps 
and then began feeding. Quiescent stages of larvae, protonymphs, male 
deutonymphs, and all the active immature instars arouse the adult males only 
momentarily, with the males moving off after only a brief moment of 
probing. Adult virgin females mated readily, in the manner described above. 
Mated females did not allow a second mating, preventing copulation by 
not responding to the male’s advances. Longevity of males was four to 21 
days, with an average of 10 days for 31 individuals observed. 

The Adult Female: Immediately after emergence the female selects a feed- 
ing site in a crevice or next to some irregularity in the bark and begins feed- 
ing. Feeding continues fairly constantly throughout the life span of the adult 
female, individuals moving only to new feeding sites or during oviposition. 
The oviposition period is preceded by a preoviposition period of three to 
four days, after which an egg is deposited every 12 to 24 hours. Cracks and 
rough surfaces of the bark are generally selected for oviposition sites. After 
oviposition, the female turns around and begins to spin a fine, loose network 


Tue Bronomics or Tenuipalpoides dorychaeta 631 


of silk over the egg. This is the only time that any of the mites were 
observed to spin silk. Grandjean (1948) states that the silk ducts open into 
the palpal thumb of the Tetranychidae. Blauvelt (1945) did not commit him- 
self on this matter; however, his diagram indicates that the common silk 
duct extends into the gnathosomal rostrum in Panonychus ulmi (Koch) 
_(=Paratetranychus pilosus Can, & Franz.). My observations of these mites 
during spinning were not conclusive, as it was not possible to see from what 
part of the gnathosoma the silk was emitted. 


There was no difference between mated and virgin females with respect 


| 


to longevity or egg-laying capacity. A maximum of 31 eggs was recorded 
from one imal with an average of 14 eggs for 30 females observed. 
‘Longevity averaged 20 days with a maximum of 39. Eggs deposited by 
virgin females developed only into males while eggs from mated females 
produced both males and females. 

The fact that most species of tetranychids commonly exhibit the pheno- 
menon of producing only males from unfertilized eggs and both males and 
females from eggs of mated females is an old and “well documented fact 
(Ewing, 1914; Garman and Townsend, 1938; Cagle, 1943, 1946, 1949; English 
and Snetsinger, 1957). Ewing (1914) suggested that the mechanism of sex 
determination in tetranychids is linked with chromosome number, haploidy 
in males and diploidy in females. Cytological confirmation of this pheno- 
menon has been demonstrated for Tetranychus telarius (Linn.) by Schrader 
(1923), where males were found to have three chromosomes and females 
SIX. 

Overwintering of this species in Kansas is accomplished by the adult 
females. Only adult females were found on the trees after the first frost in 
November, 1960, even though prior to this frost all stages were in evidence. 
A second survey in February, 1961, following -5°F temperatures, disclosed 
numerous live, feeding, adult females. These females, when brought into 
the laboratory, began feeding immediately and continuously as evidenced 
by their acquiring a green-black coloring in the idiosoma. Four females 
collected in November had a preoviposition period averaging 17 days after 
which an average of four eggs was deposited per female. Their longevity 
was 20 to 31 days after introduction into the laboratory. Seventeen females 
collected near the end of February had an average preoviposition period of 
10 days, an average of 11 eggs per female, and a longevity of 9 to 32 days, 
averaging 28. From these data it appears that metabolism is greatly reduced 
by exposure to a period of subfreezing temperatures. The longevities of 
both these groups, after being brought into the laboratory, were the same 
even though there was a lapse of three months between the two collections. 
It is ineltienee that no eggs or immature instars were present during this 
interlude. Samples of soil and leaf litter from beneath the host trees ated 


632 Tue Universiry ScIENCE BULLETIN 


to yield any T. dorychaeta individuals during the course of this study, al 
though the University of Kansas acarological collection contains one speci- 
men fen from leaf litter in November, 1960, by Dr. R. E. Beer. Eggs that 
were produced by overwintering females developed into males and females 
indicating that most, if not all, of the females had mated before the first frost 
in Nie No males were collected after this frost. 


MORPHOLOGY AND TAXONOMY 


The dorsal body chaetotaxy is constant in number and position from the 
larva to the adult, the only variation being in the size of setae, increasing in 
length and width with each succeeding instar. These setae are broadly © 
lanceolate and serrate, appearing opaque white in life. The three pairs of © 
dorsal propodosomal setae are arranged in the usual tetranychid triangular — 
pattern: 3 dorsocentral pairs, 3 pairs of dorsolaterals, and 1 pair each of © 
humeral and clunals. The sacrals are 2-paired and their marginal position — 
is considered to be a primitive condition according to Pritchard and Baker 
(1955). The ventral idiosomal setae increase in number from the larva to — 
the adult, and the chaetotaxy of this area is therefore valuable in identifying — 
the instars. The 2 pairs of post-anals are plumose and easily distinguishable — 
from the other ventral setae. There are 2 pairs of simple anal setae. The 
palpal tarsus bears 3 simple and 3 rodlike setae, lacking pronounced setal 
bases. Distally the palpal tarsus bears a small papilla, except in the adult 
male. The pretarsus of each leg bears 1 simple empodial claw, flanked later- 
ally by the reduced, padlike, true claws, each one giving rise to 2 tenent hairs 
distally. Dorsally the propodosoma is sculptured and the hysterosoma is 
wrinkled. The idiosoma is reticulated with short dashes in all stages except 
the adult male, where there are fingerprintlike striations of lines and dashes. 
Ventrally the medial portion of the idiosoma is patterned by line striations 
and the lateral margins are wrinkled. 

Adult Female (Figs. 1, 2): Length of idiosoma 331 to 375», width of 
idiosoma 218 to 312», averaging 353 x 250» for 17 individuals. 

Palpal femur with 1 dorsal, simple seta; genu with 1 dorsal, simple seta; 
tibia with 1 dorsal, one lateral, 1 medial simple seta, and a terminal claw of 
thumb-claw complex. 

Coxa I with 2 setae ventrally, the medial one simple and twice the length 

of spiculate lateral one; trochanter with 1 simple seta; femur with 1 simple, 
1 spiculate, and 1 broadly serrate seta; genu with 1 simple, 1 spiculate, and 
3 broadly serrate setae; tibia with 2 serrate, 3 spiculate, 1 simple, and 1 
short, peglike, solenidion; tarsus with 4 solenidia, 3 simple, 2 terminal, 
spatulate-plumose setae, and 2 distal sets of duplex setae. Distal members 
of duplex setae greatly exceeding proximal members in length, the lateral 
distal member longest of the 4. 


5 


Tue Bionomics or Tenuipalpoides dorychaeta 633 


Coxa IT with two ventral setae, the medial one simple and twice the length 
of spiculate lateral one; trochanter with 1 simple seta; femur with 2 serrate 
setae and 1 simple; genu with 4 serrate setae and 1 simple; tibia with 3 
serrate setae, 1 spiculate and 1 simple seta; tarsus with 3 solenidia, 2 simple, 
2 terminal spatulate-plumose setae, and 1 distal set of duplex setae, the distal 
member short and peglike while the proximal member is simple. 

Coxa III is bare; trochanter with 1 serrate seta; femur with 2 broadly 
serrate setae; genu with 1 serrate seta; tibia with 2 serrate setae, 1 spiculate 
and 1 simple seta; tarsus with 5 simple and 2 terminal spatulate-plumose 
setae. 

Coxa IV with 1 spiculate seta ventrally; trochanter with 1 spiculate seta; 
femur with 1 spiculate seta; genu with 1 serrate seta; tibia with 1 serrate, 1 
spiculate and 1 simple seta; tarsus with 5 simple and 2 terminal spatulate- 
plumose setae. 

Ventrally with 1 pair of simple setae on palpal coxa; 4 pairs of simple 
setae on ventral podosoma, the anterior pair between coxae I and II, 2 pairs 
medial to coxa III, and 1 pair medial to coxa IV; opisthosoma with 3 pairs 
of simple setae anterior to anus, these being 2 pairs of genitals plus 1 pair 
lateral to the genital aperture. 

Adult Male (Figs. 3, 4): Length of idiosoma 256 to 300u, width of idio- 
soma 156 to 169, averaging 281 x 162» for 10 individuals. 

Palpal femur with 1 peglike tactile seta dorsally; genu with 1 simple seta 
dorsally; tibia with 1 dorsal, 1 lateral, and 1 medial simple seta, and a ter- 
minal claw of thumb-claw complex. 

Coxa I with 2 ventral setae, the medial one simple and twice the length 
of spiculate lateral one; trochanter with 1 simple seta; femur with 2 simple 
setae and 1 broadly serrate seta; genu with 4 broadly serrate setae and 1 
simple seta; tibia with 1 serrate seta, 1 solenidium, 3 spiculate, and 2 simple 
setae; tarsus with 6 solenidia, 2 simple, 1 spiculate, 2 terminal, spatulate- 
plumose setae, and 2 distal sets of duplex setae, distal members being con- 
siderably longer than proximal members and lateral distal member longer 
than other 3. 

Coxa II with 2 ventral setae, the medial one simple and twice the length 
of spiculate lateral one; trochanter with 1 spiculate seta; femur with 1 
spiculate, 1 serrate, and 1 simple seta; genu with 3 serrate, 1 spiculate, and 
1 simple seta; tibia with 3 spiculate and 2 simple setae; tarsus with 5 solenidia, 
3 simple, 2 terminal, spatulate-plumose setae, and 1 distal set of duplex setae, 
the distal member peglike and shorter than the simple proximal member. 

Coxa III bare; trochanter with 1 spiculate seta; femur with 1 serrate and 
I simple seta; genu with 1 spiculate seta; tibia with 2 serrate, 1 simple, and 
I spiculate seta; tarsus with 5 simple and 2 terminal, spatulate-plumose 
setae. 


634 Tue UNIversiry ScIENCE BULLETIN 


Coxa IV with 1 simple seta ventrally; genu with 1 spiculate seta; tibia 
with 1 serrate seta, 1 spiculate seta and 1 simple seta; tarsus with 5 simple 
and 2 terminal, spatulate-plumose setae. 

Ventrally with 1 pair of simple setae on palpal coxa; ventral podosoma 
with 4 pairs of simple setae, the anterior pair medial to coxae I and II, 2 
pairs medial to coxa III and 1 pair medial to coxa IV; opisthosoma with 3 
pairs of simple setae anterior to anus. 

Deutonymph (Figs. 5, 6): Female; length of idiosoma 250 to 312, width 
162 to 193p, averaging 281 to 187» for 15 individuals. Male; length of idio- 
soma 225 to 256m, width 156 to 168», averaging 250 x 162» for 14 individuals, 

Males may be distinguished from females by their smaller body size, and 
male deutonymphs have relatively pointed abdomens in contrast to those of 
the females. 

Palpal femur with 1 simple dorsal seta; genu with 1 simple seta; tubia 
with 1 dorsal, 1 lateral, and 1 medial simple seta, and a terminal claw of 
thumb-claw complex. 

Coxa I with 2 simple ventral setae; trochanter with 1 simple seta; femur 
with 2 simple setae and a broadly serrate one; genu with 4 broadly serrate 
setae and a simple one; tibia with 3 serrate and 2 spiculate, 1 simple seta 
and 1 short, peglike, solenidium; tarsus with 4 solenidia, 3 simple, 2 terminal, 
spatulate-plumose setae, and 2 distal sets of duplex setae, the distal members 
greately exceeding proximal members in length, and distal lateral member 
exceeding the other 3 in length. 

Coxa IL with 2 ventral setae, medial member simple and twice length of 
lateral spiculate member; trochanter with 1 simple seta; femur with 1 broadly 
serrate and 1 spiculate seta; genu with 3 serrate setae, 1 spiculate and 1 simple 
seta; tibia with 3 serrate and 2 simple setae; tarsus with 3 solenidia, 3 simple 
2 terminal spatulate-plumose setae, and 1 distal set of duplex seta, the distal 
member short and rodlike and proximal member long and simple. 

Coxa III bare; trochanter with one serrate seta; femur with 2 serrate 
setae; genu with 1 serrate seta; tibia with 2 serrate and 2 simple setae; tarsus 
with 5 simple and 2 terminal, spatulate-plumose setae. 

Coxa IV with 1 simple seta; trochanter bare; femur with 1 simple seta; 
genu with I serrate seta; tibia with 1 serrate and 2 simple setae; tarsus with 
5 simple and 2 terminal, spatulate-plumose setae. 

Ventrally with 1 pair of simple setae on palpal coxa; podosoma with 4 
pairs of simple setae, the anterior pair medial to coxa II, 2 pairs medial to 
coxa III, and 1 pair medial to coxa IV; opisthosoma with 2 pairs of simple 
setae anterior to anus. 

Protonymph (Figs. 7, 8): Males and females are not distinguishable from 
each other. Length of idiosoma 187 to 224», width 125 to 156, averaging 
218 x 143 for 20 individuals. 


Tue Bronomics or Tenuipalpoides dorychaeta 635 


Palpal femur with 1 dorsal, simple seta; genu with 1 dorsal simple seta; 
tibia with 1 lateral, 1 medial and 1 dorsal, simple setae, and a terminal claw 
of thumb-claw complex. 

Coxa I with 2 simple setae ventrally; trochanter bare; femur with 1 
broadly serrate and 2 simple setae; genu with 3 broadly serrate setae and a 
simple one; tibia with 2 serrate and 4 simple setae and 1 short, peglike, 
sensory seta; tarsus with 3 solenidia, 2 simple, 2 terminal, spatulate-plumose 
setae, and 2 distal sets of duplex setae, the distal members greatly exceeding 
proximal members in length and the lateral distal member exceeding other 
3 in length. 

Coxa II with 1 spiculate seta; trochanter bare; femur with 1 simple, 1 
spiculate, and 1 broadly serrate seta; genu with 2 serrate and 2 simple setae; 
tibia with 3 serrate and 2 simple setae; tarsus with 3 solenidia, 3 simple, 2 
terminal, spatulate-plumose setae, and 1 distal set of duplex setae, distal mem- 
ber rodlike and shorter than simple proximal member. 

Coxa III bare; trochanter bare; femur with 1 serrate and 1 simple seta; 
genu with | simple seta; tibia with 1 serrate, 1 spiculate, and 1 simple seta; 
tarsus with 5 simple and 2 terminal, spatulate-plumose setae. 

Coxa IV bare; trochanter bare; femur with 1 simple seta; genu with 1 
simple seta; tibia with 1 serrate and 2 simple setae; tarsus with 3 simple and 
2 terminal, spatulate-plumose setae. 

Ventrally with 3 pairs of simple podosomal setae, the anterior pair medial 
to coxae I and II, mesal pair between propodosoma and metapodosoma, and 
posterior pair medial to coxa III; opisthosoma with 1 pair of simple setae 
anterior to anus. 

Larva (Fig. 9, 10): Males and females are not distinguishable from each 
other. Length of idiosoma 143 to 185, width 106 to 131, averaging 168 x 
118» for 10 individuals. 

Palpal femur with 1 dorsal simple seta; genu with 1 dorsal simple seta; 
tibia with 2 simple setae, and a terminal claw of thumb-claw complex. 

Coxa I with 1 simple ventral seta; trochanter bare; femur with 1 broadly 
serrate and 2 simple setae; genu with 2 serrate and 2 simple setae; tibia with 
5 simple setae and a short, peglike, solenidium; tarsus with 4 simple, 2 
terminal solenidia, and 1 distal set of simple duplex setae, the distal member 
of set considerably longer than proximal member. 

Coxa II bare; trochanter bare; femur with 1 broadly serrate and 2 simple 
setae; genu with 2 serrate and 2 simple setae; tibia with 5 simple setae; tarsus 
with 2 solenidia, 3 simple, and 1 distal set of duplex setae, the distal member 
of set rodlike and shorter than proximal simple member. 

Coxa III bare; trochanter bare; femur with | serrate and 1 simple seta; 
genu with 1 simple seta; tibia with 1 serrate and 3 simple setae; tarsus with 
3 simple and 2 terminal, spatulate-plumose setae. 


636 Tue UNiversiry SCIENCE BULLETIN 


Ventrally podosoma with 2 pairs of simple setae, anterior pair medial to 
coxae I and II, and posterior pair medial to coxa III. 

The above descriptions were made from specimens collected in Lawrence, 
Kansas on Gleditsia triacanthos L. or progeny from these specimens reared in 
the laboratory. 


Type specimens are females and males, collected in Natchez, Louisiana, 


from black locust (Robinia pseudo-acacia L.). Holotype female, type No. 
2175, in the U.S. National Museum, Washington, D.C. The known distri- 
bution of this species includes Louisiana, North Carolina, Utah and Kansas. 


BIBLIOGRAPHY 


ANpeERSEN, V. Sr. 1948. Untersuchung tiber die Biologie und Bekampfung der Obstbaumspinn- 
milbe Paratetranychus pilosus. Dissertation Bonn, As quoted in Kremer (1956). 

Biauve.t, W. E. 1945. The internal morphology of Tetranychus telarius. Cornell Univ. Agr. 
Exp. Sta. Mem. 270:25-26. 

Cacie, L. R. 1943. The life history of the spider-mite Tetranychus schoenit. Virginia Polytech. 
Inst. Agr. Exp. Sta., Tech. Bull. 87:16pp. 


Sta., Tech. Bull. 98:19pp. 

. 1949. The life history of the two-spotted spider-mite. Viriginia Polytech. Inst. Agr. 

Exp. Sta., Tech. Bull. 113:31pp. 

Camin, J. H. 1953. Observations on the life history and sensory behavior of the snake mite, 
Ophionyssus natricis (Dermanyssidae). Chicago Academy of Sciences Spec. Publ. 10: 
8-15. 

EncuisH, L. L., AND R. SNETsINGER. 1957. The biology and control of Eotetranychus multidigi- 
tuli, a spider-mite of honey locust. Jour. Econ, Ent. 50:135-141. 

Ewinc, H. E. 1914. The common red-spider or spider-mite. Oregon Agr. Coll. Exp. Sta. Bull. 
121:95pp. 

Garman, P., anv J. F. Townsenp. 1938. The biology and control of the European red-mite. 
Connecticut Agr. Exp. Sta. Bull. 418:34pp. 

Granpyean, F. 1948. Quelques Caractéres des Tétranyques. Bull. du Mus. Hist. Nat. Paris 
(series 2) 20:517-524. 

Kremer, F. W. 1956. Studies on the biology, epidemiology and control of Bryobia praetiosa 
Koch. Hofchen-Briefe, Bayer PAlanzenschutz-Nachrichten 9:189-252. 

Prircuarp, A. E., aNp E. W. Baker. 1955. A revision of the spider mite family Tetranychidae. 
Pac. Coast Ent. Soc. Mem. Ser. 2:422pp. 

Rekk, G. F. anp A. T. BacpasariAn. 1948. Dokaldy Akad. Nauk. Armenia. U.S.S.R. 9:183-186. 

Scuraver, F. 1923. Haploidie bei einer Spinnmilbe. Arch. Mikrosk. Anat. 97:610-622. 


.1946. The life history of the European Red-Mite. Virginia Polytech. Inst. Agr. Exp. | 


Tue Bronomics oF Tenuipalpoides dorychaeta 637 


Fic. 1. Tenuipalpoides dorychaeta. Dorsal aspect of adult female. 


638 Tue Universiry ScitENCE BULLETIN 


Fic. 2. Tenuipalpoides dorychaeta. Ventral aspect of adult female. 


ENG252 


Tenuipalpoides dorychaeta. Dorsal aspect of adult male. 


635 


ITY ScIENCE BULLETIN 


640 


. Ventral aspect of adult male. 


Fic. 4. Tenutpalpoides dorychaeta 


Tue Bronomics or Tenuipalpoides dorychaeta 


Figs 5 


Fic. 5. Tenuipalpoides dorychaeta. Dorsal aspect of deutonymph. 


641 


Fig.46 
of deutonymph. 


Tue Bronomics oF Tenuipalpoides dorychaeta 


Ge 7 


Fic. 7. Tenutpalpoides dorychaeta. Dorsal aspect of protonymph. 


644 


Tue Bronomics or Tenuipalpoides dorychaeta 645 


Fic. 9 & 10. Tenuipalpoides dorychaeta. Ventral (Fig. 9) and dorsal (Fig. 10) aspects of larva. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


A REVISION OF THE GENUS EXEMA 
OF AMERICA, NORTH OF MEXICO 
(CHRYSOMELIDAE, COLEOPTERA) 


By 


Jay B. Karren 


ee 
' — 


Voi. XLVI Paces 647-695 DEcEMBER 2, 1966 No. 18 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


VoL. XLVI Paces 647-695 DEcEMBER 2, 1966 No. 18 


A Revision Of The Genus Exema Of America, North 
Of Mexico (Chrysomelidae, Coleoptera)' 


By 


Jay B. Karren 


ABSTRACT 


In this revision of the genus Exema of America, north of Mexico, nine species 
are recognized and separated into three well-defined groups of one, five, and three 
species. A key to the species of this area is included. Three new species, Exema 
mormona, E. elliptica, and E. byersi are described and neotypes are designated for 
E. gibber Fabricius and E. dispar Lacordaire. Seven names are placed in synonymy 
for the first time. Although primarily taxonomic, this work brings together avail- 
able data concerning variation, distribution, host plants, and parasites of a few 
species. A total of 6,296 specimens were included in this study. 


INTRODUCTION 


The object of this paper is to redefine the genus Exema and to make 
recognizable its species from America, north of Mexico. Fourteen specific and 
two subspecific names have been proposed for the genus in this area. Ten of 
these names are suppressed as junior synonyms. The nine species recognized 
in this paper can be separated into three groups. These groups have not been 
given formal names; the divisions are made only to show the relationships of 
the species within the genus. 

Exema and its close relatives are noteworthy for their cryptic form and 
color, giving them a remarkable resemblance to caterpillar droppings, mouse 
feces, or other types of debris. The genus is closely related to Chlamisus, and 
some authors synonymize the two. The two genera are distinct, however, in 
North America; therefore, Exema is treated here as a separate genus. Species 
of Exema are found throughout the United States and in southern Canada 
from Manitoba eastward. Several of the southwestern species extend into 
Mexico and Central America. 


* Contribution number 1314 from the Department of Entomology, The University of Kansas, 
Lawrence, Kansas. 


648 Tue University ScreNcE BULLETIN ) 
| 
Brown (1943) indicated that species of Exema and Chlamisus of Canada | 
have strong monophagous tendencies. In correspondence, he suggested that ] 
undertake a study of the species of Exema and their food-plant relationships, 
Thereafter, in collecting Exema I took special note of plants on which the 
beetles were feeding. My field observations indicate that a given species feeds — 
on one or on a limited variety of plants. Two or more species have seldom 
been collected on the same food plant in the same locality. In a few cases ] 
have collected two or three specimens on “wrong” plants along with a con- © 
siderable series of each species on the respective host plants. By study of a) 
series of beetles from the same species of food plant in a locality, one is also © 
able to obtain an indication of the degree of variability within a species. 
Of the nine species in this study, nodulosa, mormona, byerst, and cana-_ 
densis definitely have monophagous tendencies. The other five have wider | 


i 
i 
Few specimens in museum collections have host labels, but the infor- 


preferences, but four of them are restricted to composites. 


mation obtained from those has been of value. Some such host records seem 
to disprove the idea of monophagy; however, since larvae often leave the food — 
plants to pupate and adults may wander to other species of plants, records: 

unsupported by observations of feeding may not represent actual hosts. A 
large series collected from a single plant species represents a fairly reliable 
host record. A list of plants on which Exema has been collected is given for 
each species. These lists may contain true host plants as well as those that ~ 
are not. 


MATERIALS AND METHODS 


This study is based upon my personal collections, those of the Snow 
Entomological Museum, and those of 48 museums listed in the acknowledg- 
ments. A total of 6,296 specimens were examined and their distribution was 
recorded. County names are placed in parentheses when they do not appear 
on specimen labels. Complete label data are given for primary types. The 
symbols in parentheses following the label data indicate the institutions where 
type material is deposited. 

Six species were observed in the field before collection. Three of these 
were observed through a complete life cycle in the laboratory. Voucher speci- 
mens of these are in my collection and that of the Snow Entomological 
Museum. 

The types of all species and synonyms were examined except those of 
Chlamys rugulosa Motschulsky, Exema dispar Lacordaire, and Clythra gibber 
Fabricius. Attempts to locate the type of rugolosa was unsuccessful. Its 
synonymy with conspersa is quite certain because it is the only species on the 
Pacific Coast, and the only one common in California. To clear up some 


A REVISION OF THE GENUs Exema or America, Nortu oF Mexico 649 


confusion and create greater stability, neotypes have been designated for the 
other two species. 

Representatives of several species were compared with a specimen (not the 
type) of Clythra gibber Fabricius in the collection of this authority by Sv. G. 
Larsson of the Universitetets Zoologiske Museum of Copenhagen, Denmark. 
The comparison made by Larsson indicates a species different from that of 
my understanding. Considering the Fabrician description and type locality, a 
neotype is designated for that species. 

Specimens were studied at 15, 60, and 120 magnifications. Drawings and 
measurements were made with an ocular grid. Overall length was measured 
from the front of the head, when recessed within the pronotum, to the tips of 
the elytra. Width was measured between the outer edges of the humeral 
umbones. Drawings of genitalia were made after clearing them in KOH. The 
genitalia were removed by placing dried specimens in KOH which was 
brought to a boil, after which the specimens were immediately transferred to 
95°% alcohol. After a few seconds in alcohol, they were removed and held be- 
neath the microscope between the thumb and index finger, and the pygidium 
was raised with a fine needle or forceps, the latter being used to remove the 
parts and to place them in hot KOH for a few minutes to clear. The genitalia 
were then immersed in alcohol for about a minute and transferred to a drop 
of glycerine in a small dish. All drawings of the dorsal aspect of the male 
genitalia were made with the apical half as nearly horizontal as possible. 
Female parts were prepared for drawing by spreading in glycerine between 
two cover-slips. The terminology used for the female parts is that used by 
Pierce (1940). 

All Exema studied have a similar pattern of tubercles to which names are 
assigned to facilitate description. Exema gibber is illustrated (Fig. 1) for the 
purpose of indicating this terminology. All descriptions were made from the 
holotype or neotype. Variation is treated in the discussion. 

Collecting specimens individually in the field was more successful than 
sweeping. After sweeping an area to determine which kind of plants had 
beetles, a methodical examination of the plants was made to procure a maxi- 
mum number of them. A net was useful under the leaf or plant to catch 
beetles that dropped. When sweeping was the only method used, many beetles 
fell to the ground as soon as the plant was disturbed and were impossible to 
locate. The ability to fold appendages compactly enables these beetles to roll 
into ground litter and debris and disappear easily. Needless to say, all “cater- 
pillar droppings” and similar objects should be examined whenever these 
beetles are being collected. Both adults and larvae occur on the same plants. 
All larvae collected were taken to the laboratory and fed with leaves of the 
plant from which they had been removed. Eventually they pupated and 


650 


Tue UNiversiry ScIENCE BULLETIN 


emerged as adults. Sometimes parasitic Ichneumonidae, Chalcidoidea, and’ 
Chalcidae emerged from the pupal cases. 


The following list of abbreviations is used throughout this paper in 


parentheses to indicate the individual, museum or University in which the 


types are housed. 
AMNH 


ANSP 
BYU 
CAS 
CDA 
CM 
CU 
FDA 
ISU 
JAW 
JBK 
KSU 
LACM 
MCZ 
MICH 
MSC 
MSU 
MZU 
NYSM 
OSU 
PU 
RU 
SCU 
UA 


U 


U 
UCR 
UK 


CB 


M 


UMIN 
U 
U 


HISTORY OF THE GENUS EXEMA IN NORTH AMERICA 


The generic name Exema was first used by Lacordaire (1848) for 16 
species, mostly South American. In his key he used the character, antennae 
dentate from the sixth segment, to distinguish Exema from Chlamys (now 
Chlamisus). This character has since been used by various authors even 
though it is not reliable in some species. For example, if this character alone 
is used, Exema gibber males (Fig. 4) would be placed in the genus Chlamisus 
while the females (Fig. 3) would remain in Exema. A few species show an 
enlarged fifth antennal segment, making it difficult to decide in which genus 
they belong. In all cases the fifth segment is smaller than the sixth, and, with 
the additional characters of the shape of the aedeagus and ejaculatory guide 
and spines or spinulae on the male sternum and tibiae, it appears wise to re- 


N 
SNM 


American Museum of Natural History 
Academy of Natural Sciences of Philadelphia 
Brigham Young University 

California Academy of Sciences 
Canada Department of Agriculture 
Carnegie Museum 

Cornell University 

Florida Department of Agriculture 
Iowa State University 

J. A. Wilcox collection 

Jay B. Karren collection 

Kansas State University 

Los Angeles County Museum 
Museum of Comparative Zoology, Harvard 
Michigan State University 

Montana State College 

Montana State University 

Museum Zoologicum Universitatis, Helsinki 
New York State Museum 

Ohio State University 

Purdue University 

Rutgers University 

Snow College, Utah 

University of Arizona 

University of California, Berkeley 
University of California, Riverside 
University of Kansas 

University of Missouri 

University of Minnesota 

University of Nebraska 

United States National Museum 


tain the name Exema for this group in North America. 


A RevISION OF THE GENUs Exema or Amegrica, Nortu or Mexico 651 


Of the 16 species treated by Lacordaire, ten were described as new. The 
other six had appeared earlier in the literature under the generic names 
Chlamys or Clythra. Only two of his species are North American. 

The earliest description of an Exema is that by Fabricius (1798), for gibber 
in the genus Clythra. Olivier (1808) transferred this species to the genus 
Chlamys and emended the specific name, making it gibbera. From 1808 until 
the erection of the genus Exema, the species remained in the genus Chlamys. 

Two other names were proposed for North American species prior to 
Lacordaire’s work but were not recognized by him as belonging to the genus. 
The first was Chlamys conspersa Mannerheim (1843), the other Chlamys 
rugulosa Motschulsky (1845), both of which Lacordaire included in the genus 
Chlamys on the basis of the original descriptions. 

Chapuis (1874) recognized the 16 species of Lacordaire and one additional 
species, Exema malayana Baly (1865). Other than giving another description 
of the genus and general distributions of the species, Chapuis’ work added 
little to the knowledge of Exema. 

In 1914, Achard summarized available information on the sub-family 
“Chlamydinae” and published the names, synonyms, and distributions of the 
species. He listed 27 species, mostly South American, but two North Amer- 
ican. 

Chlamys nodulosa Blatchley (1913) and Exema neglecta Blatchley (1920) 
were the next names proposed for North American species of Exema. The 
genus then remained unchanged until 1940, when Pierce proposed seven new 
species and two subspecies. Most of these are variations of the two western 
species. 

Since 1848, no one has redefined the genus, but authors have repeated the 
description given by Lacordaire. Many workers have felt that since the an- 
tennal character varies, the genus should be synonymized under Chlamisus 
(Chlamys of earlier authors). In fact, Chdy6 (1940) published such a proposal 
except that he recognized Chlamys as the synonym of Exema. He later (1955) 
changed this to Chlamisus with Exema as the synonym. In studying the 
North American species, I find the two genera quite distinct though closely 
related. Although the antennal character varies among the species of Exema, 
other characters that are more constant help to define the genus, and the two 
genera can be recognized easily as separate taxa in North America. More 
study on a world-wide basis is needed for a better understanding of Exema 
and its relationships to the other genera in the Chlamisinae. 


THE GENUS EXEMA 


Exema Lacordaire; 1848:844; Chapuis, 1874:204; Jacoby, 1881:89; 1908:278; Achard 1914:16- 
17; Pierce, 1940:7. Type species: Chlamys intricata Kollar, 1824 (designation by Jacoby, 1908). 


Description: 2.1-3.5 mm long. Black to dark brown, usually marked with 


652 Tue University ScrENCE BULLETIN 


yellow on head, pronotum, elytra, and legs, sometimes coppery metallic. Body 
subquadrate or subpentagonal as seen from above, distinctly convex dorsally 
and ventrally. Head deeply inserted into prothorax, front inflexed ventrally; 
face flat, coarsely punctate, each puncture with a small hair which may be 
obscure. Antenna held in deep groove on side of prosternum in preserved 
specimens, scape slightly curved, large, as long as next 4 segments combined, 
ridged on outer side; pedicel globular, larger than next 2 segments; segments 
3 and 4 subequal in length; segment 5 larger, sometimes nearly as large as 
each of the remaining segments; segments 6-10 strongly transverse, subequal 
in size, serrate, segment 11 longer, subtriangular, rounded at apex. Labrum 
slightly emarginate at apex, smooth, with a transverse row of hairs; clypeus 
not separated from frons, ventral margin shallowly emarginate. Eyes large, at 
sides of face, strongly emarginate at middle of inner margin. Pronotum 
strongly narrowed anteriorly, with more or less rounded sides; all margins 
closely paralleled by a continuous sulcus containing a row of punctures; gib- 
bose (elevated median disc); tuberculate, carinate, or both, depending on 
species; strongly rugose to punctate. Elytra elongate, tuberculate and ridged; 
sutural margin completely dentate, humeral umbone prominent, less punctate 
and tuberculate than disc of elytron. Legs short, robust, gently compressed, 
drawn into excavations on sides of thorax and abdomen in preserved spect 
mens: front tibia of male with short apical spine on inner side, usually a simr- 
lar spine on middle tibia; tarsomeres padded beneath, tarsal claws simple or 
toothed (described as appendiculate by many authors). Prosternum (called 
basisternite by Pierce) large, narrowed posteriorly, highly variable in shape; 
prosternal process fits into notch in mesosternum. Metasternum large and 
prominent, with large, round punctures. Abdomen with 5 visible sterna, Ist 
with a median longitudinal carina, deeply hollowed on each side, with a pair 
of spines or many spinulae; 5th (last) sternum much longer than 3 preceding 
combined but smaller than Ist and generally with a median fovea in both 
sexes, flatter in males; many hairs associated with foveal area in both sexes. 
Male aedeagus pistol shaped in lateral view (Fig. 9); tegmen (Fig. 9) Y- 
shaped in apical view; dorsal median orifice near middle of aedeagus; tip 
flanged, truncate, or rounded, curved ventrally; ejaculatory guide short, torso- 
shaped (Fig. 11) with a mid-ventral and apical orifice through which flagel- 
lum passes. 

Female genitalia used for covering egg with excrement, not developed into 
an ovipositor; 8th tergum with apical groove; 9th and 10th segments each 
represented by a pair of tergites and sternites in addition to pleura; auxiliary 
sclerite associated with each 9th sternite; 11th segment mostly membranous. 

Comparisons: The small third to fifth antennal segments serve for most 
species to distinguish Exema from its near relatives in CAlamisus. In all cases 
the fifth segment is smaller than the sixth. Only one species of Exema is dis- 


A REVISION oF THE GENUs Exema oF America, Nortu oF Mexico 653 


tinctly metallic, another slightly so, while many of Chlamisus are metallic. 
The elytral serration is always complete in Exema but complete or incomplete 
in Chlamisus. Species of Exema are usually smaller than those of Chlamisus. 
Males of Exema have front tibial and middle tibial (lacking in gibber) spines; 
females have neither. In Chlamisus some species have the spines in pairs, 
some single, and some have spines on the female tibiae. North American spe- 
cies of Exema have either spines or spinulae on the first and second abdominal 
sterna of the males. These spines are lacking on all Chlamisus that I have 
examined. The general shape of the aedeagus is constant in Exema, and the 
shape of the ejaculatory guide (called flagellum by Brown) is distinctive. It is 
a short, tubular, sclerotized structure as compared with the elongate, paired 
structure found in Chlamisus (see Figs. by Brown, 1943:126). More study is 
needed to determine the taxonomic value of the female structures. 


KEY TO THE NORTH AMERICAN SPECIES OF EXEMA 


I. Tarsal claws simple; cuticle all or in part black, not or only slightly me- 
tallic; pronotum at least in part rugose or striopunctate, tip of aedeagus 
Romine or sliphtly famged 2 

Tarsal claws toothed; cuticle yellow, marked with black, or metallic: 
pronotum punctate; tip of aedeagus round, truncate or highly 
BE lee ee =e st cae ie 6 


2. Frons in emargination of eye yellow; face of male mostly yellow; frontal 
slope of male pronotum always yellow; pronotum in part. strio- 
Peart area ee SE ass a A as 3 

Frons in emargination of eye black; face of male black with a yellow 
pattern (Fig. 20); frontal slope of male pronotum black or with re- 
duced yellow markings; entire pronotum highly rugose or. strio- 
RPUMCL GS | ee ee, eee Aer ete i A eee ee 5 


3. Nontubercular surface of pronotum rough, coarsely punctate anteriorly, 
punctures round and deep, especially on yellow areas of male; punc- 


Bees REION PALE POSUCTIONLY, ccc tee oc cetct ees ee eeene nee neglecta 
Nontubercular surface of pronotum not rough, striopunctate through- 
srioupuncturcs) Oval tovellipticall’. 0. cia... 4 
4. Cuticle slightly metallic; pronotal punctures elliptical, tubercles numer- 
ous, large; gibbosity high, flat in dorsal outline elliptica n. sp. 
Cuticle dull or shiny black; pronotal punctures oval to elliptical, tubercles 
sparse, small; gibbosity evenly rounded byersi n. sp. 


>. Sutural tubercle 3a (Fig. 1, su-3a) and some spots on pronotum always 
bright yellow; bend of aedeagus smooth, tip truncate ........ mormonad Nn. sp. 

Sutural tubercle 3a and the rare pronotal spots usually black, never both 
yellow; bend of aedeagus rugose, tip flanged canadensis 


654 Tue University ScIENCE BULLETIN 


6. Cuticle coppery to dark brown or black with coppery reflections; tip of 
aedeagus greatly flanged laterally and truncate with a dorsal, transverse 
row of long hairs; male with front tibial spine only; female with large 
auxiliary sclestes) (Bis: 7,aUx) 22 ee gibber 

Cuticle not coppery but yellow and black; tip of aedeagus truncate or 
rounded, with lateral and ventral hairs only; male with front and mid- 
dle tibial spines; female with small auxiliary sclerites, sometimes in- 


} 


CONSPICUOUS,” 25 fee in rr 7 
7. Tooth of tarsal claw either small or blunt (Figs. 53-55); tubercles on 
pronotum and elytra usually small and blunt . desert 
Tooth and tarsal claw large and acute (Fig. 56); tubercles on pronotum 


and elytra usually large and distimct 22-22 c- cee 8 


8. Punctures distinctly setigerous; third tarsal segment about 0.5 as long as 
fourth; tarsomeres usually brown, more yellow in specimens from the 
Paciie Coast sal Se conspersa 

Punctures not or only faintly setigerous, especially on head and pro- 
notum; third tarsal segment 0.7 as long as fourth or more; tarsomeres 
112 i are enon Pe ee ne rere Oe EE ec cccossaseoss dispar 


Exema gibber (Fabricius) 


Clythra gibber Fabricius, 1798:112; 1801:33; Zimsen, 1964:120 

Clytra gibber Fabricius; Coquebert, 1804:129. 

Chlamys gibber (Fabricius); Mliger, 1804a:126; 1804b:164; Schoenherr, 1808:343; Klug, 
1824:116-117; Dejean, 1837:439 (listed). 
Chlamys gibbera (Fabricius) ; Olivier, 1808:876. 
Chlamys nodulosa Blatchley, 1913:22-23; 1930:37, 52 (lectotype designation as Exema nodu- 

losa). 

Chlamisus nodulosus (Blatchley); Brown, 1961:971 (note on name change). 

Exema gibber (Fabricius); Lacordaire, 1848:849-852; Crotch, 1873: 30; Beutenmiiller, 1890:175 
(biology); Blatchley, 1910:1116; 1920:69-70; Chagnon, 1937-228 (misidentification); Pierce, 
1940:12: Brown, 1943:124 (of Canadian authors=E. canadensis Pierce); Wilcox, 1954:394- 
395. 

Exema gibber (Olivier; Dury, 1879-11 (probably canadensis); Henshaw, 1885:106; Smith, 
1889:215 (distribution); Linell, 1879-480; Leng, 1918:208 (synonymizes nodulosa with 
gibber); 1920:288 (listed); Blatchley, 1924:50 (note on synonymy and biology); Fattig, ” 
1948:8 (distribution). 

Exema gibbera (Fabricius); Gemminger and Harold, 1874:3308 (listed); Clavareau, 1913:222; 
Achard, 1914:17 (listed). | 

Exema nodulosa (Blatchley); Dekle, 1957:331-333 (biology). 

| 


Diagnosis: This is the easiest of the North American forms to identify.” 
The males are unique in having spinulae on the first and fifth abdominal 
sterna (Fig. 2). This is the only species which lacks a spine on the apex of the 
middle tibia. This species is more punctate and tuberculate than any of the 
others. Like conspersa and related taxa, it lacks striate or rugose lines on the 
pronotum, as opposed to canadensis and related species which have at least a 
few on the posterior half. This species is found throughout most of the south- 
ern states (Fig. 58) on oak and a few other plant genera. 


A Revision oF THE GENUs Exema or America, Nortu oF Mexico 655 


Type: Clythra gibber Fabricius, “Carolina” (probably near Charleston, 
South Carolina), Dom Bosc. The Paris museum staff is unable to locate this 
type. Blake did not find it in 1951. 

According to Blake (1952) the Bose specimens came from near Charleston, 
South Carolina between 1798 and 1800 and were taken to France and de- 
scribed by Fabricius and Olivier. Zimsen (1964:16) states that Bosc collected 
in and around New York. Upon his death, the Bosc collection was placed in 
the Museum National d’Histoire Naturelle, Paris. The type has not been 
located in that museum. The British Museum has a specimen labeled “Exema 
gibber type” which R. D. Pope (correspondence) feels is not the type but one 
sent to Olivier by Bosc and assumed to be the same as the one described by 
Fabricius in 1798. This specimen is the same as my canadensis. A speciman in 
the Fabrician collection, not the type according to Sv. Larsson of Copen- 
hagen’s Universitetets Zoologiske Museum, is similar to my byerst. Because 
of so much confusion with regard to this species, and because I feel the true 
type is lost, I am designating as a neotype a male collected in South Carolina 
(UK). I am assuming that the published record by Fabricius is the type local- 
ity. My study of byers: indicated that it is not found near South Carolina. Both 
Olivier and Lacordaire speak of the coppery color of which gibber is quite 
distinctive. There is some evidence that the type of Fabricius was used by 
these men in their descriptions. Also the darker forms of gibber agree better 
with the original description than does byersi or canadensis. Although the 
distribution of gibber is more extensive in Florida, it is found also in South 
Carolina. The species is somewhat variable in color but quite distinctive. The 
one that Blatchley described as nodulosa from Florida is a synonym of gibber. 

Description: Male neotype: 2.6 mm long; 1.8 mm wide. Dark brownish 
bronze; antennal scape pale; remaining segments becoming darker at apex; 
yellow, smooth spot on frons between eyes; yellow triangle between antennae 
above clypeus; tubercles pale and shiny; tarsomeres brown. Entire cuticle 
minutely granulate, irregularly punctate. Head coarsely punctate; large punc- 
tures with recurved hairs; depressed area between antennal bases; antenna 
sparsely clothed with short hair; segments 3, 4, and 5 subequal, segment 6 
triangular, segments 7-11 subequal; labrum smooth, slightly emarginate at 
apex, with a few ventrally projecting hairs. Pronotal gibbosity with rounded 
and acute tubercles; cephalic small and round; anterior larger and connected 
feebly to summit-1 by a small acute carina; summit-l with carina extending 
anterior to large, round lateral tubercle; summit-2 acute; marginal small; en- 
tire pronotum highly irregular and tuberculate; scutellum with rounded lat- 
eral wings; median carina small. Elytra coarsely punctate; umbone large and 
prominent; large tubercles: anterior, sutural-l, la, 2, 3, 3a, and marginal; 
small tubercles: sutural-2a, 2b, 3b, umbonal-a, marginal-a, marginal-b, and 
apical; other tubercles faintly developed; sutural-la carinate to umbonal-b. 


656 Tue Universtry ScreNce BULLETIN 


Tet 


FS 
E 
ies 
eS 
re 
— 
‘es 
\ 

c . 


A RevIsIon oF THE GENUS Exema or America, Nortu oF Mexico 657 


Legs shallowly punctate; tarsal claws broadly toothed; no spines on tibia or 
sterna. Prosternum coarsely punctate, concave; prosternal process broad, 
curved very slightly dorsally at tip; equal in length to concaved portion. Meta- 
sternal punctures large and deep, especially laterally. First abdominal sternum 
with a longitudinal carina between coxae, lateral side of sternum highly punc- 
tate with 1 large and 2 smaller tubercles; midventral area of 5th sternum flat 
and hairy. Pygidium gibbose with a pair of central, 2 elongate lateral, and 4 
basal depressions; highly punctate, hairs in punctures. Aedeagus (Figs. 9, 10) 
greatly flanged at apex, with a heavy row of long, dorsal hairs. 

Female: 3.2 mm long; 2.1 mm wide. Similar to male except lacks yellow 
on head. Tubercles are slightly less distinct. Midventral area of 5th sternum 
foveolate surrounded by hairs. Central pygidial carina less distinct. Genitalia 
as in Figs. 6-8. 

Discussion: Blatchley listed this species as 4-4.5 mm long, but males are 
2.5-3.2 mm and females 2.8-3.5 mm long. The form of the tarsal claws and the 
metallic appearance of the cuticle indicate this species’ close relationship to the 
genus Chlamisus. The males would be identified as Chlamisus by use of pre- 
vious keys, while the females would key out to Exema. This is because of the 
dimorphism of the antennae (Figs. 3-4). The species was described in Chla- 
mys and placed in Chlamisus by Brown (1961). I have returned it to Exema 
for several reasons. First, the sculpturing of the pronotum and elytra is very 
similar to that of the other members of the genus. Second, the shape of the 
ejaculatory guide (Fig. 11) is similar to that in the rest of the genus and dif- 
fers radically from that of any Chlamisus I have seen, including all species 
figured by Brown (1943). The shape of the aedeagus differs somewhat from 
the rest of Exema, but it differs equally from shapes seen in Chlamisus. Third, 
the elytral serration continues forward to the scutellum as in Exema; how- 
ever, most Chlamisus have this same character. Considering these and other 


Fic. 1. Dorsal pattern of sculpturing of Exema gibber: ac—anterior carina, ae—anterior elytral, 
alc—anterolateral carina, anp—anterior pronotal, ap—apical, ce—cephalic, dd—discal 
depression, es—elytral serration, la—lateral, ma—marginal, ma-a—marginal-a, ma-b 
marginal-b, mp—marginal pronotal, sc—scutellar, scm—scutellum, sm-l—summit-l, 
sm-2—summit-2, su-!—sutural-1, su-la—sutural-la, su-2—sutural-2, su-2a—sutural-2a, 
su-2b—sutural-2b, su-3—sutural-3, su-3a—sutural-3a, su-3b—sutural-3b, um—umbone, 
um-a—umbonal-a, um-b—umbonal-b. 


Fic. 2. Ventral aspect of Exema gibber, male: abs—abdominal sterna, ag—antennal groove, 
cc—coxal cavity, cl—clypeus, argination, eep—elytral epipleuron, el—elytron, 
fc—frontal carina, fo—fovea, fr—frons, ge—gena, lab—labrum, md—mandible, mem— 
mesoepimeron, mes—mesoepisternum, ms—metasternum, msp—male spinula, mte— 
metaepisternum, pe—pygidial carina, pe—proepisternum, pn—pronotum, pp—prosternal 
process, ps—prosternum, py—pygidium, sp—spine of male protibia, um—umbone, 
ve—vertex. 


Fic. 3. Antenna of female Exema gibber. 
Fic. 4. Antenna of male Exema gibber. 


Fic. 5. Tarsal claw of Exema gibber. 


658 Tue Universiry ScrENcE BULLETIN 


; 0.5mm. fi 


SCALE, FIGS. 6-10 


Fics. 6-8. Dorsal aspect of Exema gibber, female abdominal segments eight to eleven; ax— 
auxiliary sclerite, pl—pleurite, s—sternite, seg—segment, t—tergite. 


Fics. 9-11. Exema gibber, male genitalia; 9—lateral aspect; 10—dorsal aspect; 11—ejaculatory 
guide; ao—apical orifice, bf—basal foramen, ejd—ejaculatory duct, fg—flagellum, 
lg—lateral groove, mdp—median dorsal plate, tg—tegmen. 


characters, gibber shows affinities to Chlamisus, but it is more closely related’ 
to the genus Exema. For this reason it occupies a unique position in the genus 


A Revision OF THE GENUS Exema or America, Nortu oF Mexico 659 


as compared to the other 2 groups, which are represented by 5 and 3 species 
respectively. 

Plant records: Quercus sp., Myrica cerifer, Crataegus, pecan, and Lichi 
chinensis. Dekle (1957) also records Baccharts halimifolia, Rubus spp. (black- 
berry and dewberry), and Salzx sp. 

Specimens examined: (92 males and 100 females). FLORIDA: Alachua Co., 13 Apr.-2 Oct.; 
Bevard Co., Mar.; Dade Co., 6-12 May; DeSoto Co., 9 Apr.; Duval Co., 19 Apr.-9 May; Hardee 
Co., 20 Mar.-3 Apr.; Highlands Co., 19 Apr.-31 May; Hillsborough Co., 1 Apr.-15 May; Lake 
@o., 1-11 Mar.; Lee Co., 6 May; Levy Co., 4 Apr.; Liberty Co., 24 Apr.; Mamatee Co., 25-26 
Mar.; Monroe Co., Apr.-13 May; Okaloosa Co., 16 May; Orange Co., Apr.-6 May; Palm Beach 
Co., 7-10 Apr.; Pasco Co., 16 Apr.; Pinellas Co., 15 Apr.-26 Aug.; Putnam Co., 3-26 Apr.; 
Sarasota Co., 11 May; Seminole Co., 28 Mar.-7 May; Taylor Co., 22 Apr.; Volusia Co., 15 Apr.- 
25 May; ‘““Gunntown,” Apr.; “Haulover,” 3-17 Mar. GEORGIA: Chatham Co., 2 Apr.-27 May; 
Decatur Co., 18-21 May; Tift Co., no date. LOUISIANA: (Parish = county of other states) 
Caddo Par., 23-27 Mar. MISSISSIPPI: George Co., 7 May. PENNSYLVANIA: Philadelphia Co., 
15 June. SOUTH CAROLINA: Beaufort Co., no date. TEXAS: Brazos Co., 1 Apr.; Colorado 
Co., 24 Apr.; Goliad Co., 18 Apr.; Harrison Co., 25 Mar.; Houston Co., 2 Apr.; Jefferson Co., 27 
Apr.; Marion Co., no date; Trinity Co., 20 Mar. 


Exema mormona n. sp. 


Diagnosis: The posterior slope of the pronotum is highly rugose; the an- 
terodorsal surface is striopunctate. The posterodorsal bend of the aedeagus is 
smooth (Fig. 17) compared to highly rugose in canadensis (Fig. 25). The tip 
is more truncate and the basal half of the tegmen is much more slender than 
in canadensis, neglecta or elliptica. 

This species is found most commonly from Texas to Montana, west of the 
Great Plains and east of the Pacific coast states (Fig. 59) on Gutierrezia. 

The distribution, unusual color pattern, and host plant of this species first 
suggested the possibility of its being new. The following description is based 
upon 53 males and 54 females. 

Description: Male: 2.7 mm long; 1.7 mm wide. Shiny black with yellow 
or yellowish brown markings. Antenna yellow at base, becoming dark brown 
at apex; mouthparts dark brown, labrum light at apex; face with longitudinal 
yellow bar next to each eye, transversely connected to central yellow spot 
which merges with triangular frontal spot covering carinal area. Frontal slope 
of pronotum with bright yellow inverted V just behind antecostal suture, yel- 
low also extending laterally next to this suture for a short distance on each 
side (broken on the left side); marginal tubercle prominent, with small red- 
dish yellow spot; yellow spot on anterior pronotum just behind eyes. Sutural 
tubercle 3a bright yellow. All tibiae with subbasal and subapical yellow rings; 
tarsomeres dark brown, nearly black; femora black. 

Cuticle minutely granulate, coarsely punctate, more sparsely so on yellow 
than on black areas; head punctures large, mostly shallow, some deep ones in 
frons of emargination of eye; hairs in large punctures inconspicuous; antenna 
clothed with many short and a few long hairs, especially toward apex; seg- 
ment 5 slightly larger than 3 or 4; labrum smooth, emarginate at apex, with 


660 Tue UNiversiry ScteNcE BULLETIN 


I6 
ts} 
I7 
0.5 mm. 14 C5) 
SCALE, FIGS. 12-17 
19 
0.) mm. 


SCALE, FIGS. 18-19 


Fics. 12-19. Exema mormona n. sp.; 12—fcmale tenth and eleventh abdominal segments, 13— 
female ninth abdominal segment, 14—female eighth abdominal segment, 15—an- 
tenna, 16—lateral aspect of male aedeagus, 17—dorsal aspect of male aedeagus, 
18—claw, 19—male ejaculatory guide. 


transverse row of ventrally projecting long hairs. Pronotal gibbosity with 
rounded tubercles, cephalic and anterior pair missing, anterior carinae present, 
summit-2 present and separated from corresponding tubercle by a very broad 
U-shaped space; marginal tubercle small but prominent; entire pronotum 
with large elliptical punctures, largest posteriorly, giving a highly rugose ef- 


A Revision OF THE GENUS Exema or America, Nortu oF Mexico 661 


fect on posterior slope; areas between tubercles minutely granulate. Scutellum 
finely punctate with lateral wings reflexed and almost acute. Elytra coarsely 
punctate, punctures large and deep, each with a short curved hair; smaller 
punctures on tubercles; umbone large and prominent; large tubercles: ante- 
rior, sutural-1, 2, 3, and 3a; small tubercles: scutellar, sutural-la, 2a, 2b, um- 
bonal-a, umbonal-b, marginal and marginal-b; other tubercles very small or 
lacking; faint carina developed on mesal side between sutural-1 and 2. Legs 
coarsely punctate, tarsal claws simple; apical spine on front and middle tibiae. 
Prosternum coarsely punctate, with transverse groove on cephalic end; pro- 


sternal process long and narrow. Metasternal punctures large and deep, espe- 
cially laterally. First abdominal sternum with a longitudinal carina between 


coxal cavities, carina dividing near midlength into a faint Y-shape and termi- 
nating in 2 ventrally projecting spines on posterior margin of segment; 2nd 


segment with similar but smaller spines; punctures of Ist sternum large, 


round, shallow medially to deep on lateral margins; midventral area of 5th 
sternum flat to slightly concave, covered with long recurved hairs; lateral 
punctures large and oblong, nearly rugose. Pygidium centrally convex, faintly 
tricarinate, lateral carinae more distinct than medial, diverging ventrally. 
Aedeagus broadly truncate at tip which is narrower than the rest of structure; 
tegmen slim on the basal half. 

Female: 2.8 mm long, 1.8 mm wide; similar to male except facial yellow 
reduced, longitudinal bars next to eyes broken, and small central spot only 
other yellow on head; pronotum with several irregular yellow spots on ante- 
rior slope. Punctation, sculpturing, and vestiture similar to those of male ex- 
cept front and middle tibial spines absent, spines on Ist and 2nd abdominal 
sterna absent, and fewer hairs on 5th sternum; a few long straight hairs situ- 
ated on periphery of a slightly concave fovea bearing a few very short hairs. 
Pygidium distinctly tricarinate with lateral carinae larger than median one. 

Types: Male holotype, Hobble Creek Canyon, Uinta National Forest, 
(Utah Co.), Utah, 16 Aug. 1961, on Gutierrezia sarothrae, S. L. Wood and 
J. B. Karren (UK). Same data on female allotype, 15 4 and 15 @ paratypes 
WK JBK, CDA, BYU). 

Additional paratypes: UTAH 6¢ and 39, Diamond Fork Canyon, Utah Co., Aug. 1958, T. 
B. Moore (JBK, SCU); 3 ¢ and 3 2, Palmyra Forest Camp, Utah Co., 2 Aug. 1958, on Artemi- 
sia tridentata, T. B. Moore (JBK); 1 @ and 2 9, Honeyville, Utah Co., 18 May 1947, G. F. 
Knowlton (OSU); 1 6 and 1 @, Utah Co., 2; Sept. 1956, G. F. Knowlton (OSU); 1 2, Far West 
(Weber Co.), C. J. D. Brown (BYU); 2 6 and 1 @ labeled “Ut.” (MSU). ARIZONA: 1 9, 
Grand Canyon, South Rim, about 7,000 ft. (Coconino Co.), D. Rockefeller (AMNH); 1 92, Oak 


Creek Canyon, Cochise Co., 2 Oct. 1955, Truxal and Freeman (LACM). COLORADO: 1 4, 
Walsenburg (Huerfano Co.), 14 June 1919 (AMNH); 1 4, labeled “‘Colo.”” (UM); 1 4, Gree- 


ley (Weld Co.), 1933, Wickham Collection (USNM); 1 92, Colorado Springs, 6000-7000 ft. (El 


Paso Co.), 15 June 1930, H. F. Wickham (USNM); 1 @, labeled ‘“‘Colo.”, Wickham Collection 
(USNM). KANSAS: 2 @, labeled ‘“‘Kans.”’, Williston, R. M. Moore collection (NYSM). MIN- 
NESOTA: 1 @, Traverse Co., O. W. Oestland collection (MSC). MISSISSIPPI: 1 ¢, Wiggins 
(Stone Co.), 15 Aug. 1936, H. H. Harris (ISU). MONTANA: 1 @, 15 mi. E. Miles City, Custer 


Co., 20 June 1956, R. C. Froeschner (MSU): 1 2, Glendive, Dawson Co., 21 June 1956, R. C. 


Froeschner (MSU); 4 @, labeled “Mon.”, Horn collection (ANSP); 1 @ and 1 Q, labeled 


662 Tue Universiry ScrENcE BULLETIN 


“Mon.”’, Holland collection (CM). NEBRASKA: 1 4, Glen, (Sioux Co.), Aug. 1903 (AMNH); 
1 ¢@, Pine Ridge (Dawes Co.), July (UN). NEW MEXICO: 4 64, Santa Fe, (Santa Fe Co.), 14 
July 1934, E. L. Bell (AMNH); 1 @, Vaughn (Guadalupe Co.), 5 June 1933, R. H. Beamer 
(UK); 1g and 2 9, Maxwell (Colfax Co.), 27 June 1916 (RU); 1 ¢, Upton (Roosevelt Co.), 
19 Aug. 1949, on Gutierrezia sp., J. H. Russell (USNM). TEXAS: 1 9, Frio Co., 20 May 1948, 
D. J. & J. M. Knull (OSU); 1 ¢, Marathon, (Brewster Co.), 1-2 July 1916 (AMNH); 1 6 and 
1 2, Flatonia (Fayette Co.), July 1903, J. W. Green (CAS); 7 @ and 7 9, 10 mi. N. Pyote 
(Ward Co.), 8 July 1945, with larvae on Gutierrezia, J. H. Russell (USNM, KU, JBK). 


Discussion: This species varies little in sculpturing, but in coloration no 
two individuals are the same. The yellow on the head my be reduced to sey- 
eral spots in the females (Fig. 26). Males, however, always show more yellow 
than females (Fig. 20), and the pattern is more constant than in canadensis, 
Males always have more yellow on the pronotum than females in which the 


; 


f 


yellow may be reduced to faint spots. The yellow is quite irregular, and one — 


side of an individual may have a different pattern from the other. The yellow — 


tubercle may be reduced to just a yellow tip but is usually bright and distinct. 


t 


The body oils cause some of this yellow color to become dark in older speci-— 


mens. In one female from Texas the umbonal-a and marginal tubercle are also 
yellow. 

The shape of the prosternum is flat to concave and the process narrow to 
wide. The pygidium also varies in punctation, shape, and size of the carinae. 

Specimens of mormona show certain affinities with byers1, neglecta, and 
canadensis. The males of canadensis and mormona have similar yellow facial 
patterns, but the latter seldom has yellow on the pronotum or yellow tubercle 
3a. In a few specimens, tubercle 3a is very faintly yellow, and in others there 
are a few yellow spots on the pronotum, but the two characters never occur 
together in canadensis. Specimens of mormona can easily be distinguished 
from neglecta by the elongated punctures and less yellow on the frontal slope 
of the pronotum. The frons in the emargination of the eye is always yellow in 
neglecta and black in mormona. The aedeagus closely resembles that of neg- 
lecta. One might think that mormona is a hybrid between neglecta and 
canadensis, except for its distribution and entirely different host plants. A 
similar situation exists for byersi, discussed later. 

Plant records: Gutierrezia sarothrae and Artemisia tridentata. 

Parasites: Catolaccus aeneoviridis and Tetrastichus chlamytus. 


Exema canadensis Pierce 


Exema dispar canadensis Pierce, 1940:10-12; Proctor, 1946:187; Blackwelder and Blackwelder, 
1948:42 (listed). 

Exema canadensis Pierce; Brown, 1943:124; Blackwelder and Blackwelder, 1948-42 (listed); Dil- 
lon and Dillon, 1961:670; Wilcox, 1954:394-395. 
Diagnosis: Specimens of canadensis differ from neglecta in having long, 

slender punctures on the pronotum, giving it the most highly rugose appear- 


ance in the genus. The recurved hairs on the foveal area of the male are larger 


[ 


A ReEvIsION OF THE GENUS Exema or America, NortuH oF Mexico 663 


Pies. 20-25. Variation in yellow pattern of face of male Exema canadensis. 
) 


Fics. 26-31. Variation in yellow pattern of face of female Exema canadensis. 


664 Tue University ScIENCE BULLETIN 


and more numerous than in byersi. This species is commonly found east of the 
Great Plains from Florida into southern Canada (Fig. 60) on Solidago. 
Description: Male: 2.4 mm long; 1.6 mm wide. Shiny black with yellow 
or yellowish black markings. Antennal scape yellow, flagellum becoming dark 
brown at apex; mouthparts dark brown; face with longitudinal yellow bar 
next to each eye, transversely connected to central yellow spot which merges 
with triangular frontal spot covering carinal area. Frontal slope of pronotum 
black. All tibiae with subbasal and subapical yellow rings, anterior femur 
with basal yellow ring; tarsomeres dark brown. Entire cuticle minutely gran- 
ulate, punctate. Head coarsely punctate, deeper in frons of emargination of 
eye; hairs in large punctures inconspicuous; antenna clothed with many long 
and a few short hairs; segment 5 larger than 3 or 4; labrum smooth with 
transverse row of ventrally projecting long hairs. Pronotal gibbosity rounded, 
tubercles and carinae rounded; cephalic and anterior missing, anterior carinae, 
summit-2, marginal, and lateral tubercles present; entire pronotum with large 
elliptical punctures, largest posteriorly, giving a highly rugose effect on pos- 
terior slope. Scutellum feebly carinate, with lateral wings almost acute. Elytra 


coarsely punctate, each puncture with a short curved hair; umbone large and | 


prominent; large tubercles: anterior, sutural-1, la, 2, 3, 3a, marginal, marginal- 
b, umbonal-b; small tubercles: scutellar, sutural-2a, 2b, 3b, and apical. Legs 
with shallow setigerous punctures; tarsal claws simple; apical spine on front 
and middle tibiae. Prosternum concave, coarsely punctate, with slight trans- 
verse groove on cephalic end; prosternal process long, narrow, curved dorsally. 
Metasternal punctures large and deep, especially laterally. First abdominal 
sternum with a longitudinal carina between coxae, carina dividing near mid- 
length into a faint Y-shape and terminating into 2 ventrally projecting spines 
on posterior margin of segment; 2nd segment with similar but smaller spines; 
punctures of Ist sternum large and round, shallow medially to deep on lateral 
margins; lateral area of sternum feebly carinate; midventral area of 5th ster- 
num flat to slightly concave, covered with long recurved hairs; lateral pune- 
tures large and oblong. Pygidium slightly convex, median carina distinct, 
lateral carinae obsolete. Aedeagus narrowly flanged at tip, tegmen expanded, 
dorsal bend of aedeagus rugose. 

Female: 2.7 mm long; 1.8 mm wide. Similar to male except for facial yel- 
low reduced to spots, 4 next to eyes, 1 central, 2 on carinal area; never yellow 
on anterior slope of pronotum. Punctation, sculpturing, and vestiture similar 
to those of male except front and middle tibial spines absent, spines on sterna 
absent, and fewer hairs on a deeper fovea of 5th sternum. Pygidium distinctly 
tricarinate with lateral carinae larger than median one. 

Types: The male holotype, female allotype, 3 male and 2 female paratypes, 
from Montreal, Quebec (LACM) comprise the type series. 


A Revision oF THE GENUs Exema or America, Nortu or Mexico 665 


I have examined the holotype, allotype, male paratypes 3 and 4, and female 
paratypes 6 (CAS) and 7. All agree quite well with the description. 

Discussion: This the most abundant and wide-spread species of Exema. It 
appears to feed most frequently on Solidago, but occasionally it is found on 
other composites. The wide distribution, attributable in part to the distribu- 
tion of Solidago, may contribute to the variability of the species. I have several 
specimens that have the elytral tubercle 3a faintly yellow. Some have a yellow 
spot or two on the pronotum. Usually the femora are black with subapical 
and subbasal rings of yellow. Specimens from Pennsylvania, Ohio, and New 
York, tend to be darker than others in the species. The tibiae are darker; in 
some they are even black. The pronotal tubercles and carinae are broadly 
rounded, but larger and more numerous in specimens from the southern 
states, especially Florida. Some of these southern forms have no yellow mark- 
ings on the face and the rugose nature of the aedeagus is also very faint, but 
there are too many intermediates to consider these forms a separate species. 
There is a general pattern of yellow on the face, which may be several com- 
binations of spots and bars or the complete pattern with the broad Y-shape 
(Figs. 20-31). Very few of the extreme individuals are found, but they do 
exist and cause some difficulty in distinguishing males from females by color 
alone. In most cases the two ventral spots are lacking in the female pattern, 
but in a few these spots are faint to distinct. The yellow carina always seems 
to be present in males, but the other yellow areas may be lacking in darker 
specimens. The prosternum is concave to convex. The carinae on the pygi- 
dium vary from highly tricarinate to very smooth, males usually being 
smoother than females. 

Plant records: Solidago altissima, S. neglecta, Corylus, Sambucus cana- 
densis, chokecherry blossoms, elm, ragweed and ironweed?, Salix Spy oinela, 
strawberry, Cornus sp., Haplopappus phydocephalus, Erigeron quercifolius, 
and blackberry foliage. 

Parasites: Spilochaleis delumbis and Perilampus fulvicornis. 

Specimens examined: (568 males and 784 females). One female specimen from Blatchley’s 
collection, among those indicated as neglecta, is labeled as a “lectoparatype” of that species. 
ALABAMA: Houston Co., 18 June; Limestone Co., 27-28 June; Macon Co., 19 July; Mobile Co., 
23 Apr.-21 Sept. ARKANSAS: Clay Co., June-6 Nov. (?); Garland Co., no date. CONNECTI- 
CUT: Fairfield Co., 24 June-2 July; Litchfield Co., 21 May-28 Aug.; New Haven Co., 25 May- 
7 June; Tolland Co., 6 June-2 Oct. DISTRICT OF COLUMBIA: Washington, 23 June. FLOR- 
IDA: Alachua Co., 30 Mar.-21 Apr.; Charlotte Co., 25 Apr.; Citrus Co., 23 May; Dade Co., 14 
)Apr.-11 June; Dixie Co., 20 Apr.; Duval Co., 9 May; Gadsden Co., 1 May-12 July; Hendry Co., 
p19 Apr.-26 Feb.; Highlands Co., 6 Mar.-10 May; Lafayette Co., 16 May; Lake Co., 21 Oct.: Lee 
/Co., 31 Mar.; Levy Co., 13 Apr.-31 July; Pinellas Co., 2 Apr.-26 Aug.; Polk Co., 6 May-13 Aug.; 
Putnam Co., 24 Apr.; Seminole Co., 3 May; Volusia Co., Apr.; ‘“Gunntown,” no date. GEORGIA: 
‘State record, no date. ILLINOIS: Cook Co., 18 May-5 Aug.; Jackson Co., 3 June; Kankakee Co., 
i4 July; Lake Co., 19 Aug.; LaSalle Co., 15 May-29 July; Will Co., 13 June. INDIANA: DeKalb 
Co., 13 June; Dubois Co., 14 May; Elkart Co., no date; Knox Co., 18 Sept.; Kosciusko Co., 21 
May-24 June; Lake Co., 11 May-22 Sept.; Lawrence Co., 21 May-9 June; Orange Co., 28 May- 
9 June; Porter Co., 30 May. IOWA: Dickinson Co., 22 Aug.; Mahaska Co., 6 Aug.; Monona Co., 


6 June; KANSAS: Anderson Co., 1 July; Crawford Co., no date: Douglas Co., 4 May-11 July; 
Jefferson Co., 27 May-8 July; Labette Co., no date; Miami Co., no date: Montgomery Co., 23 


666 


Tue University ScrENCE BULLETIN 


(GEES 


36 


9 


~~ 


en od ste eocen ees oer it 


em eos Maen mangeaerea es 


SCALE, FIGS. 38-39 


34 


0.5 mm. 


SCALE, FIGS. 35-37 


o5mm 


SCALE, FIGS. 32-34 


Fics. 32-39. Exema canadensis; 32—female tenth and eleventh abdominal segments, 33—female 
ninth abdominal segment, 34—female eighth abdominal segment, 35—antenna, 
36—lateral aspect of male aedeagus, 37—dorsal aspect of male aedeagus, 38—claw, 


= ; : 
39—male ejaculatory guide. 


A RevIsION oF THE GENUs Exema oF America, Nortu oF Mexico 667 


May; Riley Co., 21 June-July; Sedgwick Co., 16 Aug. KENTUCKY: Bullitt County, 9 Mar.; Jef- 
ferson Co., 10 Aug.; LOUISIANA: (Parish=county of other states) Natchitoches Par., 16 Aug.; 
St. Charles Par., 14 Apr.; St. James Par., 7 June; St. Landry Par., June; St. Tammany Par., 25 
May-June; “Doeville,’ 6 May. MAINE: Cumberland Co., 11-15 Aug.; Hancock Co., 26 Aug.; 
Kennebec Co., 12 July; Lincoln Co., 18 June-27 Aug.; Oxford Co., 10 July; Washington Co., 
June. MARYLAND: Baltimore Co., 17 May-30 July; Charles Co., 25 May; Montgomery Co., 30 
May; Prince Georges Co., 22 May-9 July; “Glover,” no date. MASSACHUSETTS: Hampden 
Co., 28 June-24 Aug.; Hampshire Co., 9 June; Middlesex Co., 25 May-10 Oct.; Norfolk Co., 2 
Sept.; Suffolk Co., 17 May; Worcester Co., 19 June-5 Sept. MICHIGAN: Allegan Co., 1 June; 
Barry Co., 10 July; Bay Co., 1 June; Cheboygan Co., 18 June-18 July; Clinton Co., 31 Aug.; 
Ingham Co., 4-27 Sept.; Kent Co., 5 Aug.; Midland Co., 29 May-12 Aug.; Washtenaw Co., 1 
Aug.; Wayne Co., no date. MINNESOTA: Anoka Co., 14-27 Aug.; Big Stone Co., no date; 
Clearwater Co., 25 July; Hennepin Co., 28 June-24 Aug.; Ottertail Co., no date; Ramsey Co., no 
date; Red Lake Co., 20 June; Traverse Co., no date: Washington Co., 7 Aug. or 8 July. MISSIS- 
SIPPI: George Co., 5-22 June; Green Co., 23 May-13 Sept.; Grenada Co., 11 June; Jackson Co., 
12 July; Montgomery Co., 7 May; Perry Co., 23 Mar.-2 Oct. MISSOURI: Boone Co., 4-16 May; 
Howard Co., 12 May; Miller Co., 16 May; Morgan Co., 13 May; Ozark Co., 9 June; Pettis Co., 
10 June; Pike Co., 20 Apr.-28 June; St. Louis Co., no date. MONTANA: “Assinbne,” 29 Aug. 
NEBRASKA: Antelope Co., 21-22 June; Douglas Co., 20 Sept.; Holt Co., 9 Sept.; Lancaster Co., 
6 July-2 Oct. NEW BRUNSWICK: French Lake, 2 June. NEW HAMPSHIRE: Strafford Co., no 
date. NEW JERSEY: Atlantic Co., 15 June; Bergen Co., 20 June-1 Aug.; Camden Co., no date: 
Cape May Co., June; Cumberland Co., 30 Sept.; Essex Co., 4 June-27 Sept.; Gloucester Co., 31 
May; Middlesex Co., 2 June-2 Aug.; Hudson Co., no date; Passaic Co., 26 June-27 July; Somer- 
set Co., no date; Sussex Co., 24 May; Union Co., no date. NEW YORK: Albany Co., 26 May-5 
July; Bronx Co., 30 May-3 July; Cortland Co., 18 May; Erie Co., 28 June-10 Aug.; Essex Co., 
11 May-2 June; Herkimer Co., no date; Livingston Co., 29 Aug.; Monroe Co., 30 May-9 Sept.; 
Orange Co., 26-30 May; Orleans Co., 10 June; Oswego Co., 1 Aug.-15 Oct.; Putnam Co., 29 
May-June; Queens Co., 15 June-July; Richmond Co., 20 May-1 Oct.; Rockland Co., 30 May-28 
June; Saint Lawrence Co., 12 June-18 Aug.: Suffolk Co., 25 Aug.; Tompkins Co., 8 May-25 
Sept.; Westchester Co., 5 May-25 Aug. NORTH CAROLINA: Guilford Co., 21 June; Moore Co., 
Aug.; Wake Co., no date. NORTH DAKOTA: Bottineau Co., 30 July; Traill Co., 7 Aug. OHIO: 
Champaign Co., 22 May-6 June; Delaware Co., 2 May; Franklin Co., 5 June-5 Sept.; Hocking 
Co., 2 July-9 Aug.; Jackson Co., 3-6 July; Lake Co., 3 June; Lorain Co., 13 June-4 Sept.; Ross 
Co., 22 May; Scioto Co., 10 June; Summit Co., 2 July-11 Aug.; Williams Co., no date. OKLA- 
HOMA,; Jefferson Co., 18 May. ONTARIO: Arnprior, July 23 Aug.; Blackburn, 31 July; Delhi, 3 
June; Fisher Glen, 12 June; Leamington, 6 July; Ojibway, 29 Aug.; Rich Edward Co., 24 May- 
6 Oct.; Simcoe, 13 June; Toronto, 16 May; Turkey Point, 1-8 June; Walsingham, 1 June. 
PENNSYLVANIA: Bucks Co., 30 May; Dauphin Co., 17 June-27 Aug.; Delaware Co., 28 May- 
16 June; Juniata Co., 14 Sept.; Monroe Co., 21 Sept.; Northampton Co., 31 May-18 June; Phila- 
delphia Co., 25 May-21 Aug.; Pike Co., 20 May-15 Aug. QUEBEC: Hemmingford, 7 Sept.; 
Longueuil, 18 Aug.; Montreal, 10 May-20 Oct.; Outremont, 25 Aug.; Rigaud, 12 Aug. SOUTH 
CAROLINA: Jasper Co., 14 June. SOUTH DAKOTA: Auroro Co., 23 Aug.; Fall River Co., 11 
Sept.; Marshal Co., 7 Sept. TENNESSEE: Hamilton Co., 6 May. TEXAS: Brazos Co., 18 May- 
27 Sept.; Cameron Co., 30 July; Dallas Co., 27 Apr.-3 Nov.; Ellis Co., 21 Sept.; Harris Co., 3 
Aug.; Navarro Co., 7 June; Tarrant Co., 2 May. VERMONT: Windham Co., Spring. VIRGINIA: 
Alexandria, 2 June-July; Arlington Co., 30 May-14 July; Fairfax Co., 6 June-15 July; Mont- 
gomery Co., 7 May-30 June: Nottoway Co., 17 May; Prince George Co., 27 July-11 Aug.; Prin- 
cess Anne Co., 20 July. WEST VIRGINIA: Greenbrier Co.. July; Wood Co., 11 June. WISCON- 
SIN: Columbia Co., 31 Aug.; Milwaukee Co., 23 June-21 Aug.; Washington Co., 16 Aug.; Wau- 
paca Co., 7 Aug. 


Exema neglecta Blatchley 


Exema neglecta Blatchley, 1920:69; 1924:50 (biology note); Leng and Mutchler, 1927:44 
(listed) ; Blatchley, 1930:52 (lectotype designation) ; Pierce, 1940:19 (keyed); Brown, 1943:124 
(note). 


__ Diagnosis: This species is the most highly punctate of those with simple 
claws. There are rugose lines on the posterolateral slopes of the pronotum 
similar to the other species in this group, but the punctures are larger and 


more distinct. The males are easy to recognize by the large amount of yellow 


668 Tue Universiry ScrENcCE BULLETIN 


on the face and pronotum. This yellow is greatly reduced in the females. The 
flange on the tip of the aedeagus is sharp and distinct. The median dorsal 
plate is narrow. This species is found in the extreme southeastern part of the 
United States (Fig. 61) on Baccharis and other composites. 

Description: Male: 2.7 mm long; 1.8 mm wide. Black except yellow on en- 
tire face, anterior half of pronotum, spot on anterior coxa, entire anterior 
femur, subapical and subbasal rings on middle tibia and middle femur; tar- 
someres brownish yellow. Anterior pronotal yellow area spotted irregularly 
with black and black punctures. Entire cuticle minutely granulate, punctate. 
Head sparsely punctate; antenna clothed with many short and a few long 
hairs; segment 5 larger than 3 or 4, 6-11 subequal; labrum smooth and faintly 
emarginate with a few ventrally projecting hairs. Pronotal gibbosity with 
rounded tubercles and carinae; marginal present, lateral divided and carinate 
with summit-2; posterior punctures oblong, anterior punctures round. Scutel- 
lum convex, with rounded lateral wings. Elytra coarsely punctate, punctures 
large and deep, each with a short curved hair; smaller punctures on tubercles; 
umbone large and prominent; large tubercles: anterior, sutural-l, la, 2, 2a, 
2b, 3, 3a, marginal, and marginal-b; small tubercles: scutellar, umbonal-a, 
umbonal-b, sutural-lb, 3b, marginal-a, and apical. Legs shallowly punctate; 
tarsal claws simple; apical spine on front and middle tibiae. Proternum 
coarsely punctate; prosternal process narrow and long, curved dorsally. 
Metasternal punctures large and deep, especially laterally. First abdominal | 
sternum with a longitudinal carina between coxae, carina dividing near mid-| 
length into a faint Y-shape terminating in 2 ventrally projecting spines on 
posterior margin of segment; 2nd segment with similar but smaller spines; 
sides of Ist sternum irregularly carinate with large, round punctures, shallow 
medially to deep on lateral areas; midventral area of 5th segment flat with — 
short recurved hairs, lateral punctures large and round. Pygidium distinctly — 
tricarinate, lateral carinae curved and rounded. Aedeagus similar to canaden- 


sis except with a wider tip. 

Female: 2.8 mm long; 1.8 mm wide. Similar to male except yellow usually 
absent on pronotum and reduced on face, yellow frons of emargination of eye 
connected to yellow spot near eye, two spots near vertex, and a central spot 
between eyes; legs with less yellow than male. Punctation, sculpturing, and 
vestiture similar to that of male except front and middle tibial spines absent. — 
spines on Ist and 2nd sterna absent, and fewer long, straight hairs on 5th 
sternum. 

Types: Male lectotype, Dunedin, Pinellas Co., Florida, 26 Jan. 1913, W. S. 
Blatchley (PU). 1 2 lectoparatype on the same pin as the lectotype. 


EE ———————— 


Other specimens labeled lectoparatype examined from the Blatchley collection (PU): FLOR- 
IDA: 1 @, Sanford (Seminole Co.), 4 Apr. 1913; 1 ¢, Ormond (Volusia Co.), 22 Mar. 1913: 
1 $, Dunedin (Pinellas Co.), 22 Mar. 1913; 1 2, 13 Jan. 1917; 1 6, Ft. Myers (Lee Co.), 31 
May 1913; 1 9, Dunedin (Pinellas Co.), 5 Apr. 1913; 1 9, Kissimee (Osceola Co.), 16 Feb. 


A Revision oF THE GeNus Exema or America, Nortu or Mexico 669 


1913. Other specimens labeled lectoparatypes but collected after the original description was pub- 
lished (PU): 1 6, Chokoloskee (Collier Co.), 17 Mar. 1921; 1 4, Casambas (Collier Co.), 8 
Mar. 1921; 1 ¢, Royal Palm Park (Dade Co.), 21 Mar. 1924: 1 9, (canadensis), Labelle 
(Hendry Co.), 26 Feb. 1918. All were collected by W. S. Blatchley. I was able to examine all the 
types listed above. 

Discussion: The yellow areas on the pronotum have a striking, blistered 
appearance in some specimens from the South. The central Floridian forms 
have a much smoother pronotum, especially in the yellow areas of males, than 
do those from any other area. Further north this yellow in the males covers 
less of the pronotum. More elliptical punctures are found on the pronotum of 
northern forms, but deep round punctures still persist to distinguish neglecta 
from canadensis, byersi, and elliptica. The color of the legs is highly variable, 
from nearly black to completely yellow. The femora and tibiae usually have 
subbasal and subapical yellow rings so that when the legs are retracted four 
longitudinal yellow stripes are produced on the venter. There is usually a yel- 
low spot on the procoxa. 

The entire face is yellow except for a brownish area around the bases of the 
antennae and a black spot on the vertex. There may also be a few black or 
brownish punctures on the front near the vertex. 

The male sternal spines are very small and may even be absent on the 
second and third sterna. The hairs on the foveal area of the male are out- 
wardly recurved and more scattered than in canadensis. In the female the 
hairs around the fovea are straight and more numerous than in canadensis. 

The species in this group, 1.e., canadensis, mormona, neglecta, byersi, and 
elliptica, are quite similar, especially in the genitalia, and show very close re- 
lationships. For this reason I have omitted separate drawings for neglecta and 
byersi. Including them would add little of taxonomic value. The tip of the 
aedeagus in neglecta is wider than that figured for canadensis and lacks the 
rugose lines on the bend. The tegmen, ejaculatory guide, and female genitalia 
are very similar in the two species. The fifth antennal segment in neglecta is 
smaller than the sixth and similar to the fourth, while this same segment is 
noticeably larger in canadensis but not as large as the sixth segment. 

The yellow on the pronotum of males usually makes them easy to distin- 
guish from females. A number of females have a few yellow spots on the 
pronotum, and one from Carolina Beach, North Carolina, looks like a male 
in coloration. The most reliable characters to distinguish the sexes are the 
spines on the tibiae and first abdominal sternum. 

In the original description, Blatchley stated that neglecta was “common 
throughout Florida on huckleberry and other low shrubs throughout the win- 
ter months.” It is not restricted to this plant as can be seen from records. 
Brown collected and reared a good series on Baccharis, suggesting that this 


may be the preferred host. Several other collectors have taken this species on 
Baccharis. 


670 Tue University ScIENCE BULLETIN 


Plant records: Baccharis halimifolia, Chondrophora virgata, Eupatorium, 
Arundinaria (cranebrake), Solidago sp., and Strophostyles helvola. 

Specimens examined: (504 males and 376 females). ALABAMA: Baldwin Co., 10 July; 
FLORIDA: Alachua Co., 8 May-14 Nov.; Baker Co., 3 May-5 Aug.; Bay Co., 7 July; Brevard 
Co., 11 July-8 Nov.; Charlotte Co., 9 Apr.-13 Nov.; Citrus Co., 23 May; Clay Co., 7 Aug.; 
Collier Co., 8 Mar.-6 June; Dade Co., 16 Jan.-22 July; DeSoto Co., 9 Apr.; Duval Co., 9 May-1l6 
Nov.: Escambia Co., 13 Mar.; Hardee Co., 12 Apr.-15 July; Hendry Co., 26 Feb.; Hernando Co., 
13 Mar.; Highlands Co., 10 Apr.-10 May; Hillsborough Co., 13 Jan.-23 June; Indian River Co., 
27 July; Lake Co., 1 Mar.-1 Sept.; Lee Co., 1 Apr.-13 Nov.; Levy Co., 30 June-18 Oct.; Manatee 
Co., 23 Mar.-3 Sept.; Monroe Co., 28 Mar.; Nassau Co., 31 Aug.; Orange Co., 12 Jan.-19 Nov.; 
Osceola Co., 4 July; Pasco Co., 15 Apr.-7 July; Pinellas Co., Jan.-Dec.; Polk Co., 26 Mar.-10 Nov.; 
Putnam Co., 20-22 Apr., 24 Dec.; St. Johns Co., no date; St. Lucie Co., 27 July; Sarasota Co., 28 
Mar.-10 Dec.; Seminole Co., 1 Mar.-30 Dec.; Taylor Co., 20 Apr.; Volusia Co., 14-20 Apr.; 
Wakulla Co., 6 July; “Gunntown,” no date. GEORGIA: Charlton Co., 20 June-7 July; Chatham 
Co., 18 June-22 July; Glynn Co., 12-14 July; Jefferson Co., 3 May; Richmond Co., 5 Oct.; Tift 
Co., 1-10 Aug. NORTH CAROLINA: Beaufort Co., 16 June; Dare Co., 21 Aug.; Johnston Co., 
July; Moore Co., 12 May; New Hanover Co., 8-26 July; Onslow Co., 28 July; Scotland Co., 17 
Aug.; Wake Co., 7 June. SOUTH CAROLINA: Beaufort Co., 14 Apr.; Charleston Co., 14 June- 
27 July; Greenwood Co., 1-25 June; Jasper Co., 28 Oct.; Lee Co., 4 May. TEXAS: Runnels Co., 
6 June. 


Exema byersi n. sp. 


Diagnosis: Exema byersi can easily be distinguished by the yellow coloring 
on the face, which always includes the frons in the emargination of the eye, 
and many oval to elliptical punctures on the pronotum, especially on the an- 
terior half. The male aedeagus differs from that of canadensis in having a 
narrower flange and being smooth at the bend. This species is found from 
New York to Kansas and south to Texas (Fig. 61) on Gutierrezia dracuncu- 
loides and several other genera of plants. It is more common in Kansas and 
absent in the southeastern states. 

Description: Male: 2.6 mm long; 1.7 mm wide. Black with yellow or yel- 
lowish brown markings; antenna and labrum yellowish brown, mouthparts 
darker; head entirely yellow, except black genae, brownish clypeus and anten- 
nal sockets, and black punctures on vertex and front; frontal slope of prono- 
tum yellow; wide yellow band extends from summit to antecostal suture 
laterally along suture as a narrow bar to well below the middle of pronotum 
then turns posterodorsad for a short distance; front femur and all tibiae with 
subbasal and subapical yellow rings, middle femur with a wide subbasal ring, 
hind femur black, tarsomeres yellowish brown, claws black. Body surface 
finely granulate throughout; head sparsely punctate; labrum smooth, apex 
emarginate with subapical row of transverse hairs. Antennal segment 5 
slightly larger than 3 or 4, but smaller than 6, covered with a few long and 
many short hairs. Pronotum with summit-1 tubercle developed into rounded 
carina running irregularly to small anterior pronotal tubercle; lateral tubercle 
divided into a large dorsal and a small ventral, rounded tubercle; marginal 
obscure and rounded; posterolateral slope highly rugose; anterolateral pro- 
hotum covered with long elliptical punctures; punctures on yellow area oval 
or round, Scutellum with a central carina on anterior half; posterior scutellar 


A Revision oF THE GENUs Exema or America, Nortu or Mexico 671 


wings reflexed, almost acute. Umbone large and prominent; elytral punctures 
larger than those on pronotum; large tubercles: anterior, sutural-1, la, 2, 3, 3a, 
and marginal-b; small tubercles: scutellar, sutural-1b, 2a, 2b, 3b, umbonal-a, 
b, and marginal; marginal-a and apical very small; discal depression shallow; 
tarsal claws simple; legs coarsely but shallowly punctate. Prosternum coarsely 
punctate, slightly concave, with transverse anterior groove; prosternal process 
narrow. Metasternum coarsely punctate, punctures becoming larger laterally. 
First abdominal sternum with median, inverted Y-shaped carina dividing 
coxal cavities, extending to 2 small spines on posterior edge of segment, simi- 
lar pair of very short spines on 2nd sternum; Ist sternum coarsely punctate, 
laterally tuberculate; punctures of 5th sternum slightly elongate, central spot 
slightly concave, smooth, covered with short recurved hairs. Pygidium convex, 
slightly tricarinate, central carina most apparent ventrally; surface coarsely 
punctate. Aedeagus flanged at tip; basal stem of tegmen as in canadensis 
(Fig. 36). 

Female: Facial yellow reduced to 2 broken bars next to eyes, a large central 
spot, and a small spot on the frons in emargination of each eye; no yellow on 
pronotum; legs darker than those of male. Front and middle tibiae without 
apical spine; no spines on Ist and 2nd abdominal sterna, 5th sternum with 
central, slightly depressed area surrounded by a few long and short hairs; 
pygidium distinctly tricarinate. Genitalia similar to those of canadensis. 

Types: Male holotype, 8 miles north of Lawrence, Jefferson Co., Kansas, 
27 May 1965, taken on Gutierrezia dracunculoides, J. B. Karren (UK). Same 
data on allotype and 2 8 paratypes (UK, JBK). An additional 49 4 and 
59 2 paratypes were collected at the same locality on 7 June, 15 June, and 8 
July 1965 by the author (UK, JBK, BYU, RU, ANSP). A few of these speci- 
mens were reared from eggs or larvae collected in the field. 

Other paratypes, all from Kansas, as follow: 2 6 and 29, 8 miles north of Lawrence, Jeffer- 
son Co., 14 June 1964, R. C. Funk (JBK); 2 9, Mission, (Johnson Co.), 22 Aug. 1963, H. L. 
Willis (JBK). 1 8, Miami Co., 1915, R. H. Beamer; 1 &, Crawford Co., 993 ft., 1915, R. H. 
Beamer; 1 ¢, Douglas Co., 1926, R. H. Beamer; 2 9, Labette Co., 899 ft., R. H. Beamer: 4 4 
and 2°, Garnett (Anderson Co.), 8 Oct. 1926, R. H. Beamer; 2 ¢ and 2 9, Hutchinson (Reno 
Co.), 27 July 1950, taken on Echinacea angustifolia, J. G. Rozen; 2 @ and 1 @, Anderson Co., 
R. H. Beamer; 1 2, Topeka (Shawnee Co.), Aug., Popenoe; 1 9, Haskell, Douglas Co., July 
/1909 (all UK); 1 8, Medora (Reno Co.), 17 Apr. 1932, C. W. Sabrosky (MSU); 1 @, Baldwin 
(Douglas Co.), May, J. C. Bridwell (OSC); 2 ¢ and 1 92, Onaga (Pottawatomie Co:)y EF. 
Crevecoeur (CDA); 1 6 and 1 92, Onaga (Pottawatomie Co.), 29 July 1909, F. Crevecoeur 


(CAS); 3 $ and 1 2, Onaga (Pottawatomie Co.), 1 Sept. 1929, 20 June 1929, and 5 May 1923, 
\F. Crevecoeur (KSU); 1 9, Ashton collection (PU); 4 @ and 3 9, Topeka (Shawnee Co.), 1 


labeled 3 Sept. and 2 labeled Aug., Fraxinus viridis, 2 labeled Popenoe (USNM); 2 @ and 1 9, 
Riley Co., Popenoe (USNM). The type material represents all of the available specimens from 
Kansas, 

Discussion: This species very closely resembles neglecta and elliptica ex- 
cept for the elliptical punctures on the pronotum of the former and the shape 
of the pronotum of the latter. Other differences can be seen, but the above are 
the most apparent. The face of the male is entirely yellow, but there may be 


672 Tue Universiry ScreNcE BULLETIN 


a few scattered black or brown punctures, especially on the vertex. The yellow 
on the pronotum is more reduced than in neglecta, but more extensive than in 
most elliptica. The yellow on the face of the female is reduced approximately 
as shown in Fig. 21, except that the frons is entirely yellow in the emargina- 
tion of the eye, and the two ventral spots are usually missing. This pattern 
varies from the two vertical bars, more common in neglecta, to no yellow ex- 
cept for that on the frons in the emargination of each eye. 

This species looks very much like canadensis in punctation, sculpturing, 
and size. The shape of the pronotum and the pattern of the tubercles on the 
elytra are very similar in the two species. Except for the striking difference in 
coloration, many males of byers? might be mistaken for canadensis. At first, I 
felt that this species was a geographical variation of neglecta, but as more 
specimens were examined and additional collections were made in the field, 
it became apparent that the species was different. At the type locality, speci- 
mens of byersi were collected along side canadensis, the former were almost 
always found on Gutierrezia and the latter on Solidago. 

Plant records: Echinacea angustifolia, Fraxinus viridis, Apias tuberosa, 
and Gutierrezia dracunculoides. 

Specimens examined: (128 males and 143 females). ALABAMA: Mobile Co., Mar.-15 June. 
ARIZONA: Cochise Co., 7 Aug. ARKANSAS: State record, 6 Oct. INDIANA: Porter Co., 22 
July; Pulaski Co., 16 July. KANSAS: Anderson Co., 8 Oct.; Crawford Co., no date; Douglas Co., 
May-July; Harvey Co., 22 Aug.; Labette Co., no date; Jefferson Co., 27 May-8 July; Johnson 
Co., 22 Aug.; Miami Co., no date: Pottawatomie Co., 16 May-29 July; Reno Co., 17 Apr.-27 
July; Riley Co., Sept.; Shawnee Co., Aug.-2 Sept. LOUISIANA: (Parish—county of other states) 
Rapides Par., 14 Oct. MARYLAND: Baltimore Co., 6 June-26 July; Montgomery Co., May; 
Prince Georges Co., 25 May-29 June. MICHIGAN: Allegan Co., 22 July. MISSISSIPPI: George 
Co., 13 June; Jackson Co., 24 June-23 Aug. MISSOURI: State record, no date. NEW JERSEY: 
Cape May Co., June; Camden Co., 27 Aug.; Ocean Co., 31 May-11 Oct. NEW YORK: Nassau 
Co., 30 June; Suffolk Co., 31 May-25 Aug. OKLAHOMA: Muskogee Co., 6 June-19 Sept. 
PENNSYLVANIA: Allegheny Co., no date; Lebanon Co., 28 Aug. TEXAS: Brazos Co., 1 May; 


Colorado Co., 13-14 Apr.; Comal Co., no date; Harris Co., 3 Aug.; San Patricio Co., 15 June. 
VIRGINIA: Alexandria, 2 June; Campbell Co., 2 June; Fairfax Co., 9 Apr.-6 June. 


Exema elliptica n. sp. 


Diagnosis: This species is easily recognized by the high carinae. The male 
aedeagus is distinguished by the slightly flanged, rounded tip (Fig. 45) with 
very little ventral curve (Fig. 44). The tegmen is expanded on the basal half 
and truncate at the base. The female structures are also similar except for the 
highly concave outer edge of the ninth pleurite and tergite (Fig. 41). The fol- 
lowing description is based on 20 males and 20 females from nine localities in 
the southern and eastern states (Fig. 62). 

Description: Male: 2.8 mm long; 1.8 mm wide. Shiny black to slightly 
metallic, except for yellow or yellowish brown markings. Antennae yellow, 
mouthparts light brown, labrum shiny; face yellow except for light brown 
antennal bases, genae and basal clypeus; black spot or patch of black pune- 
tures on vertex. Frontal slope of pronotum with yellow band one-half width 


A ReEvIsIoN oF THE GENUs Exema or America, Nortu oF Mexico 673 


e 
ae 


} O.lmm 
SCALE, FIGS. 46-47 


0.5mm 


SCALE, FIGS. 40-45 


Fics. 40-47. Exema elliptica n. sp.; 40—female tenth and eleventh abdominal segments, 41—fe- 
male ninth abdominal segment, 4+2—female eighth abdominal segment, 43—antenna, 
| 44—lateral aspect of male aedeagus, 45—dorsal aspect of male aedeagus, 46—claw, 
47—male ejaculatory guide. 


674 Tue UNiversiry SCIENCE BULLETIN 


of head including eyes extending from antecostal suture dorsally to anterior 
carinae; slender band extending ventrally half way along suture; yellow spot 
opposite lower edge of eye. No yellow on elytra. All femora and tibiae with 
subbasal and subapical yellow rings, very wide on front and middle femora, 
leaving only a central black spot; tarsomeres yellowish brown. Head minutely 
granulate, sparsely punctate; punctures small and shallow, obscure in emargi- 
nation of eye; hairs in punctures inconspicuous or absent; antennae clothed 
with many short and a few long hairs, especially toward apex; segments 3 and 
4 subequal, segments 5 slightly larger, smaller than 6, 6 smaller than 7, 7-10 
subequal, 11 subtriangular; labrum smooth with ventrally projecting hairs. 
Pronotal gibbosity high, appears flat from lateral aspect, with large, rounded 
tubercles and carinae, cephalic and anterior missing, anterior carina large, 
summit-2 represented by outwardly curved carina, lateral tubercle divided; 
entire pronotum with large elliptical to oval punctures, deeply striopunctate 
on the flanks giving a highly rugose effect; areas between tubercles minutely 
granulate. Scutellum finely punctate, feebly carinate, scutellar wings small 
and blunt. Elytra with large, deep, setigerous punctures, smaller punctures on 
tubercles; areas between tubercles granulate; umbone large and prominent; 
large tubercles: anterior, sutural-1, la, 2, 3 expanded longitudinally, 3a, and 
marginal-b; small tubercles: scutellar (faint), umbonal-a, umbonal-b, sutural- 
2a, 2b, 3b, marginal-a, and apical. Legs with large shallow punctures; Tarsal 
claws simple; apical spine on front and middle tibiae. Prosternum coarsely 
punctate, granulate, concave, sides rounded; prosternal process long and nar- 
row. Metasternal punctures large and shallow. First abdominal sternum with 
a longitudinal carina between coxae, dividing near midlength into a Y-shape; 
2 ventrally projecting spines on posterior margin; 2nd sternum with similar 
but smaller spines; punctures of Ist sternum large and round medially, 
smaller and deeper on roughly sculptured lateral margin; mid-ventral area of 
5th sternum flat to slightly concave, covered with outwardly recurved hairs; 
lateral punctures oblong; all areas granulate between punctures. Pygidium 
convex, tricarinate, lateral carinae converging ventrally; entire surface granu- 
late with deep punctures. Aedeagus feebly flanged and rounded at apex, 
slightly curved ventrally; tegmen expanded and truncate at base. 

Female: 2.9-3.1 mm long; 1.9-2.1 mm wide. Similar to that of male except 
more variable, yellow of face and pronotum reduced, sometimes lacking on 
pronotum; facial yellow represented by spots on frons in emargination of eye 
plus those shown in Fig. 30, separate or irregularly connected. Punctation and 
sculpturing similar to those of male except front and middle tibial spines ab- 
sent; spines on Ist and 2nd abdominal sterna absent and fewer hairs on 5th 
sternum, hairs straight, a few long hairs situated on periphery of a slightly 
concave fovea covered with a very few short hairs. Pygidium distinctly tri- 
carinate, with lateral carinae rounded, much larger than median carina. 


A Revision oF THE GENUs Exema or America, Nortru oF Mexico 675 


Types: Male holotype, Anahuac (Chambers Co.), Texas, 10 Oct. 1918, on 
Iva fructescens, H.S. Barber (USNM). Same date on allotype, 3 ¢ and 3 2 
paratypes (one female collected 30 Oct.). (USNM, UK, JBK). 

Other paratypes as follows: FLORIDA: 2 9, Levy Co., no date (ANSP, MCZ). GEORGIA: 
1 2, Tifton (Tift Co.), no date, Liebeck collection (MCZ). LOUISIANA: 2 6 and 1 9, Alex- 
andria (Rapides Par.), 14 Oct. 1959, R. E. Woodruff (FDA); 6 ¢ and 5 @, Lutcher (St. James 
Par.), 7 June 1944, on Baccharis sp. (USNM, JBK). MARYLAND: 1 6, Plummer’s Island 
(Montgomery Co.), 22 Aug. 1943, R. H. Beamer (UK). SOUTH CAROLINA: 1 2, Adams Run 
(Charleston Co.), 25 June 1948, Cartwright (CDA). TEXAS: 1 @, Dickinson (Galveston Co.), 
20 Apr. 1933 (UCR); 4 6 and 5 9, Houston (Harris Co.), 25 May 1949, reared from Bac- 
charts halimifolia, J. L. Ward (USNM, NYSM, JBK). VIRGINIA: 1 @ and 1 2, Cape Henry 
(Princess Anne Co.), 28 May 1927, H. S. Barber (USNM). 

Discussion: Only 20 males and 20 females were available for study, and all 
were made part of the type series. The specimens from the United States 
National Museum were already labeled as a new species by H. S. Barber, but 
he had not yet proposed a name for it. This group had the largest series with 
host data, so I chose the type from among them. 

The coloration of the males seems to be constant except for a slight reduc- 
tion of yellow on the pronotum in some. The females, on the other hand, are 
quite variable, some having no yellow on the pronotum and others with as 
much as the males. I suppose that with more collecting the extremes will 
prove to be rare and the typical forms more common. 

This species is very close to neglecta, elliptica, and byersi, differing from 
these species in its strikingly elevated pronotum and its distribution. It is prob- 
ably sympatric with dyersi in Maryland and New Jersey in the North, and 
Louisiana, Mississippi and Alabama in the South. It is sympatric with neg- 
lecta in the southern states and has the same host plant in some cases. This is 
a species that may have developed in Florida and is now spreading northward. 
There are Florida forms of canadensis and dispar that vary from the typical 
forms of their respective species in much the same way that ellzptica differs 
from neglecta and byersi, but this divergence has not gone far enough for 
them to be called distinct species. They may be forms in which secondary in- 
tergradation has taken place; at least there are too many intermediate forms 
for them to be called species. 

Plant records: Iva fructescens and Baccharts halimufolta. 


Exema deserti Pierce 


Exema deserti Pierce, 1940:20-21; Blackwelder and Blackwelder, 1948:42 (listed). 

Exema deserti boregensis Pierce, 1940:21; Blackwelder and Blackwelder, 1948:42 (listed); (NEW 
SYNONOMY). 

Exema globensis Pierce, 1940:21-22; Blackwelder and Blackwelder, 1948:42 (listed); (NEW 
SYNONOMY). 

Exema chiricahuana Pierce, 1940:22-24; Blackwelder and Blackwelder, 1948:42 (listed); (NEW 
SYNONOMY). 

Exema parvisaxi Pierce, 1940:24; Blackwelder and Blackwelder, 1948:42 (listed); (NEW 
SYNONOMY). 


676 Tue UNiversiry ScIENCE BULLETIN 


Diagnosis: The pronotum is punctate without any trace of striae or rugose 
lines. The face of the male is always yellow except as noted in the description. 
Most specimens are more yellow than any other species in the genus; some are 
almost completely yellow. Most specimens have small tubercles and carinae. 
This species is found in the desert regions of the Southwest (Fig. 63) on a 
variety of plants. 

Description: Male: 2.6 mm long; 1.6 mm wide. Yellow marked with black; 
underside, elytral punctures, and umbone black; a few punctures on head and 
many on posterior half of pronotum black; antennal socket, condyle of scape, 
and a narrow clypeal band black; scape of antenna yellow, flagellum becom- 
ing brown at apex; mouthparts brown, labrum yellowish brown; face mostly 
yellow including vertex. Prothorax yellow with a few black punctures and 
black spots on the posterior half; prosternum with a small central yellow spot. 
Front femur mostly yellow, hind femur black with subapical yellow ring; all 
tibiae black with wide subbasal and subapical yellow rings; tarsomeres dark 
brown. Pygidium black with two lateral, yellow spots associated with yellow 
spots on the 5th sternum. Entire cuticle minutely granulate, irregularly punc- 
tate. Head with large, deep punctures, deepest in middle of vertex and frons, 
hairs in punctures inconspicuous; antenna clothed with many short and a few 
long hairs, especially toward apex; segment 5 slightly larger than 3 or 4; 
labrum smooth slightly emarginate. Pronotum gibbose with low, rounded 
tubercles, marginal broadly rounded; lateral, summit-1, and 2 feebly carinate; 
anterior Carinae present; entire pronotum covered with deep to shallow, 
rounded punctures. Scutellar wings reflexed and almost acute, distinctly cari- 
nate. Elytra coarsely punctate, shallower punctures on rounded tubercles, 
punctures setigerous; umbone large and prominent; large tubercles: anterior, 
sutural-la, 2, 2a, 3, 3a, marginal, and marginal-b; small tubercles: scutellar, 
sutural-1, 2b, 3b, and umbonal-b; other tubercles lacking. Legs feebly and 
shallowly punctate, tarsal claw feebly toothed; apical spine on front and mid- 
dle tibiae. Prosternum concave and deeply punctate; prosternal punctures 
large and deep, smaller laterally. First abdominal sternum with a longitudinal 
carina between coxal cavities, carina dividing near midlength into a faint Y- 
shape and terminating in 2 large ventrally projecting spines on posterior mar- 
gin of segment; spine feebly developed on 2nd and 3rd segments; large, round 
punctures of Ist sternum setigerous, shallow medially to deep on lateral mar- 
gins; lateral punctures of 5th sternum large, oblong, nearly rugose; midventral 
area flat, covered with long recurved hairs. Pygidium centrally convex, medial 
carina distinct, depressed areas near lateral edge. Aedeagus truncate at tip; 
tegmen expanded at base. 

Female: 2.3 mm long; 1.8 mm wide. Similar to male except darker cuticle 
in Most specimens. Punctation, sculpturing, and vestiture similar to those of 
male except front and middle tibial spines absent, spines on sterna absent, and 


A ReEvISION OF THE GENUs Exema or America, Nortu oF Mexico 677 


oS CS 


0.1 mm. 
——__________y 
SCALE, FIGS. 51, 53-56 


92 


SCALE, FIGS. 48-50, 52, 57 


Fic. 48. Exema desert; female ninth abdominal segment, yellow form. 
Fic. 49. Exema deserti; female ninth abdominal segment, black form. 
Fic. 50. Exema deserti; male aedeagus, lateral aspect. 

Fic. 51. Exema deserti; male ejaculatory guide. 

Fic. 52. Exema deserti; male aedeagus, dorsal aspect. 

Fics. 53-55. Exema deserti; variation in tarsal claws. 

Fic. 56, Exema conspersa; tarsal claw. 


Fic. 57. Exema deserti; antenna. 


678 Tue University SciENCE BULLETIN 


midventral area of Sth sternum concave, surrounded with a few long hairs. 
Pygidium feebly tricarinate with lateral carinae large and rounded. 

Type material: The species was described from a single male specimen 
from Victorville, San Bernardino Co., California, on the Mojave Desert, col- 
lected by G. P. Mackenzie, 20 May 1939 (LACM). All types of synonyms 
were examined along with most of the paratypes of Pierce’s species. The deter- 
mination labels on the types and paratypes of parvisaxt, chiricahuana, globen- 
sis, and deserti (LACM) indicates that they are subspecies of desert, yet 
Pierce (1940) treated all but boregensis as species. 

Discussion: This species shows wide variation in color and sculpturing. 
Normally, specimens have small or indistinct tubercles, but the size can vary 
as mentioned below. There is some variability in the shape of the tooth of the 
tarsal claws, the diagnostic character used in the key; however, the tooth is 
always small. The prosternum is variable in shape and sculpturing and may 
vary in color from completely yellow to black, the intermediate individuals 
having a central yellow spot of variable size. The same is true for the pygid- 
ium and the scutellum. The most diagnostic character is the tooth of the 
tarsal claws (Fig. 54). 

The color of this species is variegated yellow and black with variation 
ranging from mostly yellow to mostly black. Most populations are made up 
of specimens with more yellow than black. Several Califormia and three Utah 
specimens, which are almost entirely yellow, have only a faint tooth on the 
tarsal claws. It is so small that it appears to be missing, giving the claws a 
shape similar to that seen in specimens of Exema with a striate or rugose 
pronotum. There are also specimens of desert: with typical claws. These speci- 
mens are as yellow as those mentioned above from Utah. All of these yellow 
forms lack the dark central punctures near the vertex that are so common in 
the typical forms. The bases of the antennae and the ventral part of the clypeus 
and the genae are colored brownish yellow rather than black, and the sculp- 
turing is smoother than in the typical form. Many intermediate characters are 
present in the larger series and in individual specimens from scattered popu- 
lations. 

The three largest samples examined were one from Sabino Canyon, Ari- 
zona, and two from west of Superior, Arizona. The two Superior samples 
were collected at about the same time of the year, one in 1960 on Franserta 
ambrosioides, a plant that grows along dry stream beds and the other on Beb- 
bia juncea. In 1962, I tried collecting from Franseria, but the plants were dried 
up and no Exema were present; however, I found the beetles quite abundant 
on the roadside weed, Bebbia juncea. 1 cannot detect any morphological dif- 
ference between specimens of these two populations, but rather suspect that 
the beetles moved to the roadside weeds because of the drying up of the other 
host plant. Both populations show a large degree of variation, especially in 


A Revision oF THE GENUS Exema or America, Nortu of Mexico 679 


characters used by Pierce to distinguish his species. It is from these two popu- 
lations and the one from Sabino Canyon that I gain evidence for synony- 
mizing Pierce’s parvisaxt, chiricahuana, deserti, boregensis, and globensis with 
desert. For example, the yellow coloring on the pygidium varies from black 
to almost completely yellow. The pygidium varies from smooth and punctate 
to highly tricarinate. This condition is common in the darker forms, those 
that Pierce called chiricahuana and parvisaxi. 

This dark form has most of its yellow coloring on the face and anterior 
half of the pronotum and only a few spots on the elytra and other parts of the 
insect. For a time I felt that the dark form was a distinct species, but further 
study revealed too many intermediate forms, too many mixed samples, and a 
very irregular distribution. Many samples had approximately equal numbers 
of both dark and yellow forms. Three collections made near Sedona and 
Globe, Arizona, and one from the Jemez Mountains of New Mexico, con- 
tained dark, yellow and intermediate forms. Collections from a great many 
more localities contained only the intermediates. Such variation occurs irreg- 
ularly over the entire range of the species, although dark specimens are more 
abundant in collections from eastern Arizona and New Mexico. 

There seems to be a slight difference in the female genital structures of 
dark forms found on Gutierrezia and the yellow ones on Franseria and Beb- 
bia; therefore, I have figured both types (Figs. 48-49). Other samples contain 
many intermediate individuals, and it would be impossible to separate them 
into two distinct groups by this character. The dark form is usually the more 
roughly sculptured. 

Most specimens from a population at Mesilla Dam, New Mexico, have a 
small acute tooth on the tarsal claw (Fig. 55) as opposed to a rounded tooth 
(Fig. 53-54). The size of the tooth is within the range of variation of deserti, 
and other characters agree more closely with this species than any other. 

Some specimens of dispar, conspersa and deserti from the southwest 
are hard to distinguish because of the similarity in color and sculpturing. 
The variation of these two characters in the three species is so great that a few 
individuals in each species appear to belong to one of the other species. It is 
only with the aid of the characters used in the key or several characters used 
in combination that one can be certain of the identification. 

T have illustrated the male and female genitalia of deserti but not conspersa 

_and dispar; all are similar with no apparent taxonomic differences. 
It should also be pointed out that in Pierce’s paper the description of parvi- 
_saxt is based on “four males” but that he gives a description of a female. The 
type proves to be a female, but paratypes 2 and 3 are males. 

Plant records: Bebbia juncea, Franseria ambrosioides, Franseria deltoides, 

Larrea tridentata glutinosa, Gutierrezia sarothrae, Ambrosia psilostachya, 


680 Tue University ScIENCE BULLETIN 


Pluchea sericea, Gutierrezia lucida, Baccharis sp.. Hymenoclea salsola, and at 
black light. 


Specimens examined: (740 males and 660 females). ARIZONA: Cochise Co., 18 May-29 
Sept.; Coconino Co., 5 Apr.-13 Aug.; Gila Co., 23 Apr.-10 Aug.; Graham Co., 8 July-16 Aug.; 
Mohave Co., 4 July-26 Aug.; Maricopa Co., 5 June-31 Aug.; Navajo Co., 9 June-24 July; Pima 
Co., 20 Mar.-25 Sept.; Pinal Co., 23 Feb.-10 Aug.; Santa Cruz Co., 22 July-3 Oct.; Yavapai Co., 
14 June-19 Aug.; Yuma Co., 8-21 July. BAJA CALIFORNIA: Catavina, 29 July-2 Aug.; 10 mi. 
S. Catavina, 29 July; Cedros Island, 4 June; San Quentin, 2 Aug.; San Vincente, 8 July; Socorro, 
1 Aug. BAJA CALIFORNIA SUR: 20 mi. N. Comondu, 2 Aug.; Miraflores, 8 July; 45 mi. N. 
San Ignacio, 27 July; 24 mi. W. Santa Rosalia, 2 Aug.; Todos Santos, 15 July. CALIFORNIA: 
Imperial Co., 12-16 June; Inyo Co., 3 Apr.-9 Sept.; Kern Co., 14 July-22 Aug.; Los Angeles Co., 
9 June-10 Sept.; Monterey Co., 22 July; Riverside Co., Apr.-22 July; San Bernardino Co., 17 
Mar.-12 Nov.; San Diego Co., 6 Apr.-23 Sept.; Siskiyou Co., 23 Mar.; “Walker Pass,” 9 June. 
COLORADO: Montezuma Co., 27 July. JALISCO: S.E. slope, Mount Colima, 2 July. NEVADA: 
Elko Co., no date. NEW MEXICO: Dona Ana Co., 25 Apr.; Lincoln Co., 9 July; Otero Co., 9 
May-l1 June; Sandoval Co., 4 June-23 Aug.; Socorro Co., Aug. SONORA: 27 mi. N. Guaymas, 
16 Mar.-1I8 June; Pitiquito, 4 July; 20 mi. S.E. Sonoita. 10 June. TEXAS: Brewster Co., 10 June- 
28 July; Jeff Davis Co., 9 May-4 July; Presidio Co., 20 Feb.-May. UTAH: Emery Co., no date; 
Grand Co., 25 June; Kane Co., 16 Aug.; Sevier Co., 26 Aug.; Washington Co., 25 Apr.-30Aug.; 
“Chad’s Ranch,” 26 July. 


Exema conspersa (Mannerheim) 


Chlamys conspersa Mannerheim; 1843:311; Lacordaire, 1848:843; LeConte, 1857:24 (distribu- 
tion). 

Chlamys rugulosa Motschulsky; 1845:109; Lacordaire, 1848:844; LeConte, 1857:24 (distribution). 

Exema conspersa (Mannerheim); Crotch, 1873:30; Hubbard and Schwartz, 1878:660 (mis- 
identification) ; Beutenmuller, 1890:175 (biology, probably misidentification) ; Linell, 1897:480; 
Clavareau, 1913:221: Achard, 1914:17 (listed); Leng, 1920:288 (listed); Fall, 1927:386 
(distribution); Leonard, 1928:463 (distribution); Moore, 1937:93 (distribution and biology); 
Burks, 1940:336, 354 (parasitized by Spilochalcis sanguiniventris); Pierce, 1940:8; Brown, 
1943:124 (note on synonymy); Peck, 1963:886, 955. 

Exema conspersa (Mannerheim) probably dispar; Blatchley, 1920:69; Britton, 1920:273 (dis- 
tribution); Blatchley, 1924:50 (distribution and hosts); Johnson, 1927:114 (distribution) ; 
Proctor, 1938:56 (biology note); 1946:187 (biology note). 

Exema jenksi Pierce; 1940:13-18; Blackwelder and Blackwelder, 1948:42 (listed). NEW 
SYNONYMY). 

Exema inyoensis Pierce; 1940:19-20; Blackwelder and Blackwelder, 1948:42 (listed). (NEW 
SYNONYMY). 


Diagnosis: This species is variegated with equal amounts of black and _ 
yellow throughout the cuticle with no detectable pattern in most specimens. — 


Specimens of conspersa are usually found west of the Rocky Mountains (Fig. 
64) on species of Artemisia and closely related genera. 
Type: The type is in the Museum Zoologicum Universitatis, Helsinki, 


Finland (MZU). The redescription below was made from this type. The fol- | 


lowing labels appear on the pin: é ; Etholén (name of the ship’s captain who 
gave Mannerheim a collection of beetles from California) ; Coll. Mannerheim; 


Conspersa Mannerh. Dispar Dejean; Var. A. Lacord. Chl. rugulosa Motsch.; | 


Mus. Zool. Helsinki No. 7722. The specimen was collected by Tschernikh, 
probably around Bodega, Sonoma Co., California. 

The types of jenkst and inyoensis were also examined (LACM). The first 
agrees with the type of conspersa and the second is an example of a dark form 
of conspersa. 


A REVISION OF THE GENUs Exema or AMERICA, Nortu oF Mexico 681 


Redescription of male: 2:65 mm long; 1.6 mm wide. Variegated yellow 
and black or dark brown; face yellow to brownish yellow with many black 
punctures, antennal bases, genae and ventral clypeus black; labrum yellow, 
with brown apex; frontal half of pronotum yellow, variegated with black and 
black punctures, yellow and black about equal over most of body; femur with 
subapical yellow ring, tibia with subbasal and subapical rings; tarsomeres 
brown; irregular yellow areas on metaepisternum, Ist and 5th abdominal 
sterna, and each side of pygidium. Head minutely granulate among large 
setigerous punctures; upper part of yellow area between antennal bases with- 
out punctures; antennae clothed with many short and a few long hairs; seg- 
ments 3-5 subequal, 6 smaller than 7, 7-10 subequal, 11 subtriangular; labrum 
smooth, slightly emarginate at apex, with transverse row of about 6 ventrally 
projecting long hairs. Pronotal gibbosity with poorly defined tubercles, small 
anterior carina converging posteriorly, separated by 2 rows of punctures, sum- 
mit-2 developed into a small anterolateral projecting carina; entire pronotum 
with conspicuously setigerous punctures; areas between punctures granulate. 
Scutellum finely granulate, lateral scutellar wings small. Elytral punctures 
conspicuously setigerous, punctures large and deep, areas between tubercles 
granulate; umbone broad; large tubercles: anterior, sutural-l, 2, 3, 2a, 3a, 
umbonal-b and marginal-b; all others very small or absent; discal depression 
well developed; legs feebly punctate, tarsal claws deeply toothed, large apical 
spines on front and middle tibiae. Anterior edge of prosternum bent ventrally, 
coarsely punctate, granulate; prosternal process long and narrow. Metasternal 
punctures large and shallow. First abdominal sternite with a longitudinal 
carina between coxae, carina dividing near midlength into a Y-shape and 
terminating in 2 long, ventrally projecting spines on posterior margin of seg- 
ment; punctures of Ist and 5th sterna large, round, setigerous; midventral 
area of 5th sternum flat to slightly concave, covered with slightly recurved 
hairs. Pygidium convex, distinctly carinate, covered with setigerous punctures, 
surface granulate. Male aedeagus similar to that figured for desertz. 

Female: Slightly larger than male; yellow on face variegated with black; 
yellow pattern on pronotum more irregular than in male. Punctation and 
sculpturing as in males except front and middle tibial spines absent, spines on 
abdominal sterna absent and fewer hairs on 5th sternum. Hairs surrounding 
foveal area long and straight. Pygidium tricarinate. 

Discussion: The name conspersa has been used for most of the species in 
North America at one time or another. One reason for this is the tremendous 
variation within the species conspersa, dispar and deserti. Populations of dis- 
par from the east coast and conspersa from the west coast are very different 
from each other, but as samples are taken closer together this difference is less 
obvious, even though a large area in the great plains region seems to lack both 
species. All three of the species are difficult to distinguish in areas of the South 


682 Tue University ScrENcE BULLETIN 


and Southwest where they are sympatric. Some specimens of all three may 
look superficially alike, but by the use of the characters cited in the key a 
proper identification can be made. In some specimens of dispar the body oils 
may, in time, cause the tarsomeres and the yellow pattern on the head and 
pronotum to look brown, thus causing them to resemble specimens of con- 
spersa. The setigerous punctures on the pronotum are sometimes inconspicu- 
ous in the Utah specimens of conspersa, but are very obvious on the elytra. 

As already stated, some forms of deserti are very dark and might be con- 
fused with either conspersa or dispar. Such specimens are found in Arizona 
and Mexico. They have more yellow on the anterior half of the pronotum and 
face and less on the elytra than do forms from further north, Intermediates 
are found throughout the range of this species and into Mexico in many local- 
ities. All Mexican localities have been placed on a separate distribution list, 
but included on the map. I have several specimens from Guatemala, Costa 
Rica and Nicaragua that look like conspersa, but more are needed to be cer- 
tain of the identity. These localities appear in a third distribution list but are 
omitted from the map. 

After looking at the following list of host plants one doubts that there is 
any host specificity in this species. The problem is complicated by the fact that 
the larvae may leave the host plant to pupate and be found on any number of 
nearby plants. The adults, too, seem to wander to plants near the preferred 


host. Although there are more hosts for the species of Exema than I first sup- | 


posed, I believe that each species has one to several preferred hosts. If the 
actual host plants could be distinguished in the field from those plants on 
which the beetles are only incidentally found, specimen labels would be of 
greater taxonomic importance. 

Hosts: Artemisia tridentata, A. douglasiana, A. heterophylla, A. californica, 
Lipidospartum squanatum, Ambrosia psilostachya, Hymenoclea monogyra, 
Brickellia californica, Eupatorium adenophorum, Chrysothamus nauseosus, 
Franseria bipinnatifida, Gnaphalium decurrens, Quercus sp., Flourensia cer- 
nua, and Encelia californica. Isocoma veneta and Astragalus are reported as 
hosts by Moore (1937). 


Specimens examined: (517 males and 523 females). NORTH AMERICA: ARIZONA: Cochise 
Co., 5 June-24 Aug.; Navajo Co., 6-20 July; Pima Co., 26 May-28 June; Santa Cruz Co., 2 June- 


4 Oct.; “Atascosa Mts,” 14 July; ‘“Forestdale,” 30 June. CALIFORNIA: Alameda Co., 10 Mar.- | 


3 Oct.; Butte Co., 15 July; Calaveras Co., no date; Colusa Co., 15 Aug.; Contra Costa Co., 5 
Apr.-19 Aug.; El Dorado Co., 8 Aug.; Fresno Co., 25 July; Inyo Co., 22 July-5 Sept.; Kern Co., 
9 July-2 Sept.; Lake Co., 22 June; Lassen Co., 4 July; Los Angeles Co., 26 Jan.-24 Oct.; Merced 
Co., 30 May; Modoc Co., 20 July; Monterey Co., 4 May-28 Sept.; Orange Co., 4 May-30 July; 
Plumas Co., 25 June; Riverside Co., 2 June-29 Sept.; Sacramento Co., 7 Mar.-23 Apr.; San 
3ernardino Co., 24 Apr.-29 July; San Diego Co., 8 Feb.-7 Sept.; San Luis Obispo Co., 4 July; 
San Mateo Co., 17 Aug.; Santa Barbara Co., 30 Apr.-2 Sept.; Santa Clara Co., 2 Feb.-7 Sept.; 
Santa Cruz Co., 12 June-16 Aug.; Shasta Co., 2 June-July; Sonoma Co., 30 June-Aug.; Tulare 
Co., no date; Ventura Co., 14 Feb.; Yolo Co., 24 Mar.-24 Sept.; ““Amedae,” 20 July; “Newton,” 
14 July; “Sierra Nevada,” no date. COLORADO: State record, no date. IDAHO: Bannock Co., 
31 Aug.:; Boise Co., no date; Canyon Co., 18 June; Elmore Co., 21-30 July; Franklin Co., 13 
July-14 Aug.; Gooding Co., 15 Aug.; Idaho Co., 10 Aug.; Jerome Co., 21 June; Lemhi Co., 6 


A RevIsIOn oF THE GENUs Exema or America, Nortu oF Mexico 683 


Sept.; Nez Perce Co., 17 May; Washington Co., 26 July-1 Oct.; “Eureka,” 23 June. NEVADA: 
Humboldt Co., 29 May; Pershing Co., 27 Aug.; Washoe Co., 29 June. NEW MEXICO: State 
record, no date. OREGON: Baker Co., 6 Sept.; Deschutes Co., 30 July; Grant Co., 14 June-10 
July; Harney Co., 18 June-12 Aug.; Jefferson Co., 14 June; Lake Co., 27-28 June; Lane Co., 25 
Apr.; Malheur Co., 15-26 June; Union Co., 12 June; Wasco Co., 19 May. TEXAS: Val Verde 
Co., 14 June. UTAH: Box Elder Co., 29 Sept.; Cache Co., 2 Aug.-18 Nov.; Salt Lake Co., 14 
June; Summit Co., 12 June; Tooele Co., no date; Utah Co., 4-19 Aug.; “Raysville,” 21 Sept. 
WASHINGTON: Asotin Co., 20 Apr.; Chelan Co., 4 June; Whitman Co., 1 May; Yakima Co., 
26 May. MEXICO: BAJA CALIFORNIA: San Felipe, 15 June. BAJA CALIFORNIA SUR: Cape 
San Lucas, no date; Miraflores, 8 July; San Bartolo, 6 May; 5 mi. W. San Bartolo, 13 July; San 
Jose del Cabo, no date. COLIMA: Tonila, no date. GUERRERO: Chilpancingo, no date. HI- 
DALGO: Ixmiquilpan, 10 June. JALISCO: El Castillo, 25 Apr.; 22 mi. N.W. La Piedad, 23 
July. MICHOACAN: Patzcuaro, 26 Mar. MORELOS: near Alpuyeca, 30 Mar.-19 June; near 
Cuautla, 8 Aug.; Cuernavaca, Apr.-Nov.; 9 mi. E. Cuernavaca, 23 June; 4 mi. N.W. Cuernavaca, 
12 Apr.-17 June; 3 mi. N. Cuernavaca, 14 Mar.-10 Apr.; 3-6 mi. S. Cuernavaca, 17 Apr. 
NAYARIT: 34 mi. N. Ixtlan, Del Rio, 18 July; Jesus Maria, 26 June-27 July; Tepic, 20 July-24 
Sept. OAXACA: Oaxaca, 8 July; Rin Antonio, no date. PUEBLA: Cacaloapan, 26 Apr.; 15 mi. 
S.W. Peltalcingo, 13 Apr.; 2 mi. N.W. Tehuacan, 25 Apr. SAN LUIS POTOSI: 7 mi. E. Valles, 
29 May. SONORA: Alamos, 15-16 June; Cocospera Canyon, 8 mi. E. Imuris, no date; La Aduena, 
15 Mar.-12 June; 10 mi. E. Navajoa, 13 Aug. CENTRAL AMERICA: COSTA RICA: Ciruelas, 
Heredia, 28 Apr. GUATEMALA: N. of Cabafias, Zacapa, 10 Aug.; Coban, Alta Verapaz, 15 
June; Duenas, Suchitepequez, no date; Guatemala City, Guatemala, no date; Moca, Suchitepequez, 
21 June; Purula, Baja Verapaz, 15 June; Santa Clara in interior valley of Sierra de las Ninjas, N. 
of Cabafias, Zacapa, 10 Aug.; S. P. Yepocapa, Chimaltenango, 21 May. NICARAGUA: Chi- 
nandego, Chinandego, no date; San Marcos, Carazo, no date. 


Exema dispar Lacordaire 


Chlamys dispar Dejean; 1836:440 (nomen nudum). 

Exema dispar Lacordaire; 1848:850-852; Gemminger and Harold, 1874:3308 (listed); Dury, 
1879:11 (distribution); Dugés, 1881:5-7, fig. 1-16 (larvae); Jacoby, 1881:89; Hamilton, 
1895:339,370; Xambeu, 1899:68-69 (biology); Blatchley, 1910:1116 (distribution and_biol- 
ogy); 1920:69; Proctor, 1938:156 (biology); 1946:484, 489 (parasitized by Spilochalcis albi- 
frons and Tetrastrichus chlamytis); Pierce, 1940:9-10; Muesebeck, et al., 1951:128, 589; Peck, 
1963:128, 871-872, 955. 

Exema conspersa dispar Lacordaire; Henshaw, 1885:106 (listed); Wickham, 1896-97:153; Leng, 
1920:288 (listed); Brisley, 1925:168 (= conspersa); Leonard, 1928:462 (distribution); 
Powell, 1941:156 (male genitalia). 

Exema pennsylvanica Pierce; 1940:18-19; Brown, 1943:123-124; Proctor, 1946:187 (distribution 
and biology); Blackwelder and Blackwelder, 1948:42 (listed); Fattig, 1948:8 (distribution) ; 
Wilcox, 1954:395; Dillon and Dillon, 1961:670. (NEY SYNONOMY.) 


Diagnosis: This species is black variegated with spots of yellow on most of 
the body, and large yellow areas on the head and anterior slope of the prono- 
tum. The last tarsomere is short, only about 1/3 longer than the third. It is 
found from the Great Plains eastward (Fig. 65) on several genera of compo- 
sites, including Ambrosia, Helianthus, and Eupatorium. 

Types: Exema dispar Lacordaire, United States. This type is unable to be 
located. Exema pennsylvanica Pierce, Allegheny Co., Pennsylvania, Klages 
(LACM). 

Several different species of North American Exema have been identified 
as dispar. Although Lacordaire’s description is long and detailed, several state- 
ments make positive application of this name impossible. Unsuccessful at- 
tempts were made to locate the type in various museums in Europe. From 
these efforts it is believed that the type is lost and it is necessary to designate 
a neotype for this species in order to avoid confusion and establish stability of 
names in the genus. 


684 Tue Universiry ScriENcCE BULLETIN 


This species was first considered a synonym of conspersa (Crotch, 1873). 
Next is was applied probably to specimens of canadensis and neglecta (Blatch- 
ley, 1920). Other authors have considered it a subspecies of conspersa and one 
used dispar in place of conspersa (Jacoby, 1890). 

The color described for dispar could apply equally well to pennsylvanica, 
neglecta, byersi, or elliptica; however, the variation described suggests penn- 
sylvanica. The size given for dispar could also apply to any of the above 
species. On the basis of the shape of the tarsal claws, all species in the canaden- 
sis group key out to Poropleura in Lacordaire’s key; therefore, one would 
probably consider pennsylvanica a synonym of dispar. The description fits 
pennsylvanica very well except for the statement “elle est striée sur ses flancs” 
(it is striate on the flanks). The Latin description states “prothorace rugoso- 
punctulato” (prothorax rugose-punctate). A striate or rugose pronotum is 
found only in neglecta, byersi, canadensis, and elliptica. Lacordaire did not 
know the shape of the tarsal claw; therefore, one of these four species may be 
a synonym of dispar. Another possibility is that Lacordaire had a mixed series. 
My study indicates that this must have been the case. Lacordaire himself 
stated that there was considerable variation in the series; therefore, he desig- 
nated the most common forms as varieties. I am here designating a neotype 
for Lacordaire’s variety A. His description most nearly agrees with that of 
pennsylvanica, which now must be synonymized with dispar. 

The neotype is a male collected on Plummer’s Island (Montgomery Co.), 
Maryland, 4 May 1913, by W. L. McAtee (USNM). Maryland is chosen as 
the type locality for this species since Lacordaire’s specimens probably came 
from someplace on the eastern coast of the United States. 

Description: Male: 2.3 mm long; 1.55 mm wide. Shiny black to brown 
with yellowish or yellowish brown markings. Antenna yellow to 6th segment, 
segments 7-11 brown; mouthparts dark brown; frons yellow except emargi- 
nation of eyes and antennal sockets, clypeus, and gena black; central spot on 
vertex and area next to eyes black (similar to Fig. 25). Frontal slope of prono- 
tum yellow from antecostal suture to summit, yellow also extending laterally 
on upper 1/4 of pronotum in an irregular pattern; marginal tubercle with a 
small reddish yellow spot. Sutural tubercles 2b and 3b yellowish brown, faint 
yellowish brown on marginal tubercle and other irregular spots toward apex. 
All tibiae with subbasal and subapical yellow rings, reduced to spots on fem- 
ora; tarsomeres yellow. Entire cuticle minutely granulate, irregularly punc- 
tate. Head coarsely punctate, more sparsely so on yellow than black areas; 
punctures large, deeper on yellow areas of frons and pronotum than black 
areas; hairs in punctures inconspicuous; antenna clothed with many short 
and a few long hairs, especially toward apex; segment 5 subequal to 3 and 4, 
segments 6-11 twice as wide as 5; labrum smooth, feebly emarginate at apex, 
with ventrally projecting, long hairs. Pronotal gibbosity with rounded tuber- 


A Revision oF THE GENUS Exema or America, Nortu oF Mexico 685 


cles; cephalic and anterior pair missing; anterior carina present, connected to 
summit-1; summit-2 represented by a broken carina projecting toward lateral 
tubercle; marginal tubercle short and broad; entire pronotum with large, 
round, deep punctures inconspicuously setigerous. Scutellum feebly carinate 
with obtuse lateral wings. Elytra coarsely punctate, punctures large and deep, 
each with a short curved hair; smaller punctures on tubercles; umbone large 
and prominent, sparsely punctate but highly granulate; large tubercles: an- 
terior, sutural-1, la, 2, 2a, 3, 3a, marginal, and marginal-b; small tubercles: 
umbonal-a, umbonal-b, sutural-2b, and 3b; discal depression well developed. 
Legs shallowly punctate; tarsal claws toothed; last visible tarsomere less than 
1/3 longer than 3rd; apical spine on front and middle tibiae. Prosternum 
coarsely punctate, slightly concave; prosternal process long and narrow, 
curved dorsally. Metasternal punctures large and deep, especially laterally. 
First abdominal sternum with a longitudinal carina between coxal cavities, 
carina dividing near midlength into a faint Y-shape and terminating in 2 
ventrally projecting spines on posterior margin of segment; 2nd and 3rd seg- 
ments with similar but much smaller spines; punctures of Ist sternum large 
and round, shallow medially to deep on lateral margins; midventral area of 
5th sternum flat to slightly concave, covered with long recurved hairs; lateral 
punctures large and round. Pygidium centrally convex; faintly tricarinate, 
lateral carinae shorter but more distinct than medial, diverging slightly ven- 
trally; entire surface coarsely punctate. Male aedeagus similar to that figured 
for desertt. 

Female: 2.4 mm long; 1.6 mm wide. Similar to that of male except facial 
yellow reduced; black spot on vertex larger and extends to area between eyes; 
more black on clypeus and around antennal sockets; yellow of frons extends 
into emargination of eye. Pronotum with similar or reduced yellow areas. 
Elytra more yellowish brown than males. Punctation, sculpturing, and vesti- 
ture similar to those of male except front and middle tibial spines absent; 
spines on sterna absent and fewer hairs on 5th sternum; a few long straight 
hairs situated on periphery of a slightly concave fovea bearing a few short 
hairs. Pygidium distinctly tricarinate with lateral carinae larger than median 
one. 

Discussion: This is one of the species that has been confused with con- 
spersa because of the lack of understanding of variation in the two species and 
in the genus as a whole. Characteristics of dispar are fairly constant north of 
North Carolina, but further south there is a rapid change to a strikingly dif- 
ferent Florida form which is smaller and yellower than specimens from the 
North and in which the pronotum appears to be blistered, because of the large, 
round, yellow. tubercles. Varying degrees of this type can be found in Ala- 
bama, Mississippi, Louisiana and Texas. The yellow color persists in some 
Texas specimens, but the pronotum has only a slightly blistered appearance. 


686 Tue Universiry ScrENcE BULLETIN 


Further north, in Kansas, this blistered appearance is completely lacking and 
the pronotum has a distinct pattern rather than scattered spots of yellow. The 
apparent cline from Florida, through the Gulf states to Texas, northeastward 
to New York, and down the coast to North Carolina, could be the result of 
intergradation of a Florida form with the northern form, resulting in a highly 
variable species. 

The intermediate Texas forms are similar to the forms of conspersa from 
the eastern edge of its range, making identification of specimens from that 


area difficult. A combination of characters rather than any single character is | 


necessary to identify specimens. There are similar forms of deserti in this 
same area, but the form of the tarsal claw is a reliable character to separate 
these from dispar. 

The male and female genitalia of this species are not sufficiently different 
taxonomically from conspersa and deserti to warrant illustrating. The shape 
of the antenna (Fig. 57) and tooth of the tarsal claws (Fig. 56) are similar to 
deserti and conspersa, respectively. 

Plant records: Helianthus tuberosus, H. hirsutus, Ambrosia trifida, A. 


psilostachya, Chrysanthemum sp., Silphium, “artichoke,” Achillea, Cercis occi-— 


dentalis, strawberry, Salix sp., Verbesina, Quercus virginiana, Bidens pilosa, | 


cabbage, Ewpatorium drummondu, E. maculatum, and E. alba. 


Specimens examined: (550 males and 520 females). ALABAMA: Clay Co., 3 May; Macon | 


Co., 20 June; Mobile Co., 23 Apr.; Tuscaloosa Co., 26 Apr. ARKANSAS: Clay Co., no date; Lee 
Co., 30 June; Monroe Co., 4 July; Washington Co., 30 May-June. CONNECTICUT: Fairfield 
Co., 24 June-25 Sept. DISTRICT OF COLUMBIA: Rock Creek, 5 Apr.; Washington, 5-12 June. 
DURANGO: 10 mi. W. Durango, 12 July. FLORIDA: Alachua Co., 3 Apr.-16 Nov.; Dade Co., 
11 June; Duval Co., no date; Jackson Co., 9 July; Lake Co., 1-11 Mar.; Leon Co., 29 Oct.; 
Manatee Co., 25-29 Mar.; Palm Beach Co., 27 Mar.; Pinellas Co., 28 Jan.-26 Aug.; Seminole Co., 
15 Mar-13 June; Suwannee Co., no date; Volusia Co., 15 Apr.-16 May. GEORGIA: Clarke Co., 
22 May; De Kalb Co., 3 June; Fulton Co., 2 June; Glynn Co., Aug.; Rabun Co., 15 May. 
ILLINOIS: Alexander Co., 29 Sept.; Champaign Co., 15 Feb.; Cook Co., 1 June-11 Sept.; La 
Salle Co., 15 May-2 July; Macon Co., 8 May. INDIANA: Crawford Co., 30 Aug.; Harrison Co., 
16 June; Knox Co., 5 July-13 Sept.; Lake Co., 29 May; Lawrence Co., 8 June; Marion Co., 15 
May-8 Sept.; Posey Co., 24 Sept.; Putnam Co., 29 May; Starke Co., 18 June; Tippecanoe Co., 17 
May; “So. McAlister,” no date. IOWA: Dickinson Co., 22 Aug.; Louisa Co., June; Story Co., 18 
May-4 Oct. KANSAS: Cherokee Co., 15-31 May; Douglas Co., 23 Apr.-15 Aug.; Jefferson Co., 
21 May; Linn Co., 21 May; Montgomery Co., 26 May-2 June; Riley Co., no date; Shawnee Co., 
14 May. KENTUCKY: Edmonson Co., no date; Floyd Co., 20 June; Rowan Co., 11 June. 
LOUISIANA: (Parish=county of other states) Caddo Par., 19 Aug.; Cameron Par., 20 June-14 
Aug.; Jefferson Par., June; Natchitoches Par., 16 Aug.; Sabine Par., 16 June; St. Landry Par., 26 
Apr.; St. Mary Par., 3 May. MARYLAND: Baltimore Co., 17 June; Charles Co., 25 May; Mont- 
gomery Co., 9 Apr.-16 Sept.; Prince Georges Co., 14 May-21 July. MICHIGAN: Kalamazoo Co., 
25 June. MINNESOTA: Hennepin Co., 12 June-4 Sept.; Houston Co., 2 June; Olmsted Co., no 
date. MISSISSIPPI: Adams Co., 27 May; Grenada Co., 21 Apr.-13 May; Montgomery Co., 27 
May; Quitman Co., 8 Sept. MISSOURI: Boone Co., 11-22 May; Cooper Co., 21 May; Greene Co., 
20 July; Howard Co., 12 May; Morgan Co., 13 May; Pettis Co., 10 June; St. Louis Co., 6 Feb.- 
9 June; Shannon Co., 28 May; Wayne Co., no date. NEBRASKA: Antelope Co., 22 June; 
Cuming Co., June; Otoe Co., June; Sarpy Co., 9 May; “Sand Hills,” July. NEW JERSEY: Bergen 
Co., 14 June. NEW YORK: Erie Co., 24 Aug.; Herkimer Co., 23 May-30 June; Livingston Co., 
24 May; Monroe Co., 10 June; New York Co., 9 July; Onondaga Co., 23 June-14 Sept.; Rockland 
Co., 20 Oct. NORTH CAROLINA: Jackson Co., 1 June; McDowell Co., 25 June; Moore Co., 11 
May; Transylvania Co., 14 May-16 Aug.; Watauga Co., Aug. OHIO: Champaign Co., 10 May-31 
Aug.; Columbiana Co., no date; Delaware Co., 2 May-4 Sept.; Franklin Co., 24 May-6 June; 
Greene Co., 28 Apr.-19 Sept.; Hocking Co., 8-20 May; Lake Co., no date; Lawrence Co., 22 


| 


A Revision oF THE GENUS Exema or America, Nort oF Mexico 687 


Mar.-29 May; Lorain Co., July-12 Aug.; Ross Co., 22 May; Seneca Co., 5 Aug.; Summit Co., 12 
June; Washington Co., 15 June. ORLAHOMA: Jefferson Co., 18 May; Muskogee Co., 29 May- 
19 Sept.; Tulsa Co., 21 June. ONTARIO: Fuller, 3 mi. E. Ivanhoe, Hastings Co., 23 July-24 
Aug.; “W. Ont.,” no date. PENNSYLVANIA: Alleghany Co., 2-4 July; Dauphin Co., 24 May- 
10 July; Philadelphia Co., no date; Westmoreland Co., 10 June-26 July. SAN LUIS POTOSI: 
Huichihuayan, 25 Aug.; Tamazunchale, 7 Mar. SOUTH CAROLINA: Beaufort Co., 7 July; 
Charleston Co., 10 July; Greenville Co., 26 Aug.; Orangeburg Co., 27 June; Richland Co., 21 
Aug. SOUTH DAKOTA: Union Co., 23 July. TAUMAULIPAS: Rio Guayelajo, Victoria, 25-27 
Aug. TEXAS: Bandera Co., 10 Aug.; Bexar Co., 16 June-23 Sept.; Blanco Co., no date; Brazos 
Co., 18 Mar.-26 Sept.; Brooks Co., 20-25 July; Burleson Co., 5 Feb.; Cameron Co., 4 Apr.-27 
Nov.; Colorado Co., 19 May; Comal Co., 9 June-9 Aug.; Dallas Co., 25 Apr.-4 Nov.; Duval Co., 
26 May-13 June; Gillespie Co., 2 May-20 June; Harris Co., 6 Aug.; Harrison Co., 7 June; Hays 
Co., 5 July; Hidalgo Co., 8 May-30 July; Jim Wells Co., 24 July; Karnes Co., 23 July; Kerr Co., 
11 Apr.; Lee Co., July; Medina Co., 3 Sept.; Milam Co., 4 Apr.-3 Aug.; Parker Co., 6 May; Red 
River Co., 10 Apr.; Sabine Co., 25 Mar.; San Patricio Co., 25 Mar.; Uvalde Co., 14 June-11 July; 
Val Verde Co., 27 Aug.; Victoria Co., 11 Apr.-31 July; Wharton Co., 18 Apr.; Williamson Co., 
1 May; Zavala Co., 11 Aug. VIRGINIA: Alexandria, 22 May-9 June; Arlington, 10 May-8 July; 
Fairfax Co., 22 May-23 Sept.; Henrico Co., 15 May; Nelson Co., 28 June; Nottoway Co., 17 May; 
Princess Anne Co., July; Rockbridge Co., 5 July; Shenandoah Co., 22 May; Spotsylvania Co., 8 
June; Stafford Co., 23 May. WEST VIRGINIA: Mercer Co., 12 June; Monongalia Co., 28 June. 
WISCONSIN: Dane Co., 16 June; Grant Co., Jan. 


ACKNOWLEDGMENTS 


Sincere thanks go to the following institutions and individuals who kindly loaned material 
for this study: 

Dr. N. L. Anderson, Montana State College; Dr. G. E. Ball, Univ. of Alberta; Dr. W. F. Barr, 
Univ. of Idaho; Mr. W. J. Brown, Canada Department of Agriculture; Dr. H. R. Burks, Agricul- 
tural and Mechanical College of Texas; Dr. O. L. Cartw-ight, U.S. National Museum; Dr. L. 
Chandler, Purdue Univ.; Dr. P. J. Darlington, Museum of Comparative Zoology, Harvard Univ.; 
Dr. W. R. Enns, Univ. of Missouri; Dr. R. Fischer, Michigan State Univ.; Dr. M. A. Goodrich, 
Pennsylvania State Univ.; Dr. H. J. Grant, Academy of Natural Sciences of Philadelphia; Dr. W. 
Hackman, Museum Zoologicum Universitatis, Helsinki; Mr. W. J. Hanson, Utah State Uniy.; Dr. 
C. L. Hogue, Los Angeles County Museum; Dr. P. Hunter, Univ. of Georgia; Dr. M. T. James, 
Washington State Univ.; Dr. W. E. LaBerge, Univ. of Nebraska; Dr. J. L. Laffoon, Iowa State 
Univ.; Dr. J. Lattin, Oregon State College; Mr. H. B. Leech, California Academy of Sciences; Dr. 
K. W. MacArthur, Milwaukee Public Museum; Dr. A. T. McClay, Univ. of California, Davis; 
Mr. N. Marston, Kansas State Univ.; Dr. J. T. Medler, Univ. of Wisconsin; Mr. T. B. Moore, 
Snow College, Ephraim, Utah; Dr. L. L. Pechuman, Cornell Univ.; Dr. R. L. Post, North Dakota 
Agricultural College; Dr. J. A. Powell, Univ. of Califormia, Berkeley; Dr. L. H. Rolston, Univ. of 
Arkansas; Dr. E. I. Schlinger, Univ. of California, Riverside; Dr. J. B. Schmitt, Rutgers Univ.; 
Dr. J. A. Slater, Univ. of Connecticut; Dr. C. F. Smith, North Carolina State College; Dr. F. W. 
Stehr, Univ. of Minnesota; Dr. C. A. Triplehorn, Ohio State Uniy.; Mrs. Patricia Vaurie, Amer- 
ican Museum of Natural History; Dr. G. E. Wallace, Carnegie Museum; Dr. H. V. Weems, 
Florida Department of Agriculture; Dr. R. Wenzel, Chicago Natural History Museum; Dr. F. 
Werner, Univ. of Arizona; Dr. J. A. Wilcox, New York State Museum. 

Appreciation is expressed to Dr. R. L. McGregor, The University of Kansas, and Dr. S. L. 
Welsh, Brigham Young University, for making plant determinations. 

Thanks also go to my wife, Sarah W. Karren, for the many hours of typing, drawing, and 
constructive criticism in the preparation of this paper. 

Finally, I wish to thank the faculty of the Department of Entomology, The University of 
Kansas, for their assistance, especially Dr. George W. Byers, under whose direction this study 
Was made. 


REFERENCES 


Acnarp, J. 1914. Coleoptera Phytophaga, family Chrysomelidae, subfamily Chlamydinae & 
Sphaerocharinae. In P. Wytsman, Genera Insectorum, fac. 160:1-25, 2 col. pls. 

Arnett, R. H. 1963. The beetles of the United States. Catholic Univ. Press, Washington, D.C. 
1112 pp. 

BEUTENMULLER, W. 1890. Food habits of some Chrysomelidae. Ent. Am. 6:175-178. 


688 Tue University ScrENcE BULLETIN 


BLAcKWELpER, R. E. 1946. Check-list of the Coleopterous Insects of Mexico, Central America, 
the West Indies, and South America, part 4. Smithsonian Institution, U.S. Nat. Mus. 
Bull. 185:551-763. 
. 1957. Check-list of the Coleopterous Insects of Mexico, Central America, the West In- 
dies, and South America, part 6. Smithsonian Institution, U.S. Nat. Mus. Bull. 185:927- 
1492. 
BLAcKWELDER, R. E., and R. M. Birackwetper. 1948. Fifth Supplement 1939 to 1947 (in- 
clusive) to the Leng Catalogue of Coleoptera of America, North of Mexico. Sherman, 
New York. 87 pp. 

Brake, D. H. 1952. American Chrysomelidae in the Bosc collection. Proc. Ent. Soc. Washing- 
ton 54:57-68. 

Biatcuiey, W. S. 1910. The Coleoptera or beetles of Indiana. Bull. Indiana Dept. Geol. Nat. 
Res., No. 1, 1386 pp. 


————. 1913. On some apparently new Coleoptera from Indiana and Florida, with descriptions 
of new species. Canadian Ent. 45:21-24. 


——. 1920. Notes on the winter Coleoptera of western and southern Florida, with descrip- 
tions of new species. Canadian Ent. 52:42-46, 68-72. 

——. 1924, The Chrysomelidae of Florida. Florida Ent. 7:33-39; 49-57; 8:1-7; 17-23; 39-46. 

. 1930. Blatchleyana. Nature, Indianapolis. 77 pp. 


Bristey, H. R. 1925. Notes on the Chrysomelidae (Coleoptera) of Arizona. Trans. American 
Ent. Soc. 51:167-182. 


Britton, W. E. 1920. Check-list of the insects of Connecticut. Connecticut Geol. Nat. Hist. 
Surv. Bull. 31. 397 pp. 

Brown, W. J. 1943. The Canadian species of Exema and Arthrochlamys (Coleoptera, Chryso- 

melidae). Canadian Ent. 75:119-131. 

.1961. Notes on North American Chrysomelidae (Coleoptera). Canadian Ent. 93:967- 

Wile 

Burks, B. D. 1940. Revision of the chalcid-flies of the tribe Chalcidini in America North of 
Mexico. Proc. U.S. Nat. Museum 88:237-354. 

Cuacnon, B. 1937. Contribution at l'étude des Coléoptéres de la Province de Québec. Le Nat. 
Canadienne 64:218-228. 

Cuapuis, M. F. 1874. Histoire naturelle des insectes. Genera des Coléoptéres . . . . vol. 10, 
Paris. 455 pp. 

Cuuy6, M. 1940. Fulcidacinae of Formosa (Col. Chrysomelidae). Trans. Nat. Hist. Soc. For- 

mosa 30:265-293. 

.1955. A taxonomic study on the Chrysomelidae (Insecta-Coleoptera) from Formosa. 

IX subfamily Chlamisinae (1). Tech. Bull. Kagawa Agr. Coll. 7:51-59. 

Cravareau, H. 1913. Coleopterorum Catalogus, pars 53. W. Junk, Berlin. 278 pp. 


CoourBert DE Montsret, A. J. 1804. Illustratio Iconographica Insectorum. 3rd Decade. Didot, 
Paris, pp. 91-142. 10 col. pls. 

Crorcu, G. R. 1873. Materials for the study of the Phytophaga of the United States. Proc. 
Acad. Nat. Sci. Philadelphia 25:19-83. 

Deyean, P. F. M. A. 1836-1837. Catalogue des Coléoptéres. 3rd edition, book 5. Paris, 503 pp. 

Dexter, G. W. 1957. A leaf beetle feeding on the stems of lychee. Proc. Florida State Hort. 
Soc. 70:331-333. 

Ditton, E. S. and L. S. Ditton. 1961. A manual of common beetles of eastern North Amer- 
ica. Row, Peterson and Co., Evanston, Illinois. 884 pp. col. pls. a-d. 

Ducis, E. 1881. Métamorphosis de l’Exema dispar. Ann. Soc. Ent. Belgique 25:5-7, pl. 1, Fig. 
1-16. 

Dury, C. 1879. List of the Coleoptera observed in the vicinity of Cincinnati. Jour. Cincinnati 
Soc. Nat. Hist., Oct. 1879, pp. 1-15. 


Epwarps, J. G. 1949. Coleoptera or beetles east of the Great Plains. Edwards Brothers, Inc., 
Ann Arbor, Michigan. 181 pp., 1 pl. 


applying particularly to Western United States. Published by the author. pp. 183-212. 


Empen, F. van 1932. Die Larven von Discoloma cassideum Reitt. (Col. Colyd.) and Skwarraia | 


paradoxa Lac. (Col. Chrys.), Zool. Anz. 101:9-17. 


. 1950. A bibliographical supplement to “Coleoptera or bettles east of the Great Plains” | 


A REvIsIon oF THE GENUs Exema or America, Nortu oF Mexico 689 


Fasricius, J. C. 1798. Supplementum entomologiae systematicae, Hafniae 522 pp. 
. 1801. Systema Eleutheratorum. Kiliae, vol. 2, 687 pp. 
Fatt, H. C. 1927. Expedition of the California Academy of Sciences to the Gulf of California 


in 1921. The Chrysomelidae (Coleoptera). Proc. California Acad. Sci., 4th Ser. 16:381- 
395) 


Farric, P. W. 1948. The Chrysomelidae or leaf Beetles of Georgia. Emory University Museum 
Bull., no. 6, 47 pp. 


GemMINGER, M. and E. von Harorp. 1874. Catalogus coleopterorum hucusque descriptorum 
synonymicus et systematicus. Monachii. 11:3233-3478. 

GressitT, J. L. 1946. Chinese Chrysomelid beetles of the subfamily Chlamisinae. Ann. Ent. 
Soc. America 39:84-100. 

GressiTT, J. L. and L. Kimoro 1961. The Chrysomelidae (Coleopt.) of China and Korea, Part 
1. Pacific Insects Monograph 1A. 299 pp. 

Hamivton, J. 1895. Catalogue of the Coleoptera of southwestern Pennsylvania with notes and 
descriptions. Trans. American Ent. Soc. 22:339-341. 

Husparp, H. G. and E. A. Swartz. 1878. The Coleoptera of Michigan. Proc. American Philos. 
Soc. 17:627-666. 

Itiicer, J. C. W. 1804a. Familien, Gattungen und Horden der Kafer, Coleoptera. Magazin fur 

Insektenkunde 3:1-145. 
.1804b. Ausatze, Berichtigungen und bemerkengen zu Fabricii Systema Eleutheratorum. 
Magazin fur Insectkunde 3:146-180. 
Jacopy, M. 1881. Biologia Centrali-Americana, Insecta, Coleoptera, Cryptocephalidae, Chlamy- 
didae, Lamprosomidae, Eumolpidae 6(pt. 1) :73-144. 
. 1890. Biologia-Centrali Americana, Insecta, Coleoptera, Chlamydidae, suppl., 6(pt. 
1) :169-232. 

. 1908. Fauna of British India, Chrysomelidae, vol. 1, 534 pp. 

Jenks, G. E. 1940. Dwarfs that live in their hats. Nature Mag., pp. 337-340. 

Jounson, C. W. 1927. The insect fauna with reference to the flora and other biological features. 
In W. Proctor, Biological survey of the Mount Desert Region. Part 1. Wistar Institute, 
Philadelphia. 247 pp. 

Kuve, J. C. F. 1824. Entomologische Monographien, Berlin. 242 pp. 10 pls. 

Knocn, A. W. 1801. Neue Beytrage zur Insectenkunde. Leipzig. vol. 1, 208 pp., 9 pls. 

Kotiar, V. 1824. Monographia Chlamydum. Viennae, 49 pp., 2 pls. 

Lacorpaire, J. T. 1848. Monographie des Coléoptéres subpentaméres de la Famille des Phy- 
tophages, tome second. Mém. Soc. Roy. Sci. Liége, 5:1-890. 

LeConre, J. L. 1857. Report upon insects collected on the survey. (Reports of explorations and 
surveys for a railroad route from the Mississippi to the Pacific Ocean.) Washington. 72 
pp., 2 pls. [A preprint; the report was published in 1860.] 

LeConte, J. L. and G. H. Horn. 1883. Classification of the Coleoptera of North America. 
Smithsonian Misc. Coll. vol. 26, 567 pp. 

Lene, C. W. 1918. Notes on some changes in the list of Coleoptera. Jour. New York Ent. Soc. 

26:201-211. 

. 1920. Catalogue of Coleoptera of America, North of Mexico. Sherman, New York. 470 

pp. 

Lene, C. W. and A. J. MurcHLer. 1927. Supplement 1919 to 1924 (inclusive) to Catalogue of 
the Coleoptera of America, North of Mexico. Sherman, New York. 78 pp. 

Leonarp, M.D. 1928. A list of the insects of New York. Cornell Univ. Agric. Expt. Sta. Mem- 
oir 101, 1121 pp. 

Linett, M. L. 1897. New Species of Coleoptera of the Family Chrysomelidae with a short re- 
view of the Tribe Chlamydini. Proc. U.S. Nat. Museum 20:473-485. 

MANNERHEIM, G. C. B. von 1843. Beitrag zur Kaefer-fauna der Aleutischen Inseln, der Sitkha 
und Neu-Californiens. Bull. Soc. Imper. Nat. Moscow 16:175-314. 

Monros, F. 1952. Revision de las especies argentinas de Chlamisinae (Col., Chrysomelidae) 
(5a contribucion al conocimiento de Chlamisinae). Acta Zoologica Lilloana 10:489-672. 


Moore, I. 1937. A list of the beetles of San Diego County, California. Occas. Papers San Diego 
Soc. Nat. Hist. 2:1-109. 


690, Tue University ScrENCE BULLETIN 


MorscHoutsky, V. bE 1845. Remarques sur la collection de coléoptéres russes. Bull. Soc. Imper. 
Nat. Moscow 18:1-127, illus. 

MuesEsBEck, C. F. W., K. U. Kromsein, H. K. Townes, et al. 1951. Hymenoptera of America 
North of Mexico, Synoptic Catalog. United States Dept. Agric. Monogr. No. 2. U.S. 
Govt. Printing Office, Washington. 1420 pp. 

Navayas, E. 1946. Os gendtipos da fam, Fulcidacidae (Col. Chrysomeloidea). Papéis Avulsos 
do Departamento de Zoologia (S40 Paulo) 7:243-246. 

Ouivier, A. G. 1808. Entomologie, ou histoire naturelle des insects, avec leurs caractéres 
génériques et spécifiques, leur description, leur synonymie, et leur figure enluminée, 
Paris. Coléoptéres, vol. 6, 8. 

Peck, O. 1963. A Catalogue of the Neartic Chalcidoidea (Insecta: Hymenoptera). Canadian 
Ent., Supplement 30, 1092 pp. 

Pierce, W. D. 1940. The fauna and flora of the El Segundo sand dunes. Bull. So. California 
Acad. Sci. 39:6-27. 

Procror, W. 1938. Biological survey of the Mount Desert Region, part 6. The insect fauna, 
with references to methods of capture, food plants, the flora and other biological features. 
Wistar Institute, Philadelphia. 496 pp. 

. 1946. Biological survey of the Mount Desert Region, part 7. The Insect fauna. Wistar 
Institute, Philadelphia. 566 pp. 

Powett, E. F. 1941. Relationships within the family Chrysomelidae (Coleoptera) as indicated 

by the male genitalia of certain species. American Midland Nat. 25 148-1 95s 

SCHAEFFER, C. 1906. On new and known genera and species of the family Chrysomelidae. Sci. 
Bull. Mus. Brooklyn Inst. Arts Sci. 1:221-253. 

. 1926. New species of Boloschesis (=Chlamys) with notes on known species (Cole- 
optera; Chrysomelidae; Fulcidacinae). Proc. Ent. Soc. Washington 28:181-187. 


ScCHOENHERR, C. J. 1808. Synonymia insectorum. Stockholm. Vol. 1, pt. 2, 424 pp., 4 pls. 

SmituH, J. B. 1889. Catalogue of insects found in New Jersey. Final report of the state geologist. 
Geological Survey of New Jersey, vol. 2, part 2, 486 pp. 

Wicxnam, H. F. 1896. The Coleoptera of Canada. The Chrysomelidae of Ontario and Quebec, 
part 16. Canadian Ent. 28:151-157. 

Wicox, J. A. 1954. Leaf beetles of Ohio (Chrysomelidae:Coleoptera). Ohio Biological Survey 
Bull. 43:353-506. 


ZIMSEN, E. 1964. The type material of I. C. Fabricius. Munksgaard, Copenhagen. 656 pp. 


A RevIsION OF THE GENUs Exema or America, Nortu oF Mexico 691 


mormona 


Fic. 58. Range of Exema gibber. Each spot represents one or more collections within a county 
or at a locality. 


Fic. 59. Range of Exema mormona. Each spot represents one or more collections within a county 
or at a locality. S=state record. 


692 Tue University ScIENCE BULLETIN 


Ty 


ae 


e neglecta 


Fic. 60. Range of Exema canadensis. Each spot represents one or more collections within a 
county or at a locality. S=state record. 

Fic. 61. Range of Exema byersi and Exema neglecta. Each spot represents one or more collections 
within a county or at a locality. S=state record. 


5) 2 
A RevIsION OF THE GENUS Exema or America, Nortu oF Mexico 693 


Fic. 62. Range of Exema elliptica. Each spot represents one or more collections within a county 
or at a locality. 


Fic. 63. Range of Exema deserti. Each spot represents one or more collections within a county 
or at a locality. 


694 Tue Universiry ScrENcE BULLETIN 


64 
conspersa 


Fic. 64. Range of Exema conspersa. Each spot represents one or more collections within a county 
or at a locality. S=state record. 


A RevIsION OF THE GENUS Exema or America, Nortu oF Mexico 695 


Fic. 65. Range of Exema dispar. Each spot represents one or more collections within a county 
or at a locality. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


PUPATION SITE DIFFERENCES IN 
DROSOPHILA MELANOGASTER 


By 


Robert R. Sokal 


VoL. XLVI Paces 697-715 DECEMBER 2, 1966 No. 19 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Vo. XLVI Paces 697-715 DeEcEMBER 2, 1966 No. 19 


Pupation Site Differences In 


Drosophila melanogaster* 
By 


Ropert R. Soxar 
Department of Entomology, The University of Kansas, Lawrence 


EANTRODUCTION 


This paper describes an experiment selecting two lines of Drosophila 
melanogaster Meigen for central and peripheral pupation sites, respectively. 
The account includes the development of differences between the strains, cor- 
related characters and the results of crossing experiments. 

The study grew out of an incidental finding in a selection experiment on 
DDT-resistance of larvae of D. melanogaster, i.e., that resistant larvae in con- 
trol medium without toxicant tended to pupate at the margin of the medium- 
cylinder or on the wall of shell vials, while susceptible larvae tended to pupate 
away from the margin (Sokal and Hunter, 1954). 

The main purpose of these experiments was to be an investigation of the 
quantitative inheritance of this newly recognized character of drosophila. 
However, a number of factors, among them the complexities of measuring 
this quantal character, and environmentally caused fluctuations in phenotype, 
made the analysis and interpretation of these data extremely difficult. The 
results were put aside, but by the time new insights and new techniques 
would have placed the genetic interpretation of a repeat experiment on a 
sounder basis, my interests and activities had changed and the necessary stocks 
or facilities no longer existed in my laboratory. The reason for publishing the 
following account is, therefore, not as a definitive genetic interpretation of the 
findings, but because several aspects of this experiment are unique and should, 
consequently, be of interest to drosophila workers. 


*Contribution No. 966 from the Department of Entomology, The University of Kansas, 
Lawrence, Kansas. These studies were aided by the Office of Naval Research, Department of the 
Navy (Contract No. Nonr 171(00)). This investigation also was supported in part by a Public 
Health Service research career program award (No. 3-K3-GM-22, 021-01S1) from the National 
Institute of General Medical Sciences. 


698 Tue University ScrENCE BULLETIN 


Il. MATERIALS AND METHODS 


The flies were reared in six-dram shell vials (height 85 mm, diameter 21 
mm) filled with 4.4 cc of corn meal-agar-molasses medium (height of me- 
dium column 14 mm). Medium preparation is described in detail in Sokal 
et al. (1960). 

Medium for all lines in any one generation was prepared in one big batch 
in one pan. Occasionally it was necessary to prepare medium in two pans. In 
such cases pupation site records were taken separately for each batch, flies of 
any one line being divided equally into the two batches. 

The vials were seeded with ten eggs each from half-pint oviposition bot- 
tles, prepared by the method of Gowen and Johnson (1946). Each bottle con- 
tained only a single parent pair and the eggs from any one bottle were 
therefore full sibs. While in general only one vial was prepared from an ovi- 
position bottle, two or more vials were prepared from the same bottle when 
few flies were available or when the fecundity of a particular generation was 
unusually low. In these “replicated sibships,” vials coming from the same 
parent pair were identified as such. 

Seeded vials were placed in a temperature chamber at 250.5°C. Humid- 
ity was not controlled, but occasional records indicated a stable R.H. near 
40°/,. During periods of no selection, stocks were kept in the main laboratory 
at 25-£3°C, as were the oviposition bottles. 

When all larvae had pupated, pupation sites were recorded for each vial 
(see Figure 1). All pupae which were not touching the wall of the vial were 
called “central.” Pupae touching the wall, were termed “marginal” if partly 
submerged in the medium, or “on wall” when cemented to the wall of the 
vial and entirely free from medium. Since there were relatively few “on wall” 
pupae they were lumped for computational purposes with the “marginal” 
pupae, both being called “peripheral” pupae. The percent peripheral pupation, 


Ficure 1. Diagram explaining scoring of pupation site. 1-in medium, 2-at margin, 3-on wall; 
1 =central, 2 + 3 = peripheral. 


Pupation Sire Dirrerences iN Drosophila melanogaster 699 


the basic variable measured in the study, was computed from the number of 
peripheral pupae out of the total number of pupae. 

Pupation site readings for any one generation were always carried out by 
the same technician, although over the entire study a number of technicians 
were employed. Bias in readings among technicians was kept to a minimum 
by repeated checks. Furthermore, large sections of the study were assayed by 
the same person so that the fluctuations observed from generation to gener- 
ation cannot be attributed to differences between technicians. 

The lines were selected from the COSU-2 stock, synthesized from the 
Canton-S, Oregon-R-C, Sweden-C, and Urbana-S strains as described for 
COSU-1, by Hunter (1959). The COSU-2 stock had been mass mated in six 
culture bottles for three generations at the start of the present experiment. 

For pupation site assays and selection 100 vials of 10 eggs each were pre- 
pared. These vials ideally came from 100, but frequently fewer, parent pairs; 
the lowest number was 22. Each generation was to consist of one thousand 
individuals, but this figure was rarely reached. In each generation 100 pairs of 
flies were to serve as the parents of the next generation. Therefore rather more 
than 200 pupae were selected in the hope of obtaining 100 fertile adult pairs. 

Selection for high peripheral pupation (PP line) consisted of securing all 
pupae from those vials that exhibited 100°% peripheral pupation, whether 
based on 10 or fewer pupae. For additional individuals all vials with a per- 
centage of peripheral pupation above an arbitrary level (say 909%) would be 
chosen and all central pupae in them destroyed by a pin prick. The selection 
threshold would be lowered until a satisfactory number of peripheral pupae 
had been selected. 

Selection for central pupation (CP line) was done in an analogous manner. 
All vials exhibiting zero per cent peripheral pupation were saved in their 
entirety. Low peripheral percentage vials were selected up to a certain selec- 
tion threshold, with the peripheral pupae within them killed or disposed of. 

Adults emerging from the selected pupae were sexed every eight hours to 
ensure the virginity of the females and to prevent inbreeding among the full 
sibs of a vial. Sexed flies were stored in medium vials for several days to per- 
mit them to mature. The parents-to-be of the next generation were mated at 
random. When selection was interrupted for one generation all the flies 
emerging from the pupation site vials were sexed and mated at random to 
produce the next generation. 

The method of assaying larval resistance to DDT has been reported else- 
where (Sokal, 1959). The analysis of the mortality data was performed by 
standard probit analysis techniques (Finney, 1952), with maximum likelihood 
estimates obtained on a digital computer (Sokal, 1958). Length of larval and 


pupal periods was determined by a technique standard in this laboratory 
(Hunter, 1959; Neunes, 1962). 


700 Tue University SCIENCE BULLETIN 


To summarize the treatments carried out in the present study: Pupation 
site was assayed in generations 1-32, 40, 42, 44-55, 60-61, 75-81. Selection in both 
directions took place in generations 1-14, 16, 18-26, 28-32, 40, 44-45, 48, 51-55, 
76-81; for central pupation only in generations 15 and 27 and for peripheral 
pupation only in generation 17. Replicated sibships for both lines were run in 
generations 7, 14-15, 20, 28, 45, 48, 77 and 80; for the CP line only in gener- 
ation 5; and for the PP line only in generations 4, 6, 12, 13, 52, 54 and 79. The 
generation numbers of successful assays for DDT-resistance can be found in 
Table 3, those of length of developmental period in Section IV(7), and those 
of crosses in Section V. 


Il]. PROBLEMS OF SCALE 


It seemed proper to weight each percentage peripheral pupation by the 
number of pupae on which it was based. Moreover, since vials contained ten 
pupae, the presence or absence of one pupa made the difference between 90 
and 100° or 0 and 10% peripherality. Assuming the existence of a continuous 
distribution of phenotypes, a probability scale would lump all phenotypes 
beyond 90° and below 10% peripherality into two groups labeled as the most 
extreme phenotypes. 


While a relation between pupation site and density has been found at the’ 
higher densities (12 to 50 pupae per vial; Sokal e¢ al, 1960), these authors and 
. . . | 

Schlager (1959) were unable to demonstrate such a relation in either the 


COSU_2 or the BS-Canton strain at the lower densities employed in this study. 
Thus the relation demonstrated below for the CP and PP lines appears to 
have been acquired by them as a concomitant result of the selection procedure. 

After considerable trial and error the probit transformation was found to 
approximate a normal distribution of pupation sites and a linear regression of 
pupation site on density more closely than other scales. Furthermore, the 
problem of bias at the tails of the curve was also solved in this manner, since 
probit analysis routinely estimates the tails of the distribution during the max- 
imum likelihood method of computation. 

Data for all available generations were converted to probit peripheral 
pupation site and regressed on density, using the method of maximum likeli- 
hood (Finney, 1952). In the replicated sibships it was possible to perform an 
analysis of covariance of the data in addition to probit analysis. In the non- 
replicated sibships of the CP line only 1 of 48 regression coefficients 1s posi- 
tive; the mean regression of probit peripheral pupation on density is —0.0970. 
Of these 48 regression coefficients only 26 are significant at P S 0.05 (mean = 
—(.1255), but none of the significant ones is positive. The PP line did not 
show as clear a relationship. Of 41 regression coefficients with a mean of 
—(),0407, 8 are positive. Only 9 of the 41 coefficients are significant (mean 
—0(,0778) and 1 of these is positive. No trend in magnitude of the regression 


Pupation Sire Dirrerences 1s Drosophila melanogaster 701 


coefficients is evident in either line. In all but 8 of 89 instances there are no 
significant deviations from linearity (PS 0.05). Analyses of covariance for 
the replicated sibships bear out these findings. Among 10 CP generations the 
pooled regression within and between sibships averages —0.1181, the pooled 
significant regression —0.1660. In the PP the pooled regression is —0.0735, 
the pooled significant regression —0.1188. 

On the basis of these findings the average probit pupation sites for each 
line and generation were corrected by replacing them with the expected probit 
pupation sites at the (aimed at) density of ten pupae per vial. These estimates 
were computed from separate regression equations for each line and gener- 
ation. With the few exceptions of positive regression slopes, the correction 
resulted in a reduction of the probit peripheral pupation over the uncorrected 
value. Since the regression coefficient for the PP line was roughly only half 
that of the CP line the correction of the data also served to differentiate 
slightly the phenotypic expression of the character in the two selected lines. 
While the effects of these corrections are only modest, the regression analysis 
did permit appropriate weighting of each vial and satisfactory statistical treat- 
ment of the extreme phenotypes. 

The average probit pupation sites, the regression estimates, their 5°/ fidu- 
cial limits and the regression coefficients have been tabulated but are not re- 


produced here to conserve space. Copies of the table may be obtained from 
the author. 


IV. RESULTS OF SELECTION EXPERIMENT 
1. Over-all Trends 


Figure 2 shows the results of the selection experiment up to and including 
generation 81. The ordinate is in probits peripheral pupation estimated at 
densities of 10 pupae per vial. Differences between the two lines are graphed 
in Figure 3. 

There is an appreciable difference between the phenotypes of the two lines 
in the direction intended by selection (P of null hypothesis infinitesimal). 
Marked fluctuations in pupation site, often against the direction of selection 
(as in CP generation 7 and 10 or in PP generations 9 and 11) greatly compli- 
cated the evaluation of the results. The partial parallelism of these fluctuations 
indicates their environmental causation. Schlager (1960) was able to implicate 
one of the ingredients of the medium, molasses, as responsible for the fluctu- 
ations. Regrettably these findings came too late to modify the conduct of the 
present experiment. 


2. Evidence of Selection Pro gress 


In spite of the fluctuations between successive generations, increases of 
differences between line means, d, are evident between generations 2 and 32. 


702 Tue Universiry ScrENCE BULLETIN 


x 
65 
nee ! 
Bee Pty 
= | \f | i PP i 
x x 1 y ily ix ont x ) i 
D 55 Ro | \! yy Ve 1 Xie \ 
= EE cath \ ee aa 
a Key N y ¥ ne i pt 
x sot X/ t - x\! wi 
Ww | \| H x \] 
me lanllick MT 
4 
WW \I 
a \ CP 
E i 
fea 4.0 
e) 
je x 
a 
3.5 ‘ 
30 
[UECuOt eal ee me aaah JtIt1t 
10 20 30 40 50 60 70 80 
GENERATIONS 


Ficurer 2. Results of selection for peripheral and central pupation sites. Abscissa: generations of 
selection; ordinate: probit peripheral pupation. The means for each generation have been adjusted 
to expected values for a density of ten pupae per vial. Missing points for any one generation 
indicate no pupation site assay for that generation. Multiple points in any one generation repre- 
sent different series. Only means of consecutive generations are connected by lines. The PP means 
are connected by broken lines, the CP means by solid lines. The arrows along the abscissa indi- 
cate the generations in which selection for pupation site was practiced. A double-headed arrow 
signifies bidirectional selection, an arrow pointing up selection for peripherality, one pointing 
down for centrality. 


However, Figure 3 shows an end to the increase in the difference at approxi- 
mately generation 21. The regression coefficient of d on generation number 
from generation 2 to 21 is 0.100 + 0.0150. Visual inspection suggests a selec- 
tion limit for the PP line by generation 18. The regression coefficient of esti- 
mated pupation sites on generations 2 to 18 is 0.104 = 0.0157. Figure 2 suggests 
an end to selection progress in the CP line after generation 23, if not earlier. 
The regression coefficient of estimated pupation site on generations | to 23 is 
—(.037 + 0.0090. The response to selection is thus quite asymmetrical, being 
three times as large for peripherality as for centrality. A substantial number of 
pupation site assays of the parent COSU-2 stock, both during the course of 
this experiment and in connection with other experiments, showed no con- 


sistent time trend in pupation site, supporting the conclusion of asymmetry 
of phenotypic response. 


3. Estimation of Heritability 


Since the selection method practiced in the experiment was a combination 
of individual and family merit selection, the detailed computation of a prop- 


Pupation Sire Dirrerences 1n Drosophila melanogaster 703 


2.5 


2.0 


PROBIT PERIPHERAL PUPATION 


PITTI t testi tists ny eu oae Jtqttt 
oO. 1 ne N = 1 pees Sey eS eee 
10 20 30 40 50 60 70 80 


GENERATIONS 


Ficure 3. Deviation between PP and CP lines during the selection experiment. Only con- 
secutive generations are connected. Abscissa: generations of selection; ordinate: difference ex- 
pressed as probit peripheral pupation. Missing points for any one generation indicate no pupation 
site assay for that generation. Multiple points in any one generation represent different series. The 
arrows along the abscissa indicate the generations in which selection for pupation site was prac- 
ticed. A double-headed arrow signifies bidirectional selection, an arrow pointing up selection for 
peripherality, one pointing down for centrality. 


erly weighted standardized selection differential (7) for each generation 
would have been an extremely tedious process which, in view of the pheno- 
typic fluctuations, would have introduced an unwarranted element of refine- 
ment. Trial values computed for PP, generation 2 based on the two limiting 
assumptions of mass selection (individual merit) and family merit gave esti- 
mates of 0.20 and 0.12, respectively. Equally serious, however, was the problem 
of estimating the genetic gain, AG, between consecutive generations. The 
phenotypic fluctuations resulted occasionally in AG’s with signs opposed to 
that of their appropriate 7 or, conversely, in AG > 7. The course adopted was 
to obtain a single estimate of 4? as the regression coefficient (0.1076 = 0.0148 
for the first 21 generations) of the deviation d on the cumulative selection 
differential i, (see Figure 4). While such an estimate is of questionable value, 
it was the only approach feasible here. This 4? cannot be used for predictive 
purposes, since such an estimate of realized heritability (Falconer, 1960) is 
really little more than a function of the success of the selection. Choice of d to 
indicate genetic gain automatically eliminated the additive portion of the 
fluctuations. 


4. Components of Variation of Pupation Site 


The replicated sibships in various generations permitted analyses of vari- 


704 Tue Universiry SciENCE BULLETIN 


ance for pupation site and density and analyses of covariance for the two 
variables (Table 1). Variances among families (sibships) represent genetic 
and environmental differences among the various families used in each gener- 
ation. Environmental factors affecting the parents are unlikely to affect pupa- 
tion site. In the case of density, however, one can easily imagine that frequency 
and success of mating as well as the environment of the oviposition bottle 
(which, it will be remembered, contained a single parent pair) will contribute 
to among-family variance. 

Columns 1 and 2 in Table 1 give significance levels and percentages of 
total variation of variance components among families for density and probit 
peripheral pupation, respectively. The former show no clear trends and fluc- 
tuate markedly, while the latter increases for the first 20 generations of the 
PP with much reduced values for later generations. The CP line follows this 
general trend but is less clear, being based on fewer values. 


28 


26 


PROBIT PERIPHERAL PUPATION 


GENERATIONS 


Ficure 4. Cumulative selection differential, 72 compared with difference between means of se- 
lected lines, d. Abscissa: generations of selection; ordinate: probit peripheral pupation. Only the 
first 31 generations are shown. 


Pupation Site DirrereNces iN Drosophila melanogaster 705 


The analysis of covariance of the replicated sibships yielded a mean square 
around the average regression of pupation site on density within the families 
as an estimate of variance among vials of a single family (column 3 of Table 
1). It should represent entirely environmental differences between vials with 
an expected value of 1.0. Most values cluster around unity, with some evi- 
dence of underdispersion, especially in the PP, which is difficult to interpret. 
No trend appears in either selected line. The mean squares of family means 
(around the regression line; column 4 of Table 1) are in all but one case 
greater than those within families (column 3), showing the additional effect 
of genetic as well as environmental differences among families. 

Each generation of the non-replicated sibships (where each family is rep- 
resented by a single vial only) was divided into groups by density and sub- 
jected to an analysis of variance testing linear regression, deviations from 
linearity, densities and vials within densities. The regression coefficients have 


Taste 1. Analysis of Replicated Sibships 


among family 
probit vials means 
pupation within from 

CP density site sibship regression 

Generation proportion proportion MS MS 
> coded cope een ee ee 062 Su 978 1.699 
TS ee eee 055 .075* 1.042 1.731 
114 MAME es a cs ee 327** 2ipe* .986 ZB27 
5) cic a a I re 17 1* .160* e/95 2.383 
20) 5c) ER ene eerie B96" 051 1.165 1.117 
Sib en oe eee 045 103%" 4 1.212 
SD) 5k eee .050* .041* 1.119 1.617 
LOM el eke .091** 1073** 1.017 1.904 
i ore ees l08** 059 1.057 1.734 
(0) cocaine seen ene Oe 9 067 1.008 Pens 
PP 
Generation 

Gh xecciScnc eS ee an ee — 017 1.148 1.509 
OG satin 4* 022 O15 1.206 
i PO i wee nbs lsZe fa) 1.444 
WZ, aes Soylent slalSs .100* 1307 2.365 
U3 cocci ere! 290** 129 IZ We lZ2 
A scat Si 041 .166 437 0.873 
Ip eT ed 158 209 py? 1.301 
D i 103 seis OM 940 
28 co .129* 080 800 1.580 
31) cached ae er .069** lOSe% 1.161 23271 
13 ee .140** .108** 1.083 2.245 
peered Pe 047 837 718 1.662 
D4) ccabcuonecels een 023 101 894 1.633 
er es .095* 061 1.482 2.358 
“2 copter 081 066 542 895 
OL er ie .079 055 1.085 lo22 


| 
Explanation: “Proportion” refers to proportion of total variance represented by variance com- 
ponent among families. Starred proportions indicate significant variance components (* = 0.01< 
PS0.05; =< P==()'0I1)). 


706 Tue University SciENCE BULLETIN 


Taste 2. Mean Squares among Vials (Sibships) within Densities 
in Generations without Replicated Sibships 


Generation (Oe PP Generation CP 1p)p) 
OTR Se op May RODE Seale, 879 1.239 JD) cae We he eee 1.64 73**) ep Ob ee 
Sa ee ren ais N/A Zee aes 0/8 oon Ati See eee Seen 3.878 oo 
Ah eee tte Oe Ue [RS 2.8 ea — AD, See oa se ar eae 2-200 5% leo ee 
Die ats eee es ae Se hoyer WAS ee Pe ee 2.080*** .994 
Ook on ee 1.869*** he A Ov ks seek rene meet oe 1.418 1.987*** 
oi eters a eR 1538 I sia}) AY co Beene eas 2.449% Mes Ghat 
1 hie Xe i Saba 1362* Sys GOR Seale, Mees. ns ewe 16207 Gaeta 
1 Oy bee ee 1.3'82** 927 D0) 2 a eee PralsyPei2 AAV 
TRIM BEES Wes test os 32s PS68t** Poste Dy peewee a toe Ae 198) 1 15655" les 0% 
| y2t Delano ee ee nn 1-7 00*** ee D2 ee ee 1.085 fe 
1S} 5 Same ig Ne coe 1.604*** Ae DST 1.371 1.297 
dl Gtpesieeee es eee E05 * ee eae eee ee ee 1.894 44% IEG 84% 
le /ggeecce woe Stew oes 1.964* 2.440*** Ties Se 1.178 Pep dX oh F< 
Ih igeemeeetremarmee ses ce GO SA eS 4.0% eee renee arte 831 Dsieyey 
NOME ce sete aetna 1.163 RO 26a Dae Sees ae nee 1.614*** pa 
Aes Sere eee tes 1.256 653** Se EEE 1:865*** 1524 eee 
2D eee ee ie SE 1.298* DONT SA 60 ese eee 342 1.044 
AGS Uae i eee DDS = Oli eevee 840 2.503 
De pera ease oe zone eae eee Os Ae Teen ee 1.462 3.248*** 
TAS ge OO RO eae Re E845 * 2716 [aos Brea eee 1.620 646 
ZO enna ee x Ses 22004 ** 16192 *% TI Siapeciee tere or s We Aece 2.492*** 2°636"%% 
1 iy CRN ne SMe Sites eS 2 aA eee Bape Bee oe. 1.450** 1.907*= 
DO) See, eae a TE 1S O08*** 92235 45% fi ie ee SB 1.275 = 
30 poy oe ae ee oe 1.083 182955 
Theo eee eee 1.289* 13 05 
Lint hen ee AB Oe t= EOS 
Sc eee 2 Soe Plies 7.40% 


Explanation: Starred mean squares indicate significant heterogeneity among vials within any 


one density, (* = 001 @P 01053 ** = 000 <P 00 PaO) 


already been discussed in Section III. The tests for deviations from linearity 


are largely non-significant. The mean square among vials within densities — 


(shown in Table 2) corresponds theoretically to the mean square of family 
means around regression in the analysis of covariance. The actual values for 
these two types of mean squares show no essential difference. In the non- 
replicated sibships we cannot partition the vials mean square into environmen- 
tal and genetic components, but can test the mean square for heterogeneity, 
i.e. Whether it is greater than its expected value of 1.0. Most generations in 
Table 2 are highly heterogeneous and thus presumably possess an added, 
largely genetic component among families. There is no consistent difference 
in magnitude of these mean squares between the CP and PP lines. 


Puration Site Dirrerences 1x Drosophila melanogaster 707 


5. Relaxation of Selection 


Removal of selection pressure from generation 32 to 40 resulted in a de- 
crease of the difference in pupation site between the lines. By generation 40, 
d was reduced to the low value of 0.436. However, selection in generations 40 
and 51 to 55 increased d to values near 1.0. During a second period of relax- 
ation of selection (generation 56 to 75), each line was kept in two half-pint 
bottles with from 100 to 300 flies per bottle. The considerably more rigorous 
conditions of existence in these stock bottles should have hastened any genetic 
changes away from the selected genotype, since the specialized PP and CP 
flies were likely at a disadvantage in the struggle for survival in the bottle. By 
generation 75, d had declined to a minimal value of 0.079. The difference after 
yet another generation of relaxed selection was larger again (0.462) but still 
quite small. The high d value of 2.039 for generation 60 is presumably a case 
of gene-environment interaction, the two lines having responded differentially 
to the peripheral stimulus of the medium. 

Selection applied for six generations starting with generation 76 was im- 
mediately effective in restoring the difference between the lines to a very high 
level (d = 1.919 in generation 80). The reversion of the differentiated lines 
to their original state in the absence of selection, coupled with the speedy 
re-establishment of the differences upon resumption of selection are substanti- 
ated by Schlager (1959) who found a minimal difference (d = 0.13) in these 
same lines in generation 114 after 33 generations without directed selection. 
By selection of sublines from generation 108, he was able to re-establish an 
appreciable difference (d = 1.45) in a few generations. 


6. Fertility and Larval Survival 


The per cent survival to the pupal stage was obtained as a by-product of 
the pupation site assays. The percentage was computed only among fertile 
sibships, since totally infertile vials would, to a large measure, come from 
unfertilized females. This survival consists of two separate components: fer- 
tility in the strict sense (i.e., hatchability of eggs) and survival of larvae to 
pupation. Separation of these two components was possible only in gener- 
ations 11, 15, 60 and 61, when separate readings of the number of eggs hatched 
Were taken. On the average, half the loss from egg to pupa represents in- 
fertility of eggs and half larval mortality, but the CP line has significantly 
more egg infertility than larval mortality while the converse relation holds 
for the PP line. 

Figure 5 shows the joint variable (per cent survival to the pupal stage) for 
the entire study. A marked decline (significant at P < 0.01 by means of an 
ordering test, Quenouille, 1952) for the first 20 generations in the CP and the 
first 18 generations in the PP line corresponds well with the period of effec- 
tiveness of selection. Between approximately generation 22 and generation 32 


708 Tue Universiry ScrENCE BULLETIN 


in 100 
oO 
a 
o ; x 
4 \ A Ay 
e xe\/ \a x pe 
5 xi |\% vi V \ 
lu e i y | 
aos 
Ee ° 
e ' 
= x 
S x——x PP 
> o—e CP 
[em 
a 
S10 
AR A OR RO LR PO ge abe Utttit 
ee ee SS ed 
10 20 60 70 80 


40 
GENERATIONS 


Ficure 5. Per cent survival to the pupal stage. For a more detailed explanation of this variable 
see text. Abscissa: generations of selection; ordinate: per cent survival. Means for the PP line are 
connected by broken lines, means for the CP line by solid lines. Only means involving consecutive 
generations are connected. The arrows along the abscissa indicate the generations in which selec- 
tion for pupation site was practiced. A double-headed arrow signifies bidirectional selection, an 
arrow pointing up selection for peripherality, one pointing down for centrality. 


considerable recovery in survival appears to have taken place. In subsequent 
generations the variable fluctuated markedly. 

Significant parallelism occurs in the fluctuations of the two lines (P of the 
association = 0.05 by the medial test and < 0.01 by Tukey’s corner test). This 
parallelism must be caused by environmental factors common to each gener- 
ation and is likely to affect larval survival rather than hatchability of eggs. It 
is unlikely, however, that the decline in fertility observed in the first 20 gener- 
ations is entirely due to environmental trends. No similar long term trends in 
either hatchability or larval survival were ever observed in other work in our 
laboratory and the coincidence between the effective period of selection and 
the decline in fertility is too great to be coincidental. Attempts to relate the 
variable to selection differential, phenotypic change or fecundity were not 
successful. 

The variable described above is the same as “density” of pupating larvae. 
Thus, a density of 8.7 pupae per vial corresponds to a survival of 87 per cent. 
Two separate terms were retained for the same variable to emphasize that 
“survival” is a correlated response to selection, while “density” of pupating 
larvae (a consequence of “fertility” and “survival of larvae”) affects the ex- 
pression of the phenotype. 


7. Other Correlated Characters 

DDT resistance. Table 3 lists the LDso’s in p.p.m. DDT for the various 
generations assayed during the experiment. The higher resistance of the PP 
is established beyond reasonable doubt (P < 0.006 on hypothesis of equal re- 
sistance for the two lines). Thus, the correlated divergence in pupation site 
found when selecting for differences in resistance (Sokal and Hunter, 1954; 


| 
| 
| 
| 


Pupation Site Dirrerences iN Drosophila melanogaster 709 


Sokal, 19592) is matched by the reciprocal correlation. It should be noted, 
however, that the differences in resistance in this experiment are less than 
those found in experiments where selection for resistance was practiced. 

Length of larval period. Records for generations 11, 15, 60 and 61, suggest 
that PP flies have a longer larval period but the difference remains not proven 
(P of overall difference is between 5 and 10°% by two separate tests). 

Morphological characters. In a morphometric analysis of a number of 
drosophila strains Sokal (1959) included 14 characters of the two lines at 
various generations. CP appears to be a generally bigger line of flies. Con- 
sistency of the relations found is quite marked. 

Ethological observations. The difference between the lines are expressed 
quite early during larval life. CP and PP eggs of generation 10 were placed 
in two series of ten vials each at densities of 2 and 20 eggs per vial, respectively. 
The number of larvae visible on the surface of the medium cylinder was 
recorded at 24-hour intervals. Figure 6 shows the results of these observations. 
The general migration away from the surface before hour 72 characteristic of 
the COSU-2 strain (Sokal et al., 1960) is seen in the selected lines as well. 
However, the PP line consistently has more larvae on the surface until hour 96 
when all larvae are burrowing below the surface. These results were born out 
by a second study at generation 20. The preference of the PP larvae for the 
surface was coupled with a generally greater irritability and mobility. 

Sokal et al. (1960) were able to show that the first larvae to pupate in a 
vial tended to do so peripherally, while later pupation tended toward the cen- 
ter of the vial. In the present study (generations 11 and 15) the tendency for 


Tas_e 3. Resistance Assays of Larvae 


Relative 

Resistance 

Generation LDs0 LDso LDso CP 
A eee aT ie Nie 3.33 4.32 1.3 
00) inches SS ee ne .. 241 3/5 1.6 
Up e cee re Ae Be teed, (4.0) 5.03 1.3 
14) 7 eee EO ee ee a ee = 3.98 2) 1.8 
DS) acacia Ae eae eet (2.4) 4.78 2.0 
BIL acco a eee eee ee (4.9) (5.9) 1.2 
Bs ae ee ee 3.40 (5.3) 1.6 
GAN Noose eae a er 4.82 (5.3) 1a 
AG). See Sea eee Osage net Tire (2.0) (3.1) 1.6 
Ny MNO eee Oo a Be ete (Oa) (6.4) eS 
(7) | eae re SS A aed Be 3.18 3.87 1e2 


LDso’s are given as p.p.m. DDT in the larval medium. Values in parentheses are eyefit esti- 
mates used in cases where the data were too heterogeneous to yield an estimate of the LDso by 


the maximum likelihood method. 


710 Tue Universiry ScrENCE BULLETIN 


100 
CP e—e 
2d eae, <3 
80 
N Ww 
60 ~S oS 
fn) ES 


SSS 
< 20 eggs \ “ 


40 per vial 
\ \ 


‘ 
2 eggs .” 
per vial a 


20 


Percent of larvae on surface of medium 


24 48 72 96 
hours after vial preparation 


Ficure 6. Locations of larvae between hatching and pupation. Abscissa: hours after vial 
preparation; ordinate: per cent of larvae on surface of medium. The graph summarizes experi- 
ments at two densities, 2 per vial and 20 per vial. Larvae not on the surface are found in the 
medium or between wall of vial and medium cylinder. 


early larvae to pupate peripherally is present in both lines and the difference 
between the lines is maintained in both “early” and “late” pupae. 


V. RESULTS OF CROSSING EXPERIMENTS 


Four crossing experiments were undertaken in generations 16, 29, 45 and 
48, respectively. Pupation site assays are based on from 35 to 100 vials (most 
in excess of 50 per test). The regular assays for the selected lines, run concur- 
rently on the same batch of medium, served as standards of comparison for 
the various crosses since they were all prepared at the same time. For those 
populations where all four crosses were assayed, Friedman’s two-way analysis 
of variance yielded X? = 9.3 and 11.1 (x?.01 at 3 df. 9.21). We may con- 
clude that the differences among crosses were highly significant. Figure 7 
shows the results of the crosses expressed as deviations on a scale where the 
PP-CP distance equals one, and averaged over the available replicates. 

While crosses of the selected lines were generally intermediate between the 
parental means, several hybrid assays exceeded the phenotypic limits of the 


PupatIon Sire Dirrerences 1N Drosophila melanogaster 711 


oO 
2 2.0 
= — 
= 18 
5 16 
E a 
< 1.4 
> 
Ww 1.2 ——_ 
5 1.0 pp 
3 S a == 
fe) 
= — 
ea 
9 as 
o 4 an — 
joa 
a 2 a 
a 
i) cP 
> 
= -2 — — 
=| 
© _ 
= (CP9xPPo') =B BgxBot BoxPPd AgxPPd Agx Bo% 
Zs (PPgx cd) =A Agxad PPgxadt PP9xBo" Box Ad 
a CP9xBo* CPoxA& 
AgxCPe Box cP& 
HOMOLOGOUS HETEROLOGOUS HETEROLOGOUS 
FS 5 BACKCROSSES BACKCROSSES Fy 


Ficure 7. Graphic representation of the results of crosses between lines. The ordinate repre- 
sents an arbitrary scale expressing the deviation PP-CP as 1.0. This is done because fluctuations 
from generation to generation made the deviation between these two lines in terms of probit units 
quite variable. The horizontal lines in the graph represent the average value for a given cross 
over all its replicates expressed as a proportion of the distance PP-CP. The means of some crosses 
show higher or lower peripheral pupation than do the PP and CP lines. Labels at the bottom of 
the graph identify the crosses and are ordered in the same way as the horizontal lines indicating 
the means. Thus, among the Fs crosses the upper line represents B2: & Bé, while the lower 
horizontal line represents AQ «K Ag. 


parents. The position of the hybrid means relative to the midpoint between 
parental means appears influenced by the general phenohypic level of the 
given experiment, another example of the environmental fluctuation and 
gene-environment interaction in this study. The backcrosses resembled the 
type to which they were backcrossed. Of most interest is an apparent “pa- 
ternal” influence. Hybrid offspring are more like their fathers than their 
mothers, and backcrosses to male pure line flies are more effective than back- 
crosses to females. 


VII. DISCUSSION 
1. Pupation Site as a Heritable Trait 


The successful selection for differential pupation sites demonstrates again 
that few if any phenotypic characters of organisms resist modification under 
‘selection. Without careful analysis of any given case it is impossible to ascer- 
tain what specific structures or physiological processes of the organisms have 
been modified to produce results such as are shown in this study. 


72 Tue University ScrENcCE BULLETIN 


Humidity content of the medium and of the air surrounding the vials 


plays an important role in determining pupation site (Sokal et al., 1960), al- 


though age of the medium as such (which would be expected to have drying 
effects) did not produce changes in pupation site. If pupation site were a 
reaction to the moisture gradient from the drier periphery to the moister 


center, selection as practiced in this experiment would essentially constitute a | 
shift in the humidity preferendum of the pupating larvae. Since differences | 


in pupation site behavior between the two lines are reflected in earlier be- 
havior of the larvae it is likely, however, that the relations are more com- 
plicated. 


2. Limits to Selection 


The asymmetry of the response observed in this experiment could be | 


produced by different rates of response in the two lines as well as by asym- | 
metrical selection limits. Means for the PP line changed two and one- half | 


times as fast as those of the CP line during the period when apparent selection 


progress was being made. While in part this is due to higher selection 
intensities in the PP line, these are not sufficient to account for the differences 
in slope. 

Selection limits were apparently reached in both directions. These are 
roughly symmetrical around the probit value of 5.0, corresponding to 50 per 
cent peripheral pupation, but since the control strain from which the two lines 
originated and other control strains tested at different times have a normal 
pupation site of less than 50 per cent, i.e., a probit peripheral value of less than 
5.0, the progress of the lines is asymmetrical. There is some evidence for ex- 
haustion of the genetic variability of the lines in the lowering of the genetic 
variances after the selection limit had been reached (Table 1). Inter-vial 
variances in the non-replicated sibships (Table 2) are not similarly lowered, 
presumably owing to the weak genetic contribution to inter-vial variance. 
The lowering of genetic variability should be restricted to those loci concerned 
with pupation site, since the sample sizes employed in this experiment would 
have precluded more than a very moderate amount of inbreeding. 

While there is no reason to assume a physiological limit to the expression 
of pupation site, some asymmetry is induced by various environmental con- 
ditions. For example, Sokal et al. (1960) found that increasing the water 
content of the medium raised the probit peripheral pupation to approximately 
7.0, equivalent to approximately 98 per cent peripheral pupation, while low 

values of probit peripheral pupation rarely went below 3.5, equivalent to 9 per 
cent peripheral pupation. Could it be that the peripheral pupation of the 
earliest pupating larvae (Sokal et al., 1960) is difficult to change while the 
reactions of the later pupating larvae can be modified by selection? 


Pupation Site DirFerences IN Drosophila melanogaster 73 


3. The Crosses 


A chromosome assay was undertaken to determine the separate contribu- 
tions of each of the chromosomes and in an attempt to explain the results of 
the crosses. The results obtained suggested that the paternal effect of the back 
crosses may be largely due to overdominance of the X-chromosome in the PP 
line. However, variability among replicates was such that statistical signif- 
icance of the effects of the assay could not be established. It does not, therefore, 
seem profitable to pursue the analysis in detail. 


4. Gene Environment Interaction 


The marked fluctuations of pupation site appear due to minor chemical 
differences in the medium which are largely removed when syrup and mo- 
lasses were replaced by sucrose as a source of sugar (Schlager, 1960). When 
the environment increases peripheral pupation, the deviation in pupation site 
of the CP and PP tends to increase (e.g., generations 24 and 60). Conversely, 
environmentally caused decreases in peripheral pupation (as in generation 23) 
result in a diminution of the differences between the selected lines. Schlager 
(1959) found that the peripherality of the PP line (and of some control 
strains) responded linearly to increases in water content of the medium, while 
the CP line showed no such increase. Thus, CP appears to maintain its cen- 
trality even under unfavorable environmental conditions. Apparently the PP 
line did not develop the converse mechanism for protecting itself against 
centrality in medium containing less water. 

Environmentally caused phenotypic fluctuations of this magnitude have 
interesting implications for evolutionary theory. If natural selection in the 
field is based on phenotype, then with the effective low heritability the results 
of selection may be largely due to chance, resulting in a situation resembling 


drift. 


5. Correlated Characters 


Fertility as measured by survival to pupation decreased for approximately 
the first 20 generations, concomitant with the effective selection in the study. 
This is comparable to findings in other Drosophila selection studies such as 
those of Mather and Harrison (1949), Robertson and Reeve (1952) or Pre- 
vosti (1956, 1958). An interesting aspect of the present experiment is that in 
spite of the negative relations between pupation site and density, the periph- 
eral pupating strain did not show lower survival as might have been supposed. 

Selection for pupation site differences produced differences in resistance. 
This is a generally valid observation, having been found in other strains dif- 
fering in pupation site in this laboratory (unpublished work). Similarly, 
strains differing in resistance show differences in pupation site (Sokal and 
Hunter, 1954; and other studies on lines developed in our laboratory and in 
other laboratories). The nature of this correlation is not clear. 


714 Tue University ScIENCE BULLETIN 


The slight evidence that the PP line has a longer larval period than the 
CP line would be in agreement with the observations of Hunter (1956) who 
found strains with longer larval periods to be more resistant than those with 
shorter periods. However, the CP strain is larger in size than the PP strain. 
Since the CP larvae buried into the medium considerably earlier than the PP 
larvae and the latter apparently had a higher metabolic rate, it may be con- 
jectured that the CP larvae are larger since they feed more and earlier and do 
not use up as much food for energy as do the PP larvae. 


VII. SUMMARY 


Drosophila melanogaster reared in shell vials containing cornmeal- 
molasses-agar medium will pupate at various sites from the center of the 
medium surface to the wall of the vial above the medium. Ten eggs were 
placed in each vial. Since mortality of eggs and larvae varied and decreased 
density raised the percentage of peripheral pupation, pupation site readings 
were adjusted to a constant density of 10 pupae per vial. The variable meas- 
ured was transformed to a probit scale. Each generation was based on a large 
sample size to avoid inbreeding. 

Selection for differences in pupation site resulted in a peripherally pu- 
pating (PP) and a centrally pupating (CP) line. The PP line responded 
immediately and strikingly with a plateau reached and maintained by gener- 
ation 18, while the CP line reacted less sharply but showed selection progress 
until generation 23. The differences declined during periods of relaxed selec- 
tion but were quickly re-established when selection was renewed. 

Fluctuations in the expression of the character were caused by chemical 
instability of the larval medium. Gene-environmental interaction made these 
fluctuations deviate from parallelism. 

Fertility and larval survival declined during selection but rose again after 
the selection limits had been reached. Other characters examined for possible 
correlation with pupation site were resistance of larvae to DDT in the me- 
dium (the PP line is considerably more resistant than the CP line); length 
of developmental period (the PP possibly have a longer larval period); 14 
morphological characters (which showed that on the whole the CP flies were 
larger than the PP flies); and sex ratio (not significantly different from 50:50 
in both strains). The PP larvae took considerably longer to burrow into the 
medium than the CP larvae. 

Crosses between selected lines give intermediate values, but a strong pa- 
ternal effect was noted in all types of crosses. 


ACKNOWLEDGMENTS 


I am grateful for the assistance of a number of persons during the course of this study. Mrs. 
Dorothy Hall, Mrs. Maxine L. Howe, Paul R. Ehrlich, and Preston E. Hunter helped with the 
technical work. Jacqueline Johnson, F. James Rohlf, and Mrs. Julie Sokal helped with the com- 
putations. 


Pupation Site DirFereNces 1N Drosophila melanogaster 715 


LITERATURE CITED 


Fatconer, D. S. 1960. Introduction to quantitative genetics. Oliver and Boyd, Edinburgh and 
London. 365 pp. 


Finney, D. J. 1952. Probit analysis. Cambridge University Press. 318 pp. 


Gowen, J. W., anv L. E. Jonnson. 1946. On the mechanism of heterosis I. Metabolic capacity 
of different races of Drosophila melangaster for egg production. Amer. Naturalist. 80:149- 


179: 


Hunter, P. E. 1956. Length of larval period related to DDT-resistance in Drosophila. J. Econ. 
Entomol. 49:671-675. 


. 1959. Selection of Drosophila melanogaster for length of larval period. Z. Vererbungsl. 


90:7-28. 
Matuer, K., anp B. J. Harrison. 1949. The manifold effect of selection. Heredity 3:1-52, 
131-162. 


Neunes, H. W. 1962. Controlling fluctuations of duration of developmental period in Droso- 
phila. J. Kansas Entomol. Soc. 35:269-278. 


Prevosti, A. 1956. Resultados de la selection por alas largas y por alas cortas en dos poblaciones 
naturales de Drosophila subobscura. Genet. Iber. 8:73-115. 


——.. 1958. Respuestas correlationadas al ‘seleccionar por la longitud del ala en Drosophila 
subobscura. Genet. Iber. 10:1-55. 


QuenouiLLe, M.H. 1952. Associated measurements. Butterworths Scient. Publ., London. 242 p. 

Rogertson, F. W., ann E. Reeve. 1952. Studies in quantitative inheritance 1. The effects of 
selection of wing and thorax length in Drosophila melanogaster. J. Genet. 50:414-448. 

ScuLaceR, G. 1959. Fluctuations and gene-environment interaction of pupation site in Droso- 

phila melanogaster. M. A. Thesis. Univ. of Kansas. 51 pp. 

. 1960. Controlling fluctuations of pupation site in Drosophila, Trans. Kansas Acad. Sci. 

63:243-257. 

StecEL, S. 1956. Non-parametric statistics for the behavioral sciences. McGraw-Hill, New York, 
Toronto and London. 312 pp. 

Soxar, R. R. 1958. Probit analysis on a digital computer. J. Econ. Entomol. 51:738-739. 

- 1959. A morphometric analysis of strains of Drosophila melanogaster differing in DDT- 

resistance. J. Kansas Entomol. Soc. 32:155-172. 

——.. 1959a. Laboratory induction of changes in susceptibility levels of insects. W. H. O. Pan- 
Amer. Sanit. Bur. Rept. Seminar Suscept. of Insects to Insecticides. Panama, R. P. (26-28 
June 1958) pp. 216-225. 

SoxaL, R. R., P. R. Exnrticn, P. E. Hunter, AND G. SCHLAGER. 1960. Some factors affecting 
pupation site of Drosophila. Ann. Entomol. Soc. Amer. 53:174-182. 


SoxaL, R. R., anp P. E. Hunter. 1954. Reciprocal selection for correlated quantitative char- 
acters in Drosophila. Science 119:649-651. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


THE CLASSIFICATION OF THE 
DIPHAGLOSSINAE AND NORTH AMERICAN 
SPECIES OF THE GENUS CAUPOLICANA 
(HYMENOPTERA, COLLETIDAE) 


By 


Charles D. Michener 


‘ 
: 


VoL. XLVI Paces 717-751 DECEMBER 2, 1966 No. 20 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Vo. XLVI Paces 717-751 DECEMBER 2, 1966 No. 20 


The Classification of the Diphaglossinae and 
North American Species of the Genus Caupolicana 
(Hymenoptera, Colletidae)’ 


By 


Cuarves D. MIcHENER 


ABSTRACT 


This paper describes and provides keys for the tribes of the apoid subfamily 
Diphaglossinae. One of the tribes, the Caupolicanini, is treated in greater detail, its 
genera being characterized. The subgenera of one of the genera, Caupolicana, are 
characterized, as are the North and Central American species of the genera Caupo- 
licana and Crawfordapis. Two new subgenera, Caupolicanoides and Alayoapis, 
and one new species, Caupolicana ocellata, are described. Crawfordapis is given 
generic rank while Z:kanapis is reduced to a subgenus of Caupolicana. 

Most species of the Caupolicanini are active primarily in the early morning and 
sometimes late evening hours. Some have enlarged ocelli, presumably related to 
such a time of flight. 

This paper consists of (A) brief descriptions of the tribes of the colletid sub- 
family Diphaglossinae, (B) descriptions of the genera of the tribe Caupolicanini, 
(C) a subgeneric classification of the genus Caupolicana, and (D) an account of 
the North American (including Central American and Antillian) species of that 
genus and Crawfordapis. 


A. THE TRIBES OF DIPHAGLOSSINAE 


The Diphaglossinae is a group of large to very large colletid bees limited 
to the Western Hemisphere. It contains the Diphaglossinae plus the Caupoli- 
canini of Michener (1944); Moure (1945) has united these groups, a view- 
point in which I have agreed (e.g., Michener, 1954). Moure (1945, 1953) 
divides the subfamily into tribes which may be separated as follows: 


*Contribution number 1321 from the Department of Entomology, The University of Kansas, 
Lawrence, Kansas. 


718 Tue Universiry SciENCE BULLETIN 


Key TO THE TRIBES OF DIPHAGLOSSINAE 


1. Pre-episternal groove complete; first flagellar segment much longer than 


others, petiolate, nearly as long as to longer than scapes = Caupolicanint 
—Pre-episternal groove absent below scrobal groove; first flagellar segment 
not greatly longer than others, not petiolate, much shorter than scape. 2 
2. Notauli represented by deep grooves in anterior part of mesoscutum; malar 
space over one third as long as eye. +--+ Diphaglossini 
— Notauli weak or absent; malar space short or absent. -.....--—--------—-— 3 
Se Aroliial presen ts sec Mydrosomini 
Se yNrolinvabsents eet) oe bes ee ee Ptiloglossidiini 


TRIBE CAUPOLICANINI 


1. Lower part of face short, malar space short or absent. 2. First flagellar 
segment much longer than other segments of flagellum, nearly as long as to 


longer than scape, petiolate. 3. Notauli strong. 4. Pre-episternal groove com- 


plete. 5. Jugal lobe of hind wing over three fourths as long as vannal lobe and 
extending beyond cu-v. 6. Second submarginal cell much shorter than 1 or 3; 
first recurrent vein approximately interstitial with first transverse cubital. 7. 
Second recurrent vein more or less continued in same direction as Cur. 8. Dis- 
tal parts of wings hairless, strongly papillate, but papillae often ending in 
slender hairlike points. 9. Arolia present. 

This is the most distinctive tribe of the Diphaglossinae, as indicated by 
characters 2, 4, 5, 6, 7, and 8 above. 

Genera included in the Caupolicanini are Caupolicana, Puloglossa, and 
Crawfordapis. They range from North Carolina, Kansas, and Arizona to 
Argentina and Chile. 


TripeE DrIrPHaGLossINI 


1. Lower part of face elongate, malar space over one third as long as eye. 
2. First flagellar segment not greatly longer than others, not as long as scape, 
not petiolate. 3. Notauli strong. 4. Pre-episternal groove absent below scrobal 
groove. 5. Jugal lobe of hind wing less than half as long as vannal lobe and 
not reaching level of cu-v. 6. Submarginal cells decreasing in length from | to 
3, rarely 2 and 3 equal; first recurrent vein entering second submarginal cell 
more or less medially. 7. Second recurrent vein at a distinct angle to Cur. 8. 
Distal parts of wings with hairs, not strongly papillate. 9. Arolia present. 

This tribe is related to the following tribes as shown by characters 2, 4 
6, 7, and 8; however, it also shows some relations to the Caupolicanini, indi- 
cated especially by its similarly large and robust form and by character 3. 

Genera included in the Diphaglossini are Diphaglossa, Cadeguala, and 
Policana, all of which are Chilean. 


| THE CLassIFICATION OF THE DipHacGLossINAk 719 


TripE Myprosomini 


I. Lower part of face short, malar space short or absent. 2. First flagellar 
segment about as long as apical one and less than half as long as scape (fe- 
male) or much shorter than any others and under one fourth as long as scape 
(male), not petiolate. 3. Notauli absent. 4. Pre-episternal groove absent below 
-scrobal groove. 5. Jugal lobe of hind wing about half as long as vannal lobe 
and not reaching level of cu-y. 6. Submarginal cells decreasing in length from 

1 to 3 or 2 and 3 about equal; first recurrent vein entering second submarginal 
‘cell at base or in basal third. 7. Second recurrent vein at a distinct angle to 
Cur. 8. Distal parts of wing with hairs, not strongly papillate. 9. Arolia 
present. 

Genera included in the Mydrosomini (Moure’s Dissoglottini) are Mydro- 
soma (= Apista = Egapista = Dissoglotta) and Bicornelia. They occur in 

tropical America from Mexico to Brazil. 


TriBe PrirocLossipiNtr 


I have not seen specimens of this tribe: the following information, organ- 
ized to parallel that for the other tribes, is from Moure (1953), who gave an 
excellent account of the group. 

1. Lower part of face short, malar space short. 2. First flagellar segment 
considerably shorter than scape, shorter than (male) or equal to (female) 
flagellar segments 2 + 3, not petiolate. 3. Notauli practically absent. 4. Pre- 
episternal groove absent below scrobal groove. 5. Jugal lobe of hind wing over 
half as long as vannal lobe and nearly reaching level of cu-v. 6. Second sub- 
marginal cell shorter than 1 or 3; first recurrent vein entering second sub- 
marginal cell almost medially. 8. Distal parts of wings with hairs, not strongly 
oapillate. 9. Arolia absent. 

The only genus of Ptiloglossidiini is Puloglossidia from Argentina. I sus- 
ect that it should be included in the Mydrosomini but not having seen speci- 
nens, I retain the Ptiloglossidiini for the present. Character 9 is the only 
trikingly distinctive feature. 


B. THE GENERA OF CAUPOLICANINI 


The Caupolicanini is divisible into two large genera and one small one, 
“rawfordapis, as shown below. 


KEY To THE GENERA OF CAUPOLICANINI 


| 
- Outer hind tibial spur of male immovably fused to tibia; hind basitarsus of 
female less than twice as long as broad, second hind tarsal segment broader 
than long; metasomal terga usually weakly metallic bluish or greenish, 


(| iotnothat:tcorp ce a a ene PORE aR TERE ore we ey fp 8N Ptiloglossa 


720 Tue University ScrENcE BULLETIN 


— Outer hind tibial spur of male articulated at base; hind basitarsus of female 
more than twice as long as broad, second hind tarsal segment longer than 


broad; metasomal terga usually nonmetallic. 2 
2. Seventh sternum of male with no paired apical lobes; base of marginal cell 
prolonged as narrow sinus to apex of stigma, —......-------.- Craw fordapts 
— Seventh sternum of male with paired apical lobes; base of marginal cell not 
prolonged) as MalOW SIQUS. 222. ee Caupolicana 


Genus Ptiloglossa Smith, 1853 


Ptiloglossa (including the subgenus Priloglossodes Moure, 1945) can be 
distinguished from the rest by the following combination of characters: 

1. Clypeus clearly elevated above level of adjacent parts of face, extending 
high on face so that in female clypeoantennal distance is usually distinctly less 


than diameter of antennal socket. 2. Stigma often only half as long as pre- 
stigma, sometimes ( Ptiloglossodes ) two thirds as long as prestigma; marginal — 
cell, except in Ptiloglossodes, prolonged basally as narrow sinus to apex of | 


stigma. 3. Hind basitarsus of female less than twice as long as broad. Second 


and third hind tarsal segments of female broadly expanded above, each, or at | 


least second, broader than long. 4. Outer hind tibial spur of male immovably 
fixed to tibia, its broad base gradually expanding onto the tibia so that sharp 


line between spur and tibia does not exist (outer spur entirely absent in male | 


Ptiloglossodes). 5. Abdominal terga with metallic bluish or greenish tints (ex- 
cept in one Mexican species, P. wilmattae Cockerell). 6. Lateral extremities of 
fourth and fifth terga of male extending ventrally, usually more than other 
terga, broadly overlapped by sterna of preceding segments, and densely cov- 
ered by short erect hair of uniform length, these areas under low magnifi- 
cation appearing dull and scarcely punctate in contrast to adjacent areas. Ti 
Sixth sternum of male with posterior margin thick, thickest medially where 
there is usually a weak mid-apical angle, a sulcus at each side in margin, 
facing posteriorly; carina which delimits lower margin of sulcus projecting 
ventrally at each side as thorn-like spine (except in Péloglossodes). 8. Seventh 
sternum of male with two pairs of apical lobes, both large, broad distally, 
narrowed basally, the upper pair long pedunculate with a process arising near 
middle of peduncle (for ventral views, see Michener, 1954). 9. Eighth sternum 
of male with apical process translucent light brown, strongly curved down 
apically, with long hairs on dorsal surface. 10. Gonoforceps ending bluntly. 

Ptiloglossa is a primarily tropical genus, although a few species reach 
southern Texas and Arizona in the north and Argentina in the south. The 
genus does not occur in Chile. 


Genus Crawfordapis Moure, 1964 


The name Crawfordapis was originally proposed as a subgenus of Zr- 
kanapis but the single Middle American species seems so distinctive that 


t 


| 


l 
’ 


THE CLASSIFICATION OF THE DIPHAGLOSSINAE 72M 


generic rank is warranted for it, even though Zzkanapis is here considered a 
subgenus of Caupolicana, Crawfordapts is especially distinctive in characters 
8 and 10 described below. Characters 1 and 2 are Ptiloglossa-like as is part of 
3 (expanded second and third tarsal segments of female), and in general, 5. 
Characters 4, 6, 7, and 9 are more or less Cauwpolicana-like. As shown in the 
illustrations, there are various other distinctive features of the genitalia and 
sterna but the external features are very much like those of Caupolicana. 

1. Clypeus clearly elevated above level of adjacent parts of face, as in Prilo- 
glossa but more strikingly so. Clypeus extending rather high on face, nearly 
as in Prloglossa. 2. Stigma less than half as long as prestigma; marginal cell 
prolonged basally as a narrow sinus to apex of stigma. 3. Hind basitarsus of 
female more than twice as long as broad. Second and third hind tarsal seg- 
ments of female considerably expanded above but longer than broad. 4. Outer 
hind tibial spur of male normal, articulated at base like inner spur. 5. Abdo- 
men without or with exceedingly weak bluish or greenish tints in male, with 
distinct blue tints in female. 6. Lateral extremities of terga of male without 
areas of short, dense, erect hair. 7. Sixth tergum of male with posterior margin 
not thickened or sulcate, medially produced and slightly bifid as in Zikanaprs. 
8. Seventh sternum of male without paired lobes but with median, apical, 
slender hairy process. 9. Eighth sternum of male with apical process rather 
heavily pigmented, not downcurved, hairs of distal half shorter than width of 
process. 10. Gonoforceps ending in slender styluslike structures, therefore per- 
haps with distinguishable gonocoxites and gonostyli (Figs. 1-5). 


Genus Caupolicana Spinola, 1851 


This genus is here interpreted in a broad sense to include the species placed 
in recent years in Zikanapis and Willinkapis. Zikanapis in particular is, how- 
ever, a distinctive group which may well merit generic recognition, as is 
Caupolicanoides. The following characters separate Caupolicana from Ptilo- 
glossa. 

1. Clypeus less elevated above adjacent parts of face than in Ptiloglossa, pro- 
file of supraclypeal area often a continuation of that of clypeus. Clypeus not 
extending so high on face as in Ptiloglossa, so that in female clypeoantennal 
distance is at least as great as diameter of antennal socket. 2. Stigma slightly 
more than half as long as prestigma to as long as prestigma; base of marginal 
cell not prolonged as a narrow sinus to apex of stigma. 3. Hind basitarsus of 

female more than twice as long as broad. Second and third hind tarsal seg- 
ments of female but little expanded above, longer than broad. 4. Outer hind 
tibial spur of male normal, articulated at base like inner spur. 5. Abdomen 
without bluish or greenish tints (except in Willinkapis). 6. Lateral extremities 
of terga of male without areas of short, dense, erect hair of uniform length 


722 Tue UNiversiry ScIENCE BULLETIN 


Fics. 1-5, Crawfordapis luctuosa, male (specimen from the type series of crawfordi from 
Costa Rica). 1, genitalia; 2, lateral view of apical part of genitalia; 3, eight metasomal sternum; 


4, lateral view of apical process of eighth metasomal sternum; 5, seventh metasomal sternum. 


Fics. 6-9, Caupolicana (Caupolicanoides) pubescens, male (specimen from Concepcion, 
Chile, labeled as “typus” of herbsti by Friese, USNM). 6, genitalia; 7, eighth metasomal ster- 
num; 8, lateral view of apical process of eighth metasomal sternum; 9, seventh metasomal 
sternum. 

The genitalia and sterna are illustrated with the dorsal views at the left, ventral views at 
the right. The lateral views of genitalia and of the process of the eighth sternum are with the 
dorsal side at the right. The scale lines for each species represent 1 mm. 


except in Zikanapis, where such areas are more extensive than in Ptiloglossa. 
7. Sixth sternum of male with posterior margin not thickened or sulcate, if 
somewhat thickened, not thickest in middle; no thornlike spines. 8. Seventh 
sternum of male with upper pair of apical lobes straplike or sometimes 
broadened distally, without a median process on the peduncle except in 
Alayoapis (of course peduncle not recognizable when lobes are straplike, but 
no process in any event). Lower pair of apical lobes not or scarcely narrowed 
basally (except when broadly bifid as in C. yarrow7), often reduced so as to 
be almost unrecognizable or absent. 9. Eighth sternum of male with apical 
process heavily pigmented, not strongly curved down apically, hairs of distal 
half of process shorter than width of process. 10. Gonoforceps ending bluntly. 

Caupolicana is a genus best represented in warm temperate and subtropical 


THE CLASSIFICATION OF THE DIPHAGLOSSINAE 723 


regions of both North and South America but poorly represented or absent in 
the intervening tropical zone. The majority of the described species are from 


Chile. 
C. SUBGENERA OF CAUPOLICANA 


The genus Caupolicana can be divided into several groups, as is indicated 
below. 


Key TO THE SUBGENERA OF CAUPOLICANA 


1. Stigma slightly broader subapically than basally; marginal cell large, little 


over four times as long as wide (Fig. 10). 0... Caupolicanotides 
— Stigma parallel sided or tapering apically; marginal cell slender, five to six 
Peeesras lone as wide (Pigs 12)... oes ee 2 


2. Metasomal terga rather weakly metallic bluish; ventral apical lobe of sev- 
enth sternum of male probably represented by broad, apically rounded, 
laterally directed, heavily sclerotized lateral apical projection which is hair- 
BERGE NCSA ecole eeseect ee Willinkapis 

— Metasoma nonmetallic; ventral apical lobe of seventh sternum not heavily 
sclerotized, not hairless, variable in size and shape but not as above, some- 
SVERES. QL} GEG EM ln ee OO NEON SY Nae PE ind yt 3 

3. Sixth sternum of male with apex rounded, rarely with broad median, V- 
shaped notch but no produced region; lateral extremities of terga 2-4 with- 
out specialized regions; clypeus of male not over 0.76 times as long as 
OGLE. aed ele A Ree nee eR ea TOE cae ae | 4 

—Sixth sternum of male with weak median apical projection which has a 
broad median, V-shaped notch; lateral extremities of terga 2-4 of male with 
large areas of dense short hair of uniform length; clypeus of male about 


mpemes as long as wide. access coo ee eeeeeneeeee seen eee Zikanapis 
4. Inner orbits of male strongly converging above in male, ocellocular distance 
Batcerourth of an ocellar diameter or less. 2.2 Alayoapis 


— Inner orbits not or weakly converging above in male, ocellocular distance 
over one third of an ocellar diameter and usually nearly equal to an ocellar 
ELSI ies Bel nS NEO a yh Caupolicana s. str. 


Subgenus Caupolicana Spinola 
(Figs. 12-29) 


Caupolicana Spinola, 1851, in Gay, Historia fisica y politica de Chile, Zool., 6:212. 

Type species: Caupolicana gay! Spinola, 1851, designated by Sandhouse, Proc. U.S. Nat. Mus., 
92:534. 

Megacilissa Smith, 1853, Catalogue of the hymenopterous insects in the collection of the British 
Museum, 1:123. 

Type species: Megacilissa superba Smith, 1853, = Caupolicana fulvicollis Spinola, 1851 (mono- 
basic). 


1. Inner orbits slightly converging above (subparallel in ocellata, rather 
strongly converging in mystica) ; ocellocular distance usually slightly less than 


724 Tue Universiry ScrENcE BULLETIN 


ocellar diameter (slightly more than in dimidiata, about two thirds of an 


ocellar diameter in mystica because of unusually convergent eyes, less than — 


half an ocellar diameter in ocellata because of enormous ocelli). 2. Clypeus of 
male 0.70 to 0.76 times as long as wide. 3. Scape of male usually less than three 
times as long as wide, about three times as long as wide in adusta, quadri- 
fasciata, and hirsuta. 4. First basitarsus of male straight to slightly arcuate, as 
long as or slightly shorter than remaining tarsal segments together. 5. Hind 
tibia of male straight to arcuate. 6. Basitibial plates of male indicated only on 
posterior margin, or complete in adusta, quadrifasciata, and hirsuta. 7. Hind 
basitarsus of male four to five times as long as wide. 8. All femora of male 
with abundant long hair except middle and hind femora of electa, ocellata, 
and yarrowi which have only sparser, shorter hair; base of middle femur of 
male with area of short, often rufescent hair (this area absent in ocellata, 


small in yarrowi, but unusually large in electa). Front femur variable in 


thickness. 9. Stigma parallel sided or slightly narrower apically than basally, 
discal area narrower than or as wide as marginal thickenings; first submar- 
ginal cell shorter than to longer than second and third together (first on pos- 
terior side shorter than or equal to second and third together); marginal cell 
narrow, five to six times as long as wide; cell 1st M 2.72 to 3.61 times as long 
as broad; first recurrent vein shorter than or equal to posterior margins of 
second and third submarginal cells together, the latter shorter than or nearly 
equal to first abscissa of vein Cu: (i.e., posterior margin of cell 2nd M). 10. 
Pubescence, at least that of metasoma, partially black; terga with apical bands 
of white hair, sometimes much reduced (albiventris, adusta, electa) or absent 
(dimidiata, female of funebris); sterna two to five of male usually with 
unusually dense long hair (not or scarcely true of North American species). 
11. Metasoma nonmetallic. 12. Terga of male without lateral areas of short 
erect hair. 13. Sixth sternum of male with long hairs or with discal area of 
short hairs; posterior margin usually rounded, sometimes (ocellata, hirsuta) 
notched medially suggesting Zikanapis and Willinkapis; some species with 
a pair of preapical lobes (gayi, dimidiata, vestita, piurensis; fulvicollis has 
weak suggestions of the same lobes); surface hairs reaching posterior margin 
of sternum. 14. Dorsal apical process of seventh sternum moderately robust 
(very slender in yarrowt), not or only weakly spatulate and usually with only 
sparse, short hairs (but with long hairs in quadrifasctata and hirsuta); ventral 
apical process distinctly projecting, hairy, sometimes quite large (largest in 
quadrifasciata and hirsuta). 

The subgenus Cazpolicana proper is the largest unit of the genus. The 
diversity among the species results in forms closely approaching Caupolt- 
canoides and is discussed in connection with that subgenus. 

Included species that I have studied are: adusta Friese, 1899; albiventris 
Friese, 1904=malvacearum Cockerell, 1926; caudens Pérez, 1911; dimidiata 


Tue CLAssIFICATION OF THE DIPHAGLOSSINAE 725 


Herbst, 1917; electa (Cresson, 1878) ; fulvicollis Spinola, 1851; fenebris Smith, 
1879=cana Herbst, 1917 (male not seen); gay: Spinola, 1851; hirsuta Spinola, 
1851; Jugubris Smith, 1879; mendocina Joergensen, 1909 (male not seen); 
mystica Schrottky, 1902; ocellata new species; piurensis Cockerell, 1911; 
quadrifasciata Friese, 1898; ruficollis Friese, 1906; vestita (Smith, 1879) ; 
weyraucht Moure, 1953; yarrow (Cresson, 1875). 

Species of this subgenus are numerous in temperate South America, espe- 
cially Chile; in tropical South America they are probably restricted to arid 
regions and mountains; none are known from Central America but three 
species occur in the southern United States (north to Kansas) and northern 
Mexico. 


Caupolicanoides new subgenus 
(Figs. 6-9, 10, 48) 
Type species: Caupolicana pubescens Smith, 1879 (=C. herbsti Friese, 1904). 
The description is based on a male of this species, marked as a “typus” of 
herbsti, in the United States National Museum. 


1. Inner orbits moderately converging above in male, ocellocular distance 
equal to ocellar diameter. 2. Clypeus of male about 0.57 times as long as wide. 
3. Scape of male over three times as long as wide. 4. Front basitarsus of male 
straight, distinctly shorter than remaining tarsal segments. 5. Hind tibia of 
male straight. 6. Basitibial plate of male complete. 7. Hind basitarsus of male 
less than four times as long as wide. 8. All femora of male with abundant 
long hair, base of middle femur of male with area of short erect hair ventrally. 
Front femur over three times as long as broad. 9. Stigma slightly broader sub- 
apically than basally, discal area distinctly broader than marginal thickenings; 
first submarginal cell distinctly shorter than second and third together; mar- 
ginal cell broad (as well as unusually long), little over four times as long as 
wide; cell Ist M about 2.6 times as long as wide; first recurrent vein shorter 
than posterior margins of second and third submarginal cells which are sub- 
equal to first abscissa of Cun (i.e., posterior margin of cell 2nd M). 10. Pubes- 
cence ochraceous, on metasomal terga erect and not forming bands, especially 
long on sterna of male. 11. Metasoma nonmetallic. 12. Terga of male without 
lateral areas of short erect hair. 13. Sixth sternum of male with long hairs, 
posterior margin slightly produced medially, the apex of the projection gently 
concave; surface hairy to apex. 14. Dorsal apical process of seventh sternum 
of male spatulate, ventral apical process large, broad, hairy above, slightly 
spatulate. 

Caupolicanoides is known to me only from the type species which is from 
Chile. 

The striking features of this subgenus are the large marginal cell and large 
stigma, presumably primitive features by which C. pubescens differs from all 


726 Tue Universiry ScreNcE BULLETIN 


other Caupolicanini. Characters 3 and 10 above are the only other features by 
which Caupolicanoides differs from all species of Caupolicana s. str. 

However, in various other features Caupolicanoides falls at one extreme 
of the variation among the species of Caupolicana s. str. Character 6 is espe- 
cially striking since basitibial plates are often of importance in classification 
of bees; C. quadrifasciata and hirsuta have such plates in the male, just as 
does Caupolicanoides. In other species known to me the plates are delimited 
only posteriorly. In characters 2, 4, 5, and 7 also, as well as in those parts of 9 
not dealing with the stigma and marginal cell, Caupolicanoides is at one ex- 
treme of the variation within Caupolicana s. str. or just beyond that extreme; 
for example in character 4, the basitarsus is slightly shorter than in those 
species of Caupolicana s. str. in which it is shortest. As to character 14, the 
apical lobes of the seventh sternum are slightly broader, more spatulate, and 
more hairy than in Caupolicana proper, but are only a little different from 
those of quadrifasciata and hirsuta which are in turn closely approached by 
ruficollis. In these three species and also in albiventris the close connection 
between the lower apical lobes of this sternum and the bases of the upper ones 
is obvious, while in other species the lower lobes seem principally connected 
to the lateral margins of the sternum. The wing of /zrsuta is illustrated (Fig. 
12) to show how unlike Caupolicanoides it is in spite of certain other resem- 
blances of that species to that subgenus. 


Subgenus Willinkapis Moure 
(Figs. 11, 12) 


Willinkapis Moure, 1953, Dusenia, 4:66; Moure, 1964, Studia Ent., (n.s.) 7:453. 
Type species: Ptiloglossa chalybea Friese, 1906, by original designation. 


1. Inner orbits moderately converging above in male, ocellocular distance 


slightly less than ocellar diameter. 2. Clypeus of male about 0.65 times as long 
as wide. 3. Scape of male less than three times as long as wide. 4. First basi- 


Fic. 10. Forewings of males. Left, Crawfordapis luctuosa (specimen fom type series of 
crawford: from Costa Rica). Right, Cawpolicana (Caupolicanoides) pubescens (specimen from 
Concepcion, Chile, labeled as “‘typus” of herbsti by Friese, USNM). 


THE CLASSIFICATION OF THE DIPHAGLOSSINAE 721 


tarsus of male slightly arcuate, as long as remaining tarsal segments. 5. Hind 
tibia of male slightly arcuate. 6. Basitibial plate of male indicated only on pos- 
terior margin. 7. Hind basitarsus of male less than four times as long as wide. 
8. All femora of male with abundant long hair, less long on middle femur 
than on others; base of middle femur of male with area of short, erect hair 
ventrally. Front femur over three times as long as broad. 9. Stigma narrower 
apically than basally, discal area narrower than marginal thickenings; first 
submarginal cell equal to second and third together (first on posterior margin 
shorter than second and third); marginal cell narrow, nearly six times as long 
as wide; cell Ist M about three times as long as wide; first recurrent vein 
about equal in length to posterior margins of second and third submarginal 
cells together which are shorter than first abscissa of Cui (i.e., posterior mar- 
gin of cell 2nd M). 10. Pubescence, at least that of metasoma, largely black, 
terga with narrow marginal bands of dense hair which is sometimes white; 
sterna two to five of male with very long hair. 11. Metasoma strongly me- 
tallic blue (more metallic than in most Piiloglossa). 12. Terga of male with- 
out lateral areas of short erect hair. 13. Sixth sternum of male with hairs short 
and sparse, posterior margin medially produced, the projection emarginate 
much as in Zikanapis and Crawfordapis but projection narrower and shorter; 
surface with hairs nearly to apex. 14. Dorsal apical process of seventh sternum 
of male very slender, not at all spatulate; ventral apical process probably rep- 
resented by a broad, apically rounded, laterally directed, heavily sclerotized, 
lateral apical projection which is hairless except mesally. 

In characters 11 and 14, Willinkapis is unique among the subdivisions of 
Caupolicana. Its other features are all more or less duplicated among one or 
another of the subgenera, although the combination is not found elsewhere. 

Willinkapis, which has hitherto been accorded generic rank or placed as 
a subgenus of Zzkanapis, is known to me from two species. One is C. ( Wil- 
linkapis) chalybea (Friese) known from the cordilleran region of Argentina 
(see Moure, 1953). It ranges northward into Peru as shown by a female in the 
United States National Museum from Huanta, Andes, Peru, 2400 meters, 
March 24, 1941 (F. W. Woytkowski). From the same locality is a female of 
another, slightly smaller, species with the pubescence nearly all black. 


Subgenus Zikanapis Moure 
(Figs. 12, 30-34, 47) 

Zikanapis Moure, 1945, Arq. Mus. Paranaense, 4:147; Moure, 1964, Studia Ent., (n.s.) 7:421. 
Type species: Ptiloglossa zikani Friese, 1925, by original designation. 
Foersterapis Moure, 1964, Studia Ent., 7:441 (new synonym). 
Type species: Zikanapis foersteri Moure and Seabra, 1962, by original designation. 

1. Inner orbits moderately to strongly converging above in male, ocellocu- 
lar distance one third to one half an ocellar diameter. 2. Clypeus of male about 


0.85 times as long as wide. 3. Scape of male over three to nearly four times as 


728 Tue UNiversiry SCIENCE BULLETIN 


long as wide. 4. First basitarsus of male not or scarcely arcuate, distinctly 
shorter than remaining tarsal segments. 5. Hind tibia of male slightly arcuate. 
6. Basitibial plate of male indicated only on posterior margin. 7. Hind basitar- 
sus of male four or five times as long as broad. 8. Front femur of male with 
abundant long hair beneath, other femora with shorter hair; middle femur of 
male without area of short erect hair. Front femur three or more times as long 
as broad. 9. As in Willinkapis. 10. Pubescence largely ochraceous or in some 
South American forms extensively dark, forming apical white bands on 
metasomal terga of some species. 11. Metasoma nonmetallic. 12. Lateral ex- 
tremities of terga 2 to 4 and sometimes 5 and 6 of male with large areas 
densely covered by short erect hair of uniform length, these areas under low 
magnification appearing dull and scarcely punctate in contrast to adjacent 
areas. 13. Sixth sternum of male with hairs short, posterior margin medially 
produced and rather broadly bilobed; surface with hairs nearly to apex. 14. 
Dorsal apical process of seventh sternum of male spatulate, ventral apical 
process small and projecting but little from sternal margin. 

Characters 1, 2, and 12 are suggestive of Priloglossa, and Moure (1945) 
quite properly compared Zikanapis with Puloglossa as well as with Caupoli- 
cana. However, the majority of the characters indicate placement in Caupo- 
licana. One of the characters on which Moure placed special emphasis, the 
convergence of the eyes in Zikanapis, is weak because they are only moder- 
ately convergent in C. (Z.) elegans and in some of the other forms which he 
included in Zikanapis in 1964. 

Zikanapis, which has hitherto been accorded generic rank, contains several | 
species and seems most common in the northern and southern subtropical — 
regions, much less so in the intervening tropics although modesta is from | 
Colombia. Included species are: clypeata (Smith, 1879); elegans Timberlake, 
1965; foersteri (Moure and Seabra, 1962); funeraria (Moure, 1964); mega- 
lopta (Moure, 1948); modesta (Moure, 1964); seabrai (Moure, 1953); tucu- 
mana (Moure, 1945); and ztkani (Friese, 1925). 

The two species, foersteri and tucumana, which Moure (1964) placed in a 
subgenus, Foersterapis, of Zikanapis seem to me too similar to other Zikanapts 
to warrant separation although they do constitute a distinctive group. At the 
present stage of the development of systematics, there is no objective basis for 
such decisions and I can only say that I see no obvious advantage in separating 
Foersterapis as a named group. 


Alayoapis new subgenus 
(Figs. 12, 37-42, 48) 
Type species: Megacilissa nigrescens Cresson, 1869. 
1. Inner orbits strongly converging above in male, ocellocular distance less 
than one fourth of an ocellar diameter. Posterior margins of posterior ocelli of 


THE CLASSIFICATION OF THE DreHacLossiNaE 729 


1S 


Fic. 12. Forewings. Above left, Caupolicana (Willinkapis) chalybaea. Above right, C. 
(Caupolicana) yarrowi. Center left, C. (C.) hirsuta. Center right, C. (C.) ocellata. 
left, C. (Alayoapis) nigrescens. Below right, C. (Zikanapis) clypeata. 


Below 


male well in front of narrowest part of vertex (unlike other subgenera but like 
Crawfordapis). 2. Clypeus of male about or nearly 0.7 times as long as wide. 
3. Scape of male about three times as long as wide. 4. First basitarsus of male 
slightly arcuate, slightly longer than remaining tarsal segments. 5. Hind tibia 
of male slightly arcuate. 6. As in Willinkapts or no indication of basitibial 
plate in notabilis. 7. Hind basitarsus of male nearly five to over five times as 
long as wide. 8. Hair of femora of male short, not longer than femoral diam- 
eter, that of middle femur much shorter except in notabilis; base of middle 
femur of male without area of short erect hair ventrally. Anterior femora 
thick, not over three times as long as wide. 9. As in Willinkapis, but first sub- 
marginal cell longer than second and third together (first on posterior margin 


730 Tue UNiversiry SCIENCE BULLETIN 


equal to second and third). 10. Body with considerable dark pubescence, that 
of metasoma short; terga in some species with apical white fasciae; sterna 2-4 
of male with rather long hair. 11. Metasoma nonmetallic. 12. As in Wil- 
linkapis. 13. Sixth sternum of male with median nearly hairless area or in 
notabilis with short erect hairs; hairs of margin rather dense; posterior margin 
rounded, with hairs nearly to apex but margin proper a thin hairless trans- 
lucent flange (doubtful in notabilis). 14. Dorsal apical process of seventh 
sternum of male slightly spatulate with median lobe suggestive of Ptiloglossa 
but shorter; ventral apical process absent. 

The combination of characters 1 and 13 (rounded margin of sixth ster- 
num) is not found elsewhere in the genus. 

Alayoapis, named after Dr. Pastor Alayo D. of La Habana, Cuba, who has 
collected more specimens of the subgenus than anyone else and has provided 
me with useful information about the subgenus. It is known only from the 
islands of Cuba and Hispaniola. 


D. NORTH AMERICAN SPECIES OF THE GENERA 
CRAWFORDAPIS AND CAUPOLICANA 


This section concerns the species found not only in North America proper 
but also in Central America and in the Antilles. Probably all of the species 
are active principally in early morning and late evening, some of them when 
it seems completely dark to human observers. Nests are made in deep bur- 
rows, often and perhaps always in sandy soil. Known details of behavioral 
matters are indicated under each species. 

In the synonymies only major references are given, not mere records or 
repetitive generic changes. 


KEY TO THE SPECIES 


1. Metasomal integument red. -...............-..... Caupolicana ( Alayoapis ) notabilis 
— Metasomal integument dark brown or black... 03... 2 


2. Legs largely light red or yellowish. 2 == ee 3 | 


— Legs or at least hind leg reddish brown to black, usually more or less the 
same color as:the ‘body. .....:20.:-4.28) 2 oR 5 
3. Antenna yellowish red; ocellar diameter nearly twice maximum width of 
SCADC. secinh 2 oF ase ee ee Caupolicana (Caupolicana) ocellata 
— Antenna largely dark brown or black; ocellar diameter little if any greater 
than maximum width of scape. 2.2.1. 4 
4. Middle femur of male contorted; head and thorax with some black or 
dusky and dark tipped hairs. 2 Caupolicana ( Alayoapts ) subaurata 


— Middle femur of male not contorted; head and thorax with hair entirely — 


OchTaceOUS 29.22% 3 Se Caupolicana (Caupolicana) electa 
5. Male with lateral extremities of terga 2-4 with large areas of short, dense 


THE CLASSIFICATION OF THE DIPHAGLOSSINAE 731 


hair of uniform length, producing a distinctive dull appearance; clypeus of 
female rather flat with uniform punctation and short hairs over entire 
SIRINGS, elete ei ca oe eee ore ieam es A OF oe be 6 
—Male with lateral extremities of terga without such specialized areas: 
clypeus of female more convex with much variation in density of puncta- 
ama OU IANS: eevee oi nn 2 ee ee ee ee W, 
6. Male without metasomal bands; interocellar distance of female slightly 
more than maximum ocellar diameter. Caupolicana (Zikanapis ) clypeata 
— Male with bands of white hair at least laterally on apices of terga 2-5; inter- 
ocellar distance of female over twice maximum ocellar diameter. 
-nuncesetoncitage ieee een can 2s a ec re Caupolicana (Zikanapis ) elegans 
7. Metasomal terga without bands of pale hair. Crawfordapis luctuosa 
— Metasomal terga with apical bands of white hair... 8 
8. Hairs of thorax ochraceous, many of them dark tipped in female. 
_scats<SceobOLeoS ee ee Caupolicana (Caupolicana) yarrowt 
— Hairs of thorax with large areas of black to dark gray. 
Serer ee nak Seed ant sas Caupolicana ( Alayoapis ) nigrescens 


D1. SPECIES OF CRAWFORDAPIS 


There is only one recognized species of this genus. 


Crawfordapis luctuosa (Smith) 
(Figs. 1-5, 10, 11) 


Megacilissa luctuosa Smith, 1861, Jour. Entom., 1:150. 
Ptiloglossa crawfordi Cockerell, 1919, Proc. U.S. Nat. Mus., 55:178. 
Zikanapis (Crawfordapis) luctuosa; Moure, 1964, Studia Entom., (n.s.) 7:449. 


Crawfordapis luctuosa is a dusky haired species with light orange hairs at 
the apex and on the under surface of the metasoma. The femoral and propo- 
deal scopa of the female is largely cream colored. The species was described 
from Mexico, without other data. The type is in the British Museum (Natural 
History). The name crawfordi was based on material from “Ujurass de Ter- 
raba,” Costa Rica and the type is in the U.S. National Museum. A series of 
males in the British Museum (Natural History) is from Cerro Zunil, 4000- 
5000 feet altitude (Champion). According to Selander and Vaurie (1962) this 
locality is Volcdn Zunil, Quezaltenago, Guatemala. 

These Guatemalan males differ from Costa Rican material in being 
slightly larger with light hair mixed with the black all over the face below 
the antennae, with a band of light hair across the front of the scutum, with 
light hair on the posterior surface of the propodeum, the anterior surface of 
the first metasomal tergum, and the sides of the metasoma anteriorly, and 
with the whole venter of the metasoma covered with orange hair. Possibly 
they represent a distinct species or geographical variant. My drawings of 
genitalia and sterna are based on a specimen in the type series of crawfordi 


732 Tue University SciENCE BULLETIN 


Fic. 11. Faces, males at left, females at right. Above, Crawfordapis luctuosa; below, Cau- 
policana (Willinkapis) chalybaea. 


from Costa Rica. Moure (1964) reports specimens from Volcan Tacanas, 
Chiapas, Mexico; the Departamento de Chimaltenango, Guatemala; and Los 
Planes, El Salvador. Presumably his drawings are based on a specimen from 
one of these more northern localities and the differences between his drawings 
and mine may support the idea that different forms are involved. 


D2. NORTH AMERICAN SPECIES OF 
THE SUBGENUS CAUPOLICANA 


There are three species of this subgenus in North America, each very 
different from the others and equally different from the South American 


‘THE CLAssIFICATION OF THE DIPHAGLOSSINAE 733 


species of the subgenus. C. yarrow: and electa, however, are more similar to 
one another than either is to ocellata or any other species, and ocellata is a 
very isolated form. 


Caupolicana (Caupolicana) yarrowi (Cresson) 
(Figs. 12, 13-17, 29) 


Megacilissa yarrowi Cresson, 1875, Rep. U.S. geogr. survey west of one hundredth meridian, 


DMP 

Male: Length 17-21 mm; wing length 14-16% mm. 

1. Inner orbits converging above. Eyes closest on a line that is tangent to 
posterior margins of posterior ocelli, the latter far in front of posterior mar- 
gins of eyes; ocellar diameters about equal to maximum width of scape; 
ocellocular distance more than half of width of ocellus. 2. Basal part of labrum 
with two distinct longitudinal ridges submedially and weak longitudinal 
wrinkles laterally. 3. First flagellar segment longer than scape. 4. Anterior 
femur much thickened, especially broad basally where expanded posteriorly 
so that it is less than three times as long as broad; middle and hind femora 
progressively more slender. 5. Anterior femur with rather dense long hairs on 
lower surface, especially dense near posterobasal angle; other femora with 
only short sparse hairs ventrally except for small (one fifth as long as femur) 
basal patch of rufescent hairs on middle femur. 6. Hind basitarsus with apical 
third distinctly wider than basal third, the basitarsus little over half as long as 
the slender and distinctly curved tibia. 7. Propodeal triangle without trans- 
verse ridges. 8. Posterior margins of sternum 2 and usually 3-4 broadly emargi- 
nate. 9. Apex of sternum 6 rounded. 10. Hidden sterna and genitalia as shown 
in Figures 13 to 17. 11. Integument black, legs brownish black, under side of 
flagellum brown, tegula yellow brown. Wings light brownish with dark 
brown veins and stigma. 12. Pubescence of head white, ochraceous on vertex: 
pubescence of thorax, legs, and first tergum ochreous, paler on venter and on 
coxae, trochanters, and femora; dorsum of metasomal terga 2 to 7 with hair 
black, apical white bands on terga 2 to 4, these bands widest laterally and 
narrowed medially; metasomal sterna 1-4 with white hair, remaining sterna 
with dusky hair (see comments below on variation). 

Female: Length 18-20 mm; wing length 19 mm. 

13. Inner orbits subparallel except upper parts. 14. Ocellar diameters about 
equal to maximum width of scape; ocellocular distance greater than ocellar 
diameter. 15. First flagellar segment longer than scape, third broader than 
long and distinctly shorter than following segments. 16. Basal part of clypeus 
distinctly more shining than supraclypeal area, with distinct punctures. 17. 
Anterior coxa with apical spine covered with ochreous hair. 18. Propodeal tri- 
angle as in male (character 7). 19. Integument colored as in male (character 
11). 20. Pubescence of head white with scattered dusky hairs intermixed ex- 
cept on genal area; hair of vertex wholly dusky or black; subapical fringe of 


Tue UNIversiry ScIENCE BULLETIN 


Tl 
Go 
an 


clypeus and lower fringe of mandible rufescent. Hairs of thorax and first 
tergum ochraceous, whiter laterally, those of dorsum and pronotal tubercles 
and upper part of mesepisternum dark tipped. Hairs of basal segments of legs 
whitish or ochraceous except for areas of short reddish dusky hair on lower 
side of middle trochanter and base of middle femur; hairs of tibiae and tarsi 
dusky ochreous to almost black except that scopa of under side of hind tibia, 
like that of femur and trochanter, is white. Dorsum of metasomal terga 2-6 
with black hair except for white apical bands, slightly narrowed medially, on 
2-4, also on 1. First two sterna with ochraceous hair; sterna 3-4 with similar 
hair on posterior margins but otherwise with reddish dusky to blackish hair; 
sterna 5-6 with hair reddish dusky to black. 


This species is closest to C. electa as is shown by characters 1, 2, 3, 4, 9, 14, 
15, and 16 as well as by sternal and genital characters of the male, especially 
the ventroapical lobes of the seventh sternum. The other characters, as well as 
some aspects of those listed above, separate yarrowi from electa. 


C. yarrowi was described from New Mexico and the type is in the Acad- 
emy of Natural Sciences of Philadelphia (no. 2141). It occupies a wide range 
in the desert and semidesert areas of the southwestern United States and 
Mexico (see Figure 28). Records for Florida (Fox, 1898; Graenicher, 1930) 
are errors based on specimens of the broad banded southern form of C. electa. 


Altitude ranges from 4000 feet at Douglas, Arizona, to 5300 feet at La- 
Cueva in the Organ Mountains, New Mexico, and to 6400 feet at El Tascate, — 
Durango, and Zimapan, Hidalgo. 

Specimens have been collected as early as June 6 (Huachuca Mountains, 
Arizona) and June 23 (Tehuacdn, Puebla) but most records are in July and 
August, with one record on September 5 (Organ Mountains, New Mexico). 

There is available a single male of a probably distinctive population or 
species from the southern extremity of the range. The specimen, from the 
state of Puebla (Tehuacan, June 23, 1951, H. E. Evans), is unusually large 
with the white metasomal bands reduced; sterna 2-4 (especially 2) are un- 
usually broadly and deeply emarginate; the pubescence of the fourth and part 
of that of the third sterna is blackish like that of the following sterna instead 
of pale like that of the preceding sterna; the femora are unusually robust, the 
anterior one greatly produced posteriorly at the base and not much over twice 
as long as wide; and the middle basitarsus is broadest medially instead of 
parallel sided and of uniform width. 

Cockerell (1899) records the flight of C. yarrowi as from 5:15 to 6:15 a.m. 
(between dawn and sunrise) on September 4 and 5 in the Organ Mountains, 
New Mexico. He records a series from flowers of Datura meteloides, and two 
from Lippia wright. Probably these were nectar sources and Linsley (1960) 
and Linsley and Hurd (1959) record males of C. yarrowi taking nectar from 
Melilotus alba and Larrea divericata. At least on the latter plant the bees were 


THE CLASSIFICATION OF THE DIPHAGLOSSINAE 735 


Fics. 13-17, Caupolicana (Caupolicana) yarrow; \8-22, (G. ((C.)\ electas 23-2) GaG.) 
ocellata; males. For each species structures are genitalia; lateral view of apical part of genitalia, 
eighth sternum, lateral view of apical process of same, and seventh sternum. For further ex- 
planation see Figures 6-9. The scale line represents | mm for all three species. 


active before dawn. I have also seen the species near Douglas, Arizona, at 
about sunrise on Larrea, not collecting pollen. Specimens collected at Encar- 
naci6n de Diaz, Jalisco, on Eysenhardtia polystachya, had no pollen on the 
scopa and probably were feeding on nectar. 


Linsley and Cazier (1963) treat the pollen collecting, especially on Sola- 
num, in some detail, showing, for example, that in mid-August in southern 
Arizona on a clear morning the activity was from 5:20 to 6:50 while on an 
Overcast morning it was from 5:20 to 8:50 (sunrise both days was at approxi- 
mately 6:00). Most other pollen collecting bees on the same flowers started 


736 Tue University SCIENCE BULLETIN 


later and continued much later, but Pridoglossa jonesi Timberlake, while 
largely synchronous with C. yarrowi, started work perhaps slightly earlier 
and under overcast conditions stopped its activities considerably earlier than 
C. yarrowi. Pollen collecting, however, is not restricted to Solanum and 70% 
of the females collecting on Solanum already carried some Mentzeliatike 
pollen. At the same place where it visits Solanum in the morning, C. yarrow 
collects pollen in the evening (17:50-19:13; sunset at 18:51) from Mentzelia 
pumila, a flower that is not open in the morning. Linsley and Cazier (1963) 
also give a brief account of the scopal structure. 

Linsley (1962) records males of C. yarrow sleeping while grasping Meli- 
lotus stems with their mandibles, but he observed no aggregations of such 


bees. 


Caupolicana (Caupolicana) electa (Cresson) 
(Figs. 18-22, 29) 
Megacilissa electa Cresson, 1878, Proc. Acad. Nat. Sci. Philadelphia, p. 221. 
Caupolicana electa; Mitchell, 1960, Bees of the Eastern United States, 1:23. 
Megacilissa yarrowi; Fox, 1898, Ent. News, 9:128 (misidentification). 
Caupolicana (Megacilissa) yarrowi; Graenicher, 1930, Ann. Ent. Soc. Ame-., 23:161 (mis- 
identification). 

Male: Length 18-20 mm; wing length 15-15/4 mm. 

1. As in yarrowi. 2. Labrum with longitudinal ridges reduced to two 
gentle submedian convexities; wrinkles absent. 3. As in yarrowt. 4. Anterior 
femur thickened, less than four times as long as broad, not expanded basopos- _ 
teriorly but thickest shortly before middle; middle and hind femora about 
equal in thickness, not enlarged. 5. Anterior femur with rather dense long 
hairs on lower surface; other femora with hairs shorter and sparser ventrally | 
but less short and sparse than in yarrowz, basal patch of dense rufescent hairs 
on middle femur extending beyond middle. 6. Hind basitarsus parallel sided, 
scarcely over half as long as the slender and distinctly curved tibia. 7. Propo- 
deal triangle with a few transverse ridges, sometimes weak. 8. Posterior mar- | 
gins of sterna 2-4 transverse. 9. As in yarrowi. 10. Hidden sterna and genitalia 
as shown in Figures 18 to 22. 11. Integument black, tegula and legs yellowish 
brown, infuscated on coxae, trochanters, and femora; under side of flagellum 
dark brown, lightest apically. Wings colored as in yarrowz1. 12. Pubescence of 
head, thorax, legs, first tergum, lateroventral extremities of second tergum, 
and first two sterna ochreous, most deeply colored on dorsum of thorax; rest 
of metasoma with pubescence fuscous or black, very narrow apical bands of 
white pubescence on terga 2-4, sometimes reduced to lateral parts of 2 and 
absent on 3 and 4; or in specimens from southern Florida, these bands as in 
yarrout. 

Female: Length 17-18 mm; wing length 19 mm. 

13. Inner orbits slightly converging below except upper parts which con- 
verge above. 14. As in yarrow. 15. First flagellar segment longer than scape, 


THE CLassIFICATION OF THE DIPHAGLOSSINAE 737 


third slightly longer than broad. 16. As in yarrowt. 17. Anterior coxa without 
spine. 18. Propodeal triangle as in male (character 7). 19. Integument colored 
as in male (character 11) but legs reddish brown rather than yellowish brown. 
20. Pubescence as in male (character 12) but that of tarsi and outer surfaces 
of tibiae deep ochreous, that of area of basitibial plate brown, that of sterna 
23 reddish dusky, lateral extremities of sterna 2-4 with long ochraceous hair. 
The one female studied from the main range of the species has the apical 
tergal bands present; in southern Florida they are broad as in the male from 
that area. 

This species is most similar to C. yarrowi. The similarities and differences 
are summarized under that species. 

Caupolicana electa was described from Georgia and the type (no. 2140) is 
in the Academy of Natural Sciences of Philadelphia. It is known from sandy 
areas in the eastern lowlands from North Carolina to Georgia, Alabama, and 
northwestern Florida (Fig. 28). Localities are Mobile, Alabama; Crestview, 
Okaloosa County, Florida; Southern Pines and Lakeview, Moore County, 
North Carolina; and Harnett County, North Carolina. In addition, at least 
two specimens have also been collected in Dade County, southern Florida 
(Fox, 1898; Graenicher, 1930). 

The specimens from southern Florida have broad white tergal bands and 
look superficially like C. yarrow1, which accounts for their misidentification 
by Fox and Graenicher. They are also smaller and more slender than typical 
electa. In other features, however, they agree with electa. The probable speci- 


- fh 


Ne es te a ) 4/ & 
<< Poe tO A eis 


So 


— — SN —— —_ 


Fic. 28. Map showing distribution of the North American species of Caupolicana s. str. 
Dots in Mexico and the west show records of C. yarrowi; dots in the east, C. electa: crosses held 
together by shading, C. ocellata. 


738 Tue UNiversiry ScIENCE BULLETIN 


men (a male) which Fox recorded merely from Florida is in the American 
Museum of Natural History, labeled as from Biscayne Bay. The specimen 
taken by Graenicher was from South Miami. 

The season of flight seems to be autumnal (September and October). 
Flower records on specimens seen are yellow Gerardia (—=Aureolaria) and 
Trichostemma. The flight period is reported as “around sunrise” and late 
afternoon or dusk (Mitchell, 1960; Graenicher, 1930). 


Caupolicana (Caupolicana) ocellata new species 
(Figs. 12, 23-27, 29) 


Male: Length 14-15 mm; wing length 1144-12 mm. 

1. Inner orbits parallel except for upper parts which converge. Eyes closest 
ona line that passes through posterior ocelli, the latter extending back almost 
as far as posterior margins of eyes; ocelli enormous, diameters much greater 
than maximum width of scape, ocellocular distance less than half width of 
ocellus. 2. As described for yarrow7 but ridges weaker and wrinkles fewer and 
coarser. 3. First flagellar segment about as long as scape. 4. Femora of about 
equal thickness, anterior one about four times as long as broad. 5. Anterior 
femur with rather dense long hair on lower surface, middle and hind femora 
with ventral vestiture progressively sparser and shorter; patch of rufescent 
hair at base of middle femur absent. 6. Hind basitarsus parallel sided, dis- 
tinctly over half as long as tibia. 7. Propodeal triangle as in yarrow. 8. As in 
electa. 9. Apex of sternum 6 with median broadly V-shaped emargination. 
10. Hidden sterna and genitalia as shown in Figures 23 to 27. 11. Integument 
black; labrum, basal half of mandible, antenna, tegula and legs yellow brown 
or testaceous; coxae, trochanters, and femora darker brown. Wings nearly 
clear, veins and stigma brown. 12. Pubescence white, slightly ochraceous on 


dorsum of thorax and on terga 6-7; dorsal surfaces of terga 2-5 with hair black | 


or fuscous except for broad apical white bands, not narrowed medially, on 
terga 1-5. 

Female: Length 15-16 mm; wing length 11-12 mm. 

13. As in electa. 14. Ocellar diameter much greater than maximum width 
of scape; ocellocular distance over half ocellar diameter. 15. First flagellar seg- 


ment shorter than scape, third as broad as long or longer than broad. 16. Basal | 
part of clypeus dull and granular like supraclypeal area. 17. As in yarrowt. | 


18. Propodeal triangle as in male (character 11). 19. Integument as in male 
(character 11), but pale areas, including lower part of clypeus, reddish brown 
rather than yellow brown. 20. Pubescence white, slightly ochraceous on dor- 
sum of thorax, on outer surfaces of tibiae and tarsi, on under surface of mid- 
dle trochanter and base of femur, and on terga 5-6; dorsal surfaces of terga 2-4 
with hair dusky except for broad apical white bands on terga 2-4, a narrower 
one on 1; metasomal sterna with hair ochreous. 


THE CLASSIFICATION OF THE DIPHAGLOSSINAE 739 


Caupolicana ocellata is one of the most strikingly distinct species of Caz- 
policana, Characters 1, 3, 4, 5, 9, 14, and 15, are especially distinctive. 


I am pleased to accept the specific name suggested by Padre J. S. Moure 
when he saw this remarkable species; the name is given in reference to the 
enormous ocelli. 


The species occurs in sandy areas in the southern high plains and adjacent 
semidesert areas from Kansas to Chihuahua (Fig. 28). Type material is as 
follows: 


Holotype female and five female paratypes: Three miles south of Garden 
City, Finney County, Kansas, September 3, 1951 (C. D. Michener and W. E. 
LaBerge). One female paratype, same data but 5 miles south of Garden City. 
Three female paratypes, type locality (labeled Garden City), August 29, 1952 
(W. E. LaBerge). Allotype male and one male paratype: Albuquerque, New 
Mexico, July 23, 1950 (R. H. Beamer, H. O. Wright). The above are all from 
the Snow Entomological Museum of The University of Kansas. One female 
paratype: Samalayuca, Chihuahua, June 24, 1947 (D. Rockefeller Expedition, 
C. D. Michener), in the American Museum of Natural History. Forty-seven 
male and 11 female paratypes: 6.7 miles south of Manahans, Ward County, 
Texas, June 1-2, 1964 (Peter H. Raven) in the collection of the California 
Insect Survey, University of California, Berkeley. 

A paratype taken with the holotype in Kansas is in the collection of Padre 
16S. Moure, Curitiba, Brazil; another is in that of Dr. C. A. C. Seabra, Rio 
de Janeiro. 

Females have been recorded only on Dalea lanata; those from both Kan- 
sas and Texas were on this flower. Males have been taken on another small 
legume, Petalostemum flavescens (in New Mexico) and on an onagraceous 
flower, Gaura coccinea (in Texas). 

Specimens from Kansas were taken from early to mid morning and a 
female was seen to enter her nest after 9:00 a.m. on a sunny day. 

In Texas observations are available on the times of day when specimens 
were collected on June 2, 1964, at flowers of Dalea lanata, thanks to Dr. Peter 
H. Raven of Stanford University. The specimens were collected and thus re- 
moved from the population; figures for later in the day might have been 
higher had this not been the case. However, Table 1 clearly shows that C. 
ocellata was active early in the morning and late in the evening and that the 
other principal visitors to the same flowers, Martinapts luteicornis (Cockerell) 
and Agapostemon texana Cresson, appeared later and disappeared earlier, 
having little overlap with the Campolicana. Dr. Raven arrived at the site at 
5:00, when it was barely light, and found the Caupolicana common; they 
were much rarer by 5:45. He writes that at 20:10 they became abundant again 
and continued until 20:40 when it was pitch dark and he had to use a flash 
light to see the bees. After that time the bees were scarce or absent. 


740 Tue Universiry ScIENCE BULLETIN 


TasLe 1. Times of Capture of Principal Visitors to Flowers of Dalea lanata at 
Manahans, Ward Co., Texas, on June 2, 1964, by Dr. Peter H. Raven. Sunrise was 
about 5:40; sunset at 19:46. 


Caupolicana Martinapis Agapostemon 
ocellata luteicornis texanus | 

re) Le i) ore a) eg °F 
523 026: 0 0" = 22 ee eee 22 ] rae Bed ws3 ae 58 
(SONOS AR 0) OV are Ree ad aac) a a Sie 5 3) 
No observations made 63 
723 05820 Ole: eee 2 = 3) i 4 1 65 
SsQOSSES OD a ae eee eee eee ee ] =e 20 3 6 2 68 
Discontinuous : Bhs ae it Ee =e a4 active allday to95 
observations 
MisheSXO SIAN) ea a eee 6 7 1 7 ] 89 
192 OOS1O Age ae ee ee 6 1 4 88 
No observations made 
NO) 45=2. 05 0 Ol eeseee oe eke 4 cool pes ep ae 
20): 0,.022.0):5 0 eae eee eae 2 a me he Be Bes 83 


Of the eight males taken on Gaura coccinea, four were collected from 7:30 
to 8:15, four from 15:45 to 20:30. 

Nests were found by me in loose sand of dunes stabilized by vegetation, 
three miles south of Garden City, Kansas. They entered horizontal surfaces 
among small plants in the vicinity of the flowering Dalea lanata. 1 dug two 
nests. One seemed to consist of only a burrow although the bees were badly 
tattered and no males were taken, suggesting that the season was well ad-. 
vanced (September 3). The other was 60 cm deep, vertical, rather straight. At 
the bottom it turned sharply to one side, extended laterally for 8 cm to a 
vertical cell lined with a membrane similar to that of other colletid cells and 
partially filled with liquid provisions. Another provisioned cell was found 
about 8 cm in another direction from the bottom of the vertical burrow; the 
horizontal burrow leading to it had been completely filled with sand. These 
very scanty data suggest basic nest structure reasonably similar to that of other 
Caupolicana species whose nests have been described. 


D3. NORTH AMERICAN SPECIES OF ZIKANARFIS 


The North American Zikanapis are similar to the South American C. (Z.) 
megalopta (Moure), being large bees with largely ochraceous pubescence.’ 
They are quite different from the type species, C. (Z.) zikani (Friese), which 
has much black pubescence. The North American species agree in the tufts 
of long hairs, sigmoidly curved, arising from the lateral extremities of the fifth 
sternum of the male, and less conspicuously from the fourth. They also agree 
in the rather flat clypeus of the female (flatter in elegans) with abundant, 
rather uniform punctation, finely roughened and dull surface between punc- 


Tue CLASSIFICATION OF THE DIPHAGLOSSINAE 741 


Fic. 29. Faces of North American species of Caupolicana s. str., males at left, females at 
right. Above, C. yarrowi; center, C. electa; below C. ocellata. 


tures, and abundant erect hairs or bristles, less than half as long as the longest 
hairs of the face, with apices commonly bent rather sharply downward: 
The genitalia and hidden sterna of the two North American species are 


742 Tue UNIversirty SCIENCE BULLETIN 


similar. They are illustrated for clypeata; those of elegans have been studied 
only in a dry state and differ from those of clypeata as follows: penis valves 
and gonoforceps somewhat less broad apically, the former with the preapical 
angle (basal to slender, downward directed apex) more rounded; apical proc- 
ess of eighth sternum not quite so broad medially, the apical part parallel 
sided. 


Caupolicana (Zikanapis) clypeata (Smith) 
(Figs. 12, 30-34, 47) 
Megacilissa clypeata Smith, 1879, Descriptions of new species of Hymenoptera in the collection 
of the British Museum, p. 59. 


Caupolicana clypeata; Cockerell, 1905, Trans. Amer. Ent. Soc., 31:343. 
Zikanapis (Zikanapis) clypeata; Moure, 1964, Studia Ent., (n.s.) 7:439. 


Male: Length 15-17 mm; wing length 1244-13 mm. 

1. Eyes strongly converging above, upper interocular distance about as 
long as scape; interocellar distance about equal to maximum ocellar diameter; 
ocellocular distance clearly less than half an ocellar diameter. 2. Penis valves 
widest subapically. 3. Integument black, under side of flagellum, clypeus, 
tegula, anterior and middle legs, and hind tarsi variably light brown, some- 
times considerably infuscated; posterior margins of metasomal segments 
broadly transparent brownish. 4. Pubescence ochraceous, sometimes slightly 
dusky on vertex; shorter hairs of dorsum of metasoma, especially on second 
and third terga, slightly dusky; no pale fasciae on metasoma; hairs of outer 
surface of middle tibia slightly dusky, those of outer surfaces of hind tibia and 
basitarsus dusky or blackish. 

Female: Length 16-18 mm; wing length 1114-12 mm. 

5. Interocellar distance slightly more than maximum ocellar diameter. 6. 
Integument black, under side of flagellum especially apically and tegula light 
brown; fore and middle legs with considerable brownish color. 7. Pubescence 
of head largely whitish except that of clypeus which is brown; fuscous hairs 
intermixed with white on rest of face and genal areas; hair of vertex fuscous. 
Hair of dorsum of thorax ochraceous with dusky tips; hair of sides and venter 
whitish, apices dusky on upper parts of sides. First metasomal tergum and 
first two sterna with hair pale ochraceous or whitish, remaining sterna with 
hair ochraceous and somewhat infuscated except for very long, plumose, pale 
ochraceous hairs laterally on sterna 3 and 4; second to fourth terga with hair 
except at extreme sides blackish, and with apical bands of appressed white 
hair, narrowed medially; fifth and sixth terga with strongly infuscated, deep 
ochre hairs. Hairs of legs ochraceous, those of outer sides of tibiae and tarsi 
infuscated (slightly so on foreleg), hind tibial hairs of basitibial region and 
extending to middle of tibia black, scopa including hairs of inner sides of hind 
tibiae white. 


THE CLASSIFICATION OF THE DIPHAGLOSSINAE 743 


This species was described from Oaxaca, Mexico. It is not clear whether 
this means the city or elsewhere in the state. I have examined the type in the 
British Museum (Natural History). 

A series was collected on a misty morning in complete darkness at Tux- 
pan, Michoacan, September 1, 1962, by D. H. Janzen. The vicinity consists of 
cultivated land. Seven males were taken at car lights between 5:10 and 5:20 
a.m., two males between 5:20 and 5:35 a.m., two males between 5:35 and 5:45 
a.m. One male was taken on flowers of Salvia at the side of the road at first 
light (still much too dark to see the bee) at 5:45 a.m., another at 6:00 a.m. 
Two females were taken at the car lights, both at 5:50 a.m. 

A single female was taken, with no details as to collecting time, 2 miles 
east of Lake Patzcuaro, Michoacan, July 25, 1954 (J. W. MacSwain). 

Moure (1964) records specimens from Amula (now Almolonga, 9.5 km 
northwest of Chilapa), Guerrero, at 6000 feet altitude and 11 miles southwest 
of Acambaro, Guanajuato, August 17, 1954. 


Caupolicana (Zikanapis) elegans Timberlake 


Caupolicana elegans Timberlake, 1965, Jour. New York Ent. Soc., 73:46. 

Male: Length 14-16 mm; wing length 1014-13 mm. 

1. Eyes moderately converging above, upper interocular distance about 1.6 
times as long as scape; interocellar distance about 1.5 times maximum ocellar 
diameter; ocellocular distance slightly less than half an ocellar diameter. 2. 
Genitalia and hidden sterna similar to those of clypeata, the most conspicuous 
difference being that the penis valves are widest medially. 3. Integument black, 
underside of flagellum apically, tegula, and distal parts of tarsi light brown; 
fore and middle legs sometimes brownish; margins of metasomal segments 
broadly translucent. 4. Pubescence ochraceous, scattered dusky hairs on face 
along inner orbits (scarcely noticeable in type) and some fuscous hairs on 
vertex (not in type); hairs of dorsum of metasomal terga 2 to 4 black or fus- 
cous, those of 5 and 6 reddish fuscous, posterior margins of terga 2 to 5 with 
bands of white appressed hair, these bands present only laterally on 2 and 3, 
broken medially on 4, and continuous on 5; hairs of outer surfaces of middle 
and hind tibiae and hind basitarsus blackish. 

Female: Length 17 mm; wing length 114% mm. 

5. Interocellar distance over twice maximum ocellar diameter. 6. Integu- 
mental coloration as in male (character 3) but fore and middle legs black 
except for brownish small segments of tarsi. 7. Pubescence of head and thorax 
as described for clypeata. Pubescence of metasoma as described for clypeata but 
hairs of first tergum with dusky apices; sternal pubescence slightly paler, the 
very long hair at sides of sterna 3 and 4 whiter and more extensive; white 
fasciae on terga 2 to 4 broken medially, that on 4 narrowly so; hairs of terga 


744 Tue University ScrENcCE BULLETIN 


Fics. 30-34, Caupolicana (Zikanapis) clypeata, male. 30, genitalia; 31, lateral view of — 
apical part of same; 32, eighth sternum; 33, lateral view of apical process of same; 34, seventh 
sternum. 


Fics. 35-36, Caupolicana (Alayoapis) notabilis, male. 35, apical process of eighth sternum; 
36, apical processes of seventh sternum. These drawings were sketched from dry preparations 
in the British Museum (Natural History). 


Fics. 37-40, Caupolicana (Alayoapis) subaurata; 41-44, C. (A.) nigrescens; males. For 
each species structures are genitalia (somewhat flattened artificially in nigrescens), eighth ster- 
num, lateral view of apical process of same, and seventh sternum. 

For further explanation see Figures 6-9. The scale lines do not apply to Figures 35 and 36. 


5 and 6 largely reddish black. Hairs of legs as in clypeata but those of outer 
surfaces of middle and hind tibiae and tarsi black. 

This species was described from one male taken near Portal, Cochise 
County, Arizona, at about 10:00 a.m. It is preserved in the American Museum 
of Natural History and was lent me for study by Dr. J. G. Rozen, Jr. It is 
smaller than the other specimens and with less dark hair on the head. 


THE CLASSIFICATION OF THE DIpHAGLOSSINAE 745 


— 


Fic. 47. Faces of North American species of the subgenus Zikanapis, females at right. 
Above, C. (Z.) clypeata; below, C. (Z.) elegans. 


The only other known specimens are a pair taken on highway 150, 37 
miles west of Tehuacan, Puebla, Mexico, August 31, 1962, on flowers of Salvia, 
the male at 6:00 p.m., the female at 6:40 p.m. (D. H. Janzen). 


DF. THE SPECIES OF ALAYOAPIS 


This subgenus is known from three species, the only Antillean representa- 
tives of the subfamily, which are characterized below: 


Caupolicana (Alayoapis) nigrescens (Cresson) 
Megacilissa? nigrescens Cresson, 1869, Trans. Amer. Ent. Soc., 2:295. 
Male: Length 15-16 mm; wing length 13-14 mm. 
1. Eyes very strongly converging above, upper interocular distance less 


! 
746 Tue University ScrENCE BULLETIN 


than length of scape, eyes closest at point over an ocellar diameter behind | 
ocelli; interocellar distance about equal to ocellar diameter; ocellocular dis- 
tance less than one sixth ocellar diameter. 2. Front coxa not spined; front fe- 


mur about three times as long as wide. 3. Front tibia about 4.4 times as long 
as broad with some hairs as long as tibial diameter. 4. Middle femur with hair 
much shorter than that on other femora; femur not contorted. 5. Mediotarsal 
segments of middle leg nearly symmetrical. 6. Inner hind tibial spur gently 
curved medially. 7. First sternum with median apical spine; second and third 
sterna with margins recurved. 8. Hidden sterna and genitalia as shown in 
Figures 41 to 44. 9. Black, flagellum dark brown beneath, paler apically, apex 
of last segment reddish. Legs brown to reddish, more or less infuscated, coxae, 
trochanters, mediotarsi and distitarsi blackish. Tegula piceous to black. 
Wings brownish, slightly darker than in yarrowi. 10. Hair of face mixed 
black and white. Hair of vertex black, of occiput, genal areas, and under side 
of head white. Hair of prothorax white, blackish on posterior lobe of pro- 
notum. Mesonotum with hair dark gray to blackish except for broad white 
band across front of scutum, narrower white band along scuto-scutellar suture, | 
and white along posterior edge of scutellum and along lateral edge of scutum 
mesad from tegula. Pleural and ventral areas of thorax with dark gray to 
blackish hairs except for large dull white lateroventral mesepisternal area and 
dull white on metapleuron. Metanotum and posterior part of propodeum | 
with white hairs, long dorsolateral hairs of propodeum partly black. Coxae _ 
and middle and hind trochanters with hairs dark gray to black; rest of legs 
with hairs pale yellowish in some lights, blackish in others, darkest on pos- 
terior femora, forelegs with the most yellowish. First tergum with hairs white 
anteriorly and laterally, black dorsally, with a small dorsolateral indication of 
apical white band. Remaining terga with hairs black, white at extreme sides 
of second, short on terga 2 to 4 with narrow, bright, white apical bands (nar- 
rower than in yarrowt). First sternum with hair white, others with hair dark 
gray. 


Female: Length 16-18 mm; wing length 11 mm. 


11. Basal part of labrum raised to form two strong longitudinal submedian 
carinae. 12. Maximum ocellar diameter less than ocellocular distance which is 
subequal to interocellar distance. 13. Apical spine of strigilis shorter than rest 
of strigilis. 14. Integumental coloration as described for male (character 9) but 
hind leg more consistently black. 15. Hair of face mixed black and white, 
sparse on shining (but punctured and minutely roughened) clypeal disc; 
median part of frons with hairs all white; hair of vertex black forward to and 
including transverse band of dense black hair between anterior and posterior 
ocelli, wholly black hair extending down in upper paraocular areas; subapical 
fringe of clypeus and lower fringe of mandible rufescent fuscous; genal and 
hypostomal areas with hair white. Thoracic pubescence (in poor condition in 


THE CLASSIFICATION OF THE DIPHAGLOSSINAE 747 


available specimen) has a pattern similar to that of male. Hairs of legs black, 
reddish on fore legs and small segments of tarsi; femoral scopa white. Tergal 
pubescence as described for male but white band of tergum 2 present only 
laterally, of 3 broken medially. Sternal pubescence dusky reddish, whitish at 
extreme sides. 

Caupolicana nigrescens is most similar to subaurata. Characters 1, 2, 4 
(contorted femur), 5, 12, and the color differences are among the striking 
distinguishing features. 

This species was described from Cuba without further data but the type, 
which is still in good condition in the Gundlach collection in Havana, bears 
the number 293; Dr. Pastor Alayo D. has looked this up in Gundlach’s manu- 
script catalogue and found the notation “Yateras, Ote., XI.” It is interesting 
that the type locality for this species and its relative, sebaurata, should be the 
same place in Oriente Province. 

Dr. Alayo writes of this species, “These bees are dwellers in the most dense 
forests of Cuba, specially in the mountains of Oriente Province, and I do not 
remember to have found any specimen in the lowlands.” 

Specimens have been studied by me or reported to me by Dr. Alayo (from 
his collection and that of Dr. F. de Zayas M.) from the following localities, 
all in Oriente Province, Cuba: 

Sierra Cristal, Mayari, May, 1955 (Zayas); Piloto, Moa, June, 1954 (Zayas 
and Alayo); Loma del Gato, Hongolosongo, Sierra del Cobre, June, July, 
September (Zayas and Alayo); same locality, September, 1935, October 1-2, 
i255, 2600 to 3325 feet altitude (J. Acuna, S. C. Bruner, L. C. Scaramuzza, 
collectors); Pico (or Alto de) Cardero, Macizo del Turquino, June, 1963 
(Alayo). 


Caupolicana (Alayoapis) subaurata (Cresson) 
(Figs. 37-40, 45, 48) 
Megacilissa? subaurata Cresson, 1869, Trans. Amer. Ent. Soc., 2:296. 

Male: Length 16 mm; wing length 13!4 mm. 

1. Eyes strongly converging above, upper interocular distance greater than 
length of scape, eyes closest at point less than ocellar diameter behind ocelli; 
interocellar distance greater than ocellar diameter; ocellocular distance over 
one fourth ocellar diameter. 2. Front coxa with short apical spine; front femur 
of male less than three times as long as wide. 3. Front tibia about four times 
as long as broad with some hairs nearly as long as tibial diameter. 4. Middle 
femur with hair much shorter than that on other femora; femur contorted 
(Fig. 45). 5. Mediotarsal segments of middle leg with posterior lobes much 
larger than anterior lobes. 6. Inner hind tibial spur strongly curved medially. 
7. As in nigrescens. 8. Hidden sterna and genitalia as shown in Figures 37 to 
40. 9. Black, flagellum as in nigrescens. Legs red, mediotarsi blackish, espe- 
cially on mid and hind legs. Tegula reddish brown. Wings as in nigrescens. 


748 THe UNIversiry SCIENCE BULLETIN 


10. Hair color pattern as in nigrescens but pale hair ochre instead of white, 
dark hairs dark only apically so that gray color is light and somewhat ochrace- 
ous, pale band on scuto-scutellar suture weak, pale ventrolateral mesepisternal 
area larger than in nigrescens so that midventral dark region is reduced and 
ventral and lateral pubescence of thorax may appear entirely ochraceous; hairs 
of legs ochreous. First metasomal tergum with ochraceous hairs anteriorly and 
laterally, gray or blackish dorsally except for complete narrow apical band of 
white. Remaining terga with black hairs, becoming reddish on 5 and 6 and 
extreme sides of 2, narrow apical white bands (narrower than in nigrescens) 
on 2-4 as well as 1. Sterna 1 and 2 with ochraceous hair, others with light 
ochreous gray hair. 
Female: Length 17 mm; wing length 11 mm. 


11. Basal part of labrum raised to form two strong longitudinal submedian 
carinae. 12. Maximum ocellar diameter about equal to ocellocular distance 
which is less than interocellar distance. 13. Apical spine of strigilis longer than 
rest of strigilis. 14. Integumental coloration as in male (character 9). 15. Pubes- 
cence of head as described for female nigrescens but pale hairs ochraceous, 
dark ones mostly dusky rather than black, at least basally; transverse band of 
dense hair between ocelli mixed dark and light, as are hairs of upper para- 
ocular areas; subapical fringe of clypeus and lower fringe of mandible cop- 
pery. Thoracic hair coloration as in male, the striking pattern of nigrescens 
being only weakly evident. Hair of legs ochreous, even fulvous on tibiae and 
basitarsi. Tergal pubescence as described for male. Sternal pubescence dusky 
reddish, fulvous laterally. 


/ 


This species is closest to C. nigrescens. The most distinctive features are 
listed under that species. 


This species was described from Cuba without further data. The type, 
which is in good condition in the Gundlach collection in Havana, bears the 
number 292. Dr. Pastor Alayo D. has found that this number in Gundlach’s 
manuscript catalogue refers to the notation “Yateras, Ote., XI.” This is also 
the type locality for nzgrescens although other records are for localities differ- 
ent than those where nigrescens has been taken. Caupolicana subaurata may 
well be wider ranging both ecologically and geographically than nigrescens. 

Specimens have been studied by me or reported to me by Dr. Alayo (from 
his collection and that of Dr. F. de Zayas M.) from the following localities: 

Oriente Province: Puerto Boniato, Santiago de Cuba, October, 1943 
(Alayo); Moa, Baracoa, June, 1954 (Alayo and Zayas); Capitolio, Rio Yara, 
1150 feet altitude, May 18, 1948 (Zayas). 

Pinar del Rio Province: Mogote de Xyla, Couret, carretera a Luis Lazo, 
November, 1956 (Jaume); San Vicente, Vifales, May, 1956 (Alayo and 
Zayas) ; Sierra Cajalbina, June, 1956 (Zayas) ; Rancho Mundito, Sierra Rangel, 
June, 1950 (Zayas). 


‘THE CLASSIFICATION OF THE DIPHAGLOSSINAE 749 


46 


Fics. 45-46, Posterior and ventral views of middle femora of males of Cuban Alayoapis. 
Posterior views are above, with the dorsal surfaces uppermost; ventral views are below with the 
anterior surfaces uppermost. 45. C. (A.) subaurata; like letters indicate like localities. 46. C. 
(A.) nigrescens. 


Caupolicana (Alayoapis) notabilis (Smith) 
Megacilissa notabilis Smith, 1861, Jour. Ent., 1:149. 
Male (the following is based on notes made from male type in the British 
Museum) : 


1. About as in nigrescens. 2. Front coxa not spined; front femur less than 
three times as long as wide. 3. Front tibia about six times as long as broad, 
with only very short hairs. 4. Middle femur with hair rather long and dense 
like that of hind femur; femur not contorted. 5. Not seen. 6. Inner hind tibial 
spur nearly straight. 7. First sternum not spined; second and third sterna with 
margins transverse. 8. Hidden sterna as in sebaurata except as shown in Fig- 
ures 35 and 36; genitalia as in swbaurata but distal halves of parameres densely 
covered on dorsal and outer surfaces with black, plumose hairs (cleared 
preparation not made). 9. Black, legs and tegula dark brown, presumably 
faded from blackish. Wings light brownish, slightly darker apically than in 
yarrow1. Integument of metasoma red. 10. Hair of head white, light brownish 
on vertex, lower part of clypeus, and mandibles; intermixed long fuscus hairs 
on these areas and elsewhere below antennae; a tuft of fuscous or black hairs 
near eye margin at level of anterior ocelli. Hair of thorax gray, fuscous or 
darker gray on anterior lateral parts of scutum, anterior face of mesepister- 
num, scutellum, and posterior parts of sides of thorax; coxae and trochanters 
and much of anterior femora with gray hair, rest of hair of legs mostly fus- 
cous. Metasomal hair fuscous, that of first segment gray ventrally and lat- 
erally. 


Female: Length 17-19 mm; wing length 12-1314 mm. 

11. Basal part of labrum with broad even median convexity. 12. Maximum 
ocellar diameter less than ocellocular distance which is slightly less than inter- 
ocellar distance. 13. Apical spine of strigilis about as long as rest of strigilis. 
14. Integumental coloration as in male (character 9) (tegula and legs blackish 
in more recently collected material). 15. Hair of head as described for female 
nigrescens but with some white hairs around ocelli and intermixed black hairs 
on genal and hypostomal areas. (Black hairs only dusky in female from Brit- 


750 Tue Universiry ScrENcE BULLETIN 


ish Museum approximately as old as male type.) Hair of thorax as in male, 
varying to all nearly black. Hair of legs black or nearly so, reddish on under 
sides of some tarsal segments, scopa grayish because axes of hairs are black but 
branches colorless. Metasomal hair as in male but fuscous hairs black in 
fresher material. 

This species is known only from the Dominican Republic in the island of 
Hispaniola. The type which is a male, and one female, both from Santo Do- 
mingo, are in the British Museum. Six females in the United States National 
Museum were taken at Constanza, Dominican Republic, May 27, 1927 (A. 
Wetmore). Dr. Wetmore writes (77 litt.) that his collecting was at about 4000 
feet altitude in an area of pines mixed with patches of rain forest. The bees 
were collected near midday, on flowers growing in an opening among the 
pines. 


LITERATURE CERED 


CocKERELL, T. D. A. anp Witmatte Porter. 1899. Contributions from the New Mexico 
Biological Station—VII. Observations on bees, with descriptions of new genera and 
species. Ann. Mag. Nat. Hist. (7)4:403-421. 

Fox, W.J. 1898. (No title). Ent. News 9:128. 

Friese, H. 1898. Monographie der Bienengattungen Megacilissa, Caupolicana, Diphaglossa und 
Oxaea. Ann. K. K. Naturhist. Hofnus. [Wien] 13:59-86. 

GRAENICHER, S. 1930. Bee-fauna and vegetation of the Miami Region of Florida. Ann. Ent. 
Soc. Amer. 23:153-174. 

Linstey, E. G. 1960. Observations on some matinal bees at flowers of Cucurbita, Ipomoea and 
Datura in desert areas of New Mexico and Southeastern Arizona. Jour. New York Ent. 
Soc. 68:13-20. 


. 1962. Sleeping aggregations of aculeate Hymenoptera 
55:148-164. 

Linscey, E. G. anp M. A. Cazier. 1963. Further observations on bees which take pollen from 
plants of the genus Solanum (Hymenoptera: Apoidea). Pan-Pac. Ent. 39:1-18. 


II. Ann. Ent. Soc. Amer: 


Linscey, E. G. ano P. D. Hurp. 1959. Ethological observations on some bees of southeastern 
Arizona and New Mexico (Hymenoptera: Apoidea). Ent. News 70:63-68. 


Micuener, C. D. 1944. Comparative external morphology, phylogeny, and a classification of 
the bees (Hymenoptera). Bull. Amer. Mus. Nat. Hist. 82:151-326. 


. 1954. Bees of Panama. Bull. Amer. Mus. Nat. Hist. 104:1-176. 


Moure, J. S. 1945. Contribuicao para o conhecimento dos Diphaglossinae, particularimente 
Priloglossa (Hym.-Apoidea). Arg. Mus. Paranaense 4:137-178. 


. 1953. Notes sobre Colletidae Sul-Americanos (Hymenoptera-Apoidea). Dusenia 4:61-78. 


. 1964. As espécies de Zikanapis, com a descric¢ao de dois novos subgéneros e duas espécies 
novas (Hymenoptera, Apoidea). Studia Ent. (n.s.) 7:417-458. 

SELANDER, R. B. anp P. Vaurie. 1962. A gazetteer to accompany the “Insecta’’ volumes of the 
“Biologia Centrali-Americana.”’ Amer. Mus. Novitates 2099:1-70. 


Tue CLASSIFICATION OF THE DIPHAGLOSSINAE qo 


Fic. 48. Faces of the subgenera Alayoapis and Caupolicanoides, females at right. Above, 
C. (A.) nigrescens; center, C. (A.) subaurata; below left, C. (C.) pubescens male (from Con- 


| aon Chile, labeled as “typus’” of herbsti by Friese, USNM); below right, C. (4.) norabilis, 
emale. 


me 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


A REVISION OF 
THE BEE GENUS CALLIOPSIS 
AND THE BIOLOGY AND ECOLOGY OF 
Cc. ANDRENIFORMIS 
(HYMENOPTERA, ANDRENIDAE) 


By 


Alvin F. Shinn 


Marine Biolog'cal Laboratory 
LIBRARY 


JAN 1 6 1967 
WOODS HOLE, MASS. _ 


Vor. XLVI Paces 753-936 Jan. 6, 1967 No. 21 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Wor. XLVI Paces 753-936 JAN. 6, 1967 No. 21 


A Revision of the Bee Genus Calliopsis 
and the Biology and Ecology of C. andreniformis 
(Hymenoptera, Andrenidae)' 


ALVIN F. SHINN” 


ABSTRACT 


This is a taxonomic revision of the bee genus Calliopsis; information concerning 
distribution and biology is incorporated in the treatment of each species. Calliopsis 
occurs throughout North and Central America from southern Canada to southern 
Panama. 

Four subgenera are recognized, including one new one: Callopsima. Thirty- 
eight species are included. One species is placed in synonymy; one is removed 
from synonymy; one is left as a nomen nudum; one is declared a nomen dubium; 
and two are transferred to other genera. About 7,000 specimens were examined. 

Twenty-four new species are described: C. granti, C. mouret, C. peninsularts, 
C. sonora, C. empelia, C. zora, C. helenae, C. rogert, C. syphar, C. limbus, C. gilva, 
C. fulgida, C. yalea, C. rozeni, C. pectidis, C. timberlake1, C. unca, C. azteca, C. 
crypta, C. deserticola, C. hurdi, C. quadridentata, C. kucalumea, and C. michenert. 

The biology and ecology of Calliopsis andreniformis were studied intensively 
at nesting sites in Lawrence, Kansas, and auxiliary observations were made on the 
species at Nacogdoches, Texas. It is primarily a summer bee which is active from 
May to September, and it forages mostly on legumes, especially alfalfa and clovers. 
It passes the winter underground as a prepupa. Emergence is in May about two 
weeks after the start of transformation to the pupa. Females typically make nest 
burrows in hard-packed clayey soil near or among clovers, and excavated soil is 
left as a closed mound on top of each burrow. The finished nest burrow of the fe- 
male is a slanting, winding tunnel with one to ten short lateral burrows radiating 
around it at successively deeper levels. Each lateral burrow ends in a polished, 
waxed cell containing a spherical pollen ball within a thin, transparent membrane, 


*Contribution number 1324 from the Department of Entomology, The University of Kansas, 
Lawrence. 
“Present address: Oak Ridge National Laboratory, Oak Ridge, Tenn. 37830. 


754 Tue University SCIENCE BULLETIN 


and one frankfurter-shaped white egg atop it. The tunnel is four millimeters in 
diameter, and is up to 162 millimeters deep. In Kansas two generations per season 
are usual: in eastern Texas there are three generations. 

A grid of squares each 30 centimeters on a side was fixed in place to cover two 
nesting areas of 41 square meters which included about 250 nests during one sea- 
son. The succession of nesting phenomena, and physical and biotic factors were 
studied. Male and female burrows are aggregated in distinctly separate areas. Both 
sexes spend the night in their burrows, and marked individual males and females 
returned repeatedly to their same overnight burrows. The bees are active in winds 
up to 32 kilometers per hour, air temperatures of 27 to 40°C., and soil surface tem- 
peratures up to 54°C. The stimulus for daily emergence is a combination of soil 
temperature and soil moisture. Watered plots showed that rainfall increased bur- 
rowing activity. 

Males patrol fairly well delineated areas. Each selects a resting place on the 
ground—a twig, stone, or leaf—and sets forth in a definite flight pattern, returning 
and alighting repeatedly at the same place. They often fly out of their flight pattern 
to make a sortie over nearby clover flowers. Mating takes place at the flowers or on 
the ground near the female burrows. An intruder in the male flyway is engaged — 
in an aerial “dogfight,” the two tumbling over each other, falling to the ground, 
kicking up dust, and biting one another. The intruder 1s repulsed. This is the first — 
example of territorial behavior in the Andrenidae. 

The female emits an odor of oil of lemongrass when slightly squeezed. The 
first mounds of female burrows contain this odor, but those of the males lack it. 
Experimentation suggests that its biological role is in aiding females to recognize 
their nest. 

Parasites of Calliopsis are the bees, Holcopasites calliopsidis, H. illinoiensis, H. 
arizonicus, and Sphecodes spp.; the pyemotid mite Trochometridium tribulatum; 
and the molds Penicillium cyclopium, Aspergillus flavipes, and A. sydow1. An 
undetermined species of robber fly is a predator. Bombylius ater bombards with 
her eggs those open burrows where wind has blown away the tumulus, and 
bombyliid larvae were found in a few Calliopsis cells with no traces of the imma- 
ture Calliopsis. 


INTRODUCTION 


The genus Calliopsis includes 38 described species of solitary, ground-nest- 
ing bees having white, cream or yellow areas on the head, legs, and mesosoma, 
and lacking such color on the metasoma. As the genus is currently inter- 
preted, its species occur only, and are distributed widely, in North and Cen- 
tral America. Four subgenera are recognized and described. The state of the 
genus remained confused until Michener’s decisive handling of it in Muese- 
beck, Krombein, and Townes (1951). In that publication numerous species of 
North America north of Mexico erroneously described as Calliopsis were 
transferred to their correct generic assignment. The only reliable published 
key to a circumscribed group of Calliopsis species is that of Mitchell (1960) 
which includes the three species known from the eastern United States. Other 


A REVISION OF THE BEE GENUS ISS) 


keys are totally useless because 24 of the present species were unrecognized by 
the authors of the keys. Most of the original descriptions are inadequate in 
the light of present knowledge, and therefore all species are redescribed 
herein. The types deposited in the British Museum are unavailable to me, but 
excellent notes on certain characteristics of them have been furnished, on my 
request, by Padre J. S. Moure. 

The investigation of the biology and ecology of Calliopsis andreniformis, 
the type species of the genus, was undertaken to form a frame of reference for 
similar work on other species. An attempt was made to collect meaningful 
data on diverse phases of the bee’s biology and ecology so that even meagre 
data on other species would have a basis for comparison. The results of the 
investigation reveal striking resemblances to the biology of Andrena, Perdita, 
and Nomadopsis as discussed in appropriate places. My field observations 
have convinced me that C. andreniformis is potentially a valuable pollinator of 
small-flowered clovers, e.g., alfalfa. Its wide distribution, intense activity, 
great range of flowers visited, and the larger numbers in the field than mu- 
seum collections indicate, all spell an important role for this species in the 
maintenance of native vegetation, especially among the Leguminosae and 
Compositae. 


ACKNOWLEDGEMENTS 


I express my thanks to Professor W. E. LaBerge who first drew my atten- 
tion to Professor Charles D. Michener’s interest in the Calliopsis problem 
(Professor Michener was in Brazil at the time). I am sincerely grateful to 
Professor Michener for initial outlining of the systematics and suggestions for 
the biological phase of the study, as well as his long-continued interest, en- 
couragement, and specific help in preparation of the manuscript. To Padre J. 
S. Moure of the University of Parana, Curitiba, Brazil, I am indebted for 
illuminating conversations which established the base for my understanding 
of the South American Panurgine genera which are closely allied to Calliopsis, 
and for the time generously spent in making tedious measurements and obser- 
vations on Calliopsis types in the British Museum as mentioned above. 

I am grateful also to the National Science Foundation for support of the 
behavioral parts of this study, through a grant (G11967) for Professor C. D. 
Michener to the University of Kansas, and to the Society of the Sigma Xi for 
a Sigma Xi-RESA Grant-in-Aid in the summer of 1958 which made possible 

comparative biological observations on Calliopsis teucrit and Nomadopsis 
scitula at Coaldale, Colorado. 

I gratefully acknowledge the aid of the Oak Ridge National Laboratory, 

operated by Union Carbide Corporation for the U.S. Atomic Energy Commis- 
sion, in the preparation of the final copies of this manuscript. 


756 Tue UNiversity SCIENCE BULLETIN 


Thanks are extended to the following collectors and museum curators who | 
have lent so generously of their own material or of specimens in their charge: 

A.H. Alex, A & M College of Texas, College Station; G. E. Ball, Univer- 
sity of Alberta, Edmonton; R. D. Bird, Field Crop Insect Laboratory, Depart- 
ment of Agriculture, Brandon, Manitoba; G. E. Bohart, U.S.D.A. Wild Bee 
Pollination Investigations, Utah State University, Logan; G. D. Butler, Jr., 
and F. G. Werner, University of Arizona, Tucson; Paul Christian, Univer- 
sity of Louisville, Kentucky; E. F. Cook, University of Minnesota, St. Paul; 
W. L. Brown, Jr., Cornell University, Ithaca, New York; H. Denmark, Flor- 
ida State Plant Board, Gainesville; W. Enns, University of Missouri, Colum- 
bia; H. E. Evans, Museum of Comparative Zoology, Harvard, Cambridge, 
Massachusetts; R. C. Froeschner, U.S. National Museum, Washington, D.C.; | 
D. W. Fronk, University of Wyoming, Laramie; W. W. Gibson, Stephen F. 
Austin State College, Nacogdoches, Texas; H. J. Grant, Jr., Academy of Nat- | 
ural Sciences of Philadelphia, Pennsylvania; C. Hopla, University of Okla- 
homa, Norman; P. D. Hurd, Jr., E. G. Linsley, H. V. Daly, University of | 
California, Berkeley; K. E. Hyland, Jr., University of Rhode Island, Kings- 
ton; M. T. James, State College of Washington, Pullman; G. F. Knowlton, 
Utah State University, Logan; J. N. Knull, Ohio State University, Columbus; 
J. B. Kring, Connecticut Agricultural Experiment Station, New Haven; K. 
V. Krombein, U.S. National Museum, Washington, D.C.; W. E. LaBerge, 
University of Nebraska, Lincoln; the late R. Lambert, Insect Systematics and 
Biological Control, Canada, Department of Agriculture, Ottawa, Ontario; F. 
A. Lawson, University of Wyoming, Laramie; K. McArthur, Milwaukee 
Public Museum, Wisconsin; J. A. Mathewson, N. Scituate, Rhode Island; C. 
D. F. Miller, Entomology Research Institute, Canada Department of Agricul- 
ture, Ottawa, Ontario; T. B. Mitchell, North Carolina State College, Raleigh; 
B. E. Montgomery, Purdue University, Lafayette, Indiana; R. A. Morse, Cor- 
nell University, Ithaca, New York; L. W. Quate, Bishop Museum, Honolulu, 
Hawai; C. W. Rettenmeyer, Kansas State University, Manhattan; H. G. 
Rodeck, University of Colorado Museum, Boulder; J. G. Rozen, Jr., and the 
late H. F. Schwarz, American Museum of Natural History, New York, New 
York; E. S. Ross and G. I. Stage, California Academy of Science, San Fran- 
cisco; the late H. C. Severin, South Dakota State College, Brookings; L. H. 
Shinners, Southern Methodist University, Dallas, Texas; M. E. Smith, Uni- 
versity of Massachusetts, Amherst; R. R. Snelling, Los Angeles County Mu- 
seum, Los Angeles, California; P. H. Timberlake, University of California, 
Riverside, California; H. K. Townes, Museum of Zoology, University of 
Michigan, Ann Arbor; G. E. Wallace, Carnegie Museum, Pittsburgh, Penn- 
sylvania; L. O. Warren, University of Arkansas, Fayetteville; R. Wenzel, 
Chicago Natural History Museum, Illinois; J. A. Wilcox, New York State 


A Revision OF THE BEE GENUS qD>/ 


Museum, Albany; I. H. H. Yarrow, British Museum (Natural History), Lon- 
don, England; F. N. Young, Indiana University, Bloomington. 

I thank particularly Dr. Karl V. Krombein for comparing specimens with 
types in his care; Professor T. B. Mitchell for his loan of homotypes of some 
species not available to me; Professor P. H. Timberlake for his magnanimous 
gesture in lending me his entire collection of Calliopsis including type speci- 
mens and specimens of species he recognized as new some years ago. Recog- 
nition should be accorded my son, Roger, for his assistance in the field and 
laboratory. 

My cousin, Dr. Sylvia Earle Taylor, merits a special note of thanks for 
contributing the fine scratchboard drawings of Calliopsis andreniformis and 
C. anomoptera. To my sister, Mrs. Virginia S. Griggs, I am happy to extend 
thanks for aid in the exhaustive checking, ordering, and recording of collec- 
tion data from the specimens used in the study. 

Lastly, it gives me pleasure to acknowledge the moral and financial help 
of both my parents and my wife’s parents in the last stages of the study, for 
without their aid the completion of the work would have been much delayed. 


DISTRIBUTION 


Ecologically, the distribution of the genus Calliopsis is virtually unknown. 
The bits of information available suggest a group which nests in hard-packed 
clayey soil less, usually much less, than 100 m from its pollen source. As 
exceptions to generalities, C. (Verbenapis ) nebraskensis has been taken dig- 
ging in sand dunes in Minnesota, and Calliopsts andreniformis was taken by 
my son, Roger, nesting in loam beside salt marshes at Hancocks Bridge, New 
Jersey, where it used the pollen of the Copper Mallow (Malva neglecta). 

Flower preferences are relatively pronounced for each subgenus, and dif- 
fer among them. Although Calliposts s.s. is widely polylectic, it has been col- 
lected mostly on Leguminosae, especially the small-flowered clovers. Perts- 
sander occurs principally on Euphorbiaceae. Calliopsima is found primarily 
on the Compositae, particularly the Astereae and Heliantheae, with many 
records for Heterotheca and Grindelia. Verbenapis, as its name implies, is an 

oligolege of Verbena. 

Pertssander is apparently restricted to arid areas, but the other subgenera 
are amply represented in mountains, deserts, plains, and cultivated land, but 
not in heavily forested areas. 

Seasonal distribution is somewhat different among the subgenera: Call1- 
opsis s.s. has its peak activity soon after summer begins; Verbenapis shows 
maximum activity shortly afterwards, about midsummer; and Pertssander 
and Calliopsima have their peaks in late summer. It should be borne in mind 
that these seasons, of course, occur at different calendar months depending 


758 Tue University ScIENCE BULLETIN 


upon latitude and altitude. Table 1 gives the seasonal distribution of the spe- 
cies of Calliopsts. 


Tasre 1. Seasonal Distribution of Species of Calliopsis 


Species Jan. Feb. Mar. Apr. May June July Aug. Sept. Oct. Nov. Dec. 
1. andreniformis 
2. teucrit 
3. granti a 
4. rhodophila 
5. mouret 
7. squamifera 
6. hondurasica ee ee a 
8. peninsularis 
9. sonora 
10. empelia 
ll. zora 
12. helenae 
13. anomoptera 
14. rogert 
15. syphar 
16. limbus 
17. gilva 
18. fulgida 
19. yalea 
20. rozent 
21. coloradensis 
22. pectidis 
23. timberlaket 
24. bernardinensis a 
25. unca 
26. crypta Pe, Eco: 
27. azteca 
28. chlorops 
29.  coloratipes 
30. deserticola 
31. pugionts 
32. hurdi 
33. quadridentata 
34. kucalumea 
35. verbenae 
36. nebraskensis 
37. hirsutifrons 
38. micheneri 


Geographically, the genus occurs from coast to coast in North America and 
from latitude 50° North in Canada to latitude 8° North in Panama. No 
specimens are known from the northwestern United States and adjacent Can- 


A REVISION OF THE BEE GENUS 759 


ada, viz., northern California, Oregon, Washington, and British Columbia. 
There is a similar lack of specimens from El Salvador, British Honduras, and 
Nicaragua, presumably because of lack of collecting. 

Calliopsts s.s. and Calliopsima are represented throughout the range of the 
genus. The distribution of Perissander is limited to southeastern Arizona and 
the Sonoran Desert in México, Arizona, California, and Baja California. Ver- 
benapts has been collected only about Mexico City northward to New Mexico 
and Texas and in the plains states east of the Rocky Mountains with an iso- 
lated population from northern New Jersey. I expect it will eventually be 
found in most of the eastern United States. 

The richest concentration of species in each of the subgenera occurs in 
Mexico or the arid southwestern United States. These regions are well known 
for their diverse habitats and abundant opportunities for geographic isolation. 


PHYLOGENY AND SYSTEMATICS 


There is no fossil record of these bees nor is there such a record for any 
panurgine bee. This situation forced an interpretation of phylogenetic rela- 
tionships deduced from species at one level in time, and this interpretation is 
based primarily upon comparative morphology of the adults supplemented by 
the available distributional data. The resultant diagram of relationships is 
given in Fig. 1. 

In Fig. 1 the species syphar and yalea are depicted with a relationship to 


syphar rhodophila andreniformis bernardinensis pugionis 


rogeri mourel deserticola 


anomoptera 
granti pectidis coloradensis coloratipes 


limbus 


yalea 
timberlakei 


1 teucril chlorops 
7 peninsularis 


rozeni crypta 
Squamiteray 9" 8 Nf NS R= = azteca 
empelia 


sonora 


kucalumea 


helenae 
¢ 


< 


nebraskens!s 


PERISSANDER 


quadridentata 


verbenae 
michener! 
hondurasica hirsutifrons hurdi 
| | 
ei VERBENAPIS Cae ne es 
SHOPOE UM like ACAMPTOPOEUM- like 
ancestor ances 


Fic. 1. Diagram of relationships of the genus Calliopsis. 


760 Tue UNIversiITy SCIENCE BULLETIN 


Perissander. 1 believe they belong there, but this will only be determined 
upon discovery of the males. A dashed line indicates uncertainty. 

Four genera of panurgine bees are closely related to Calliopsis: Nomadop- 
sis, Hypomacrotera, Acamptopeum, and Liopoeum. Rozen (1951) has shown 
that the generic limits of these bees are questionable, and I agree with him. 
Various combinations of presumably important characters which appear in 
Calliopsis are found in these allied genera. The unsettled status of the genera 
makes it impossible to state which subgenera of Calliopsis have more primi- 
tive characters, and, moreover, it is not possible to make decisions upon which 
species within a subgenus are more primitive than others because particular 
specializations in each species are counterbalanced by other specializations in 
other species. This is the same predicament encountered by Rozen (1958) in 
his revision of Nomadopsis. 

Three main types of male genitalia are represented among, and are charac- 
teristic of, the subgenera of Calliopsis, as explained below. 

Type 1, Calliopsis s.s. and Pertssander: penis short, thick; penis valve 
about twice penial length, in form of a thin, broad sheet which is usually 
wider medially, convex dorsally with intricate folds, lines, and thickenings, 
and hollowed ventrally; volsellae small, well-separated, knoblike or lobelike. 
See Figs. 8-57. 

Type 2, Calliopsima: penis long, slender; penis valve narrow, tubular, 
slightly exceeding length of penis, having an expanded, flattened, dorsally di- 
rected terminal portion; volsellae large, well-separated, elongate. See Figs. 58- 
122) 

Type 3, Verbenapis: penis long, thick; penis valve narrow, medio-ven- 
trally concave, terminal portion directed dorsally, tip bent toward mesal line; 
volsellae small with projections so close as to give the impression of a single 
structure. See Figs. 123-143. 

Primarily on the basis of a study of the genitalia and sterna of panurgine 
bees which are likely relatives of Calliopsis, I have concluded that the genera 
most closely related to it are the North American Nomadopsis and Hypo- 
macrotera, and the South American Acamptopoeum and Liopoeum. The 
superficially somewhat similar South American genera Spinoliella, Callony- 
chium, and Arhysosage (=Ruiziella Timberlake) do not belong in the same 
group. Hypomacrotera and Micronomadopsts parallel Calliopsis s.s. in male 
genitalia and sterna, yet are closer to Verbenapis on the basis of external 
characteristics. 

The diagnosis excludes all panurgine genera except the South American 
Acamptopoeum Cockerell and some species of Liopoeum Friese. This 1s 
deliberately contrived because until a thorough review of the classification of 
the panurgine genera closely related to Calliopsis is completed, I am inclined 


to favor the possibility suggested by Michener (1944:246) that the “South 


A REVISION OF THE BEE GENUS 761 


American genus Parafriesea=Acamptopoeum .. . is probably a mere sub- 
genus of Calliopsis.” | believe the species included by J. S. Moure (1956, per- 
sonal communication) in Liopoeum represent more than one genus, but those 
considered by him to be Acamptopoeum are a clearcut group of closely related 
species. The differences and similarities among the subgenera of Calliopsis 
seem to be of the same order of magnitude as those between Acamptopoeum 
and any one of them. In any event, if my interpretation of the content of 
Calliopsis presented herein proves acceptable, then surely Acamptopoeum 
must fall as a subgenus of the older Callopsis. Padre Moure, however, would 
unhesitatingly accord generic rank to Calliopsis s.s., Calliopsima, Perissander, 
and Verbenapis (1957, personal communication). With such a classification, 
Acamptopoeum would also be of full generic rank. An investigator’s position 
on a problem like this depends upon personal psychology, philosophy, and 
taste. 

Species from South America which most closely resemble the genus 
Calliopsis as herein constituted are Liopoewm hirsutulum Friese, 1908 (not 
Spinola, 1851), and two apparently undescribed species of ?Liopoeum (one 
may be Camptopoeum laetum Vachal) from Argentina. The latter two 
species may prove eventually to represent yet another group intermediate 
among Calliopsis s.s., Calliopsima, Liopoeum, and Acamptopoeum. Color 
patterns and head dimensions of Calliopsima are similar to those of Acamp- 
topoeum. Verbenapis is most similar, among South American bees, to Lio- 
poeum trifasciatum (Spinola, 1851); a number of important external charac- 
ters are shared. These are, for example, placement of cream areas on face in 
the female; shape of face; presence of a few curled hairs on front tarsomeres 
24; similar propodeum; and similarly shaped and sloped metasomal tergum 
1. The genitalic pattern of L. trifasciatum is that of Calliopsis s.s., however, 
although three or four metasomal terga bear yellow maculae. Liopoeum is a 
more likely progenitor of Nomadopsis than is Spinoliella. No counterparts of 
Perissander are known to me from South America. 

The absence of closely related forms in the Palearctic region and the pau- 
city of species in the northern and eastern United States, together with the 
numerous closely related forms from South America, argue for an origin of 
Calliopsis in South America with migrations northward through Central 
America and Mexico to the United States and Canada. 

Taxonomically useful characters at the subgeneric level were relatively few 
with the females, but included hair color of the prepygidial and pygidial fim- 
briae; length of mouthparts; sculpture of the dorsal enclosure of the propo- 
deum; extent of light color on the paraocular area; and relationships among 
the minimum interocular distance, clypeocellar distance, and flagellar length. 
The most reliable and hence the most useful bases for subgeneric groups, as 
well as for discrimination of species, were the male genitalia and sterna 5, 6, 7, 


762 Tue UNIvErRSITY SCIENCE BULLETIN 


and 8. Other useful ones for subgeneric grouping of the males were length of 
mouthparts, sculpture of the dorsal enclosure of the propodeum, extent of 
light color in the paraocular areas, color of subantennal plates, and presence or 
absence of scutellar and metanotal hair pads. No reliable differentiating char- 
acter was found for the females of Perissander; the male must be known to 
make a positive subgeneric assignment. 

Considerable difficulty was experienced in finding reliable characters to 
separate females of several closely related species. The most important eri- 
terion in every such case was the sculpture of the dorsal enclosure of the 
propodeum or the punctation and ground character of the integument of the 
scutum and metasomal terga | and 2. 

The bases for the matching of the sexes, among sympatric species, are pri- 
marily the similarity of the punctation of the scutum and of metasomal terga 
1 and 2 and the striking resemblance of the mouthparts other than the mandi- 
bles. Matching of the sexes proved troublesome in the case of the species 
crypta and rozent. Most students of the wild bees assume that the capture of 
male and female bees in copulo is virtually proof that they are the same spe- 
cies. The following paragraph, however, shows that this may not be the case. 

Initial analysis of the species in Calliopsima led me to group females of 
chlorops, crypta, and rozeni all under chlorops. This grouping was based on 
too few specimens for the recognition of crypta and rozent. Specimens re- 
ceived recently have clarified the status of the three species involved. J. G. 
Rozen and G. I. Stage, both competent collectors, supplied mating pairs of 
Calliopsis including a male of rozeni in each case. Rozen’s patr, however, in- 
cluded a female different from the one in Stage’s pair. On the basis of the 
close similarity of both punctation and vestiture of Stage’s male and female, 
I concluded that they were the same species and that Rozen’s female repre- 
sented crypta. Is it possible that females of these sympatric species resemble 
cach other so closely that their respective males may make an error in identity 
and copulate with the wrong female? 

Characters useful at the specific level included various quantitative dimen- 
sions and ratios among them, as well as qualitative characters, as noted in the 
keys and taxonomic treatments of the species. The black or dark brown in- 
tegumental color is subject to fading, e.g., the female type of tewcrit was de- 
scribed as “black .. . clypeus entirely black . . . tegulae shining piceous. . . .” 
The specimen is now brown to light brown. Fading is equally striking in 
older specimens of chlorops, coloradensis, nebraskensis, and verbenae, among 
others. The white, cream, or yellow maculations do not seem to change in 
quality. Because of fading, differences in darkness of ground color are de- 
scribed only in extreme cases, and the integument is to be taken as dark, non- 
metallic, for all species unless otherwise described. 


A Revision OF THE BEE GENUS 763 


DESCRIPTIVE TERMINOLOGY AND METHODS 


AspreEVIATIONS. These are held to a minimum and include the following: 

2-5, 1-4, 9-wt, etc. Read as “2 to 5 inclusive, 1 to 4 inclusive,” or “distance 
from position 2 to position 5, distance from position 9 to position wt,” etc. 

mio—minimum interocular distance. 

mm=millimeter(s). 

mow=—middle ocellar width(s). 

N.A.=not available, used where a structure cannot be examined or 
measured, usually because of its absence, the fragility of the specimen, or the 
unavailability of the specimen to me. 

pwa=puncture width(s) apart. 

S.W.R.S.=Southwestern Research Station of the American Museum of 
Natural History, 5 miles west of Portal, Chiricahua Mountains, Cochise 
County, Arizona. 

wing, 12-13/13-14—forewing, the ratio of the distance between 12 and 13 
to the distance between position 13 and 14 (cf. Fig.5). 

Terminotocy. The terminology followed in the systematic portion of the 
paper is basically that of Michener (1944). An explanation or equivalent of 
certain terms which are used in this work or have been used in older works 
treating Calliopsis species is given below. The terms used here are italicized. 

apicotarsus: collective term for the last four tarsomeres; used in apposition 
to basitarsus. 

light coloration on the subantennal plates=dog-ear marks (Cockerell). 

dorsal enclosure of the propodeum—disc of metathorax (Robertson)= 
horizontal (usually somewhat declivous) enclosure of the propodeum=basal, 
or upper, portion of the propodeal triangle. It differs markedly from the re- 
mainder of the propodeal triangle by its sculpturing. 

eccentric punctures—punctures deviating from radial symmetry in being 
shallow at one end, becoming deeper with a vertical, or nearly vertical, wall 
at the other end. 

galeal gap—the distance between the tip of the galea in repose and the base 
of the prementum. 

-mere—sufhx used with tarsus and flagellum to designate real or apparent 
segments, e.g., tarsomere. 

orbital convergence ratiothe ratio of the middle ocellar interocular dis- 
tance (at the level of the lower border of the middle ocellus, Fig. 3,T) to the 
minimum interocular distance. It is used as a measure of the degree of ven- 
tral convergence of the inner orbits. 

pebbled=a type of integumental sculpturing which resembles the surface 
of pebble-grained leather. 

pronotal lobes=tubercles—the lobelike, dorsolateral projections of the 
pronotum. 


ae eae 
| 


Ee 


A Revision oF THE BEE GENUS 765 


propodeal flats—the posterolateral faces of the propodeum, immediately 
adjacent to the propodeal triangle, where such faces are relatively plane, some- 
times medially convex, areas. 

scutellar and metanotal hair pads=calluslike areas of mesoscutellum and 
metanotum of certain Calliopsis males (Timberlake) well-defined patches 
of dense, extremely short, profusely branched—hence mosslike—pubescence, 
which occupy, respectively, the depression between the disc of the scutellum 
and the base of the hind wing, and the lateral portions of the metanotum. 

scutellum—mesoscutellum. 

scutum—mesoscutum—thoracic dorsulum (Cockerell). 

sternum—metasomal sternum, where unqualified. 

tergum—metasomal tergum. 


MeasvreMents. Head and galeal measurements are shown in Pigs: 2-4. 

All measurements are given in millimeters and were made at 30X mag- 
nification with a Bausch & Lomb StereoZoom'®? Microscope and an ocular 
micrometer with each micrometer unit equal to 17 microns or 0.017 mm. The 
precision of the decimal ratios can be inferred from the number of digits 
given, the last digit being the doubtful one. 

Lengths of scutum, scutellum, metanotum, and propodeum are measured 
along the median longitudinal line. The intertegular distance is the mini- 
mum distance measured transversely across the scutum between the edges of 
the concavities of the thorax which receive the tegulae. Forewing length is 
length including tegula. Hindwing length is from the junction of the wing 
with the thorax to the wing tip. It is as good an indicator of general body size 
as the forewing. 

The forewing of a male Calliopsis andreniformis is illustrated in BigeD: 
An arbitrary numbering system was used in the analysis of similarities and 
differences of the wings within and among species. Lengths of wing veins 
Were measured from the midpoints of the vein intersections shown. The 
length of the marginal cell was taken as the maximum length measured be- 
tween the inside edges of the bordering veins. Distance from the tip of the 
marginal cell to the wing tip was measured at 9-wt (Fig. 5). 

Abdominal width is the maximum width and invariably occurs on meta- 
somal tergum 3. 


Fics. 2-4. Measurements of head and galea. 2-3: A, frontal line; B, antennocular distance, mini- 
mum; C, clypeal length, median; D, clypeal width, the lateral limit usually mesal to the 
lowest point on the orbit; E, clypeocellar distance; F, eye length; G, eye width, facial view; 
H, eye width, basal, the straight line distance between anterior and posterior orbital margins; 
I, head length; J, head width, maximum: K, inner subantennal sutural distance, minimum; 
L, inner subantennal suture, vertical length; M, interantennal distance; N, interocellar dis- 
tance; O, interocular distance, minimum; P, ocellocular distance, minimum (=ocellocular 
line); C + E, ocellolabral distance; Q, outer ocellar distance; R, outer subantennal sutural dis- 
tance; S, outer subantennal suture, vertical length; T, interocular distance at level of median 
ocellus. 4: Galea, lateral view, showing measurement of length. 


766 Tue University SCIENCE BULLETIN 


Fic. 5. Forewing of male Calliopsis andreniformis. Wing dimensions are obtained by measuring 
point to point, 0-wt, wing length including tegula; 1-2, length of vein C; 2-3, length of 
stigma; 2-5, length of prestigma; 2-6, length of stigma from base to vein r; 3-4, length of 
marginal cell (=2nd Ri) along costal margin of wing; 3-4-9, length of vein Ri; 5-10, basal 
side of Ist submarginal cell (=Ist abscissa of Rs); 6-7, length of crossvein r; 6-9, length of 
marginal cell; 7-8, length of costal side of 2nd submarginal cell (=2nd abscissa of Rs); 8-9, 
length of free part of marginal cell (=3rd abscissa of Rs); 10-11, length of posterior side of 
Ist submarginal cell (=Rs+M); 10-15, length of basal vein (=Ist abscissa of M); 11-12, 
distance from Ist transverse cubital or intercubital to Ist recurrent vein (=2nd abscissa of M); 
11-14, length of posterior side of 2nd submarginal cell (=2nd to 4th abcissae of M); 12-13, 
distance between anterior ends of recurrent veins (=3rd abscissa of M); 12-16, length of first 
recurrent vein (=Ist m-cu); 13-14, distance between transverse cubital and 2nd recurrent 
veins (—4th abscissa of M); 13-19, length of 2nd recurrent vein (=2nd m-cu); 15-16, length 
of Ist abscissa of Cu; 15-17, distance from cu-v to basal vein; 15-18, length of Cu; 16-18, 
length of 2nd abscissa of Cu; 17-20, length of transverse medial crossvein (=cu-v); 18-19, 
length of Cur; 18-21, length of Cuz; 20-21, length of 2nd abscissa of V. 


The word “distance” as well as the dimensional units are usually omitted 
from various facial and body measurements to save needless repetition, and 
where units are not specified they are millimeters, e.g., “clypeocellar 0.85" 
means that the clypeocellar distance is 0.85 mm. 

Basitarsal length in Verbenapis is measured along the lateral median line 
and hence excludes the ventral apical prolongation present in this subgenus; 
basitarsal length in the other subgenera is measured from extreme base to 
apex which includes any apical prolongation present. 


Notes on TECHNIQUES. In various places a magnification is specified for an 
observation by placing it in parentheses following the observation. For exam- 
ple, “facial fovea shiny, unsculptured (15> ).” This is done because terms 
such as shiny, impunctuate, polished, unsculptured, etc., are sometimes ap- 
plicable only within a certain range of magnification, and the above statement, 
while true, might also read “facial fovea dull, roughened (20Kies 

Optimum lighting is diffused daylight, and the Bausch & Lomb Fluores- 
cent [luminator used in this study yielded 500 foot candles of illumination at 
the specimen. Integumental punctation and propodeal sculpture will be ob- 
scured or distorted for some species with undiffused light. 

Accurate measurements of mouthparts may occasionally be necessary. 
Where relaxation in a moist chamber followed by manual manipulation failed 
to extend the mouthparts as desired, I used the following method: relaxed 
specimen in a moist chamber; dissected away mouthparts and placed them in 


A ReEvIsION OF THE BEE GENUS 767 


cold lactophenol solution; brought to boiling point and held there until struc- 
tures extended; rinced tn hot glycerin; allowed to cool; transferred to fresh 
glycerin for measurements. This method has worked on specimens 80 years 
old. 

The same technique produces cleared genital and sternal preparations 
which retain their deposited melanins and fine hairs. I find this superior to 
the caustic potash treatment except that occasionally the genital capsule has 
entrapped air which, on expansion and escape, tears the capsule. 

Descriptions. The description of a new species is based upon the holotype 
and allotype with parenthetical notes referring to paratypes which differ sig- 
nificantly from the type. Variation of other specimens is discussed under 
Remarks or Geographic Variation. 

Redescription of a previously described species is treated in a similar man- 
ner when the type(s) is available. Otherwise the redescription is based on 
metatypes, homotypes, notes made on types by a reliable hymenopterist, topo- 
types, or the most representative specimen among all available, with paren- 
thetical notes referring to specimens which differ significantly from the de- 
scribed specimen. 

The initial part of the description of each sex of a species gives measure- 
ments which are general indicators of the size of the bee. Least reliable is the 
length of the bee, because it is dependent upon the degree of extension of both 
head and metasomal terga. Clypeal length is the measurement showing the 
highest correlation with the length of the bee. 

A number is given to each character or group of characters described for 
each subgenus and each species. This numbering is used to facilitate com- 
parisons among the taxa. Where a particular number is absent for a given 
species, the character is not of comparative importance for the subgeneric or 
species group in which the species occurs. 

Descriptions of the species ordinarily omit verbal treatment of the geni- 
talia and sterna because the illustrations are adequate in themselves. Details 
of those which are of particular significance in separation of closely allied 
species are designated by an arrow on the illustration which points to the 
diagnostic feature. 

Detailed collection data are given for new species, but such data are given 
only for the uncommon redescribed species. 

Distributional maps depict the known range of each species. Collections 
and biological data are too few to attempt prediction of the total range of most 
species. Each symbol for a species represents one or more specimens collected 
at that locality, and in some cases represents several collections made at nearby 
localities. This procedure is justified on a small scale map provided that the 
collections so plotted are from similar localities. Of course, every locality is 
listed in the written treatment. 


768 Tue UNiversiry SCIENCE BULLETIN 


TAXONOMIC TREATMENT 
Genus CALLIOPSIS Smith 


Type species. Calliopsis andreniformis Smith, 1853, by designation of Ashmead, 1889, Trans. 
Amer. Ent. Soc. 26:85. 

Diacnosts: (a) American panurgine bees with white, cream or yellow 
integumental markings on head, mesosoma, and their appendages, on a black 
ground, metasoma without such light markings; (b) marginal cell truncate 
apically, its length along costal margin of wing beyond apex of stigma sub- 
equal to or less than distance from its tip to tip of wing, except in anomoptera; 
(c) two submarginal cells, lengths along posterior margins subequal; (d) sec- 
ond submarginal cells along longitudinal axis of wing much longer than 
broad; (e) middle tarsus of male longer than hind tarsus; (f) horizontal por- 
tion of propodeum impunctate with various sculptural patterns or smooth and 
shiny with depressions; (g) metasomal terga with conspicuous, although not 
necessarily complete, bands of light hair along posterior margins (1X). 


Femate. Length 5-10 mm. Integument sculptured or smooth, unsculp- 
tured (120X). 

Head. Integumental background color black, occasionally metallic. Light 
colored areas: (4) paraocular area; (2) clypeus; (3) labrum, sometimes ab- 
sent; (4) supraclypeal area, except usually absent in Verbenapis; (5) suban- 
tennal plate, often absent, and variable intraspecifically; (6) mandible, basal 
position, or absent, tip reddish or reddish brown. (7) Scape, pedicel, and dor- 
sal surface of flagellomeres 1-4 brown to black, ventral surface usually much 
lighter with darker color of flagellomeres 1-3 (sometimes 4) encircling part 
way ventrad, flagellomeres 5-10 of uniform color ventrally. (8) Hair of head 
variable from black to white, but hair of gena always white. (9) Punctures 
usually finer than in corresponding male. (10) Puncture size and density 
variable. (11) Frontal line with lower portion moderately elevated, narrowly 
sulcate, rarely weakly carinate. (12) Clypeus evenly convex or with lower 
half of disc flattened or medially concave, slightly to strongly protuberant 
beyond anterior edge of eye in profile. (13) Inner orbits convergent below: 
orbital convergence ratio 1.05-1.30. Facial fovea shallow, indistinct to deep, 
distinct; narrow, linear to broad, ovoid. (14) Galea variable in length, shape, 
and sculpture; galeal gap absent to about 3.8 mow. Glossa cylindrical, flabel- 
late (flat, truncate, flabellum absent in squamifera, peninsularis, and anomop- 
tera group of subgenus Perissander. (15) Head width/head length, 1.2-1.5. 
(16) Median apical border of clypeus below orbit by 0.5-1.3 mow. (17) Ra- 
tios of eye length, mio, and flagellar length variable. (18) Interocellar, ocel- 
locular, antennocular, interantennal variable with respect to each other. (19) 
Ocellolabral subequal to or greater than clypeal width. (20) Clypeocellar 
greater than outer subantennal sutural. Outer subantennal sutural 2 times or 


A Revision oF THE BEE GENUS 769 


more length of inner subantennal sutural. (21) Basal labial palpomere from 
1.5-6.0 times length of others combined. (22) Flagellomere 1 longer than 2, 
2 shorter than 3, 3 shorter than 4. Flagellum with maximum width subequal 
to mow. Flagellar length 1.8-2.8 times length of scape, relatively uniform 
within a subgenus. 

_Mesosoma. (23) Integumental background color black, occasionally metal- 
lic. Light colored areas: medially interrupted stripe along posterior dorsal 
border of pronotum except in Verbenapis; pronotal lobe, often; scutellar crest, 
often. (24) Scutal and scutellar hair variable in length, density, and color. 
Mesepisternal hair white, usually longest of mesosoma, flowing, with numer- 
ous minute branches. Posterior half of scutellum with many long, erect hairs. 
Metanotal hair white to fulvous, with long, medial hairs directed posteriorly 
but curving upward apicad. A narrow strip of white to fulvous, densely 
branched, minute hairs clustered along scutellar crest. Metanotum with simi- 
lar strip along posterolateral border, of greater extent, hair sparser, longer. 
Scutellar and metanotal hair pads absent or obsolete. Propodeal hair white. 
(25) Scutal disc with punctures and character of interspaces variable. (26) 
Dorsal enclosure of propodeum completely smooth to variously sculptured, 
only rarely with a few punctures. (27) Legs with light color usually the same 
as on face. Foreleg with light coloration. (28) Middle leg with light colora- 
tion. Spur no more than 0.8 times basitarsal length, pectination variable (ab- 
sent in squamifera). (29) Hind leg dark except in the closely related species 
pectidis, timberlaket, and bernardinensis. (30) Tegula usually transparent, at 
least posteriad, often with anteroapical patch of light color. Humeral plate 
white to brown. (31) Wing iridescent, colorless or faintly smoky to brown in 
the apical region beyond cells. Costal vein progressively darker apicad. Sub- 
costal vein brown, darkest in the wing. Stigma slightly wider, or narrower, 
than prestigma, width including costal vein in both cases. (32) Marginal cell 
6-9 and 3-4 variable with respect to 9-wt, see Fig. 5. 

Metasoma. (33) Integumental background color of terga black (except 
orange to reddish on some terga of anomoptera), occasionally with metallic 
tints, of sterna black to brown. (34) Tergal hair bands white or fulvous, 
denser laterad. Band of tergum 1 usually sparse or interrupted medially. Sub- 
erect hair of discs of terga 4-5 black to white. Prepygidial and pygidial fim- 
briae smoky to white. (35) Tergum 1 with punctures variable. (36) Tergum 
2 with punctures of median area finer, denser than tergum 1. Pygidial plate 
present, distinct, narrowly rounded apically. (37) Sternal color testaceous to 
black, light coloration absent. Sternum 6 usually with a median, clear (rarely 
extremely dark) circular or subcircular area in the apical sclerotized portion 

(indistinct in some specimens). 


Mate. Length, 4-9 mm. Integument as in female. 
Head. Integumental background color as in female. Light colored areas: 


770 Tue UNIversiry SCIENCE BULLETIN 


(1) paraocular area; (2) clypeus, except for narrow, testaceous to black apical 
border and two tiny, testaceous to black wedge-shaped marks always present 
near dorsolateral corners of median portion; (3) labrum, rarely absent; (4) 
supraclypeal area (sometimes absent in Verbenapis); (5) subantennal plate 
(absent in Verbenapis); (6) as in female. (7) As in female but of lighter hue 
than corresponding female. (8,9,10) As in female. (11) Frontal line with 


lower portion moderately to strongly elevated, either narrowly sulco-carinate 
or sharply carinate. (12) As in female. (13) Inner orbits more strongly con-_ 


vergent below than in female: orbital convergence ratio about 1.15-1.45. Fa- 
cial fovea obsolete to distinct; smaller than that of female; shape similar to 
that of female. (14) As in female. (15) Head width/head length about 1.10- 
1.50. (16, 17) As in female. (18) Interocellar variable with respect to ocel- 
locular, both greater than either antennocular or interantennal; antennocular 
greater than interantennal. (19,20,21) As in female. (22) As in female but 
flagellomere 1 sometimes subequal to or shorter than 2. Flagellum with maxi- 
mum width subequal to mow. Flagellar length 2.8-4.2 times length of scape. 

Mesosoma. (23) As in female except all subgenera exclusive of Verbenapis 
have some species with more or less extensive light color on sterna. (24) 
Scutal, scutellar, mesepisternal and metanotal hair similar to that of female. 
Scutellar and metanotal hair pads present in Calliopsis s.s., absent in other 
subgenera. (25,26) As in female. (27) Legs with light color usually the same 
as on face. Foreleg with more extensive light coloration than in female. (28) 
Middle leg with more extensive light coloration than in female. (29) Hind 
leg with much more extensive light coloration than in female except in Ver- 
benapis where leg may be completely brown or have only basitibial plate with 
light coloration. (30,31,32) As in female. 

Metasoma. (33,34) As in female. (35) Tergum 1 with punctures of 
median area as in female but rarely sparse. (36) As in female but often less 
distinct. (37) As in female but rarely with some yellow. (38) Genitalia with- 
out gonostyli, penis not fused with penis valves. Sterna and genitalia distine- 
tive for the subgenera as described. 


Discussion. The diagnosis excludes the distantly related genus Perdita by 
the combination of characters (b,c,d), and usually (a); the closely related 
Nomadopsis by the combination of (a,b,g); and the equally closely related 
Hypomacrotera by the combination of (b,e,g). 


Keys To THE Species oF Calliopsis 


The following keys are separated by sexes because of dimorphism in all 
species. Every attempt was made to avoid characters requiring dissection, but 
this was not always possible. The section on Descriptive Terminology and 
Methods should be consulted before use of the keys, with special attention be- 


| 


A Revision oF THE BEE GENUS 771 


ing given to the part on measurements. It has been possible for the most part 
to place related species together, and this should help in the final determina- 
tion which might have to be based upon the description of the species. 

The male terminalia are illustrated by drawings which portray their aver- 
age condition, but minor variations are to be expected. In some instances, an 
arrow points to a particularly significant condition which is mentioned in the 
keys or in the descriptive treatment of the species. 


2(1). 


B(2). 


Mates 


Dorsal area of propodeum unsculptured, highly polished, with 
median depression with longitudinal, low ridge giving im- 
pression of twin pits; subantennal plates black (subgenus 
VGRICLIE SD) | sa a OREN Raters | oS Z 

Dorsal area of propodeum sculptured, or, if polished, lacking 
kind of depression described above; subantennal plates vari- 
IIS Un, C06 hen cone a Re 5 

Tegula with opaque white patch on outer anterior portion; 
pygidial plate with sides at angle of less than 35°, length of 
margined portion of plate about 2.0 times width at base; depth 
of incision between lobes of sternum 6 twice width of a 
Holceemeree ee ye Gk micheneri 

Tegula without white coloration; pygidial plate with sides at 
angle of 40° or more, length of margined portion of plate less 
than 1.5 times width at base; depth of incision between lobes 
of sternum 6 equal to or less than width of a lobe 

Hind basitarsus light brown to dark brown; elevation bordering 
anterior edge of posterior depressed margin of metasomal ter- 
gum 1 absent medially (use 20 or less); punctures of medial 
area of dorsum of tergum 1 finer than mesoscutal punctures, 
IMO sthyy2 Or aMore pWay. aan 24)! e211 oee eee nebraskensis 

Hind basitarsus whitish or cream colored; elevation bordering 
anterior edge of posterior depressed margin of metasomal ter- 
gum 1 entire; punctures of medial area of dorsum of tergum 1 
as large or larger than mesoscutal punctures, mostly 1-2 pwa.. 4 

Clypeus entirely cream colored with two small brownish trian- 
gular clypeal dots; supraclypeal area usually with a cream dot: 
eye length exceeding minimum interocular distance: hind- 
wainigel ene thy 2./-3.3 mums 2 as Se ee ee hirsutifrons 

Clypeus cream colored except for triangular brown dorsolateral 
corners; supraclypeal dot absent; eye length less than mini- 
mum interocular distance; hindwing length 3.5 mm. or 


oS) 


ROTC?) * tn tie ti se 1 eee a Oe ee ee .. verbenae 


Tie Tue University SCIENCE BULLETIN 


5(1). Without velvety patches of dense, short, moss-like hairs on the 
lateral portions of the scutellum and metanotum (Perissander 
and Calhopsima wins) Soaps ee 

With velvety patches of dense, short, moss-like hairs on the lat- 
eral portions of the scutellum and metanotum (Calliopsis 
SS) de ae Re eee eee 

6(5). Sternum 6 with a pair of long, mesolateral, posteriorly directed, 
directed, subacicular processes; ratio of length of middle tibia 
to length of middle basitarsus less than 0.95; mesopleural 
punctures very fine, some indistinct, mostly more than 2 pwa; 
galea usually not exposed beyond closed mandibles, but may 
extend as much as 1 mow in gilva; penis valve broad, width 
one-third or more of length; penis short, scarcely reaching to 
midlength of penis valve (Pertssander) — 1 = eee 

Sternum 6 with a pair of short, mesolateral, variously directed, 
variously shaped (but never subacicular) processes; ratio of 
length of middle tibia to length of middle basitarsus greater 
than 1.00; mesopleural punctures large, deep, distinct, mostly 
less than 2 pwa; galea always exposed beyond closed mandi- 
bles, length exposed usually more than 2 mow; penis valve 
narrow, width about one-eighth of length; penis long, almost 
reaching apex of penis valve (Calliopsima) _.-------------- 

7(6). At least terga 1-3 reddish orange; forewing with tp bent 
abruptly posteriad and drawn out into a rounded apex (Fig. 


10 


BD) scars ees ete, ae RA anomoptera 


Terga 1-3 black, or black with metallic tints; forewing with tip 
normal’. (Pig, Myicseee Ae ee 

8(7). Thoracic dorsum and terga 1-4 with brassy tints (30); wing 
tip distinctly brown to the naked eye; sternum 8 in ventral 

view with apical portion paddle-shaped, the paddle portion 


8 


about as lone as broad... =.25235 2 ee rogerl 


Thoracic dorsum and terga 1-4 without metallic tints (30); 
wing tip clear to smoky to the naked eye; sternum 8 in ventral 
view with apical portion distinctly longer than broad _.. 

9(8). Scutal punctures with smooth, shiny interspaces (30%); hind 
tarsomere 2 cylindrical to club-shaped in dorsal view; eye 


length about one-sixth greater than mio .....22 2 ee gilva 


Scutal punctures with interspaces roughened; hind tarsomere 2 
equilaterally triangular in dorsal view; eye length subequal 


to MiO .....<...:cock. teats we ei limbus 


10(5). Scape with ventral surface largely yellow 2... ere 
Scape with ventral surface entirely’dark 222 


11 


11(10). 


12(11). 


13(12). 


14(13). 


15(10). 


16(15). 


17(16). 


A ReEvIsION OF THE BEE GENUS 773 


Scape entirely yellow, except sometimes with a small triangle of 
light brown with base at mesolateral apex of scape, apex of tri- 
angle attenuate ventrally and not reaching middle of scape; 
metanotal hair pads dark brown to black (pale brown in 
Baton Rouge, La., specimen), small, separated by 3 mow; 
hind femur often with a brown patch posteriorly . andreniformis 

Scape partly yellow on ventral surface and sometimes on dorsal 
surface, too, with at least a large area of the mesolateral apical 
corner brown or black (if brown area is apparently small, then 
flagellum much longer than head); metanotal hair pads tan or 
gray, or if brown then separated by less than 1 mow; hind 


I TVOOUID NIU), ee cee an eae PR Fy Eee A Ce Sete JL 12 
Tegula with a yellow patch; tiny species with hindwing length 

(oritorewing 1-9) trom! 1.90 to 2.30 mm squamifera 
Tegula brown; larger species with hindwing length (or fore- 

wvimeel-9))§2 65: minor lancer). 7 ba eee 13 
Metanotal hair pads tan to pale brown, separated by more than 

SPAM ON epee sence Ne Ree) Pu JAE hondurasica 
Metanotal hair pads brown or gray, separated by less than 1 

TN ONY meses ea ees oD a LE te ee Se ee eee ee 14 


Flagellar length subequal to head length, about one-fifth longer 
than head width, about 3.5 times scape length; metanotal hair 
pacsibrowis sonora, Mexicoi -...04- 2 3 sonora 

Flagellar length about one-fourth longer than head length, sub- 
equal to head width, about 4.0 times scape length; metanotal 
hair padsicray; southeastern Arizona 2.22 ae empelia 

Integument with metallic cobalt blue tints on dark frons, vertex, 
and thoracic dorsum, and metallic green tints on metasomal 
terga; light markings of face, pronotal lobe, and interrupted 
posterior pronotal stripe, white or cream color; light markings 
giles wellow:stateor Mexico... eee mourel 

Integument either non-metallic or with brassy tints especially on 
lower frons, vertex and thoracic dorsum, and with or without 
brassy tints on metasomal terga; light markings of face, pro- 
notal lobe, interrupted posterior pronotal stripe, and legs con- 


colorouss yellow 4.21.25 ie ee ee cee 16 
Metanotal hair pads ligulate, confined to metanotum, separated 

byptwo-thirds Or Mere ION. so eee ee ee teucril 
Metanotal hair pads large, subquadrate or oval, covering dorsal 

propodeum, contieuous,at least posterioly 2) 17 


Integument black; scutellar hair pads separated by less than 2 
mow; metanotal hair pads subquadrate; galea aciculate, 


774 Tue UNiversiry SCIENCE BULLETIN 


smooth, shiny, length exposed beyond closed mandibles about 
three times galeal gap; punctures of disc of scutum fine, about 
2 pwa, interspaces simooth sees eek ce ee rhodophila 
Integument black with brassy tints, especially on head and 
thoracic dorsum; scutellar hair pads separated by more than 
2 mow; metanotal hair pads oval; galea narrowly rounded 
but not aciculate, uniformly pebbled, length exposed beyond 
closed mandibles subequal to galeal gap; punctures of disc of 
scutum larger, about 1 pwa, interspaces shiny but finely rough- 
énéd; more'so anteriad.. 3 eee eee grant 
18(16). Pedicel with anterolateral surface yellow; scape entirely yellow 
or at most with brown apical rim or brown line on posterior 
surface; all coxae with at least a spot of yellow about half the 
area of median ocellus 20000. 2 19 
Pedicel brown (1 specimen of bernardinensis with anterolateral 
surface yellow) ; scape, posteriorly and at least mesoapical cor- 
ner anteriorly, brown; all coxae brown, or at most a tiny patch 
of yellow on the foré Coxaé 22... 20 
19(18). Scape entirely yellow or at most with light brown apical rim; 
ratio of maximum length of middle tibia to maximum length 
of middle basitarsus 1.00-1.08; scutal punctures not visible 
from directly above because of profuse branching of dense, 
short (about 1 mow), hairs (30%); punctures of scutal disc 
mostly less than 1 pwa 2.222.542: 52 ee pectidis 
Scape yellow except for brown apical rim and narrow brown 
triangle with apex attenuated toward base of scape on pos- 
terior surface; ratio of maximum length of middle tibia to 
maximum length of middle basitarsus 1.11-1.29; scutal punc- 
tures visible from directly above, hairs with very short 
branches, longer (mostly 2 mow) (30); punctures of scutal 


disc:mostly 1-2 pwa 22.::2:4:<2e a timberlaket 
20(18). Scape entirely brown to black 22 21 
Scape with yellow on anterior surface .2..2-2-2 25 

21(20). Front apicotarsus yellow; marginal cell 6-9 about one-sixth to 
one-fifth longer than 9-wt. —...0:s:s.0s i232 22 

Front apicotarsus brown; marginal cell 6-9 about one-fourth to 
one-third longer than 9:wt) 220200 23 


22(21). Expanded midlateral portion of metasomal sternum 8 bearing 
sharp points at posterior corners; posteroventral projections of 
base of genital capsule small, short, rounded —.......- chlorops 

Expanded midlateral portion of metasomal sternum 8 smoothly 


23(21). 


24(23). 


25(20). 


26(25). 


27 (26). 


-28(25). 


A REVISION OF THE BEE GENUS 775 


rounded at posterior corners; posteroventral projections of 
base of genital capsule large, long, sharply pointed crypta 
Tegula with yellow macula; tip of humeral plate yellow; ex- 
panded midlateral portion of metasomal sternum 8 smoothly 
rounded at posterior corners; anterior surfaces of front and 
hind tibiae entirely yellow; genital capsule with four postero- 
Ventralaprojections trom base... nen ee, quadridentata 
Tegula without yellow macula; humeral plate entirely brown; 
expanded midlateral portion of metasomal sternum 8 bearing 
sharp (sometimes tiny!) points at posterior corners; anterior 
surfaces of front and hind tibiae with large areas of brown; 
genital capsule with two posteroventral projections from base 24 
Outer surface of hind basitarsus brown; basal labial palpomere 
with length 3 or more times combined length of remaining 
palpomeres (1.14:0.36); length of galea exposed beyond closed 
mandibles about 4 times galeal-gap se Ze hurae 
Outer surface of hind Reena yellow with brown border; 
basal labial palpomere less than 2.6 times combined length of 
remaining palpomeres (0.78 :0.34); length of galea exposed be- 


yond closed mandibles less than 2 times galeal gap kucalumea 
Twin mesolateral posterior projections of metasomal sternum 6 
flat (readily observable on intact specimens) _-.------- 26 


Twin mesolateral posterior projections of metasomal sternum 6 
bent distinctly ventrad thus forming a pair of short to rather 
OM eyorOne setae ae 0 ee 28 

Front and middle apicotarsi bright yellow; anterior surface of 
hind tibia yellow; anterior surface of hind basitarsus yellow, 
often with partial or complete brown border... rozeni 

Front and middle apicotarsi testaceous to brown; anterior sur- 
face of hind tibia half or more brown; anterior surface of hind 
basitarsusibrowi <5... 5 2) wel ol ae ee ee 27, 

Length of galea exposed beyond closed mandibles 3.0-4. 5 times 
galeal gap; interantennal more than 1.8 times galeal gap; mow 
less than maximum flagellar diameter; tegula and tip of hum- 
eral plate with yellow maculation; Colorado Desert .... deserticola 

Length of galea exposed beyond closed mandibles less than 
25 times galeal gap; mow greater than maximum flagellar 
diameter; tegula and humeral plate brown; southwestern 
alifornia’ 42-.5..2.0 250s oe ls ee pugionis 

Hair bands of metasomal terga 1-4 complete, dense, snow- 
white, hairs erect with somewhat decumbent apices, profusely 
branched, integument below bands not or barely visible from 


776 Tue UNIversity ScIENCE BULLETIN 


above; punctures of entire middle third of metasomal tergum 
1 uniformly crowded, about one-third pwa; metasomal ster- 
num 7 with median portion broad apically (Fig. 78); south- 
westermiCalitormiays 25:0 205 eee ee ee bernardinensis 
Hair bands of metasomal terga 1-4 never all complete, hairs 
sparse, whitish, appressed, branches virtually nonevident 
(30%), integument below readily visible from above; punc- 
tures of entire middle third of metasomal tergum 1 not un1- 
formly crowded, more than one-half pwa; metasomal sternum 
7 with median portion narrowed apically (Fig. 59); not 
known from) Califormia 22.0). 24. ee 29 
29(28). Fore tibia with posterior surface yellow or yellow at apex and 
base with a patch of brown medially; hind tibia with posterior 
surface yellow or yellow with a median patch of brown, 
rately brown patch at tibial apex 2 30 
Fore tibia with posterior surface brown or brown except for 
basal area subequal to 2 times area of median ocellus; hind 
tibia with posterior surface brown or brown with yellow outer 


Margin widening into a yellow basal area 1... 2 32 
30(29). Dorsal surface of metasomal tergum 1 with a wide, impunctate, 
shiny area 0.5-1 mow in length adjacent to upper rim of an- 

terior declivity 2. <.0.:0- 0522 ee uzteca 
Dorsal surface of metasomal tergum 1 punctate all the way to 

upper rim of anterior declivity ee 3 


31(30). Medial half of dorsal enclosure of propodeum with ridges 
strongly vermiform; punctures of dorsal median fifth of meta- 
somal tergum 1 large, deep, distinct, contiguous; volsella 
with posterolateral corner bent sharply dorsad, hence called 
“hooked”; metasomal sternum 6 with lateral margin at base of 
each posterorior projection slightly swollen (Fig. 84) —... unca 

Medial half of dorsal enclosure of propodeum with ridges 
straight or slightly bowed; punctures of dorsal median fifth 
of metasomal tergum 1 small, usually shallow, often eccentric 
0.5 or more pwa; volsella with posterolateral corner not 
“hooked”; metasomal sternum 6 with lateral margin at base of 
each posterior projection not swollen (Fig. 64) coloradensts 

32(29). Hind ubia with posterior surface all brown; metasomal sternum 
6 with lateral margin at base of each posterior projection 
slightly swollen (Fig: 88). 22 crypta 

Hind tibia with posterior surface brown with a yellow border 
along outer margin usually widening into a yellow area basad; 


34(33). 


2). 


3(2). 


A REVISION OF THE BEE GENUS FIL. 


metasomal sternum 6 with lateral margin at base of each pos- 
ichlom jrofectionenot swollen 2 2s (ieee ee 33 
Dorsal surface of metasomal tergum 1 with a wide, impunctate, 
shiny area 0.5-1 mow in length adjacent to upper rim of an- 
HeTi@reclecliyity ieee... Re. Act 4 ci ae eee ee ee A azteca 
Dorsal surface of metasomal tergum 1 punctate all the way to 
Upper timvot anterior declivity 1.2.0). ee at 
Metasomal tergum 1 with large, coarse punctures less than 1 
pwa laterad; about 8 punctures per 0.01 mm? in middle of ter- 
gum; dorsal enclosure of propodeum with a carinate posterior 
border, median portion of about 1 mow with quite vermiform 
ridges, remaining ridges longitudinal, relatively straight .. chlorops 
Metasomal tergum 1 with small, fine punctures mostly more 
than 1 pwa, about 10 punctures per 0.01 mm? in middle of ter- 
gum; dorsal enclosure of propodeum with at most a weak 
carina only along median portion of posterior border, median 
portion of about 1 mow with longitudinal ridges relatively 
straight and similar to remaining ridges coloratipes 


FEMALES 


Dorsal area of propodeum unsculptured, highly polished, with 
median depression with longitudinal, rounded ridge giving 
the impression of twin pits; subantennal plate black or dark 


brown: (Subgenus Verbenapis) 2 
Dorsal area of propodeum sculptured, without such a depres- 
sion; subantennal plate variable in color 0.0.0 5 


First metasomal tergum with a shallow concavity in median 
dorsal area, sharply sloped anteroventrally; punctures of me- 
dian dorsal area of tergum 1 very few, 3 or more pwa on a 
very shiny ground; metasomal tergum 2 with a distinct 
median convex bump; fore basitarsus, ratio length/width, 
5h ee ME eee weet rN nebraskensts 

First metasomal tergum without such a concavity, the dorsal 
area a smooth, continuous curve from side to side; punctures 
of median dorsal area abundant, 2 or less pwa on a moderately 
shiny ground; metasomal tergum 2 without a distinct median 
convex bump; fore basitarsus, ratio length/width, 5.5 or more 3 

Mandible with basal half brown or black, or with a tiny, indis- 
tinct yellowish spot at extreme outer base; dorsal median 
margin of clypeal cream coloration straight or shallowly 
CONCAVE veritas SE Foi Be cnet ee | ieee ns woe ee ee a pe ver benae 


5(1). 


6(5). 


Z(G): 


NON, 


9(8). 


10(8). 


THe UNIvERSITY SCIENCE BULLETIN 


Mandible with basal fourth to half cream colored; dorsal median 
margin of clypeal cream coloration strongly convex upward 4 
Tegula transparent dark brown; hindwing length less than 3.6 
mm; fore basitarsus, ratio length/width, 8.0-9.5 ....... hirsutifrons 
Tegula transparent smoky straw color with opaque cream macu- 
lation anteriorly; hindwing length equal to or more than 4.0 
mm; fore basitarsus, ratio length/width, 4.5-5.5 _.... micheneri 
Prepygidial and pygidial fimbriae smoky, reddish-brown, or 
black; disc of metasomal terga 3-5 bearing numerous short, 


black hairs <...4:t.a to dee 6 
Prepygidial and pygidial fimbriae fulvous or white; disc of 
metasomal terga 3-5 bearing fulvous or white hairs —....... 23 
Galea in repose extending beyond closed mandibles 1.25 mow or 
MONG sitions Mae A ee 7 
Galea in repose extending beyond closed mandibles 1 mow or 
less (Perissander, and Calliopsts s.s.im patt)) 2. 15 


Medium sized species, hindwing length 3.80 mm or less, mio 
less than 1.36 mm; maxillary palpomere 2 shorter than, or at 
most equal to maxillary palpomere 3 (30); lateral portion of 
clypeus brown to black (Calhopssss. )_... 8 

Large sized species, hindwing length 3.90 mm or more, mio 
more than 1.50 mm; maxillary palpomere 2 distinctly longer 
than maxillary palpomere 3 (30>); lateral portion of clypeus 


yellow (Calhopsima, in part) —.220.2402002-2 32 
Tegula with at least a dot of yellow ........ PN ccc 9 
Tegula without yellow, straw color to black 202222222 10 


Disc of scutum with fine punctures more than | pwa, interspaces 
shiny, becoming only faintly roughened anteriorly; flagellar 
length subequal to mio; yellow area of tegula larger than 
median ocellus;: Texas, Coahuilat 2. helenae 

Disc of scutum with coarse punctures 1 pwa, interspaces sha- 
greened, becoming strongly roughened anteriorly; flagellar 
length one-sixth shorter than mio; yellow area of tegula 
smaller than median ocellus; southeastern Arizona _.......... empelia 

Galea highly polished, tapered to a sharp point (20), length 
exposed beyond closed mandibles about 3 mow; metasomal 
tergum 1 highly polished with very fine, sparsely distributed 
punctures about 3 pwa on the disc, posterior area impunctate 
(20) cae ae ede rhodophila 

Galea lightly pebbled, at least on apical fourth, and tapered to a 
blunt, rounded tip, length exposed beyond closed mandibles 
15-25 mow; metasomal tergum 1 dull, or if polished, then 


11(10). 


12(11). 


13(12). 


14(11). 


15(6). 


16(15). 


A Revision oF THE BEE GENUS 779 


punctures fairly evenly distributed, usually 2 or less pwa on 
disespesterion areal punctate (20) eee ee 11 
Scutum and metasomal tergum | with ground dulled by abun- 
dant fine roughening; basal labial palpomere with stout setae 
ventrally; frons without brassy, metallic tints (hondurasica 
AU OU IN OA) heres ew eS eee er le ele 12 
Scutum and metasomal tergum 1 with ground smooth, some- 
times shiny, or if roughened, then roughening confined to an- 
terior portion of scutum; frons with weak, brassy, metallic 
LATTES perenne alent neo, cd oe Sc acct ne tO 14 
Hairs of stipes straight or slightly curved from bases to apices; 
punctures of scutum coarse; less than 0.5 pwa on disc with 
punctures becoming much more crowded anteriorly .......... _ zord 
Hairs of stipes, at least many of them, more or less abruptly bent 
over at the tip, frequently curled at the tip; punctures of scu- 
tum fine, mostly more than 1 pwa with punctures becoming 
somewhat more crowded anteriorly |...) we ee 13 
Flagellar length about one-tenth longer than mio; larger species, 
hindwing length about 3.5 mm, head length 1.62 mm or more, 
interteeular 39! mm or more ......0. ee hondurasica 
Flagellar length subequal (one-twentieth more or less) to mio; 
smaller species, hindwing length about 2.9 mm, head length 
153 mm or less, intertegular 1.33 mm or less _...2.........----2--- sonora 
Declivity of metasomal tergum 1 with a dull, satiny surface; 
punctures of disc of metasomal tergum 1 fine, shallow, with 
moderately shiny interspaces, actual diameter about 8 » or 
less, 2 pwa on anterior portion of dorsum; Rocky Moun- 
tain states, southeastern Canada, central and eastern United 
SSticlie speaes Rey sceae shee Se CR 2 es ts UE olen = eee andreniformis 
Declivity of tergum 1 finely lineolate with a high polish; punc- 
tures of disc of metasomal tergum 1 fine, deep, with highly 
polished interspaces, actual diameter about 12 », 1 pwa on 
anterior portion of dorsum; central Mexico, Arizona, New 


Mexico. Colorado n= a fe teucrit 
Mesotibial spur, untoothed. (60 ><). ae es 16 
Mesotibial spur with at least 2 apical teeth 200 17 


Mesotibial spur three-fourths or more of length of middle basi- 
tarsus; middle tibial length subequal to middle basitarsal 
lengths Arizona: cs <5 25 oregano ae ee squamifera 

Mesotibial spur less than 0.6 times length of middle basitarsus; 
middle tibial length 0.3 times longer than middle basitarsus; 
Bay ae Califoniniay > tite eo erect taa eta ey Rae peninsularis 


780 


1713). 


18(17). 


19(18). 


20(19). 


21(20). 


i) 
ine) 


(20). 


Tue University SCIENCE BULLETIN 


Mesotibial spur with 2-4 coarse teeth on apical 0.4, the basal 0.6 
perfectly. bdre’s:.2.2i. ei ee See gilva 
Mesotibial spur‘with 7'or more fime teeth 22 eee 18 
Metasomal terga 1-4 largely orange or reddish-orange . anomoptera 
Metasomal terga 1-4 black or metallic brassy 19 
Scutum with all interspaces between punctures finely, distinctly 
roughened; marginal cell 6-9 less than 9-wt; ratio of length of 
flagellum to lengthofiscape; 2/0-2:8 se ee limbus 
Scutum either with al! interspaces between punctures smooth, 
or only roughened anteriorly; marginal cell 6-9 more than 
9-wt; ratio of length of flagellum to length of scape, 2.0-24 20 
Smaller species, hindwing length (forewing 1-9 gives same 
measurement) 2.0-2.7 mm; median portion of dorsal enclosure 
of propodeum with fine lines originating posteriorly, fanning 
out anterolaterally toward base of propodeum 21 
Larger species, hindwing length (forewing 1-9 gives same 
measurement) 3.0 mm or larger; dorsal enclosure of propo- 
déum not as above 0. 22..c.n cen ee 22 
Brassy, metallic colored integument on frons, scutum, scutellum, 
and metasomal terga ...52....2. rogert 
Black or dark brown integument on frons, scutum, scutellum, 
and metasomiall terval 22). syphar 
Mesotibial spur with 6-8 teeth; facial fovea long, linear, 4-5 times 
longer than wide, deeply impressed with distinct lateral mar- 
gin; posterior portion of metasomal tergum 1 with mirror-like 
polish, virtually impunctate; southeastern Arizona and south- 
western New Mexic0.2.0....2.2)22... = fulgida 
Mesotibial spur with about 15 teeth; facial fovea moderately 
long, broader at midlength, 3-4 times longer than wide, shal- 
lowly impressed with relatively indistinct lateral margin; pos- 
terior portion of metasomal tergum 1 shiny, completely punc- 
tates Michoacan: «2.5... ..2 8.252020. 6 yalea 
Tibiae with dorsal or anterior surfaces mostly yellow —........... 24 
Tibiae with dorsal surfaces bearing yellow color only at extreme 
bas€S: #2062 A oe ee 
Scutal hair dense, appearing nap-like (10%) and obscuring 
punctation because of its profuse, relatively long, branches; 
scutal length one-tenth or more greater than eye length; 
front basitarsal length equal to or greater than front 
apicotarsts: 0. is et te eS pectidis 
Scutal hair thin (10) with punctures readily visible; scutal 


25(23). 
26(25). 


27(26). 


28(25). 


29(28). 


30(29). 


32(7). 


A ReEvISsION oF THE BEE GENUS 781 


length subequal to eye length; front basitarsal length about 


one-tenth shorter than front apicotarsus tumberlakei 
At least basal fifth of mandible brown or black . 26 
At least basal fifth of mandible yellow or cream color... 28 
Both long and short scutal hairs fulvous 21, 
Long scutal hairs brown, short ones fulvous crypta 
Metasomal tergum 1 with large, deep, punctures, sparser 

OiVEG es II gy Pe ee Ree Meg Es rozeni 


Metasomal tergum 1 with very fine punctures, not sparser medi- 
aliysoncaudenser medially 2:22. 0). eee coloradensis 

Metasomal tergum 1 with punctures very dense, crowded medi- 
ally, less than 0.5 pwa, about 11 punctures per 0.01 mm? of 


Tie cMIGNAAL Ae Se Coe. 129 ll ee Penna ee eel bernardinensis 
Metasomal tergum 1 with very sparse punctures, from 0-3 punc- 
tunessper O0I m= of median area... =e ee 2 


Labrum brown; the two lateral brown bars of median disc of 
clypeus joined by a strip of brown at least along dorsal side of 
preapical groove of clypeus; facial light color a deep, rich 
femonbyellow ent .e 5.0 nue ae a ee pugionis 
Labrum cream color, pale yellow, or rarely, mostly brown; the 
two lateral brown bars of median disc of clypeus not veined 
facial light color cream or very pale yellow «0. 30 

Punctures of metasomal tergum 1 moderately large, usually 
sparse medially, but may be rather regularly spaced medially 
becoming sparser laterally; horizontal enclosure of propodeum 
distinctly longer medially than laterally, bearing longitudi- 
nally vermiform ridges which are fairly readily distinguish- 
alle anon cach other (30x )) eo chlorops 

Punctures of metasomal tergum 1 very fine, always sparser 
medially than laterally, interspaces highly polished; horizontal 
enclosure of propodeum with median third of about equal 
length throughout, bearing either straight, longitudinal ridges 
well separated from each other, or densely packed, very fine, 
longitudinally vermiform ridges difficult to distinguish from 
eachother: (30< )! .2uc:55 22 2252 ee Sl 

Disc of clypeus with tiny, twin, wedge-shaped, light brown spots 
or very light brown, narrow bars; basal labial palpomere one- 
tenth to one-fifth shorter than clypeocellar coloratipes 

Disc of clypeus with large, twin, dark brown bars; basal 
labial palpomere one-tenth to one-fifth longer than clypeo- 
Cella i en Se ee rt: Lee ee deserticola 

Largest species of ihe genus, hindwing length 4.9 mm; basal 


Tue UNiversiTy SCIENCE BULLETIN 


“I 
(oe) 
ie) 


labial palpomere one-sixth longer than clypeocellar; forewing 

3-4 about one-tenth less than 9-wt; disc of metasomal tergum 

l-finely roughened 2.32 2 eee hurdi 
Moderately large species, hindwing length 3.8-4.4 mm; basal 

labial palpomere subequal to clypeocellar to one-tenth less; 

forewing 3-4 subequal to or less than 9-wt; disc of metasomal 


tercumelsmooth; shiny. = kucalumea 


Subgenus CALLIOPSIS Smith 


Calliopsis Smith, 1853, Catalogue of Hymenoptera in the British Museum 1:128; Michener, 1951, 
in Muesebeck et al., U.S. Dept. Agric., Monogr., No. 2:1103; Mitchell, 1960, North Carolina 
Agric. Exp. Sta. Tech. Bull. No. 141:288-294. 

Type species. Calliopsis andreniformis Smith, 1853, by designation of Ashmead, 1899, Trans. 
Amer. Ent. Soc., 26:85. 

This subgenus is closer to Perissander than to the other subgenera. Its 
closest South American relatives seem to be apparently unnamed Argentinian 
species of Liopoeum. Two groups of species fall together rather naturally: 
the andreniformis group, including teucru, mouret, rhodophila, wilda, and 
helenae; and the hondurasica group, including sonora, empelia, squamifera, 
peninsularis, and probably zora. 

This is the most widespread subgenus. It occurs from Panama to Canada, 
and from the eastern United States west to Utah and southwest to Nevada, 
California, and the tip of Baja California Sur. The species favor Legumi- 
nosae, especially Trifolium, Melilotus, Medicago, and Psoralea, with next 
choices being Verbenaceae and Compositae. 

Calliopsis differs from the other subgenera in the male by the prominent 
scutellar and metanotal hair pads, and in the female by the combination of the 
length of galea exposed beyond closed mandibles being greater than 1 mow 
(except in squamifera), the prepygidial and pygidial fimbriae being smoky to 
dark reddish brown, and in both sexes by the tumid paraocular area. 

Femate. Length, 5.0-7.5 mm. 

Head. Yellow areas: (1) paraocular area below a line originating between 
middle of outer subantennal suture and near upper end of suture and extend- 
ing diagonally upward ending on orbit approximately at level of upper rim of 
antennal socket, usually below level of facial fovea (above in helenae), or 
yellow area reduced to small patch along orbit; lower inner paraocular area 
adjoining junction of frontoclypeal suture and outer subantennal suture 
tumid, most strongly in andreniformis, least in hondurasica; (2) clypeus with 
a median longitudinal band originating at frontoclypeal suture, ending at or 
slightly above clypeal apex; lateral areas brown except in helenae and some 
rhodophila which have several blotches of yellow; (3) absent on labrum ex- 
cept on apex of labral plate in zora and helenae; (4) supraclypeal area pen- 


A Revision oF THE BEE GENUS 783 


tagonal or semilunar with apex approximately at midlevel of antennal socket; 
(5) subantennal plate highly variable from totally dark to totally light (6) 
absent on mandible. (7) Scape and pedicel black, ventral surfaces of flagel- 
lomeres 1-2 black, of 3-4 black with median portions tan, of remaining ones 
tan; dorsal surface of basal flagellomeres dark brown, flagellomeres becoming 
progressively lighter brown apicad. (8) Hair of vertex mixed colorless and 
light brown, or mixed fulvous and dark brown, of frons white or fulvous, of 
clypeus pale brown to black and coarser than on other areas, of gena white. 
(10) Punctures along ocellocular line fine, interspaces variable; impunctate 
area lateral to posterior ocellus shiny to exceedingly dulled by roughening; 
punctures of frons with interspaces sometimes metallic, shiny to very dull 
(30). (11) Frontal line with lower portion a narrow sulcus rising gradu- 
ally to a summit on frontal prominence between antennal sockets slightly 
above their midline. (12) Clypeus with punctures of disc somewhat eccentric, 
or at least with sloping sides like a volcano cone. Clypeus with disc convex 
medially. Projections beside median apical emergination of clypeus pointed. 
(13) Inner orbits slightly convergent below. Facial fovea shallow to deep, 
linear to broadened medially and narrowed at both ends. (14) Galeal length 
variable, length of galea exposed beyond closed mandibles variable from ab- 
sent to within 1 mow of base of prementum. (15) Head width/head length 
1.25-1.60. (17) Eye length variable with respect to mio, but equal to or less 
than flagellar length (except much greater in empelia), and subequal to outer 
subantennal sutural. (18) Interantennal to antennocular variable, interanten- 
nal less than 2 mow. (19) Ocellolabral equal to or greater than mio. (21) 
Basal labial palpomere 1.5-3.8 times length of others combined. (22) Flagel- 
lomere 1 with length variable with respect to flagellomere 9. Flagellar length 
1.8-2.5 times length of scape. 

Mesosoma. (23) Light colored areas: medial interruption of pronotal 
stripe 1-4 mow; apex of pronotal lobe; scutellar crest. (24) Scutal and scutel- 
lar hairs of two kinds, longer ones fulvous to brown, shorter ones fulvous. 
(25) Scutal disc with punctures usually distinct, interspaces highly polished 
to extremely dulled by minute roughening. (26) Dorsal enclosure of propo- 
deum variously sculptured, always with some longitudinal ridges. (27) Legs 
with light color the same as on face. Foreleg with basal spot of yellow on tibia, 
and sometimes on extreme apex of femur. (28) Middle leg with basal spot of 
yellow on tibia, sometimes on extreme apex of femur; spur with many fine 
teeth (bare in sguamifera). (29) Hind leg brown. (30) Tegula brown, with 
anterior yellow spot in some species. Humeral plate brown. (31) Wing 
smoky apically beyond cells. Stigma brown. (32) Marginal cell 6-9 subequal 
to, or longer than, and 3-4 shorter than 9-wt. 

Metasoma. (34) Tergal hair bands white to fulvous (orange fulvous in 
some andreniformis), dense, appressed. Band of tergum 1 broadly inter- 


784 Tue University SCIENCE BULLETIN 


rupted, of tergum 2 narrowly interrupted; other tergal bands entire (all at 
10). Suberect hair of discs of terga 45 dark brown to black. Prepygidial 
and pygidial fimbriae smoky to dark brown. (35) Tergum 1 with punctures 
of median area variable, uniformly to irregularly distributed to partially im- 
punctate; interspaces exceedingly dull to highly polished; punctures smaller 
than on scutum. Declivity of tergum 1 usually finely lineolate, somewhat dull 
(shiny in rhodophila and most specimens of teucrit). 


Mate. Length, 4.0-6.5 mm. 

Head. Yellow areas (white in mourei and squamifera): (1) paraocular 
area below a relatively straight, or a dorsally convex line, from upper outer 
edge of antennal rim tangent to or slightly indented by lower end of facial 
fovea, ending on orbit at or slightly above lower end of facial fovea; lower 
half of paraocular area tumid, most strongly so near junction of outer suban- 
tennal suture and frontoclypeal suture; (2) clypeus except a clear to black 
border on lateral portion of apex; frontoclypeal suture yellow to black; (3) 
labrum, except for 3 small circular or subcircular clear to black spots, one in 
each laterobasal corner and a median one approximately in center of labral 
plate; (4) supraclypeal area apex from slightiy below midlevel of antennal 
socket to about 0.5 mow above upper edge of socket; (5) subantennal plate 
except sometimes black near junction of outer subantennal suture and fronto- 
clypeal suture; (6) mandible basal fourth or more, some species with anterior 
and posterior border brown to black. (7) Scape and pedicel all yellow, yellow 
and brown, or all black; flagellomeres as in female, except some species with 
yellow on ventral surface of first two or three flagellomeres, remainder having 
successively more tan toward apex of flagellum. (8) Hair of vertex variable, 
of frons white or fulvous, of clypeus colorless, fulvous, brown, or black. (10) 
Punctures of face variable. (11) Frontal line with lower portion a sharp high 
carina or a relatively low, extremely finely sulcate carina. (12) Clypeus with 
punctures of disc as in female except finer; median portion strongly protuber- 
ant, usually protruding 0.17 mm or more in front of eye in lateral view. (13) 
Inner orbits strongly convergent below. Facial fovea usually with ill-defined 
mesal border. (14) Galeal length and length of galea exposed beyond closed 
mandibles as in female, galeal gap greater than inner subantennal sutural, but 
less in rhodophila. (15) Head width/head length 1.15-1.38. (17) Eye length 
greater than mio, considerably less than flagellar length. (18) Interantennal 
greater than antennocular, and less than 2 mow. (21) As in female. (22) As 
in female except flagellar length about 3.0-4.1 times length of scape. 

Mesosoma. (23) As in female. (24) Scutal and scutellar hair all light, or 
in some species with longer hair brown, other hair as in female. Scutellar 
hair pads well developed, never contiguous, composed of densely packed, 
short, profusely branched hairs occupying lateral portion of scutellum, some- 
times part of median portion also. Metanotal hair pad much greater in extent 


A ReEvIsION oF THE BEE GENUS 785 


than on scutellum, contiguous medially in granti and rhodophila. (25) Scutal 
disc with punctures as in female. (26) Dorsal enclosure of propodeum vari- 
ously sculptured, with at least some longitudinal ridges which may be some- 
what vermiform or reticulate. (27) Legs with light color the same as on face 
except in mouret and squamifera. Foreleg color pattern variable. (28) Middle 
leg color pattern variable. (29) Hind leg color pattern variable. (30) Tegula 
brown, with anterior yellow spot in squamifera and rarely in andreniformis. 
Humeral plate brown. (31) Wing as in female but tip much darker, pale 
brown. Stigma as in female. (32) Marginal cell 6-9 subequal to or greater 
than, and 3-4 equal to or less than 9-wt; 11-12 shorter than 13-14. 
Metasoma. (34) As in female but hair bands much less dense. (35) As in 
female but size of punctures variable with respect to scutal punctures. 
Declivity of tergum 1 variable. (36) Pygidial plate ill-defined, disc irregularly 
concave, usually with median apical emargination. (37) Sterna brown to 
black (rarely, with median preapical yellow spot in andreniformis). Sternum 


fd 
m4) 
"Ai SYLVIA E. TayLog 


Fic. 6. Lateral view of male Calliopsis (C.) andreniformis Smith. 


786 Tue UNIversiTy ScIENCE BULLETIN 


5 produced posteriorly into a long, tapered point. Sternum 6 with apical mar- 
gin produced into a pair of long, aciculate, spine-like processes. Sternum 8 
with a relatively broad, apical paddleike portion. (38) Sterna and genitalia 
as illustrated (Figs. 8-45). 


CALLIOPSIS (CALLIOPSIS) ANDRENIFORMIS Smith 
(Figs. 5,6,8-13; Map 1) 


Calliopsis andreniformis Smith, 1853, Catalogue of Hymenoptera in the British Museum, 1:128, 
female; Cockerell, 1897, Canad. Ent., 29:290; Cockerell, 1898, Trans. Amer. Ent. Soc., 
25:196: Robertson, 1898, Trans. Acad. Sci. St. Louis, 8:48; Ashmead, 1899, Trans. Amer. 
Ent. Soc., 26:85 (type species of Calliopsis); Bridwell, 1899, Trans. Kansas Acad. Sci., 16:210; 
Cockerell, 1899, Ent. News, 10:3; Graenicher, 1905, Bull. Wisconsin Nat. Hist. Soc., 3:159; 
Lovell and Cockerell, 1906, Psyche, 13:113; Swenk and Cockerell, 1907, Ent. News, 18:178; 
Cockerell, 1909, Ann. Mag. Nat. Hist., (8)4:28; Tucker, 1909, Trans. Kansas Acad. Sci., 
22:282: Smith, 1910, Ann. Rep. New Jersey State Mus. 1909: 691; Graenicher, 1911, Pub. 
Mus. City Milwaukee Bull., 1:238; Crawford, 1913, Canad. Ent., 45:271; Gibson, 1913, 44th 
Ann. Rep. Ent. Soc. Ontario: 20; Viereck, 1916, im Britton, Connecticut Nat. Hist. Surv. 
Bull., 22:722; Washburn, 1919, Minnesota Agric. Exp. Sta., Jour. series, No. 156:229; Stevens, 
1919, Canad. Ent., 51:210; Britton, 1920, Connecticut Geol. Nat. Hist. Surv. Bull. 31:345; 
Rau, 1922, Trans. Acad. Sci. St. Louis, 24(7):33; Robertson, 1922, Psyche, 29:168; Reinhard, 
1924, Ann. Rep. Smithsonian Inst., 1922, Publ. Number 2738:371-373 (biol.); Lovell, 1925, 
Maine Naturalist, 5:7; Leonard, 1926, Cornell Univ. Agric. Exp. Sta. Mem., 101:1021; Rob- 
ertson, 1926, Psyche, 33:118 (biol.); Robertson, 1928, Flowers and Insects, p. 10+ (biol.); 
Hendrickson, 1930, Iowa State Coll. Jour. Sci., 4(2): 163 (biol.); Pearson, 1933, Ecol. 
Monogr., 3:387, 409-11, 418 (biol.); Graenicher, 1935, Ann, Ent. Soc. Amer., 28:303; Cock- 
erell, 1936, Amer. Mus. Novitates, 831:3; Ainslie, 1937, Canad. Ent., 69(5): 97-100: (biol.); 
Brimley, 1938, Insects of North Carolina, p. 453; Procter, 1938, Biological Survey of the 
Mount Desert Region [Maine], 6:442 and 1946, 7:504 (biol.); Lovell and Lovell, 1939, 
Rhodora, 41:185 (biol.); Timberlake, 1947, Pan-Pac. Ent. 23:29; Crandall and Tate, 1947, 
Jour. Amer. Soc. Agronomy, 39:161-163 (biol.); Michener, 1947, Amer. Midl. Nat., 38:477; 
Stevens, 1950, North Dakota Agric. Exp. Sta. Bimonthly Bull., 12:93-94; Michener, 1951, in 
Muesebeck er al., U.S. Dept. Agric. Monogr. No. 2:1103; Rozen, 1951, Jour. Kansas Ent. Soc., 
24:142; Mitchell, 1956, Jour. Elisha Mitchell Sci. Soc., 72(2)207 (biol.); Montgomery, 1957, 
Proc. Indiana Acad. Sci., 66:129 (biol.); Wille, 1958, Ann. Ent. Soc. Amer., 51(6):544 
(anat.); Mitchell, 1960, North Carolina Agric. Exp. Sta. Tech. Bull. No. 141:288-289, 291- 
294; Byers, 1962, Jour. Kansas Ent. Soc., 35:320 (biol.). 

Calliopsis flavipes Smith, 1853, Catalogue of Hymenoptera in the British Museum, 1:128, male. 

Calliopsis lepidus Cresson, 1878, Trans. Amer. Ent. Soc., 7:68, female; Cockerell, 1898, Trans. 
Amer. Ent. Soc., 25:196; Cresson, 1916, Mem. Amer. Ent. Soc., 1:122. 

Panurgus vernalis Provancher, 1882, Nat. Canad., 13:204; 1883, Faune entomologique du Canada, 
Hyménoptéres, 704. 

Calliopsis flavifrons; Banks, 1912, Ent. News, 23:107 (misidentification). 

Calliopsis rhodophila; Stevens, 1919, Canad. Ent., 51:210 (misidentification ). 


The female of andreniformis was considered to resemble females of the 
genus Andrena, hence its name. Its closest relative is C. teucrit. The male of 
andreniformis is distinguished readily from that of teacrit by its entirely yel- 
low scape, whereas that of teacrii is brown. The female of andreniformis is 
distinguished from teverti only with difficulty. The anterior slope of meta- 
somal tergum 1 has a satiny sheen on the surface in andreniformis but has a 
finely lineolate polish in tewcrii; the integument of the anterior portion of the 
dorsum of the scutum is brassy colored in andreniformis, non-metallic in 
teucril. 


A RevIsIon oF THE BEE GENus 787 


Femare. Length, 7.0 mm (type 7.4 mm); forewing length, 4.8 mm (type 
5.0 mm) ; hindwing length, 3.4 mm; clypeal length, 0.54 mm (type 0.51 mm); 
scutal length, 1.31 mm. 

Head. Yellow areas: (1) paraocular area, except lowermost corner, below 
a line originating about midlevel of antennal socket and extending diagonally 
upward to a point on the orbit slightly below or slightly above level of upper 
rim of antennal socket and below lower margin of facial fovea (to almost all 
black); area below level of middle portion of frontoclypeal suture tumid: (2) 
clypeus with narrow, longitudinal stripe medially (to all black, rarely); (4) 
supraclypeal area in a semilunar shape (absent, rarely); (5) absent on suban- 
tennal plate (to completely yellow). (8) Hair of vertex mixed colorless and 
brown, of frons fulvous, both longer than long hairs of scutum; of clypeus 
black. (10) Punctures along ocellocular line 2-3 pwa with smooth interspaces; 
impunctate area laterally adjacent to posterior ocellus finely roughened; punc- 
tures of frons adjacent to upper portion of frontal line deep, distinct, less than 
I pwa, interspaces brassy, faintly roughened (30>). (12) Clypeus with punc- 
tures of disc approximately equal in size to the average frontal puncture, deep, 
3-4 pwa, interspaces finely roughened (to barely discernibly roughened in 
part) (30). (13) Orbital convergence ratio as 1.31:1.21, 1.08 (type 1.33:1.26, 
1.05). Facial fovea with indistinct mesal border, broader medially, narrowed 
above and below. (14) Galea shiny despite minute pebbling; length inter- 
mediate between antennocellar and clypeocellar; galeal gap exceeds length of 
galea exposed beyond closed mandibles. (15) Head width/head length 2.16: 
1.75, 1.23 (type 2.21:1.84, 1.20). (17) Eye length, mio, and flagellar length as 
1.28:1.21:1.26 (type 1.28:1.26:N.A.). (18) Interocellar, ocellocular, antennocu- 
lar, and interantennal as 0.31:0.39:0.31:0.32 (type 0.29:0.39:0.32:0.31). (19) 
Ocellolabral greater than clypeal width 1.45:1.36, 1.07. (20) Clypeocellar to 
outer subantennal sutural as 0.90:0.71, 1.27 (type 0.90: N.A.). (21) Basal 
labial palpomere 3.6 (type 3.0) times length of others combined, 40:11 (type 
36:12). (22) Flagellar length about 2.3 times length of scape, 1.26:0.54 (type 
29 ':0.57). 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal 
stripe about 2.5 times mow. (24) Scutal and scutellar hairs of two kinds, 
longer ones brown (to fulyous). Hair of metanotum fulvous. (25) Scutal 
disc with punctures fine, about 2 pwa, interspaces shiny, becoming exceed- 
ingly fine and crowded to less than 1 pwa anteriad, interspaces finely rough- 
ened and faintly (to strongly) brassy colored. (26) Dorsal enclosure of pro- 
podeum relatively dull with fine, irregularly longitudinal ridges, median 
portion not distinctly bordered posteriorly, length not much exceeding length 
of lateral area. (27) Legs with light color the same as on face. (28) Spur with 
about 25 fine, short teeth: spur length half length of middle basitarsus, 0.51: 
0.80 (type 0.51:0.75). (30)Tegula dark brown. Humeral plate dark brown. 


788 Tue UNiversity ScrENCE BULLETIN 


(32) Marginal cell 6-9 subequal to and 3-4 less than 9-wt, 1:00:0.80:1.04 (type 
1.05:N.A.:1.02). 

Metasoma. (35) Tergum 1 with punctures of median area extremely fine, 
about 2 pwa, interspaces shiny (or with a silken sheen). Declivity of tergum 
1 with a fine, silken sheen. 


Maze. Length, 6.0 mm (type 6.2 mm); forewing length, 5.0 mm (type 
4.6 mm); hindwing length, 2.86 mm; clypeal length, 0.44 mm (type 0.53 
mm); scutal length, 0.95 mm. 

Head. Yellow areas: (1) paraocular area below an ascending diagonal 
line from upper outer edge of antennal rim tangent to (or indented by) lower 
end of facial fovea, ending on orbit slightly above level of end of facial fovea; 
(2) clypeus; (4) supraclypeal area extending well above level of upper anten- 
nal rims (0.2-0.5 mow); (5) subantennal plate and adjacent antennal rim; 
(6) mandible, basal two-thirds; (7) scape (sometimes with small, brown, 
trianguliform area on inner upper corner). (8) Hair of vertex, frons, and 
clypeus fulvous (clypeus sometimes with a few apically light brown hairs to 
many completely light brown hairs), of gena white. (10) Punctures along 
ocellocular line minute, about 3 pwa (to 1 pwa), interspaces shiny; impunc- 
tate area adjacent to lateral border of posterior ocellus finely roughened; 
lower, dark portion of frons heavily roughened. (11) Frontal line with lower, 
yellow, carinate portion ending at midlevel of antennal sockets. (12) Clypeus 
strongly protuberant with medial portion between clear clypeal dots raised 
somewhat conically (sometimes raised as an indistinct, short longitudinal 
ridge). (13) Orbital convergence ratio 1.12:0.87, 1.29. (14) Galea dull, com- 
pletely pebbled, length as in female; galeal gap subequal to length of galea 
exposed beyond closed mandibles. (15) Head width/head length as 1.85:1.53, 
1.21 (type 2.16:N.A.). (17) Eye length, mio, and flagellar length as 1.11:0.87: 
1.38 (type 1.21:1.04:N.A.). (18) Interocellar, ocellocular, antennocular, and 
interantennal as 0.26:0.34:0.17:0.26 (type 0.29:0.39 :0.24:0.24). (19) Ocellolab- 
ral greater than clypeal width, 1:22:1.14, 1.07 (type 1.43:N.A.). (20) Cly- 
peocellar to outer subantennal sutural as 0.77:0.58, 1.33 (type 0.90:N.A.). 
(21) Basal labial palpomere 1.5-1.8 times length of others combined. (22) 
Flagellar length 3.0-3.4 times length of scape, 1.38 :0.43. 

Mesosoma. (23) Yellow areas: large patch between middle coxal cavities 
truncate anteriorly 1 mow or more posterior to front coxae; smaller area be- 
tween hind coxal cavities contiguous anteriorly with mesosternal patch; 
medial interruption of pronotal stripe about 2 mow. (24) Scutal and scutellar 
long and short hairs fulvous (white in faded specimens). Scutellar hair pad 
dark brown, barely extending to mesal end of yellow of scutellar crest. Meta- 
notal hair pad dark brown, mesal margins separated by about 3 mow. (25) 
Scutal disc with punctures fine, 1 pwa or less, interspaces finely roughened but 
relatively shiny (30%), becoming very fine and crowded anteriorly. (26) 


A Revision oF THE BEE GENUS 789 


Dorsal enclosure of propodeum with fine, reticulate ridges medially, and 
distinct, fine roughening laterally. (27) Foreleg yellow (sometimes with base 
of coxa brown, apicotarsus testaceous) ; mediotarsomeres somewhat dorsoven- 
trally flattened, widths about 0.75 times lengths. (28) Middle leg yellow 
(often with base of coxa brown, femur with basal brown spot, apicotarsus 
testaceous) ; mediotarsomeres cylindrical, widths about half of length. (29) 
Hind leg yellow (often with variable amounts of brown on base of coxa, 
inner surface of trochanter, base of femur especially inner or posterior surface; 
posteroventral surface of basitarsus and entire apicotarsus testaceous), medio- 
tarsi similar to front mediotarsi but larger. (30) Tegula brown with anterior 
yellow spot (to all brown). Humeral plate brown (to testaceous). (32) 
Marginal cell 6-9 greater than, and 3-4 considerably less than 9-wt, 0.80:0.66: 
0.77. 

Metasoma. (35) Tergum 1 with punctures of median area very minute, 
smaller than on scutum, 1-2 pwa, interspaces shiny. 


Tyre Marerrar. Holotype female of andreniformis, No. 17.a.1798, and 
holotype male of flavipes, No. 17.a.1799, both from East Florida and both col- 
lected by E. Doubleday in 1838, are in the British Museum (Natural History). 
The type of C. lepidus, from Georgia, collected by Morrison, is in the Acad. 
emy of Natural Sciences of Philadelphia, Pennsylvania. The type of vernalis 
is presumably in the Provincial Museum, Quebec, Canada. The above de- 
scriptions are primarily based upon specimens from northern Georgia with 
comparative measurements for the female type of andreniformis and the male 


type of flavipes. 


Distrisgution. This is the most widespread species in the United States. 
It occurs in the entire eastern United States and southeastern Canada and the 
Maritime Provinces. With the exception of a mating pair of specimens from 
the southwestern corner of Utah, all records are from east of the Rocky Moun- 
tains and its western boundary is the Rocky Mountain Front Range from 
Montana to Colorado, southeast to central Oklahoma, south to East Texas and 
the Texas Gulf coast. 

It has been collected from early April to late September in the southern 
states, but mostly in June and July. In the northern states it has been taken 
from early June to early October, but mostly in July. Doubtless, further col- 
lecting will find it with an extreme as late in the southern states as in the 
northern ones. 

The highest altitude at which it is recorded is about 5700 feet at the foot of 
the Front Range of the Rocky Mountains at Boulder, Colorado. 

Ecologically, it occurs in bare, clayey soil (loam, clay loam, silt loam) 
where small-flowered leguminous plants, especially of the clover group, thrive 
during its active season. It is likely to be found on almost any school campus, 


790 Tue UNIversiry SCIENCE BULLETIN 


playground, or sports field within its range. For this reason I suggest the 
common name, the Campus Bee. 

I have not examined the type material in the British Museum but I have 
examined the type of C. lepidus as well as Prof. T. B. Mitchell’s homotypes 
of andreniformis and flavipes. In addition, I have examined approximately 
5,000 specimens from the localities shown in Map 1. 


The following western localities seem worth indicating in detail: CoLorapo: Alamosa; Boul- 
der; Golden, South Table Mountain; Longmont; White Rocks, near Valmont. Montana: Forsyth. 
Uran: Parowan Canyon. The other localities are cited in a thesis on deposit in the library of The 
University of Kansas. 


Grocrarnic Variation. The most striking variation in the males is in the 
coloration of middle and hind trochanters and femora. Males from east of the 
Mississippi exhibit a higher incidence and greater amount of dark color on 
these parts. The amount of yellow increases to a maximum with a specimen 
from near Parowan, Utah. The most obviously variable feature in the females 
is the amount of yellow on the face. It parallels the situation in the male legs. 
No males exhibit all brown middle or hind femora, and no females exhibit all 
black faces. Some females from North Carolina and Vermont, however, have 
the yellow reduced to a mere spot on the paraocular area and a line or dot on 
the disc of the clypeus. Clypeal hairs of both male and female show a similar 
trend although they are subject to considerable fading after preservation. 

Bionomics. This is fully discussed in the section on biology and ecology of 
C. andreniformts. 

Frower Recorps. Achillea, Ailanthus altissima, Ammania coccinea, 
Amorpha canescens, Anaphalis margaritacea, Asclepias, Aster ericoides vil- 
losus, Bidens artstosa, Boltonia asteroides, Brassica, Castanea pumila, Ceano- 
thus americanus, Chrysanthemum leucanthemum, Chrysopsis, Cleome, Con- 
volvulus arvensis, C. sepium, Coreopsis palmata, C. tripteris, Cryptotaenia 
canadensis, Cucumis, Desmodium marilandicum, D. paniculatum, Dianthera 
americana, Epilobium, Erigeron, Eryngium yuccifolium, Gerardia tenuifolia, 
Geum album, Gillenia stipulacea, Hedeoma pulegioides, Hedyotis nigricans, 
H. purpurea, Helenium, Hypericum perforatum, Lindernia dubia riparia 
(Raf.) (=Ilsyanthes riparia), Lespedeza capitata, L. procumbens, L. repens, 
L. reticulata, Ligustrum, Lippia lanceolata, Lycopus sinuatus, Lythrum ala- 
tum, Medicago sativa, Malva neglecta, M. rotundifolia, Melilotus alba, M. 
officinalis, Nepeta cataria, Oenothera laciniata, Oxalis dillenu, O. stricta, 
Penstemon, Petalostemon candidum, Polygala sanguinea, Polygonum buxt- 
forme, P. convolvulus, P. pennsylvanicum, Portulaca, Potentilla monspelien- 
sis, P. recta, Prunella vulgaris, Prunus, Psoralea onobrychis, P. tenuiflora, P. 
tenuiflora floribunda, Pycnanthemum flexuosum, P. pilosum, P. virginianum, 
P. lanceolatum, Raphanus sativus, Rhus glabra, Rosa (wild), Rubus, Rud- 
beckia triloba, Serinea oppositifolia, Sisymbrium repardum, Solanum caro- 
linense, Solidago canadensis, Spiranthes gracilis, Stachys palustris, Stellaria, 


A ReEvIsION OF THE BEE GENUS 791 


) 
* 
\2 


1: { 
fel ke lat 
BE, Her mentee maculatum 
Ae 
Sr ae -} 
helenae ~ 


seers 
Ss! Aa Lect 
Cy eee 


Map |. Map showing the known distributions of Calliopsis (Calliopsis) andreniformis Smith, 
C. (C.) rhodophila Cockerell, C. (C.) helenae Shinn, C. (C.) teucrii Cockerell, C. (C.) 
mouret Shinn, and C. (C.) granti Shinn. The presumptive collection locality for Acampto- 
poeum maculatum (Smith) is also shown. 


Strophostyles pauciflora, Stylosanthes biflora, Symphoricarpos, Tamarisk, Tri- 
folium hybridum, T. pratense, T. repens, Verbena bracteata, V. bracteosa, V. 
hastata, V. stricta, V. urtictfolia, Verbesina helianthoides, Veronica spicata, 
Victa. 

CALLIOPSIS (CALLIOPSIS) TEUCRII Cockerell 


(Figs. 14-17; Map 1) 


Calliopsis teucrii Cockerell, 1899, in Cockerell and Porter, Ann. Mag. Nat. Hist. (7) 4:412, 
female; Cockerell, 1906, Trans. Amer. Ent. Soc., 32:299; Michener, 1951, 7” Muesebeck ef al., 
U.S. Dept. Agric., Monogr. No. 2:1103. 

Calliopsis lepida var. a Cockerell, 1901, Ann. Mag. Nat. Hist., (7) 7:128, male, female. 

Calliopsis lepida Cockerell, 1906 (not Cresson, 1878), Trans. Amer. Ent. Soc., 32:299. (mis- 
identification ) 

Calliopsis sp., Wille, 1956, Univ. Kansas Sci. Bull., 38:453, 474-475, female (thoracic muscu- 
lature). 


Closest to andreniformis, the females of which are virtually identical with 
teucri1, and are best separated by the key characters. The male of teucri has 
a brown scape, but that of andreniformis is yellow. 


792 Tue Universiry SciENCE BULLETIN 


Fremate. Length, 6.7 mm; forewing length, 4.6 mm; hindwing length, 3.2 
mm, clypeal length, 0.46 mm, scutal length, 1.02 mm. 

Head. Yellow areas: (1) paraocular area with a small patch bordering 
orbit below midlevel of antennal socket (to yellow, except lowermost corner, 
below a line originating at middle of, or slightly above, outer subantennal 
suture and extending diagonally upward ending on orbit at or below midlevel 
of antennal socket, well below lower rim of facial fovea, and lower than 
andreniformis in most cases); (2) absent on clypeus (to broad longitudinal 
median stripe from frontoclypeal suture to about 0.5 mow above preapical 
groove); (4) supraclypeal area with 2 minute dots (to full semilunar area 
reaching to slightly below midlevel of antennal socket); (5) absent on suban- 
tennal plate (to completely yellow). (8) Hair of type faded. Hair of fresh 
specimens: on vertex mixed brown and fulvous, on frons fulvous, on clypeus 
blacker, coarser, thicker than in andreniformis. (10) Punctures along ocel- 
locular line 2-3 pwa, interspaces smooth; impunctate area adjacent to lateral 
border of posterior ocellus almost imperceptibly roughened (30) and quite 
shiny; punctures of frons adjacent to upper portion of frontal line as in 
andreniformis except larger and more distinctly roughened (30). (11) 
Frontal line with lower portion as in andreniformis except more distinct. (12) 
Clypeus with punctures of disc larger than frontal punctures, deep, 2-3 pwa, 
interspaces shinier than in andreniformis with extremely fine roughening. 
(13) Orbital convergence ratio as 1.19:1.11, 1.08. Facial fovea with distinct 
mesal border, linear but slightly wider below, narrower than in andreniformis. 
(14) Galea somewhat shiny but finely pebbled; length as in andreniformis 
(except in Flagstaff, Arizona, specimens where length exceeds clypeocellar 
1.19:0.94); galeal gap to length of galea exposed beyond closed mandibles, 
variable. (15) Head width to head length as 1.96:1.55, 1.26 (1.20-1.28). (17) 
Eye length, mio, and flagellar length as 1.11:1.11:1.21(1.24:1.19:1.29). (18) 
Interocellar, ocellocular, antennocular, and interantennal as 0.29 :0.36 :0.28 :0.28 
(0.31 :0.39:0.31:0.31). (19) Ocellolabral greater than clypeal width, 1.29:1.19, 
1.08. (20) Clypeocellar to outer subantennal sutural as 0.83:0.63,1.32. (21) 
Basal labial palpomere about 4.0(3.9-4.8) times length of others combined. 
(22) Flagellar length about 2.4 times length of scape, 1.21:0.49. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal stripe 
about 2 mow. (24) As in andreniformis except long scutal and scutellar hairs 
always brown. (25) Scutal disc with punctures slightly larger than in 
andreniformis, 1 pwa, interspaces shiny, becoming finer and crowded ante- 
riorly, but less so than in andreniformis, less than 1 pwa, no discernible 
roughening (30%) (to very slight), faintly to strongly greenish metallic 
colored. (26) Dorsal enclosure of propodeum relatively dull, with finer, more 
regularly longitudinal ridges than in andreniformis; median portion more 
distinctly bordered posteriorly, and proportionately longer in relation to lateral 


A ReEvIsION OF THE BEE GENUS 793 


areas than in andreniformis. (27) Legs with light color the same as on face. 
Foreleg with yellow like that of andreniformis. (28) Middle leg and spur as 
in andreniformis except spur length about 0.6 (to 0.7) of length of middle 
basitarsus, 0.47:0.73. (30) Tegula brown. (32) Marginal cell 6-9 longer than, 
and 3-4 shorter than 9-wt, 0.99 :0.83 :0.94. 

Metasoma. (34) Tergal hair bands white. Bands of terga 1-2 as in 
andreniformis. (35) Tergum 1 with punctures of median area fine, mostly 1 
pwa, interspaces highly polished. Declivity of tergum 1 with a high polish. 


Mate. Length, 5.5 mm; forewing length, 4.3 mm; hindwing length, 3.13 
mm; clypeal length, 0.48 mm; scutal length, 0.78 mm. 

Head. Yellow areas: (1) paraocular area as in andreniformis but more 
horizontally truncate above; (2) clypeus as in andreniformis except apical 
border black; (4) supraclypeal area as in andreniformis except height of area 
lowered with decreasing latitude, upper limit of area below middle of anten- 
nal socket near southern limit of range; (5) subantennal plate as in andreni- 
formis. (7) Scape and pedicel brown. (8) Hair on vertex and frons fulvous, 
on clypeus brown (to black), on gena white. (10) As in andreniformis except 
impunctate area usually very shiny despite minute roughening, and punc- 
tures of lower, dark portion of frons larger than in andreniformis. (11) 
Frontal line with lower, carinate portion ending slightly above midlevel of 
antennal socket. (12) Clypeus as in andreniformis. (13) Orbital convergence 
ratio as 1.09:0.85,1.28. (14) Galea dull, completely pebbled, somewhat ligulate 
in repose; galeal gap subequal to (to less than) length of galea exposed beyond 
closed mandibles. (15) Head width to head length as 1.77:1.56,1.13. (17) Eye 
length, mio, and flagellar length as 1.02:0.85:1.53. (18) Interocellar, ocellocu- 
lar, antennocular, and interantennal as 0.27:0.34:0.17:0.26, close to andreni- 
formis. (19) Ocellolabral greater than clypeal width, 1.24:1.11,1.12. (20) 
Clypeocellar to outer subantennal sutural as 0.77:0.61,1.25. (21) Basal labial 
palpomere 1.4-2.0 times length of, others combined. (22) Flagellar length 
about 3.6 times length of scape, 1.53:0.43. 


Mesosoma. (23) Light areas yellow; medial interruption of pronotal 
stripe about 1.5 mow. (24) Scutal and scutellar hair of two kinds, one short, 
fulvous, the other long, brown. Metanotal hair fulvous (to dirty white). 
Scutellar and metanotal hair pads as in andreniformis except metanotal hair 
pads much closer, 1.5 mow between mesal margins. (25) Scutal disc with 
punctures larger than andreniformis, 1 pwa or less, interspaces smooth (to 
finely roughened), becoming crowded anteriad. (26) Dorsal enclosure of 
propodeum with larger, thicker, reticulate ridges medially, with shorter longi- 
tudinal ridges laterally partly hidden by metanotal hair pad. Median portion 
of enclosure distinctly larger and longer than in andreniformis. (27) Legs 
with light color the same as on face. Foreleg with yellow on coxal apex, 


794 Tue UNIversity SCIENCE BULLETIN 


ventral trochanteral surface, basal half, more or less, of femur, tibia entirely 
except for posterior brown patch, tarsus entirely; mediotarsus as in andrent- 
formis. (28) Middle leg colored like foreleg except apicotarsus testaceous; 
mediotarsomeres as in andreniformis. (29) Hind leg colored like middle leg 
except yellow greatly reduced on trochanter, and brown patch present on pos- 
terior tibial surface; apicotarsus testaceous; mediotarsomeres as in andrent- 
formis. (30) Tegula brown. (31) Wing tip more distinctly brown apically 
beyond cells than in andreniformis. (32) Marginal cell 6-9 greater than, and 
3-4 subequal to 9-wt, 0.97 :0.85 :0.85. 

Metasoma. (34) Tergal hair bands white. Bands of terga 1-2 more 
broadly interrupted than in andreniformis. (35) As in andreniformis. (37) 
Sterna brown. 


Tyre Mareriat. Holotype female from Las Vegas, New Mexico, July 11 
(T. D. A. Cockerell), on Teucrium laciniatum, is at the University of Cali- 
fornia, Riverside, California. It has the yellow markings of the face greatly 
reduced compared to usual specimens, and is smaller than the average as well. 

The above description of the male is based primarily on a specimen from 
Coaldale, Colorado, 7800 ft., Aug. 4, 1957 (C. D. Michener). 


Distrigution. Colorado to Oaxaca. The species is widespread in central 
México but no specimens bridge the gap from Fresnillo, Zacatecas to San_ 
Jose, New Mexico. This seems likely to be a collecting bias rather than a real 
gap in distribution. 

Seasonal distribution is June to Aug. 31 with one late record of Sept. 24, 
1938, at Agua Fria, Hidalgo (L. J. Lipovsky). 

The numerous records of the bee at altitudes of 5,000 to 8,000 ft. may indi- 
cate it is a mountain form that may, indeed, be absent from the desert parts 
of Durango and Chihuahua. 


In addition to the type material approximately 305 specimens have been studied from the 
following localities: ARIZONA: Flagstaff (4 mi. N.; 6 mi. W.). CoLorapo: Animas, 6600 ft: 
Coaldale, 7800 ft., Fremont Co.; Eleven-Mile Canyon, 8000 ft., Park Co.; Estes Park; Florissant; 
Ouray, 8500 ft.; Pagosa Springs, 7500 ft.; Ridgway, 7000 ft. New Mexico: Las Vegas; Ruidoso; 
Santa Fe; San Jose; Sapello. Disrrrro FeperaL: Contreras. Duranco: Nombre de Dios, 5900 ft. 
Guerrero: Acapulco. Hiatco: Agua Fria; Epazoyucan; Jacala (21 mi. S.W.; 7, 24 mi. N.E.), 
5000 ft.; Lagunilla; Pachuca (14 mi. S.W.), 7500 ft.; Tepeapulco (3 mi. N.); Tizayuca (13.5 
mi N.E.), 7700 ft. Jatisco: San Juan de los Lagos; Tepatitlan. Mexico: Chapingo; Tepexpan, 
6900 ft. MicHoAcan: Morelia (22 mi. W.), 6800 ft.; PAtzcuaro; Tzitzio (3 mi. N.), 5500 ft. 
Moretos: Cuernavaca. Nuevo Leon: El Cercado (4 mi. W.). Oaxaca: Oaxaca (14 mi. E.), 5000 
ft.; Tamazulapan (2 mi. N.W.), 6000 ft. PuesLa: Chila (5 mi. S.), 5700 ft.; Huachinango (4 mi. 
S.W.), 5700 ft.; Teziutlin (5 mi. N.E.), 5100 ft. QueRETARO: San Juan del Rio (10 mi. E.), 
6500 ft. San Luis Porost: Ciudad del Matz (5 mi. E., 4700 ft.; 20 mi. N.E., 3000 feet). 
TLaxcaLa: Apizaco (8 mi. W.; 8 mi. W.N.W.), 8200 ft. Zacatecas: Fresnillo (9 mi. S.). 


Fiower Recorps. Asclepias, Convolvulus incanus, Heterotheca chrysops- 
idis, Melilotus officinalis, Oxalis, Potentilla, Stylosanthes, Taraxacum, Teu- 
crium laciniatum. 


A Revision oF THE BEE GENUS 795 


CALLIOPSIS (CALLIOPSIS) GRANTI, new species 
(Figs. 18-21; Map 1) 


The species is named in honor of the late Dr. Harold J. Grant, Jr., my long 
time friend and companion in the field. The species is intermediate between 
teucrit and rhodophila, but is closer to the latter. It is easily separated from 
both species by the key characters. 


Mate. Length, 5.2 mm; forewing length, 4.2 mm; hindwing length, 3.02 
mm; clypeal length, 0.39 mm; scutal length, 0.80 mm. Integument of head, 
mesosoma, and first metasomal tergum with faint brassy, metallic tints. 

Head. Yellow areas: (1) paraocular area as described for subgenus, upper 
boundary somewhat convex dorsally; (2) clypeus, dots black; border of lat- 
eral portion of apex black; frontoclypeal suture black. (4) Supraclypeal area, 
summit below midlevel of antennal socket; (5) subantennal plate, except 
lowermost outer corner triangularly black; (6) mandible with anterior and 
posterior borders black. (7) Scape, pedicel, flagellomeres 1 and 2, black. (8) 
Hair of vertex and frons fulvous, of clypeus black. (10) Punctures of upper 
frons larger than in teucri, interspaces shiny. (11) Frontal line with lower 
portion a low, finely sulcate, carina. (13) Orbital convergence ratio as 1.05: 
0.83,1.26. (14) Galea strongly pebbled, but little shiny, shinier than teucriz, 
with rounded, blunt tip; galeal gap estimated to be shorter than length of 
galea exposed beyond closed mandibles, 0.36:0.37. (15) Head width to head 
length as 1.67:1.43,1.17. (17) Eye length, mio, and flagellar length as 0.95: 
0.83:1.19. (18) Interocellar, ocellocular, antennocular, and interantennal as 
0.27 :0.32:0.17:0.27. (19) Ocellolabral equal to clypeal width, 1.12:1.12,1.00. 
(20) Clypeocellar to outer subantennal sutural as 0.73:0.54,1.34. (21) Basal 
labial palpomere 1.4 times length of others combined. (22) Flagellar length 
about 3.2 times length of scape, 1.19:0.37. 

Mesosoma. (23) Light areas yellow. (24) Scutal and scutellar hairs short, 
white and long, brown. Scutellar hair pads black, separated by 0.37 mm, 
slightly more than 2 mow, somewhat farther than in rhodophila. Metanotal 
hair pads black, subquadrate, contiguous posteriorly, separated anteriorly 
leaving a triangular area from which arises a tuft of long white hairs. (25) 
Scutal disc with punctures slightly larger than in fewer, much larger than in 
rhodophila, 1 pwa or less, interspaces shiny. (26) Dorsal enclosure of pro- 
podeum covered by metanotal hair pad except for small median area which 
is abruptly elevated. (27) Legs with light color the same as on face. Foreleg 
yellow except dark brown on basal half of coxa, dorsal half of trochanter; 
anterior half, passing dorsally to posterior three-fourths, of femur, and most 
of posteroventral surface of tibia; mediotarsus testaceous, distitarsus brown. 
(28) Middle leg colored like foreleg. (29) Hind leg colored like middle leg 
except brown patch on posterior surface of tibia reduced and located medially. 


796 Tue University ScrENcE BULLETIN 


(30) Tegula brown. (32) Marginal cell 6-9 greater than, and 3-4 subequal to 
9-wt, 0.94 :0.80 :0.82. 

Metasoma. (34) Tergal hair bands very sparse on all terga. Band of ter- 
gum | represented by only a small lateral group of sparse hairs. (35) Tergum 
1 with punctures of median area scarce, larger than on scutum, irregularly 
distributed, of area immediately laterad larger than on scutum, about 1 pwa, 
interspaces highly polished; punctures shallow, some eccentric, giving pock- 
marked appearance suggesting deformity but bilaterally symmetrical. Decliv- 
ity of tergum 1 finely lineolate, very shiny. 


Type Marertar. Holotype male from Apizaco (8 miles W.N.W.), 
Tlaxcala, 8200 feet, June 18, 1961 (University of Kansas Mexican Expedition), 
on Stylosanthes, is in the Snow Entomological Museum of The University of 
Kansas, Lawrence. 


CALLIOPSIS (CALLIOPSIS) RHODOPHILA Cockerell 
(Figs. 22-25; Map 1) 


Calliopsis andreniformis rhodophila Cockerell, 1897, Proc. Acad. Nat. Sci. Philadelphia, 49:350, 
male, not female; zdem, Bull. Univ. New Mexico, 24:19; idem, 1898, Trans. Amer. Ent. Soc. 
25:196. 

Calliopsis rhodophila; Cockerell, 1898, Bull. Denison Univ., 11:52: Birkman, 1899, Ent. News, 
10:244 (rec., Fedor, Lee Co., Texas, det. Friese); Cockerell, 1902, Amer. Naturalist, 36:810; 
idem, 1906, Trans. Amer. Ent. Soc., 32:299; dem, Bull Amer. Mus. Nat. Hist., 22:440; zdem, 
1921, Amer. Mus. Novitates, 24:14; idem, 1922, 40:4; Timberlake, 1947, Pan-Pac. Ent., 
23:29; Michener, 1951, 2 Muesebeck e¢ al., U.S. Dept. Agric., Monogr. No. 2:1103. 


Closest to teucri and grant but fairly easily separated from them by the 
long, sharply pointed galeae which are slim and tapering towards their apices, 
and the galeae appear to be somewhat converging towards each other. The 
few specimens of this bee I have examined have come from widely separated 
geographic areas, and the species appears to be rare. 


Fremare. Length, 6.8 mm; forewing length, 4.8 mm; hindwing length, 3.43 
mm; clypeal length, 0.51 mm; scutal length, 1.12 mm. 

Head. Yellow areas: (1) paraocular area, except lowermost corner, below 
a sinuous line originating at middle of, or slightly above outer subantennal 
suture and extending upward ending on orbit at or above midlevel of anten- 
nal socket, usually not touching lower border of facial fovea, but if touching 
lower border of fovea then ending on orbit above lower border of fovea, — 
usually below upper rim of antennal socket; (2) clypeus, a median, longi- 
tudinal stripe about as wide as width of yellow supraclypeal area, upper por- 
tion somewhat constricted just below base; lower margin about 0.5 mow 
above clypeal apex; (4) supraclypeal area broadly semilunar to broadly pen- 
tagonal, extending to midlevel, or above, of antennal socket; (5) usually ab- 
sent on subantennal plate, but sometimes on lower half. (8) Hair like andren-_ 
iformis, except clypeal hairs finer. (10) Punctures along ocellocular line 


A Revision oF THE BEE GENUS 797 


sparse, very fine, with shiny interspaces, impunctate area adjacent to lateral 
border of posterior ocellus shiny (30>); punctures of frons adjacent to upper 
portion of frontal line deep, distinct, 1-2 pwa, interspaces smooth, moderately 
shiny (30). (12) Clypeus with punctures of disc larger than frontal punc- 
tures, deep 2-3 pwa, interspaces shiny, highly polished, except faintly rough- 
ened beside subantennal plate. (13) Orbital convergence ratio as 1.28:1.17, 
1.09. Facial fovea linear, distinctly bordered medially. (14) Galea smooth, 
shiny; apex pointed (30>) in contrast to varying degrees of broadness in 
other species of Calliopsis s.s.; length only slightly less than antennocellar; 
galeal gap less than length of galea exposed beyond closed mandibles. (15) 
Head width to head length as 2.11 :1.58,1.33. (17) Eye length, mio, and flagel- 
lar length as 1.19:1.19:1.22. (18) Interocellar, ocellocular, antennocular, and 
interantennal as 0.34:0.34:0.29:0.32. (19) Ocellolabral subequal to clypeal 
width, 1.34:1.36. (20) Clypeocellar to outer subantennal sutural as 0.83:0.70, 
1.19. (21) Basal labial palpomere about 2.0 times length of others combined; 
ventral hairs fine, straight. (22) Flagellar length about 2.3 times length of 
seape, 1.22:0.54. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal 
stripe about 2 mow. (24) As in andreniformis. (25) Scutal disc with punc- 
tures fine, deep, distinct, about 3 pwa, interspaces shiny, becoming crowded 
anteriorly to 1 pwa with shiny interspaces. (26) Dorsal enclosure of propo- 
deum shiny, with obliquely longitudinal ridges laterally, straight medially; 
posterior border carinate, a small medial portion of border raised distinctly 
upward; median portion longer than lateral portion. (27) Foreleg without 
(or with) yellow spot on extreme apex of femur. (28) Middle leg colored 
like foreleg; spur length about 0.6 length of middle basitarsus, 0.49 :0.82. (30) 
Tegula brown. (32) Marginal cell 6-9 greater than, and 3-4 less than 9-wt, 
1.07 :0.88 :1.02. 

Metasoma. (34) Tergal hair bands white. (35) Tergum 1 with punctures 
of median area exceedingly fine, 2-3 pwa anteriorly to impunctate posteriorly, 


interspaces highly polished. 


Mate. Length, 5.9 mm; forewing length, 4.3 mm; hindwing length, 3.10 
mm; clypeal length, 0.46 mm, scutal length, 0.75 mm. 

Head. Yellow areas: (1) paraocular area as in teucri1; (2) clypeus, dots 
pale brown; lateral border of apex black; frontoclypeal suture black; (4) 
supraclypeal area, summit between midlevel and upper border of antennal 
socket; (5) subantennal plate; (6) mandible with anterior and_ posterior 
borders black. (7) Scape and pedicel black; flagellomeres 1-2 partly black. 
(8) Hair on vertex and frons white, on clypeus black. (10) Punctures of 
upper frons largest of the andreniformis group, contiguous, shallow, indis- 
tinct, interspaces dulled by fine roughening. (11) Frontal line with lower 
portion a high, sharp, non-sulcate carina, distinctly higher than other species 


798 Tue University SCIENCE BULLETIN 


of subgenus. (13) Orbital convergence ratio as 1.17:0.88,1.33. (14) Galea 
highly polished, smooth, rather slim, galeal gap slightly more than half (to a 
third) of length of galea exposed beyond closed mandibles. (15) Head width 
to head length as (1.77:1.51,1.17). (17) Eye length, mio, and flagellar length 
as 1.09:0.88:1.48. (18) Interocellar, ocellocular, antennocular, and interanten- 
nal as 0.31:0.32:0.19:0.27. (19) Ocellolabral subequal to clypeal width (1.21: 
1.19), 1.02. (20) Clypeocellar to outer subantennal sutural as 0.78 :0.58,1.35. 
(21) Basal labial palpomere 1.3 times length of others combined. (22) Flagel- 
lar length about 3.5 times length of scape, 1.48:0.43. 

Mesosoma. (23) Light areas yellow. (24) Scutal and scutellar short hairs 
white, long hairs white (sometimes a few fulvous to brown on scutum pos- 
teriorly, and on scutellum). Scutellar hair pads black, apart by 0.31 mm, 
about 2 mow, somewhat closer than in granti. Metanotal hair pads as in 
granti except triangular area with apex more acute, pads larger, more exten-_ 
sive. (25) Scutal disc with punctures finer than in ¢eucri1, mostly 2 pwa, inter- 
spaces shiny, smooth (sometimes faintly roughened). (26) Dorsal enclosure — 
of propodeum as in granti. (27) Legs with light color the same as on face. 
Foreleg colored as in granti. (28) Middle leg like grant:. (29) Hind leg col- 
ored like granti, except tibial patch black, located medially (or apically). (32) 
Marginal cell 6-9 greater than, and 3-4 subequal to 9-wt, 0.94 :0.83 :0.85. 

Metasoma. (34) Tergal hair pads as in granti. (35) Tergum 1 with 
punctures of median area regularly distributed, finer than on scutum, ex- 
tremely fine, about equal in size to those in andreniformis, but much more 
distinct, further apart at 3 pwa, interspaces highly polished. Declivity of ter- 
gum | highly polished. 

Type Marerrat. Holotype male from Santa Fe, New Mexico, July (T. D. 
A. Cockerell, Cockerell No. 3844), on Sphaeralcea, is in the U.S. National 
Museum, U.S. N. M. Type No. 3701. The female allotype, same label data 
except Cockerell No. 3842, U. S. N. M. Type No. 3702, is actually teucrit. 

The above description of the female is based primarily on a specimen 
from Santa Fe, New Mexico, July (T. D. A. Cockerell, Ckll. No. 3843), on 
Sphaeralcea! 

Distrisution. The southwestern United States south to slightly north of 
central Mexico. Seasonal occurrence is from May to August. The highest 
altitudinal collection record is 5900 ft., and other records suggest a bee adapt- 
able to low mountains. 

In addition to the type material, 7 males and 18 females have been studied from the following 
localities; ARIZONA: Chino Valley. CALtForNIA: Kelso (7 mi. S.), Providence Mts., San Bernardino 
Co.; Mt. Laguna, San Diego Co.; Pifion Flat, San Jacinto Mountains, Riverside Co. NEw MExIco: 
Las Vegas; Ruidoso, Lincoln Co.; Santa Fe. Uran: Orderville, Kane Co. Duranco: Nombre de 
Dios, 5900 ft. Zacatecas: Fresnillo (9 mi. S.E.; 1.5, 5, and 9 mi. S.). 

Discussion. Too few specimens are available for positive conclusions, but 
the Arizona specimen, a female, is considerably duller than the others. The 


A RevIsIoN OF THE BEE GENUS 799 


galeal gap in males is proportionately greater in specimens from New Mexico 
(half of length of galea exposed beyond closed mandibles) than in those from 
Utah, California, and Mexico (a third of length of galea exposed beyond 
closed mandibles). Males from New Mexico are shorter than those from other 
states. The hair pads show very little variation. 

Frower Recorps. C. rhodophila has been collected on Sphaeralcea in Cali- 
fornia in May, in New Mexico in July, Aug., Sept., and in Zacatecas in Aug. 
One record for Solidago is from Zacatecas in Aug. 


CALLIOPSIS (CALLIOPSIS) MOUREI, new species 
(Figs. 26-29; Map 1) 


I take great pleasure in naming this distinctive species in honor of Padre 
Jesus S. Moure, C. M. F., of the University of Parana, Curitiba, Brazil, who 
gave me a solid foundation for understanding the panurgine bees of South 
America most closely related to Calliopsis. 

This species may be closest to rhodophila, but is easily distinguished from 
all other species of Calliopsis s.s. by the white facial color in combination with 
the cobalt blue color of the head and mesosoma. 


Mate. Length, 5.2 mm; forewing length, 4.1 mm; hindwing length, 3.01 
mm; clypeal length, 0.44 mm; scutal length, 0.87 mm. Integumental back- 
ground color of head and mesosoma metallic cobalt blue, of metasoma metallic 
greenish. 

Head. White areas: (1) paraocular area as described for subgenus, upper 
boundary dorsally convex; (2) clypeus, dots testaceous; border of lateral por- 
tion of apex black; frontoclypeal suture light; (4) supraclypeal area, summit 
0.5 mow above upper edge of antennal socket; (5) subantennal plate; (6) 
mandible with anterior and posterior border black. (7) Scape, pedicel, and 
flagellomere 1 black; flagellomere 2 partly black. (8) Hair on vertex, frons, 
— clypeus, and gena white. (10) Punctures of upper frons fine, 2-3 pwa, inter- 
spaces shiny, cobalt blue, minutely roughened (30%). (11) Frontal line with 
lower portion a medium-height, non-sulcate carina. (13) Orbital convergence 
ratio as 1.11:0.82,1.35. (14) Galea finely pebbled, dull, width somewhat slim 
as in rhodophila, tip very narrowly rounded, intermediate between rhodophila 
and granti; galeal gap slightly shorter than length of galea exposed beyond 
closed mandibles, 0.36:0.37. (15) Head width to head length as 1.73:1.50,1.16. 
(17) Eye length, mio, and flagellar length as 1.02:0.82:1.39. (18) Interocellar, 
ocellocular, antennocular, and interantennal as 0.32:0.31:0.17:0.26. (19) Ocel- 
lolabral greater than clypeal width, 1.22:1.11,1.11. (20) Clypeocellar to outer 
subantennal sutural as 0.78:0.56,1.39. (21) Basal labial palpomere 1.8 times 
length of others combined. (22) Flagellar length about 3.6 times length of 

scape, 1.39 :0.39. 


800 Tue University SCIENCE BULLETIN 


Mesosoma. (23) Light colored areas: of pronotal stripe, pronotal lobe, 
meso- and metasternum, white; of scutellar crest, yellow. (24) Scutal and 
scutellar hair white. Scutellar hair pads black, widely separated, confined to 
lateral scutellar areas, separated by 0.49 mm. Metanotal hair pads black, large, 
broadly oval, separated by 0.14 mm, less than 1 mow. (25) Scutal disc with 
punctures finer than in rhodophila, distinct, mostly 2 pwa, interspaces shiny, 
smooth. (26) Dorsal enclosure of propodeum with median portion exposed, 
lateral portion only covered anteriorly by metanotal hair pad; seven slightly 
irregularly longitudinal ridges on median portion, interspaces shiny; less 
distinct, lower ridges on lateral portion, interspaces shiny posteriorly, rough- 
ened anteriorly. (27) Legs with light color the same as on face except yellow 
on front and middle coxae. Foreleg colored as in granti except dark tibial 
patch much smaller. (28) Middle leg colored as in granti except dark tibial 
patch much smaller. (29) Hind leg colored as in grant: except dark ubial 
patch much smaller. (32) Marginal cell 6-9 greater than, and 3-4 slightly less 
than 9-wt 0.87 :0.75 :0.80. 

Metasoma. (34) Tergal hair bands as in granti. (35) Tergum 1 with 
punctures of median area regularly distributed, finer than on scutum, slightly 
larger than in rhodophila, 2-3 pwa, interspaces smooth shiny. Declivity of 
tergum | roughened, somewhat shiny. 

Type Materia. Holotype male from Tepexpan, state of México, Aug. 12, 
1954 (University of Kansas Mexican Expedition), is in the Snow Entomo- 
logical Museum of The University of Kansas, Lawrence. 

Discussion. This striking species is the only one of the genus to display 
a metallic cobalt blue integument. Nothing is known of its biology. 


CALLIOPSIS (CALLIOPSIS) HONDURASICA Cockerell 
(Figs. 30-33; Map 2) 


Calliopsis hondurasica Cockerell, 1949, Proc. U.S. Nat. Mus., 98 (3233): 437; Michener, 1951, in 
Muesebeck et al., U.S. Dept. Agric., Monogr. No. 2:1103; Michener, 1954, Bull. Amer. Mus. 
Nat. Hist., 104(1):36-38. 

Calliopsis andreniformis; Cockerell, 1932, Bull. Brooklyn Ent. Soc., 27:11 (misidentification). 
The closest relative of this species is sonora from which it is differentiated 

in the male by the small tan scutellar and metanotal hair pads, and in the 
female by the key characters of couplet 14. In Louisiana and Texas it may be 
mistaken for andreniformis, from which it is readily distinguished in the 
male by the much longer flagellum, and in the female by the distinctive fine 
puncturing on a finely roughened ground. 


Femate. Length, 7mm; forewing length, 4.9 mm; hindwing length 3.45 
mm; clypeal length, 0.49 mm; scutal length, 1.29 mm. 
Head. Yellow areas: (1) paraocular area as in rhodophila but yellow not 


A Revision oF THE BEE GENUS 801 


J 


sonora. 


Map 2. Map showing the known distributions of Calliopsis (Calliopsis) empelia Shinn, C. (C.) 
sonora Shinn, C. (C.) zora Shinn, and C. (C.) hondurasica Cockerell. 


touching lower border of facial fovea; (2) as in rhodophila; (4) supraclypeal 
area, broadly semilunar, extending to below midlevel of antennal socket; (5) 
absent on subantennal plate (to all yellow). (8) Hair like andreniformis, ex- 
cept clypeal hairs fulvous (to brown). (10) Punctures along ocellocular line 
3 pwa, very fine, with dull roughened interspaces; impunctate area laterally 
adjacent to posterior ocellus dull (30>); frontal punctures adjacent to upper 
portion of frontal line finer than in rhodophila, distinct, somewhat shallower, 
1-2 pwa, interspaces dull, roughened (30%). (12) Clypeus with punctures of 
disc much larger than upper frontal ones, 2-3 pwa, interspaces roughened ex- 
cept shiny in median apical portion. (13) Orbital convergence ratio as 1.38: 
1.31,1.05. Facial fovea shallow, broadened medially (to almost borderless in 
some Panamanian specimens). (14) Galea smooth, shiny; length slightly less 
than, to subequal to, antennocellar; galeal gap more than twice length of galea 
exposed beyond closed mandibles, 0.61 :0.26. Some hairs of stipes bent abruptly 
at tip and often curled in a plane perpendicular to the shaft. (15) Head 
width to head length as 2.30:1.72,1.34. (17) Eye length, mio, and flagellar 
length as 1.24:1.31:1.41. (18) Interocellar, antennocular, and interantennal 
as ().37:0.41:0.36:0.34. (19) Ocellolabral greater than clypeal width, 1.45:1.38, 
1.05. (20) Clypeocellar to outer subantennal sutural as 0.94:0.75,1.25. (21) 


802 Tue UNIvERSITY SCIENCE BULLETIN 


Basal labial palpomere 1.5 times length of others combined; many ventral 
hairs, thickened to stout setae. (22) Flagellar length about 2.3 times length 
of scape, 1.41:0.61. 

Mesosoma. (23) Light area yellow; medial interruption of pronotal stripe 
about 2.5 mow. (24) Longer scutal and scutellar hairs only slightly darker 
than shorter hairs. Hair of metanotum white. (25) Scutal disc with punc- 
tures very fine, deep, distinct, about 2 pwa, interspaces extremely dulled by 
very fine roughening. (26) Dorsal enclosure of propodeum shiny, with nu- 
merous vermiform ridges, posterior border carinate; median portion about 
same length as lateral portion. (27) Foreleg without yellow spot on femur. 
(28) Middle leg colored like foreleg; spur length about 0.6 times length of 
middle basitarsus. (30) Tegula brown. (32) Marginal cell 6-9 greater than, 
and 3-4 less than 9-wt, 1.11:0.85:1.00. 

Metasoma. (34) Tergal hair bands white. (35) Tergum 1 with punctures 
of median area exceedingly fine, 3 pwa, interspaces dull, finely roughened. 


Mate. Length, 6.0 mm; forewing length, 4.5 mm; hindwing length, 3.40 _ 
mm, clypeal length, 0.44 mm, scutal length, 1.05 mm. 

Head. Yellow areas: (1) paraocular area as described for subgenus, upper 
boundary a relatively straight line; (2) clypeus, dots testaceous; lateral border 
of apex pale; frontoclypeal suture pale; (4) supraclypeal area, truncate dor- 
sally (to somewhat emarginate) at midlevel of antennal socket; (5) subanten- 
nal plate; (7) scape, except a roughly triangular area of brown, base of tri- 
angle along apex of scape anteriorly and posteriorly, apex of triangle near 
middle of scape; pedicel, on lateral half; flagellomeres 1 and 3 partly brown. 
Scape unusually long. (8) Hair on vertex, frons, and clypeus fulvous. (10) 
Punctures of upper frons finest in subgenus, 2 pwa, interspaces smooth. (11) 
Frontal line with lower portion a relatively high (lower than in rhodophila, 
higher than in moure7), sharp carina. (13) Orbital convergence ratio as 1.33: 
1.04,1.28. (14) Galea shiny, smooth at tip, lightly pebbled, dulled basad, slim, 
tip narrowly rounded; galeal gap greater than (to subequal to) length of 
galea exposed beyond closed mandibles, 0.46:0.26. (15) Head width to head 
length as 2.13:1.63,1.30. (17) Eye length, mio, and flagellar length as 1.17: 
1.04:2.16. (18) Interocellar, ocellocular, antennocular, and interantennal as 
0.32:0.41:0.26:0.31. (19) Ocellolabral slightly less than clypeal width, 1.29:1.34, 
0.96. (20) Clypeocellar to outer subantennal sutural as 0.85:0.70,1.22. (21) 
Basal labial palpomere 1.4 times length of others combined. (22) Flagellar 
length about 4.1 times length of scape, 2.16:0.53. 

Mesosoma. (23) Light areas yellow. (24) Scutal and scutellar hair white 
to pale fulvous. Scutellar hair pads tan, small, confined to outer half of lateral 
portion of scutellum; often hidden by base of wing. Metanotal hair pads tan, 
oval, separated by 0.39 mm or about 2 mow. (25) Scutal disc with punctures 
fine, mostly 1.0-1.5 pwa, interspaces finely roughened, somewhat shiny. (26) 


A Revision oF THE BEE GENUS 803 


Dorsal enclosure of propodeum with indistinct reticulated ridges, interspaces 
strongly roughened, somewhat shiny. (27) Legs with light color the same as 
on face. Foreleg yellow except basal half of coxa dark brown, distitarsus tes- 
taceous. (28) Middle leg yellow except dorsal half of coxa dark brown, apico- 
tarsus testaceous. (29) Hind leg yellow except base of coxa, dorsal apex of 
femur, and basal rim (often entire base) of basitibial plate brown, apicotarsus 
testaceous. (32) Marginal cell 6-9 greater than, and 3-4 slightly less than 9-wt, 
1.02:0.75 :0.88. 

Metasoma. (34) Tergal hair bands as in granti. (35) Tergum 1 with 
punctures of median area exceedingly fine, distinct, 2-3 pwa, interspaces finely 
roughened, dull. (37) Sterna brown. 

Type Materia. The female holotype taken at Zamorano, Honduras, in 
December, is in the U.S. National Museum (U.S.N.M. No. 58444). Zamorano 
is S.S.E. of Tegucigalpa, in the mountains, at latitude 14.01° N., longitude 
87.01" W. The above description of the male is principally based on a speci- 
men from Amatitlan, Guatemala, 4000 ft., July 6, 1947 (C. and P. Vaurie). 

Disrripution. This is the most widespread species in the genus. It occurs 
from Central Panama to northwestern Louisiana. It is remarkable that 
despite the huge area covered by this bee, it is instantly recognizable and 
varies relatively little geographically. Specimens from Chiapas are darker 
than others, with the male pronotal lobes almost entirely dark; they also show 
considerable pebbling which dulls the galea, whereas others are smooth and 
very shiny. 

This bee is the only species of Calliopsis which has been collected in every 
month of the year, owing to its occurrence in tropical latitudes. Likely enough 
its activity span in any one area is little more than in its relatives: a few 
months. Collections in Panama tend to confirm this opinion: they date only 
from November to February. Michener (1954) suggests the limitation of 
hondurasica activity by rainfall rather than temperature. C. hondurasica flies 
during the early part of the dry season when moist conditions with clear, 
sunny days bring numerous flowering plants into bloom. 

In addition to the type, specimens have been studied from the following localities: Louisiana: 
Robson, U.S.D.A. Pecan Field Station. Texas: Brownsville; Cameron Co.; Giddings, Lee Co.; 
Hidalgo; Mission (State Park on Rio Grande nearby), Hidalgo Co.; Progreso; Richmond, Fort 
Bend Co.; Riogrande (5 mi. E.); San Manuel (10 mi. S.); Santa Maria; Southmost, Cameron Co.; 
Sweeny. Cutapas: Comitan (15 mi. N.W.); San Cristébal de las Casas (39 mi. E.); Simojovel 
(4 mi. S.). Guerrero: Amula, 6000 ft.; Chilpancingo, 3700 ft.; Iguala (13 mi. N.), 3900 ft.; 
Taxco (19.5 mi. N.E.), 4800 ft.; Tepetlapa, 3000 ft. MicHoacan: Apdatzingan (4, 11 mi. E.). 
MoreEtos: Yautepec (4 mi. §.W.), 3800 ft. Nayarir: San Blas (5 mi. E.), 25 ft. Oaxaca: Oaxaca. 
San Luts Potrost: Ciudad del Maiz (5 mi. E.), 4700 ft.; El Naranjo (3.4 mi. N.E.), 800 ft.; El 
Salto, 1800 ft. Tamautipas: El Limén; Jiménez (22 mi. S.); Padilla; Tampico. Veracruz: Cér- 
doba; Gutiérrez Zamora (4 mi. E.), 100 ft.; Nautla; Puente Nacional (4 mi. W.), 900 ft. 
Yucatan: Temax. GuaTeMara: Amatitlin, 4000 ft.; Guatemala City, 5000 ft. Honpuras: Agua 
Azul; Zamorano. Costa Rica: Playa del Coco. Panama: Chillibre, Panama Province; David; Old 


Panama; Panama City; Pueblo Nuevo, Panama Province; Salanas. Cana Zone: Ancon Hill; 
Chiva Chiva; Corozal; Fort Clayton; Summit. 


804 Tue UNiversiry ScIENCE BULLETIN 


Bronomics. Professor Alvaro Wille of the University of Costa Rica at San 
Jose has kindly furnished, through Professor Charles D. Michener, fragmen- 
tary notes on nesting sites, nest density, and the form of the burrow. These 
are discussed in comparison with andreniformis in connection with biology 
and ecology of the latter. 

Frower Recorps. Cassia, Kallostroemia hirsutissima, Lippia, Nama 
undulatum, Phyla strigosa, Teucrium, Trifolium repens, Verbena. 


CALLIOPSIS (CALLIOPSIS) SQUAMIFERA Timberlake 
(Figs. 42-45; Map 3) 


Calliopsis squamifera Timberlake, 1947, Pan-Pacific Ent., 23:28, male; Michener, 1951, in 

Muesebeck e¢ a/., U.S. Dept. Agric., Monogr. No. 2: 1103. 

Timberlake considered this one of the species of Calliopsis s.s. and I con- 
cur, but it has characters which link it as an intermediate between Calliopsis 
s.s. and Perissander. 

It is closest to C. peninsularis, which will doubtless be proven a Calliopsis 
s.s. when the male becomes known, but is readily distinguished by the meso- 
tibial spur being bare and 0.75 times or more the length of the middle basitar- 
sus (less than 0.60 in peninsularts). 


Femate. Length, 5.0 mm; forewing length, 3.3 mm; hindwing length, 
2.40 mm; clypeal length, 0.34 mm; scutal length, 0.85 mm. 

Head. White areas: (1) paraocular area, except for lowermost corner, 
below a sinuous line originating at about middle of outer subantennal suture 
and extending to a point on orbit about 0.4 eye length below summit of eye; 
(2) clypeus with a broad, inverted T, the base bordering the supraclypeal area, 
the crossbar bordering the narrow, testaceous apical margin to the extent of 
the total width of clypeal emargination (reduced in some specimens to a nar- 
rowed, vertical part of the T); (4) supraclypeal area to about level of middle 
of antennal socket, dorsal margin evenly convex; (5) subantennal plate (to all 
black). (8) Hair on vertex fulvous, longer than long scutal hairs, on frons 
and clypeus fulvous. (10) Punctures of vertex and along ocellocular line with 
interspaces roughened (30). Median punctures of frons deep, subconfluent, 
interspaces shiny (30>). (11) Frontal line with lower portion narrowly sul- 
cate, becoming obsolete, ending in a slightly raised prominence at level of 
upper rim of antennal socket. (13) Orbital convergence ratio as 1.05 :0.94,1.13. 
(14) Galea finely pebbled, hidden with mandibles closed. (15) Head width 
to head length as 1.73:1.22,1.42. (17) Eye length, mio, and flagellar length as 
1.00:0.94:1.02. (18) Interocellar, ocellocular, antennocular, and interantennal 
as 0.31:0.29:0.27:0.22. (19) Ocellolabral greater than clypeal width, 1.05:0.94, 
1.13. (20) Clypeocellar to outer subantennal sutural as 0.71:0.51,1.40. (21) 


A Revision oF THE BEE GENUS 805 


Basal labial palpomere 0.8 times length of others combined. (22) Flagellar 
length about 2.2 times length of scape, 1.02:0.46. 

Mesosoma. (23) Light areas yellowish; medial interruption of pronotal 
stripe about 3 mow. (24) Scutal and scutellar hair fulvous apicad, light brown 
basad, darker than hairs of vertex. Hair of metanotum fulvous. (25) Scutal 
disc with punctures finer and deeper than on vertex, 1-2 pwa, interspaces shiny 
(60). (26) Dorsal enclosure of propodeum generally dull, slightly decli- 
vous, median portion with fine, close, vermiform, interrupted ridges, inter- 
spaces roughened, ridges becoming longitudinal and several ridge widths 
apart laterad, interspaces roughened. Proposed triangle adjacent to enclosure 
dull, roughened. (27) Legs with light color more yellowish than on face. 
Foreleg with cream (to yellow) on dorsal apex of femur and knee of tibia. 
(28) Middle leg colored like foreleg; spur entirely smooth, dark brown, length 
very long, about 0.8 length middle basitarsus, 0.53:0.60. (30) Tegula trans- 
parent, testaceous, with small anterior patch of cream (to yellow) color. Hu- 
meral plate testaceous. (32) Marginal cell (6-9) subequal to, and 3-4 shorter 
than 9-wt, 0.70:0.61:0.71. 

Metasoma. (34) Tergal hair bands white. Band of tergum 1 narrowly (to 
broadly) interrupted medially, others entire. Suberect hair of disc of tergum 
4 fulvous (to brownish), of disc of tergum 5 white (to fulvous). (35) Ter- 
gum 1 with punctures of median area exceedingly fine, dense, regularly 
spaced, 1-1.5 pwa, interspaces shiny though minutely roughened (60), the 
tergum with a silky sheen (15). Declivity of tergum 1 moderately shiny (to 
very shiny) though finely roughened. 


Mate. Length, 4.0 mm (type 4 mm); forewing length, 2.9 mm (type 2.8 
mm); hindwing length, 2.2 mm; clypeal length, 0.32 mm; scutal length, 1.33 
mm. 

Head. White areas: (1) paraocular area below a line originating at dor- 
solateral rim of antennal socket and extending diagonally upward, passing 
below the facial fovea, to a point on orbit about 0.4 eye length below summit 
of eye; (2) clypeus, except for apical pale testaceous margin; (4) supraclypeal 
area, pentagonal shape, to level of upper rim of antennal socket; (5) suban- 
tennal plate. (7) Scape broadly yellow (to yellowish white) in front, the 
mesoapical portion brown; pedicel brown with a lateral yellow spot; dorsal 
surface of flagellum dusky, darker basad, with brown extending part way 
round flagellomeres 1, 2, and 3; ventral surface of flagellum tan, flagellomere 
1 with a lateral yellow spot. (8) As in female. (10) Punctures of vertex and 
along ocellocular line with interspaces shiny but faintly roughened (60%). 
Median punctures of frons shallow, contiguous, dull. (11) Frontal line with 
lower portion carinate. (13) Orbital convergence ratio as 0.92:0.68,1.36. (14) 
Galea as in female. (15) Head width to head length as 1.39:1.12,1.24. (16) 
As in female. (17) Eye length, mio, and flagellar length as 0.83:0.68:1.41, 


806 Tue UNiversiry SCIENCE BULLETIN 


(18) Interocellar, ocellocular, antennocular, and interantennal as 0.26:0:26: 
0.15:0.19. (19) Ocellolabral greater than clypeal width, 0:92:0.82,1.13. (20) 
Clypeocellar to outer subantennal sutural as 0.60 :0.43,1.42. (21) Basal labial 
palpomere 0.7 times length of others combined. (22) Flagellar length about 
4.0 times length of scape, 1.41:0.36. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal 
stripe about 2 mow. (24) Scutal and scutellar hairs whitish. Scutellar hair 
pads pale grayish-brown (to tan), their combined width slightly less than 
width of median portion of scutellum. Metanotal hair pads same color, 
broadly oval, failing to meet medially by less than mow, covering metanotum 
except for a median triangular area with base at anterior border of meta- 
notum. Hair of metanotum white. (25) Scutal disc with punctures finer, 
deeper than on vertex, 2 or more pwa, interspaces shiny (60%). (26) Dorsal 
enclosure of propodeum with fine ridges originating anteriorly at the median 
line, each curving laterally along the width of the enclosure, giving the im- 
pression of tightly stretched, wrinkled skin. Remainder of propodeal triangle 
shiny. (27) Legs with light color bright yellow. Foreleg entirely yellow. (28) 
Middle leg with basal half of coxa dark, remainder of leg yellow with disti- 
tarsus pale testaceous. Lengths of tibia, basitarsus, and apicotarsus as 0.80: 
0.85:1.00. (29) Hind leg colored like middle leg except distitarsus brown. 
Mediotarsomeres triangular in shape. (30) As in female. (32) Marginal cell 
lengths similar to female, 0.66 :0.58 :0.65. 

Metasoma. (33, 34) As in female. (35) As in female except punctures 2 
pwa. 

Type Materia. Holotype male from Picacho Pass (5 mi. S.E. of Picacho) 
[summit is at 1800 ft. altitude], Pinal Co., Arizona, Aug. 7, 1940 (P. H. Tim- 
berlake), on Euphorbia, is in the Timberlake collection at the University of 
California, Riverside, California. The above description of the male is based 
primarily on a topotypical specimen collected by C. D. Michener in company 
with Prof. Timberlake at the time of collection of the holotype. 

Disrrisution. Known only from southern Arizona but to be expected 
from southwestern New Mexico and northern Sonora, southeastern Cali- 
fornia, and Baja California. 

Specimens have been studied from the following localities: Arizona: Apache (5 mi. SES 4 
mi. S.W.), Cochise Co., Aug. 7-11, on Euphorbia, (13 mi. S.W.), Aug. 19-26, on Euphorbia, 
(5 mi. S.W.), Aug. 17, on Baccharis glutinosa; Douglas, Aug. 26 (3 mi. N.; 17 miles E.), Cochise 
Co., Aug. 8, on Euphorbia; Picacho Pass (5 mi. S.E. of Picacho), Pinal Co., Aug. 7, on Euphor- 
bia: Portal (2 mi. N.E.), Cochise Co., Sept. 14; Roll, Yuma Co., July 11-Aug. 11, Oct. 20, on 
Euphorbia and Alfalfa; Tempe, Aug. lL. 

Grocrapnic VariaTIoN. Specimens from Roll in southwestern Arizona 
differ from those in middle southern and southeastern Arizona by having 
denser, shorter, more plumose hair on the face, thorax, and abdominal hair 


A REVISION OF THE BEE GENUS 807 


bands. Females from Apache and Douglas have greatly reduced areas of 
cream color on the head. Other minor differences in certain head and tho- 
racic dimensions occur, but without males I must consider these within the 
range of variation of the species. 

Discussion. Most of the flower records are from Euphorbia, a biological 
datum bolstering the morphological data which indicate a relationship with 
Perissander. 

Bronomics and Frower Recorps. Nothing is known of the bionomics of 
squamifera, but it has been taken on each one of the favorite plant families 
for the genus: Compositae, Baccharis glutinosa; Euphorbiaceae, Euphorbia; 
and Leguminosae, Alfalfa, Medicago sativa. 


@® squamifera 
@-0-f: 1 


* peninsularis 


O fulgida 


Map 3. Map showing the known distribution of Calliopsis (Calliopsis) peninsularis Shinn, C., 
(C.) squamifera Timberlake, C. (Perissander) syphar Shinn, C. (P.) fulgida Shinn, and C. 
(P.) yalea Shinn. 


808 Tue University SCIENCE BULLETIN 


CALLIOPSIS (CALLIOPSIS) PENINSULARIS, new species 
(Map 3) 


The specific name is from the Latin, peninsula, because of occurrence on 
the peninsula of Baja California. 

The species resembles C. squamifera and C. syphar. It is distinguished 
from the former by the much shorter, pectinate mesotibial spur, and from 
the latter by the distinctly reticulated ridges on the dorsal enclosure of the 
propodeum. 


Femare. Length, 5.5 mm; forewing length, 3.6 mm; hindwing length, 
2.50 mm; clypeal length, 0.37 mm; scutal length, 0.90 mm. 

Head. Cream colored areas: (1) paraocular area below a line originating 
at about middle of outer subantennal suture and extending dorsally, inclining 
slightly laterad, to about 1 mow above antennal socket then convexly curved 
upward passing just below facial fovea and ending on orbit about 0.4 times 
eye length below summit of eye; (2) clypeus with a broad, inverted T, the 
base bordering the supraclypeal area, the crossbar bordering the narrow, tes- 
taceous apical margin for almost full width of clypeus, thus isolating two sub- 
triangular brown areas with bases along subantennal plates, and two tiny 
patches of brown in the extreme lower corners of clypeus; (4) as in squamu- 
fera; (5) subantennal plate. (8) As in squamifera. (10) As in squamifera 
except roughening confined to vertex and virtually undetectable (30). (11) 
As in squamifera. (13) Orbital convergence ratio as 1.05:1.00,1.05. Facial 
fovea ellipsoidal, distinct, width a fourth to third of length. (14) As in 
squamifera except glossa less broad medially. (15) Head width to head 
length as 1.80:1.28,1.41. (17) Eye length, mio, and flagellar length as 1.04: 
1.00:1.05. (18) Interocellar, ocellocular, antennocular and interantennal as 
0.31 :0.31:0.26:0.24. (19) Ocellolabral greater than clypeal width, 1.09:1.05,1.02. 
(20) Clypeocellar to outer subantennal sutural as 0.71:0.54,1.31. (21) Basal 
labial palpomere 0.8 times length of others combined. (22) Flagellar length 
about 2.3 times length of scape, 1.05:0.46. 

Mesosoma. (23) Light areas cream colored; medial interruption of pro- 
notal stripe about 1.5 mow. (24) Scutal, scutellar, and metanotal hair fulvous 
(may be faded!). (25) As in squamifera except punctures apicad from an- 
terior end of parapsidal line becoming crowded to subconfluent with inter- 
spaces roughened. (26) Dorsal enclosure of propodeum with medial portion 
dull, bearing a distinct reticulum of ridges, lateral portion having about 7 
longitudinal ridges with interspaces shiny (30%). Area of propodeal triangle 
adjacent to lateral areas of enclosure roughened but relatively shiny. (27) 
Legs with light color the same as on face. Foreleg with cream color on dorsal 
apex of femur and a slanting patch to slightly less than midpoint of tibia. 


A ReEvIsION OF THE BEE GENUS 809 


(28) Middle leg colored like foreleg except tibial patch extends only slightly 
beyond knee; spur finely, evenly pectinate, testaceous, its length to length of 
basitarsus as 0.34:0.61,0.55. (30) Tegula transparent, almost colorless, with 
small anterior patch of cream color. (32) Marginal cell 6-9 and 3-4 greater 
than 9-wt, 0.77 :0.71:0.70. 

Metasoma. (34) Tergal hair bands white, hairs longer, less dense, and less 
plumose than in squamifera or in syphar. Prepygidial and pygidial fimbriae 
smoky, denser and longer than squamifera. (35) Tergum 1 with punctures of 
median area fine, larger than squamifera, dense, regularly spaced, 1.0-1.5 pwa, 
interspaces shiny (30>), the tergum without a silken sheen (15). Declivity 
of tergum | shiny, finely lineolate. 

Tyre Marertat. Holotype female from San Ignacio (15 mi. N.), Baja 

California Sur, Sept. 29, 1941 (E. S. Ross and R. M. Bohart), is in the 
California Academy of Sciences, San Beanie. 


CALLIOPSIS (CALLIOPSIS) SONORA, new species 
(Figs. 34-37; Map 2) 


The specific name is from the state of Mexico in which it occurs. It is 
closest to hondurasica. The female is distinguished from hondurasica by the 
smaller size and proportionately shorter flagellum relative to the minimum 
interocular distance. The male is distinguished by the smaller size and by the 
brown or gray metanotal hair pads a 1 are separated by less than 1 mow. 


Femare. Length, 6.0 mm; forewing length, 4.2 mm; hindwing length, 2.90 
mm; clypeal length, 0.46 mm, scutal length, 1.16 mm. 

Head. Yellow areas: (1) paraocular area below a sinuous line originating 
at about middle of outer subantennal suture and passing below facial fovea 
ending on orbit slightly below facial fovea, well above antennal socket; (2) as 
in hondurasica except median stripe only 0.25 mow from clypeal apex; (4) as 
in hondurasica; (5) subantennal plate except uppermost portion. (8) Hair as 
in hondurasica. (10) Punctures along ocellocular line 1-2 pwa, fine but larger 
than hondurasica, interspaces dull but less so than in hondurasica; impunctate 
area laterally adjacent to posterior ocellus dull, roughened, but less so than in 
hondurasica; frontal punctures adjacent to upper frontal line as in honduras- 
ica except interspaces very faintly metallic (30). (12) Clypeus with punc- 
tures of disc larger than upper frontal ones, 2-3 pwa, interspaces roughened, 
except shiny in median apical portion. (13) Orbital convergence ratio, as 
1.22:1.19,1.03. Facial fovea distinct, shorter than hondurasica, linear. (14) 
Galea smooth, shiny except faintly roughened basad; length shorter than 
antennocellar; galeal gap greater than length of galea exposed beyond closed 
mandibles. Hair of stipes as in hondurasica. (15) Head width to head length 
as 2.04:1.50,1.36. (17) Eye length, mio, and flagellar length as 1.12:1.19:1.21. 


810 THe UNIVERSITY SCIENCE BULLETIN 


(18) Interocellar, ocellocular, antennocular, and interantennal as 0.37:0.37: 
0).29:0.32. (19) Ocellolabral greater than clypeal width, 1.29:1.24,1.04. (20) 
Clypeocellar to outer subantennal sutural as 0.83:0.65,1.29. (21) Basal labial 
palpomere about 1.4 times length of others combined; ventral hairs as in 
hondurasica except smaller in diameter and fewer present. (22) Flagellar 
length about 2.2 times length of scape, 1.21 :0.54. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal 
stripe about 2 mow. (24) As in hondurasica. (25) Scutal disc with punc- 
tures finer, shallower than in hondurasica, about 2 pwa, interspaces as in 
hondurasica except finer roughening. (26) Dorsal enclosure of propodeum 
as in hondurasica. (27) Foreleg as in hondurasica. (28) Middle leg colored 
like foreleg; spur length about 0.6 times length middle basitarsus, 0.44 :0.73. 
(30) Tegula brown. (32) Marginal cell 6-9 greater than, and 3-4 less than 
9-wt, 0.90:0.70:0.85. 

Metasoma. (34) Tergal hair bands white. (35) Tergum 1 with punctures 
of median area exceedingly fine, 2-3 pwa, interspaces dull, finely roughened. 


Mate. Length, 5.9 mm; forewing length, 4.0 mm; hindwing length, 2.81 
mm; clypeal length, 0.43 mm; scutal length, 0.85 mm. 

Head. Yellow areas (dull, not shiny like hondurasica): (1) paraocular 
area as described for subgenus, upper boundary relatively straight; (2) 
clypeus, dots colorless; frontoclypeal suture pale; (4) supraclypeal area, ex- 
tending higher than in hondurasica, onto lower edge of frontal carina, summit 
slightly below level of upper border of antennal socket; (5) subantennal plate. 
(7) Scape, as in hondurasica except apex of triangular brown area only about 
0.4 distance towards base of scape. Scape with length proportionately shorter 
than in hondurasica. (8) Hair as in hondurasica. (10) Punctures of upper 
frons fine, less so than Aondurasica, less distinct than in hondurasica, 1 pwa, 
interspaces smooth, rather dull. (11) Frontal line with lower portion as in 
hondurasica. (13) Orbital convergence ratio as 1.17:0.99,1.19. (14) As in 
hondurasica. (15) Head width to head length as 1.87:1.53,1.22. (17) Eye 
length, mio, and flagellar length as 1.05:0.99:1.56. (18) Interocellar, ocellocu- 
lar, antennocular, and interantennal as 0.31:0.36:0.22:0.26. (19) Ocellolabral 
less than clypeal width, 1.22:1.26,0.97. (20) Clypeocellar to outer subantennal 
sutural as 0.80:0.61,1.30. (21) Basal labial palpomere 1.3 times length of others 
combined. (22) Flagellar length about 3.5 times length of scape, 1.56:0.44. 

Mesosoma. (23) Light areas yellow. (24) As in Aondurasica. Scutellar 
hair pads dark grayish brown, confined to entire lateral portion of scutellum, 
not completely hidden by base of wing. Metanotal hair pads same color, large, 
oval, separated by less than 1 mow. (25) Scutal disc punctures larger than in 
hondurasica, shallower, 1 pwa or less, interspaces as in hondurasica. (26) 
Dorsal enclosure of propodeum covered by metanotal hair pad except for 
medial portion which is gradually elevated; a single, prominent, mesally 


A REVISION OF THE BEE GENUS 811 


oblique ridge on each side of the center line, enclosing area with several 
small, indistinct ridges. (27) Legs with light color the same as on face. Fore- 
leg yellow except base of coxa brown. (28) Middle leg yellow except base of 
coxa brown, apicotarsus testaceous. (29) Hind leg yellow except base of coxa 
brown, dorsal apex of femur, and basal rim of basitibial plate testaceous, 
apicotarsus testaceous. (32) Marginal cell 6-9 subequal to, and 3-4 less than 
9-wt, 0.77 :0.61 :0.78. 

Metasoma. (34) As in granti. (35) Tergum 1 with punctures of median 
area very fine, indistinct, interspaces roughened, dull. (37) Sterna brown. 

Type Materia. Holotype male, from Rio Mayo, Sonora, Aug. 25, 1935 
(collector unknown), is in Dr. G. E. Bohart’s collection, Logan, Utah. Allo- 
type female and one female paratype, from Rio Mayo, San Bernardino, 
Sonora, same date (J. J. du Bois), are at the University of California, River- 
side, California. 

Discussion. Apparently this species is an offshoot of hondurasica adapted 
to the arid Sonoran area where hondurasica does not occur. 

One female specimen, from Canipole (about 100 mi. S.E. of San Ignacio), 
Baja California Sur, Oct. 2, 1941 (E. S. Ross and R. M. Bohart), may be this 
species. Its salient differences from the type material of sonora are the pres- 
ence of a yellow spot on the tegula, basal labial palpomere 2.2 times length of 
others combined and 0.66 times length of scape, galeal length equal to scape 
length (much greater in sonora), interantennal less than antennocular, anten- 
nocular greater than inner subantennal sutural, scutal punctures shallower, 
less distinct, interspaces smoother and duller. Other characteristics are vir- 
tually identical, and this is so for numerous body measurements except the 


highly variable head width. 


CALLIOPSIS (CALLIOPSIS) EMPELIA, new species 
(Figs. 38-41; Map 2) 


The specific name is from the Greek, empelios, meaning gray, which is 
applied with reference to the large gray metanotal hair pads. 

It is closest to sonora and the female is distinguished from it by the yellow 
dot on the brown tegula, by the extremely dull interspaces between scutal 
punctures, by the non-metallic frons, and by the flagellar length being mark- 
edly shorter than the minimum interocular distance. 


Femare. Length, 7.2 mm; forewing length, 4.4 mm; hindwing length, 
2.40 mm; clypeal length, 0.53 mm; scutal length, 1.27 mm. 

Head. Yellow areas: (1) paraocular area as in sonora; (2) clypeus as in 
hondurasica; (4) supraclypeal area as in sonora; (5) subantennal plate en- 
tirely. (8) Hair faded, but as in sonora. (10) Punctures along ocellocular line 
coarser, shallower than others in hondurasica group, less than 1 pwa, inter- 


812 Tue UNiversiry SCIENCE BULLETIN 


spaces very dull, most coarsely roughened of hondurasica group; impunctate 
area dull, roughened; upper frontal punctures less than 0.5 pwa, interspaces 
heavily roughened. (12) Clypeus with punctures of disc about size of upper 
frontal punctures, 2-3 pwa, interspaces roughened, except shiny in median 
apical portion. (13) Orbital convergence ratio as 1.31:1.26,1.04. Facial fovea 
distinct, broader medially, tapering strongly above. (14) Galea faintly pebbled 
(30>); length less than antennocular; galeal gap subequal to length of galea 
exposed beyond closed mandibles. (15) Head width to head length as 2.30: 
1.72,1.34. (17) Eye length, mio, and flagellar length as 1.26:1.26:1.02. (18) 
Interocellar, ocellocular, antennocular, and interantennal as 0.36 :0.39 :0.32:0.31. 
(19) Ocellolabral greater than clypeal width, 1.45:1.39,1.03. (20) Clypeocellar 
to outer subantennal sutural as 0.92:0.70,1.31. (21) Basal labial palpomere 1.8 
times length of others combined; ventral hairs somewhat thickened but much 
less than in hondurasica. (22) Flagellar length about 1.8 times length of 
scape, 1.02:0.56. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal stripe 
about 1.5 mow. (24) As in andreniformis. (25) Scutal disc with punctures of 
moderate size, much larger and closer than preceding species, 1 pwa or less, 
interspaces heavily roughened. (26) Dorsal enclosure of propodeum as in 
hondurasica except dull and more sharply carinate posteriorly. (27) Foreleg 
with yellow on extreme apex of femur. (28) Middle leg without yellow on 
extreme apex of femur; spur length about 0.64 times length of middle basi- 
tarsus. (30) Tegula brown with anterior yellow dot. (32) Marginal cell 6-9 
greater than, and 3-4 less than 9-wt, 1.11 :0.85:1.00. 

Metasoma. (34) Tergal hair bands white. (35) Tergum 1 with punctures 
of median area fine, about 1 pwa, interspaces heavily roughened, dullest of 
subgenus. 


Mate. Length, 5.3 mm; forewing length, 4.0 mm; hindwing length, 2.89 
mm; clypeal length, 0.41 mm; scutal length, 0.85 mm. 

Head. Yellow areas: (1) paraocular area as described for subgenus, upper 
boundary relatively straight; (2) clypeus, dots black; frontoclypeal suture 
pale; (4) supraclypeal area, summit at level of upper border of antennal 
socket; (5) subantennal plate. (7) As in hondurasica, except brown area of 
scape extending to base of scape. (8) Hair as in hondurasica. (10) Punce- 
tures of upper frons almost as fine as in hondurasica, less distinct than in 
hondurastca, 1-2 pwa, interspaces smooth, dull. (11) Frontal line with lower 
portion as in sonora except lower. (13) Orbital convergence ratio as 1.11:0.90, 
1.22. (14) Galea shiny, faintly pebbled, broader than hondurasica and sonora, 
tip less narrowly rounded; galeal gap greater than length of galea exposed 
beyond closed mandibles. (15) Head width to head length as 2.13:1.63,1.30. 
(17) Eye length, mio, and flagellar length as 0.99 :0.90:1.79. (18) Interocellar, 
ocellocular, antennocular, and interantennal as 0.29:0.34:0.20:0.24. (19) Ocel- 


A ReEvIsIon oF THE BEE GENUS 813 


lolabral slightly less than clypeal width, 1.16:1.19,0.97. (20) Clypeocellar to 
outer subantennal sutural as 0.75 :0.56,1.33. (21) Basal labial palpomere about 
1.2 times length of others combined. (22) Flagellar length about 4.0 times 
length of scape, 1.79:0.44. 

Mesosoma. (23) Light areas yellow. (24) As in hondurasica. Scutellar 
hair pad as in sonora except dark gray. Metanotal hair pad as in sonora ex- 
cept dark gray, contiguous medially on posterior border, separated anteriorly 
leaving a triangular area with broad apex from which arises a tuft of long 
white hairs. (25) Scutal disc with punctures as in sonora except interspaces 
shinier. (26) Dorsal enclosure of propodeum as in sonora. (27) Legs with 
light color the same as on face. Foreleg yellow except extreme base of coxa 
brown. (28) Middle leg colored like foreleg except apicotarsus testaceous. 
(29) Hind leg as in hondurasica. (32) Marginal cell 6-9 greater than, and 3-4 
less than 9-wt, 0.82:0.61 :0.77. 

Metasoma. (34) Tergal hair bands sparse but lateral portions somewhat 
more plumose and denser than other species of hondurasica group. (35) 
Tergum 1 with punctures of median area very fine, indistinct, interspaces 
roughened, somewhat shiny. (37) Sterna brown. 

Tyre Mareriar. Holotype male, from Mt. Lemmon Road, Mt. Lemmon, 
Arizona, 3500 ft., Aug. 15, 1954 (R. M. Bohart), and allotype female, from 
Douglas, Arizona, Aug. 2, 1940 (W. W. Jones), are in the Timberlake collec- 
tion at the University of California, Riverside, California. 

Discusston. The sexes were associated on the basis of punctation, sculp- 
turing, similarity of mouthparts, and size difference as compared with that 
for sonora and hondurasica. The species probably occurs in that rich area for 
the genus—the Chiricahua Mountain region in the vicinity of Portal, Arizona. 


CALLIOPSIS (CALLIOPSIS) ZORA, new species 
(Map 2) 


The specific name is from the Greek zoros, meaning strong, and is applied 
because this species is the largest and most robust bee in its subgenus. 

Apparently closest to empelia from which it is distinguished by the com- 
pletely brown tegula, by the moderately shiny interspaces between the scutal 
punctures, and by the faint greenish metallic color on the frons. 


Femate. Length, 7.5 mm; forewing length, 5.0 mm; hindwing length, 3.62 
mm; clypeal length, 0.58 mm; scutal length, 1.38 mm. 

Head. Yellow areas: (1) paraocular area as in sonora but ending on orbit 
at about level of upper rim of antennal socket; (2) clypeus with median 
stripe as in hondurasica except reaching to preapical groove; (3) labrum with 
median apical spot on labral plate; middle portion of labral plate strongly 
depressed, sunken well below usual level in Calliopsis s.s.; (4) supraclypeal 


814 Tue University SCIENCE BULLETIN 


area, apex reaching to slightly above midlevel of antennal socket; (5) suban- 
tennal plate as in empelia. (8) Hair as in hondurasica. (10) Punctures along 
ocellocular line fine, 2 pwa, interspaces faint greenish metallic, moderately 
shiny; impunctate area laterally adjacent to posterior ocellus dull, finely 
roughened; upper frontal punctures extremely fine, indistinct, shallow, less 
than 1 pwa, interspaces faint greenish metallic, interspaces smooth, moder- 
ately shiny. (12) Clypeus with punctures of disc larger than upper frontal 
punctures, 2-3 pwa, interspaces roughened beside subantennal plate, shiny 
elsewhere (30%). (13) Orbital convergence ratio as 1.45:1.33,1.09. Facial 
fovea with indistinct mesal border, broader medially narrowed above and 
below. (14) Galea faintly pebbled basad; length subequal to antennocellar; 
galeal gap and length of galea exposed beyond closed mandibles N. A. (15) 
Head width to head length as 2.36:1.87,1.36. (17) Eye length, mio, and 
flagellar length as 1.33:1.33:1.33. (18) Interocellar, ocellocular, antennocular, 
and interantennal as 0.34:0.44 :0.32:0.36. (19) Ocellolabral greater than clypeal — 
width, 1.55:1.43,1.08. (20) Clypeocellar to outer subantennal sutural as 0.97: 
0.78,1.24. (21) Basal labial palpomere about 1.5 times length of others com- 
bined; ventral hairs thickened but less than in hondurasica, only 4-5 present. 
(22) Flagellar length about 2.2 times length of scape, 1.33:0.61. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal stripe 
about 4 mow. (24) Longer scutal and scutellar hairs dark brown, longer and 
more prominently branched than in preceding species. Hairs of metanotum 
fulvous, others as in hondurasica. (25) Scutal disc with punctures moderately 
large, about as in empelia, less than 0.5 pwa, interspaces shiny although finely 
roughened, punctures becoming crowded anteriorly, to contiguous. (26) 
Dorsal enclosure of propodeum shiny, with numerous fine, reticulated ridges; 
posterior border carinate, a small median portion of border raised distinctly 
upward; median portion about as long as lateral portion. (27) Foreleg with- 
out yellow on femur. (28) Middle leg colored like foreleg; spur length about 
0.65 times length middle basitarsus. (30) Tegula blackish brown. (32) Mar- 
ginal cell 6-9 less than, and 3-4 less than 9-wt, 1.04:0.88:1.09. 

Metasoma. (34) Tergal hair bands white. (35) Tergum 1 with punctures 
of median area fine, about 1 pwa, interspaces moderately shiny, finely 
roughened. 

Typr Martertat. Holotype female, El Salto (6 mi. N.E.), Durango, 
México, $500 ft., Aug. 10, 1947 (Gertsch, David Rockefeller Expedition), 1s 
in the American Museum of Natural History. 


CALLIOPSIS (CALLIOPSIS) HELENAE, new species 
(Map 1) 


This species is named for my wife, Helen, who has helped in many ways 
to make the completion of this revision possible. The species has no apparent 


A REVISION OF THE BEE GENUS 815 


close relative and is quite distinctive by virtue of the yellow markings on the 
lateral portions of the clypeus and on the tegula. 


Femae. Length, 6.5 mm; forewing length, 4.1 mm; hindwing length, 
2.98 mm; clypeal length, 0.48 mm; scutal length, 1.16 mm. 

Head. Cream colored areas: (1) paraocular area, below a sinuous line 
originating in upper half of outer subantennal suture and extending diagon- 
ally upward to a point on orbit about 0.4 of eye length below summit of eye; 
(2) clypeus with a broad, vertical median area from base bordering supracly- 
peal and subantennal areas to preapical groove and with irregular splotches 
contiguous to median area; (3) labrum with spot on apex of labral plate; (4) 
supraclypeal area to level of middle of antennal socket; (5) subantennal plate; 
(6) mandible possibly with a faint basal spot. (8) Hair on vertex, frons, and 
clypeus fulvous. (10) Punctures of vertex and along ocellocular line with 
interspaces shiny (30%). Median punctures of frons coarse, deep, less than 1 
pwa, interspaces shiny (30%). (11) Frontal line with lower portion sulcate 
ending at level of middle of antennal socket. (13) Orbital convergence ratio 
as 1.22:1.19,1.03. Facial fovea elongate, length about 1.5 mow. (14) Galea 
enturely pebbled, galeal gap less than length of galea exposed beyond closed 
mandibles, 0.31:0.58. Glossa, length 1.62 (paratype). (15) Head width to 
head length as 2.06:1.65,1.25. (17) Eye length, mio, and flagellar length as 
1.22:1.19:1.21. (18) Interocellar, ocellocular, antennocular, and interantennal 
as 0.34 :0.34:0.31:0.30. (19) Ocellolabral greater than clypeal width, 1.34:1.29, 
1.32. (21) Basal labial palpomere about 2.0 times length of others combined. 
(22) Flagellar length about 2.3 times length of scape, 1.22:0.54. 

Mesosoma. (23) Light areas cream colored; medial interruption of pro- 
notal stripe about 2 mow. (24) Scutum and scutellum with short and long 
fulvous hairs (long hairs may be faded brown hairs!). Hair of metanotum 
fulvous. (25) Scutal disc with punctures fine, 2-3 pwa, interspaces shiny, 
smooth (30), but roughened on anterior slope of scutum. (26) Dorsal en- 
closure of propodeum with median portion a dull reticulum of close trans- 
verse ridges, the lateral portion shiny with ridges becoming longitudinal. 
(27) Legs with light color the same as on face. Foreleg with cream color on 
dorsal apex of femur and knee of tibia. (28) Middle leg colored like foreleg; 
spur finely, uniformly pectinate; length of spur to basitarsal length of para- 
type as 0.46:0.71,0.69. (30) Tegula transparent testaceous with anterior patch 
of cream color. (32) Marginal cell 6-9 and 3-4 both shorter than 9-wt, 0.82: 
0.75 :0.87. 

Metasoma. (34) Tergal hair bands moderately dense, white, appressed. 
(35) Tergum 1 with punctures of median area finer than on scutum, mostly 2 
pwa, interspaces shiny (30%). Declivity of tergum 1 finely lineolate. (37) 
Sternum 6 brown without a median, clear, subcircular area. 


816 Tue Universiry ScIENCE BULLETIN 


Type Mareriat. Holotype female, Alice (10 mi. S.), Texas, July 17, 1956 
(University of Kansas Mexican Expedition), is in the Snow Entomological 
Museum at The University of Kansas. One female paratype, Piedras Negras, 
Coahuila, 1300 ft., Aug. 25, 1949 (G. M. Bradt), is at the American Museum 
of Natural History. 


Subgenus PERISSANDER Michener 


Perissander Michener, 1942, New York Ent. Soc. Jour., 50:275; Michener, 1951, zn Muesebeck 
et al., U.S. Dept. Agric., Monogr. No. 2:1104. 
Type species. Calliopsis anomoptera Michener, 1942, monobasic and original designation. 
This subgenus seems to be a derivative of the hondurasica group of Calli- 
opsis s.s. The entire subgenus is restricted to the arid northwestern Mexico 
and southwestern United States. The species prefers flowers of Euphorbia to 
which the short, stubby mouthparts seem adapted. Calliopsis gilva, C. ful- 
gida, and C. yalea are slightly different from the anomoptera group (which 
includes rogeri, syphar, and limbus) largely by reason of the elongate cylin- 
drical glossae. The floral records for gilva and fulgida include Verbesina and 
Tidestromia, members of the Compositae and Amaranthaceae, respectively, 
which also suggest an ecological separation from the anomoptera group. 
Perissander differs from the other subgenera as follows. Males lack the 
scutellar and metanotal hair pads, yet have the type of genital capsule de- 
scribed for Calliopsis s.s. Both males and females have the galeae hidden in 
repose by the closed mandibles, or in gilva, fulgida, and yalea have the length 
of the galea exposed beyond closed mandibles less than 1 mow. The peculiar, 
flat, short, truncate glossa of the anomoptera group occurs elsewhere only in 
squamifera and peninsularts. 


Femate. Length, 4.5-7.5 mm. 

Head. Light colored areas: (1) paraocular area below a line originating 
on outer subantennal suture slightly below its upper origin, and extending 
laterally from 0.50-0.75 times width of paraocular area, thence dorsally to, or 
close to, lower margin of facial fovea, thence laterally ending on orbit at level 
of it; in gilva, upper border runs sinuously from about midpoint of outer sub- 
antennal suture (to near upper end) to lower border of facial fovea ending at 
level above it on orbit. (2) Clypeus with a median area with base adjacent to 
frontoclypeal suture, shape variable, from a relatively narrow horizontal strip 
or trianguliform area, to longitudinal band with laterally expanded base, 


length about 0.66 times length of clypeus; gva with spot in lower lateral | 


corner; (3) absent on labrum; (4) supraclypeal area pentagonal or semilunar 
with apex at or below midlevel of antennal socket; (5) subantennal plate; 


(6) absent on mandible (basal dot in fulgida and yalea). (7) Scape and 


pedicel brown to black, flagellum dark above, tan below on at least apical 7 
flagellomeres. (8) Hair on vertex, frons, and clypeus variable, of gena white. 


A Revision oF THE BEE GENUS 817 


(10) Punctures along ocellocular line variable, interspaces variable; impunc- 
tate area lateral to posterior ocellus shiny (roughened and dull in limbus); 
punctures of frons with shiny interspaces (roughened in limbus). (11) 
Frontal line with lower portion a sharp carina or a finely sulcate carina. (12) 
Clypeus with punctures of disc somewhat eccentric; disc very little protuber- 
ant; apical portion adjacent to medial area bent about 45° posteriad as seen 
from below, projection of apical border smoothly rounded, not tooth-like. 
(13) Inner orbits subparallel to divergent below. Facial fovea shallow to deep, 
smaller than in other subgenera, usually oval. (14) Galea short, not visible 
with mandibles closed except in gilva, fulgida, and yalea, where length of 
galea exposed beyond closed mandibles is less than 1 mow. Glossa flat, short, 
truncate, abruptly broadened medially, flabellum absent ee in gilva, ful- 
gida, and yalea which have short, cylindrical, flabellate glossae. (15) Head 
width/head length 1.3-1.5. (17) Eye length slightly less nian mio, subequal to 
or more than a tenth less than aeeeiie length. (18) Antennocellar about 
twice antennocular; interocellar two-thirds of, to equal to ocellocular. (19) 
Ocellolabral equal to or greater than mio, subequal to clypeal width. (21) 
Basal labial palpomere slightly less than, or subequal to, length of others com- 
bined in anomoptera group, about 1.7 times length of others in gilva. (22) 
Flagellomere 1 slightly shorter than flagellomere 9. Flagellar length 2.1-2.8 
times length of scape. 

Mesosoma. (23) Light colored areas: medial interruption of pronotal 
stripe 2-3 mow; apex of pronotal lobe; scutellar crest. (24) Scutal and scutel- 
lar hairs of two kinds, longer ones fulvous to black, shorter ones fulvous. (25) 
Scutal disc with punctures distinct, interspaces polished to dulled by fine 
roughening. (26) Dorsal enclosure of propodeum variously sculptured, al- 
ways with some ridges, often quite low. (27) Legs with light color the same 
as on face. Foreleg as in Calliopsis s.s. (28) Middle leg as in Calliopsis s.s., 
except spur of gilva almost bare, with a few coarse teeth apicad. (29) Hind 
leg brown (yellow band on trochanter of yalea). (30) Tegula brown to 
transparent straw color, with anterolateral patch of yellow or cream color. 
Humeral plate brown, or testaceous with apical half yellow. (31) Wing at 
least partly smoky apically beyond cells. Stigma brown. (32) Marginal cell 
6-9 variable, 3-4 less than 9-wt. 

Metasoma. (34) Tergal hair bands white, dense, appressed. Band of ter- 
gum | broadly irene medially, of tergum 2 much less so, of other terga 
complete, except all bands complete on rogers, white to fulvous. Suberect hair 
of discs of terga 4-5 dark brown to black. Prepygidial and pygidial fimbriae 
smoky to dark brown. (35) Tergum 1 with punctures of median area sub- 
equal to or smaller than those on scutum, uniformly distributed, distinct, 
interspaces dull to shiny. Declivity of tergum 1 variable. (37) Sterna orange- 
red, brown, or black. 


818 Tue UNIversITy SCIENCE BULLETIN 


Mate. Length, 4.05.5 mm. 

Head. Light colored areas: (1) paraocular area upper limit varying be- 
tween lower and upper border of facial fovea, in latter case indented by facial 
fovea; (2) clypeus; (3) labrum; (4) supraclypeal area apex between three- 
fourths length of scape above upper rim of antennal socket to about 0.02 mm 
below middle ocellus, joined with paraocular area to make a continuous light 
colored area; (5) subantennal plate; (6) base of mandible; (7) scape, pedicel 
(tan dorsally in limbus), flagellomere 1, remaining flagellomeres tan. (8) 
Hair of vertex, frons, and clypeus white or fulvous. (10) Punctures of upper 
frons, fine, dense, fairly regularly distributed, 1 pwa or less. (11) Frontal line 
with lower portion as in female. (13) Inner orbits moderately convergent 
below. Facial fovea tiny, less than area of middle ocellus in some cases. (14) 
Galea as in female. (15) Head width/head length 1.2-1.4. (17) Eye length 
subequal to or greater than mio; flagellar length variable. (18) Interocellar 
to ocellocular variable; antennocular equal to or greater than interantennal. 
(21) Basal labial palpomere 0.5-1.8 times length of others combined. (22) As 
in female but flagellar length 3.1-3.8 times length of scape. 

Mesosoma. (23) Light colored areas: as in female but also ventral surface 
of mesosoma extending upward to lower third or half of mesepisternum. 
(24) Hair as in female except long and short scutal and scutellar hairs con- 
colorous. Scutellar and metanotal hair pads absent. (25) Scutal disc with 
punctures deep and distinct, or shallow and indistinct; interspaces smooth or 
finely roughened. (26) Dorsal enclosure of propodeum with fine ridges, ver- 
tical triangular portion delimited very faintly. (27) Legs with light color the 
same as on face. Foreleg yellow except minute spot of brown on posterior 
dorsal surface of trochanter, distitarsus testaceous. (28) Middle leg colored 
like foreleg except brown trochanteral spot larger, apicotarsus testaceous. (29) 
Hind leg colored like middle leg. (30) Tegula brown to colorless with an- 
terior patch or spot of yellow. Humeral plate basally brown, apically yellow. 
(31) Wing with apical portion beyond cells clear to distinctly light brown to 
naked eye. Costal vein testaceous to yellowish white. Stigma pale testaceous. 
(32) As in female. 

Metasoma. (34) Similar to female. (35) Tergum 1 with punctures vari- 
able. (36) Pygidial plate sunken medially, broadly rounded to somewhat 
truncate apically. (37) Sternal color similar to that of female. Shapes of 
sterna 5, 6, and 8 similar to those in Calliopsis s.s. (38) Sterna and genitalia 
as illustrated (Figs. 46-57). 

CALLIOPSIS (PERISSANDER) ANOMOPTERA Michener 
(Figs. 7, 46-49; Map 4) 
Calliopsis (Perissander) anomoptera Michener, 1942, Jour. New York Entom. Soc., 50: 2752s 


Michener, 1951, in Muesebeck et al., U.S. Dept. Agric., Monogr. No. 2: 1104; Krombein, 1961, 
Ent. News, 72: 82-83 (biology). 


A REVISION OF THE BEE GENUS 819 


The specific name obviously was applied because of the anomalous shape 
of the apical portion of the wing of the male which is unique among the 


Fic. 7. Lateral view of male Calliopsis (Perissander) anomoptera Michener. Note the great 
elongation of the middle tarsus which is characteristic of the subgenus. 


820 Tue UNIversITY SCIENCE BULLETIN 


Apoidea. The closest relative is rogert from which anomoptera is readily dis- 
tinguished by the larger size and by the orange or orange-red metasomal terga. 
In fact, anomoptera is distinguished from all other species of Calliopsis by the 
orange or orange-red metasoma. 


Femare. Length, 5.5 mm; forewing length, 3.6 mm; hindwing length, 
2.4 mm; clypeal length, 0.34 mm; scutal length, 0.78 mm. 

Head. Integumental background color black (faintly metallic in some 
specimens). Yellow areas: (1) paraocular area below a sinuous line originat- 
ing slightly below upper origin of outer subantennal suture and extending 
concavely upward to ventral margin of facial fovea and laterally to a point on 
the orbit about same level which is about 0.4 of eye length below summit of 
eye; (2) clypeus with a subtriangular spot with base bordering supraclypeal 
area and apex reaching about 0.6 of distance to clypeal apex (to all yellow 
except for brown splotching along apical margin medially; (7) scape brown 
(some specimens with extreme apex and base including basal bulb pale tes- 
taceous). (8) Hair of vertex, frons, and clypeus fulvous. (10) Punctures of 
midvertex and along ocellocular line fine, 1-2 pwa, interspaces shiny (30X). 
(11) Frontal line with lower portion carinate, ending in a slightly raised 
prominence at level of upper rim of antennal socket. (13) Orbital conver- 
gence ratio as 1.00:0.97,1.04. Facial fovea ellipsoidal, short, only slightly longer 
than mow, length to width as 0.17:0.09. (15) Head width to head length as 
1.60:1.16,1.39. (17) Eye length, mio, and flagellar length as 0.90:0.97:0.90. 
(18) Interocellar, ocellocular, antennocular, and interantennal as 0.27:0.44: 
().27:0.22. (19) Ocellolabral subequal to clypeal width, 0.97:0.99. (20) Clypeo- 
cellar to outer subantennal sutural as 0.63:0.46,1.37. (21) Basal labial palpo- 
mere subequal to length of others combined. (22) Flagellar length about 2.1 
times length of scape, 0.90:0.42. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal 
stripe about 2 mow. (24) Scutum and scutellum with numerous short, ful- 
vous hairs and less numerous, long, brownish hairs. Hair of metanotum ful- 
vous medially, white laterad. (25) Scutal disc with punctures fine, 1-2 pwa, 
interspaces shiny, smooth (30>). (26) Dorsal enclosure of propodeum with 
curved lines radiating anterolaterad from the median apical border, inter- 
spaces shiny, remainder of propodeum smooth, highly polished. (27) Foreleg 
with yellow on extreme dorsal apex of femur (sometimes absent) and basal 
third to half of tibia. (28) Middle leg colored like foreleg except tibial yellow 
slightly less extensive; spur finely, evenly pectinate, testaceous, its length 
slightly less than half of basitarsal length, 0.29:0.60. (30) Tegula transparent, 
straw colored with anterior patch of yellow. Humeral plate testaceous with 
small patch of light color. (32) Marginal cell 6-9 and 3-4 both longer than 
9-wt, 0.76 :0.70 :0.66. 


A ReEvIsION oF THE BEE GENUS 821 


Metasoma. (33) Terga 1-4 reddish-orange, sometimes with irregular 
patches of reddish-black, tergum 4 often with posterior margin black; terga 
5-6 black. (35) Tergum 1 with punctures of median area much finer than 
on scutum, densely, regularly distributed, less than 1 pwa, interspaces shiny 
(30). Declivity of tergum 1 somewhat shiny. (37) Sterna orange except 
sternum 6 dark brown to black. 


Mate. Length, 5.1 mm; forewing length, 4.5 mm; hindwing length, 3.0 
mm; clypeal length, 0.34 mm; scutal length, 0.75 mm. 

Head. Yellow areas: (1) paraocular area ending between middles of 
facial foveae, indented by them, or extending above facial foveae, surrounding 
them continuous with supraclypeal area; (4) supraclypeal area and entire 
frons to about 0.02 mm below middle ocellus. (13) Orbital convergence ratio 
as 1.00:0.87,1.15. Facial fovea ovoid, tiny, a fourth to third area of middle 
ocellus. (15) Head width to head length as 1.65:1.24,1.33. (17) Eye length, 
mio, and flagellar length as 0.94:0.87:1.29. (18) Interocellar, ocellocular, 
antennocular, and interantennal as 0.27:0.29:0.22:0.20. (19) Ocellolabral sub- 
equal to clypeal width, 0.99:1.00,1.00. (20) Clypeocellar to outer subantennal 
sutural as 0.65:0.46,1.41. (21) Basal labial palpomere about 0.66 times length 
of others combined. (22) Flagellar length about 3.6 times length of scape, 
1.29 :0.36. 

Mesosoma. (23) Yellow area of mesepisternum usually covering approxi- 
mately lower half, sometimes with yellow reaching pronotal lobe. (24) Short 
scutal hairs many-branched, appearing moss-like and partially hiding surface 
viewed from directly above. (25) Scutal disc with punctures fine, deep, dis- 
tinct, 2 pwa, interspaces smooth, shiny. (26) Dorsal enclosure of propodeum 
declivous, sunken, with rolled porterior border; median portion much longer 
than lateral portions, with fine ridges originating in strong emargination 
along posterior border, diverging laterally towards anterior border; lateral 
portions with low, fine ridges separated by about 2 ridge widths, interspaces 
slightly roughened but shiny. (28) Length of middle tibia, basitarsus, medio- 
tarsus, and apicotarsus as (0.95:1.21:1.38:1.65), basitarsus distinctly shorter 
than mediotarsus, ratio of tibia to basitarsus 0.79. (30) Tegula colorless to 
pale straw color with anterior yellow patch. (31) Forewing of unique shape 
among Apoidea (Fig. 6), apical portion beyond marginal cell bent abruptly 
posteriad, drawn into a broad, posteriorly-directed tip, brown apical portion 
in form of a band. Costal vein yellowish white. (32) Marginal cell long, 
very slender, parallel-sided apically, 6-9 about 3 times, and 3-4 about 2 times 
greater than 9-wt, 1.09:0.85:0.39; 11-12 less than 0.7 times 13-14. 

Metasoma. (33) Terga 1-3 orange, sometimes with reddish-brown areas; 
tergum 2 with lateral fovea reddish-brown; terga 4-6 reddish-brown to black. 
(35) Tergum 1 with punctures of median area fine, distinct, regularly dis- 


822 Tue University SCIENCE BULLETIN 


tributed, 0.5 pwa or less, interspaces smooth, dull (30%). (37) Sterna orange, 
except sternum 8 dark brown. 

Type Marertat. Holotype male and allotype female from Picacho Pass, 
Arizona, Aug. 7, 1940 (C. D. Michener), on a small Euphorbia, are at the 
American Museum of Natural History. The above description of the female 
is primarily based on a specimen from Douglas (1 mi. E), Arizona, Aug. 16, 
1962 (M. Statham), and that of the male on a paratype specimen. 

Distrisution. The southwestern United States and northwestern Mexico, 
including Baja California. It occurs from the first week of July to the last 
week of September in New Mexico and Arizona, and has been taken in Cali- 
fornia from early September to early October. The latest date of capture is at 
La Paz, Baja California Sur, on Oct. 10, 1955 (F. X. Williams). It is a 
sonoran desert form which occurs where Euphorbia flourishes. 

Approximately 255 specimens were examined from the following localities which include the 


type: Arizona: Apache (5 mi. S.E.), Cochise Co.; Brenda (2 mi. W.), Yuma Co.; Douglas (1, 17 
mi. E.; 3 mi. N.; 16 mi. N.E.), Cochise Co.; Gila Bend (18 mi. S.), Maricopa Co.; Kingman (10 


cit: anomoptera 
Chaar 030005 | ! 


YX) OD rogeni 


7 \ iS: » @ gilva 


N\A A limbus 


Map 4. Map showing the known distributions of Calliopsis (Perissander) anomoptera Michener, 
C. (P.) rogeri Shinn, C. (P.) gilva Shinn, and C. (P.) limbus Shinn. 


A RevIsION oF THE BEE GENUS 823 


mi. N.W.), Mohave Co.; Lowell (5 mi. E.), Cochise Co.; Picacho Pass; Portal, Chiricahua Mts., 
Cochise Co.; Sabino Canyon, near Tucson; Santa Rita Mts., 5000 ft. to 8000 ft.; Sells, Pima Co.; 
Tucson (10, 23 mi. S.); Wenden; Willcox, Cochise Co. CarirorniA: Palm Springs, Riverside Co.; 
Twentynine Palms. New Mexico: Granite Pass, Hidalgo Co.; Rodeo (11, 18 mi. N.), Hidalgo 
Co.; Rodeo, Cienaga Lake, Hidalgo Co.; Willow Creek Mts. Baya Cauirornia: La Paz. Sonora: 
Sonoita (26 mi. E.). 

Bionomics. This species is one of several which favor Euphorbia for a 
food plant for both pollen and nectar. Krombein (1961) reported that he 
found it in abundance on Euphorbia albomarginata from July 23 to 31, 1959, 
and that P. D. Hurd collected it at the same place and flower August 9 to 15, 
1958, both at Portal, Arizona, 4000 ft. altitude, near the S. W. R. S., American 
Museum of Natural History. Krombein (op. cit.) mentioned that only males 
were taken on July 23 and 24, most of them newly emerged, and that freshly 
emerged females were present on the flowers on July 26th. Two of the 
females captured bore pollen masses composed entirely of Euphorbia pollen. 

Frower Recorps. Cladothrix lanuginosa, Eriogonum, Euphorbia albo- 
marginata, E.. capitellata, E. hirtella, E. pleniradiata, E. polycarpa hirtella, E. 
polycarpa typica, Lepidium thurberi, Tidestromia. 


CALLIOPSIS (PERISSANDER) ROGERI, new species 
(Figs. 50-53; Map 4) 


I take pleasure in naming this smallest species of Calliopsis after my son, 
Roger, who has helped me in many ways during the study, both in the field 
and the laboratory. 

The species is closest to anomoptera and syphar, but both sexes are readily 
distinguished from anomoptera by their black, metallic integument, and total 
absence of any trace of integumental orange-red color. The male of syphar is 
unknown but the female of rogeri differs from that of syphar by the strong 
brassy tints on head, mesosoma, and especially the metasomal terga. 


Femate. Length, 4.6 mm; forewing length, 3.1 mm; hindwing length, 2.2 
mm; clypeal length, 0.31 mm; scutal length, 0.82 mm. Integumental back- 
ground color black with faint greenish metallic tint on head, strong brassy 
metallic tint on mesosomal dorsum, very strong brassy tint on metasomal 
terga. 

Head. Yellow areas: (1) paraocular area as described for subgenus; (2) 
clypeus with T-shaped median figure, ventral end of shaft of T extending to 
middle of clypeus (T often filled in to form a triangular patch). (8) Hair of 
vertex, frons, and clypeus fulvous. (10) Punctures along ocellocular line fine, 
dense, regularly distributed, 1 pwa, interspaces shiny. (11) Frontal line with 
lower portion a finely sulcate carina. (13) Orbital convergence ratio as 1.02: 
0.94,1.09. Facial fovea ovoid, short, only slightly longer than mow. (15) Head 
width to head length as 1.56:1.12,1.39. (17) Eye length, mio, and flagellar 


824 Tue UNiversiry SclIENCE BULLETIN 


length as 0.83 :0.94:0.88. (18) Interocellar, ocellocular, antennocular, and inter- 
antennal as 0.31 :0.25:0.27:0.18. (19) Ocellolabral equal to clypeal width, 0.95: 
0.95,1.00. (20) Clypeocellar to outer subantennal sutural as 0.65 :0.44,1.46. (21) 
Basal labial palpomere about 0.85 times length of others combined. (22) 
Flagellar length about 2.2 times length of scape, 0.88 :0.39. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal stripe 
about 2 mow. (24) Scutal black hairs twice length of short, fulvous hairs, or 
more. (25) Scutal disc with punctures almost hidden, viewed from above, by 
plumosity of short, fulvous hairs; punctures fine, deep, distinct, 1-2 pwa, inter- 
spaces shiny, metallic. (26) Dorsal enclosure of propodeum with fine, low 
ridges originating along posterior, median border, passing anterolaterally in 
parallel, symmetrical, curves. (27) As in anomoptera but only basal fourth to 
third of tibia yellow. (28) Middle leg colored like foreleg; spur evenly, finely 
pectinate, length of spur about half of basitarsal length, 0.27:0.56. (30) Tegula 
brown with anterolateral patch of yellow. Humeral plate brown. (32) Mar- 
ginal cell 6-9 subequal to, and marginal cell 3-4 distinctly shorter than 9-wt, 
0.68 :0.53 :0.66. 

Metasoma. (34) All tergal hair bands complete. (35) Tergum 1 with 
punctures of median area smaller than on scutum, very fine, denen regularly 
distributed, 0.5 pwa, interspaces moderately shiny. Declivity of tergum 1 dull. 


(37) Sterna black (to dark brown). 


Mate. Length, 4.5 mm; forewing length, 3.3 mm; hindwing length, 2.15 
mm; clypeal length, 0.34 mm; scutal length, 0.65 mm. 

Head. Yellow areas: (1) paraocular area extending to lower border of 
facial fovea, indented by it (usually), continuous with supraclypeal area; (4) 
supraclypeal area extending above to same level as (to slightly below) para- 
ocular yellow, both well below middle ocellus. (11) Frontal line with lower 
portion a sharper carina than in anomoptera. (13) Orbital convergence ratio 
as 0.95 :0.82,1.16. Facial fovea ovoid to tear drop shaped, area about half area 
of middle ocellus. (15) Head width to head length as 1.50:1.11,1.35. (17) Eye 
length, mio, and flagellar length as 0.82:0.82:1.11. (18) Interocellar, ocellocu- 
lar, antennocular, and interantennal as 0.26:0.29:0.23:0.17. (19) Ocellolabral 
greater than clypeal width, 0.94:0.90,1.03. (20) Clypeocellar to outer suban- 
tennal sutural as 0.60:0.53,1.40. (21) Basal labial palpomere about 0.85 times 
length of others combined. (22) Flagellar length about 3.4 times length of 
scape, 1.11:0.32. 

Mesosoma. (23) Yellow area of mesepisternum covering somewhat less 
than lower half, less extensive than in anomoptera. (24) As in anomoptera 
except denser. (25) As in anomoptera but punctures 1 pwa. (26) Dorsal en- 
closure of propodeum subhorizontal, not sunken, posterior border weakly or 
not defined; median portion as in anomoptera, lateral portions roughened 
with few mesally oblique ridges. (28) Lengths of middle tibia, basitarsus, 


A Revision oF THE BEE GENUS 825 


mediotarsus, and apicotarsus as 0.78 :0.97 :1.19:1.36, basitarsus distinctly shorter 
than mediotarsus, ratio of tibia to basitarsus 0.81, about as in anomoptera. 
(30) Tegula pale straw color with anterior yellow patch. (31) Forewing 
normal, brown apical portion in form of a round spot, slightly flattened on 
side toward wing base. Costal vein pale testaceous. (32) Marginal cell 6-9 
longer than, and 3-4 shorter than 9-wt, 0.71:0.58:0.66; 11-12 equal to or only 
slightly shorter than 13-14. 

Metasoma. (33) Terga black with brassy tints. (34) As in female. (35) 
Tergum 1 with punctures of median area fine, shallower and not so distinct 
as in anomoptera, less regularly distributed, 0.5 pwa or less. Interspaces 
roughened, dull (30%). (37) Sterna brown to black. 

Tyre Marerrar. Holotype male and allotype female from Douglas (1 mile 
E.), Cochise Co., Arizona, Aug. 16, 1962 (M. Statham), are at the American 
Museum of Natural History. 

Twelve male and 10 female paratypes were collected at the following localities: Arizona: 
Apache (5 mi. S.E.), Cochise Co., 1 male, Aug. 11, 1958 (D. D. Linsdale), 5 males, same date 
(R. M. Bohart), Euphorbia; idem (14 mi. S.W.), 2 males, 2 females, Aug. 7, 1961 (J. G. Rozen), 
Euphorbia; Douglas (1 mi. E.), Cochise Co., 2 males, 1 female, Aug. 16, 1962 (M. Statham); 
idem (3 mi. N.), 4 males, 1 female, Aug. 4, 1961, Euphorbia, 2 males, 2 females, Aug. 8, 1961 
(all J. G. Rozen); idem (17 mi. E.), 1 male, Aug. 4, 1958 (D. D. Linsdale); zdem, 4 males, 1 
female, Aug. 8, 1958, Euphorbia, 1 male, same date, Lepidium thurberi (all P. D. Hurd); dem, 
2 females, Aug. 8, 1958 (R. M. Bohart), Euphorbia; Lowell (5 mi. E.), Cochise Co., 1 male, 
Aug. 15, 1958 (P. M. Marsh); Portal, Cochise Co., 1 male, Aug. 12, 1958 (P. D. Hurd), 
Euphorbia; idem, | female, July 23-31, 1959 (K. V. Krombein), Euphorbia albomarginata; idem 
(2 mi. N.E.), “Site B”, 1 female, Sept. 25, 1961 (M. A. Cazier); Willcox, Cochise Co., 1 male, 
Aug. 14, 1958 (D. D. Linsdale), Euphorbia; idem (1 mi. S.), 1 female, Sept. 8, 1959 (G. I. 
Stage). Paratypes will be deposited in the collection of P. H. Timberlake, University of California, 
Riverside, California, the California Academy of Science, the University of California at Berkeley, 
the American Museum of Natural History, and the personal collections of Dr. Karl V. Krombein 
of the U.S. National Museum, Mr. Roy E. Snelling, Los Angeles, California, and The University 
of Kansas, Lawrence. 

DistripuTion. Known only from the southeastern corner of Arizona but 
probably will be found in at least the adjacent parts of México and New 
Mexico. Its flight season is in August and September. 

Frower Recorps. Euphorbia albomarginata, E. sp., and Lepidium thur- 
beri. It has been taken with Calliopsis anomoptera on these plants. A study 


of the competition between these two species would be enlightening. 


CALLIOPSIS (PERISSANDER) SYPHAR, new species 
(Map 3) 


The specific name is derived from the Greek, sypharos, meaning wrinkled 
skin, in allusion to the appearance of the dorsal enclosure of the propodeum 
which suggests the wrinkled character of the elephant’s skin. The specimens 
exhibit strong similarities to C. (Perissander) roger, notably in the charac- 
teristic of the dorsal enclosure of the propodeum, though expressed more 
strongly than in rogert. C. syphar is readily distinguished from roger: by the 


826 Tue UNiversiry SciENCcE BULLETIN 


non-metallic integument, larger size, proportionately longer mesotibial spur, 
and by the marginal cell 6-9 being about the same length as 3-4 and 9-wt. 


Femare. Length, 5.5 mm; forewing length, 3.5 mm; hindwing length, 2.5 
mm; clypeal length, 0.39 mm; scutal length, 0.85 mm. 

Head. Cream colored areas: (1) paraocular area below a line originating 
at upper end of outer subantennal suture and extending horizontally to about 
midline of area thence curving sharply upward to level of lower margin of 
facial fovea, ending on orbit about half eye length below summit of eye. (2) 
Clypeus with a broad T-shaped maculation (to only the vertical stem of T 
present), the crossbar of the T along the border of the supraclypeal area (to 
along supraclypeal area and both subantennal plates), the vertical stem extend- 
ing about two-thirds of distance to clypeal apex. (8) Hair of vertex fulvous 
(to light brown), twice length of longer scutal hairs, of frons and clypeus 
fulvous. (10) As in rogeri. (11) Frontal line with lower portion a barely 
discernible (30) sulcus becoming obsolete about midlength of scape, frontal 
prominence with a minute point. (13) Orbital convergence ratio as 1.11:1.07, 
1.02. Facial fovea ovoid, very shallow, width about half length, shallower and 
broader than rogert. (15) Head width to head length as 1.75:1.22,1.43. (17) 
Eye length, mio, and flagellar length as 0.97:1.07:1.05. (18) Interocellar, 
ocellocular, antennocular, and interantennal as 0.34:0.31:0.31:0.22. (19) Ocel- 
lolabral slightly greater than clypeal width, 1.09:1:04,1.05. (20) Clypeocellar 
to outer subantennal sutural as 0.70:0.51,1.36. (21) Basal labial palpomere 
about 1.1 times length of others combined. (22) Flagellar length about 2.4 
times length of scape, as 1.05:0.44. 

Mesosoma. (23) Light areas cream colored; medial interruption of pro- 
notal stripe about 2 mow. (24) As in rogeri. (25) Scutal disc with punctures 
finer than in rogerz, deep, about 2 pwa, interspaces shiny, smooth. (26) Dorsal 
enclosure of propodeum with fine lines radiating anterolaterad from the 
median apical border as in roger: but more pronounced, numerous, longer, 
and originating along a broader part of rear border of dorsal enclosure, inter- 
spaces roughened, and remainder of propodeal triangle finely roughened. (27) 
Foreleg with cream color on dorsal apex of femur and basal third of tibia, 
tarsus brown. (28) Middle leg colored like foreleg; spur linely, evenly pec- 
tinate, testaceous, its length about 0.6 times basitarsal length, 0.37:0.61. (30) 
Tegula transparent, almost colorless, with small anterior patch of cream color. 
Humeral plate testaceous. (32) Marginal cell 6-9 and 3-4 greater than 9-wt, 
0.73 :0.71 :0.68. 

Metasoma. (35) Tergum 1 with punctures of median area smaller than on 
scutum, densely, regularly distributed, 1 pwa or less, interspaces shiny (30). 
Declivity of tergum 1 finely lineolate, shiny. (36) As in rogert. 


A REVISION OF THE BEE GENUS 827 


Tyre Marteriar. Holotype female from San Ignacio (15 mi. N.), Baja 
California Sur, Sept. 29, 1941 (E. S. Ross and R. M. Bohart) is in the Cali- 
fornia Academy of Sciences, San Francisco. 

Five female paratypes are from the localities below: Baya Ca irorNnia 
Sur: Canipole, 2 females, Oct. 2, 1941 (Ross and Bohart); La Paz, 1 female, 
Oct. 10, 1955 (F. X. Williams) ; San Ignacio (15 miles N.), 1 female, Sept. 29, 
1941 (Ross and Bohart) ; San Pedro, 1 female, Oct. 7, 1941 (Ross and Bohart). 


CALLIOPSIS (PERISSANDER) LIMBUS, new species 
(Map 4) 


The species name is from the Latin, /imbus, meaning border, in reference 
to its occurrence near the Mexican and United States borders. It is closest to 
C. rogeri but the male is readily distinguished from rogeri by the faintly 
smoky wing tip and the finely roughened interspaces between mesoscutal 
punctures. The abdomen of the male holotype is missing. Both the female 
allotype and single female paratype are the same species, but since they were 
not collected with the male, the association of the sexes is tentative. The 
female of limbus differs from that of roger: by the higher ratio of flagellar 
length to scape length (average 2.7 to 2.3 for rogeri), by the marginal cell 6-9 
being less than 9-wt, and by the lack of metallic sheen on the dorsum of the 
thorax and abdomen. 


Femate. Length, 4.4 mm; forewing length, 3.1 mm; hindwing length, 2.19 
mm; clypeal length, 0.34 mm; scutal length, 0.82 mm. 

Head. Pale yellow areas: (1) paraocular area as described for subgenus; 
(2) clypeus, narrow strip extending along frontoclypeal suture between mid- 
dles of subantennal plates, with a broad, ventrally directed median projection 
of light color. (8) Hair of vertex, frons, and clypeus white. (10) Punctures 
along ocellocular line as in roger: except interspaces roughened, impunctate 
area beside lateral ocellus roughened. (11) Frontal line with lower portion a 
low, sharp carina. (13) Orbital convergence ratio as 0.97:0.95,1.02. Facial 
fovea elliptical, minute, area smaller than that of lateral ocellus; length 
slightly greater than mow. (14) Glossa narrower medially than that of rogerz, 
shape somewhat intermediate between rogert and fulgida. (15) Head width 
to head length as 1:53:1.16,1.32. (17) Eye length, mio, and flagellar length as 
0.87 :0.95:1.02. (18) Interocellar, ocellocular, antennocular, and interantennal 
as 0.29 :0.29:0.27:0.20. (19) Ocellolabral slightly greater than clypeal width, 
1.00:0.95,1.05. (20) Clypeocellar to outer subantennal sutural as 0.66 :0.44,1.50. 
(21) Basal labial palpomere 1.8 times length of others combined. (22) Flagel- 
lar length about 2.5 times length of scape, 1.02:0.39. 

Mesosoma. (23) Light areas pale yellow; medial interruption of pronotal 
stripe 1-2 mow. (24) Scutal dark hairs only slightly longer than short, fulvous 


828 Tue UNIversITy SCIENCE BULLETIN 


hairs. (25) Scutal disc with punctures not obscured by hair; punctures finer 
than in roger1, 2 pwa, interspaces faintly roughened, shiny, becoming crowded 
with more distinct roughening of interspaces anteriad. (26) Dorsal enclosure 
of propodeum with about 18 longitudinal or obliquely longitudinal ridges 
with interspaces heavily roughened. (27) Foreleg with yellow on extreme 
dorsal apex of femur and base of tibia. (28) Middle leg colored like foreleg; 
spur evenly, finely pectinate, length of spur about 0.6 (to 0.5) times basitarsal 
length, 0.31:0.51. (30) Tegula pale brown with anterolateral patch of pale 
yellow. Humeral plate pale brown. (32) Marginal cell 6-9 less than, and 3-4 
much less than 9-wt, 0.63 :0.53 :0.68. 

Metasoma. (35) Tergum 1 with punctures of median area very fine, dense, 
regularly distributed, 0.5 pwa, interspaces roughened, quite dull. Declivity of 
tergum | as in rogert. (37) Sterna black. 


Mare. Length, N. A. (abdomen missing) ; forewing length, 3.3 mm; hind- 
wing length, 2.4 mm; clypeal length, 0.34 mm; scutal length, 0.78 mm. 

Head. Yellow areas: (1) paraocular area as in roger? but separated from 
supraclypeal area by intervening black less than 0.02 mm wide at antennal 
socket; (4) supraclypeal area extending above to a level 0.8 mow below sum- 
mit of paraocular yellow. (11) Frontal line with lower portion as in rogert. 
(13) Orbital convergence ratio as 1.04:0.90,1.15. Facial fovea elliptical, area 
about half of area of middle ocellus. (15) Head width to head length as 1.53: 
1.22,1.25. (17) Eye length, mio, and flagellar length as 0.88:0.90:1.14. (18) 
Interocellar, ocellocular, antennocular, and interantennal as 0.29 :0.32:0.26 :0.22. 
(19) Ocellolabral longer than clypeal width, 1.02:0.97,1.05. (20) Clypeocellar 
to outer subantennal sutural as 0.68:0.46,1.48. (21) Basal labial palpomere 
about 1.1 times length of others combined. (22) Flagellar length about 3.2 
times length of scape, 1.14:0.36. 

Mesosoma. (23) Yellow area of mesepisternum covering less than lower 
one-third, less extensive than either rogeri or anomoptera. (24) As in rogert. 
(25) Scutal disc with punctures fine, shallow, larger than roger: or anomop- 
tera, 1 pwa, interspaces finely roughened, moderately shiny. (26) Dorsal 
enclosure of propodeum subhorizontal, slightly sunken, posterior border de- 
fined by low carina; median and lateral portions with numerous, slightly 
vermiform, longitudinal ridges separated by 1-2 ridge widths, dull. (28) 
Lengths of middle tibia, basitarsus, mediotarsus, and apicotarsus as 0.83 :0.94: 
0).99:1.17, basitarsus only slightly shorter than mediotarsus, ratio of tibia to 
basitarsus 0.89, higher than in rogeri and anomoptera. (30) Tegula pale straw 
color with anterior yellow patch. (31) Forewing normal, apical portion 
beyond cells clear to naked eye, slightly smoky beyond marginal cell (30X). 
Costal vein pale testaceous. (32) Marginal cell 6-9 and 34 both shorter than 
9-wt, 0.68 :0.56:0.75; 11-12 slightly longer than 13-14. 

Metasoma. Missing. 


A REVISION oF THE BEE GENUS 829 


Tyre Materia. Holotype male from Sabino Canyon, near Tucson, Ari- 
zona, July 31, 1941 (L. H. Banker), and allotype female from Granite Pass, 
15 mi. N.E. Rodeo, Hidalgo Co., New Mexico, Aug. 25, 1958 (P. D. Hurd), 
on Euphorbia, are in the Snow Entomological Museum at The University of 
Kansas, Lawrence, and at the University of California, Berkeley, respectively. 
One female paratype, Mt. Lemmon Road, Arizona, 3500 ft. altitude, Aug. 15, 
1954 (R. M. Bohart), is at the University of California, Riverside. 

DisrripuTion. Known only from southeastern Arizona and southwestern 
New Mexico. 

Frower Recorps. Euphorbia. 


CALLIOPSIS (PERISSANDER) GILVA, new species 
(Figs. 54-57; Map 4) 


The specific name from the Latin, gilvws, meaning pale yellow, is applied 
in reference to the large amount of pale yellow color on the male face, an- 
tenna, and legs. 

The species is closest to limbus. The female is readily separated from all 
others of the genus by its unique mesotibial spur which is bare except for 2-4 
coarse teeth on the apical four-tenths; both male and female are separated 
from limbus by the shiny, non-roughened interspaces on the head and scutum. 
The male is readily separated from limbus also by the non-metallic integu- 
ment on the body. 


Fremare. Length, 7.5 mm; forewing length, 4.4 mm; hindwing length, 3.1 
mm; clypeal length, 0.49 mm; scutal length, 1.19 mm. 

Head. Yellow areas: (1) paraocular area below a sinuous line originating 
at about middle (to upper origin) of outer subantennal suture and extending 
obliquely upward ending on orbit about 0.4 of eye length below summit of 
eye, slightly above lower rim of facial fovea; (2) clypeus with a vertical stripe 
from margin below supraclypeal area to within less than mow of apex, width 
of stripe slightly less than width of supraclypeal area (and a small dot in 
lower corner of clypeus in some specimens); (8) Hair of vertex and clypeus 
both brown and fulvous, of frons fulvous. (10) Punctures of midvertex with 
interspaces shiny, along ocellocular line 2 pwa, interspaces shiny (30). 
Median punctures of frons fine, deep, mostly 1 pwa, interspaces shiny (30). 
(11) Frontal line with lower portion sulcate, ending at level of middle of 
antennal socket. (13) Orbital convergence ratio as 1.36:1.34,1.01. Facial fovea 
elongate, length about 2 mow. (14) Galea pebbled apically, shiny medially, 
length exposed beyond closed mandibles to galeal gap as 1.10:0.56. (15) Head 
width to head length as 2.23:1.62,1.38. (17) Eye length, mio, and flagellar 
length as 1.24:1.34:1.22. (18) Interocellar, ocellocular, antennocular, and inter- 
antennal as 0.37 :0.39:0.37:0.31. (19) Ocellolabral less than clypeal width, 1.34: 


830 Tue UNiversity ScIENCE BULLETIN 


1.41,0.95. (20) Clypeocellar to outer subantennal sutural as 0.85:0.70,1.22. 
(21) Basal labial palpomere about 1.5 times length of others combined. (22) 
Flagellar length about 2.2 times length of scape, 1.22:0.57. 

Mesosoma. (23) Light areas yellow; medial interruption of pronotal stripe 
about 2 mow. (24) Scutum and scutellum with similarly sized, short fulvous 
hairs and brown hairs twice as long. Hair of metanotum fulvous except long, 
posteriorly directed hair whitish. (25) Scutal disc with fine punctures mostly 
1 pwa, interspaces shiny, smooth (30). (26) Dorsal enclosure of propodeum 
a reticulum of fine, vermiform ridges, dull medially but with interspaces shiny 
laterad, remainder of propodeal triangle shiny. (27) Foreleg with yellow on 
extreme dorsal apex of femur and knee of tibia. (28) Middle leg colored like 
foreleg; spur bare basally, with three sharp teeth on apical 0.4; length of spur 
about 0.7 times basitarsal length, 0.56:0.83. (30) Tegula transparent testaceous 
with anterior patch of yellow. Humeral plate testaceous, broadly margined 
with yellow. (32) Marginal cell 6-9 slightly longer than, and 3-4 much shorter 
than 9-wt, 0.90:0.65 :0.88. 

Metasoma. (35) Tergum 1 with punctures of median area finer than on 
scutum, somewhat irregularly distributed, less than 1 pwa, interspaces shiny 
(30). Declivity of tergum 1 finely lineolate. (37) Sterna brown. 


Mate. Length, 4.7 mm; forewing length, 3.7 mm; hindwing length, 2.6 
mm; clypeal length, 0.37 mm; scutal length, 0.80 mm. 

Head. Yellow areas: (1) paraocular area as in limbus (to broadly con- 
tinuous with supraclypeal area); (4) supraclypeal area extending above to 
level of summit of paraocular yellow (to 0.02-0.03 mm below). (11) Frontal 
line with lower portion as in rogerz. (13) Orbital convergence ratio as 0.99: 
().82,1.20. Facial fovea elliptical, less distinct than in other Perissander species, 
area as in limbus. (15) Head width to head length as 1.65:1.24,1.33. (17) Eye 
length, mio, and flagellar length as 0.95:0.82:1.39. (18) Interocellar, ocellocu- 
lar, antennocular, and interantennal as 0.31:0.31:0.20:0.20. (19) Ocellolabral 
greater than clypeal width, 1.05:0.97,1.09. (20) Clypeocellar to outer suban- 
tennal sutural as 0.68:0.44,1.54. (21) Basal labial palpomere about 1.7 times 
length of others combined. (22) Flagellar length about 3.7 times length of 
scape, 1.39:0:37. 

Mesosoma. (23) Yellow area of mesepisternum covering approximately 
lower half (variable, from less than lower third to more than lower half). 
(24) Short scutal hairs not many-branched, not moss-like nor partially hiding 
surface viewed from directly above. (25) As in anomoptera. (26) Dorsal 
enclosure of propodeum subhorizontal, slightly sunken, with rolled posterior 
border, area much reduced compared with other Perissander species; median 
and lateral portions with numerous, slightly vermiform, longitudinal ridges 
separated by 1 ridge width, moderately shiny. (28) Lengths of middle tibia, 
basitarsus, mediotarsus, and apicotarsus as 0.97:1.11:0.87:1.09, basitarsus dis- 


A ReEvIsION OF THE BEE GENUS 831 


tinctly longer than mediotarsus, slightly longer than (to subequal to) apico- 
tarsus, ratio of tibia to basitarsus 0.87. (30) Tegula pale straw color with 
anterior yellow patch (to spot). (31) Forewing normal, apical portion beyond 
cells faintly smoky to naked eye. Costal vein yellowish white basally to tes- 
taceous apically. (32) Marginal cell 6-9 subequal to, and 3-4 much shorter 
than 9-wt, 0.78 :0.58 :0.77. 

Metasoma. (35) Tergum 1 with punctures of median area fine, larger 
than in other species of Perissander, deep, distinct, regularly distributed, 0.5-1 
pwa, interspaces smooth, shiny (30%). Declivity of tergum 1 somewhat 
shiny. (37) Sterna brown. 

Tyre Marertat. Holotype male from Douglas (17 mi. E.), Cochise Co., 
Arizona, August 8, 1958 (R. M. Bohart), on Euphorbia, is in the collection of 
P. H. Timberlake at the University of California, Riverside, California. Allo- 
type female from Douglas (1 mi. E.), Arizona, August 16, 1962 (M. 
Statham), is at the American Museum of Natural History. 

Two male and seven female paratypes were collected from the following localities: ARIZONA: 
Douglas (1 mi. E.), 4 females, Aug. 16-17, 1962 (M. Statham); Quijotoa (30 mi. E.), Pima Co., 
] male, 1 female, Aug. 28-29, 1927 (Cornell University, Lot 542 Sub 336); Tucson (10 mi. S.), 
1 male, Aug. 7, 1940 (C. D. Michener), on Verbesina. New Mexico: Rodeo (4.5 mi. N.), Hidalgo 
Co., 1 female, Aug. 21, 1962 (J. G. Rozen, M. Statham, S. J. Hessel); idem (4.8 mi. N), 1 
female, Sept. 4, 1961 (P. D. Hurd), on Tidestromia lanuginosa. Paratypes will be deposited at 


Cornell University, the Snow Entomological Museum of The University of Kansas, the University 
of California at Berkeley, the American Museum of Natural History, and in the author’s collection. 


Distrigution. Known only from southeastern Arizona and southwestern 
New Mexico. Its flight season is August and early September, which is similar 
to that of rogeri and limbus. 

Frower Recorps. Euphorbia, Verbesina, and Tidestromia lanuginosa. 
These records represent three families, viz., Euphorbiaceae, Compositae, and 
Amaranthaceae. Although taken mostly on Euphorbia, this bee apparently 
visits other plants, perhaps more extensively than Ezphorbia. Its mouthparts 
are short as in the other species of Perissander, but the glossa is cylindrical and 
the labial palp is much longer and with segments of different proportions 
than the other species. A comparison of mouthparts of Ewphorbia-visiting 
bees to discover the adaptations necessary for visiting this flower would be 
interesting. 


CALLIOPSIS (PERISSANDER) FULGIDA, new species 
(Map 3) 


The specific name is from the Latin fulgidus, meaning shining, and is 
applied because of the overall high lustre of the black integument—the shini- 
est of the genus. Its closest relative is gi/va, from which it is easily distin- 
guished by the impunctate posterior area on the first metasomal tergum. It is 
distinguished from all other species of Calliopsis s.s. and Pertssander by the 
impunctate posterior area on metasomal tergum 2 as well. A fundamental 


832 Tue UNiversiry SCIENCE BULLETIN 


difference between fulgida and gilva occurs in the mouthparts. C. fulgida has 
a cylindrical, flabellate glossa which is shorter than the maxillary palp, 
whereas gilva has the same type of glossa but it is distinctly longer than the 
maxillary palp. 

Femare. Length, 6.4 mm; forewing length, 4.3 mm; hindwing length, 3.1 
mm; clypeal length, 0.48 mm; scutal length, 1.02 mm. 

Head. White to cream colored areas: (1) paraocular area below a sinuous 
line originating at about middle (to upper origin) of outer subantennal suture 
and extending diagonally upward ending on orbit about 0.4 of eye length 
below summit of eye, slightly above level of lower rim of facial fovea; (2) 
clypeus with a median T-shaped mark (one arm of crossbar partially missing 
in type), the crossbar adjacent to supraclypeal and subantennal areas, the 
vertical portion reaching to preapical groove; small dots adjacent to apical 
border laterally; (6) mandible with a faint basal dot. (8) Hair of vertex and 
clypeus both brown and fulvous, of frons fulvous. (10) Punctures of mid- 
vertex with interspaces shiny, along ocellocular line 3 pwa, interspaces shiny 
(30>), both areas with punctures finer than in gi/va. Median punctures of 
frons finer than gilva, 2 pwa, interspaces shiny (30). (11) Frontal line with 
lower portion sulcate, much wider and deeper than in gilva. (13) Orbital 
convergence ratio as 1.24:1.21,1.03. Facial fovea as in gilva. (14) Galea 
pebbled, length exposed beyond closed mandibles to galeal gap as 0.07:0.56 
(0.09:0.44). (15) Head width to head length as 2.06:1.51,1.36. (17) Eye 
length, mio, and flagellar length as 1.16:1.21:1.09. (18) Interocellar, ocellocu- 
lar, antennocular, and interantennal as 0.39:0.32:0.34:0.31. (19) Ocellolabral 
slightly shorter than clypeal width, 1.28:1.29,0.99. (20) Clypeocellar to outer 
subantennal sutural as 0.77:0.54,1.40. (21) Basal labial palpomere about 1.5 
times length of others combined. (22) Flagellar length about 2.1 times length 
or scape, 1.092051 

Mesosoma. (23) Light areas cream colored; medial interruption of pro- 
notal stripe 1.8 mow. (24) As in gilva except brown hairs faded to fulvous on 
type. (25) Scutal disc with fine punctures 2-3 pwa, interspaces highly pol- 
ished (30%). (26) Dorsal enclosure of propodeum with longitudinal, shiny 
ridges and interspaces, remainder of propodeal triangle highly polished. (27) 
Foreleg with cream color as in gilva. (28) Middle leg colored like foreleg; 
spur with about 7 fine, uniformly-spaced teeth, length of spur about 0.6 times 
basitarsal length, 0.43:0.68. (30) Tegula transparent testaceous with anterior — 
patch of cream color. Humeral plate as in gilva. (32) Marginal cell 6-9 longer 
than, and 3-4 shorter than 9-wt, 0.90:0.77:0.85. | 

Metasoma. (35) Tergum 1 with punctures of anterior half of median area 
finer than on scutum, 1-3 pwa, of posterior half virtually absent, interspaces 
highly polished. Declivity of tergum 1 smooth, finely punctate. (37) As in 
gilva. 


oe) 
Oo 
eS) 


A REVISION OF THE BEE GENUS 


Tyre Matertav. Holotype female, from Rodeo (4.8 mi. N.), Hidalgo Co., 
New Mexico., Sept. 4, 1961 (P. D. Hurd), on Tidestromia lanuginosa, is in 
the collection of the California Insect Survey, University of California, Berke- 
ley. One female paratype from Portal (5 mi. E.), Arizona, Sept. 16, 1956 (G. 
E. Bohart), on Euphorbua, is in the collection of the collector at Logan, Utah. 

Discussion. P. D. Hurd has taken both gilva and fulgida on Tidestromia, 
and both have been taken on Euphorbia, records which support the placement 
of fulgida as a close relative of gilva. 


CALLIOPSIS (PERISSANDER) YALEA, new species 
(Map 3) 


The specific name is from the Greek, yaleos, meaning highly polished. 
The species has no close affinities and although the facial marking is similar 
to C. syphar, it is readily distinguished from syphar by the much larger size. 
The stout setae of the ventral surface of the basal labial palpomere are similar 
to those of the Calliopsis hondurasica group, but they are much shorter, 
thicker, and straighter. 

Femare. Length, 7.5 mm; forewing length, 5.2 mm; hindwing length, 3.5 
mm; clypeal length, 0.59 mm; scutal length, 1.36 mm. 

Head. Yellowish areas: (1) paraocular area below a line originating on 
outer subantennal suture about at level of middle of antennal socket and ex- 
tending dorsally 0.03 mm then laterally slightly more than half width of area, 
curving concavely upward to just below facial fovea and ending on orbit about 
at level of lower border of facial fovea, 0.44 times length of eye below summit 
of eye; lower mesal portion of area somewhat tumid; (2) clypeus with a 
broad T-shaped area adjacent to subantennal plates and supraclypeal area, the 
lower part of vertical bar 2 times mow, extending to within 0.5 mow of clypeal 
apex; (6) mandible, a more or less indistinct spot. (8) Hair of vertex brown, 
of frons white, of clypeus both fulvous and brown. (10) Punctures of mid- 
vertex with interspaces roughened, along ocellocular line of mixed sizes, 1-2 
pwa, interspaces slightly roughened. (11) Frontal line with lower portion a 
sulcus, frontal prominence ending about level of middle of antennal socket. 
(13) Orbital convergence ratio as 1.53:1.46,1.04. Facial fovea as in syphar. 
(14) Galea shiny, length exposed beyond closed mandibles to galeal gap as 
0.15:0.65. (15) Head width to head length as 2.50:1.87,1.32. (17) Eye length, 
mio, and flagellar length as 1.39:1.46:1.33. (18) Interocellar, ocellocular, an- 
tennocular, and interantennal as 0.43:0.43:0.39:0.36. (19) Ocellolabral less 
than clypeal width, 1.55:1.63,0.95. (20) Clypeocellar to outer subantennal 
sutural as 0.97:0.77,1.26. (21) Basal labial palpomere about 1.7 times length of 
others combined; seven stout setae ventrally on basal five-eighths, two addi- 
tional setae paired apically, the mesoapical seta longer, thicker, all setae pro- 


834 Tue UNiversiry SCIENCE BULLETIN 


portionately shorter and thicker than those of Calliopsis (C). hondurasica 
group. (22) Flagellar length about 2.1 times length of scape, 1.33:0.65. 
Mesosoma. (23) Light areas yellowish; medial interruption of pronotal 
stripe about 2 mow. (24) Scutal and scutellar longer hairs brown, others N. 
A. Hair of metanotum white. (25) Scutal disc with punctures very fine, 1 
pwa, interspaces shiny. (26) Dorsal enclosure of propodeum with medial 
portion dull with fine, low, interconnecting ridges, lateral portions shiny with 
a few prominent, shiny ridges. (27) Legs with light color the same as on face. 
Foreleg with a narrow yellow band on ventral apex of trochanter, yellow on 
dorsal apex of femur and basal fourth to third of tibia. (28) Middle leg col- 
ored like foreleg; spur finely, evenly pectinate, testaceous, its length about 0.5 


times basitarsal length, 0.49:0.97. (29) Hind leg brown except trochanter col- 


ored as in foreleg but band narrower. (30) Tegula transparent, testaceous, 
with small anterior patch of yellow. Humeral plate brown. (32) Marginal 
cell 6-9 longer than, and 3-4 slightly shorter than 9-wt, 1.05 :0.90:0.92. 


Metasoma. (35) Tergum 1 with punctures of median area about the same — 


size as on scutum, densely, regularly distributed, 1-2 pwa, interspaces shiny 
(30). Declivity of tergum 1 rather dull with numerous fine punctures. (37) 
Sterna dark brown. 

Tyre Marteriat. Holotype female, from Apatzingan (11 miles E.), 
Michoacan, Aug. 20, 1954 (E. G. Linsley, J. W. MacSwain, and R. F. Smith), 
is in the California Academy of Sciences, San Francisco. 


Subgenus CALLIOPSIMA, new subgenus 


Type species. Calliopsis rozent Shinn. 


This subgenus is composed of closely related species which are sharply 
separated from the other subgenera. No specimens which might represent 
annectent forms have been discovered. Several species of the South American 
panurgine genus Acamptopoeum have color markings almost identical with 
the males of many species of Calliopsima. The deep, distinct punctures with 
smooth, shiny interspaces found in Calliopsima are also typical of Acampto- 
poeum. Calliopsima scems to have a mixture of the characters found in 
Acamptopoeum and Calliopsis s.s., and is judged to be closest to Calliopsis 5.5. 

Four groups of species occur within this subgenus. They are: 1) the 
crypta group including rozeni, unca, azteca, and chlorops; 2) the pectidis 
eroup including timberlakei and bernardinensis; 3) the coloratipes group 10- 
cluding deserticola, pugionis, and possibly coloradensis; and 4) the Aurdi 
group including quadridentata and kucalumea. 

Calliopsima occurs only from Canada to near the border between Mexico 
and Guatemala. The locality for Acamptopoeum which is closest to this area 
is for A. colombiensis Shinn (Shinn, 1965) in northern Colombia. | The 


: 
i 


A ReEvIsIoN oF THE BEE GENUS $35 


record of A. maculatum (Smith) in Florida is almost certainly erroneous. | 
Calliopsima predominately visits flowers of the Compositae and Leguminosae. 

The males of Calliopsima differ from those of Calliopsis and Perissander 
by the relatively flat clypeus, the non-tumid paraocular area, and the different 
shapes of the posterior projections of sterna 5 and 8. Sternum 5 of Calliopsima 
has a relatively large median, posterior projection which is club-shaped or 
parallel-sided; sternum 8 has a long, slim club-shaped projection. The fe- 
males differ from those of Calliopsis and Perissander by the white to amber 
prepygidial and pygidial fimbriae (smoky in Awrdi group) and the light hairs 
of the discs of terga 4-5 (a few brown hairs occasionally). Both sexes differ 
from the above subgenera by having maxillary palpomere 2 longer than 3, 
and by having coarse, deep, pleural punctures with smooth, usually shiny 
interspaces. 


Femate. Length 6.5-10.0 mm. 

Head. Light colored areas: (1) paraocular area below a sinuous line 
originating between middle of outer subantennal suture and upper end of 
suture and extending diagonally upward to lower inner margin of facial 
fovea, thence laterally, tangent to fovea, ending on orbit well above level of 
upper border of antennal socket, usually about level of middle of facial fovea; 
very shiny; (2) clypeus variable from completely yellow except for two small 
clypeal dots, to yellow on lateral portions with a median longitudinal band 
extending from frontoclypeal suture a variable distance towards the clypeal 
apex; (3) labrum variable, all black to yellow; (4) supraclypeal area pen- 
tagonal to trianguliform with apex between middle of and a third mow above 
antennal socket; (5) subantennal plate variable, black to yellow; (6) mandible 
with basal portion variable. (7) Scape reddish brown to blackish brown dor- 
sally (except basal and apical yellow areas in timberlakez), pedicel and dorsal 
surface of flagellum brown to black extending part way onto ventral surface 
of flagellomeres 1-4, or latter all black, remaining flagellomeres tan ventrally. 
(8) Hair of vertex colorless or mixed light and dark, or all dark, of frons 
whitish, of clypeus fulvous. (10) Punctures along ocellocular line medium 
sized to very fine. Interspaces shiny; impunctate area, lateral to posterior 
ocellus shiny; punctures beside lower half of frontal line fine to large, inter- 
spaces smooth, dull to shiny. (11) Frontal line with lower portion a narrow, 
sulcus, sometimes interrupted, rising gradually to a low summit on frontal 
prominence between antennal sockets slightly above their midline. (12) 
Clypeus with punctures of disc more distinct than in Calliopsis s.s., clypeus 
relatively much flatter than in Calliopsis s.s. but somewhat protruding in 
deserticola. Projections beside apical emargination of clypeus smoothly 
rounded. (13) Inner orbits slightly convergent below. Facial fovea deep, 
upper end slightly below middle ocellus, lower end about 1 mow above level 
of upper border of antennal socket, somewhat tear-drop shaped or broadened 


836 Tue UNiversiry SciENCE BULLETIN 


medially, tapered above and below. (14) Galea of moderate length, propor- 
tionately longer than in Calliopsts s.s. (except hondurasica group). Galeal gap 
less than inner subantennal sutural. (15) Head width/head length 1.2-15. 
(17) Eye length subequal to or less than mio or flagellar length. (18) 
Interocellar subequal to ocellocular except greater in pectidis group; anten- 
nocular greater than interantennal (less than in deserticola). (19) Ocello- 
labral subequal to or greater than clypeal width except distinctly less in hurd. 
(21) Basal labial palpomere 2.1-3.3 times length of others combined. Mazil- 
lary palpomere 2 longer than 3. (22) Flagellomere 1 about twice length of 
flagellomere 2, subequal to flagellomere 9; flagellar length 2.0-2.3 tmes 
length of scape. 

Mesosoma. (23) Light colored areas: medial interruption of pronotal 
stripe 0.5-3 mow, variable within species; apex pronotal lobe, sometimes 
absent; scutellar crest variable. (24) Scutal and scutellar hairs of two kinds, 
longer ones fulvous to black, shorter ones fulvous. Hairs of scutellar crest and 


of posterolateral border of metanotum as in Calliopsis s.s. (25) Scutal disc 


with punctures deeper, mostly larger, more distinct than in Calliopsis s.s., 
interspaces smooth, either shiny or dull. (26) Dorsal enclosure of propodeum 
with longitudinal ridges, sometimes vermiform, posterior border at least cari- 
nate laterally, medial portion usually prolonged posteriorly, interspaces shiny. 
(27) Foreleg with at least basal spot of yellow on tibia, on apex of femur 
(absent in hurdi and kucalumea), and on base of basitarsus in pectidis group. 
(28) Middle leg colored like foreleg. Spur with extremely minute teeth or 
without evident teeth (20), covered with fine, abundant, short fulvous or 
white hair. (29) Hind leg usually brown except yellowish on apex of femur 
and base of tibia in pectidis, timberlakei, and some specimens of bernardinen- 
sis. (30) Tegula colorless to dark brown with light colored anterior patch or 
spot. Humeral plate brown or light colored apically. (31) Wing slightly 
smoky (to naked eye or at 30>) apically beyond cells. Stigma testaceous to 
brown. (32) Marginal cell 6-9 greater than (subequal to in pectidis), and 3-4 
much less than (subequal to or greater than in deserticola, bernardinensis, and 
kucalumea) 9-wt. 

Metasoma. (34) Tergal hair bands white, appressed, dense to sparse. 
Band of tergum 1 broadly interrupted, of tergum 2 less so (except both may 
be continuous in pectidis and bernardinensis). Suberect hair of discs of terga 
4.5 whitish to fulvous, a few brown hairs in some cases. Prepygidial and 
pygidial fimbriae white to fulvous (smoky in hurd: group). (35) Tergum 1 
with punctures of median area as in Calliopsts s.s. but interspaces shiny to 
highly polished (dull in Aurd7), puncture size variable with respect to punc- 
tures on scutum. Declivity of tergum 1 with a highly polished mirror-like 
surface, bearing few to no punctures. 


A Revision oF THE BEE GENUS 837 


Mate. Length, 5.2-8.0 mm. 

Head. Yellow areas: (1) paraocular area as in female, except upper 
border usually straight, mesal origin of dorsal boundary line usually higher 
than in corresponding female, yellow ending in more of a point on orbit; (2) 
clypeus with testaceous apical border; (3) labrum; (4) supraclypeal area as 
in female; (5) subantennal plate in some cases with small black area, black 
border, or all black (some Auwrd1); (6) mandible a basal spot to basal two- 
thirds; (7) scape entirely, through lesser amounts of yellow, to all brown; 
pedicel, sometimes; flagellum as in female except lighter throughout, flagel- 
lomeres 1-3 sometimes yellow. (8,10) As in female. (11) Frontal line with 
lower portion a slightly elevated, low, non-sulcate, somewhat rounded, never 
sharp, rather broad ridge, summit as in female. (12) As in female. (13) 
Inner orbits moderately to strongly convergent below. Facial fovea usually 
distinct. (14) As in female. (15) Head width/head length 1.30-1.48. (17) 
Eye length greater than (barely so in Aurdi and quadridentata) mio, and 
much less than flagellar length. (18) Interocellar subequal to ocellocular 
(greater in tamberlakei, bernardinensis, less in hurd1, kucalumea, quadriden- 
tata); antennocular less than interantennal; antennocellar subequal to outer 
subantennal sutural. (21) Basal labial palpomere 1.6-3.2 times length of others 
combined. Maxillary palpomere 2 longer than 3. (22) As in female except 
flagellar length 2.8-3.3 times length of scape. 

Mesosoma. (23) Yellow areas as in female, except more extensive, richer 
yellow. (24) As in female unless otherwise stated. (25) Scutal disc with 
punctures deeper, larger, more distinct than in Calliopsis s.s., more abundant 
than in female, interspaces smooth, dull or shiny. (26) Dorsal enclosure of 
propodeum similar to that of female except median portion more prolonged 
and ridges usually straighter, farther apart. (27) Legs with light color the 
same as on face. Foreleg with highly variable color pattern among species. 
(28) Middle leg colored like foreleg, but less extensive light color. (29) Hind 
leg usually colored like middle leg. (30) As in female (azteca, hurdi, and 
kucalumea without light color on tegula). (31) As in female. (32) Marginal 
cell 6-9 greater than, and 3-4 equal to or less than (more than, in Aurdi, 
kucalumea, quadridentata) 9-wt; 11-12/13-14 variable. 

Metasoma. (34) As in female. (35) Tergum 1 with punctures of median 
area as described for Calliopsis s.s., tergum sometimes dull; puncture size 
smaller than to larger than that of scutum. Declivity of tergum 1 usually 
smooth, dull, sometimes shiny, bearing several to many punctures (none in 
azteca) (30). Pygidial plate plane to convex, not abruptly sunken medio- 
apically as in Calliopsts s.s. (37) Sterna brown to testaceous. Sternum 5 pro- 
duced posteriorly into a blunt, rounded, often club-like projection. Sternum 
6 with a broadly concave, median bilobed portion, each lobe flattened from 
base to apex or apex bent abruptly ventrad to produce a pair of ventral 


838 Tue UNiversity ScIENCE BULLETIN 


“prongs” usually tilted slightly towards each other. Sternum 8 with a long, 
median projection drawn gradually or abruptly into a terminal club. (38) 
Sterna and genitalia as illustrated (Figs. 58-122). 


CALLIOPSIS (CALLIOPSIMA) ROZENI Shinn 
(Figs. 58-62; Map 5) 
Calliopsis rozeni Shinn, 1965, Amer. Mus. Novitates, 2211:2. 


This species is named for Dr. Jerome G. Rozen, Jr., who has lent numerous 

specimens for my study of the genus. The species is closest to C. coloradensis 
and C. kucalumea. Superficially, it bears a striking resemblance to C. erypta. 
The male of rozeni has the tips of the projections beside the apical portion of 
metasomal sternum 6 flat, whereas in coloradensis, kucalumea, and crypta 
they are bent ventrad. The female of rozenz is distinguished from crypta by 
the fine head punctures described in (10) below and by the possession of ful- 
vous long hairs on the scutum and scutellum, whereas in crypta the head 
punctures are coarse and the long hairs on the scutum and scutellum are 
brown. The female of coloradensis has fine, dense punctures, regularly spaced 
about 2 pwa on the disc of metasomal tergum 1, but those of kucalumea are 
regularly spaced less than 1 pwa, and those of rozeni are larger, sparse, irregu- 
larly spaced about 0.5-3 pwa. 


Femate. Length, 8.0 mm; forewing length, 5.5 mm; hindwing length, 3.8 
mm; clypeal length, 0.60 mm; scutal length, 1.43 mm. 

Head. Cream colored areas: (1) paraocular area below a sinuous line 
originating at middle of outer subantennal suture and extending to lower 
inner margin of facial fovea, thence bordering fovea ventrally, ending on 
orbit slightly below middle of fovea; (2) clypeus except for narrow brown 
apical border and two vertical bars of brown originating at dorsolateral cor- 
ners of clypeal emargination, extending dorsally about five-sevenths of median 
length of clypeus; (3) absent on labrum; (5) subantennal plate (to all black 
or with irregularly shaped cream colored area); (6)absent on mandible. (7) 
Flagellomeres 1-4 brown and tan ventrally. (8) Hair of vertex fulvous. (10) 
Punctures beside lower half of frontal line large, 1-2 pwa, interspaces shiny. 
(13) Orbital convergence ratio as 1.68:1.55,1.09. (14) Galea finely pebbled, 
galeal gap subequal to length of galea exposed beyond closed mandibles, 0.37: 
0.39. (15) Head width to head length as 2.64:1.82,1.45. (17) Eye length, mio, 
and flagellar length as 1.39:1.55:1.39. (18) Interocellar, ocellocular, antennocu- 
lar, and interantennal as 0.48:0.48:0.41:0.36. (19) Ocellolabral equal (to sub- 
equal) to clypeal width, 1.53:1.53. (20) Clypeocellar to outer subantennal 
sutural as 0.92:0.82,1.12. (21) Basal labial palpomere about 2.8 times length of 
others combined. (22) Flagellar length about 2.0 times length of scape, 1.38: 
0.68. 


4 


A ReEvIsion OF THE BEE GENUS $39 


Mesosoma. (23) Cream colored areas: a dot on right pronotal lobe (a 
dot on each, or none); scutellar crest. (24) Scutal and scutellar hair fulvous, 
long hair concolorous with short. (25) Scutal disc with punctures about the 
same size as on midvertex, 0.5 pwa or less, interspaces shiny (30). (26) 
Dorsal enclosure of propodeum somewhat declivous with sharp, carinate 
posterior border, enclosure with longitudinally vermiform ridges medially, 
straighter longitudinal ridges laterad. (27) Legs with light color the same as 
on face. Foreleg with cream color on apex of femur and knee of tibia and ad- 
jacent area subequal to it. (28) Spur with many short hairs, finely pectinate 
(to no distinct teeth); spur length less than half of length of middle basitarsus, 
0.44:0.94. (30) Tegula transparent testaceous. Humeral plate with cream- 
colored apex. (31) Stigma tan. (32) Marginal cell 6-9 longer than, and 3-4 
shorter than (to equal to) 9-wt, 1.17:0.99 :1.02. 

Metasoma. (34) Suberect hairs of discs of terga 4-5 white. (35) Tergum 
1 with punctures of median area smaller than on scutum, deep, irregularly 
spaced, sparse medially, dense laterally, less than 1 pwa. 


Mate. Length, 6.5 mm; forewing length, 4.7 mm; hind wing length, 3.2 
mm; clypeal length, 0.55 mm; scutal length, 1.27 mm. 

Head. Yellow areas: (1) paraocular area below diagonal line originating 
at upper rim of antennal socket to lower margin of facial fovea ending on 
orbit about third of eye length below summit of eye; (6) mandible basal half 
to two-thirds; (7) scape with broad lateroventral stripe; a mere dot on latero- 
ventral surface of flagellomere 1. (10) Punctures beside lower half of frontal 
line large, 1-2 pwa, interspaces shiny. (13) Orbital convergence ratio as 1.53: 
1.14,1.34. (14) As in female but galeal gap slightly more than half of length 
of galea exposed beyond closed mandibles (0.26:0.44). (15) Head width to 
head length as 2.35:1.70,1.38. (17)) Eye length, mio, and flagellar length as 
1.31:1.14:1.53. (18) Interocellar, ocellocular, antennocular, and interantennal 
as 0.43 :0.41:0.29:0,32. (19) Ocellolabral greater than clypeal width, 1:41:1.26, 
1.12. (20) Clypeocellar to outer subantennal sutural as 0.82:0.71,1.14. (21) 
Basal labial palpomere about 2.6 times length of others combined. (22) 
Flagellar length about 3.0 times length of scape, 1.55:0.51. 

Mesosoma. (24) Scutal and scutellar hair pale grey (to pale fulvous), 
otherwise as in female. (25) Scutal disc with punctures slightly larger than on 
midvertex, 0.5-1 pwa, interspaces shiny (30). (27) Foreleg with yellow on 
dorsoapical third of femur, anterior surface of tibia, basitarsus and second 
tarsomere entirely; third to fifth tarsomeres successively darker, testaceous to 
light brown; posterior surface of tibia brown. (28) Middle leg colored like 
foreleg but yellow on dorsoapical fifth of femur; lengths of tibia, basitarsus, 
and apicotarsus as 1.09:0.87:1.00. (29) Hind leg colored like middle leg, nar- 


row brown margin may occur on anterior surface of basitarsus, apicotarsus 


840 Tue University ScIENCE BULLETIN 


brown. (32) Marginal cell 6-9 greater than, and 3-4 less than 9-wt, 0.99:0.83: 
0.92. 

Metasoma. (34) Suberect hairs of discs of terga 45 fulvous. (35) Tergum 
1 with punctures of median area larger than in female, smaller than on scu- 
tum, deep, fairly regularly spaced, less than 1 pwa, interspaces shiny. 

Tyre Materia. Holotype male, from Rodeo, Hidalgo County, New Mex- 
ico, August 22, 1962 (J. G. Rozen, M. Statham, S. J. Hessel) on Heterotheca 
subaxillaris, and allotype female, from Portal (5 mi. W.), S. W. R. S,, 
Cochise Co., Arizona, 5400 ft., Aug. 10, 1956 (Ellen Ordway), on Melilotus 
alba, are at the American Museum of Natural History. 

In addition, 154 male and 120 female paratypes are from the following localities: ARIZONA: 
Apache and vicinity; Skeleton Canyon, Benson, Bisbee, Portal and vicinity, all in Cochise Co.; 
Flagstaff; Mt. View; Phoenix; Santa Rita Mountains; San Xavier Mission; Superior, Pinal Co.; 
Tucson; Warren; Yuma: Yuma Test Station. New Mexico: Granite Pass, Peloncillo Mts., Hidalgo 
Co.; High Rolls, Otero Co.; Lordsburg; Filmore Canyon and Soledad, Organ Mountains; Road 
Forks, Hidalgo Co.; Rodeo and vicinity, Hidalgo Co.; Roswell; Socorro. Texas: Coopers Store, 
Big Bend National Park, Brewster Co.; Davis Mts., Jeff Davis Co.; Marathon (20 mi. S.); Stone- 
wall: Uvalde. Cotnuanua: Salaices. Coanuita: La Gloria, south of Monclova, 3300 ft.; Paila, 3900 
ft.; Piedras Negras (192 km. S.), 1300 ft. 

Distriution. The southwestern United States and north central Mexico. 
In addition to the type material, 4 females have been examined from 5 miles 
west of Junction, Kimble Co., Texas, April 15, 1961 (Rozen and Schrammel). 

This species is active in April, May, June, July, August, September, and 
October. 

Gerocrapuic Variation. Three characters are conspicuously variable in fe- 
male rozeni: 1) the amount of yellow color on the pronotal lobe, 2) the 
amount of yellow on the subantennal plate, and 3) the density of the pune- 
tures of the median area of metasomal tergum 1. The first two do not appear 
to be clinal, but the third may be so. Specimens from Texas exhibit denser 
punctation than more western ones, and females from Junction, Texas, have 
been excluded from the type series primarily for this reason since no males 
were collected to confirm the determination. 

Discussion. It is unusual for two solitary bees of different species to be 
taken in copulo, and this has almost always been interpreted to mean that the 
two specimens are not, indeed, different species. One male of rozeni was col- 
lected in copulo with a female of crypta by J. G. Rozen (Portal, S. W. R. S., 
Arizona, September 14, 1962, on Heterotheca subaxillaris), and one pair of 
rozeni was taken in copulo by G. I. Stage, Rodeo (2.5 miles north), New 
Mexico, September 7, 1959, on Baileya pleniradiata. Both collectors had ob- 
served the specimens carefully prior to collection. Inasmuch as I have an in- 
terspecific mating pair of Calliopsis chlorops with Calliopsis coloradensis, | 
believe the phenomenon may be more common than preserved specimens in- 
dicate. Nevertheless, I have not yet seen a specimen of either sex of any of 
these species which can be considered an intermediate, or hybrid, form. It is 


A REVISION OF THE BEE GENUS 841 


} 


\ \ 2 
> 
51S ae mG an < 
\ J 
y 
‘ »\ 
, pie 
‘ 


pugionistay 
/ 


30K 
deserticola 


“@ quadridentata rir 
"5 ae = 


azteca # en 
100 90 


Map 5. Map showing the known distributions of Calliopsis (Calliopstma) coloradensis Cresson, 
C. (C.) coloratipes Cockerell, C. (C.) pugionis Cockerell, C. (C.) deserticola Shinn, C. (C.) 
rozem Shinn, C. (C.) quadridentata Shinn, C. (C.) azteca Shinn, and C. (C.) hurdi Shinn. 
The presumptive collection locality for Calliopsis flavifrons Smith is also shown. 


possible that the males can be deceived by some of the females of these closely 
related and superficially indistinguishable pairs of species. I conclude that no 
viable offspring are produced. 

Firower Recorps. Baccharis, Baileya pleniradiata, Chamaesaracha con- 
loides, Eriocarpum gracile, Eriogonum, Gaillardia, Helianthus, Heterotheca 
subaxillaris, Hymenoxys odorata, Melilotus alba, Parkinsonia, Pectis papposa, 
Psilostrophe coopert, Sphaeralcea emoryi?, Verbesina exauria. 'Vaken pri- 
marily on Heterotheca subaxillaris which is a favorite flower among its rela- 
tives also. 


CALLIOPSIS (CALLIOPSIMA) COLORADENSIS Cresson 
(Figs. 63-66; Map 5) 


Calhiopsis coloradensis Cresson, 1878, Trans. Amer. Ent. Soc., 7:63, female, male; Patton, 1879, 
Bull. U.S. Geol. Survey, 5:366; Cockerell, 1897, Canad. Ent., 29:290; 1897, Bull. Univ. New 
Mexico, 24:19; 1897, Proc. Acad. Nat. Sci. Phila., 49:350; 1898, Trans. Amer. Ent. Soc., 
25:196, 1898, Bull. Sci. Labs. Denison Univ., 11:52; 1898, Zoologist, (4) 2:313; 1899, Ent. 
News, 10:4; Bridwell, 1899, Trans. Kans. Acad. Sci., 16:210; Cockerell, 1901, Ann. Mag. 
Nat. Hist., (7) 7:128; Cockerell and Atkins, 1902, Ann. Mag. Nat. Hist., (7) 10:44; Cock- 


842 Tue UNIversiry SCIENCE BULLETIN 


erell, 1906, Trans. Amer. Ent. Soc., 32:299; 1906, Bull. Amer. Mus. Nat. Hist., 22:440; Swenk 

and Cockerell, 1907, Ent. News, 18:178; 1908, Canad. Ent., 40:147-148; Crawford, 1912, 

Canad. Ent., 49:359; Robertson, 1914, Ent. News, 25:70; Cresson, 1916, Mem. Amer. Ent. 

Soc., 1:115; Stevens, 1919, Canad. Ent., 51:210; Cockerell, 1919, Jour. N.Y. Ent. Soc., 27:299; 

1921, Amer. Mus. Nov., 24:13; Robertson, 1922, Psyche 29(4):169; Robertson, 1926, 

Ecology, 7:379 (anthecology) ; Hicks, 1926, Univ. Colo. Studies, 15:223; Robertson, 1928, 

Flowers and Insects, p. 10+ (anthecology); 1929, Psyche, 36(2):115 (anthecology); Stevens, 

1950, North Dakota Agric. Exp. Sta. Bimon. Bull., 12:90,93 (biol.); Rozen, 1951, Jour. Kans. 

Ent. Soc., 24(4):142+(male genitalia); Linsley, 1958, Hilgardia, 27:561; Mitchell, 1960, 

North Carolina Agric. Exp. Sta. Tech. Bull. No. 141:288-289, 291-294. 
coloradensis coloratipes; Pierce, 1904, Stud. Univ. Nebraska, 4:23. (misidentification ) 
coloratipes fedorensis; Cockerell, 1909, Ann. Mag. Nat. Hist. (8) 4:28. (misidentification) 
coloradensis fedorensis; Cockerell, 1921, Amer. Mus. Novitates, 24:14. 

C. sp.?, Michener, 1947, Amer. Midl. Nat., 38:447. 

The species is closest to rozeni and coloratipes. It is separated in the male 
from them by the long ventral prongs on sternum 6, by having the volsella 
much expanded, and by having the punctures of tergum 1 much smaller than 
in rozeni and larger than in coloratipes. It is separated in the female from 
coloratipes by the dark mandibular base, from rozeni by the much finer 
punctures of tergum 1, and from both species by the regularly distributed 


punctures of tergum 1, which are sparse medially in them. 


Femate. Length, 8.8 mm; forewing length, 5.8 mm, hindwing length, 4.1 
mm; clypeal length, 0.62 mm; scutal length, 1.60 mm. 

Head. Cream colored areas: (1) as in rozeni; (2) clypeus with a broad T 
with convex crossbar bordering entire length of horizontal portion of fronto- 
clypeal suture (to somewhat less), bottom of the T with adjacent small spot 
(more typically, clypeus light colored except for testaceous apical border and 
wide, twin, vertical bars of brown arising from its median portion); (3) 
absent on labrum; (5) subantennal plate with dot (to entirely); (6) absent 
on mandible. (7) As in rozeni. (8) Hair of vertex fulvous. (10) Punctures 
beside lower half of frontal line slightly smaller than in rozenz, 1-2 pwa, inter- 
spaces shiny. (13) Orbital convergence ratio as 1.82:1.70,1.07. (14) Galea 
finely pebbled, dull, galeal gap about half length of galea exposed beyond 
closed mandibles, (0.31:0.58). (15) Head width to head length as 2.77:1:89, 
1.47. (17) Eye length, mio, and flagellar length as 1.46:1.70:1.48. (18) Intero- 
cellar, ocellocular, antennocular, and interantennal as 0.51 :0.51:0.48 0.43. (19) 
Ocellolabral slightly greater than clypeal width, 1.62:1.56,1.03. (20) Clypeo- 
cellar to outer subantennal sutural as 0.99:0.87,1.14. (21) Basal labial palpo- 
mere about 2.5 times length of others combined. (22) Flagellar length about 
2.1 times length of scape, 1.48:0.71. 

Mesosoma. (23) Yellowish areas: apex of pronotal lobe, scutellar crest. 
(24) As in rozeni, except hairs longer. (25) Scutal disc with punctures 
slightly larger than those of vertex, 1 pwa or less. (26) As in rozeni except 
ridges narrower, more abundant, shinier. (27) Foreleg with yellow at apex of 
femur and knee of tibia. (28) Spur exceedingly finely pectinate on apical half, 
bearing about 10 short teeth, barely distinguishable (30X); spur length about 


A REVISION OF THE BEE GENUS 843 


half of length of middle basitarsus, 0.48:0.94. (30) Tegula transparent, light 
amber. Humeral plate with anterior half yellow. (31) Wing clear (to faintly 
smoky apically, not so smoky as in rozen/). Stigma tan. (32) Marginal cell 
6-9 greater than, and 3-4 less than 9-wt, 1.33:0.90:1.14. 

Metasoma. (34) As in rozeni but hairs longer. (35) Tergum 1 with pune- 
tures of median area slightly finer than on scutum, fairly regularly spaced, 
2-3 pwa. 


Mate. Length, 6.3 mm; forewing length, 4.9 mm; hindwing length, 3.52 
mm; clypeal length, 0.58 mm; scutal length, 1.41 mm. 

Head. Yellow areas: (1) paraocular area below diagonal line originating 
at upper end (or between upper end and middle) of outer subantennal suture 
and extending tangent to facial fovea ending on orbit below midlevel of fovea, 
angle of upper corner about 40° (or 10°); (4) supraclypeal area in some 
cases with black border between frontoclypeal suture and yellow supraclypeal 
area; (5) subantennal plate, sometimes reduced to basal spot; (6) mandible 
basal half; (7) scape with ventral surface yellow except dark brown mesally 
and apically (through various states of reduction of yellow to all brown 
scape); yellowish ventral dots on flagellomeres 1-2 (sometimes absent). 
(10) As in female. (13) Orbital convergence ratio as 1.60:1.22,1.30. (14) As 
in female. (15) Head width to head length as 2.41:1:70,1.42. (17) Eye length, 
mio, and flagellar length as 1.33:1.22:1.51. (18) Interocellar, ocellocular, anten- 
nocular, and interantennal as 0.46:0.46:0.31:0.36. (19) Ocellolabral greater 
than clypeal width, 1.43:1.26,1.13. (20) Clypeocellar to outer subantennal 
sutural as 0.85 :0.68,1.25. (21) Basal labial palpomere about 2.0 times length of 
others combined. (22) Flagellar length about 2.9 times length of scape, 1.51: 
053. 

Mesosoma. (25) Scutal disc with punctures slightly larger than in rozenz, 
0.5-1 pwa, interspaces shiny. (27) Foreleg with yellow on dorsoapical half (or 
less) of femur, tibia, basitarsus and second tarsomere, remaining tarsomeres 
tinged testaceous. (28) Middle leg colored like foreleg but less yellow on 
femur, brown patch on posterior surface of tibia (sometimes); lengths of 
tibia, basitarsus, and apicotarsus 1.04:0.94:0.99. (29) Hind leg colored like 
middle leg. (32) Marginal cell 6-9 greater than, and 3-4 less than 9-wt, 1.12: 
0.87 :0.99, 

Metasoma. (35) Tergum 1 with punctures of median area slightly smaller 
than on scutum, deep, crowded medially, contiguous, interspaces shiny. 

Type Marertay. Lectotype female, 1 female and 2 male paratypes, Colo- 
rado (Ridings and Morrison), with no further label data, are at the Academy 
of Natural Sciences of Philadelphia, Types 2187, 2187.2, 2187.4, and 2187.5, 
respectively. Paratype 2187.3, same data, is a female specimen of C. chlorops. 
I suggest that the indefinite type locality be restricted to the Colorado coun- 
ties of Denver, northern half of Jefferson, and eastern half of Boulder, where 


844 Tue University SciIENCE BULLETIN 


specimens taken agree exactly with the types. The description of the male is 
principally based on paratype No. 2187.2. 

Distripution. Alberta to southern Utah and southern New Mexico, east 
to the Mississippi River, thence through eastern Texas and the Gulf States to 
the east slope of the Appalachian Mountains in North Carolina. Not yet 
found in Florida, although it has likely occurred there in the past (cf. remarks 
regarding Calliopsis flavifrons Smith under the following section, Geographic 
variation). It is a late summer-early autumn bee collected between June 27, 
1918, at Winnfield, Louisiana, and Oct. 10, 1961, at Nacogdoches, Texas, but 
most records are in July and August. 

In addition to the type specimens, about 490 others have been studied from the following 
localities; ALBERTA: Lethbridge; Medicine Hat; Scandia; Tilley. Manirosa: Balmoral; Stony 
Mountain; Stormy Mountain; Winnipeg. AcABaAMa: Pickett Springs, Montgomery Co, ARKANSAS: 
Chessman Ferry, Stone Co.; Fayetteville, Washington Co.; Polk Co. Cotorapo: Boulder; Denver; 
Hoehne; Limon; Loveland; Roggen; Wray. Grorcia: Augusta, Richmond Co.; Cartersville, Bar- 
tow Co.; Nacoochee Valley. IpAHo: Downey; Fort Hall. Ivvrors: Carlinville. lowa: Sioux City. 
Kansas: Blue Rapids; Dodge City (8 mi. N.E.); Douglas Co.; Garnett; Hutchinson; Lakin (4 mi. 
E.); Riley Co.; Scott Co., 2970 ft.; Sherman Co., 3690 ft.; Smith Co., 1800 ft.; Stafford Co.; Yates 
Center. Lourstana: Winnfield, Winn Co. Minnesora: Detroit; Moorhead; Powder Plant Woods, 
Ramsey Co.; University Farms, Ramsey Co, Mississippi: Hattiesburg; Utica. Missouri: Conway 
(10 mi. N.); Gilmore; High Hill; Lebanon (12 mi. E.). Montana: Bozeman; Hill Co.; Mis- 
soula; Pompey’s Pillar, Yellowstone Co.; Pondera Co. Nesraska: Crofton (2 mi. W.; 7 mi. 
N.W.): Gordan, Sheridon Co.; Harrison (13 mi. N.); Lincoln; West Point. NEw Mexico: Organ 
Mountains, Filmore Canyon, Dona Ana Co., 5400 ft.; Roswell (5 mi. E.), Chaves Co. NortH 
Carotina: Aberdeen, Moore Co. NortH Daxora: Beach, Bismarck; Carpio; Crary; Devils Lake; 
Edgeley; Fargo; Glen Ullin (10 mi. E); Grand Forks; Granville; Hatton; Jamestown; Lakota; 
McKenzie; Mandan; Minot; Mott; Perth; Sentinel Butte; Valley City; Williston. SourH Dakota: 
Geddes (5 mi. E.); Fort Thompson. Texas: Brazos Co.; Fedor; Lee Co.; Nacogdoches. Uran: 
Farr West; Garfield; Knaub; Magna; Petersboro; Price; Topaz; Williard. Wyominc: Waltman, 
Natrona Co.; Yellowstone National Park. 

Grocrapuic Variation. This species is more variable than any other spe- 
cies of Calliopsima. The subgenus as a whole consists of rather more closely 
related species than does either Calliopsis s.s. or Perissander. It is about the 
same in this respect as Verbenapis whose species are possibly even more closely 
related than those in Calliopsima. 

The most obvious variation is in size. The size grades from the large 
robust specimens of Canada and the Rocky Mountain States to the small speci- 
mens from the Missouri and Mississippi River valleys from Missouri to 
Mississippi. 

Seven characteristics were investigated for use as total size indicators be- 
cause of the high variability of total length. These characteristics were meas- 
ured for 33 male specimens from all parts of the range of the species and the 
correlations were calculated among them with the results shown in Table 2. 

These results suggest that scutal length, clypeal length, and flagellar 
length are most highly correlated with all the other variables, and that eye 
length has the smallest correlation with all the others. 


Based upon these data and similar ones for C. andreniformis, wing length, 


A Revision oF THE BEE GENUS 845 


Taste 2. Correlation Coefficients of Characteristics for Expressing Total Size of a 


Bee Based on Calliopsis coloradensis 


Hindwing Scutal Head Clypeo- — Clypeal Eye Flagellar 
Length Length Length cellar Length Length Length 

Hindwing 
Lene eae 1.000 884 817 .870 900 639 864 
Scutal 
L@metdi) Seana 1.000 883 .888 948 693 935 
Head 
(LO@WyS(” pee 1.000 HIS) .888 637 830 
Clypeo- 
adler 22 eer 1.000 863 657 .838 
Clypeal 
[L (STO Ue re 1.000 653 931 
Eye 
Mem oth fees 2c----. 1.000 .639 
Flagellar 
\L@S(U 0) eee 1.000 


Multiple Correlation 
of each variable with 


allithe others) .........- 292 197), 90 Soul 297 70 295 


scutal length, and clypeal length were chosen as suitable measures of the total 
size of a bee. These are the introductory measurements given for each species 
treated in this genus. 

A summary of the salient differences between male topotypical specimens 
of coloradensts and those of the Missouri and Mississippi River basins is given 
in Table 2. The measurements are for the male allotype from Colorado, and 
for a Missouri River specimen from Gilmore, Missouri. The latter is repre- 
sentative of the specimens from the two river systems and from the south- 
eastern United States. I do not feel that there is an adequate basis to consider 
these latter specimens as a species separate from coloradensis. 1 would like to 
see more specimens and comparative ecological data before reaching a deci- 
sion on their status. If they eventually are shown to be a distinct species there 
is every reason to use the available name Calliopsis flavifrons Smith, 1853. 
Smith’s description of the flavifrons collected in East Florida would fit only 
this form of Calliopsis which is the only representative of its subgenus in the 
southeastern United States. Specimens from localities between Colorado and 
the Mississippi are somewhat variable with respect to the characters given 
below. However, no clear clinal changes have been uncovered although the 
average size of males increases steadily from Missouri to Colorado and from 
North Carolina to Texas. The typical southeastern male is only slightly more 
than three-fourths the length of the typical Colorado male. 


846 Tue University SCIENCE BULLETIN 


Taste 3. Comparison of Rocky Mountain forms of C. coloradensis with forms 


from the southeastern United States 


Character Colorado Missouri 

Felines yy7i' 0 ora] etn thn eee 1.17 0.97 
Hlead/slen oth gs eet ee Be ee ee renee 1.70 1.36 
Inner subantennal sutural/galeal gap ................----. Sil =] 
Galealilengthi/clypeocellan es. nee Sil =] 
Basal labial palpomere length/clypeocellar ........... == <l 
Byenlengthyscutalaleng theses == eens sens <1 (1.33:1.41) =] (1.02:1.04) 
Eye profile width/antennocellar __.................--.--. S1 (0.71:0.73) >1 (0.61:0.53) 
Eye profile width/outer subantennal sutural —_.. >1 (0.73 :0.68) <1 (0.53:0.56) 
Midocellar interocular/flagellar length —.............. Sil Clase 1) <al (EZ 631254) 
Fore basitarsal length/clypeocellar .....................-.--- <1 (0.85:0.94) <1 (0.68:0.71) 
Dorsal propodeal enclosure, hind border Rolled anteriad, or 

Ofsmed taney porto nips eee eee arene ee Weakly carinate obscurely carinate 
Buncturesyo tarknom sues eae eee Finer Coarser 
Hindiatibiasmposteniorsuibace ses se seee eee Mostly yellow Mostly brown 
Outline of vertex) in) facial aspect 22-2 se Flat Arched 
Penis valve length/volsella length -....................-.- Greater Lesser 
Apodemies of penisinval vies ssssesmee er Shorter Longer 


Frower Recorps. Anthems cotuia, Aplopappus pluriflorus, Aster dumo- 
sus, A. praetus, Bidens laevus, B. aristosa, Boltonia asteroides, Chrysopsis, 
Chrysothamnus nauseosus, Coreopsis tripteris, Eriocarpum gracile, Grindelia 
perennis, G. squarrosa, Helenium nudiflorum, Heterotheca subaxillaris, Ma- 
crotera, Rudbeckia triloba, Silphium, Solidago rigida, S. serotina. 

Robertson (1922) studied the anthecology of this species at Carlinville, — 
Illinois. His results are given below: 


FEMALES COLLECTING POLLEN MALES SUCKING NECTAR MALE-FEMALE IN COPULO 
Astereae: 


Boltontia asteroides Boltonia asteroides Boltonia asteroides 

Solidago canadensis Solidago canadensis Solidago canadensis 
Heliantheae: 

Bidens artstosa Bidens artstosa Bidens artstosa 

Rudbeckia triloba Rudbeckia triloba 


Coreopsis tripteris 


CALLIOPSIS (CALLIOPSIMA) PECTIDIS Shinn 
(Figs. 67-71; Map 6) 
Calliopsis pectidis Shinn, 1965, Amer. Mus. Novitates, 2211:10. 


This species is named for one of the genera of plants, Pectis, used as its 
food source, and is adopted from the manuscript name proposed for it by 


A REVISION OF THE BEE GENUS 847 


Prof. P. H. Timberlake who recognized it as a new form more than a decade 
ago. The closest relatives are C. timberlakei and C. bernardinensis. All three 
species comprise a closely related group which has close afhinities with rozen7. 

The male is separated from other species of Calliopsima by the yellow 
scape with light brown apical rim and by the shape of the ventral protuber- 
ance of the apical portion of the penis valve (Fig. 71). The female is distin- 
guished by the presence of large amounts of cream coloration on the tibiae 
and basitarsi as described in (27), (28), and (29) below, and by the charac- 
teristic nap-like vestiture of the scutum as described in (24) below. 


Femate. Length, 7.7 mm; forewing length, 5.0 mm; hindwing length, 3.4 
mm; clypeal length, 0.56 mm; scutal length, 1.49 mm. 

Head. Cream colored areas: (1) paraocular area as described for sub- 
genus, sinuous line originating at a point about two-thirds up outer subanten- 
nal suture and ending on orbit at about level of middle of fovea; (2) as in 
rozent except brown clypeal bars narrower, extending dorsally about five- 
eighths (or less) of median length of clypeus; (3) labrum except for testa- 
ceous rim of labral plate and median area apical to it; (5) subantennal plate; 
(6) mandible about basal fourth; (7) scape with small apicoventral dot, base 
testaceous ventrally; pedicel and flagellum as in rozeni except lighter brown. 
(8) Hair color as in rozent but all areas with shorter, more plumose hair. (10) 
Punctures beside lower half of frontal line larger than in rozent, 1-2 pwa, in- 
terspaces shiny. (13) Orbital convergence ratio as 1.60:1.51,1.05. (14) Galea 
shinier, more finely pebbled than in rozenz. (15) Head width to head length 
as 2.58:1.79,1.45. (17) Eye length, mio, and flagellar length as 1.36:1.51:1.45. 
(18) Interocellar, ocellocular, antennocular, and interantennal as 0.53:0.43: 
0.41:0.37. (19) Ocellolabral greater than (to subequal to) clypeal width, 
1.48 :1.45,1.02. (20) Clypeocellar to outer subantennal sutural as 0.90:0.75,1.20. 
(21) N. A., measured on paratype of equivalent size: basal labial palpomere 
about 2.3 times length of others combined. (22) Flagellar length 2.1 times 
length of scape, 1.45:0.68. 

Mesosoma. (23) Cream colored areas: apex of pronotal lobe, scutellar 
crest. (24) Hair about half length of that of rozen1, more plumose. Scutal and 
scutellar hair color as in crypta except long dark hairs of lighter brown hue; 
short hairs dense, profusely branched concealing scutal punctures, appearing 
like a closely applied nap at 10>, highly distinctive, unique in its group. (25) 
Scutal disc with punctures contiguous, smaller than in rozeni. (26) As in 
rozeni but enclosure shorter. (27) Foreleg with cream color at apex of femur, 
most of anterior and posterior surfaces of tibia dorsally, and basal half of an- 
terior surface of basitarsus. (28) Spur length half of length of middle basi- 
tarsus, 0.41:0.82. (29) Hind leg colored like foreleg but anterior surface of 
basitarsus entirely light with brown rim. (30) Tegula hyaline laterally, straw 


848 Tue UNiversiry ScrENcE BULLETIN 


color posteriorly. Humeral plate as in rozent. (31) As in rozent. (32) Mar- 
ginal cell 6-9 less than (to subequal to), and 3-4 shorter than 9-wt, 0.92:0.82: 
0.95: 

Metasoma. (34) Tergal hair bands denser than in rozenz, more plumose, 
hence much more distinct. Suberect hairs of discs of terga 45 brownish 
(almost always whitish). (35) As in rozeni but much smaller, regularly 
spaced, less than 1 pwa. 

Mate. Length, 7.2 mm; forewing length, 4.3 mm; hindwing length, 2.9 
mm; clypeal length, 0.56 mm; scutal length, 1.37 mm. 

Head. Yellow areas: (1) paraocular area below a dorsally convex line 
originating at upper rim of antennal socket and extending to lower inner 
margin of facial fovea ending on orbit at about level of lower margin of 
fovea; (6) as in rozeni; (7) scape entirely (or with a longitudinal streak of 
brown on dorsal surface); pedicel and flagellomere 1 on ventrolateral sur- 
faces. (10) As in rozeni. (13) Orbital convergence ratio as 1.46:1.14,1.28. 
(14) Asin female. (15) Head width to head length as in rozenz, 2.35:1.70,1.38. 
(17) Eye length, mio, and flagellar length as 1.22:1.14:1.62. (18) Interocellar, 
ocellocular, antennocular, and interantennal as 0.41 :0.43:0.28 :0.32. (19) Ocel- 
lolabral greater than clypeal width, 1.38:1.14,1.21. (20) Clypeocellar to outer 
subantennal sutural as 0.80:0.68,1.18. (21) N. A., taken on paratype of equiv- 
alent size: basal labial palpomere about 1.6 times length of others combined. 
(22) Flagellar length about 3.3 times length of scape, 1.62:0.49. 

Mesosoma. (25) As in rozeni but smaller, less than half pwa, interspaces 
shiny. (27) Foreleg with coxa yellow on apical half of ventral surface, 
trochanter with yellow patch apicoventrally, femur with yellow on apical 
three-fourths of anterodorsal surface and apical third of posterior surface, 
tibia and tarsomeres 1-4 yellow, tarsomere 5 testaceous. (28) Middle leg with 
coxal yellow reduced to apical fourth of ventral surface, trochanter with 
apicoventral yellow patch smaller than that of foreleg, femur and tibia with 
yellow pattern similar to that of foreleg, tarsomeres 1-3 yellow, tarsomeres 4-5 
successively darker testaceous; lengths of tibia, basitarsus, and apicotarsus as 
0.95 :0.90:0.95. (29) Hind leg with yellow color pattern of trochanter, femur, 
and tibia similar to those of middle leg; coxa with yellow as on front 
trochanter; tarsomeres 1-2 yellow, tarsomeres 3-5 successively darker testa- 
ceous. (32) Marginal cell 6-9 equal to (to subequal to), and 3-4 shorter than 
9-wt, 0.87 :0.78 :0.87. 

Metasoma. (34) As in female but very few erect, brownish hairs. (35) 
Tergum 1| with punctures of median area smaller than on scutum or in rozent, 
regularly spaced, contiguous, interspaces shiny. 

Type Materia. Holotype male and allotype female, from Portal (2 miles 
N.E.), Cochise Co., Arizona, August 21, 1962 (J. G. Rozen, M. Statham, S. J. 


Hessel), are at the American Museum of Natural History. 


A REVISION OF THE BEE GENUS 849 


In addition 65 males and 47 female paratypes are from the following localities: Arizona: Ajo; 
Aguila; Brenda (2 mi. W.), Yuma Co.; Congress (4 mi. S.W.); Florence Junction (3.1 mi. S.), 
Pinal Co.; Gila Bend and 28 mi. E., Maricopa Co.; Portal and vicinity, Cochise Co.; Salome (2.3 
mi. N.). Cavirornia: Blythe (10 mi. N.); Julian (12 mi. E.); San Diego Co.; Twentynine 
Palms. New Mexico: Carrizozo, Lincoln Co. ; Lordsburg (11 mi. N.W.); Road Forks; Rodeo. 
Baya CaLiForNIA Sur: San Pedro; Sierra de la Laguna Mountains, Big Canyon (about latitude 
23° 34’ N., longitude 110° 00’ W.). 


| 


bernardinensis @px 


i) 


dis 


Map 6. Map showing the known distributions of Calliopsis (Calliopsima) chlorops Cockerell, C. 
(C.) bernardinensis Michener, C. (C.) pectidis Shinn, and C. (C.) crypta Shinn. 


850 Tue UNIversITY SCIENCE BULLETIN 


Discussion AND DistrisuTion. Dr. Paul D. Hurd, Jr., has informed me (in — 
litt.) that two female Holcopasites arizonicus (Linsley) were taken 1n associa- 
tion with a female C. pectidis by Dr. Mont A. Cazier, 2 mi. N.E. of Portal, 
Arizona, Sept. 24, 1961. This is the second species record of a Calliopsis- 
Holcopasites association, the other being of H. calliopsidis (Linsley) with C. 
andreniformis Smith. C. pectidis is apparently widespread from southern 
New Mexico-to the Mohave and Colorado Deserts in California and south to 
near the tip of Baja California. I anticipate its discovery in the Sonoran Desert 
of northern Mexico. The species is active from August to October. 

Fiower Recorps. Aplopappus, Baccharis, Baileya pleniradiata, Helianthus, 
Heterotheca subaxillaris, Hymenothrix wislizeni, Melilotus alba, Pectis an- 
gustifolia, P. papposa, Tidestromia lanuginosa, Verbesina encelioides. Seven | 
of the same genera and four of the same species are visited by pectidis and 
rozeni. 


CALLIOPSIS (CALLIOPSIMA) TIMBERLAKE, new species 
(Figs. 72-76; Map 7) 


The specific name is given to honor Mr. P. H. Timberlake, Associate 
Entomologist, Emeritus, of the University of California, Riverside, who lent 
me his entire collection of Calliopsis including several new, unpublished spe- 
cies he had earlier recognized. 

Calliopsis timberlakei is closest to pectidis. It is separated in the male by 
the sparser, longer, scutal hair which does not conceal the scutal punctation, 
and by the partially yellow pedicel. The female is distinguished by the en- 
tirely yellow clypeus which lacks brown clypeal bars. 


Femate. Length, 7.7 mm; forewing length, 4.8 mm; hindwing length, 3.47 
mm; clypeal length, 0.58 mm; scutal length, 1.33 mm. 

Head. Yellow areas: (1) paraocular area as described for subgenus, sinu- 
ous line originating near upper end of subantennal suture; (2) clypeus except 
for brown markings near anterior tentorial pit, black subtriangular area with 
anterior tentorial pit as center, and testaceous apical border; (3) labrum ex- 
cept brown spot medioapically beyond labral plate (sometimes along apical 
border of labral plate); (5) subantennal plate; (6) mandible basal half; (7) 
scape a tiny area on extreme dorsolateral surface and at base, flagellomeres 1-4 
tan ventrally. (8) Hair of vertex fulvous. (10) Punctures beside lower half 
of frontal line slightly smaller than in coloradensis, 1-4 pwa, interspaces shiny. | 
(13) Orbital convergence ratio as 1.56:1.48,1.06. (14) Galea finely pebbled, 
shiny basally, dull apically. (15) Head width to head length as 2.47:1.72,1.44. 
(17) Eye length, mio, and flagellar length as 1.34:1.48:1.38. (18) Interocellar, 
ocellocular, antennocular, and interantennal as 0.49:0.41:0.39:0.36. (19) Ocel- | 
lolabral greater than clypeal width, 1.48:1.41,1.05. (20) Clypeocellar to outer 


A REvIsION OF THE BEE GENUS $51 


subantennal sutural as 0.90:0.77,1.18. (21) Basal labial palpomere about 2.5 
times length of others combined. (22) Flagellar length about 2.2 times length 
of scape, 1.36:0.61. 

Mesosoma. (23) Yellowish areas: apex pronotal lobe; scutellar crest. (24) 
Color as in rozent, length of shorter hairs intermediate between rozeni and 
coloradensis. (25) Scutal disc with punctures finer than in rozeni and ber- 
nardinensts, 0.5-1 pwa. (26) Dorsal enclosure of propodeum as in rozeni 
except ridges more regularly spaced, slightly farther apart. (28) Foreleg with 
yellow at dorsal apex of femur, dorsal surface of tibia, and base (to all) of 
basitarsus. (28) Middle leg colored like foreleg but tibial and basitarsal yellow 
less extensive, not reaching apices of segments; spur length half of length of 
middle basitarsus, 0.43:0.85. (29) Hind leg with yellow patch on femur dorso- 
subapically and on tibia anterobasally. (30) Tegula testaceous. Humeral 
plate yellowish apically. (31) Wing not noticeably smoky to naked eye. 
Stigma testaceous. (32) Marginal cell 6-9 greater than, and 3-4 less than 9-wt, 
0.66 :0.77 :0.94. 

Metasoma. (34) As in rozenz but hair longer and denser. (35) Tergum 1 
with punctures of median area smaller than on scutum, deep, fairly regularly 


spaced, 1-2 pwa. 


Mate. Length, 5.5 mm; forewing length, 4.0 mm; hindwing length, 2.80 
mm; clypeal length, 0.51 mm; scutal length, 1.12 mm. 

Head. Yellow areas: (1) paraocular area below diagonal line originating 
at upper end of outer subantennal suture and extending dorsolaterally tangent 
to facial fovea, ending on orbit at about midlevel of facial fovea; (6) mandible 
basal half; (7) scape, except for brown dorsoapical band with streak; pedicel, 
ventrolaterally; flagellomeres 1-3, ventrally. (10) As in female. (13) Orbital 
convergence ratio as 1.33:1.02,1.30. (14) As in female. (15) Head width to 
head length as 2.11:1.46,1.44. (17) Eye length, mio, and flagellar length as 
1.17:1.02:1.43. (18) Interocellar, ocellocular, antennocular, and interantennal 
as 0.41 :0.37 :0.27 :0.34. (19) Ocellolabral greater than clypeal width, 1.28:1.11, 
1.15. (20) Clypeocellar to outer subantennal sutural as 0.77:0.63,1.21. (21) 
Basal labial palpomere about 1.8 times length of others combined. (22) 
Flagellar length about 3.2 times length of scape, 1.43:0.44. 

Mesosoma. (25) Scutal disc with punctures 1 pwa or less, interspaces dull. 
(27) Foreleg with yellow on coxa apically, femur dorsoapically, extending 
onto anterior and posterior surfaces, tibia (a dot of brown on posterior sur- 
face sometimes), basitarsus and mediotarsus, distitarsus testaceous. (28) Mid- 
dle leg colored like foreleg but less yellow on femur and entire apicotarsus 
testaceous; lengths of tibia, basitarsus, and apicotarsus as 0.85:0.66:0.78. (29) 
Hind leg colored like middle leg. (32) Marginal cell 6-9 greater than, and 
3-4 less than 9-wt, 0.83 :0.70:0.78. 


852 Tue UNIversiry SCIENCE BULLETIN 


—— 


Metasoma. (34) As in female. (35) Tergum 1 with punctures of median 
area medium sized, less than 1 pwa, interspaces shiny. 

Type Marertat. Holotype male and allotype female, from Three Rivers 
(7.5 m. S.), Otero Co., New Mexico, Sept. 9, 1961 (P. D. Hurd), on Gutier- 
rezia microcephala, are in the collections of the California Insect Survey, — 
University of California, Berkeley. 


In addition, 24 males and 37 female paratypes are from the following localities: ARIZONA; i 
Chambers, 1 female, Sept. 19, 1938 (I. H. McCracken), Helianthus; Coconino Co., 1 female, 
Aug. 19, 1927 (P. A. Readio); Petrified Forest, 1 female, Aug. 27, 1931 (P. H. Timberlake), — 
Gutierrezia sarothrae; Taylor, Navajo Co., 1 male, Sept. 14, 1961 (P. D. Hurd), Gutierrezia; — 
Tucson (9 mi. S.£.), 1 female, Sept. 3, 1961 (P. D. Hurd), Bahia absinthifolia. NEw Mexico: — 
Carlsbad (5 mi. N.), Eddy Co., 1 female, Sept. 21, 1956 (J. W. MacSwain); Carrizozo, | male, — 
1 female, Sept. 10, 1961 (P. D. Hurd), Gutierrezia microcephala; idem (8 mi. N.), 1 male, — 
Aplopappus spinulosus; Correo, 1 female, Sept. 4, 1930 (Timberlake), Gutierrezia sarothrae; — 
Laguna, 2 males, Sept. 4, 1930 (Timberlake), [socoma wrighti; Mesilla Park (8.5 mi. E.) Dona ; 
Ana Co., 1 male, Sept. 5, 1961 (P. D. Hurd), Gutrerrezia lucida; Rinconada, 1 female, Sept. 26 — 
(T. D. A. Cockerell, No. 5547), tall Brgelovia; Roswell, Chaves Co., 2 males, Sept. 11, 1961 © 
(P. D. Hurd), Gutierrezia longifolia; Roswell, 3 males, 1 female, Sept. 12, 1937 (R. H. Crane § 
dall); Three Rivers (7.5 mi. $.), Otero Co., 11 males, 27 females, Sept. 9, 1961 (P. D. Hurd), 
Gutierrezia microcephala. Texas: The Basin, Big Bend National Park, 1 male, 1 female, in copulo, 
Oct. 4, 1956 (J. W. MacSwain). Uran: Wildcat Canyon (N. of Beaver), 1 male, Sept. 7, 1954 
(G. F. Knowlton), Chrysothamnus nauseosus. 


Distripution. Extreme western Texas, northwest through New Mexico to 
northern Arizona, the range is allopatric to that of the close relative pectidis. 
C. timberlakei occurs generally in more mountainous localities, whereas — 
pectidis is more of a desert form. | 

Frower Recorvs. Haplopappus spinulosus, Bahia absinthifolia, Bigelovia 
(tall), Chrysothamnus nauseosus, Gutierrezia longifolia, G. microcephala, G. 
sarothrae, Isocoma wrightii. C. timberlake: shares one genus, Haplopappus, 
with C. pectidis, and one genus, Gutierrezia, with bernardinensis. 


ree 


— 


CALLIOPSIS (CALLIOPSIMA) BERNARDINENSIS Michener 
(Figs. 77-82; Map 6) 
Calliopsis bernardinensis Michener, 1937, Ann. Mag. Nat. Hist., (10)19:323; Michener, 1951, i 

Muesebeck er al., U.S. Dept. Agric., Monogr. No. 2:1103. 

This species is closest to timberlakei and pectidis. The male is distin 
cuished by the densely punctured, dull, tergum 1, with interspaces about 0.33 q 
pwa, by the flagellum with posterior surface brown, and by the large patches | 
of brown color on the posterior surfaces of the tibiae. The female is distin-- 

; : 4 | 
euished by the completely yellow labrum and the brown hind leg which” 
sometimes bears a small yellow spot near the base of the tibia beside the basi- ~ 


tibial plate. 
Fremate. Length, 7.6 mm; forewing length, 5.2 mm; hindwing length, 3.50 
mm; clypeal length, 0.58 mm; scutal length, 1.33 mm. 


Head. Yellow areas: (1) as in timberlakei; (2) as in timberlakei; (3) — 
labrum entirely; (5,6,7,8) as in timberlakei. (10) Punctures beside lower half 


A ReEvIsION OF THE BEE GENUS 85 


erlakei , 


f: timb 


Map 7. Map showing the known distributions of Calliopsis (Calliopsima) timberlake: Shinn, C. 
(C.) unca Shinn, and C. (C.) kucalumea Shinn. 


of frontal line smaller than those of pectidis or timberlake, 1 pwa or less, 
interspaces shiny. (13) Orbital convergence ratio as 501431 (st) eas 
in timberlakei. (15) Head width to head length as 2.48:1.77,1.40. (17) Eye 
length, mio, and flagellar length as 1.34:1.43:1.36. (18) Interocellar, ocellocu- 
lar, antennocular, and interantennal as 0.48:0.44:0.39:0.39. (19) Ocellolabral 
greater than clypeal width, 1.50:0.80,1.87. (20) Clypeocellar to outer suban- 


854 Tue University SCIENCE BULLETIN 


tennal sutural as 0.92:0.77,1.20. (21) Basal labial palpomere about 2.5 times 
length of others combined. (22) Flagellar length about 2.2 times length of 
scape, 1.36:0.61. 

Mesosoma. (23) Yellowish areas: apex pronotal lobe; scutellar crest. (24) 
Color as in timberlakei, length of shorter hairs intermediate between pectidis 
and timberlakei; some specimens with shorter hairs dense, profusely branched, 
obscuring scutal surface as in pectidis. (25) Scutal disc with punctures smaller 
than in pectidis, larger than in temberlakei, slightly more than 0.5 pwa. (26) | 
Dorsal enclosure of propodeum as in timberlakei except ridges slightly further 
separated. (27) Foreleg with yellow on femur dorsoapically, dorsal surface of 
tibia, basal half or less of anterior surface of basitarsus. (28) Middle leg with 
yellow on femur dorsoapically, on basal half of dorsal surface of tibia, some- — 
times a spot on basitarsus basally; spur length about half of basitarsal length, 
0.39:0.80. (29) Hind leg brown, rarely with yellow basal spot anterior to 
basitibial plate. (30) Tegula pale straw color to colorless. Humeral plate 
yellow apically. (31) As in timberlaket. (32) Marginal cell 6-9 longer than, 
and 3-4 equal to 9-wt, 1.11 :0.94:0.94. 

Metasoma. (34) Tergal hair bands entire although somewhat sparser 
medially and sparser than in pectidis; suberect hair of discs of terga 4-5 white. 
(35) Tergum 1 with punctures of median area smaller than on scutum, 
smaller than in pectidis, larger than in timberlakei, deep, regularly spaced, 
05-1 pwa. 


Mate. Length, 6.4 mm; forewing length, 4.2 mm; hindwing length, 2.80 
mm; clypeal length, 0.51 mm; scutal length, 1.10 mm. 

Head. Yellow areas: (1) paraocular area as in timberlakei; (6) mandible 
basal half; (7) scape, except dorsally; pedicel, ventrolaterally, much less dis- | 
tinct than in timberlakei (absent in some cases); flagellomere 1, ventrally 
(sometimes). (10) Punctures beside lower half of frontal line slightly larger 
than in timberlakei, 0.5 pwa or less, interspaces shiny. (13) Orbital con- 
vergence ratio as 1.34:1.00,1.34. (14) As in female. (15) Head width to head | 
length as 2.07:1.53,1.36. (17)) Eye length, mio, and flagellar length as 1.17: 
1.00:1.50. (18) Interocellar, ocellocular, antennocular, and interantennal as 
0.39:0.37 :0:22:0.32. (19) Ocellolabral greater than clypeal width, 1.28:1.11,) 
1.15. (20) Clypeocellar to outer subantennal sutural as 0.77:0.63,1.22. (21)) 
Basal labial palpomere about 2.3 times length of others combined. (22) | 
Flagellar length about 3.4 times length of scape, 1.50:0.44. 

Mesosoma. (25) Scutal disc with punctures fine, slightly larger than in 
timberlakei, mostly 0.5 pwa, interspaces moderately shiny. (27) Foreleg with 
yellow as in timberlakei but sometimes absent on coxa, tibia except large pos- 
terior patch of brown reaching almost to apex. (28) Middle leg colored like 
foreleg but yellow absent from coxa, apicotarsus pale brown; lengths of tibia. 
basitarsus, and apicotarsus as 0.87:0.73:0.77. (29) Hind leg colored like mid. | 


| 
| 


A REVISION OF THE BEE GENUS 855 


dle leg, brown patch of tibia covering most of posterior surface. (32) Mar- 
ginal cell 6-9 greater than, and 3-4 subequal to 9-wt, 0.87 :0.78 :0.80. 

Metasoma. (35) Tergum 1 with punctures of median area largest, deepest 
of the pectidis group, crowded, contiguous, giving surface a definite bumpy 
character. 

Type Mareriar. Holotype male and allotype female, from Erwin Lake, 
San Bernardino Mountains, California, Aug. 22, 1932 (C. D. Michener), are 
at the California Academy of Science, San Francisco. I have not seen the 
types. The above description of the female is principally based on a paratype 
with the same label data as the type, while that of the male is based on a 
specimen from the Upper Santa Ana River, San Bernardino Co., California, 
Sept. 1, 1946 (Grace H. and John L. Sperry), on Senecio 1onophyllus. 

Disrrisution. A late-summer, early-autumn bee. 

Fourteen males and eight females have been studied from the following localities, which in- 
clude the type locality: CaLirornia: Erwin Lake, San Bernardino Mts., Aug. 16, 22, 1932 (C. D. 
Michener); Riverside, Sept. 26-28, 1934 (P. H. Timberlake), on Gutierrezia sarothrae; Santa Ana 
River (upper), San Bernardino Co., Aug. 23, 1946 (Grace H. and John L. Sperry), on Senecio 
tonophyllus. 

Discussion. This species is probably a form isolated from an originally 
continuous range of C. timberlaket. 

Frower Recorps. Gutierrezia sarothrae and Senecio tonophyllus, both 
Compositae. Gutierrezia is also used by timberlake. 


CALLIOPSIS (CALLIOPSIMA) UNCA, new species 
(Figs. 83-86; Map 7) 


The specific name is from the Latin wncus, a hook, named for the hook- 
like projection on the outer posterior corner of the male volsella. It is closest to 
crypta and chlorops but is distinguished by the shape of the male volsella, and 
by the large amount of yellow on the ventral surface of the scape as described 


below in (7). 


Mate. Length, 6.8 mm; forewing length, 4.5 mm; hindwing length, 3.10 
mm; clypeal length, 0.54 mm; scutal length, 1.19 mm. 

Head. Yellow areas: (1) paraocular area below diagonal line originating 
at antennal socket above upper end of outer subantennal suture and extending 
concavely dorsolaterally tangent to facial fovea, ending on orbit slightly above 
level of lower border of facial fovea; (6) mandible basal half; (7) scape ex- 
cept dorsal surface and apex ventromesally; flagellomeres 1-2 both with tiny 
patch ventrally. (8) Hairs of vertex fulvous with dark tips, of frons and 
clypeus white, more plumose than in chlorops or in crypta. (10) Punctures 
beside lower half of frontal line larger than those of azteca or chlorops, about 
the same size as those of crypta, 1 pwa, interspaces shiny. (13) Orbital con- 


856 Tue University SCIENCE BULLETIN 


vergence ratio as 1.43:1.11,1.29. (14) Galea shiny, apical pebbling barely 
perceptible (30%). (15) Head width to head length as 2.23:1.60,1.39. (17) 
Eye length, mio, and flagellar length as 1.21:1.11:1.56. (18) Interocellar, ocel- 
locular, antennocular, and interantennal as 0.43:0.41:0.27:0.34. (19) Ocello- 
labral less than clypeal width, 1.16:1.21,0.96. (20) Clypeocellar to outer sub- 
antennal sutural as 0.80:0.66,1.20. (21) Basal labial palpomere about 2.4 times 
length of others combined. (22) Flagellar length about 3.0 times length of 
Scape, oS: 0ie 

Mesosoma. (24) ) Scutal and scutellar shorter hairs grayish-fulvous, longer 
hairs amberisk, dark when viewed from behind. (25) Scutal disc with punc- 
tures deepest, most distinct in its group, relatively large, 0.33 pwa, interspaces 
moderately shiny. (26) Dorsal enclosure of propodeum with carinate border, 
shiny (30%) with high, narrow, strongly vermiform longitudinal ridges 
medially, somewhat straighter longitudinal ridges laterally, median portion 
produced posteriorly. (27) Foreleg with yellow as in bernardinensis but ab- 
sent from coxa. (28) Middle leg with yellow on femur dorsoapically, tibia 


except large median brown patch on posterior surface, basitarsus and follow- | 
ing two tarsomeres, last two testaceous; lengths of tibia, basitarsus, and apico- 


tarsus as 1.02:0.78:0.94. (29) Hind leg colored like middle leg but tibia with — 


only tiny, pale brown patch medially. (30) Tegula transparent light brown | 


with anterior yellow patch. Humeral plate yellow apically. (32) Marginal 
cell 6-9 and 3-4 both less than 9-wt, 0.88 :0.78 :0.94. 

Metasoma. (34) Tergal hair bands indistinct, hairs fine, appressed, whit- 
ish. Suberect hairs of discs of terga 5-6 white. (35) Tergum 1 with pune- 
tures of median area smaller than on scutum, deep, distinct, about 0.33 pwa, 
interspaces moderately shiny. 

Tyre Marertat. Holotype male, from Bingham (3 mi. W.), Socorro Co., 


New Mexico, Sept. 12, 1961 (P. D. Hurd), on Baileya pleniradiata, is the | 


property of the California Insect Survey, University of California, Berkeley. — 


Distriwution. The type locality of this species is a relatively short dis- 


tance from the site of the world’s first atomic bomb explosion near Alamo- — 
gordo, New Mexico, in July, 1945. It seems unlikely, however, that this man- © 
made radiation has created a mutated population from which this unique ~ 
specimen has arisen. Persistent collecting from central to southern New — 
Mexico and adjacent Mexico will probably be necessary to get a good study — 


series of specimens. 


CALLIOPSIS (CALLIOPSIMA) AZTECA, new species 
(Figs. 91-94, Map 5) 


The specific name is from the Nahuatlan, Azteca, meaning an Indian of 
the Nahuatlan tribe which founded the Mexican Empire, and is given because 


A REvIsIOoN OF THE BEE GENUS 857 


of the occurrence of the species in the heart of the former Aztec territory in 
Mexico. It is closest to wnca but is easily distinguished by the wide, impunc- 
tate shiny area adjacent to the upper rim of the anterior declivity of meta- 
somal tergum 1, by the yellow color on the scape being reduced to a small 
ventrobasal patch, and by the much finer, more widely separated punctures of 
the frons at the middle of the frontal line. 


Mate. Length, 6.0 mm; forewing length, 4.1 mm; hindwing length, 3.0 
mm; clypeal length, 0.46 mm; scutal length, 1.10 mm. 

Head. Yellow areas: (1) paraocular area below diagonal line originating 
just below upper end of outer subantennal suture and extending convexly 
dorsolaterally tangent to facial fovea, ending on orbit slightly above level of 
lower border of facial fovea; (6) as in wnca; (7) scape a small patch on ventral 
surface basally; flagellomeres 1-3 with tiny, successively larger patches ven- 
trally. (8) Hairs of vertex and frons fulvous, of clypeus whitish. (10) Punce- 
tures beside lower half of frontal line, finest in the subgenus, somewhat irreg- 
ularly distributed, 2-4 pwa, interspaces shiny (30); punctures half to third 
of diameter of those of wnca, much farther apart. (13) Orbital convergence 
ratio as 1.29:0.99,1.30. (15) Head width to head length as 2.06:1.53,1.35. (17) 
Eye length, mio, and flagellar length as 1.22:0.99:1.39. (18) Interocellar, 
ocellocular, antennocular, and interantennal as_ 0.34:0.37:0.24:0.27. (19) 
Ocellolabral greater than clypeal width, 1.24:1.04,1.19. (20) Clypeocellar to 
outer subantennal sutural as 0.78:0.51,1.53. (21) Basal labial palpomere about 
2.0 times length of others combined. (22) Flagellar length about 3.0 times 
length of scape, 1.39:0.46. (24) Scutal and scutellar shorter and longer hairs 
whitish to fulvous, longer hairs amberish, not dark when viewed from behind. 
(25) Scutal disc with punctures deep, distinct, medium sized, about two-thirds 
diameter of those of wnca, 0.33 pwa, interspaces moderately shiny. (26) As in 
unca but ridges lower, fewer medial vermiform ridges, more longitudinal 
ridges laterally. (27) Foreleg with yellow on femur dorsoapically extending 
onto anterior and posterior surfaces, less extensive than that of wnca, tibia 
except larger brown patch on posterior surface, basitarsus and following 
tarsomere, other tarsomeres pale testaceous. (28) Middle leg colored like 
foreleg but less yellow at apex; lengths of tibia, basitarsus, and apicotarsus 
as 0.90:0.78:0.87. (29) Hind leg colored like foreleg but brown patch of hind 
tibia much less extensive separated from base of tibia by 2 mow, tarsus brown. 
(30) Tegula transparent testaceous without yellow maculation. Humeral 
plate brown. (32) Marginal cell 6-9 and 3-4 both greater than 9-wt, 0.90:0.82: 
0.78. 

Metasoma. (34) As in wnca. (35) Tergum 1 with punctures of median 
area smaller than on scutum, shallow but distinct, 1.5-3 pwa, interspaces 
moderately shiny. Declivity of tergum | shiny, impunctate. 


858 THE UNIVERSITY SCIENCE BULLETIN 


Type Mareriat. Holotype male, from Acatlan, Puebla, Sept. 11, 1948 (H. 
O. Wagner), is at the University of Michigan, Ann Arbor. 

Remarks. This species bears on the outer posterior corner of the male 
volsella a hook-like projection which is like that of wnca. 


CALLIOPSIS (CALLIOPSIMA) CRYPTA Shinn 
(Figs. 87-90; Map 6) 
Calliopsis crypta Shinn, 1965, Amer. Mus. Novitates, 2211:15. 


The specific name from the Greek kryptos, meaning hidden, is applied 
because this species remained mixed in a series of specimens of C. chlorops 
and C. rozeni for a long time prior to its recognition. 

The differentiation of crypta from rogent is discussed under the latter. The 
species is closest to C. chlorops and C. kucalumea. The male of erypta has the 
expanded middle section of sternum 8 with smoothly rounded posterior cor- 
ners, whereas chlorops bears a tiny, posteriorly-directed, sharply-pointed proc- 
ess at each corner. The female of crypta has the mandibular base black or 
brownish black, whereas chlorops has the base cream colored. C. crypta is best 
distinguished from C. kucalumea by the key characters. 


Femae. Length, 8.6 mm; forewing length, 5.5 mm; hindwing length, 3.8 
mm; clypeal length, 0.60 mm; scutal length, 1.49 mm. 

Head. Cream colored areas: (1,2) as in rozeni; (3) dot (sometimes ab- 
sent) dorsally near apex of labral plate; (5) absent on subantennal plate; (6) 
absent on mandible. (7) As in rozeni except tan areas of flagellomeres 1-4 
smaller. (8) Hair of vertex mostly brown (view with integument as back- 
ground). (10) Punctures beside lower half of frontal line larger than those of 
rozent, less than 1 pwa, interspaces smooth, dull. (13) Orbital convergence 
ratio as 1.68:1.51,1.11. (14) As in rozenz. (15) Head width to head length as 
2.60:1.82,1.43. (17) Eye length, mio, and flagellar length as 1.39:1.51:1.50. | 
Flagellum slightly longer than in rozenz. (18) Interocellar, ocellocular, anten-_ 
nocular, and interantennal ratios similar to those of rozenz, as 0.46:0.49:0.43: 
0.36. (19) Ocellolabral distance slightly greater than clypeal width, 1.53:1.45, 
1.06. (20) Clypeocellar to outer subantennal sutural as 0.94:0.85, 1.10. (21) As 
in rozent. (22) Flagellar length about 2.2 times length of scape, 1.50:0.68. 

Mesosoma. (23) Cream colored areas: as in rozeni except dot on each 
pronotal lobe. (24) As in rozenz but short hairs of scutum fulvous, long hairs 
brown; short hairs and lateral long hairs of scutellum fulvous, other long hairs 
brown. (25) Scutal disc with punctures contiguous, larger than in rozenz, and 
larger than on midvertex, interspaces dull. (26) As in rozeni but more 
ridges and enclosure appearing somewhat duller although with interspaces 
shiny. (27) As in rozent. (28) As in rozeni except spur length to length of 


A Revision oF THE BEE GENUS 859 


middle basitarsus as 0.46:0.94. (29,30,31) As in rozenz. (32) As in rozent but 
ratio as 1.14:0.97:1.11 (some specimens with ratios almost identical to those in 
rozent). 

Metasoma. (34) As in rozenz but disc of tergum 4 with at least ten brown 
hairs. (35) Tergum 1 with punctures of median area smaller than on scutum, 
larger than in rozenz, dense, fairly regularly spaced, 1 pwa or less. 


Mate. Length, 7.3 mm; forewing length, 4.8 mm; hindwing length, 3.4 
mm; clypeal length, 0.52 mm; scutal length, 1.27 mm. 

Head. Yellow areas: (1) paraocular area below diagonal line originating 
on outer subantennal suture at level of middle of antennal socket (or below) 
and extending to lower margin of facial fovea ending on orbit slightly above 
level of lower border of facial fovea; (6) as in rozeni. (7) Scape entirely 
black (basal ventral dot or streak of yellow to a basal, ventral, narrow yellow 
stripe strongly attenuate apically); a dot on lateroventral surface of flagello- 
mere 1. (10) As in female. (13) Orbital convergence ratio as 1.51:1.14,1.33. 
(14) As in rogeni. (15) Head width to head length as 2.33:1.63,1.43. (17) 
Eye length, mio, and flagellar length as 1.21:1.14:1.62. (18) Interocellar, 
ocellocular, antennocular, and interantennal as 0.43:0.43 :0.31:0.34. (19) Ocel- 
lolabral greater than clypeal width, 1.34:1.14,1.18. (20) Clypeocellar to outer 
subantennal sutural identical to that in rozenz, 0.82:0.71,1.14. (21) N. A., 
measured on paratype: basal labial palpomere about 2.4 times length of others 
combined. (22) Flagellar length about 3.1 times length of scape, 1.62:0.53. 

Mesosoma. (24) As in rozent, except scutal and scutellar short hairs ful- 
vous, long hairs whitish (fulvous, or brown as seen from behind). (25) 
Scutal disc with punctures larger than in rozeni or chlorops, mostly 0.5 pwa 
(or less), interspaces shiny. (26) Dorsal enclosure of propodeum with fewer, 
higher ridges than that of rozeni, median ridges more vermiform and lateral 
ridges less straight than those of rozen1. (27) As in rozent. (28) Middle leg 
yellow as in rozeni, but tibia with anterior surface splotched with brown; 
lengths of tibia, basitarsus, and apicotarsus about as in rozent, 1.02:0.85:1.09. 
(29) Hind leg yellow as in rozeni but slightly darker, apicotarsus brown. (32) 
Marginal cell 6-9 greater than, and 3-4 less than 9-wt, 0.99:0.83 :0.90. 

Metasoma. (34) Suberect hairs of disc of terga 4-5 brownish. (35) Ter- 
gum 1 with punctures of median area subequal to those on scutum, larger 
than those of rozeni, contiguous, becoming no more than 0.5 pwa laterad, 
interspaces shiny. (38) Plane of each ventral prong of sternum 6 perpendic- 
ular to surface of sternum whereas ventral prongs of other species of Calliop- 
sima are tilted towards sagittal plane of body. 

Tyre Mareriat. Holotype male, from Rustler Park (near Apache), 
Chiricahua Mts., Cochise County, Arizona, Sept. 5, 1962 (J. G. Rozen, M. 
Statham, S. J. Hessel), and allotype female, from Portal (5 mi. W.), S.W.RS., 


860 Tue UNIversIty ScIENCE BULLETIN 


Cochise County, Arizona, 5400 ft., September 3, 1962 (J. G. Rozen, M. 
Statham), are at the American Museum of Natural History. 

In addition 17 male and 28 female paratypes are from the following localities: Arizona: 
Montezuma Pass, Huachuca Mountains, Cochise Co., 6500 ft.; Portal (5 mi. W.); Rustler Park, 
Chiricahua Mountains, Cochise Co., 8500 ft. CumauAHua: Santa Barbara, 6200 ft. 

Discussion aNp DistrisutTion. An apparent case of interspecific mating 
with male rozenz is discussed above under that species. The one record from 
Mexico indicates that crypta is not strictly an endemic species of southeastern 
Arizona. 

Fiower Recorps. Cirsium, Helianthus, and Heterotheca subaxillaris. Cal- 
liopsis crypta shares Helianthus and Heterotheca flowers with both C. rozeni 
and C. pectidis. 


CALLIOPSIS (CALLIOPSIMA) CHLOROPS Cockerell 
(Figs. 95-98; Map 6) 


Calliopsis chlorops Cockerell, 1899, in Cockerell and Porter, Ann. Mag. Nat. Hist., (7)4:413, 
male; Cockerell and Atkins, 1902, Ann. Mag. Nat. Hist., (7) 10:44; Cockerell, 1906, Trans. 
Amer. Ent. Soc., 32:300; 1908, Canad. Ent., 40:148; 1921, Amer. Mus. Novitates, 24:14. 
This species has affinities with coloratipes, but the preponderance of mor- 

phological similarities are with crypta and wnca from which it is differentiated 

as discussed under those species. 


Femae. Length, 7.9 mm; forewing length, 5.6 mm; hindwing length, 3.65 
mm; clypeal length, 0.56 mm; scutal length, 1.34 mm. 

Head. Yellowish areas: (1) paraocular area as described for subgenus, 
sinuous line originating at about middle of outer subantennal suture, ending 
on orbit slightly below level of middle of facial fovea; (2) clypeus, with two 
longitudinal brown, to black, bars arising above apical margin of disc, or on 
apical margin, at angles of median emargination, reaching upward close to, to 
contacting, frontoclypeal suture along subantennal plate and supraclypeal 
area; frontoclypeal suture black; (3) labrum entirely, to only on median por- 
tion of labral plate; (5) absent on subantennal plate, to almost imperceptible 
dot at 30%; (6) mandible basal half. (7) Flagellomeres 1-3 dark ventrally, 
with tiny dot of tan, flagellomere 4 tan ventrally. (8) Hair of vertex fulvous. 
(10) Punctures beside lower half of frontal line smaller than those of crypta, 
larger than those of coloratipes, 1 pwa or more, interspaces dull. (13) Orbital 
convergence ratio as 1.51:1.48,1.02. (14) Galea shiny basally, finely pebbled, 
dull apically. (15) Head width to head length as 2.57:1.72,1.49. (17) Eye 
length, mio, and flagellar length as 1.29:1.48:1.43. (18) Interocellar, ocellocu- 
lar, antennocular, and interantennal as 0.46:0.46.037:0.36. (19) Ocellolabral 
greater than clypeal width, 1.50:1.41,1.06. (20) Clypeocellar to outer suban- 
tennal sutural as 0.94:0.78,1.20. (21) Basal labial palpomere 2.8 times length 


A RevIsION OF THE BEE GENUS 861 


of others combined. (22)Flagellar length about 2.2 times length of scape, 1.43: 
0.65. 

Mesosoma. (23) Yellowish areas: apex pronotal lobe; scutellar crest. 
(24) Scutal and scutellar longer hairs fulvous (to amber). Shorter hairs more 
plumose than those of crypta but scutal surface readily visible. (25) Scutal 
disc with punctures slightly larger than on frons, deep, crowded, contiguous 
to 0.5 pwa, interspaces dull. (26) Dorsal enclosure of propodeum with fine, 
longitudinally vermiform ridges, interspaces shiny, medial portion prolonged 
posteriorly, posterior border carinate. (27) Foreleg yellowish at dorsal apex 
of femur, knee of tibia and adjacent area subequal to it. (28) Spur length 
slightly more than half length of middle basitarsus, 0.44:0.83. (30) Tegula 
colorless to pale straw color. Humeral plate yellowish apically. (31) Stigma 
testaceous. (32) Marginal cell 6-9 greater than, and 3-4 less than 9-wt, 1.24: 
0.99 :1.05. 

Metasoma. (34) As in crypta but brown hairs often absent. (35) Tergum 
1 with punctures of median area larger than on scutum, usually sparsely, 
irregularly distributed, but rarely relatively regularly distributed medially 
about 1-2 pwa, interspaces highly polished, mirror-like. 


Mate. Length, 5.3 mm; forewing length, 4.1 mm; hindwing length, 2.85 
mm; clypeal length, 0.51 mm; scutal length, 1.05 mm. 

Head. Yellow areas: (1) paraocular area below diagonal line originating 
at a point on outer subantennal suture fourth its length below its summit (or 
from upper end of suture) and extending tangent to facial fovea ending on 
orbit slightly above level of lower border of facial fovea (to about middle) ; 
(5) subantennal plate in some cases with a black triangular area in lowermost 
outer corner including anterior tentorial pit; (6) mandible basal half; (7) 
scape, a tiny basal dot (to narrow band or streak reaching as high as middle 
of scape to all brown). (10) Punctures beside lower half of frontal line of 
moderate size, smaller than those of rozeni, unca, and crypta, larger than 
those of azteca, coloradensis, or pectidis group, 0.5-1.5 pwa, interspaces shiny. 
(13) Orbital convergence ratio as 1.29:1.02,1.27. (14) Galea lightly pebbled, 
dull, tip narrowly rounded. Galeal gap about half length of galea exposed 
beyond closed mandibles, (0.22:0.48). (15) Head width to head length as 
1.96:1.41,1.39. (17) Eye length, mio, and flagellar length as 1.07:1.02:1.38. 
(18) Interocellar, ocellocular, antennocular, and interantennal as 0.44:0.44: 
0.27:0.31. (19) Ocellolabral greater than clypeal width, 1.22:0.97,1.26. (20) 
Clypeocellar to outer subantennal sutural as 0.71:0.60,1.20. (21) Basal labial 
palpomere 2.4 times length of others combined. (22) Flagellar length about 
2.9 times length of scape, 1.38:0.48. 

Mesosoma. (24) Scutal and scutellar hair grayish (to fulvous), longer 
hairs dark in some specimens. Shorter hairs not so dense nor so plumose as to 
hide scutal surface. (25) Scutal disc punctures of moderate size, smaller than 


862 Tue UNiversiry SCIENCE BULLETIN 


in crypta, mostly contiguous, interspaces shiny. (26) Dorsal enclosure of 
propodeum concave inward, declivous, bearing distinct, fine, relatively 
straight, ridges, median portion with vermiform ridges (in some cases). (27) 
Foreleg with yellow on anterior dorsoapical third (to half) of femur, anterior 
surface of tibia, all of basitarsus, apicotarsus testaceous. (28) Middle leg 
colored like foreleg but femoral yellow reduced to sixth (to fourth) of femur; 
lengths of tibia, basitarsus, and apicotarsus as 0.82:0.68:0.78. (29) Hind leg 
colored like foreleg but posterior surfaces of tibia and basitarsus all brown 
(to about four-fifths brown). (32) Marginal cell 6-9 and 3-4 greater than 9-wt, 
0.92 :0.85 :0.80. 

Metasoma. (35) Tergum 1 with punctures of median area larger than 
those of scutum, smaller than those of crypta or rozent, deep, about 0.5 pwa, 
interspaces shiny. 

Type Mareriat. Holotype male, from Las Vegas, New Mexico, Aug. 9 
(W. Porter), on Grindelia squarrosa, is at the University of California, River- 
side. A microscope slide of the mouth parts, U.S. N. M. type slide No. 2390, 
is in the United States National Museum. The above description of the fe- 
male is principally based on a specimen from Morley, Colorado. The male 
holotype is exceptionally small. 

Distrrisution. The species occurs in the Front Range of the Rocky Moun- 
tains from northern New Mexico to northern Wyoming, in the eastern part 
of the Great Basin, and the upper and middle portion of the Colorado Pla- 
teau. It is allopatric with wnca and crypta, but in the western and southern 
part of its range it is sympatric with coloratipes, and also largely with colora- 
densis. It has been collected between July 10 at Petersboro, Utah, and Sept. 
27 at Embudo, New Mexico. The highest elevations for it are 9300 and 9200 
ft. at Ward, Colorado, and at Warner Ranger Station near Moab, Utah, 
respectively, 


In addition to the type, approximately 115 males and females were examined from the follow- 
ing localities: Arizona: Del Rio Verde River vicinity, Yavapai Co.; Eagar; Flagstaff, Coconino 
Co.; Kirkland, Yavapai Co.; Prescott; Springerville (32 mi. W.), Apache Co. Cotorapo: Boulder, 
5500 ft.; Coaldale, 7800 ft.; Colorado Springs; Cortez; Durango; Florissant; Manitou; Meeker, 
6200 ft.; Morley; Peaceful Valley; Ute Creek, Sage Flats; Ward, 9300 ft. IbaHo: Downey; Frank- 
lin, Franklin Co.; Nevapa: Ely (9 mi. W.); Glorieta; Las Vegas; Pecos; Raton; Santa Fe. Uran: 
Aspen Grove; Ballard; Devils Slide, Summit Co.; Garfield; Logan; Magna; Moab (28 mi. E.S.E.), 
Warner Ranger Station, Grand Co., 9200 ft.; Morgan, Morgan Co.; Murray; Petersboro, Cache 
Co.; Salt Lake City, 5000 ft.; Sandy, Salt Lake Co.; Spanish Fork, Utah Co.; Wellsville, Cache 
Co. Wyominec: Wheatland; Yellowstone National Park. 


Flower Recorps. Haplopappus gracilis, Chrysopsis, Grindelia squarrosa, 
Medicago sativa, Ratibida, Verbesina, Viguiera annua. 


CALLIOPSIS (CALLIOPSIMA) COLORATIPES Cockerell 
(Figs. 99-192; Map 5) 


Calliopsis flavifrons race coloratipes Cockerell, 1898, Bull. Denison Univ. Sci. Labs., 11:52, male; 
idem, Bull. Univ. New Mexico, 1:52. 


A REvIsION oF THE BEE GENUS 863 


Calliopsis coloradensis coloratipes; Cockerell, 1900, Entomologist, 33:64, female; idem, 1906, 
Trans. Amer. Ent. Soc., 32:300. 
Calliopsis coloratipes; Cockerell, 1908, Canad. Ent., 40:148; 7dem, 1921, Amer. Mus. Novitates, 
Densilcns 
Outside of its own group this species is close to coloradensis and has mor- 
phological similarities to chlorops. Within its own group it is closest to 
deserticola. The male is easily distinguished by the mesally-tilted, conspicuous 
ventral prongs of sternum 6 whereas prongs are absent in deserticola. The 
female is separated from deserticola only with difficulty, but the dorsal enclo- 
sure of the propodeum is very shiny in coloratipes and quite dull in deserticola 
(20); moreover, the ridges are straight and 1-2 ridge widths apart in 
coloratipes, but are vermiform and run together in deserticola. 


Femae. Length, 7.6 mm; forewing length, 5.3 mm; hindwing length, 3.71 
mm; clypeal length, 0.54 mm; scutal length, 1.36 mm. 

Head. Cream colored areas: (1) paraocular area as described for sub- 
genus, dorsal boundary a straight, to sinuous, line originating at upper end 
of outer subantennal suture, ending on orbit at about level of middle of facial 
fovea; (2) clypeus as in chlorops except many specimens without brown bars, 
with only two small, brown clypeal dots; labrum all yellow to all brown; (5) 
subantennal plate except for subtriangular black area at lower end of outer 
subantennal suture, to mostly dark; (6) mandible basal half. (7) Flagello- 
meres 1-4 black with tiny amount of tan. (8) Hair of vertex fulvous. (10) 
Punctures beside lower half of frontal line fine, 2-4 pwa, interspaces dull (to 
shiny). (13) Orbital convergence ratio as 1.60:1.50,1.07. (14) Galea finely 
pebbled, dull. Galeal length 1.10. (15) Head width to head length as 2.52: 
1.72,1.46. (17) Eye length, mio, and flagellar length as 1.26:1.50:1.31. (18) 
Interocellar, ocellocular, antennocular, and interantennal as 0.48 :0.46:0.39 :0.36. 
(19) Ocellolabral greater than clypeal width, 1.46:0.94, 1.56. (20) Clypeocellar 
to outer subantennal sutural as 0.92:0.80,1.15. (21) Basal labial palpomere 2.4 
times length of others combined. (22) Flagellar length about 2.3 times length 
of scape, 1.31:0.56. 

Mesosoma. (23) Cream colored areas: apex pronotal lobe; scutellar crest. 
(24) Scutal and scutellar longer hairs fulvous (to amber). Shorter hairs more 
plumose than those of chlorops, in some cases hiding scutal surface. (25) 
Scutal disc with punctures fine, about twice diameter of those on frons, finer 
than in chlorops, deep, crowded, mostly 0.5 pwa, interspaces shiny. (26) 
Dorsal enclosure of propodeum with relatively straight shiny ridges separated 
by 1-2 ridge widths, interspaces shiny, median portion prolonged only slightly, 
posterior border carinate; a highly polished, impunctate band on vertical sur- 
face of propodeal triangle adjacent to enclosure. (27) Foreleg cream colored 
at dorsal apex of femur and basal third to half of dorsal tibial surface. (28) 
Middle leg colored like foreieg but about half as much cream color. Spur 


864 Tue UNIversiry SCIENCE BULLETIN 


length less than half length of middle basitarsus, 0.44:0.92. (30) Tegula 


testaceous. Humeral plate cream colored apically. (31) Stigma testaceous. — 


(32) Marginal cell 6-9 and 3-4 both greater than 9-wt, 1.12:1.00:0.94. 
Metasoma. (34) Suberect hairs of discs of terga 4-5 white. (35) Tergum 


1 with punctures of median area smaller than on scutum, finer than in 


chlorops, very sparsely, irregularly distributed to virtually absent, tergum 
highly polished, mirror-like. 


Mate. Length, 6.5 mm; forewing length, 4.4 mm; hindwing length, 4.0 
mm; clypeal length, 0.53 mm; scutal length, 1.22 mm. 
Head. Yellowish areas: (1) paraocular area below diagonal line originat- 


ing at upper end of outer subantennal suture (to fourth length of suture below — 
upper end), ending on orbit below level of middle of facial fovea; (6) man- _ 


dible basal half; (7) scape anterior surface except for narrow brown strip 
mesally, slightly widened mesoapically. (10) Punctures beside lower half of 


frontal line extremely fine, 1-2 pwa, interspaces shiny. (13) Orbital con- | 
vergence ratio as 1.45:1.09,1.33. (14) Galea pebbled, dull apically, tip nar-— 
rowly rounded. Galeal gap about half length of galea exposed beyond closed — 


mandibles, 0.22:0.44. (15) Head width to head length as 2.26:1.55,1.46. (17) | 


Eye length, mio, and flagellar length as 1.24:1:09:1.45. (18) Interocellar, 
ocellocular, antennocular, and interantennal as 0.46:0.39:0.26:0.31. (19) Ocel- 
lolabral greater than clypeal width, 1.28:1.16,1.10. (20) Clypeocellar to outer 
subantennal sutural as 0.75:0.66,1.13. (21) Basal labial palpomere about 2.2 
times length of others combined. (22) Flagellar length about 2.9 times 
length of scape, 1.45:0.49. 

Mesosoma. (24) Scutal and scutellar shorter hairs fulvous, plumose, tend- 
ing to hide scutal surface. (25) Scutal disc with punctures smaller than those 
of chlorops, mostly 0.5 pwa, interspaces shiny. (26) Dorsal enclosure of 
propodeum as in chlorops except ridges finer, straighter, much shinier; me- 
dian portion barely (or not) produced posteriorly, medial length about 0.26, 
or about 1.5 mow. (27,28,29) As in chlorops but posterior surface of hind tibia 
sometimes half yellow basally; lengths of tibia, basitarsus, and apicotarsus as 
0).87:0.77:0.77. (32) Marginal cell 6-9 greater than, and 3-4 subequal to 9-wt, 
1.04 :0.83 :0.82. 

Metasoma. (35) Tergum 1 with punctures of median area smaller than 
those of scutum, smaller than those of chlorops, about 0.5 pwa, interspaces 
shiny. 

Tyre Marertat. Holotype male, from Mesilla, New Mexico, Aug. 12 (T. 
D. A. Cockerell), is at the University of California, Riverside. The above 
description of the female is principally based on a specimen of a pair taken 7m 
copulo at Mesilla Park, New Mexico, Sept. 1 (T. D. A. Cockerell), on 


Bigelovia wrighti. 


A REVISION OF THE BEE GENUS 865 


DisrripuTion. The southern Great Basin and the Colorado Plateau. Its 
range is partially sympatric with coloradensis and chlorops, but it is wholly 
allopatric with deserticola and pugionis. It has been collected from July 2, at 
Carson City, Nevada, to Oct. 23, at Phoenix, Arizona, a distinct difference 
from its nearest relative, deserticola, which 1s a spring bee. 

In addition to the type approximately 41 specimens have been examined from the following 
localities: Arizona: Apache (6 mi. S.E.), Skeleton Canyon; Cochise (2 mi. N.); Florence; Phoe- 
nix; San Xavier Mission; Tempe; Tucson; Willcox; “Southern Arizona.” Nevapa: Carson City, 
Ormsby Co.; Yerrington (8.5 mi. S.), Lyon Co. New Mexico: Jemez Springs, Sandoval Co.; 
Mesilla; Mesilla Park. Uran: Hinckley, Millard Co.; Milford, Beaver Co.; Topaz, Juab Co. 

GrocGRAPHIC VARIATION. Specimens from Utah and Nevada have a longer 
galea, a more protruding clypeus in the female, and tend to be darker than 
those from Arizona and New Mexico. Two male specimens from Willcox, 
Arizona, Aug. 1956 (Ellen Ordway) and 1958 (E. G. Linsley), the latter on 
Baileya pleniradiata, differ from other coloratipes by having white face mark- 
ings with scape yellow, a combination also found in squamuifera. 

Discussion. C. coloratipes appears to have given rise to deserticola in the 
southwestern extreme of its range. The basal cream color of the mandible 
and the sparsely punctate median portion of tergum 1 indicate a relationship 
to chlorops. The cream colored subantennal plates of coloratipes are corre- 
spondingly black in chlorops and either black or partly cream-colored in 
coloradensis. The smaller punctures of coloratipes and the straight-ridged 
character of the dorsal enclosure of the propodeum betoken a relationship 
with coloradensis. 

Frower Recorps. Asclepias subverticillata, in copulo on both Bigelovia 
hartwegi and B. wright, Chrysothamnus sp., C. nauseosus consimilis, 
Heterotheca. 


CALLIOPSIS (CALLIOPSIMA) DESERTICOLA, new species 
(Figs. 103-106; Map 5) 


The specific name is compounded from the Latin desertum, a waste place, 
and -cola, dwelling in, in reference to its occurrence in the Colorado Desert in 
California. The species was recognized as new by P. H. Timberlake a num- 
ber of years ago, and his manuscript name is used here. 

Judged by the male it is closest to pugionis, but judged by the female it is 
closest to coloratipes. The male is distinguished by the ratio of the length of 
galea exposed beyond closed mandibles to the galeal gap being 3.5-4.5, whereas 
in pugionis it is 2-3; C. deserticola is easily distinguished from coloratipes by 
the absence of ventral prongs on sternum 6. The female is easily distinguished 
from pugionis by its cream colored, rather than deep lemon yellow, face. It is 
distinguished from coloratipes only with difficulty as discussed under that 
species. 


866 Tue UNIversity ScrENCE BULLETIN 


Femate. Length, 7.5 mm; forewing length, 4.8 mm; hindwing length, 3.30 | 
mm; clypeal length, 0.60 mm; scutal length, 1.22 mm. 

Head. Pale yellowish areas: (1) paraocular area, as described for sub- — 
genus, sinuous line originating about middle of outer subantennal suture 
(lower than in coloratipes!) and ending on orbit about 0.33 (or less) times 
length of facial fovea above its lower border; (2) as in chlorops; (3) labrum 
entirely; (5) subantennal plate a spot touching lower border of antennal — 
socket; (6) mandible basal half. (7) As in coloratipes but lighter. (8) Hair of 
vertex pale fulvous. (10) Punctures beside lower half of frontal line exceed- 
ingly fine, slightly finer than those of pugionis, shallow, 1-3 pwa, interspaces 
highly polished. (13) Orbital convergence ratio 1.45:1.38,1.05. (14) Galea 
very finely pebbled, moderately shiny. Galeal length 1.19. (15) Head width 
to head length as 2.36:1.73,1.36. (17) Eye length, mio, and flagellar length 
as 1.26:1.38:1.31. (18) Interocellar, ocellocular, antennocular, and interanten- — 
nal as 0.43:0.41:0.37:0.37. (19) Ocellolabral greater than clypeal width, 1.48: 
1.45,1.02. (20) Clypeocellar to outer subantennal sutural as 0.88:0.77,1.15. 
(21) Basal labial palpomere about 2.2 times length of others combined. (22) 
Flagellar length about 2.3 times length of scape, 1.31:0.56. 

Mesosoma. (23) Yellowish areas: apex pronotal lobe (in some cases); 
scutellar crest. (24) Scutal and scutellar hairs pale fulvous. Shorter hairs as 
plumose as in chlorops, tending to hide scutal surface; longer hairs shorter, 
more plumose than in coloratipes. (25) Scutal disc with punctures finer, 
shallower and farther apart, 1-2 pwa, than in coloratipes, interspaces shiny. 
(26) Dorsal enclosure of propodeum with numerous close-packed, ultra fine, 
vermiform ridges, dull (20); median portion scarcely (or not at all) pro- 
longed posteriorly, posterior border non-carinate; an impunctate, highly 
polished adjacent band of propodeal triangle present as in coloratipes. (27) 
Foreleg yellowish at dorsal apex of femur, knee of tibia and adjacent area sub- 
equal to it. (28) Middle leg colored like foreleg but about half as much yel- 
low color; spur length about half of length of middle basitarsus, 0.43:0.85. 
(30) Tegula testaceous. Humeral plate yellowish apically. (31) Stigma 
brown. (32) Marginal cell 6-9 and 3-4 both greater than 9-wt, 1.12:1.00:0.94. 

Metasoma. (34) Tergal hair bands sparser, hairs slimmer, less plumose 
than in coloratipes. Suberect hairs of terga 45 white. (35) Tergum 1 with 
punctures of median area much smaller than on scutum, finer than in colora- 
tipes, very sparsely irregularly distributed to virtually absent; tergum highly 
polished, mirror-like. 


Mare. Length, 7.3 mm; forewing length, 4.8 mm; hindwing length, 3.25 
mm; clypeal length, 0.56 mm; scutal length, 1.21 mm. 

Head. Yellow areas: (1) paraocular area as in coloratipes but dorsal 
boundary line may extend along orbit only to level of lower border of facial 
fovea; (5) subantennal plate sometimes with yellow reduced by subtriangular 


A Revision oF THE BEE GENUS 867 


black area around anterior tentorial pit; (6) mandible basal half; (7) scape as 
in coloratipes, flagellomere 1 yellow ventrally. (10) Punctures beside lower 
half of frontal line exceedingly fine, larger than those of azteca or deserticola, 
1-2 pwa, interspaces shiny. (13) Orbital convergence ratio as 1.39:1.16,1.21. 
(14) Galea pebbled, dull, tip broadly rounded. Galeal gap about fourth of 
length of galea exposed beyond closed mandibles, 0.15:0.68. (15) Head width 
to head length as 2.26:1.72,1.31. (17) Eye length, mio, and flagellar length as 
1.22:1.16:1.55. (18) Interocellar, ocellocular, antennocular, and interantennal 
as 0.46 :0.39 :0.27 :0.34. (19) Ocellolabral greater than clypeal width, 1:38:1.22, 
1.13. (20) Clypeocellar to outer subantennal sutural as 0.82:0.71,1.14. (21) 
Basal labial palpomere about 2.1 times length of others combined. (22) 
Flagellar length about 3.1 times length of scape, 1:55:0.49. 

Mesosoma. (24) Scutal and scutellar hair grayish white, shorter hairs 
somewhat plumose, tending to hide scutal surface as in coloratipes. (25) 
Scutal disc with punctures smaller than in coloratipes, mostly 1-1.5 pwa, inter- 
spaces shiny. (26) Dorsal enclosure of propodeum declivous, duil, bearing 
very fine, obscurely vermiform, ridges; enclosure shorter than others in its 
group, about 1 mow medially, median portion not produced posteriorly. (27) 
Foreleg as in coloratipes but posterior surface of basitarsus sometimes testa- 
ceous. (28) Middle leg with yellow on dorsoapical fourth (to third) of femur, 
a stripe along tibia and basitarsus anteroventrally, apicotarsus testaceous; light 
coloration sometimes as unevenly splotched areas; length of tibia, basitarsus, 
and apicotarsus as 0.99:0.88:0.95. (29) Hind leg colored like middle leg but 
less extensive yellow, absent from basitarsus (in some cases), tarsus brown; 
light coloration sometimes as unevenly splotched areas. (32) Marginal cell 
6-9 and 3-4 both greater than 9-wt, 1.12:0.94 :0.82. 

Metasoma. (35) Tergum 1 with punctures of median area subequal to 
those of scutum, smaller than those of coloratipes, 0.5-1 pwa, interspaces shiny. 
Declivity of tergum 1 with more than 25 fine, deep, distinct punctures, mostly 
grouped laterally below dorsal margin. 

Type Marerrar. Holotype male, from Painted Gorge, southeast corner 
of Anza Desert State Park north of Coyote Wells, Imperial Co., California, 
April 12, 1949 (R. A. Hoch), on Encelia farinosa, and allotype female, Box 
Canyon, near Mecca, Riverside Co., April 14, 1935 (P. H. Timberlake), on 
Encelia farinosa, are at the University of California, Riverside. Paratypes are 
at the Snow Entomological Museum of The University of Kansas and in the 
author’s collection. 

In addition, 12 male and 4 female paratypes are from the following localities: CALIFORNIA: 
Box Canyon, Riverside Co., 5 males, 3 females, April 14, 1935 (P. H. Timberlake), on Encelia 
farinosa; Needles, San Bernardino Co., 3 males, 2 females, April 3, 1951 (P. D. Hurd), 2 males, 


same data except (E. G. Linsley); Painted Gorge, Imperial Co., 2 males, April 12, 1949 (R. A. 
Hoch), on Encelia farinosa. 


868 Tue UNIversITY SCIENCE BULLETIN 


CALLIOPSIS (CALLIOPSIMA) PUGIONIS Cockerell 
(Figs. 107-110; Map 5) 
Calhiopsis pugionis Cockerell, 1925, Proc. California Acad. Sci., (4) 14:197, female. 


The specific name is from the Latin pugio, meaning dagger, and refers to 
the dagger-shaped, yellow median line on the clypeus of the female holotype. 
The species is closest to deserticola. C. pugionis is distinguished in the male 
by the very deep, distinct, large, contiguous punctures of the median portion 
of tergum 1 whereas those of deserticola are not nearly so deep nor distinct, 
are of medium size and are mostly 1 pwa. The female is easily distinguished 
by the deep lemon yellow facial coloring in contrast with the cream or pale 
yellowish facial coloring of deserticola. Both sexes are distinguished from 
deserticola by the galeal gap being greater than the middle ocellar width, and 
by the galeal length being only about six-sevenths that of deserticola. 


Femate. Length, 8.0 mm; forewing length, 5.5 mm; hindwing length, 3.67 
mm; clypeal length, 0.58 mm; scutal length, 1.32 mm. 

Head. Deep lemon yellow areas: (1) paraocular area as described for 
subgenus, sinuous line originating slightly above middle of outer subantennal 
suture, ending on orbit at level of lower third or less of facial fovea; (2) 
clypeus with wide, twin vertical brown or black bars arising near angles of 
median apical emargination, frequently joined by subapical brown band, 
reaching close to frontoclypeal suture, the bars separated by a dagger-shaped 
yellow area less than 1 mow in width; brown bars strikingly similar to those 
in coloradensis; (3) labrum almost entirely to absent; (5) subantennal plate 
with small spot, or absent; (6) mandible basal fifth to absent. (7) Flagello- 
meres 1-4 brown and tan ventrally. (8) Hair of vertex amber with brownish 
tips. (10) Punctures beside lower half of frontal line exceedingly fine, finest 
in the subgenus, deep, 2-4 pwa, interspaces highly polished. (13) Orbital 
convergence ratio as 1:58:1.48,1.07. (14) Galea finely pebbled, dull. Galeal 
length about six-sevenths that of galeal length of deserticola, 1.02. (15) Head 
width to head length as 2.48:1.80,1.38. (17) Eye length, mio, and flagellar 
length as 1.36:1.48:1.33. (18) Interocellar, ocellocular, antennocular, and inter- 
antennal as 0.49:0.48:0.41:0.37. Antennocular equal to or greater than inter- 
antennal! (19) Ocellolabral equal (to subequal) to clypeal width, 1.50:1.50, 
1.00. (20) Clypeocellar to outer subantennal sutural as 0.92:0.78,1.17. (21) 


Basal labial palpomere about 2.5 times length of others combined. (22) | 


Flagellar length about 2.2 times length of scape, 1.33:0.60. 

Mesosoma. (23) Yellow areas: apex pronotal lobe; scutellar crest (some- 
times), or testaceous. (24) Scutal and scutellar longer hairs amberish, apical 
portion frequently brownish. Shorter hairs plumose as in deserticola, tending 
to hide scutal surface; longer hairs as in deserticola. (25) Scutal disc with 
punctures finest in subgenus, deep, 1 pwa or less, interspaces shiny. (26) 


A REVISION OF THE BEE GENUS 869 


Dorsal enclosure of propodeum with somewhat fewer, more distinct, very 
fine vermiform ridges, usually dull but shinier, than in deserticola; median 
portion slightly prolonged posteriorly, posterior border carinate; impunctate, 
polished adjacent band of propodeal triangle narrower than in deserticola. 
(27) Foreleg deep lemon yellow with same extent as yellowish on foreleg of 
deserticola. (28) Middle leg deep lemon yellow with same extent as yellowish 
on middle leg of deserticola; spur length half of length of middle basitarsus, 
0.44:0.88. (30) Tegula testaceous. Humeral plate yellow to testaceous apically. 
(31) As in deserticola. (32) Marginal cell 6-9 greater than, and 3-4 less than 
9-wt, 1.09:0.90:1.04. 

Metasoma. (34) Tergal hair bands denser than in deserticola. Suberect 
hairs of terga 4-5 white. (35) Tergum 1 with punctures of median area much 
smaller than on scutum, about same size or larger than those in deserticola, 
very sparsely, irregularly distributed to virtually absent; tergum highly pol- 
ished, mirror-like. 


Mate. Length, 7.5 mm; forewing length, 5.3 mm; hindwing length, 3.6 
mm; clypeal length, 0.61 mm; scutal length, 1.34 mm. 

Head. Deep lemon yellow areas: (1) as in deserticola; (6) mandible 
basal half; (7) scape as in deserticola (variable, to only basal spot with streak 
extending dorsally from it). (10) Punctures beside lower half of frontal line 
larger than those of deserticola, smaller than those of coloratipes, mostly 1 
pwa, interspaces shiny. (13) Orbital convergence ratio as 1.58:1.22,1.29. (14) 
Galea pebbled, dull, tip broadly rounded. Galeal gap about half length of 
galea exposed beyond closed mandibles, 0.26:0.58. (15) Head width/head 
length as 2.52:1.87,1.34. (17) Eye length, mio, and flagellar length as 1.43: 
1.22:1.62. (18) Interocellar, ocellocular, antennocular, and interantennal as 
0.44 :0.46:0.27:0.36. (19) Ocellolabral greater than clypeal width, 1.50:1.39, 
1.07. (20) Clypeocellar to outer subantennal sutural as 0.88:0.75,1.18. (21) 
Basal labial palpomere about 2.4 times length of others combined. (22) 
Flagellar length about 3.0 times length of scape, 1.62:0.54. 

Mesosoma. (24) Scutal and scutellar shorter hairs grayish white, non- 
plumose, not hiding scutal surface as in deserticola. (25) Scutal disc with 
punctures larger than in deserticola, mostly 1-1.5 pwa, interspaces shiny. (26) 
Dorsal enclosure of propodeum declivous, dull, bearing fine (but coarser than 
in deserticola), obscurely vermiform ridges; enclosure not produced posteri- 
orly but long, length about 0.31, or 1.7 mow. (27) Same note as for deserticola. 
Foreleg colored like that of deserticola but basitarsus may be brown. (28) 
Middle leg colored like that of deserticola except tarsus may be brown; lengths 
of tibia, basitarsus, and apicotarsus as 1.12:0.92:0.99. (29) Hind leg colored 
like that of deserticola. (32) Marginal cell 6-9 longer than, and 3-4 shorter 
than 9-wt, 1.12:0.90:0.94. 


870 Tue UNIversity SCIENCE BULLETIN 


Metasoma. (35) Tergum 1 with punctures of median area larger, deeper 
than those of scutum, very coarse, larger than those of coloratipes or deserti- 
cola, contiguous, interspaces shiny, but tergum zm” toto rather dull. Declivity 
of tergum 1 with more than 25 coarse deep, distinct punctures, mostly grouped 
about third of length of declivity below the dorsal margin. 

Type Mareriat. Holotype female, from Soboba Springs, Riverside Co., 
California, June 3, 1917 (E. P. Van Duzee), is Type No. 1657 at the Cali- 
fornia Academy of Sciences, San Francisco. The above description of the 
female is principally based on a specimen from Palm Springs, Taquitz Can- 
yon, California, April 16, 1938 (R. M. and G. E. Bohart) while that of the 
male is based on a specimen from Riverside, California, May 3, 1928 (P. H. 
Timberlake), on garden Coreopsis. 

DisrripuTion. Known only from San Diego, Riverside, and Los Angeles 
counties in southwestern California. It is a spring bee collected from April 6 
to June 14. 


I have not examined the type but have studied approximately 70 males and females from the 
following localities: CaLtForniA: Acton, Mint Canyon, Los Angeles Co., May 3, 1936 (E. G. 
Linsley), on Chaenactis; Dulzura, June 14, 1917; Newhall, Los Angeles Co., April 20, 1940 (R. 
M. Bohart); Palm Springs, Andreas Canyon, Riverside Co., March 24 and April 11, 1936 (P. H. 
Timberlake), Encelia farinosa; idem, April 10, 1936, Encelia farinosa, April 6, 1939 (both C. D. 
Michener); idem, April 7, 1940 (R. M. Bohart); Palm Springs, Palm Canyon, April 15, 1938 
(R. M. and G. E. Bohart), Encelia farinosa; Palm Springs, Taquitz Canyon, April 16, 1938 (R. 
M. and G. E. Bohart); Gavilan (between Perris and Lake Mathew), Riverside Co., 2200 feet, 
May 15, 1936 (P. H. Timberlake), Chaenactis artemisiaefolia; idem, but no altitude given, April 
18, 1940 (P. H. Timberlake), Oenothera ocitebrana; Riverside, April 25, 1929, April 28, 1934, 
Aug. 5, 1938, all on Chaenactis glabrisscula, May 3, 1928, in copulo on garden Coreopsis, May 17, 
1929, June 7, 1927, both on Coreopsis, May 12, 1932, May 13, 15, 16, 1929, all on Coreopsis 
lanceolata, May 31 and June 9, 1927, Hemizonia wright (all P. H. Timberlake). 

Discussion. Although Cockerell considered this species as nearest to 
coloradensis it is really a rather distant relative. The color pattern of the 
legs of the male is very close to that of deserticola, and the color patterns of 
the legs of both species are remarkably similar to those of Awrdi and kucalu- 
mea. | believe that pugionis is a relict of a formerly widespread population of 
deserticola. 

FLower Recorps. Chaenactis artemisiaefolia, C. glabrisscula, Coreopsis 
(cultivated), C. lanceolata, Encelia farinosa, Hemizonia wright, Oenothera 


ocitebrana. 


CALLIOPSIS (CALLIOPSIMA) HURDI, new species 
(Figs. 111-114; Map 5) 


This species, the largest one of the genus, is named for Dr. Paul D. Hurd, 
Jr., in grateful acknowledgement of his loans of many important specimens 
for my study of the genus. 

The closest relative is Rucalumea from which it is best distinguished by the 
key characters in couplet 24 for males, and in couplet 32 for females. 


A REVISION OF THE BEE GENUS 871 


Femates. Length, 9.8 mm; forewing length, 7.0 mm; hindwing length, 5.00 
mm; clypeal length, 0.65 mm; scutal length, 1.78 mm. 

Head. Pale yellow areas: (1) paraocular area as described for subgenus, 
sinuous line originating below middle of outer subantennal suture and end- 
ing on orbit at about level of third length of facial fovea above its lower bor- 
der; (2) clypeus with wide, twin, vertical black bars arising on clypeal apex 
near angles of median apical emargination, reaching close to frontoclypeal 
suture, the bars separated by a finger of yellow 1 mow in width; color pattern 
similar to that of coloradensis; (3) absent on labrum; (5) absent on suban- 
tennal plate; (6) absent on mandible. (7) Flagellomeres missing. (8) Hair 
of vertex brown. (10) Punctures beside lower half of frontal line of medium 
size about the same size as those of chlorops, 1 pwa, interspaces shiny. (13) 
Orbital convergence ratio as 1.96:1.82,1.07. (14) Galea finely pebbled, slightly 
shiny. Galeal length greatly exceeding that of kucalumea, 1.39:0.99. (15) 
Head width to head length as 3.06:2.12,1.44. (17) Eye length to mio as 1.63: 
1.82, flagellum missing. (18) Interocellar, ocellocular, antennocular, and inter- 
antennal as 0.49 :0.58 :0.49:0.46. (19) Ocellolabral less than clypeal width, 1.79: 
1.96,0.91. (20) Clypeocellar to outer subantennal sutural as 1.12:0.95,1.18. 
(21) Basal labial palpomere about 3.3 times length of others combined. (22) 
Length of scape 0.82, flagellum missing. 

Mesosoma. (23) Yellow areas: absent from apex of pronotal lobe and 
scutellar crest. (24) Scutal and scutellar longer hairs black. (25) Scutal disc 
with punctures small, deep, crowded to contiguous, interspaces shiny, scutal 
surface completely visible from above. (26) Dorsal enclosure of propodeum 
relatively dull with many longitudinally vermiform, distinctly separated 
ridges; median portion prolonged posteriorly, posterior border carinate. (27) 
Foreleg with yellow tibial knee. (28) Spur length slightly more than half of 
length of middle basitarsus, 0.58:1.10. (30) Tegula brown. Humeral plate 
brown. (31) Wing brownish. Stigma brown. (32) Marginal cell 6-9 greater 
than, and 3-4 less than 9-wt, 1.48 :1.16:1.33. 

Metasoma. (34) Suberect hairs of disc of tergum 4 brown, of disc of ter- 
gum 5 smoky. (35) Tergum 1 with punctures of median area smaller than 
on scutum, absent from narrow band immediately behind declivity, 2 pwa, 
interspaces minutely roughened, dull. Area immediately behind declivity 
laterally, impunctate, high polished. 


Mate. Length, 8.0 mm; forewing length, 5.6 mm; hindwing length, 4.0 
mm; clypeal length, 0.63 mm; scutal length, 1.44 mm. 

Head. Yellow areas: (1) paraocular area below diagonal line originating 
slightly below middle of outer subantennal suture and ending on orbit at 
about level of lower border of facial fovea; (3) labrum except coarsely punc- 
tate apical area (to all yellow): (5) absent from subantennal plate (to yellow 


872 Tue UNIVERSITY SCIENCE BULLETIN 


except for subtriangular black area around anterior tentorial pit); anterior 
tentorial pit smaller, more distinct than in other groups of Calliopsima; (6) 
mandible basal fourth (to almost absent). (7) Scape brownish black. (10) 
Punctures beside lower half of frontal line of moderate size, larger than in 
chlorops, mostly 1 pwa, interspaces shiny. (13) Orbital convergence ratio as 
1.60:1.33,1.27. (14) Galea almost imperceptibly roughened (30>), shiny, tip 
narrowly rounded. Galeal gap fourth of length of galea exposed beyond 
closed mandibles, (0.17:0.68). (15) Head width to head length as 2.58 :1.87, 
1.38. (17) Eye length, mio, and flagellar as 1.36:1.33:1.75. (18) Interocellar, 
ocellocular, antennocular, and interantennal as 0.44 :0.51:0.32:0.39. Interocellar 
less than ocellocular! (19) Ocellolabral greater than clypeal width, 1.56:1.41, 
1.11. (20) Clypeocellar to outer subantennal as 0.94:0.82,1.14. (21) Basal 
labial palpomere about 3.2 times length of others combined. (22) Flagellar 
length about 2.9 times length of scape, 1.75:0.61. 

Mesosoma. (24) Scutal and scutellar shorter hairs grayish white, longer 
hairs brown; longer hairs of metanotum and propodeum more plumose than 
in quadridentata or kucalumea. (25) Scutal disc with punctures of moderate 
size about 0.017-0.025 in diam., deep, distinct, mostly less than 0.5-1 pwa, 
interspaces shiny. (26) Dorsal enclosure of propodeum subhorizontal, bear- 
ing strong distinct, straight, longitudinal ridges laterally with shiny inter- 
spaces; median portion with slightly vermiform ridges; median portion pro- 
duced posteriorly; length about 0.24, or about 1.3 mow. (27) Same note as for 
deserticola. Foreleg colored like that of pugionis except light color less exten- 
sive, yellow confined to basal half of anterior surface of basitarsus. (28) 
Middle leg with yellow on tibial knee; lengths of tibia, basitarsus, and apico- 
tarsus as 1.09:0.97:1.05. (29) Hind leg with yellow on disc of basitibial plate 
(absent sometimes) and irregular splotch and/or dot on anterior surface of 
tibia preapically (usually). (30) Tegula brown without yellow patch. Hu- 
meral plate brown. (32) Marginal cell 6-9 and 3-4 greater than 9-wt, 1.21: 
1.07 :0.88. 

Metasoma. (35) Tergum 1 with punctures of median area subequal to 
those of scutum, size intermediate between those of coloratipes and chlorops, 
05-1 pwa, interspaces shiny (anterior median portion sometimes finely 
roughened). 

Tyre Marertat. Holotype male and allotype female, from Soyalo (7 mi. 
S.E.), Chiapas, (R. C. Bechtel and E. I. Schlinger), are at the University of 


California, Berkeley. Of four male paratypes, same data as above except | 


Simojovel (4 mi. S.W.), two are at the California Academy of Sciences, San | 
Francisco, one is at the Snow Entomological Museum of The University of | 


Kansas, Lawrence, and one is in the author’s collection. 


Distripution. This species represents the southernmost record for the | 


subgenus Calliopsima. Since habitats similar to those known for Calliopsima 


A REVISION OF THE BEE GENUS 873 


occur far south of this, and since I feel that Calliopsima is an intruder from 
South America, I anticipate discoveries of other species farther south when 
collections are made during the proper season. 


CALLIOPSIS (CALLIOPSIMA) QUADRIDENTATA, new species 
(Figs. 115-118; Map 5) 


The specific name is compounded from the Latin words quatuor, four, 
and dentata, toothed, and is given in reference to the four tooth-like projec- 
tions of the genital capsule, a condition unique in the genus and possibly in 
the Panurginae. 

It is probably closest to Aurdi but is easily separated from it by the yellow 
coloration on the anterior portion of the tegula and the more extensive yellow 
of the anterior surfaces of the middle and hind tibiae as well as the larger 
punctures of tergum 1 and larger ridges of the dorsal enclosure of the 
propodeum., 

The male of this species is separated from kucalumea by the very long 


galea and labial palp as described in (14) and (21) below. 


Mate. Length, 6.5 mm; forewing length, 5.1 mm; hindwing length, 3.60 
mm; clypeal length, 0.56 mm; scutal length, 1.27 mm. 

Head. Yellow areas: (1) paraocular area as in Aurdi but boundary line 
ending on orbit slightly above level of lower border of facial fovea; (6) 
mandible basal half. (7) Scape brownish black. (10) As in Aurdi but larger. 
(13) Orbital convergence ratio as 1.55:1.16,1.34. (14) Galea finely, distinctly 
pebbled, relatively shiny, tip narrowly rounded. Galeal gap about fourth of 
length of galea exposed beyond closed mandibles, 0.17:0.68. (15) Head width 
to head length as 2.33:1.72,1.38. (17) Eye length, mio, and flagellar length as 
1.21:1.16:1.68. (18) Interocellar, ocellocular, antennocular, and interantennal 
as 0.43:0.46:0.29:0.34. Interocellar less than ocellocular! (19) Ocellolabral 
greater than clypeal width, 1.41:1.16,1.22. (20) Clypeocellar to outer subanten- 
nal sutural as 0.85:0.70,1.22. (21) Basal labial palpomere of paratype 3.2 times 
length of others combined. (22) Flagellar length about 3.3 times length of 
scape, 1.68:0.51. 

Mesosoma. (24) Scutal and scutellar shorter hairs grayish-white, not hid- 
ing scutal surface, longer hairs amberish with tips darkened. (25) Scutal disc 
with punctures slightly larger than in Awrdz, deeper, more distinct, contiguous, 
interspaces shiny but scutum appearing rather dull (20). (26) Dorsal en- 
closure of propodeum declivous, intermediate between hurd: and kucalumea; 
pattern of ridges as in Aurdi except finer, closer; median portion, somewhat 
produced, length about 0.26, or 1.5 mow. (27,28,29) Front, middle, and hind 
legs colored like those of pagionis. (30) Tegula brown with yellow anterior 


874 Tue UNIvERSITY SCIENCE BULLETIN 


patch. Humeral plate with tiny apical yellow dot. (31) Wing clear, stigma 
testaceous. (32) Marginal cell 6-9 and 3-4 both greater than 9-wt, 1.07:0.88: 
0.78. 

Metasoma. (34) As described for female of subgenus. (35) Tergum 1 
with punctures of median area subequal to those of scutum, as large as those 
of crypta, but punctures of lateral area smaller than those of crypta, 0.5 pwa, 
interspaces shiny. 

Tyre Materia. Holotype male, from Lagos de Moreno, Jalisco, 6400 ft., 
Aug. 21, 1954 (C. D. Michener and party), is at the Snow Entomological 
Museum of The University of Kansas, Lawrence, and one male paratype, 
same data, is in the author’s collection. 


CALLIOPSIS (CALLIOPSIMA) KUCALUMEA, new species 
(Figs. 119-122; Map 7) 


The specific name is an acronym formed from Kansas University and 
California University Mexican Expeditions and is given to this species in 
recognition of the valuable specimens for this study which have come from 
several separate expeditions into Mexico made by the two universities. 

The species is closest to Aurdi and is best separated from it by the key 
characters in couplet 24 for males, and in couplet 32 for females. 


Femate. Length, 9.0 mm; forewing length, 6.3 mm; hindwing length, 4.35 
mm; clypeal length, 0.60 mm; scutal length, 1.68 mm. 

Head. Pale yellow areas: (1) as in Aurdi; (2) clypeus with twin black 
bars as in Aurdi (to median portion of clypeus all black except streak of yellow 
along frontoclypeal suture bordering supraclypeal area); (3,5,6) as in hurdt. 
(7) Flagellomeres 1-2 black ventrally, 3-4 mostly black ventrally with some 
tan. (8) Hair of vertex fulvous, with brownish tips (all fulvous, or all 
brownish). (10) Punctures beside lower half of frontal line smaller than 
those of hurd, deep, 1-2 pwa, interspaces shiny. (13) Orbital convergence 
ratio as 1.77:1.63,1.08. (14) Galea finely pebbled, slightly shiny. Galeal length 
much less than in Aurdi, 0.99:1.39. (15) Head width to head length as 2:72: 
1.96,1.39. (17) Eye length, mio, and flagellar length as 1.38:1.63:1.55. (18) 
Interocellar, ocellocular, antennocular, and interantennal as 0.49 :0.54:0.49 :0.46. 
(19) Ocellolabral less than clypeal width, 1:60:1.63,0.98. (20) Clypeocellar to 
outer subantennal sutural as 1.00:0.88,1.13. (21) Basal labial palpomere about 
2.3 times length of others combined. (22) Flagellar length about 2.2 times 
length of scape, 1.55:0.71. 

Mesosoma. (23) As in Aurdi. (24) Scutal and scutellar longer hairs 
amberish to brownish. (25) Scutal disc with punctures as in Aurdi, scutal 
surface less visible than in Aurd: because of greater plumosity of scutal shorter 
hairs. (26) Dorsal enclosure of propodeum slightly shinier than in Aurdi, 


A REVISION OF THE BEE GENUS 875 


ridges fewer, straighter; median portion prolonged posteriorly; posterior bor- 
der carinate. (27) Foreleg with yellow tibial knee. (28) Spur length more 
than half of length of middle basitarsus, 0.53:0.97. (30) Tegula brown. 
-Humeral plate brown with apical yellow spot. (31) Wing brownish, clearer 
than in Aurdi. Stigma brown. (32) Marginal cell 6-9 greater than, and 3-4 
subequal to 9-wt, 1.38:1.17:1.16. 

Metasoma. (34) Suberect hairs of disc of tergum 4 fulvous, of disc of 
tergum 5 mixed white and fulvous. (35) Tergum 1 with punctures of median 
area smaller than on scutum, deep, mostly less than 1 pwa. 


Mate. Length, 7.2 mm; forewing length, 5.4 mm; hindwing length, 3.70 
mm; clypeal length, 0.58 mm; scutal length, 1.43 mm. 

Head. Yellow areas: (1) paraocular area as in Aurdi; (5) subantennal 
plate, a tiny spot (to all yellow except for black subtriangular area around 
anterior tentorial pit); (6) mandible basal half. (7) Scape as in guadriden- 
tata. (10) As in hurd: but slightly larger, deeper, more distinct. (13) Orbital 
convergence ratio as 1.68:1.24,1.36. (14) Galea finely pebbled (30), shiny 
(to dull), tip narrowly rounded. Galeal gap to length of galea exposed beyond 
closed mandibles variable, 0.31:0.51(0.39:0.43). (15) Head width to head 
length as 2.50:1.84,1.36(2.40:1.62,1.48). (17) Eye length, mio, and flagellar 
length as 1.34:1.24:1.70. (18) Interocellar, ocellocular, antennocular, and inter- 
antennal as 0.42:0.48 :0.31:0.34. Interocellar less than ocellocular! (19) Ocello- 
labral greater than clypeal width, 1.48:1.19,1.24. (20) Clypeocellar to outer 
subantennal sutural as 0.90:0.77,1.18. (21) Flagellar length about 3.3 times 
length of scape, 1.70:0.56. 

Mesosoma. (25) Scutal disc with punctures smaller than in guadridentata, 
larger than in Aurdi, 0.5 pwa to contiguous, interspaces shiny, but scutum 
rather dull because of close, deep punctures (20). (26) Dorsal enclosure of 
propodeum sharply declivous; pattern of ridges as in Aurdi; median portion 
but little produced, length about 0.29, or 1.7 mow. (27) Same note as for 
deserticola. Foreleg colored like that of pugionis. (28) Middle leg colored 
like that of hurd; lengths of tibia, basitarsus, and apicotarsus as 1.07 :0.92:0.99, 
ratios which are almost identical with those of hurdi. (29) Hind leg with 
yellow on basitibial plate, irregular splotching along anterior surfaces of tibia 
and basitarsus, latter with black border. (30) Tegula brown without yellow 
patch. Humeral plate brown. (32) Marginal cell 6-9 and 3-4 both greater 
than 9-wt, 1.24:1.09:0.94. 

Metasoma. (35) Tergum 1 with punctures of median area subequal (to 
slightly smaller) to those of scutum, intermediate in size between those of 
kucalumea and hurd1, contiguous, interspaces shiny. 

Type Mareriat. Holotype male and allotype female, from Coyotes, 
Durango, 8300 ft., Aug. 8, 1947 (C. D. Michener, David Rockefeller Expedi- 
tion), are at the American Museum of Natural History, New York. 


876 Tue UNIVERSITY SCIENCE BULLETIN 


In addition four male and six female paratypes are from Mexican localities as follows: Atlaco- 
mulco (22 mi. N.), México, 8100 ft., 3 males, Aug. 18, 1954 (C. D. Michener and party); 
Durango (10 mi. W.), Durango, 1 female, July 12, 1954 (E. I. Schlinger); El Salto (6 mi. 
N.E.), Durango, 8500 ft., 1 female, Aug. 10, 1947 (M. A. Cazier, David Rockefeller Expedition) ; 
Fresnillo (9 mi. S.), Zacatecas, 1 male, Aug. 10, 1954 (E. G. Linsley, J. W. MacSwain, R. F. 
Smith), 1 female, same locality, Aug. 20, 1954 (J. W. MacSwain), on Haplopappus gracilis; 
Toluca (24.5 mi. N.W.), México, 1 female, July 30, 1962 (Naumann and Marston); Zacatlan 
(20.7 mi. N.W.), Puebla, 7950 ft., 2 females, Aug. 22, 1962 (University of Kansas Mexican 
Expeditions). Paratypes are at the University of California, Berkeley; the Snow Entomological 
Museum of The University of Kansas, Lawrence; the American Museum of Natural History, New 
York; and in the author’s collection. 


Subgenus VERBENAPIS Cockerell and Atkins 


Verbenapis Cockerell and Atkins, 1902, Ann. Mag. Nat. Hist., (7) 10:44; Michener, 1951, in 

Muesebeck et a/., U.S. Dept. Agric., Monogr. No. 2:1103. 

Type species. Calliopsis verbenae Cockerell and Porter, 1899, monobasic. 

This subgenus is closer to Calliopsima than to the others. As discussed 
under the genus, it has characters in common with Hypomacrotera and the 
South American Liopoeum and Acamptopoeum. The four species are inter- 
related, but verbenae and micheneri are close to each other as are hirsutifrons 
and nebraskensis. Depending upon what characteristics are chosen, any one 
may be thought of as somewhat singular with respect to the other three, a 
situation which precludes any definitive statement on their phylogeny at 


present. There are no records from Canada, nor from any southern state east — 


of the Mississippi River. They are apparently oligolectic on Verbena flowers, 


and indeed are particularly adapted to collect pollen from them (see under 


bionomics of nebraskensis). 


Verbenapis differs from the other subgenera as follows: propodeum with | 
impunctate, shiny, unsculptured dorsal area bearing a median depression with © 


a longitudinal, low ridge giving the impression of twin pits; posterior dorsal 
margin of the pronotum without pale maculation; front tarsi of females with 
rows of long, stout, curled hairs ventrally; male sternum 5 with posterior 
margin concave medially; sternum 6 with rounded lobes instead of pointed 
projections; areas of the male tergum 7 lateral to pygidial plate dropping 
sharply downward toward the median sagittal plane of the body and acting as 
a sheath for the enclosed genitalia. 


Femare. Length, 6.4-8.8 mm. Integumental background color of head, 
mesosoma, and metasomal terga black, of metasomal sterna black to brown. 
Integument almost entirely shiny, non-metallic, rarely some roughening 
present. 

Head. Cream colored areas: (1) paraocular area with a triangular patch 
in lower ventral corner; (2) clypeus, cream color variable in extent and less 


than in male; (3) labrum, usually labral plate only, absent on punctate apical — 


portion; (4) absent on supraclypeal area except sometimes a tiny dot; (5) 
absent on subantennal plate, but indications of potential cream color in some 


A Revision oF THE BEE GENUS 877 


specimens; (6) mandible, basal portion. (7) Scape, pedicel, and flagellomeres 
1-4 dark brown to black, upper surface of flagellum dark brown, lower sur- 
faces of flagellomeres 5-10 tan. (8) Hair of vertex, frons, gena, and lateral 
angles of clypeus white, of disc of clypeus sometimes fulvous. (10) Punctures 
along ocellocular line medium-sized to very fine, interspaces smooth, moder- 
ately shiny; impunctate area lateral to posterior ocellus shiny; punctures of 
frons near middle of frontal line mixed tiny and large, interspaces smooth, 
moderately shiny. (11) Frontal line with lower portion a raised, narrow 
sulcus. (12) Clypeus with lower half of disc flattened or medially concave. 
Clypeus with a median subapical portion, usually transparent, overhanging 
the preapical groove from which arise long, amber hairs. (13) Inner orbits 
slightly convergent below. Facial fovea shallow, shiny, indistinctly sculp- 
tured (45), elongate elliptical, upper limit slightly below level of middle 
ocellus, lower limit above level of upper rim of antennal socket. (14) Galea 
long, relatively slender; galeal gap absent or less than mow, length of galea 
exposed beyond closed mandibles usually extending to or beyond base of 
prementum. (15) Head width/head length 1.3-1.5. (17) Eye length less than 
mio, subequal to basal labial palpomere, equal to or less than flagellar length. 
(18) Interantennal greater than antennocular and more than twice diameter 
of antennal socket; antennocellar more than twice antennocular. (19) Ocello- 
labral equal to or greater than mio. (21) Basal labial palpomere 3.0-5.5 times 
length others combined. (22) Flagellomere 1 equal to or longer than flagel- 
lomere 9 (in Airsutifrons distinctly shorter). Flagellar length 2.3-2.5 times 
length of scape. 

Mesosoma. (23) Light color absent along posterior dorsal border of 
pronotum. Some specimens with indistinct yellowish tan dash on medial 
apical rim of pronotum and similar dash at extreme posterolateral edge of 
pronotum. (24) Dorsum with hair white to fulvous, white elsewhere. (25) 
Scutal disc with punctures few, scattered, becoming larger, denser laterad, 
then abruptly smaller and denser along edges, interspaces shiny. (26) Dorsal 
enclosure of propodeum impunctate, shiny, unsculptured, bearing a median 
depression with a low, longitudinal ridge giving the impression of twin pits. 
(27) Foreleg with cream coloration. Front tarsus with suberect, thick, apically 
hooked hairs, amber basad, colorless apicad. Front mediotarsus with flat- 
tened, expanded tarsomeres each with a posteroventral prolongation. Front 
distitarsus strikingly different from other distitarsi: base laterally compressed, 
dorsal surface shiny, hairless, except for a few straight apical hairs. Tibial 
spurs white. (28) Middle leg with cream coloration; mediotarsus cylindrical. 
Hind leg dark; mediotarsus cylindrical. (30) Tegula and humeral plate 
without cream maculation (present in michenert). (31) Wing colorless. 
Stigma testaceous to brown. (32) Marginal cell 6-9 subequal to longer than, 
and 3-4 shorter than 9-wt. 


878 Tue University SCIENCE BULLETIN 


Metasoma. (34) Tergal hair bands white, band of tergum 1 often sparse — 
or absent, of terga 2-4 always present. Suberect hair of discs of terga 4-5 long, 
white, usually moderately abundant. Prepygidial and pygidial fimbriae white, _ 
occasionally appearing pale fulvous because of enmeshed pollen or dust 
particles. 

Mate. Length, 5.8-8.8 mm. 

Head. Cream colored areas: (1) paraocular area a triangular patch in — 
lower ventral corner; (2) clypeus, mostly to completely, with apical, narrow — 
brown border; (3) labrum, on labral plate to entire impunctate basal portion, 
absent on punctate apical portion; (4) as in female; (5) absent on subantennal 
plate; (6) mandible basal fourth to two-thirds. (7) Scape, pedicel and 
flagellomeres as described for female but lighter. (8) As in female but more 
intensely fulvous where female has fulvous hair. (10) Punctures as de- 
scribed for female. (11) Frontal line with lower portion a carina with narrow, 
obsolete sulcus or none. (12) Clypeus as described for female. (13) Eyes 
bulging, inner orbits moderately to strongly convergent below. (14) Galea as _ 
in female. (15) Head width/head length 1.30-1.40. (17) Eye length variable — 
with respect to mio and flagellar length. (20) Clypeocellar greater than outer 
subantennal sutural except subequal in micheneri. (21) Basal labial palpo- 
mere as in female. (22) Flagellomere 1 variable with respect to 9. Flagellar 
length 3.4-4.1 times length of scape. 

Mesosoma. (23) As in female. (24) Dorsum with hair as in female. 
Scutellar and metanotal hair pads absent. (25) Scutal disc with punctures as 
in female except finer. (26) As in female. (27) Legs with white, cream, or 
yellowish coloration. Tibial spurs white. Foreleg with cream coloration on 
femoral apex and stripe along anterior aspect of tibia from knee apicad, inter- 
rupted subapically in nebraskensis and micheneri. (28) Middle leg with 
cream coloration at least on knee. Middle tibia with dorsoapical projection. 
(29) Hind leg without light coloration (except in Airsutifrons). Hind medio- 
tarsus and distitarsus shorter than those of other legs. (30,31,32) As in female. 

Metasoma. (34) As described for female but hair bands sparser. (36,37) 
Sternum 5 with posterior margin concave medially. Sternum 6 with a bilobed 
median portion. Sternum 8 with a long, slender, clublike median projection. 
(38) Sterna and genitalia as illustrated (Figs. 123-143). 


CALLIOPSIS (VERBENAPIS) VERBENAE Cockerell and Porter 
(Figs. 123-128; Map 8) 


Calliopsis verbenae Cockerell and Porter, 1899, Ann. Mag. Nat. Hist., (7)4:412; Cockerell, 1906. 
Trans. Amer. Ent. Soc., 32:300; Michener, 1951, 72 Muesebeck ef al., U.S. Dept. Agric., 
Monogr. No. 2:1103 (synonym of Airsutifrons). 


Closest to nebraskensis and micheneri. The female is distinguished from 
nebraskensis by having the brown color of the basal half of mandible with at 


A ReEvIsION OF THE BEE GENUS 879 


most a tiny area of yellow at the extreme base; by the clear oval area on the 
sclerotized apical portion of sternum 7 being a fourth or more wider than 
width of median ocellus; by the front basitarsus being about 6 times longer 
than wide; and by the shape, vestiture, and punctation of metasomal terga 
1-2. The male is distinguished from nebraskensis by the key characters; by 
the eye length being less than the clypeal width; by the clypeal length being 
more than the scape length; and by the sterna and genitalia as illustrated. The 
male is separated from micheneri by the key characters; by a higher ratio of 
head length to minimum interocular and to intertegular; by the outer suban- 
tennal sutural being distinctly less than the clypeocellar; and by the sterna 
and genitalia, as illustrated. 


Femate. Length, 8.5 mm; forewing length, 5.7 mm; hindwing length, 3.9 
mm; clypeal length, 0.59 mm; scutal length, 1.32 mm. 

Head. Cream colored areas: (1) paraocular area a triangular patch in 
lower corner below line originating at junction of outer subantennal suture 
and frontoclypeal suture and extending diagonally ending on orbit slightly 
below level of lower rim of antennal socket; (2) clypeus, apical half or less of 
median portion with a median dorsal rectangular or concave emargination 
(usually), cream color extending onto the posteriorly bent sides of clypeus; 
(3) labral plate; (4) absent on supraclypeal area; (6) mandible a tiny area, 
often absent, at extreme base. (10) Punctures of ocellar triangle and im- 
mediately posteriad, 1 pwa or less. Lower dark paraocular area with punc- 
tures distinct, subequal to those of subantennal plate, 1 pwa; of light area few, 
variable in size. (13) Orbital convergence ratio as 1.67:1.50,1.11. Lower 
median border of supraclypeal area elevated slightly above level of adjacent 
clypeus. (14) Galea unsculptured (to finely pebbled at 30, especially apicad 
and basad); galeal gap absent, tip of galea in repose extending just beyond 
base of prementum. (15) Head width to head length as 2.55:1.84,1.39(1.30- 
1.41). (16) Eye width in profile about 2.5 times protrusion of clypeus beyond 
it. (17) Eye length, mio, and flagellar length as 1.36:1.50:1.36. (18) Interocel- 
lar, ocellocular, antennocular, and interantennal as 0.48:0.53:0.34:0.44. (19) 
Ocellolabral subequal to clypeal width, 1.51:1.53. (20) Clypeocellar to outer 
subantennal sutural as 0.92:0.85,1.08. (21) Basal labial palpomere 3.8 (3.3- 
4.0) times length of others combined. (22) Flagellar length about 2.3 times 
length of scape, 1.36:0.60. 

Mesosoma. (25) Scutal disc with punctures between anterior ends of 
parapsidal lines of variable size but mostly larger ones, 1-2 pwa. Punctures 
immediately below episternal scrobe oblong, 1 pwa or less; anterior to scrobe 
eccentric, less than 1 pwa to contiguous, on a very shiny ground. Punctures 
of propodeal flats round, very deep, smaller than mesepisternal punctures, 
larger than scutellar punctures, less than 1 pwa. (27) Legs with cream color 
the same as on face. Foreleg with cream color on knee. Hooked hairs of front 


880 Tue University SCIENCE BULLETIN 


tarsus slenderest of the subgenus. Front basitarsus narrow, length/width 
from 55 to 65. Front and middle mediotarsi with lengths subequal and 
exceeding hind mediotarsi 0.56:0.56:0.53. Front distitarsus fifth shorter than 
middle distitarsus which is subequal to hind distitarsus. (28) Middle leg 
with cream color on knee. (30) Tegula transparent dark brown. Humeral 
plate lighter brown than tegula. (31) Stigma three tmes as long as wide. 


(32) Marginal cell 6-9 longer than, and 3-4 shorter than 9-wt, 1.26:1.02:1.14 | 


(29 aac), 

Metasoma. (34) Hair bands of terga 1-3 dense, of tergum 4 sparser. (35) 
Tergum 1 with punctures of median area subequal to scutal punctures, 1-2 
pwa, becoming scattered laterad and posteriad, but dense and regularly spaced 
anterolaterad, about 1 pwa, interspaces shiny. Declivity of tergum 1 sharply 
distinct from dorsal portion, shiny, punctures about 15 » in diameter. Raised 
dorsal surface at anterior edge of posterior depressed margin of tergum 1 
complete from side to side. (36) Tergum 2 with punctures of median area 
irregularly spaced, about 1 pwa, sparser and larger laterad. Tergum 2 without 
a median bulge. Pygidial plate with dark brown margin, either light tan or 


dark brown center, about one-fifth longer than basal width, sides forming — 


about 45° angle, tip narrowly rounded. (37) Sternum 6 with large, oval 


median clear area in apical sclerotized plate, this area wider than long, 0.26-_ 


0.29 by 0.17-0.20, length of area 0.35-0.40 times length of sclerotized plate. 


Mate. Length, 7.0 mm; forewing length, 5.2 mm; hindwing length, 3.6 | 


mm; clypeal length, 0.51 mm; scutal length, 1.33 mm. 


Head. Cream colored areas: (1) paraocular area a triangular patch in 
lower corner below a line from a point about a third down on inner orbit to — 


about midpoint of outer subantennal suture (or to vicinity of anterior tentorial 
pit); (2) clypeus except for a brown triangular area in dorsolateral corner 
below subantennal plate, apical border brown laterally, amber medially; (3) 
labrum; (4) absent on supraclypeal area; (6) mandible basal half (or more). 
(10) Punctures similar to female except light portions of paraocular area and 
clypeus well punctured. (13) Orbital convergence ratio as 1.53:1.24,1.22. (14) 
Galea with medial portion shiny, basal and apical portions roughened more 
distinctly than in female; galeal gap absent, tip of galea in repose extending 
slightly beyond base of prementum. (15) Head width to head length as 


2.26:1.72,1.32. (17) Eye length, mio, and flagellar length as 1.21:1.24:1°72am) 


(18) Interocellar, ocellocular, antennocular, and interantennal as 0.43:0.46: | 


0.27:0.37. (19) Ocellolabral greater than clypeal width, 1.36:1.31,1.04. (20) 
Clypeocellar to outer subantennal sutural as 0.87:0.75,1.16. (22) Flagellar 
length about 3.7 times length of scape, 1.73:0.48. 

Mesosoma. (25) Punctation as in female except scutal disc with punc- 
tures finer, shallower, and less dense than in female or in nebraskensis, 2 pwa 


A REVISION OF THE BEE GENUS 881 


between anterior ends of parapsidal lines, mesepisternal punctures rounder. 
(27) Legs with light color in part different from that on face. Foreleg with 
yellowish stripe on tibia reaching apex; basitarsus and second tarsomere 
white on anterior surface, others straw color. (28) Middle leg with ventral 
apical rim of yellow on trochanter; yellow macula on dorsal surface of 
femoral apex and knee; basitarsus colored like that of foreleg, apicotarsus 
light tan, darker than that of foreleg; lengths of tibia, basitarsus, and apico- 
tarsus as 1.00:0.68:0.82. (29) Hind leg colored like middle leg but without 
femoral and tibial light color (basitibial plate may bear indistinct yellow on 
basal half medially). (30) As in female. (31) Wing 12-13/13-14 about 5. 
Stigma about 5 times as long as wide. (32) Marginal cell 6-9 longer than, and 
3-4 shorter than 9-wt, 1.12:0.97 :1.07. 

Metasoma. (34) Tergal hair bands as in female but sparser. (35) Tergum 
1 with punctures of median area subequal to scutal punctures but deeper, 
evenly spaced on dorsum, about 1 pwa. Declivity of tergum 1 as in female 
except at least twice as many punctures, finer than on dorsum. Raised dorsal 
surface as in female. (36) Tergum 2 with punctures of median area finer than 
on 1, sparser, but fairly evenly distributed. Pygidial plate with dark brown (to 
black) flat margin, broadly rounded at tip. 

Discussion. Too few specimens of this species are available to make 
meaningful statements on geographic variation, but the one Mexican female 
is conspicuously larger than the females from the United States. No data on 
the bionomics of the species have been published; Robertson’s notes on 
Verbenapis actually pertained to nebraskensis (qwv.). 

Tyre Marertat. The holotype female taken at Las Vegas, New Mexico, 
Aug. 9 (W. Porter), on Verbena stricta, has not been located. One male para- 
type, same collection data, is in the Academy of Natural Sciences of Philadel- 
phia and two male paratypes, same data as above, one in the U. S. National 
Museum, one in the Museum of Comparative Zoology, Harvard University. 
The above description of the female is principally based on a specimen from 
Pecos, New Mexico, July 7 (T. D. A. Cockerell), on Verbena bipinnatifida, 
while that of the male is based on the first paratype listed above. 

DistripuTioN. Western Texas to southeastern Arizona and south to 
Durango, Mexico. It has been collected in Texas in April only, but elsewhere 
from July 7 to September 7. 

In addition to the type material, 9 females and 2 males have been examined. The localities 
below include the type locality: Arizona: Portal (5 mi. W.), Chiricahua Mts., Aug. PY I ye} (GX 
A. Opler). New Mexico: Ft. Union (2.4 mi. S.), Mora Co., July 9, 1959 (Ray F. Smith); Las 
Vegas, 3 males, 1 female, Aug. 9 (W. Porter), on Verbena stricta, and 1 male, same data, on 
Sphaeralcea lobata; Pecos, 1 female, July 7 (T. D. A. Cockerell), on Verbena bipinnatifida; Pecos, 
1 male, 1 female, July 19 (Cockerell and Porter), on Verbena macdougalli; Rodeo (2.5 mi. N.), 
Hidalgo Co., Sept. 7, 1959 (D. D. Linsdale). Texas: Davis Mts., 1 female, April 17, 1954 (R. H. 


Beamer), on Chamaesaracha conloides, 1 female, same date (L. D. Beamer). Duranco: El Tas- 


cate, 6400 ft., July 28, 1947 (C. D. Michener). 


882 Tue UNIveErsSITY SCIENCE BULLETIN 


~ hirsutifrons #=hie 


\ 


Map 8. Map showing the known distributions of Calliopsis (Verbenapis) hirsutifrons Cockerell, 
C. (V.) verbenae Cockerell and Porter, C. (V.) micheneri Shinn, and C. (V.) nebraskensis 
Crawford. 


CALLIOPSIS (VERBENAPIS) NEBRASKENSIS Crawford 
(Figs. 129-133; Map 8) 


Calliopsis verbenae var. nebraskensis Crawford, 1902, Canadian Ent., 34:240; Graenicher, 1910, 
3ull. Pub. Mus. Milwaukee, 1:238. 

Calliopsis verbenae nebraskensis; Swenk and Cockerell, 1907, Ent. News, 18:179; Graenicher, 
1935, Ann. Ent. Soc. Amer., 28:303. 

Verbenapis verbenae; Robertson, 1914, Ent. News, 25:72; 1922, Psyche, 29:171; 1926, Ecology, 
7:379; 1929, Flowers and Insects, pp. 10, 216-218; Robertson, Psyche 36:115-116; Pearson, 
Ecol. Monogr., 3:378. (misidentifications) 

Calliopsis nebraskensis; Crawford, 1915, Proc. U.S. National Museum, 48:179; Cockerell, 1916, 
Ann. Mag. Nat. Hist., (8)17:279; Rau and Rau, 1916, Jour. Anim. Behav., 6:368; Stevens, 
1919, Canadian Ent., 51:210; Cockerell, 1921, Amer. Mus. Nov., No. 24:14; Rau, 1922, 
Trans. Acad. Sci. St. Louis, 24:33; Cockerell, 1928, Univ. Colorado Studies, 16:103; Stevens, 


A ReEvIsION OF THE BEE GENUS 883 


1950, Bull. North Dakota Agric. Exp. Sta., 12:93; Michener, 1951, 72 Muesebeck ef al., U.S. 
Dept. Agric., Monogr. No. 2:1103; Mitchell, 1960, North Carolina Agric. Exp. Sta. Tech. 
Bull., No. 141, vol. 1:288, 291-294. 

Calliopsis sp.; Washburn, 1919, Minnesota Agric. Exp. Sta., Jour. series, Paper No. 156:229 

(dorsal view of female). 

Females differ from other Verbenapis in the shape and punctation of 
metasomal terga 1 and 2. Females are distinguished from verbenae by having 
the basal third to half of the mandible cream colored; by the clypeal width 
being subequal to the flagellar length; by the front basitarsus being about four 
times longer than wide; and by the finer punctation throughout. The male 
is distinguished from verbenae as discussed under that species; from Aursutt- 
frons, with which it shows strong affinities based on similarities of facial 
structure and dimensions, by the key characters, by the lower ratio of head 
length to intertegular distance, by the flagellar length being subequal to the 
head length, by the higher value of wing ratio 12-13/13-14, and by the sterna 
and genitalia, as figured. 


Femae. Length, 7.5 mm; forewing length, 5.4 mm; hindwing length, 3.7 
mm; clypeal length, 0.51 mm; scutal length, 1.33 mm. 

Head. Cream colored areas: (1) paraocular area as in verbenae (triangu- 
lar patch frequently truncated dorsally); (2) clypeus as in verbenae (upper 
median emargination of cream colored area may be absent, wedgeshaped, 
rectangular, evenly concave, or sinuous) ; (3,4) as in verbenae; (6) mandible, 
basal third. (10) Punctures as in verbenae except somewhat finer, less shiny. 
(13) Orbital convergence ratio as 1.58:1.36,1.16. Lower median border of 
supraclypeal area slightly below level of adjacent clypeus. (14) Galea un- 
sculptured (30), narrower and more sharply pointed than in verbenae; 
galeal gap absent, tip of galea in repose just extending to or beyond base of 
prementum. (15) Head width to head length as 2.43:1.70,1.43(1.36-1.44). 
(16) Eye width in profile 4.0(3.5-4.0) times protrusion of clypeus beyond it, 
clypeus more abruptly protuberant than verbenae. (17) Eye length, mio, and 
flagellar length as 1.31:1.36:1.34. (18) Interocellar, ocellocular, antennocular, 
and interantennal as 0.46 :0.46 :0.34:0.37. (19) Ocellolabral greater than clypeal 
width, 1.43:1.36,1.05. (20) Clypeocellar to outer subantennal sutural as 0.92: 
0.77,1.20. (21) Basal labial palpomere about 3.5 times length of others com- 
bined. (22) Flagellar length about 2.2 times length of scape, 1.34:0.61. 

Mesosoma. (25) Scutal disc with punctures as in verbenae except finer and 
shallower. Punctures immediately below episternal scrobe round to ovate, 1 
pwa or slightly more; anterior to scrobe as in verbenae. Punctures of pro- 
podeal flats as in verbenae except finer, mostly 1 pwa. (27) Legs with cream 
color the same as on face. Foreleg with cream color on knee. Front basitarsus 
broader and shorter than in verbenae, length/width from 3.5 to 4.0. Middle 
mediotarsus fifth its length longer than front mediotarsus, and tenth its 
length longer than hind mediotarsus. Front distitarsus about fourth shorter 


884 Tue Universiry SciENcE BULLETIN 


than middle distitarsus, which is subequal to hind distitarsus. (28) Middle 
leg dark. (30) Tegula as in verbenae. Humeral plate lighter than tegula and 
lighter than in verbenae. (31) Stigma about 3.5 times as long as wide. (32) 
Marginal cell 6-9 longer than, and 3-4 shorter than (to subequal to) 9-wt, 
210397099: 

Metasoma. (34) Band of tergum 1 absent, of terga 2-3 present (but much 
sparser than in verbenae), of tergum 4 denser than on 3. (35) Tergum 1 
with punctures of median area much finer than scutal punctures, 3-6 pwa, 
becoming scattered laterad and posteriad. Declivity of tergum 1 not sharply 
distinct from dorsal portion, shiny, punctures ultra fine, less than 10 » in 
diameter. Raised dorsal surface at anterior edge of posterior depressed margin 
of tergum 1 absent medially, entire median portion sloping anteriorly down- 
ward with central portion concave. (36) Tergum 2 with punctures of median 
area dense, regularly spaced, 1-2 pwa, punctures scattered laterad. Tergum 2 
with median area a gently sloping bulge. Pygidial plate similar to verbenae 
but sides forming a smaller angle, about 40°, and tip more narrowly rounded. 
(37) Sternum 6 with tiny, round median clear area in apical sclerotized plate, 
diameter 0.09-0.12, length of area about 0.20-0.33 length of sclerotized plate. 


Mate. Length, 7.0 mm; forewing length, 5.2 mm; hindwing length, 3.6 
mm; clypeal length, 0.48 mm; scutal length, 1.41 mm. 

Head. Cream colored areas: (1,2,4) as in verbenae; (3) labrum except 
brown in dorsolateral corners (frequently entire basal half); (6) mandible 
basal third only. (10) Punctures coarser, more abundant, and more regularly 
spaced than in female or in verbenae; punctures below middle ocellus larger, 
much denser than in verbenae, 1 pwa or less, subequal in size to those of sub- 
antennal plate. (13) Orbital convergence ratio as 1.60:1.12,1.42. (14) Galea 
shiny, unsculptured (30%), but with barely detectable sculpture at 45%; 
galeal gap absent, tip of galea in repose extending slightly beyond base of 
prementum. (15) Head width to head length as 2.33:1.72,1.36. (17) Eye 
length, mio, and flagellar length as 1.22:1.12:1.80. (18) Interocellar, ocellocu- 
lar, antennocular, and interantennal as 0.43:0.46:0.26:0.32. (19) Ocellolabral 
ereater than clypeal width, 1.36:1.19,1.14. (20) Clypeocellar to outer suban- 
tennal sutural as 0.88:0.71,1.24. (22) Flagellar length about 3.5 times length 
of scape, 1.80:0.51. 

Mesosoma. (25) Punctation as in female except scutal disc with pune- 
tures larger, denser than in either female or verbenae, 1 pwa or less between 
anterior ends of parapsidal lines, punctures immediately below episternal 
scrobe about 2 pwa. Punctures of propodeal flats finer than in verbenae, 1 
pwa or more. (27) Legs with light color in part different from that on face. 
Foreleg with yellow stripe on tibia interrupted preapically; basitarsus pale 
whitish on anterior aspect, apicotarsus straw color. (28) Middle leg with only 
faint indication of yellow on dorsal surface of femoral apex and with (or 


A REVISION OF THE BEE GENUS 885 


without) light color on knee; basitarsus pale amber to light brown, apico- 
tarsus light brown; lengths of tibia, basitarsus and apicotarsus as 0.94 :0.73:0.75. 
(29) Hind leg entirely light brown. (30) As in female. (31) Wing 12-13/ 
13-14 about 10. Stigma about 4 times as long as wide. (32) Marginal cell 6-9 
and 3-4 longer than 9-wt, 1.16:0.99:0.94. 

Metasoma. (34) Hair bands of terga 1 and 4 obsolete, of terga 2-3 very 
sparse. Suberect hair of disc of tergum 5 shorter than in verbenae. (35) 
Tergum 1 with punctures of median area extremely fine, about third of 
diameter of scutal punctures, evenly spaced on dorsum, 3 pwa. Declivity of 
tergum 1 similar to female. (36) Tergum 2 with punctures of median area 
larger than on tergum 1. Pygidial plate with black upturned margin, nar- 
rowly rounded at tip. (38) Penis longer in comparison to penis valve than in 
verbenae. 

Type Mareriat. Holotype female, U. S. N. M. No. 5237, from Lincoln, 
Nebraska, July 7, 1901 (J. C. Crawford, No. 427), on Verbena, one female 
paratype, same data except Aug. 4, 1901 (No. 527) and two male paratypes, 
same data as holotype but Crawford No. 425 on one, no Crawford No. on the 
other, are at the U. S. National Museum. One male paratype, same data as 
holotype except Crawford No. 420, is at the Academy of Natural Sciences of 
Philadelphia, and one male paratype, same data as holotype except Crawford 
No. 423, is at the University of California, Riverside. Description of the 
female is principally based on the female paratype specimen listed above, 
while that of the male is based on the Crawford paratype No. 420. 

Distrisution. North Dakota south to Colorado and Kansas, east to 
Arkansas and northeast following the Prairie Peninsula and Great Lake states 
(if the range is continuous) all the way to northern New Jersey. 

In addition to the type material, 150 males and 65 females were examined from the following 
localities which include the type: ARKANsAs: Marion Co. Cotorapo: Wray, 3700 ft. ILLrNots: 
Adams Co.; Carlinville; Chicago; Grafton; Havana; Havana, Devil’s Hole; Meredosia (Sand Pit); 
Moline. Iowa: Ames; Boone (13 mi. N.W.); Gilbert (4 mi. E).; Iowa Co.; Johnson Co.; Sioux 
City. Kansas: Baldwin; Douglas Co.; Lawrence; Linn; Washington Co.; Riley Co.; Sherman Co. 
Micuican: Ionia Co.; Lapeer Co. MinNesora: Browns Valley; Fridley Sand Dunes, Anoka Co.; 
Hennepin Co.; Princeton; Sargents Bluff; Sucker Lake, Ramsey Co. Missourt: St. Louis. NE- 
BRAsKA: Halsey; Hamlet; Lincoln; Louisville; Monroe Canon, Sioux Co.; Neligh; Northport (2 
mi. E.); Omaha; Valentine (27 mi. S.), Cherry Co. New Jersey: Closter, Bergen Co. NortH 
Daxora: Valley City. SourH Daxora: Hot Springs; Philip; Rapid Canyon; Whitewood. Wiscon- 
sin: Genoa, Vernon Co.; Maiden Rock, Pierce Co.; Milwaukee Co.; Prescott, Pierce Co.; Rutledge, 
Grant Co.; Waukesha Co.; Wyalusing, Grant Co. 

Remarks. The specimens at hand are homogeneous for obvious external 
characters. Although specimens from North Dakota and Minnesota are 
larger than those from more southern localities, the difference is not statis- 
tically significant. The easternmost specimens, from northeastern New Jer- 
sey, differ from others in the combination of piceous integument and darker 

colors in all areas; more deeply fulvous hair of vertex, thoracic dorsum and 
abdominal bands; and smaller head length and abdominal width despite an 


S86 Tue University SCIENCE BULLETIN 


average length hindwing. The affinities of nebraskensis seem to be mostly | 
with verbenae, yet the striking similarity in facial dimensions and ratios which 
groups nebraskensis with hirsutifrons and verbenae with micheneri suggests 
that nebraskensis is not a lineal descendant from either verbenae or hirsuti- 
frons, and that its ancestor is some extinct or uncollected form. 

Bronomics. Most of the flower records are for Verbena species; one each 
for Ambrosia and Asclepias are probably accidental. I have seen mixed pollen _ 
loads which incorporated pollen of Verbenaceae and Leguminosae. Robertson 
regarded this species as oligolectic on Verbena spp. and (1914) described the 
action of the females in obtaining pollen of Verbena: “Anthers of Verbena 
are included in a slender tube and above them is a circle of hairs. ... Ordinary 
bees can only collect pollen which adheres to their tongues. Verbenapis — 
verbenae | misidentification for nebraskensis | has its front tarsi provided with 
curled bristles. When collecting pollen the bee thrusts both front legs into the 
tube of the corolla and drags out the pollen with its front tarsi.” Examination 
of specimens supports this observation, for pollen was often found packed 
between the curled hairs of the front tarsi. Robertson (1929) records the 
species at Carlinville, Illinois, as active from June 28 to September 11 which | 
gives the flight period of the bee as 75 days, or well within the period of 
anthesis of Verbena from May 22 to October 5. Both sexes were taken by 
Robertson (1922) on Verbena hastata, V. stricta, and V. urticifolia (in copulo 
on the last two). 

Robertson (op. cit.) found Calliopsis (Calliopsis) andreniformis visiting 
Verbenaceae, too: Lippia lanceolata and Verbena bracteosa, neither of which — 
were visited by nebraskensis. At Lawrence, Kansas, C. andreniformis males | 
visit Verbena sp., and at Nacogdoches, Texas, Verbena tenutsecta, but I have — 
not seen females at Verbena at either locality. 

Rau and Rau (1916) and Rau (1922) give notes on the nesting of 
nebraskensis at St. Louis, Missouri; it is similar to that of andreniformts. 
Hard-packed bare soil is selected for nest construction, e.g., a baseball dia- 
mond! A burrow is dug with a single or double entrance, in the latter case, 
the openings about two inches apart, and kept covered by a mound of “fine 
dust.” Mating was quietly effected on the ground and copulation lasted per- 
haps five minutes. The same sort of buzzing, tumbling, and whirling in small 
circles by pairs on the dusty ground was seen for nebraskensis as 1 have seen 
with pairs of male andreniformis. Cockerell has observed similar behavior by 
males of Nomadopsis zebrata and mating pairs of Calliopsis coloratipes, and 
Rozen (1958) has noted it for Nomadopsis a. anthidius and N. michenert. In 
the last instance, Rozen found mating pairs rather than males alone. The 
Raus state that this tumbling did not immediately precede or follow mating, | 
but they were not able to catch pairs to determine sexes. Sphecodes spp. enter 
the burrows freely and are suspect, but unproven, parasites. Schecodes 


A REVISION OF THE BEE GENUS 887 


brachycephalus Mitchell and Sphecodes sp. are discussed as possible parasites 
of Calliopsts andreniformis in the section on the biology of that species. 

Frower Recorps. Literature and label records include Ambrosia and 
Asclepias, both of which are probably accidental, Medicago sativa, Verbena 
hastata, V. stricta, V. urticifolia, Vernonia. 


CALLIOPSIS (VERBENAPIS) HIRSUTIFRONS Cockerell 
(Figs. 134-138; Map 8) 


Calliopsis hirsutifrons Cockerell, 1896, Canad. Ent., 28:158; 1897, Bull. Univ. New Mexico, 

24:19; Michener, 1951, 72 Muesebeck e¢ al., U.S. Dept. Agric., Monogr. No. 2:1103. 

Readily distinguished from other Verbenapis by the small size. It is 
closest to nebraskensis on the basis of numerous head dimensions and ratios 
as well as the nature of the clear, median disc of the metasomal sternum 6 of 
the female. 

It has a punctation pattern similar to that of verbenae and michenert, and 
the shape of the metasomal terga is closer to them than to nebraskensis. The 
concave nature of the medial basal margin of metasomal sternum 8 distin- 
guishes the male from other Verbenapis. In the female the first labial palpo- 
mere is a fourth to a third shorter than the flagellar length, but in the other 
species it is subequal to or up to a third longer than the flagellar length. The 
female differs from nebraskensis in having the base of the middle tibia yellow, 
and from verbenae in the more extensive cream colored area on the mandible, 
and from both in the shape of the clypeal cream colored area as described in 


(2) below. 


Femate. Length, 6.5 mm; forewing length, 4.5 mm; hindwing length, 3.1 
mm; clypeal length, 0.44 mm; scutal length, 1.19 mm. 

Head. Cream colored areas: (1) paraocular area similar to verbenae and 
nebraskensis but dorsal apex of triangular patch ends at or above level of 
lower rim of antennal socket (may reach to level of lower border of facial 
fovea); (2) clypeus, the median portion of cream colored area produced 
dorsally to within half mow of frontoclypeal suture, the lateral portion ex- 
tended to within less than 1 mow of anterior articulation of mandible; (3) 
labrum entirely or labral plate only; (4) absent on supraclypeal area (small 
dot may occur in impunctate lower median part adjacent to frontoclypeal 
suture); (6) mandible basal fourth. (10) Punctures of ocellar triangle fine, 
distinct, 1 pwa or more; punctures immediately posteriad larger, less than 1 
pwa. (13) Orbital convergence ratio as 1.39:1.22,1.14. Lower median border 
of supraclypeal area at same level as adjacent clypeus. (14) Galea unsculp- 
tured; galeal gap subequal to mow. (15) Head width to head length as 2.14: 
1.53,1.40(1.39-1.45). (16) Eye width in profile about 6.5 times protrusion of 
clypeus beyond it. (17) Eye length, mio, and flagellar length as 1.14:1.22:1.24. 


888 Tue University SCIENCE BULLETIN 


(18) Interocellar, ocellocular, antennocular, and interantennal as 0.41:0.39: 
0.32:0.36. (19) Ocellolabral greater than clypeal width, 1.29:1.19,1.08. (20) 
Clypeocellar to outer subantennal sutural as 0.85:0.65,1.31. (21) Basal Jabial 
palpomere about 3.2 times length of others combined. (22) Flagellar length 
about 2.4 times length of scape, 124-0518 

Mesosoma. (25) Scutal disc with punctures between anterior ends of 
parapsidal lines distinct, mostly large, about 1 pwa. Punctures immediately 
below episternal scrobe deep, round, 1 pwa; anterior to scrobe, round, less 
than 1 pwa. Punctures of propodeal flats as in verbenae but finer, less than 1 
pwa. (27) Legs with cream color the same as on face. Foreleg with cream 
color on knee. Front basitarsus long, narrow, length/width about 8.0 to 9.4. 
Front, middle, and hind mediotarsi as in nebraskensis. Front distitarsus third 
shorter than middle distitarsus which is subequal to hind distitarsus. (28,30) 
As in verbenae. (31) Stigma about 3 times as long as wide. (32) Marginal 
cell 6-9 longer than, and 3-4 shorter than 9-wt, 1.02:0.82:0.87. 

Metasoma. (34) Bands of terga 1-4 quite dense, least so on tergum Il. 
(35,36) Tergal punctures finer than in verbenae, tergum 1 with punctures of 
median area more scattered, but punctation pattern similar. Pygidial plate 
similar to nebraskensis. (37) Sternum 6 with round, median clear area 1n 
apical sclerotized plate, diameter of area less than mow (about 0.15:0.17), and 
about third of length of sclerotized plate. 


Mate. Length, 6.0 mm; forewing length, 4.3 mm; hindwing length, 2.9 
mm; clypeal length, 0.41 mm; scutal length, 1.10 mm. 

Head. Cream colored areas: (1,3) As in verbenae; (2) clypeus except for 
narrow dark strip below subantennal plate; (4) supraclypeal area, a medial 
round spot touching frontoclypeal suture; (6) mandible basal third. (10) 
Punctures similar to those of female but different in manner described for 
male verbenae. (13) Orbital convergence ratio as 1.36:0.97,1.40. (14) As in 
female. (15) Head width to head length as 1.99:1.43,1.40. (17) Eye length, 
mio, and flagellar length as 1.02:0.97:1.72. (18) Interocellar, ocellocular, an- 
tennocular, and interantennal as 0.36:0.41:0.20:0.32. (19) Ocellolabral greater 
than clypeal width, 1.17:0.97,1.21. (20) Clypeocellar to outer subantennal 
sutural as 0.77:0.61,1.25. (22) Flagellar length about 4.0 times length of scape, 
72043. 

Mesosoma. (25) Punctation as in female. (27) Legs with light color in 
part different from that on face. Foreleg as in verbenae. (28) Middle leg 
colored like verbenae but tip of tibia cream colored, and apicotarsus dark 
brown. (29) Hind leg as in verbenae except apicotarsus dark brown. (30) 
As in female. (31) Wing 12-13/13-14 about 8.5. Stigma about 3.5 times as 
long as wide. (32) Marginal cell 6-9 and 3-4 longer than 9-wt, 0.94:0.85 :0.80. 

Metasoma. (34) Hair bands of terga 1-3 distinct, sparse medially, of 
tergum 4 obsolete. Suberect hair of disc of tergum 5 shorter than in verbenae. 


A ReEvIsION OF THE BEE GENUS 889 


(35) Tergum 1 with punctures of median area larger and deeper than scutal 
punctures, evenly spaced on dorsum, 1-2 pwa. Declivity of tergum 1 similar 
to female. (36) Tergum 2 with punctures of median area as in verbenae. 
Pygidial plate usually darker peripherally with an irregular submarginal line 
of coarse, oblong punctures. 

Type Materia. Holotype male, U.S. N. M. No. 5821, from Albuquerque, 
New Mexico, middle Aug., 1895 (T. D. A. Cockerell, No. 4527), is at the 
U.S. National Museum. The description of the female is principally based on 
a specimen from Garfield, New Mexico, July 16, 1952 (R. H. and L. D. 
Beamer, W. E. La Berge, and Cheng Liang), while that of the male is based 
on a specimen from Hot Springs, New Mexico, July 22, 1950 (R. H. Beamer). 

Distrisution. North central New Mexico, east of the Rio Grande River, 
to El Paso, Texas, south across the Mexican desert to the state of México. 


In addition to the type material, 24 males and 9 females have been examined from the locali- 
ties listed below which include the type: New Mexico: Albuquerque, middle Aug.; Garfield, July 
16; Hot Springs, July 22; Moriarty, June 24; Radium Springs, July 16. Texas: El Paso, July 11. 
CuinuAnua: Villa Ahumada, 5700 ft., July 28. Duranco: San Juan del Rio, 5200 ft., July 30. 
Guanajuato: Leon (1 mi. N.W.), Aug. 9. Hipatco: Ixmiquilpan, 5200 ft., July 29; Lagunilla, 
June 14. Mexico: Tepexpan, 6500 ft., Aug. 12. 


Remarks. A female specimen from Leén, Guanajuato, has clypeus with 
only a lateral cream spot, labrum and mandible completely brown. A male 
specimen from San Juan del Rio, Durango, is the largest specimen and ex- 
hibits other differences in structure that may indicate a population worthy of 
subspecific recognition. Unfortunately the sterna and genitalia are not in good 
enough condition to make positive statements. The genital capsule is flatter, 
penis valve rather broader distally, volsella more sharply pointed, and penis 
slightly shorter in comparison to the penis valve than in other specimens. No 
biological data are available for the species. 


CALLIOPSIS (VERBENAPIS) MICHENERI, new species 
(Figs. 139-143; Map 8) 


This large, distinctive species is named in honor of Dr. Charles D. 
Michener who discovered it and who has done much to aid in the present 
understanding of this genus. The species is distinguished from others of the 
subgenus by the occurrence of a triangular area of opaque cream coloration 
on the outer apical half of the tegula and by the high ratio of the first labial 
palpomere to the length of the others combined, usually about 5. The female 
is separated from verbenae by having finer punctures with wider interspaces, 
and by the basal outer half of the mandible being cream colored. The male 
is readily known by the large size and the pointed, narrow pygidial plate. 


Femace. Length, 8.3 mm; forewing length, 6.0 mm; hindwing length, 4.1 
mm; clypeal length, 0.53 mm; scutal length, 1.44 mm. 
Head. Cream colored areas: (1) paraocular area as in Airsutifrons but 


890 THe UNIVERSITY SCIENCE BULLETIN 


triangle less extensive, ending dorsally at (or slightly above) level of upper 
portion of frontoclypeal suture, well below level of lower rim of antennal 
socket; (2) clypeus as in Airsutifrons except lateral cream color extends to 
within half (or more) mow of anterior articulation of mandible; (3,4) as in 
hirsutifrons; (6) mandible basal half (to third). (10) Punctation similar to 
verbenae but finer. Lower dark paraocular area with punctures distinct, finer 
than those of subantennal plate, 2-3 pwa. (13) Orbital convergence ratio as 
1.60:1.41,1.13. Lower median border of supraclypeal area below level of ad- 
jacent clypeus. (14) Galea finely, distinctly pebbled on entire anterior surface 
(30); galeal gap absent, tip of galea in repose extending slightly beyond 
base of prementum. (15) Head width to head length as 2.47:1.80,1.37 (1.34- 
1.37). (16) Eye width in profile about 2.5 times protrusion of clypeus beyond 
it. (17) Eye length, mio, and flagellar length as 1.34:1.44:1.33. (18) Intero- 
cellar, ocellocular, antennocular, and interantennal as 0.44:0.46:0.32:0.44. (19) 
Ocellolabral less than clypeal width, 1.46:1.53,0.96. (20) Clypeocellar to outer 
subantennal sutural as 0.94:0.82,1.14. (21) Basal labial palpomere 3.5 (3.555) 
times length of others combined. (22) Flagellar length about 2.3 times 
length of scape, 1.33:0.58. 

Mesosoma. (25) Scutal disc with punctures between anterior ends of 
parapsidal lines uniform, fine, 2 pwa. Punctures immediately below episternal 
scrobe round, 1.5-2.5 pwa; anterior to scrobe round, 1 pwa. Punctures of pro- 
podeal flats round, finer than in others of the subgenus, subequal in size to 
scutal punctures, about 2 pwa. (27) Legs with light color the same as on face. 
Foreleg colored as in Airsutifrons. Front, middle, and hind mediotarsi and 
distitarsi as in verbenae. (28) As in verbenae. (30) Tegula transparent 
smoky tan with anterior patch of opaque cream coloration; humeral plate 
with anterior half cream colored, posterior half brown. (31) Stigma about 4 
times as long as wide. (32) Marginal cell 6-9 longer than, and 3-4 shorter 
than 9-wt, 1.22:0.94:1.19. 

Metasoma. (34) About as in verbenae. (35) Tergum 1 with punctures of 
median area almost as fine as in Airsutifrons, diameter subequal to scutellar 
punctures, tergum generously covered including posterolateral areas, punc- 
tures 1-3 pwa. (36) Tergum 2 with punctures of median area as in verbenae 
but 1-2 pwa anteromedially. Pygidial plate similar to Airsatifrons but more 
slender. (37) Sternum 6 with oval, median clear area in apical sclerotized 
plate, this area wider than mow, wider than long, about 0.24:0.19, and about 
third of length of sclerotized plate. 


Mate. Length, 8.8 mm; forewing length, 6.1 mm; hindwing length, 4.1 
mm; clypeal length, 0.60 mm; scutal length, 1.59 mm. 

Head. Cream colored areas: (1,2,3,4) as in Airsutifrons; (6) mandible as 
in nebraskensis. (10) Punctures similar to those of female but different in the 
manner as described for verbenae. (13) Orbital convergence ratio as 1.77:1.48, 


A Revision OF THE BEE GENUS 891 


1.19. (14) Galea lightly, distinctly roughened as in female; galeal gap absent, 
tip of galea in repose extending beyond base of prementum by length of 
second labial palpomere (or more). (15) Head width to head length as 2.70: 
2.01,1.35. (17) Eye length, mio, and flagellar length as 1.45:1.48:1.87. (18) 
Interocellar, ocellocular, antennocular, and interantennal as 0.49 :0.49 :0.32:0.48. 
(19) Ocellolabral less than clypeal width, 1.36:1.58,0.86. (20) Clypeocellar 
subequal to outer subantennal sutural, 0.94:0.95. (22) Flagellar length about 
3.6 times length of scape, 1.87:0.53. 

Mesosoma. (25) Punctation as in female. (27) Legs with light color the 
same as on face except tibial stripe yellowish. Foreleg with color as in 
nebraskensis except trochanter as in verbenae, knee cream colored, tarsus 
brown. (28) Middle leg with trochanter as in verbenae, knee cream colored, 
tarsus brown. (29) In a few specimens, hind leg with medial portion of basi- 
tibial plate whitish with black (to dark brown) border. (30) As in female. 
(31) Wing 12-13/13-14 about 5.5. Stigma about 5 times as long as wide. (32) 
Marginal cell 6-9 longer than, and 3-4 shorter than 9-wt, 1.26:1.00:1.17. 

Metasoma. (34) Hairiest of the subgenus. Hair bands of terga 1-4 dis- 
tinct, sparse. Suberect hair of disc of tergum 5 denser and much longer than 
other species in the subgenus. (35) Tergum 1 with punctures of median area 
subequal to scutal punctures, deeper and finer than in /ursutifrons or verbenae, 
1-2 pwa. Declivity of tergum 1 similar to female. (36) Tergum 2 with punc- 
tures of median area as described for verbenae. Pygidial plate twice as long 
as broad at base with extremely narrow, black, slightly upturned border; sides 
forming 25° (to 30°) angle, tip very narrowly rounded. (37) Medial lobes of 
sternum 6 separated by a deep cleft, depth about twice (or more) width of 
one lobe. (38) Penis valve rounded at apex with distinct subapical pointed 
protuberance. 

Tyre Mareriav. Holotype male and allotype female from Carrizo Springs, 
Dimmit Co., Texas, April 14, 1949 (C. D. Michener and R. H. Beamer), on 
Verbena cloveri, are in the Snow Entomological Museum of The University 
of Kansas, Lawrence. Two female and 25 male paratypes bear the same col- 
lection data; two male paratypes, same locality, were taken March 28, 1946, by 
C. D. Michener who informs me that they were visiting the same species of 
Verbena. 

Remarks. Known only from the type locality in Texas. Its host flower 
is endemic to Texas. 


NOMEN NUDUM ww CALLIOPSIS 
CALLIOPSIS BRIDWELLI Bridwell 
Calliopsis bridwelli Bridwell, 1899, Trans. Kansas Acad. Sci., 16:210. 


This name was published in a list of insects of Kansas and must represent 
a misidentification of a common species. 


892 Tue UNIVERSITY SCIENCE BULLETIN 


NOMEN DUBIUM tw CALLIOPSIS 
CALLIOPSIS FLAVIFRONS Smith 


Calliopsis flavifrons Smith, 1853, Catalogue of Hymenoptera in the British Museum, 1:129, male; 
Cresson, 1887, Trans. Amer. Ent. Soc. Suppl. :245; Cockerell, 1897, Bull. Univ. New Mexico, 
24:26; 1898, Trans. Amer. Ent. Soc., 25:197; 1905, Trans. Amer. Ent. Soc., 31:321, female; 
Sandhouse, 1943, Proc. U.S. Nat. Mus., 92:531 (=Camptopocum prinit Holmberg, 1884) 
(misidentification); Michener, 1951, 72 Muesebeck et al., U.S. Dept. Agric., Monogr. No. 
2:1103; Mitchell, 1960, North Carolina Agric. Exp. Sta. Tech. Bull. No. 141:293. 


Smith’s description of the male is given below. 


“Mare. Length 4 lines [=8.47 mm]. Black, the face, scape of the anten- 
nae in front*, labrum and mandibles yellow, the latter have a black line on 
their inferior margin and are ferruginous at the tips; the flagellum pale tes- 
taceous beneath. Thorax, the disk thinly covered** with pale ochraceous 
pubescence; the anterior and intermediate tibiae in front yellow; all the tibiae 
and tarsi have a pale glittering pubescence; the apical joints of the tarsi ferru- 
ginous; wings subhyaline, iridescent, nervures fuscous, the tegulae testaceous. 
Abdomen short and somewhat globose, delicately punctured, the margins of 
the segments have on each side a short fascia of white pubescence. 

“Obs. The fascia on the abdomen probably in very recently disclosed 
specimens would be entire. Hab. East Florida. Coll. of F. Smith. [Collector 
E. Doubleday, Esq. |” 

Cockerell (1905) described a female in the British museum collection 
bearing the flavifrons type label. The specimen he described was undoubtedly 
C. andreniformis Smith. The difference in sex shows that he did not have 
flavifrons before him. 

At least three thoroughly competent bee specialists (Cockerell, Mitchell, 
and Moure) have failed to find the type of this species in the British Museum 
or among the Smith types at the Hope Museum in Oxford. 

The original description is inadequate to determine the species, and 
flavifrons must remain a nomen dubium. 


SPECIES ERRONEOUSLY ASSIGNED TO CALLIOPSIS 


Perhaps a score of species occurring outside the United States and Canada 
have been described as Calliopsis. All but two have subsequently been trans- 
ferred to other genera, mostly to related panurgine genera. The following 
two species are here asisgned their proper genera. 


*When compared to Smith’s description of Calliopsis flavipes this character is a clue that 
Smith described either the form of C. coloradensis Cresson from the southeastern United States, or 
an exceptional specimen of Acamptopoeum accidently introduced to Florida by man. Italics mine. 

** The disc of the thorax is thinly covered with pale ochreous pubescence in the specimens of 
C. coloradensis from the southeastern United States but all known species of Acamptopoeum have 
the disc of the thorax densely covered. 


A Revision oF THE BEE GENUS 893 


I am indebted to Padre J. S. Moure, as mentioned elsewhere, for his notes 
and measurements on the first species, and also for advising me of the correct 
placement of the second species based on his examination of the type 
specimen. 


ACAMPTOPOEUM MACULATUM (Smith) new combination 
(Map 1) 


Calliopsis maculatus Smith, 1853, Catalogue of Hymenoptera in the British Museum, 1:129, male; 
Cockerell, 1889, Trans. Amer. Ent. Soc., 25:196; Cockerell, 1901, Ann. Mag. Nat. Hist., 
(7)7:129 (transferred to Spinoliella); Cockerell, 1905, Trans. Amer. Ent. Soc., 31:321, fe- 
male; Michener, 1951, 72 Muesebeck e¢ al., U.S. Dept. Agric., Monogr. No. 2:1103; Mitchell, 
1960, North Carolina Agric. Exp. Sta. Tech. Bull. No. 141:293. 


The description of the female type by Smith is given below. 


“FemMace. Length 4 lines [=8.47 mm.]. Black, the sides of the clypeus, a 
coronet-shaped spot above, on each side of it a minute lunate spot, and another 
at the base of the mandibles, yellow; the flagellum fulvo-testaceous beneath. 
Thorax, the disk has a fulvous pubescence, on the metathorax and beneath it 
is a griseous; the wings hyaline, iridescent, their apical margins having a slight 
fuscous cloud, the tegulae testaceous; the tibiae and tarsi above have a glitter- 
ing pale yellow pubescence, on the tarsi beneath it is bright fulvous, the claw- 
joints ferruginous; the calcariae [tibial spurs | ferruginous, the extreme base 
of the anterior and intermediate tibiae yellow. Abdomen ovate, the margins 
of the segments narrowly testaceous, and having a marginal fascia of pale 
ochraceous pubescence, on the two basal segments usually nearly obliterated. 
Hab. East Florida. Coll. of F. Smith.” [Collected by Edward Doubleday, 
Esq., in 1838. ] 


Supplementary description and measurements by Professors Padre J. S. 
Moure, T. B. Mitchell (1960), and Charles D. Michener (1965) are given 
below. 


Femate. Length, 9.2 mm; forewing length, 7.4 mm; clypeal length, 0.65 
mm. 

Head. Yellow areas: (1) paraocular area a very small mark in lowermost 
corner and short, very narrow stripe adjacent to frontoclypeal suture running 
ventrolaterad from anterior tentorial pit; (2) clypeus on each side laterad 
from labral notch, dividing it about equally into thirds, middle third black; 
(3) absent on labrum; (4) supraclypeal area a spot; (5) subantennal plate a 
small spot; (6) mandible a basal spot. (7) Scape and pedicel black; flagello- 
mere | with a vague yellowish mark on ventral (frontal) surface. (10) Punc- 
tures of frons and gena exceedingly minute, barely visible (30>), interspaces 
shiny; (11) Frontal line with lower portion a sharp raised carina. (12a) 


894 Tue University SclENcE BULLETIN 


Clypeus with punctures of disc deep, distinct, close, interspaces shiny. (12b)* 
Labrum with area basad from high premarginal carina punctate, area apicad 
to carina smooth, shiny, impunctate. (13a)* Inner orbits only slightly con- 
vergent below, orbital convergence ratio as 2.02:1.97,1.03. (13b) Facial fovea 
shallow, broad medially, narrowed almost to acute angle above and below. 
(15) Head width to head length as 3.20:2.18,1.47; abdominal width 3.25. 
(17)* Eye length, mio, and flagellar length as 1.60:1.97:1.76. (18) Interocellar, 
ocellocular, antennocular, interantennal, and antennocellar as 0.51 :0.61:0.54: 
0.44:0.71. Antennocular subequal to width of eye in frontal view. (19)* 
Ocellolabral shorter than clypeal width, 1.73:1.98,0.87. (20) Clypeocellar to 
outer subantennal sutural as 1.09:0.44,2.46. Outer subantennal sutural :inner 
subantennal sutural:width of subantennal plate as 0.44:0.22:0:31. Clypeal 
length 0.65. (22) Lengths of flagellomeres 1-4 and 10 as 0.20:0.14:0.16:0.16: 
0.31. Flagellar width less than mow, 0.20:0.22. *Flagellar length about 2.1 
times length of scape, 1.76:0.84. 

Mesosoma. (23) Light areas absent. (24) Hair of dorsum abundant, 
rather short, and whitish. (25) Scutal and scutellar punctures very fine, close, 
deep and distinct, interspaces shiny. (26) Dorsal enclosure of propodeum 
slightly longer than mow, 0.26:0.22, surface unsculptured, not excavated, 
highly polished. (27) Foreleg with yellow spot on distal end of femur and 
base of tibia. (28) Middle leg with two yellow spots on femur and one at 
base of tibia. Lengths of tibia, basitarsus, and spur as 1.43:1.12:1.12 (spur as 
long as basitarsus!). Spur with several coarse teeth on apical two-thirds. (29) 
Hind leg dark. (31) Stigma 4.33 times as long as broad (includes costal vein), 
0).88:0.20. (32) Marginal cell 6-9 4 times as long as broad, 1.73:0.43, and 
greatly exceeding 9-wt, 1.31. Prestigmal length 5-2,0.34. Wing 10-11/11-14 as 
1.00:1.14; 11-12/13-14 as 0.22:0.14. Hindwing with 11 hamuli. 

Metasoma. (34) Suberect hair of disc of tergum 5 dark in part. (35,36) 
Terga 1 and 2 with punctures of median area minute, close, interspaces shiny. 

Tyre MareriaL. Holotype female, No. 17.a.1800, from East Florida, col- 
lected by E. Doubleday in 1838, is in the British Museum (Natural History). 

Distrisution. If this specimen was really captured in East Florida, it is 
presumably because it had been recently introduced there by man’s activities. 
No other specimens of Acamptopoeum have ever been taken north of Colom- 
bia, South America. 

Discussion. The specimen is unquestionably a panurgine bee closely 
related to Calliopsis, and the list below summarizes certain key characteristics 
found in maculatum as to their occurrence in species of the listed genera and 
constitutes the evidence for my transfer of maculatum to the genus Acampto- 
poeum. The numbers refer to characters indicated in the description above; 


* Measurements kindly made by Prof. Charles D. Michener, July 1965. 


A ReEvIsION OF THE BEE GENUS 895 


+ means that the character of maculatum occurs in the genus; — means that 
the character does not occur in the genus. 


Number Calliopsis Acamptopoeum — Hypomacrotera Liopoeum 

-- + + —- 

2 + + -- ~- 
10 — a — 
11 — f — — 
12b == ak. ea 3 
13b — + — — 
24 = a = = 
28 — oo -- — 


Items 12b and 28 seem absolutely diagnostic for Acamptopoeum. Item 28 
shows a mesotibial spur as long as the middle basitarsus and coarsely toothed 
on the apical two-thirds. In Calliopsis only squamifera has a spur almost as 
long as the basitarsus, and it is completely bare (30). 


BICOLLETES SPEGAZZINI (Jorgensen) new combination 
Calliopsis spegazzini Jorgensen, 1912, Zool. Jahrb., Syst., 32:121. 


Padre J. S. Moure has determined the correct generic assignment from an 
examination of the type specimen. 


BIOLOGY AND ECOLOGY OF CALLIOPSIS ANDRENIFORMIS 


Intropuction. This work on Calliopsis andreniformis Smith is meant to 
furnish a frame of reference for comparative studies in the biology and ecology 
of other species of Calliopsis and other burrowing bees. This is the most 
widespread and commonest member of the genus in the United States and 
Canada and is an apparently effective pollinator of many wild plants as well 
as alfalfa and various cultivated clovers. 

Most of the observations were made on the campus of The University of 
Kansas at Lawrence, Kansas, with additional ones on the campuses of the 
public schools and of Stephen F. Austin State College at Nacogdoches, Texas; 
at Hancocks Bridge, New Jersey; and at Knoxville, Tennessee. 

The Campus Bee has been proposed in the generic revision as the common 
name for this species because of its ubiquity on campuses of colleges and 
schools. Its closest relatives are C. teucrit and C. hondurasica, whose nest 
burrows are shown with those of C. andreniformis in Figs. 148-153. 

Specimens documenting this study will be deposited in the Snow Entomo- 
logical Museum of The University of Kansas, Lawrence, Kansas. 

Techniques used in the study were in part obtained from Michener, Cross, 
Daly, Rettenmeyer, and Wille (1955). Weston dial thermometers were used 


THe Universiry SciENCE BULLETIN 


896 


genital capsule (also lateral for | 


and 5 (also 7 for andreniformis). 8-13. C. 


C. granti. 22-25. C. rhodophila. 


ventral view of sterna 8, 6, 


Male terminalia. Subgenus Calliopsis. Dorsal view of 
andreniformis), 
el 


Fics. 8-25. 


18-21. 


C. teucri. 


andreniformis. 


897 


A REVISION OF THE BEE GENUS 


0.5 mm 


= WN 
us 
5S 
ae 
V9 
rs 
4S) 
i] . 
nN 
ale 
art 
Qn 
a 
9 . 
Pris} 
£3 
qs 
bo. 
aS} 
wt 
O98 
SS 
eae 
20 
= 
20 
3° 
qm 


Subgenus Calliopsis. 


view of sterna 8, 6, and 5. 26-29. C. mouret. 


38-41. C. empelia. 42-45. C. squamifera. 


Fics. 26-45. Male terminalia. 


Tue University SCIENCE BULLETIN 


898 


0.5 mm 


Perissander. Dorsal view of genital capsule and ventral 


Subgenus 


Aale terminalia. 


6-57. N 
view of sterna 8, 6, 


4 
I 


Fics. 


C. gilva. 


and 5. 46-49. C. anomoptera. 50-53. C. rogeri. 54-57. 


899 


A ReEvIsIONn oF THE BEE GENUS 


Fics. 58-76. Male terminalia. Subgenus Calliopsima. Ventral view of genital capsule (also lateral 


view for pectidis and timberlake1) and ventral view of sterna 8, 6, and 5 (also 7 rozent). 
58-62. C. rozent. 63-66. C. coloradensis. 67-71. C. pectidis. 72-76. C. timberlakei. 


900 Tue University ScIENCE BULLETIN 


~ 
(ee, 
“I 
@o 


83 


- 


88 


93 | ) 
94 
0.5 mm 
91 


Fics. 77-94. Male terminalia. Subgenus Calliopsima. Ventral view of genital capsule (also lateral 
view for bernardinensis) and ventral view of sterna 8, 6, and 5 (also 7 for bernardinensis). 
77-82. C. bernardinensis. 83-86. C. unca. 87-90. C. crypta. 91-94. C. azteca. 


901 


A REVISION OF THE BEE GENUS 


3-106. C. deserticola. 


view of sterna 8, 6, and 5. 95-98. C. chlorops. 99-102. C. coloratipes. 10 


Fics. 95-110. Male terminalia. Subgenus Calliopsima. Ventral view of genital capsule and ventral 
107-110. C. pugionis. 


Tue UNIversiTry SCIENCE BULLETIN 


902 


Ventral view of genital capsule and ven- 


119-122. C. 


111-114. C. hurdi. 115-118. C. quadridentata. 


tral view of sterna 8, 6, and 5. 


kucalumea. 


Fics. 111-122. Male terminalia. Subgenus Calliopsima. 


A ReEvIsION OF THE BEE GENUS 903 


< 


123 


. 
nh 


29 
134 135 


0.5 mm 


Fics. 123-143. Male terminalia. Subgenus Verbenapis. Lateral and ventral views of genital cap- 
sule and ventral view of sterna 8, 7, and 6 (also 5 for verbenae only). 123-128. V. verbenae. 
129-133. C. nebraskensis. 134-138. C. hirsutifrons. 139-143. C. micheneri. 


for taking soil temperatures; fine copper screen cones were used for retaining 
bees as they came out of their burrows. Quick-drying acetate paints were used 
for marking individual bees, and roofing nails with yellow-painted heads 
bearing red numbers marked the locations of nest entrances. A string grid 
was fixed permanently in place for use in mapping the location of burrows. 
Excavations of nests were usually made by carving away the clayey soil with 
a sharp blade. 


904 Tue University SCIENCE BULLETIN 


Because of its efficiency in pollinating legumes as well as a wide variety of — 


other plants (see Table 8), Calliopsis andreniformis doubtless has an im- 
portant ecological role in the maintenance of native flora. No quantitative 
assessment of this role has ever been made. Nonetheless, Crandell and Tate 


(1947) report that H. M. Tysdal observed (Unpublished, 1942) that the 


species was primarily responsible for very good seed set in alfalfa at Lincoln, 


Nebraska, over several seasons. They note, however, that C. andreniformis — 


visits alfalfa flowers at a slower rate than Megachile and must be present in 
large numbers to effect a good seed set. Moreover, adults are present for a 


relatively short time, late June and July, and this limits their effectiveness. 


Their utilization in pollination would necessitate regulation of the flowering 


time of alfalfa or other plants involved to coincide with the period of adult | 


activity. The same authors point out that this was not difficult at their nur- 
series in Lincoln, Nebraska, since the alfalfa blooms over a relatively long 
period, 

My observations suggest that andreniformis would be a valuable pollinator 


of Trifolium repens and Trifolium procumbens, as well as alfalfa. Fur- | 


ther discussion on the anthecology of the species appears under Flower 
Relationships. 
The principal nesting habitat for andreniformis is bare, clayey soil, fully 


exposed to the sun, with leguminous plants of the clover group nearby. The — 
farthest from a clover supply that any nesting site was located was about 59 m. 
In a tangled growth of virgin prairie on the campus of The University of | 
Kansas, nests were made at the bases of the tall grasses and legumes in loamy — 


soil. Even here preference was given to those small areas where the plants 
gave the thinnest cover, and the sun penetrated to the surface at least four 
hours daily. I have often found nests of the species in city ball parks or on 
school grounds, and these are the first places to look when in a new locality to 
see if the species occurs there in any numbers. Although the nesting sites 


usually have a slight slope, some very large aggregations have been found in — 
flat fields, newly sown in clover and grass, e.g., football practice fields at The ~ 


University of Kansas and at Stephen F. Austin State College, Nacogdoches, 
Texas. 
The extremes of the flight period for the species are from early April to 


late September in the southern states, with peak activity in June and July; in 


the more northerly states it is from early June to early October, with peak 
activity in July and August. 

A summary of the life history is as follows. The species over-winters as 
prepupae in closed, wax-lined cells in the ground. In spring the bees meta- 
morphose and emerge as the active adults of the overwintering generation. 
Aggregations of males and females segregate by sex in digging their individ- 
ual overnight burrows. In about a week mating is observed and females 


A ReEvIsION OF THE BEE GENUS 905 


begin nest construction by, 1) making a cell at the bottom of the overnight 
burrow; 2) beginning a new shaft near the mid-point of the overnight bur- 
row; or, 3) abandoning the overnight burrow and digging a new one. Fe- 
males typically construct two nests of three or four cells each, provisioning 
each cell with a spherical ball of leguminous pollen and laying one egg atop it. 
In about two weeks these eggs have given rise to the adults of the first sum- 
mer generation. This generation in turn constructs nests and lays eggs, some 
of which give rise to the second generation and some of which enter diapause 
to overwinter as prepupae. Apparently most of the eggs laid by the second 
generation produce young which enter diapause, for the adult population 
drops drastically by the end of July. Individual females, however, lay about 
the same number of eggs, seven, as the females of the overwintering genera- 
tion. A few adults provision nests as late as September. The last of the 
adults dies by early autumn. 

Relatively few data are available on the continuity of nesting sites of C. 
andreniformis. Ainslie (1937) followed the progress of an “exceedingly 
active colony ... in hard trodden ground in front of a grocery store at Sioux 
City, Iowa ... three years under observation and study, then for unknown 
reasons colonies could not be found where they formerly were populous.” 
Crandall and Tate (1947) had an aggregation under study for two consecu- 
tive years. Michener and Rettenmeyer (1956) mention a nesting site for 
andreniformis at Lawrence, Kansas, which apparently persisted from 1949- 
1953, and had not returned by 1959. My observations in Kansas and Texas 
indicate that andreniformis will stay at least three years in the same nesting 
site if flowers are available for pollen. 

Srupy Sires. Sites used in the study are listed and briefly described below. 
Major nesting sites for C. andreniformis on The University of Kansas campus 
were designated Horseshoe, Sunnyside Field, Athletic Field, Gym Area 
(West Stake, East Stake, Gym Path), Prairie Acre, Curb, Corner, Malott 
Path, and Smoke Stack; major sites at Nacogdoches, Texas, on the campus of 
Stephen F. Austin State College and vicinity, were designated Stadium, 
Raguet School, and High School. 

Horseshoe is the only locality found by me that sloped northward. It was 
about 91 m from Michener and Rettenmeyer’s (1956) site, which also sloped 
northward. The soil at Horseshoe is clayey, probably a clay loam, and is 
covered by fine stones and pebbles to a depth of about five mm. The top 2.5 
cm of soil are black with tan soil below. The area surrounding it is somewhat 
woodsy with considerable shade, but Horseshoe is in the sun from 10:30 a.m. 
till about 5 p.m. After rains, Horseshoe takes longer to dry out than any other 
locality where I have found andreniformis. Moreover, the dried soil splits and 
cracks more than at any other locality. Much Trifolium repens and T. 
pratense were there along with lesser amounts of T. procumbens. C. andrent- 


906 Tue UNIversiTy SCIENCE BULLETIN 


formis was first taken one week later than at the localities on south-facing 
slopes. Halictus, Agapostemon, and Lasioglossum are also common on 
flowers here. 

Sunnyside Field, Athletic Field, Gym Area, Malott Path, and Smoke Stack 
are sites with clayey soil, less dense than at Horseshoe, and in general resem- 
ble neglected lawns where the grass is relatively sparse and considerable 
Trifolium repens grows. Halictus, Lasioglossum, Augochlora, and Apis were 
common. 

Prairie Acre is a remnant of the original prairie of the region, with abun- 
dant wild flower cover and native grasses. Psoralea tenuiflora was common 
along with Asclepias tuberosa and other milkweeds. The nesting sites were 
at the bases of grasses and flowers where the growth was slightly sparser than 
elsewhere. Megachile, Lasioglossum, and Apis were the commonest bees. 

Curb and Corner were bare, with many small and large stones that inter- 
fere with burrowing. Lasioglossum was the only other bee there. 

The Texan localities, Stadium, Raguet School, and High School, were 
very similar, and much like the description above for Sunnyside Field e¢ al. 


Fic. 144. Diagram of the East and West Stake plots at the Gym Area. The dashed lines enclose 
areas continuously exposed to sunlight from 10 a.m. to 3 p.m., whereas the remainder of the 
area was intermittently shaded by the leaves of the trees whose positions are shown on the 
diagram. 


A ReEvIsION OF THE BEE GENUS 907 


Panurginus polytrichous and Lasioglossum were the commonest bees at the 
Calliopsis nesting sites. 

In order to study the seasonal development of nesting sites, two plots were 
chosen for intensive study during the summer of 1957 and 1958 in Kansas. 
Both were subdivisions of the Gym Area, called West Stake and East Stake 
(Figs. 144, 145, 146). A grid of squares 30 cm on a side was laid out on both 
areas with cord within about 3 days of the start of nesting activity on June 28, 
1957. Arrangements were made to prevent cutting of grass by mowers in the 
area, and disturbance was otherwise minimized. The areas were close enough 
together that I could check activity on East Stake with binoculars while at- 
tending to other matters at West Stake. No emergence holes were present; 
the bees had just started this as a new area of aggregation and the full 
progression could be followed. 

The soil at andreniformis nesting sites was of the loam, clay loam, or clay 
textural classes. At the Gym Area the soil became so hard upon drying that 
a scalpel was necessary to whittle it away in excavating burrows. The pH of 
the soil at the West and East Stakes was 7.43 to 8.79. Values for all other 
sites fell within these limits. 

DistripuTion oF Nests. Fig. 145 shows the distribution of nests in the 
West Stake plot, and a tabulation is made of them in Table 4. The density of 
nests varied with the barrenness of the soil. Bare soil, area 4, had about 6 times 
as many nests per unit of surface as the grassy areas 1 or 2, the vegetation of 


Oe 
BERMUDA 


0.1 m2 


Fis. 145. Distribution of nests among different types of vegetational cover in the West Stake 
plot. The principal vegetation for zones 1 to 4 is indicated. 


908 Tue UNIversIty SCIENCE BULLETIN 


which covered a relatively large area of the soil. The density of nests in area 
3 was about twice that in areas 1 or 2, and the vegetation of area 3 covered 
considerably less of the soil than that of areas 1 or 2. The density of nests was 
significantly different among the areas of the West Stake (Chi square equals 
70.8, P<0.001). 


Tasce 4. Distribution of Nests of Calliopsis andreniformis in the West Stake Plot 
(See Fig. 145) 


AREA Number No. of 
Area in of Nests per 
No. Dominant plant m? Nests m? 
le Dieitanuidecum Dens en eran ates 1.36 3 DM 
Da G NOG OMA GCN LOT meme ne een ene ee 6.65 15 2.26 
SOW RONAN LUA OF 77 Cue nena 16239 66 4.93 
43 Bai Ge Soc ooe este ncpseatcsecees casero te toes ss neteee eo scteeeerceeee 3292 56 14.29 
TOW ATES Ss 22. Se ERY EER See 25-32 140 


X= 70:8 2 01001: 


Calliopsis andreniformis will persist in a nesting site until the site becomes 
more than 809% covered with vegetation, provided that its pollen source re- 
mains available. The nesting sites at Lawrence held good populations of the 
species despite grass-sowing. When the normally preferred, hard-packed 
clayey soil sites acquired a cover of grass and clover, the bees accepted sites 
with surface layers of coarse sand 25 mm deep, and dug down through it to 
reach suitably firm soil. In observations at Lawrence from 1957-1959, and at 
Nacogdoches from 1960-1962, there was no evidence of appreciable fluctuation 
in numbers of nests or of social parasites where the physical state of the nesting 
site remained relatively unchanged. Where grass cover left less than about 
one-fifth bare area and had little or no admixture of Trifolium repens, C. 
andremiformis clearly migrated elsewhere. This occurred at the Texan nesting 
sites, Raguet School and High School, and at the Kansan sites, Horseshoe, 
Sunnyside Field, Malott Path, and the East Stake plot, all of which had pre- 
viously held good populations of the specics. Males remained at Prairie Acre 
despite the high density of herbage. 

The most spectacular effect upon populations of C. andreniformis was 
brought about by man’s activities. The following locations are close together 
on the south side of the campus at Lawrence. A football practice field at 
Lawrence having approximately one nest with five larvae per 3.7 m? was 
plowed to 15 cm depth, reconditioned, and sown with grass in May. 1958. In 
the summer of 1958 only ten nests were found in this field which carried at 
least 4500 bees prior to the reconditioning. Beside a curbing more than 200 
nests were dug up in placing an electrical conduit, with a loss of about 1000 
bees. When the West Stake plot was run over by a truck approximately 600 


A ReEvIsION oF THE BEE GENUS 909 


Fic. 146. Distribution of nests in the East Stake plot of the Gym Area. The letter D signifies a 
relatively dense growth of Cynodon dactylon. 


bees were killed. Razing of buildings in the vicinity of the West Stake plot 
destroyed several thousand more. Thus between 7,000 and 10,000 C. andreni- 
formis bees were destroyed in one season, the overwintering generation of 
1957-1958, and the species was scarce in this locality in the summer of 1958. 

Where C. andreniformis depends largely on Trifolium repens, it suffers 
when this clover is cut and its nectar and pollen become unavailable. This 
delays the time sequence of nest construction. On the other hand, cutting 
makes the clover season longer, and if the clover is not all cut at one time, the 
net effect is to extend the total length of the bee’s active season. 

A study of the changing pattern of sunlight and shade revealed that every 
nest burrow was located so as to receive continuous sunlight for 5 hours be- 
tween 8:30 a.m. and 4:30 p.m. Especially striking was the absence of nests in 
bare ground along the southern strip of the East Stake. This seemed to be 
ideal nesting territory but andreniformis did not nest there, presumably be- 
cause two ash trees cast a strip of shade there as the day progressed. I watched 
females come down to the shady area and investigate it for digging. Each 
often returned several times but eventually went a few inches north into the 
East Stake area and burrowed there (Fig. 146). 

The bees are active up to 54°C soil surface temperature. At this tempera- 
ture, however, they plunge quickly into their burrows and seldom loiter about 
on the hot surface. They fly. or attempt to fly, in winds up to 32 km per hour, 
but gusty winds of 48 km per hour discourage them and they spend much 
time in their burrows. Air temperatures between 27°C to 40°C characterized 
their most active flying periods with diminishing activity as temperatures fell 


910 Tue UNIversITy ScIENCE BULLETIN 


to 21°C at which most flying ceased. In the soil they are sluggish at such tem- 
peratures. The area receiving continuous sunlight from 10 a.m. to 3 p.m. is 
shown in Fig. 144. Many suitable nesting sites were available for andrent- 
formis, and numerous small aggregations of nests were scattered around the 
campus, mostly on the southward slopes. Other nesting sites were found 
within the city of Lawrence, but human activities usually ruined them for 
study. 

Although nests of andreniformis characteristically are aggregated, there 
are always some that are rather isolated from others. Although it does not 
seem to hide its burrow entrances deliberately, it does succeed in doing so very 
effectively on occasions. About 10°% sink their shafts against and under a 
twig or a leaf that is fairly well anchored in the ground. If large enough, such 
an object covers the tumulus. I have seen nests under prostrate dandelion 
leaves, aluminum chewing gum wrappers washed into the soil, and stones 
about 2 cm in diameter. Rozen (1958) has described a similar tendency in 
Nomadopsis. 

At Horseshoe I initially netted andreniformis females but could not locate 
their tumuli. Since the resting burrows of males were readily located, the 
female burrows were expected nearby. I found them at the bottom of soil 
cracks as large as 150 mm deep by 13 mm wide. This location would have 
serious disadvantages when the next rain came, and likely would wipe out 
many of them as the cracks closed. 

SEQUENCE OF NESTING IN RELATION To Puysicat Facrors. The sequence of 
nesting activity at the Gym Area in the East and West Stake plots, including 
the air and soil temperatures and rainfall is given in Fig. 147. The first nest 
burrows appeared after a rainfall of 1.6 cm which followed a dry period of 12 
days during which less than 2.1 cm of rain fell. It seems likely that this sud- 
den, drenching rain, which fell mostly between 8 and 11 a.m., triggered the 
nest-building activity at the Gym Area. 

The first burrows appeared on June 29th and all activity ceased after 
August 13th, a span of 47 days. The bees which arrived to nest were doubtless 
the first generation (offspring of the overwintering generation). Emergence 
of the first summer generation is between four and six weeks after emergence 
of the overwintering generation. Most likely these bees emerged from a re- 
vegetated area which was bare the year before and migrated to the bare Gym 
Area. 

The rainfall of their first four days in the area made burrowing easy, and 
the total number of active nests rose rapidly to a peak on the sixth day. As the 
soil dried and soil temperatures rose, burrowing activity dropped to a mini- 
mum until the next soaking rain on the 13th and 14th days, June 9th and 
10th. Again the total number of active nests rose to a maximum on the sixth 
day following the rain. The second peak of activity is much broader than the 


A REVISION OF THE BEE GENUS 911 


first and represents the emergence of the second spring generation plus fur- 
ther burrowing of the first spring generation. Most burrowing activity had 
died down by the 27th day, July 24. 

I decided to wet the plots using a garden hose; enough water was sprayed 
to moisten the soil to a depth of six to seven cm, that is, to the level of the first 
to third cells below the ground. I believe the third, weak peak of burrowing 
activity is attributable to this watering, but I had no control test for it. Fifty- 
five dormant nests were reactivated. 


DAYS OF ACTIVITY 
2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 


Tatil eal acaleaeT T = ] 
total number of | | | | | | | | | | 
iT | | occupied nests | | | [| Heal aR 
a | aan AnPPnal lmieleaelratalh teielt lala 
total number of | ) te || 
A | | nests with fresh 
70 sleaicataatael eal alee 
| 
| 
” 60 err 
i= | 
ul | 
ive) 
z 
50 
aL 
oO 
40 
a 
Ww 
co 
= 
>) 
z 
— 
— 
te 
Zo ptf jj} {ttt ++tititto 
s } 11 l+edaily watering period ++ 
HS) 
oo 
n . 
a + + + 
os ee 
Sie | 
i ! fastens 
29 #1 3 5 UTD CT Jina at JS 7A cer 2c Wen ae477t be eh) 3 5 7 Sie 13 
JUNE JULY AUGUST 


Fic. 147. Graph of nesting activity at the Gym Area. The solid black area gives a measure of the 
nests which are not abandoned, but are inactive for a particular day. The white area includes 
nests previously begun and currently active, while the stippled area shows the number of 
active nests on a particular day which are new burrows. In the rainfall diagram, C means 
cloudy, PC means partly cloudy, and days without a symbol were clear. The daily maximum 
and minimum air temperatures are connected by solid lines; the dashed line connects values 
for maximum soil temperatures at a depth of 12 cm. The question marks signify that no value 
was recorded for that date. 


912 THe UNIVERSITY SCIENCE BULLETIN 


Sprinc Emercence. The nest burrow fills up completely with soil during 
the winter months of rain and snow, freezing and thawing. The adult 
emerges by crawling through the soil plug in the neck of the cell; it does not 
break through the waxed wall. Apparently it displaces dirt behind it into the 
cell as it digs its way out. Emergence holes are close to the original burrow 
entrance. Males and females appear almost simultaneously, but more males 
appear to emerge than females. 

The factors which influence emergence of the overwintering brood are soil 
temperature, in turn influenced by slope, exposure, and rainfall. As men- 
tioned previously, bees nesting on a south slope are out a week earlier than 
those on a north slope. 

The emergence of the overwintering brood was best correlated with the 
flowering of Trifoliam repens (White Dutch Clover, the common lawn 
clover) at Lawrence, Kansas; Nacogdoches, Texas; and Knoxville, Tennes- 
see. The species usually emerged about a week after the first flowering of T. 
repens. At Lawrence, T. procumbens, Melilotus officinalis, and Verbena 
bracteata flowered within a week of first emergence of C. andreniformis in 
1957-1959. This close synchrony did not hold true for T. procumbens and M. 
officinalis in Vexas, 1960-1962, or in Tennessee, 1963-1964, and the last species 
was not found in the Texas or Tennessee study areas. 

Prenestinc Activiry aNnp Restinc Burrows. The prenesting activity 
lasted 7-12 days at different sites. Females and males flew about the area from 
which they emerged, resting on the ground or on sticks, twigs, and light- 
colored objects. Females were especially attracted to the painted yellow heads 
of the roofing nails used to mark burrows. 

Although some 40 marked female bees which emerged from an established 
aggregation in the Gym Area dug burrows there, some bees possibly migrated 
out of their original area either by intent or because they did not know the 
landmarks well enough to return. I lost 10 out of 20 marked males and fe- 
males at Smoke Stack in 1958 within 10 days after their emergence (death or 
predation could, of course, have caused the loss). 

At night the females rest in vertical or slightly slanting, closed burrows 
which they dig about 50 mm deep. The males do not rest in burrows with the 
females but make their own, which are usually about 25 mm deep. When 
the nesting site is established, males rarely have overnight burrows among 
those of the females, or even on the fringe of the site. They usually are as 
much as 6 m from the nearest female’s nest. 

At Prairie Acre 12 tumuli of males were crowded into an area of 0.046 m?, 
some within 25 mm of each other, with a 10 mm depth. Many more tumult 
were scattered around the edges of a patch of Psoralea tenuiflora. The bur- 
rows were initiated between tangled roots. The ground stayed damp because 


A RevIsION OF THE BEE GENUS 913 


of the density of the vegetation. I trampled down some of the grass so that 
sun would shine to the ground, and many more tumuli appeared in this spot 
than in the less accessible and damper areas. 

Two males at Horseshoe were timed from the start of digging of their 
burrows until the last movement was seen in the tumulus covering the bur- 
row. A large male began at 3.25 p.m. and finished 34.5 minutes later. The 
resultant conical tumulus was 15 mm wide and 9 mm high at the center. A 
smaller male began at 4:28 p.m., same date, and finished 36.5 minutes later at 
a depth of 25 mm. Both males and females return to their same burrows night 
after night. The female usually deepens her burrow into a nest but may dig 
a new burrow or make another tunnel down from the original entrance. 

Both sexes dig with their mandibles and their front feet, passing the dirt 
along under them and giving it an expeditious kick out of the burrow. As 
dirt accumulates and begins to drop into the burrow, the bees apparently 
pass the dirt upward as before and then press it upward with the tip of the 
abdomen. 

Night burrows of male and female can be distinguished because the 
tumulus of the female is much the larger of the two. The average dimen- 
sions of 104 tumuli cast up by females by 7 p.m. of the first night’s digging of 
a burrow were 23.3 mm wide by 6.1 mm high. The comparable figures for 9 
males were 11.4 mm wide by 4.5 mm high. The largest tumulus of a male 
was 13 x 18 x 4 mm high at the center. The smallest tumulus of an overnight 
burrow of a female was 20 x 12 x 4 mm high, with nearly all others being 
larger than 20 mm diameter by 7 mm high at the center. The periphery of 
the tumulus of andreniformis is not always round; a slightly compressed 
circular outline or very broad oval frequently results from the way the soil is 
pushed out of the burrow. 

Martine anp Mate Partrotinc. Mating takes place on clovers (rarely) or 
other flowers or at the nesting site. Stevens (1950) took mating pairs in 
North Dakota on Lactuca pulchella (Blue Lettuce) in early July. Mating at- 
tempts at clovers usually are met with evasive action or resistance by the fe- 
male. At the nesting site a male advances from behind, mounts the female, 
usually biting at her wings. Copulation lasts 2 minutes or less. Sometimes the 
two fly away in copulo. 1 have only a single record of a female which was 
mated more than once during the brief time I kept her in sight, and she did 
not resist a second mating. Often the first male on the female is pulled off as 
a succession of males try to mate with her. Females with huge pollen loads 
mate in the nest area. 

The males patrol fairly well prescribed routes which they seem to consider 
their own property. They choose a favorite spot in this route—a twig, stone, 
or leaf—and set forth in a definite flight pattern, then return and alight re- 
peatedly at the same place. If another male intrudes upon this flyway, an 


914 Tue University SCIENCE BULLETIN 


aerial “dog-fight” occurs. The two rise vertically, tumbling over each other in 
aerial acrobatics. If one catches hold of the other they fall to the ground, kick- 
ing up dust and biting each other. The intruder soon leaves, as I know from 
four marked males that became involved in such fights. This seems to be a 
clear instance of territorial behavior. The flyway is from 2 to 5 m long, and 
is not necessarily straight. Two males may have adjacent flyways with one 
loop of one overlapping that of the other; if they chance to meet a fight is 
sure to ensue. Males captured and marked with paint return to their flyways 
as if nothing had occurred. Two records of different males give some idea of 
the amount ot time spent in a flyway versus that spent flying over nearby 
clover. One male spent three minutes in the flyway, one and a half minutes 
over clover, then ten minutes back in the flyway. During the ten minutes in 
the flyway it rested on the ground for periods of 11, 8, 3, 6, and 6 seconds 
respectively. A second male alternated between flyway and clover as follows: 
5, 2, 6, 5, and 3 minutes respectively. 

I have never seen a male try to copulate with any insect or object other 
than a female Calliopsis. An apparently non-receptive female is sometimes 
pursued from flower to flower by a number of males. 

Nest Structure AND Construction. In beginning her burrow, andrent- 
formis shows a characteristic pattern which I have not seen in other bees. She 
bites the soil into a sloping arc to make her entranceway, then digs the shaft 
more or less perpendicular to the ground level. The female may begin several 
entrances, some clockwise and some counterclockwise, before proceeding with 
the sinking of the shaft. Females with full pollen loads dig burrows with 
front legs and mandibles. Wille and Michener (1951) give a female’s progress 
as 44 mm in 30 minutes digging. The rate of progress depends upon soil mois- 
ture. They do not carry water but wait until the soil is moist enough to work. 
Hence, immediately after a rain, numerous tumuli appear. Many are made 
by bees that have not dug previously, but some are made by females that have 
provisioned three or four cells (evidence from four marked females). They 
may start a new nest rather than sink the first one deeper. This seems a waste 
of energy, for the average depth to the top of the first cell is 54 mm, whereas 
later (lower) cells are made with an average vertical distance between them 
of only 7 mm, usually at different compass bearings from the main shaft. 

The tumulus thrown up seems always somewhat moist, and it may be that 
it is at this time that the female imparts an odor to the tumulus, possibly by 
adding secretions to the soil to soften it. The tumulus is always a closed 
mound, the entrance of the burrow being filled by the loose soil. Wind or 
rain will eventually obliterate the tumulus, but the upper 25 mm or more of 
the burrow remains filled with loose dirt. The female may chew out some of 
the upper wall of the burrow to keep an adequate dirt fill in the shaft. 


A REVISION OF THE BEE GENUS 915 


Pickles (1940) reported that the tumulus of Andrena armata is more alka- 
line than the ordinary soil from which it is dug. His published data, how- 
ever, are not statistically significant for the soil of the mounds, although the 
central core is decidedly more alkaline than the ordinary soil or the soil of the 
rest of the mound. Twenty-six paired samples of soil dug from burrow walls 
and the tumuli from the same nests from several localities were tested for dif- 
ferences in pH. A glass electrode pH meter and soil:water ratios from 1.10 
to 1:17.5 were used for the tests. The mean and standard error of the pH of 
the soil of the burrow walls was 8.169+0.0590 and that of the tumuli was 
8.198 0.0666; the difference is not statistically significant. The range in pH 
of the samples was between 7.4 and 8.6. 

Diagrams of the nests of C. andreniformis, C. hondurasica, and C. teucrii 
are shown in Figs. 148-153, and Table 5 gives dimensions for nests of C. 
andreniformis. The diameter of the main shaft of andreniformis is some- 
times slightly constricted at ground level by 0.5 to 1.0 mm, and this applies to 
the only nest of ¢eacrit found. Professor Alvaro Wille of the University of 


Tas.e 5. Dimensions of Burrows of Calliopsis andreniformis 


(All dimensions are in mm) Standard 

N Mean Error 
Diameter of burrow, 25 mm below ground level _............. 18 3.688 0.1291 
Same, male burrow, 10 mm below ground level —... | 2 NA 
Mengthwotaplug atentrance ses. 14 32.0 3.04 
Wenthetontoprottirsticell) Ree 2 ee ee 70 55.16 1.416 
Denthytomtop) of second) cell) 222.2 49 62.90 1.426 
Depthytoltopror. third cell! 2. pe 31 72.68 2.087 
Wepthatoptopros courthy cell 2 22 ee 22 80.32 2.365 
Depthatontoprot fetth cell 2 2 eee 16 83.81 2.261 
Depthetor top of sixth cell pss sess ee 9 86.80 235i 
Mepth to topiok seventh cell) 2... 2 occ cee eee eceeeeeeeee ee 9 83.22 0.707 
Depthytostopyotreirginthy cell 2. = ee 9 90.89 1.871 
Depthytoutoproteminthucelll eee 28.99 97.66 2.483 
Depth to top of tenth cell 0... Cea aa ey es nA pee 3} 106.00 1.155 
Minimum depth tol top of first ‘cell 22 1 34 NA 
Maximum depth to bottom of last cell 0 ae 1 162 NA 
Weighted mean number of cells per nest 0... 70 35375 NA 
Wepth of female overnight burrow? <2 -..cc..cc:.------2eceenc eect eec ences see 15 51.20 2.590 
Mepthwot malesovernight buriow —.2..2 222 ee eee 26 26.00 27 
Wengthvotlaterallibunro wae oe ee eee 16 8.00 0.713 
Domne term Ole lea. stent ene Bee Ae, A eee ee 17 3.871 0.0757 
Mette thw iy Cellet sere emi ye ea ee tee ee eee 19 7.842 0.2356 
Diameter of first night’s tumulus, female _..................2000.22..---- 20 22.90 0.894 
Dyameter, of first might’s tumulus, male: .22...22 9 11.56 1.291 
Maximum height of first night’s tumulus, female — 20 6.75 0.673 


Maximum height of first night’s tumulus, male 9 4.56 1232 


916 THe Universiry ScteNcE BULLETIN 


A REVISION OF THE BEE GENUS 917 


Costa Rica drew the nest diagrams of hondurasica based on excavations he 
made in a little used road at Playa del Coco, Guanacaste Province, Costa Rica. 
The only dimensions he recorded were the depths of cells. The architecture of 
all three species is closely similar so far as is known. No data are available on 
the cells of the nests of tewcri, but the diameter of the burrow is 4 mm, which 
falls within the range for that of andreniformis. 

After the entrance has been made the nest may be continued vertically 
downward or may slope; in the latter case it may be irregularly helical. If 
helical, the burrow usually follows the direction of the arc set at the entrance 
hole. Only one cell is placed at the end of each relatively short, lateral burrow. 
If the main burrow is a helix, the cells may be built off from it regularly clock- 
wise or counterclockwise as the helix goes deeper, but some cells show regres- 
sions (Fig. 150, cell 4), and others are built one directly below another. 

Only two nests of andreniformis with a double entrance were discovered 
in the course of the study. Several nests had a single opening with a slanting, 
blind burrow of about 50 mm depth to one side slightly below the entrance. 
Probably these were cases where females dug new burrows rather than deep- 
ening their night burrows of the prenesting period. The only nest positively 
identified as that of C. teucrit had both a double entrance and a blind burrow 
Eig. 153). 

One nest was provisioned by two different females at the same time. 
Unfortunately this was one of the casualties when a truck ran over the West 
Stake site and I do not know the architecture of the nest. Two instances were 
found, in August 1957, of females taking over abandoned, provisioned nests. 
The second females added 3 and 4 cells respectively to nests of 5 cells and 3 
cells. This was determined by dusting bees and tumuli with metallic red 
powder when they first took over the nests, and then observing which cells 
of the excavated nests showed the metallic powder. 

The cells for the developing young are short, oval, and the short lateral 
burrows to which they are attached often slant wp from the main shaft. The 
cell slants slightly downward toward its apex. The neck is closed with soil, 
packed tightly and filling the lateral burrow (Fig. 149). The interior of the 
cell is beautifully smoothed and coated with a wax-like substance which is 
insoluble in cold carbon tetrachloride, cold 95°% ethyl alcohol, and cold water, 
but dissolves instantly in ethyl ether. The thin, pliable coating of the 
spherical pollen ball in the cell is demonstrated by placing the ball in glacial 


Fics. 148-153. Diagrams of nests of three species of Calliopsis. Fic. 148. C. andreniformis, 
showing regions in a burrow. Fic. 149. C. andreniformis, a typical nest for the species. Fic. 
150. C. andreniformis, top view showing disposition of cells about the central burrow; cell 1 
is uppermost and this nest is a “clockwise”? nest except for cell 4. Fics. 151, 152. C. hon- 
durasica, as interpreted from field sketches from Prof. Alvaro Wille. Fic. 153. C. teucrit, 
Coaldale, Colorado. 


918 Tue University ScIENCE BULLETIN 


acetic acid and slicing it with a sharp blade. The internal pollen mass retracts 
from the coating and is easily seen. 

Cells are made and provisioned from the upper level downward. The time 
required to complete a nest varies with the kind of weather, and, on The 
University of Kansas campus, with the frequency and thoroughness of mow- 
ing the grass and clover mixed with it. Infrequent mowing allowed the bees 
to have clover supplies for relatively long periods of time between drastic 
drops in their major pollen source. 

A female of andreniformis customarily makes two nests. The average 
number of cells per nest is about 3.4. This figure is derived from excavations 
of burrows where activity had ceased, and may be biased by accidental death 
of the maker (see Table 5). It is my impression that nests made later in the 
season when the ground is hard and dry are more frequently deeper with 
more cells per nest, but I lack adequate data to substantiate this impression. 
The largest number of cells found in a single nest was ten, in a nest dug in 
Malott Path, August 1957. 

Other than the chasing of Holcopasites out of the nesting area by the male 
andreniformis, I have seen nothing which would be construed as a protective 
tactic. The female did not even try to re-establish the tumulus over her bur- 
row when I pushed it aside, but the tumulus did not seem to discourage social 
parasites anyway, with the possible exception of Parabombylius, which only 
flips eggs into burrows without tumuli. 

Immature Sraces. Table 6 gives sizes and weights of eggs and pollen 
balls. The eggs are slightly arcuate upward when in position on the pollen 
ball, have rounded ends and their middle portions may contact the pollen ball 
(Fig. 154) or may be raised above it. At a temperature of 30°C six eggs 


Taste 6. Physical Data on Adults, Eggs, and Provisioning of Nests of 


C. andreniformis 


Standard 

N Mean Error 
Wieightiot renal eleni casee menace aamne enn noues 23 12235 0.367 
Wietch tio tenia] egy cece eae eee eee ne 5 6.440 0.1311 
Bengthwotseg osetia eae eee ann eee 6 122.95 0.1213 
Wadthxof egos mim oe. es ee eee ee ee 6 0.3666 0.02646 
Length of pollen-collecting trips, minutes ............ 13 70.5 9:97 
Time spent in nest between pollen-collecting trips, 

Minutes se BORE ere ide oo Tey ey Ne hn 2 15 138 2.964 
Pollentloads(bothvlegs) in mg eee eee i 1529 0.1460 
Bt Dhesow lope saa¥er ee ee 7 10.86 0.407 
Pollen! ball idiameter simi 22) ee ee 11 2.475 0.1567 
Calculated number of trips for one pollen ball -.... 7 7.10 0.71 


A Revision oF THE BEE GENUS 919 


Fic. 154. Top view of the wax-lined cell made by Calliopsis andreniformis showing the pollen 
ball and egg in place. Note the mandibular markings in the wall of the lateral near its junction 
with the neck of the cell. The broader end of the egg is the anterior end and faces the opening 
of the cell. 


hatched in between 24 and 48 hours. The exact time of hatching is difficult 
to determine because the chorionic membrance is not visibly shed, but rather 
seems to be absorbed. 

The egg is oriented with the broader, head end toward the entrance, or 
plug, of the cell (Fig. 154). The posterior end is somewhat tapered. When 
the egg becomes a first instar larva, the larva seems to spread a secretion on the 
pollen ball surface, then uses its mandibles to break into the pollen ball. Per- 
haps it is a combination of a secretion and the action of the mandibles that 
opens up the enveloping coat of the ball and allows the larva to begin feeding. 
As the larva continues to feed, it eats downward on the pollen ball so that it 
eventually rests on its back, rear toward the entrance of the cell, a vestige of 
the pollen ball on its venter, and head toward the apex of the cell. Wille and 
Michener (1951:56,59) made similar observations. 

After the larva has fully fed it reorients with its head toward the cell 
entrance and rests for 2-4 days. Defecation then begins, with the excrement 
being plastered on the distal, upper end of the cell wall (Fig. 149) by the 
lateral and circular movements of the posterior end of the larva. Several fully 
fed larvae were placed in gelatin capsules which approximated the curvatures 
of their cells. They performed the above process visibly within the capsules. 


920) Tue University SCIENCE BULLETIN 


Larvae do not have to have a wall against which to void excrement. They 
will do so lying on their backs in the small plaster rearing cells described 
under laboratory rearing methods. In the latter case the excrement is voided 
as a stringlike mass which coils itself under the larva. Under such conditions 
the excrement often is like a string of toad eggs, each segment somewhat 
flattened and about 0.33 mm long. This long golden length is followed by 
pasty, gray matter which is smeared uniformly over the other excrement on 
the cell wall. 

Molting larvae presumably eat their shed skins as none are found in the 
cell. It takes the developing larva 6-8 days to consume the pollen ball at a 
temperature of 27°C. After defecation the larva can either become quiescent 
for 3-5 days and then pupate, or it may go into diapause. Fig. 155 shows a 
lateral view of two overwintering prepupae, and Michener (1953a) described 
and figured the prepupal head, mandible, spiracle, and full lateral view. The 
pupal stage lasts about 10 days in the summer. After the pupal skin is shed, 
the adult remains quiescent, and in about 36-48 hours a gray liquid suspen- 
sion is voided. The adult usually remains inactive for a day or so under labo- 
ratory conditions; after that it becomes active. 

An attempt was made to differentiate male and female prepupae on mor- 
phological or morphometric bases. No significant external morphological dif- 
ference was found between the sexes. There was so much overlap of prepupal 
weight, width of head capsule, width of body, length of body, and inter- 
tubercular distance that these measurements were impractical to use for 
separation of sexes. 

Laporatory Reartnc. Immature specimens of C. andreniformis from egg 
to prepupal stage were taken from cells and placed in various containers at 
different relative humidities and temperatures in order to rear them to adult- 
hood. No eggs or young larvae were reared through to adults. The earliest 
stage which was reared to an adult was one in which the pollen ball was about 
two-thirds consumed. 

The most convenient rearing container for a large number of specimens 
was a cigar box coated with paraffin and with its top replaced by a piece of 
glass for observation of the specimens. The cigar box can be made airtight 
with putty. It may be important to isolate specimens from each other if mold 
spore contamination is widespread. In order to handle individuals easily, 
small rectangular blocks 18 x 25 x 9 mm deep were made of plaster of Paris 
and a depression 8 x 12 x 5 mm deep of the approximate shape of the lower 
half of a bee cell was excavated in it by means of a hand-held grinding tool. 
The depression of the block was coated with parafin. The upper surfaces sur- 
rounding the depression would take pencilled identification symbols. With 
this system individuals could be easily shifted from container to microscope 
stage with the least disturbance. 


A REVISION OF THE BEE GENUS 921 


Obtaining specimens for laboratory work was an exasperating undertak- 
ing. Excavation was not difficult, but opening the cell containing the devel- 
oping bees without getting grains of dirt on the pollen ball, egg, early 
developing larva, or, worst of all, in the moisture condensate often found in 


Fic. 155. Vertical section through two cells of a nest of Calliopsis andreniformis showing the 
overwintering prepupae in place. Note that their heads are towards the openings of the cells. 
The larval feces are seen near their posterior ends on the upper distal surfaces of the cells. 


922 Tue UNIversity ScIENCE BULLETIN 


the cell around the bottom of the pollen ball, required painstaking precautions 
and luck! If a small piece of dirt lodged on the egg it could often be removed 
with very fine jewelers forceps. If it fell on the first stage larva a watery 
“blister” appeared and the larva died; if on the pollen ball, it was usually 
quickly dispersed on the wet surface and mold invariably grew there, killing 
the immature form; and, if in the cell condensate, mold usually resulted, but 
even if not, the larva ceased to feed when it came in contact with the dirt- 
contaminated pollen. If the pollen ball dried a little too much, the larva 
refused to feed and died. If the spherical pollen ball was deformed in being 
removed from the cell, the same thing occurred. It should not be surprising 
that I have the impression that C. andreniformis is a sensitive, fastidious 
species. 

To cope with these difficulties, the following techniques were used suc- 
cessfully. A petri dish was designed to serve as a receiving vessel for imma- 
ture specimens in the field. The lower half of a standard petri dish was half- 
filled with melted beeswax (parafhin did not give sufficient friction for this 


surface). A flat piece of galvanized sheet metal was cut to form a V-point, — 
the angle of the V such that a pollen ball resting in the conical cut made in | 


the wax by this tool would be supported at or below its middle. The V-point 
was simply pressed lightly into the solidified beeswax and slowly turned to 
give the necessary conical excavation. Other cone-shaped depressions were 


sunk deeper to receive water which maintained the pollen balls at their 


proper moisture during their time in the field. It was necessary to shade the 
petri collection dish for otherwise the temperature would have risen high 
enough to kill the specimens. 

A device was made to remove pollen balls with no resulting distortion of 
shape. A circular loop 1.5 mm in diameter was formed at the end of each of 
two number one stainless steel insect pins whose heads were removed. These 


were cemented to the tips of a pair of fine forceps with the planes of the loops — 


parallel to each other and to the sides of the forceps. This afforded such 
steady handling that moisture condensate could be blotted with ease from the 
bottom of the pollen ball with lens tissue during transfer from nest cell to 
petri collection dish. Once set in the conical depression, the pollen ball did 
not shift in orientation. Larvae suddenly exhumed from complete darkness 
into full sunlight continued to consume pollen normally. 

Pollen balls kept in a closed container of 93°% relative humidity (saturated 
solution of NHyH2POs,), with daytime temperature 30°C, dropping to 18°C 
nightly, shrunk in diameter from 2.83 mm to 250 mm in 48 hours. The 
feeding larva contracted, became motionless, and mold growth began in 24 
more hours. 

Pollen balls with eggs or early larvae kept in a closed container at 100% 


(olo)= 


relative humidity (distilled water) and 30°C did well until a sudden drop in 


A REvIsION OF THE BEE GENUS 923 


room temperature precipitated moisture on them. This invariably resulted 
in cessation of feeding and mold formation shortly thereafter. Mold forma- 
tion followed cessation of feeding in so many instances that I am led to specu- 
late that there may be a cause and effect relationship. Why should moisture 
condensate in the laboratory disrupt feeding when obviously moisture con- 
densate in the cell in the field does not do so? The answer may lie in the kind 
of container I used: a flat-topped one. The plane surface allowed coalescence 
of condensate and several /arge drops fell on top of the pollen ball and Jarva. 
In the rounded, waxed cell, moisture condensate rolled down the rounded 
sides of the cell and collected around the bottom of the pollen ball; neither 
pollen ball above nor specimen were wetted. Had I copied nature in making 
a waxed, domed roof for my rearing chamber at high humidity I would per- 
haps have been successful in carrying through these early stages which require 
such high humidity, yet apparently cannot stand moisture condensing directly 
on them. Gaseous mold inhibitors, or possibly a fine film of a liquid one 
applied to the pollen ball, would probably defeat the mold problem. 

Datzy Activity. Daily activity was dependent primarily on soil moisture, 
soil temperature, and air temperature. If the ground was soaked, the bees 
stayed in their burrows even though the weather was very good. Obviously 
they could not struggle through the mud. Although the greatest number of 
andreniformis were on the wing when the weather was clear and bright, 
many of them collected pollen on days when the light level was down to 860 
milliphots in contrast to 8600 to 12,900 milliphots on the best days. Cloudy 
weather, even complete overcast (30 milliphots), did not make them cease 
activity altogether. Possibly they could sense a real storm coming, for I 
never saw any of them working just prior to a downpour. Both males and 
females spend much time at the nesting site, flying about and landing here 
and there. Their antennae are held straight and move side-to-side in tandem, 
or to and fro when they settle in the area. The female ventures only a short 
distance from her burrow on the day after digging it. This seems strange 
because she has apparently gotten thoroughly acquainted with the area 
around it by actually walking over it time and time again before deciding 
just exactly what spot suits her. She is often seen to push her head and part of 
her thorax out through the tumulus and remain there moving her antenna. 
She eventually emerges but may do nothing more than walk around the 
tumulus scratching about like a cat. She may come out and immediately go 
back in. After the first day following burrow construction, the female may fly 
out of the area, presumably on foraging flights. A common action was noticed 
by Ainslie (1937), and anyone studying females at the nest will soon see it 
again and again; females often fly immediately above the ground surface if 
it is bare, and they create little whirlwinds of dust with their rapidly vibrating 
wings. Upon leaving their nests they may hover a few mm above the 


924 Tue UNIversITY SCIENCE BULLETIN 


tumulus in the manner of hover flies, pivoting slowly about the tumuli, then — 


fly one or two somewhat circular patterns over the tumuli and fly out of the 
area, 

The female’s orientation flight upon first leaving the nest in the morning 
is very short. She bursts quickly through the tumulus, circles a few times at 
rapidly increasing distances from the nest and is gone. I was never able to 


follow her closely enough to sketch the motions, but the orientation flight is — 


not intricate. 

On returning, she may fly straight to the tumulus or, more usually, she 
flies back with her pollen load to the general vicinity of the nest and alights 
here and there. Eventually she goes into her nest. I never saw a female mis- 


taking another female’s nest for her own. Related bees such as Andrena and 


Perdita may have considerable trouble finding their own nests (Michener and 
Rettenmeyer, 1956; Michener and Ordway, 1963) and may make errors in 
entering another’s nest before finding their own. Individual female Calliopsis 
nests may have distinctive odors that prevent such errors. 

Opor oF CaLLiopsis AND ORIENTATION ABOUT THE Nests. A distinctive odor 
resembling that of oil of lemongrass (Guenther, 1950) was first noticed when 
a female andreniformis on her first emergence was captured at 9 a.m., July 24, 
1957; she secreted a large drop of clear fluid from her mouthparts onto my 
fingers. Rough handling elicits this odor from females but gentle handling 
does not do so. Specimens captured with an aerial net impart their lemony 
scent to the net. I have tried to provoke its release in males but have noticed 


only the faintest odor from one specimen, a fresh male dug out of its over- 


night burrow at 6 p.m., May 26, 1958. Dr. J. G. Rozen, Jr., has informed me 
(1963) that both C. crypta and C. rozeni give off a lemony odor; they are 
members of a different subgenus. 

The odors of non-A pis bees have been a neglected facet of bee biology. G. 
E. Bohart has told me that his field experience indicates that most genera of 
solitary bees have distinctive odors. Bees with which a lemon-like odor has 
been associated are listed in Table 7. 

Malyshey (1913) found that Russian Ceratina with atrophied stings de- 
fend themselves and their nests by secreting “a yellow saliva, vigorously 
scenting of lemon. If a Ceratina is seized across the body with a forceps, it 
hastily strives to place its fragrant secretion upon the instrument by means of 
its fore and then middle legs. When the Ceratina are expelling earwigs from 
their nests this odour of lemon . . . is sometimes perceptible.” 

Moure (1958) described Warwick Kerr’s discovery that the meliponid bee 
Lestrimelitta limao uses its scent glands to give off a strong odor of lemon 
which is used to invade and rob the nests of the meliponine bees, Trigona 
emerina, remota, and testaceicornis. The strong lemony odor “overpowered 
the normal odor of the nest. . .. Once the robbed colony has lost control of 


A ReEvIsION OF THE BEE GENUS 


925 


Taste 7. Bees Which Produce a Lemon-like Odor 


Author Odor Locality Bee 
Shuckard (1866) Citron England Prosopis (=Hylacus) 
Ferton (1901) Vervain France Colletes 
= lemony Prosopis (=Hy!aeus) 
Andrena 
Panurgus 
Malyshev (1913) Lemon Russia Ceratina 
Schwarz (1948) Lemon-like Central Trigona testaceicornis 
& South Lestrimelitta limao 
America 
Shinn Oil of Colorado Calliopsis teucrit 
Lemongrass; Kansas C. andreniformis 
Citral Tennessee C. andreniformis 
Texas C. andreniformis 
Rank Oil of Colorado Nomadopsis scitula scitula 
Lemongrass 
Moure (1958) Lemon Brazil Lestrimelitta imao 
Kullenberg (1953) Lemon scent: Sweden Prosopis (=Hylaeus) 
zitronellol- Halictus 
geraniol- Heriades 
citral Megachile 
Andrena 
Nomada 
Apis mellifera (Italian) 
Rozen (1963) Lemon Arizona Calliopsis crypta 
Calliopsis rozent 
Rozen (1965a) Lemon Switzerland Melitturga clavicornis 
Weaver, Weaver, Geraniol- Texas Apis mellifera 
and Law (1964) Citral 
the odor, the Lestrimelitta workers can enter and leave freely. . . . (the lemony 


odor) seems to be the important factor in orienting the arrival of the robbers, 
guiding them to the correct entrance, even if other nest entrances are close 
oe.” 

The above references then give three functions for the lemony odor: de- 
fense; offense; guidance towards location of others of their group. The first 
function is likely in andreniformis as indicated above, and I suspect the 
presence of the last. The tumuli of the females have this odor, and it may be 
a factor which influences the bees to build their nests in aggregations. Most 
females have no trouble finding their own nests, even though several other 
entrances may be only an inch or so away. A female may plunge straight into 
her tumulus, dig furiously downward, and disappear in a few seconds. 

To test whether she detected her individual scent spectrum at close quar- 
ters to the tumulus, I performed the following experiments, Each individual 


926 Tue UNIversITY SCIENCE BULLETIN 


of a number of pairs of bees which nested very close together in the East 
Stake plot was differentially marked. A period was chosen when two bees of 
a pair having tumuli of similar size left their nests. Then the tumuli were 
exchanged and arranged approximately as found. When a female returned to 
the nest, she exhibited one of five actions: 

(1) Flew directly to her nest with neighbor’s tumulus, dug down and 
deposited pollen in the nest as though undisturbed by the exchange of tumult: 
5 individuals. 

(2) Flew directly to her nest with neighbor’s tumulus, but was disturbed, 
buzzed around tumulus and adjacent tumuli, and dug into her true tumulus 
over neighbor’s nest, but came right up again. Found her own nest after fly- 
ing around, deposited pollen load in nest: 4 individuals. 

(3) Same as (2), except dug into several tumuli including her own, taking 
more than 10 minutes to find her own nest: 1 individual. 

(4) Came into area, settled nearby (usually about a foot distant from her 
nest), made several sorties over tumuli including her own, then landed on 
hers. Dug right in, but came up again. Flew away, returned to her nest 
with neighbor’s tumulus over it, dug down, deposited pollen: 4 individuals. 

(5) Came into area without pollen, investigated several tumuli, left area; 
back in the proper nest the next morning (1 individual never returned): 3 
individuals. 

To the five individuals of (1), and the one of (3) a scented tumulus was 
unimportant in finding their true burrows. To the four specimens of (2), 
and the four specimens of (4) odor apparently played a part in the decision 
as to which tumulus was theirs, but the real test was whether the burrow it- 


self was theirs or not. If not, they did not deposit pollen in it. Unfortunately, | 


exactly the right conditions for this experiment were rarely encountered, and 
more trials will be necessary to draw definite conclusions. 

Apparently landmarks are important to andreniformis in finding its own 
burrow, but it also seems likely from these trials that scent sometimes plays a 


role when the bee alights on the tumulus. I think it possible that some bees | 


depend more upon physical-object orientation at close range, while others pay 
more heed to the scent of the tumulus. The males return to their same bur- 
rows night after night in some cases, and their tumulus is apparently un- 
scented. It is possible that odor may also function as a sex attractant, but I 
have no evidence for such a role. 

Newly emerged females in the laboratory give off the scent; males do not. 


If it were directly from their pollen ball or nectar, the males would be ex- | 
pected to furnish the scent, too. The one male that gave a very faint scent 


may have been scented by a female. 
C. andreniformis uses the nectar of Verbena, which possesses the lemon- 
odored citral in its pollen and leaves (Guenther, 1952), but the clovers used 


A REVISION OF THE BEE GENUS 927 


for pollen and nectar lack terpenes whose odors resemble citral. The odor of 
C. andreniformis is also reminiscent of citronellol and citronellal. The last 
two compounds are not widespread among plants, but it is not impossible that 
any of the three compounds could result from the bee’s metabolic breakdown 
of products present in clover. Citronellol in crocodile musk has been attrib- 
uted to the breakdown of cholesterol. 

Although the terpenes above are powerful scents which diffuse quickly, 
female tumuli at the West Stake plot continued to give the odor for more 
than two weeks (not tested after that length of time). A cat and dog repel- 


lent, Chaperone‘®? 


, whose principal active ingredient is oil of lemongrass, 
persists this long on soil or longer when applied to garbage cans which are 
left out in the sun and rain. 

Frower Revationsuips. Robertson (1929) listed 51 species of flowers vis- 
ited by Calliopsis andreniformis at Carlinville, Illinois. It has been taken on 
flowers of 98 species in all. It is certain, however, that a fair number of the 
records from the literature and specimen labels are for flowers that play little 
or no role in the ecology of the bee. 

In general, andreniformis gathers pollen and sucks nectar primarily from 
species of Leguminosae, Compositae, Verbenaceae, Oxalidaceae, and Mal- 
vaceae, in descending order of frequency of use according to my observations. 
The last family is especially favored in the salt marsh areas of southern New 
Jersey where both Malva rotundifolia and M. neglecta are abundant. 

In visiting flowers of Trifolium, Lespedeza, Psoralea, and Melilotus, the 
bee holds onto the wings of the legume blossom with its feet, slides the frons 
down against the standard, and thrusts the proboscis into the flower to get the 
nectar. Pollen is carried by the females on the hind tibiae exclusively; I have 
not observed the method of collecting the pollen. Flight ranges are usually 
short because the bee picks the closest suitable nesting site to its flower. They 
will fly at least 45 m from their nest in search of pollen, and they probably 
fly farther if their source of pollen is destroyed. 

Flowers used by andreniformis as determined by direct observation or by 
examination of pollen loads or pollen balls are listed in Table 8 and include 
Robertson’s (1929) records as well as any others where the nature of the visit 
to the flower is stated. The composition of the loads of pollen was examined 
for only 17 specimens—too few for generalization. These 17 pollen loads were 
of all Leguminosae, all Malvaceae, or mixed Malvaceae-Compositae. 

Both male and female bees feed on pollen. Taniguchi (1956) records this 
habit for Japanese Andrena, and Michener and Rettenmeyer (1956) state that 
the females of A. erythronu regularly eat pollen. The species was never seen 
to suck up dew, and their water requirement is probably met by nectar 
collection. 


928 Tue UNIVERSITY SCIENCE BULLETIN 


Taste 8. Known Pollen Sources of Female Calliopsis andreniformis 


FAMILY SPECIES 


GCOMPOSIHUAE ee ee AS ICMICTICOLILC RUZILOSUS 
Erigeron canadensis 
Verbesina helianthoides 


CON MOL VIWIEA CRNA oe ee eee ee ee Convolvulus sp. 

TAB TARTAR eee atte ona e ore e e ee Lycopus sinuatus 
Pycnanthemum pilosum 

LE GUIMIIN© SAE 8 20 ees ncaa ace reac gees desea easese scene mew re Desmodium marilandicum 


Desmodium paniculatum 
Melitotus alba 

Melitotus officinalis 
Psoralea onybrychis 
Trifolium procumbens 
Trifolium repens 
Trifolium pratense 


TE AYR EeLRUA Os EA oo To I es Me eee ee RN Ce Seas Lythrum alatum 
MAIEWiA CHALE se aemeeec ees BE plate As Grn alae ee eee Malva neglecta 

M. rotundifolia 
@ CAMEO ANG BA eo oases Se re ee Oxalts stricta 
ILO) ACV ANIEVAN ©) BAN Si Seo cee eee eae eee ee Polygala sanguinea 
PRO TRY GOINIA GC ESAT i cara coe a ea a eee se ee ee Polygonum buxiforme 
RSTO TVA EVAN ese ee ee ae ee aos oc nee cae ee Hedyotis purpurea 
S ©IR © eS TIWIIEZANINTUA © ANE wa eee Gerardia tenuifolia 
AVABIRIB ENA GEA NE oe ee eo ee eee eee Lippia lanceolata 


Verbena bracteata 
Verbena hastata 
Verbena urticifolia 


* These families are the most important pollen sources. 


The emerging overwintering generation depends largely on Trifolium 
repens, T. procumbens, Melilotus officinalis, and Verbena bracteata for its 
foraging. These plants begin blooming in early May. By about the first week 
in June Psoralea, Polygonum, Erigeron, and Convolvulus are in use for pollen 
and nectar. By the middle of June Melilotus alba and Oxalts stricta are flow- 
ering and are used for pollen and nectar. The second generation visits mostly 
Trifolium repens and Melilotus alba, as do any third generation bees which 
may emerge (rare). I never saw andreniformis make any visits to the numer- 
ous flowers of Taraxacum officinalis which bloomed in the vicinity of every 
nest site I observed. 

Sociat Parasites. The social parasites most in evidence at all localities 
where nest sites were located were bees of the genus Holcopasites. ‘These 
ubiquitous parasites have been reported by Ainslie (1937) to parasitize C. 
andreniformis. Holcopasites calliopsidis parasitizes andreniformis at Law- 
rence, Kansas; Nacogdoches, Texas; Knoxville, Tennessee; and apparently in 


A REVISION OF THE BEE GENUS 929 


Jowa and Illinois. Here recorded for the first time is the fact that Holcopasites 
illinoiensis (determination by P. D. Hurd, Jr.) parasitizes C. andreniformis 
during the same period as H. calliopsidis does, at Nacogdoches, Texas. 

Since there are only three species of Calliopsis in the eastern United States, 
and there are as many as 13 species of Holcopasites (Muesebeck, Krombein, 
and Townes, 1951), it seems likely that some of the species will be found to 
parasitize other panurgine bees closely related to Calliopsis. Linsley, Mac- 
Swain, and Smith (1956) reported the association of H. arizonicus with a 
Pseudopanurgus in Mexico; I reported (Shinn, 1965) P. D. Hurd’s record of 
the association of this species with Calliopsis pectidis in Arizona; and Rozen 
(1965b) reported the strong likelihood that this species parasitizes Pseado- 
panurgus in Arizona. He also reported finding larvae of H. null: in nests of 
C. crypta and suggested that it also attacks C. rozent. 

Despite excavations of nests of C. andreniformis which had been repeat- 
edly visited by H. calliopsidis, I failed to find the eggs of the parasite. Rozen 
(1965b) found the eggs of H. knulli inserted at an angle under a U-shaped 
flap of the cell wall in the cells of C. crypta and Pseudopanurgus aethiops. 
Most of the eggs were positioned in the roof or side of the cell. 

At some of the nesting sites H. calliopsidis emerged at the same time as 
the first andreniformis males, but at others it emerged 7-10 days later. Males 
and females appeared virtually simultaneously. The species disappeared from 
nesting sites about the time that the bulk of female andreniformis disappeared. 
I have seen no individuals persist as late as some of the female host bees, which 
may live well into September. Possibly calliopsidis never has more than two 
generations, whereas andreniformis may have a weak third or fourth genera- 
tion, depending upon latitude. 

H. calliopsidis spends much time walking around the andreniformis nest 
sites. It walks in a singular fashion; Holcopasites is the only known bee 
which tucks its wings between metasoma and legs and drags them in the dust 
as it busily examines the ground. I believe this is the first notice of this un- 
usual habit. Rozen (1965b) described H. arizonicus, H. insoletus, and H. 
knulli as flying slowly over the nesting sites, stopping frequently on tumuli or 
at the edges of stones under which there might be burrow entrances. This 
difference in behavior from that of H. calliopsidis may be a species difference 
or possibly a function of the density of nests. 

H. calliopsidis flew over the nest sites with seemingly little regard for the 
presence of andreniformis females. It would occasionally arrive at a tumulus 
simultaneously with a female andreniformis. No conflict was observed, but 
andreniformis preceded calliopsidis into the burrow. The male andreniformis 
does not tolerate the presence of calliopsidis. It chases the latter out of its area. 

The female calliopsidis burrows down through the host’s tumulus and 
may spend as much as half an hour inside. On several occasions (at least 5), 


930 Tue UNIverRsITY SCIENCE BULLETIN 


the female andreniformis returned and entered the nest during this time, but 
there was no indication of combat. The pollen-laden andreniformis spent the 
necessary time unloading pollen, emerged for another pollen-collecting trip, 
and the calliopsidis subsequently emerged. 

In the light of Rozen’s (1965b) observations that Holcopasites lays its eggs 
before Calliopsis provisions its cells, it would be interesting to determine if 
adult females of Holcopasites eat pollen from partially provisioned cells of 
Calliopsis. This might explain some of the long underground stays I re- 
corded for H. calliopsidis. 

The developmental time for calliopsidis may be shorter than that of 
andreniformis because in two nests the shallower cells contained fully fed 
larvae of Calliopsis which had not yet defecated while three deeper cells con- 
tained prepupae of Holcopasites. 

FH. calliopsidis has much longer diurnal flight activity than andreniformis. 
It flies from 8 a.m. to 6:30 p.m., whereas andreniformis is on the wing from 
about 8:30 a.m. to 4:00 p.m. Part of this difference in time is spent by the 
females of andreniformis in digging their burrows. The total active hours of 
andreniformis are greater than calliopsidis, because the former often continues 
digging until 8 p.m. 

At least 30°% of the nest burrows in the West Stake plot (Fig. 146) were 
visited by calliopsidis, but less than 4°% of all the dug cells of andreniformis 
contained calliopsidis larvae (Table 10). Apparently many visits are made 
before calliopsidis finds a cell suitable for laying her egg. A single female may 
briefly visit several nests repeatedly—a few seconds each—before remaining 
underground in one of them for a longer time of a minute or more. 

Frequent mating of Holcopasites took place at the Calliopsis nesting sites. 
I never saw mating attempts on flowers. The male calmly mounted the fe- 
male, which did not struggle to escape, and copulation lasted 1-2 minutes on 
the ground, the female not attempting to fly. At each nesting site the para- 
sites were readily netted and were marked with quick-drying colored airplane 
dope. They returned repeatedly to the same site, but ten of twenty marked 
at the Gym Area disappeared within a week’s time. Possibly they disperse to 
other nesting sites. 

Dr. Howell V. Daly found a few calliopsidis asleep at 6 p.m., July 14, 1957, 
on the acuminate tips of the grass, Setaria glauca (det. R. N. McGregor) on 
The University of Kansas campus. They slept holding by their mandibles 
alone, their wings tucked between the metasomal sternum and hind legs 
(Fig. 156). Rozen (1965b) described the same sleeping position for H. 
insoletus and knulli. The depth of sleep in calliopsidis was such as to permit 
the grass plant on which it was sleeping to be dug up and transported 1 km 
by foot to a laboratory for photographing without disturbing the bee. More 


A ReEvIsIoN OF THE BEE GENUS 


Fic. 156. Photograph of Holcopasites calliopsidis asleep on the upright, acuminate tip of a blade 
of the grass Setaria glauca. Note the characteristic position of the wing which hes between the 
hind leg and the metasomal sternum. 


932 Tue University SciENCE BULLETIN 


than a dozen calliopsidis bees which I marked with colored paints returned to 
sleep in the same vicinity, sometimes on the same blade of grass. 

Sphecodes bees were found at every nesting site I studied. They were 
much scarcer than Holcopasites, but | made no quantitative comparisons. 
They investigated the andreniformis burrows and entered them, staying as 
long as ten minutes (14 timings, mean 4.861.015). I found no larval 
Sphecodes in any andreniformis cell; however, the larva from a cell of 
andreniformis described as that of Holcopasites by Michener (1953a) is 
totally unlike that of other pasitine bees, and Dr. J. G. Rozen, Jr., (in litt.) 
considers it to be a Sphecodes. There is therefore little doubt that Sphecodes 
does parasitize Calliopsts. 

At the Stake nest plots two sizes of Sphecodes were present and were 
presumably two species. Both visited the Calliopsis burrows, although the 
smaller one made most of its visits to Lasioglossum nests in the area. Mitchell 
(1956) suspected that Sphecodes brachycephalus is a probable parasite of C. 
andremformis. He also writes (1960) that his personal observations in nest- 
ing areas suggest both Calliopsis and Perdita as hosts for Sphecodes. Ainslie 
(1937) observed Sphecodes flying around and active in nest sites of C. andrent- 
formis at Sioux City, Iowa. Rau and Rau (1916) found Sphecodes sp. in 
constant attendance at the nest sites of Calliopsis nebraskensis near St. Louis, 
Missouri, and considered them as visitors to the nests. 

Villa sinuosa and Parabombylius ater were the only bee flies whose actions 
suggested that they were parasites of C. andreniformis. The former was rare, 
but the latter was a frequent visitor at the Kansas nesting sites and was very 
abundant at the Texas nesting sites. Several bombyliid larvae were exca- 
vated from andreniformis cells, but none pupated. 

Several counts of P. ater were made. On July 4 and July 16, 1957, five and 
seven individuals were simultaneously active in the West Stake nest plot, or 
one per 5.1 m* and one per 3.6 m*, respectively. At the Stadium nesting site 
in Texas on May 10, 15, and 17, 1962, there were 12, 16, and 20 individuals 
simultaneously active in an area of comparable size, or one per 18.6, 14.8, and 
12.0 m’, respectively. The bee flies hover about 25 mm from a nest whose 
tumulus has been removed, and they flick tiny white eggs into the exposed 
burrow entrance. Presumably the larva burrows down through the dirt-filled 
upper portion of the nest entrance, enters a cell and eats pollen and Calliopsis 
larva, for cells containing bee fly larvae have no traces of anything else in 
them. Parabombylius ater appears about the same time as C. andreniformis; 
its peak of abundance is about coincident with mid-season for its host, and its 
population appears to drop sharply after this. I interpret this to mean that it 
has only one generation in Kansas. The fact that several larval specimens 
dug up in early June did not metamorphose, whereas prepupal Calliopsis did, 
tends to support this interpretation. 


A ReEvIsION OF THE BEE GENUS 933 


Orner Associates. The pyemotid mite, Trochometridium tribulatum 
(Cross, 1965) was discovered in two cells of one nest on The University of 
Kansas campus. Both fungus and pollen were present in these cells. No 
traces of the bee larvae were found. Krombein (1961) points out that several 
families of mites contain species that are parasites of solitary wasps and 
bees, and that several species of Pyemotes kill and feed on the more or less 
helpless immature stages of many insects. This may be the case with 
Trochometridium. 

The mites were collected July 22, 1957, from cells 73 and 90 mm below 
ground level from a block of soil which had been recently brought to the 
laboratory. Several hundred mites were in the two cells taken together. A 
huge gravid female was in one cell and almost all the mites were in the egg 
stage. The mites were transferred to a covered Syracuse dish for observation 
at room temperature (29.5°C). On July 23, most of the mites were in a ball, 
but four walking specimens had appeared. On July 24, twelve walking fe- 
males were present, and on July 25 two males appeared. Females were 
distinguished from males by a straight white line on the dorsum, by their 
smooth contour in lateral view, by their more elongate form, and their 
smoother locomotion. The male bore a broader white line on the dorsum 
with several strong constrictions along its border, exhibited a posterodorsal 
tubercle in lateral view, was broadly oval in outline, and moved slowly and 
clumsily. Its locomotion was mostly by means of the middle two pairs of legs, 
for the front and hind pairs were borne aloft. By July 27 most eggs had 
hatched, but mold was forming on the unhatched eggs. Only a few males 
had appeared, the bulk of the specimens being females. Inasmuch as Dr. 
Earle A. Cross had recognized them as a new genus and species, the speci- 
mens were preserved on July 27. Ainslie (1937) found larvae of C. andreni- 
formis infested with mites of the genus Pygmephorus (=Pigmeophorus 
Banks, 1904) as determined by H. E. Ewing. He gave no estimate of the 
extent of the infestation, but I infer it was relatively minor. Crandall and 
Tate (1947) state that late in the season of C. andreniformis at Lincoln, 
Nebraska, “... many of the cells containing larvae were infested with mites. 
... They listed the mites as: Pediculoides americanus (Banks) and Tyro- 
phagus sp., both determined by E. W. Baker, and Lohmannia sp., deter- 
mined by H. E. Ewing. 

Three species of fungi were identified from prepupal C. andreniformis: 
Penicillium cyclopium, Aspergillus flavipes, and A. sydowi. Dr. Robert 
Lichtwardt kindly determined the molds. The molds came from specimens 
in the soil block discussed above, from the West Stake plot, and from speci- 
mens reared in the laboratory. Only bees in the rearing boxes kept at 88°/ 
relative humidity (maintained by BaCly solution) and above developed mold 


934 Tue UNIversity ScIENCE BULLETIN 


Tasie 9. Contents of Calliopsis andreniformis Cells Dug During Summer and 
Early Autumn 


Number Percentage 

GNARL EN ORM a sea lity came TANNIN el U1 meee ee ae 192 77.4 
Ghandrenijormis) dead moldyseiminmatin eee ee ee 22 8.9 
Emptya, waxedttcellls ees. seee tee ee ee ee ee eee 12 4.8 
Holcopasites oni phecodessprepupac eee oe ae 10 4.0 
Béehliéess Wana meee noes eae ee ee ee Le 20 e oa ee 8 3.2 
Mites, Trochometridium tribulatum __... BE tee ea 4} 1.6 

6 tall Seen eee fee Se ee Sr Se rks eee Pure Teac toseacdtap ss be seet steered oese eeseee 248 99.9 


growth. Two bee fly larvae, surrounded by moldy bee prepupae, did not sup- 
port a mold growth. 

Penicillium cyclopium is worldwide and found on many different sub- 
strates, for example, rotting bulbs of Liliaceae, mildewing tentage, soil and 
decaying vegetation, and in bee hives (Raper and Thom, 1949). Both 
Aspergillus species are cosmopolitan. A. flavipes is particularly common in 
scil and decomposing organic materials, and A. sydowi is known from soil 
of several eastern states and from beehives in Michigan (Thom and Raper, 
iA): 

A small asilid fly was present at three nest sites at Lawrence, Kansas. It 
looked remarkably like a female of C. andreniformis and acted and flew like 
the female as well. Its only known prey were two male andreniformis, which 
tempts speculation that it may be aided in obtaining its prey by its superficial 
resemblance to the female. 

Although the bee-predator wasps of the genus PAilanthus were seen about 
the West Stake nesting plot as well as several others, they were never seen to 
take andreniformis, or for that matter, any prey. Reinhard (1924) records 
Philanthus gibbosus as using andreniformis for prey: Philanthus cells yielded 
1 female and 5 males of andreniformis and 325 specimens of halictid bees. It 
seems that andreniformis, representing only 1.8°% of the prey of Philanthus, 
could well be considered as accidental prey. Philanthus captures its prey on 
flowers, and a list of flowers given by Reinhard as used by P. gibbosus includes 
some flowers used by andreniformis: Achillea millefolium, Erigeron, and 
Polygonum. These flowers are rarely used by andreniformis, but much by 
Halictidae. This, too, lends support to my supposition that andreniformis is 
an exceptional prey for P. gibbosus rather than one that is customarily used at 
a low percentage. 

A Nysson wasp was much in evidence about andreniformis nests at the 
West Stake plot, often entering nests, but never emerging with any plundered 
prey or external evidence of pollen. Rau (1922) recorded the entrance of 


A REVISION OF THE BEE GENUS 935 


Nysson raui into the burrow of C. (Verbenapis) nebraskensis. 1 have no 
other observations about the wasps’ action in connection with andreniformis 
nests. 

White podurid collembolans were sometimes found in nests. A clamydid 
beetle larva which encases itself in a mud cocoon occasionally spent the night 
in burrows whose upper portions were free of loose, excavated dirt. 

Table 9 lists the contents of cells from apparently completed nests which 
were dug during summer and autumn. 


LITERATURE Cle) 


Arnstiz, C. N. 1937. Notes on the biology of two panurgine bees. Canadian Entomol. 
69:97-100. 


CranpbacL, B. H. and H. D. Tate. 1947. The bee Calliopsis andreniformis as a factor in alfalfa 
seed setting. Jour. Amer. Soc. Agronomy 39:161-163. 


Cross, Earte A. 1965. The generic relationships of the family Pyemotidae (Acarina: Trom- 
bidiformes). Univ. Kansas Sci. Bull. 45:29-275. 


Ferton, C. H. 1901. Notes détachées sur l’instinct des Hyménoptéres melliféres et ravisseurs 
avec la description de quelques espéces. Ann. Soc. Entomol. France 7():83-148. 


GuENTHER, E. 1950. The essential oils. Vol. 4, Van Nostrand Co., Inc., N.Y., xiv + 752 pp. 
. 1952. Idem. Vol. 5. xvii + 507 pp. 


KroMseIn, Kart V. 1961. Some symbiotic relations between Saproglyphid mites and solitary 
Vespid wasps. Jour. Washington Acad. Sci. 51:89-92. 


KuLtensurc, B. 1953. Some observations on scents among bees and wasps (Hymenoptera). 
Entomol. Tidskr., Stockholm 74:1-7. 


Linstey, E. Gorton. 1958. The ecology of solitary bees. Hilgardia 27 :543-599. 


Linsey, E. G., J. W. Macswain, and Ray F. SmirH. 1956. Association of Holcopasites with 
Pseudopanurgus in Mexico. Pan-Pacific Entomol. 32:82. 


LoveLt, Harvey, and JoHN H. Loverr. 1939. Pollination of Verbena hastata. Rhodora 
41:183-186. 
MatysHey, S. I. 1913. Life and instincts of some Ceratina-bees (Hymenoptera, Apidae). A 


comparative and experimental study. Trudy Russkago Entomologicheskago Obshchestva, 
St. Petersburg 40:35-60. 

MICHENER, CHaRLES D. 1953. Comparative morphological and systematic studies of bee larvae 
with a key to the families of hymenopterous larvae. Univ. Kansas Sci. Bull. 35:987-1102. 

MICHENER, C. D., E. A. Cross, H. V. Daty, C. W. RETTENMEYER, and A. Witte. 1955. Addi- 
tional techniques for studying the behavior of wild bees. Insectes Sociaux 2:237-246. 

MicHener, C. D. and C. W. RetrenMeyer. 1956. The ethology of Andrena erythroni with 
comparative data on other species (Hymenoptera, Andrenidae). Uniy. Kansas Sci. Bull. 
37 :645-684. 

MicHener, CuHarces D, and ELLEN Orpway. 1963. The life history of Perdita maculigera 
maculipennis (Hymenoptera:Andrenidae). Jour. Kansas Entomol. Soc. 6:34-45. 

MicHENER, CHarces D, and Atvaro Witte. 1961. The bionomics of a primitively social bee, 
Lasioglossum inconspicuum. Univ. Kansas Sci. Bull. 42:1123-1202. 

MircuHett, T. B. 1956. New species of Sphecodes from the eastern United States. Jour. Elisha 
Mitchell Sci. Soc. 72:206-222. 

——. 1960. Bees of the eastern United States. Vol. 1. North Carolina Agric. Exp. Station 
Tech. Bull. No. 141, 538 pp. 

Mourg, J. S. 1958. Evolutionary problems among Meliponinae (Hymenoptera, Apidae). Proc. 
Tenth Inter. Congr. Entomol. 2:489-493. 

MueseBEck, C. F. W., K. V. Krompein, and H. K. Townes. 1951. Hymenoptera of America 
north of Mexico. U.S.D.A. Agriculture Monograph No. 2, Washington, D.C., 1420 pp. 


936 Tue University SCIENCE BULLETIN 


Picxets, W. 1940. Observations on the soil of the mounds of the mining bee Andrena armata 
Gmelin (fulva Schrank). Entomol. Monthly Mag. 76:230-231. 


Raper, KenNetH B. and CHArtes THom. 1949. A manual of the Penicillia. Williams and 
Wilkins Co., Baltimore, Maryland, ix + 875 pp. 

Rav, Pum and Neue Rav. 1916. Notes on the behavior of certain solitary bees. Jour. Anim, 
Behav. 6:367-370. 

Rav, Pum. 1922. Ecological and behavior notes on Missouri insects. Trans. Acad. Sci. St. 
Louis 24:1-71. 


REINHARD, Epwarp G. 1924. The life history and habits of the solitary wasp Philanthus gib- 
bosus. Smithsonian Inst. Ann. Rep. for 1922, Publ. 2738:363-376. 


Rozertson, C. 1929. Flowers and insects. Carlinville, Illinois, 221 pp. 


Rozen, Jerome G. Jr. 1954. Morphological description of the larva of Oreopasites vanduzeet 
Cockerell. Pan-Pacific Entomol. 30:203-207. 


———. 1958. Monographic study of the genus Nomadopsis Ashmead (Hymenoptera:An- 
drenidae). Univ. California Pubs. Ent. 15:1-202. 


———. 1963. Personal communication. 


——. 1965a. The biology and immature stages of Melitturga clavicorms (Latreille) and of 
Sphecodes albilabris (Kirby) and the recognition of the Oxaeidae at the family level 
(Hymenoptera, Apoidea). Amer. Mus. Novitates No. 2224:1-18. 


—. 1965b. Biological notes on the cuckoo bee genera Holcopasites and Neolarra (Hy- 
menoptera: Apoidea). Jour. New York Ent. Soc. 73:87-91. 


ScHwarz, H. F. 1948. Stingless bees of the western hemisphere. Bull. Amer. Mus. Nat. Hist. 
90:xvint + 546 pp. 

SHucKarD, W.E. 1866. British bees. L. Reeve and Co., London, 371 pp. 

SHinn, A. F. 1965a. The bee genus Acamptopoeum: Diagnosis, key, and a new species 
(Hymenoptera:Andrenidae). Jour. Kansas Ent. Soc. 38:278-284. 

——. 1965b. Descriptions of three new species of the bee genus Calliopsis (Hymenoptera, 
Andrenidae). Amer. Mus. Novitates No. 2211:1-19. 

SreveNns, O. A. 1950. Native bees. North Dakota Expt. Station. Bimonthly Bull. 12:90-98. 

Swenk, M. H. and T. D. A. Cockerett. 1907. The bees of Nebraska. II. Entomol. News 
18:178-179. 

TanicucHl, Setsu. 1956. Biological studies on the Japanese bees. II. Request in the flower- 
visiting of infrasocial bees. Sci. Reports Hyogo Univ. Agric. (Series: Agric. Biol.) 2:37-51. 

THomM, CHARLES and KENNETH B. Raper. 1945. A manual of the Aspergilli. Williams and 
Wilkins Co., Baltimore, Maryland, ix + 373 pp. 

Weaver, N., EvizapeTH C. WEAvER, and JoHN H. Law. 1964. The attractiveness of citral to 
foraging honey bees. Texas Agric. Expt. Station Prog. Rep. No. 2324:1-7. 

Witte, A. and C. D. MicHener. 1951. Unpublished notes from the files of the Bee Biology 
Project, Univ, Kansas, Dept. Entomology. 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


THE CULICOIDES OF NEW MEXICO 
(DIPTERA: CERATOPOGONIDAE) 
(Diptera: Ceratopogonidae) 


By 


William R. Atchley 


Marine see Laboratory 
LIBRARY 


MARS 1967 
WOODS HOLE, MASS. 


VoL. XLVI Pacss 937-1022 Marcu 3, 1967 No. 22 


ANNOUNCEMENT 


The University of Kansas Science Bulletin (continuation of the Kansas Uni- 
versity Quarterly) is issued in part at irregular intervals. Each volume contains 
300 to 700 pages of reading matter, with necessary illustrations. Exchanges with 
other institutions and learned societies everywhere are solicited. All exchanges 
should be addressed to 


Liprary OF THE UNIVERSITY OF Kansas, 
LAawreENCcE, Kansas 66044 


PUBLICATION DATES 


The actual date of publication (z.e., mailing date) of many of the volumes of 
the University of Kansas Science Bulletin differs so markedly from the dates on 
the covers of the publication or on the covers of the separata that it seems wise to 
offer a corrected list showing the mailing date. The editor has been unable to 
verify mailing dates earlier than 1932. Separata were issued at the same time as 
the whole volume. Beginning with Volume XLVI, publication was by separate 
numbers and the date on each number is the actual publication date. 


Vol. XX—October 1, 1932. Vol. XXXIV,Pt. I—Oct. 1, 1951. 
Vol. XXI—November 27, 1934. Pt. II—Feb. 15, 1952. 
Vol. XXII— November 15, 1935. Vol. XXXV,Pt. I—July 1, 1952. 
Vol. XXIII—August 15, 1936. Pt. II—Sept. 10, 1953. 
Vol. XXIV—February 16, 1938. Pt. III—Nov. 20, 1953. 
Vol. XXV—July 10, 1939. Vol. XXXVI,Pt. I—June 1, 1954. 
Vol. XXVI—November 27, 1940. Pt. II—July 15, 1954. 
Vol. XXVII,Pt. I—Dec. 30, 1941. Vol. XXXVII, Pt. I—Oct. 15, 1955. 
Vol. XXVIII, Pt. I—May 15, 1942. Pt. II—June 29, 1956. 

Pt. II—Nov. 12, 1942. Vol. XXXVIII, Pt. I—Dec. 20, 1956. 
Vol. XXIX, Pt. I—July 15, 1943. Pt. Il—March 2, 1958. 

Pt. II—Oct. 15, 1943. Vol. XXXIX—Nov. 18, 1958. 
Vol. XXX, Pt. I—June 12, 1944. Vol. XL—April 20, 1960. 

Pt. II—June 15, 1945. Vol. XLI—Dec. 23, 1960. 
Vol. XXXI, Pt. I—May 1, 1946. Vol. XLII—Dec. 29, 1961. 

Pt. II—Nov. 1, 1947. Vol. XLII—Supplement to, June 28, 1962. 
Vol. XXXII—Nov. 25, 1948. Vol. XLIII—Aug. 20, 1962. 
Vol. XXXIII, Pt. I—April 20, 1949. Vol. XLIV—Sept. 1, 1963. 

Pt. II—March 20, 1950. Vol. XLV—June 7, 1965. 

AOR Soc Waratah ce enone R. C. JacKson 
Editorial Board ........ GeorceE Byers, Chairman 


KENNETH ARMITAGE 
CHARLES MICHENER 
Pau Kiros 
RICHARD JOHNSTON 
DELBERT SHANKEL 


THE UNIVERSITY OF KANSAS 


SCIENCE BULLETIN 


Voit. XLVI Paces 937-1022 Marcu 3, 1967 INO; 22 


The Culicoides of New Mexico (Diptera: Ceratopogonidae)' 
By 


WituraM R. ArcHLey 
Department of Entomology 
University of Kansas 


ABSTRACT 


A taxonomic review is provided for the biting midges of the genus Culicoides 
(Diptera: Ceratopogonidae) occurring in New Mexico. The descriptions of females 
and males, if known, of 23 species are given of which 4, Culicoides (Drymodes- 
myia) byerst, C. (Oecacta) doeringae, C. (O.) novamexicanus and C. (O.) sub- 
letter are described as new to science. Culicoides pampoikilus Macfie, previously 
reported from Central America, is recorded from New Mexico and constitutes a 
new United States record. Biological as well as morphological evidence is presented 
to support the relegation of C. vari pennis australis Wirth and Jones to synonomy. 
Culicoides tenuilobus Wirth and Blanton is elevated from synonomy. 


INTRODUCTION 


The genus Culicoides, an almost worldwide group of minute hemato- 
phagous gnats, is a conspicuous part of most environments. The anthro- 
pophilic behavior of these flies makes them readily perceivable to even the 
most unobserving “naturalist.” 

It has only been in the last 40 years or so that students of the Cerato- 
pogonidae have devoted their energies to the taxonomy of these small flies. 
The Palearctic fauna has been well studied through the efforts of the fol- 
lowing workers: Goetghebuer (1920, 1933-1934) first. monographed the 
Belgian species and later more extensively studied the Palearctic fauna, while 
Kieffer (1925) and Edwards (1926, 1939) treated the French and British 


‘Contribution No, 1333 from the Department of Entomology, University of Kansas. 


938 Tue University SciENCE BULLETIN 


species. Such workers as Callot and Kremer are continuing the studies of 
the French fauna, while Kettle and Lawson are among the more recent con- 
tributors of the studies relating to the British species. The revision by 
Campbell and Pelham-Clinton (1960) represents the most up-to-date treat- 
ment of the British fauna. Among the more notable Russian contributors 
to the taxonomy of the Palearctic Culicoides have been Gutsevich and 
Dzhafarov. 

Carter, Ingram and Macfie (1920), DeMeillon (1936-1961) and Fiedler 
(1951) have contributed to our knowledge of Culicoides in the Ethiopian 
Region, although Macfie’s later contributions were by no means restricted 
to this area. 

The Australian species have been recently monographed by Lee and 
Reye (1953). The Oriental Region has long been ignored by ceratopogonid 
workers and, with the exception of a reference by Causey (1938) to species 
occurring in Siam, little is known of the Culicoides of this region. Tokunaga 
(1932 to date) and Arnaud (1956) have conducted extensive taxonomic 
studies on the species in Japan and surrounding areas, and, in addition, 
Tokunaga has recently presented excellent studies of the Micronesian and 
Papuan fauna. 

Valuable contributions have been made to the taxonomy of the Neo- 
tropical species by Barbosa (1943-1952), Fox (1942-1955), Lane (1944-1961), 
Lutz (1912-1914), Macfie (1937-1948), Ortiz (1949-1954), and Vargas (1944- 
1960). The excellent monograph of the Panamanian species by Wirth and 
Blanton (1959) provides a valuable addition to our knowledge of the 
Neotropical species. 

Within the boundaries of the continental United States the Culicoides of 
California, Florida, Wisconsin, Oklahoma and New York have recently 
been the subjects of rather extensive taxonomic studies. A perusal of the 
literature pertaining to the Culicoides of the United States, however, reveals 
a conspicuous void regarding the culicoid fauna of the southwestern states. 

Information regarding the Culicoides of the southwestern United States 
can be found in the works of James (1941, northern Colorado), Fox (1946, 
Utah), Wirth (1952, California), and Khalaf (1952, 1957, Oklahoma), and 
Wirth (1955), Wirth and Bottimer (1956), and Jones and Wirth (1958) have 
studied the Texas species. Occasional records of southwestern species are 
given by Foote and Pratt (1954) in their treatment of the Culicoides of the 
eastern United States. Wirth and Hubert’s (1960) study of the copiosus 
group of Culicoides includes distribution records for species occurring in 
California, Arizona and Texas, as well as some Mexican records. 

The Catalog of North American Diptera, North of Mexico (1965) lists 
21 species occurring in Texis, and Khalaf (1957) lists 22 valid species in 
Oklahoma. This compares quite unfavorably with 9 species recorded in the 
literature as occurring in New Mexico. New Mexico, the point of fusion 


Tue Cuticomwes or New Mexico 939 


of four physiographic regions, i.e., Southern Rocky Mountains, Great Plains 
Province, Colorado Plateau and the Basin and Range Province (Fenneman, 
1931), presents a diverse ecological situation. Because of this diversity, one 
would expect a varied and interesting fauna of Culicoides. 

The primary objective of this study is the presentation of a taxonomic 
treatment of the Culicoides of New Mexico. It is hoped that this endeavor, 
when added to the works of Wirth (1952) and Khalaf (1957), will form 
the basis for a future extensive taxonomic investigation of the Culicoides 
of the western United States. Such a study is of paramount importance if 
serious investigations involving the public health significance of this group 
in the western United States are to be continued. 


ACKNOWLEDGMENTS 


I would like to extend my sincere thanks to Dr. George W. Byers, under 
whose direction this study was made, for his enthusiasm toward this en- 
deavor and consideration in the many problems brought to him. To Dr. 
James E. Sublette, who first suggested this problem, directed its initiation and 
provided me with many specimens, I offer my deep appreciation. 

I wish to thank Dr. Joseph H. Camin, who read and criticized the 
manuscript, Dr. Charles D. Michener for his many suggestions relating to 
problems of systematics, John A. Hendrickson and Jack L. Boese, who read 
and suggested improvements on various facets of this paper. 

Special thanks are due Dr. Willis W. Wirth, of the United States National 
Museum, for his gracious assistance in confirmation of my determinations, 
loan of material and answers to many perplexing taxonomic problems. His 
assistance in this study has been of immeasurable benefit. 

I would like to gratefully acknowledge my indebtedness to the follow- 
ing members of the New Mexico Department of Public Health for their 
cooperation and assistance in securing light trap material: Robert W. Allen, 
Cade L. Lancaster, George May, Bryan Miller, Hiram T. Miller, Bill F. 
Mittag, Richard M. Swartz and Edwin Walrath. Thanks are due Dr. 
Robert H. Jones, United States Department of Agriculture, for the loan of 
specimens for comparison, and Dr. Cluff Hopla, who searched for type 
material at the University of Oklahoma Museum. To the many others who 
contributed materially with reprints of their studies in Culicoides taxonomy 
and who otherwise made this study easier, I offer my sincerest thanks. 

Special appreciation is given to the American Museum of Natural History 
for their financial assistance from the Theodore Roosevelt Memorial Fund. 

In conclusion, I would like to express my deep obligation to my wife, 
Wilinda, for her encouragement and assistance throughout the duration 
of this study. 


940) Tue University SciENCE BULLETIN 


HISTORICAL REVIEW 


One of the first accounts of Culicoides was given as early as 1713 by 
Derham (Edwards, 1939) in his excellent description of a small gnat re- 
ferred to as Culex minimus nigricans maculatus sanguisuga. Concerning this 
small gnat, Derham states, “It is spotted with blackish spots, especially on the 
wings, which extend a little beyond the body. It comes from a little slender 
Eel-like Worm, of a dirty white Colour, swimming in stagnating Waters 
by a wrigling Motion. 

“Its Aurelia (pupa) is small, with a black Head, little short Horns, a 
spotted, slender rough Belly. It lies quietly on the top of the Water, now 
and then gently wagging itself this way, and that. 

These Gnats are Greedy Blood-suckers, and very troublesome where 
numerous....” 

In spite of this early beginning the Ceratopogonidae were shunned by 
early workers because of their small size and the consequent difficulties im- 
posed by their taxonomy. Only one species of Culicoides, then called Culex 
pilicarius, was known to Linneaus. Winnertz’s (1852) monograph of the 
European species represented the first major attempt at the classification of 
the Ceratopogonidae. Both Winnertz and Coquillett, who first studied 
the North American fauna, included all the known species in the genus 
Ceratopogon. 

It was Kieffer who first began to define generic limits in the Cerato- 
pogonidae and in 1906 resurrected the generic name Culicoides Latreille 
(1809). Following the work of Kieffer, Malloch (1917) proposed the re- 
cognition of the Ceratopogonidae as a distinct family from the Chironomidae. 

Hoffman (1925) reviewed the North American species of Culicoides, 
described several new species and included a key for females. The 1937 
revision of the North American Culicoides by Root and Hoffman included 
additional new species and distribution records, keys for female characters 
and male genitalia and illustrations of male genitalia of most of the known 
species. 

With the exception of a few isolated descriptions of new species or 
species lists of very limited geographical areas, the next taxonomic works 
were those of Johannsen (1943, 1952) and Wirth (1952). Johannsen’s (1943) 
generic synopsis and list of the North American species of Ceratopogonidae 
filled a long neglected taxonomic gap. The monograph on the Cerato- 
pogonidae (—Heleidae) of California by Wirth (1952), with its excellent 
species descriptions, keys and overall systematics, stands as the outstanding 
North American treatment of this family. 

Foote and Pratt (1954) treated the species occurring in the eastern United 
States; however, the usefulness of this paper is somewhat limited, due to 
the inadequacy of the species descriptions. These authors were among the 


Tue Cuticomes or New Mexico 941 


first to attempt classification of females on the basis of structural characters, 
instead of relying completely on the traditional wing and mesonotal patterns. 

Khalaf (1952, 1957), studying the seasonal incidence of the Culicoides of 
Oklahoma, described the relative abundance and distribution of these flies 
and provided descriptions of new species and keys to the Oklahoma 
Culicoides. As in the case of Foote and Pratt, Khalaf supplemented his 1957 
key to external characters and male genitalia with a key to slide-mounted 
females. This is of significance to the problem at hand, since many of the 
species found in Oklahoma also occur in New Mexico. 

The catalogue of the bloodsucking midges of the Americas by Fox (1955) 
includes a compilation of the known species, together with keys and a 
geographical index; however, more up-to-date information regarding distri- 
bution records, classification and synonymies is now available in the Catalog 
of North American Diptera (Stone et al., 1965). 

Considerable emphasis has recently been placed on the taxonomy of species 
groups, resulting in revisions of the copiosus group by Wirth and Hubert 
(1960), the eastern species of the piliferus group by Wirth and Hubert 
(1962), and the odsoletus group of eastern U.S. by Jamnback and Wirth 
(1960). Although these studies represent a strong beginning, there exists a 
great need for studies involving the western species of the piliferus and 
obsoletus groups, as well as the crepuscularts group, or the subgenus Se/fia, 
to name only a few pressing taxonomic problems. 

The study of the immature stages of Culicoides in the United States has 
recently been given a needed stimulus through studies by Jones and Williams, 
although our knowledge still lags behind that of the British, who have 
amassed considerable literature relating to immature stages of various 
Palearctic species through the excellent works of such entomologists as 
Kettle, Lawson and Parker. 


ECONOMIC IMPORTANCE 


The bloodsucking behavior of Culicoides (and two related genera of 
Ceratopogonidae) has gained them considerable notoriety. When these 
small gnats (commonly known as punkies, moose-flies, no-see-ums or sand- 
flies, although the last name has been applied also to Phlebotomus) occur in 
large numbers, their hematophagous habits result in great annoyance both 
to man and his animals. To many people the mention of “biting gnats” 
brings to mind painful, vicious, invisible and insatiable bloodsuckers. As 
Reye (1964) has so ably stated, “the abrupt fall in morale and the strong 
desire to be elsewhere which they (Culicoides) engender, are difficult to 
convey to those who have not experienced them.” The irritation resulting 
from the bites of these small flies has been described as worse than that of 
mosquitoes, and the reaction by some people to the saliva of Culicoides may 


942 THe UNIversiry SCIENCE BULLETIN 


be quite severe. Their small size enables them to pass through netting or 
screens which would normally restrain a mosquito. Several authors have 
cited evidence of depressed development and property value in coastal and 
resort areas where high infestations of these gnats occur. In some areas 
outdoor work is made impossible due to the severity of attack of Culicoides. 
It was probably these conditions which prompted Kettle (1962) to comment 
that while “one midge is an entomological curiosity, a thousand can be 
hell!” 

It was Sharp (1928) who made the first definite association between 
pathogen transraission and Culicoides when he reported the development of 
the microfilariae Acanthocheilonema perstans in Culicoides austent and C. 
grahami in Cameroon, West Africa. This was later confirmed by Hopkins 
and Nicholas (1952); however, Henrard and Peel (1949) and Chardome and 
Peel (1949) doubted the validity of Sharp’s conclusions when they demon- 
strated C. grahami to be an intermediate host of Dipetalonema streptocerca in 
the Belgian Congo. Duke (1954) confirmed the earlier experiments on the 
development of D. streptocerca in wild C. grahami but, in addition, showed 
the flies will take up, at the same time, fully representative numbers of A. 
perstans microfilariae when the host is infected with both parasites. 

Steward (1933) demonstrated the transmission of Onchocerca reticulata 
(as O. cervicalis), the principle causitive agent of fistulous withers in horses, 
by C. nubeculosus in England. This was followed by a report by Dampf 
(1936) of the development of O. cervicalis and O. volvulus in Culicoides in 
Mexico; however, later experimentation by Gibson and Ascoli (1952) failed 
to show Culicoides as a vector of O. volvulus. Buckley (1938) has shown 
Onchocerca gibson, occurring in cattle in Malaya, to be transmitted by the 
bite of Culicoides. 

Culicoides was demonstrated by Buckley (1934) to be an intermediate 
host of Mansonella ozzardi, a common human filarial blood parasite in the 
British West Indies. Robinson (1961) reported the early development of an 
avian filarial worm in C. crepuscularis, which had fed on an infected starling. 
Allergic dermatitis, Queensland itch, of horses in Australia has been shown 
by Riek (1954) to be caused by hypersensitivity to bites of Culicoides brevi- 
tarsis. 

More recently Fallis and Wood (1957) have established that Culicoides 
downest is an intermediate host of the protozoan Haemoproteus nettionis 
occurring in ducks in Canada, and Fallis and Bennett (1960) have demon- 
strated sporogony of H. canachites in Culicoides sphagnumensis. Akiba 
(1960) has incriminated C. arakawae in the transmission of Leucocytozoon 
caullery: to chickens in Japan, and Garnham et al. (1961) have shown 
Culicoides to be involved in the transmission of Hepatocystis kochi in 
monkeys in Africa. 


THe Cuticomwsrs or New Mexico 943 


The transmission of fowlpox virus by Culicoides has been suggested by 
Tokunaga (1937), and bluetongue virus has been shown to be transmitted 
to sheep by the bite of Culicoides by du Toit (1944), Price and Hardy (1954) 
and Foster, Jones and McCrary (1963). The virus of eastern encephalitis has 
been isolated from an unknown species in Culicoides in southern Georgia by 
Karstad et al. (1957). Levi-Castillo has described the isolation of Venezuelan 
encephalitis from Culicoides in Ecuador (cf. Karstad et al. 1957). 

The beneficial aspect of ceratopogonids has been greatly overshadowed 
by their bloodsucking habits. Recent authors (Macfie, 1944; Posnette, 1944; 
Warmke, 1951, 1952; Saunders, 1959; Wirth, 1956) have shown cerato- 
pogonids to be involved in the pollination of cacao and para rubber trees 
in tropical America. Wirth (1956) has shown that two species of culicoid 
flies, C. gamaicensis and C. diabolicus, are involved in rubber tree pollina- 
tion in Brazil and Guatemala. 

As of this time there have been no published reports involving the eco- 
nomic significance of this group in New Mexico. I have accumulated definite 
man-biting records for two species, C. reevest Wirth and C. obsoletus 
(Meigen), and have taken engorged females of C. stellifer (Coquillett) on 
two occasions from light traps located in horse and cattle pens. H. T. 
Miller, Vector Control Section of the New Mexico Department of Public 
Health, informs me that Culicoides constitute a considerable annoyance in 
the lower Rio Grande Valley of southern New Mexico. 


METHODS AND MATERIALS 


Most of the material used in this study was collected at light, either by 
New Jersey light traps, or by illuminating a white, vertically oriented bed- 
sheet with a 15-watt fluorescent tube. The remainder was obtained by 
sweep-net collection and rearing of immature stages. Some discussion of 
the relative merits of the two methods of collection at light might be 
worthwhile. 

By operating the 15-watt light tube from the battery of a vehicle, the 
collector is relatively mobile and may sample isolated areas where collection 
with a standard light trap is impossible. The collector is also able to dis- 
criminate in his collecting, i.e., he can aspirate relatively pure samples of 
small flies from the sheet, thus excluding the beetles, moths, etc., common 
in collections made by light trap, which may damage ceratopogonids. One 
limitation in the use of the light tube is the endurance of the collector. 

It was discovered that the collecting of gnats from the sheet by aspira- 
tion could be supplemented by sweeping with a fine mesh insect net at 
various levels over the light tube. Examination of segregated catches made 
at the same locality, for example, on the Jemez River, Sandoval County, 
showed that only one species, Culicoides hieroglyphicus Malloch, was col- 


944 THe UNIversiTy SCIENCE BULLETIN 


lected by aspiration from the sheet; however, by sweeping at various levels 
over the light tube, small numbers of C. crepuscularis Malloch, C. haema- 
topotus Malloch, C. palmerae James, C. sitiens Wirth and Hubert, C. 
varupenms sonorensis Wirth and Jones and C. n. sp. nr. villosipennis were 
obtained. 

After collection the material was stored in 70% ethyl alcohol. For study 
purposes alcoholic material is greatly preferred over pinned specimens. The 
loss of the color pattern of the mesonotal disc is over-compensated by the 
clarity of structures and the ease with which alcoholic material can be 
manipulated. Dried specimens are very difficult to handle due to their 
extreme fragility, and palps, antennae and legs are easily broken and lost 
during handling. With many of the more common species, identification 
can often be made on the basis of an alcoholic specimen, thereby alleviating 
the laborious task of slide mounting the specimen. Large numbers of 
alcoholic specimens can be stored in a small amount of space as compared 
to the amount of space required for a similar number of pinned specimens. 
Jones (1955) states that 4,000 specimens of a medium sized species will only 
half fall a vial measuring 21 x 70 mm. 

In spite of the advent of the so-called “biological species concept,” the 
criteria for specific and subspecific categories are ordinarily still morphologi- 
cal. Since the morphological discontinuities between individuals or groups 
of individuals must be analyzed and interpreted by the taxonomist, taxonomic 
procedures have been made more and more complex and sophisticated in 
order to reduce the amount of subjective interpretation. 

It has become increasingly apparent in recent taxonomic works involving 
the genus Culicoides that the traditional approach of using wing and 
mesonotal disc patterns to differentiate species is no longer sufficient. Re- 
fined taxonomic methods have shown that many of what were considered 
for many years to be species are actually species complexes. Pointed speci- 
mens, therefore, although they show mesonotal disc patterns, etc., are of 
considerably less taxonomic value than dissected slide-mounted specimens. 
Many important taxonomic characters, e.g., antennal sensoria, mandibular 
teeth, spermathecal structure, are at best not easily discernible on pointed 
specimens but are readily observable on dissected specimens. 

There are several good methods currently in use for slide mounting 
ceratopogonids. The technique employed in this study was to clear the 
specimens in 10°/ potassium hydroxide. The specimen to be mounted was 
first transferred from 70° alcohol to 95 or 100% alcohol, where one wing 
was dissected away and mounted under a separate cover glass, using Dia- 
phane, a synethetic resinous medium. The wing can be removed in 70% 
alcohol and passed through 95 or 100°% alcohol and then mounted; however, 
on several occasions when attempts were -made to transfer the wing to the 
higher gradient of alcohol, it became torn or otherwise damaged. 


Tue Cuticomes or New Mexico 945 


After the wing was removed, the specimen was placed in 10°% potassium 
hydroxide for clearing. I found chemical spot plates useful for clearing 
specimens. The specimen was retained in the KOH solution, which may 
be heated or used cold, until the proper degree of clearing had been achieved. 
The amount of time needed to clear a specimen properly depends on the 
amount of sclerotization, as well as size, and can only be learned through 
experience. Next, the specimen was washed in distilled water to which a 
drop of 10% glacial acetic acid had been added to stop the macerating process. 
After washing, the specimen was dehydrated in absolute alcohol and trans- 
ferred to a drop of mounting medium on the slide, where the head and 
abdomen (or genitalia, if the individual is a male) were dissected away from 
the thorax under a wide field microscope. On the slide, the head was 
oriented dorsal surface up, the antennae and palps arranged and the mouth- 
parts manipulated in such a fashion as to make the mandibular teeth visible. 
Special attention must be given to mounting the head and genitalia in order 
that they present a good symmetrical view. The coverglass was then ap- 
plied, the slide labeled and then put away, preferably in an oven, to dry. 

The use of Diaphane instead of balsam is advantageous, since it does 
not become tacky during arrangement of the dissected parts of the fly on 
the slide and permits mounting of the wing directly from alcohol. 

A second technique for mounting ceratopogonids, preferred by Dr. Willis 
Wirth, is the phenol-balsam method. Specimens are placed overnight in a 
warm saturated solution of liquid phenol, prepared by mixing phenol 
crystals in absolute alcohol. They are then transferred to a solution con- 
taining equal parts Canada balsam and liquid phenol and, after standing 
a few minutes, are transferred to the slide for dissection. Small pieces of 
broken coverglass are added to prevent excessive flattening by the coverslip. 
The slides are then oven dried, and pure balsam is occasionally added to 
replace evaporating phenol. This method allows the preparation of large 
numbers of specimens at one time. 

Many times, when attempting to observe the pattern of pale spots on 
the wings of Culicoides, one is handicapped by the lack of contrast between 
the pale spots and the remainder of the wing. I have found it very useful 
to observe the wing pattern by a false dark field illumination of a phase 
contrast microscope. This is accomplished by turning the phase condenser 
one or two stops too far while the low power objective is in position. With 
this false dark field microscopy the pale spots, which are due to the scarcity 
or absence of microtrichia, appear distinctly dark against a white background 
and are easily delimited (Figs. 1, 2). Even very small or indistinct spots 
are evident when viewed under these conditions. This technique, when 
supplemented by bright field illumination for studying the color pattern, 
presents an excellent method for examining Culicoides wings. 


Tue University ScriENcE BULLETIN 


Fic. 1. Phase contrast dark field of wing of Culicoides byersi, new species. Fic. 2. Phase 
contrast dark field of wing of Culicoides sublettei, new species. 


Types of the species described as new in this paper will be deposited in 
the United States National Museum, and as many paratypes as possible will 
be deposited in the Snow Entomological Museum of the University of 
Kansas and the American Museum of Natural History. A determined series 


of the New Mexico species will be deposited in the Snow Entomological 
Museum. 


All measurements used were made with an ocular grid calibrated with a 
stage micrometer and most were taken from slide-mounted specimens which 
had been cleared in potassium hydroxide or from specimens cleared in 


Tue Cuticomwes oF New Mexico 947 


phenol. Measurements, when possible, were made of a series of specimens 
so as to give some insight into the variation of the structures. The results 
are presented as the mean followed (in parentheses) by the range of varia- 
tion and N, the number of specimens measured; for example, 1.68 (1.5-1.8; 
N=3). Measurements not given in this fashion refer to single specimens. 

Drawings were made with the aid of a Bausch and Lomb V-H micro- 
projector. No attempt was made to correct the symmetry of most of the 
drawings, since this could have led to distortion of some of the structures. 

The illustration of the dark and pale wing spots in this paper was achieved 
by using Craf-Tone, a commercial stippling paper. This technique is limited 
in that one cannot show decreasing or increasing intensity of the wing color. 
This, however, is not of great taxonomic value and the ease by which this 
material can be applied compared to stippling by hand greatly overshadows 
this limitation. 


MORPHOLOGY 


The head is subspherical, anteriorly flattened and possesses a broad 
postocciput. The vertex is undifferentiated and usually contains scattered 
setae. The eyes are large reniform structures and are bare or rarely have short 
pubescence among the ommatidial facets. The eyes may be contiguous or 
separated; the degree of marginal contact or separation is of importance in 
distinguishing some species. The ocelli are more or less poorly differentiated. 
The frons can be arbitrarily delimited posteriorly in most species by a 
transverse transocular suture, which arose secondarily. It probably gives 
support in the area of the eyes and is quite evident, particularly in species 
with rather widely separated compound eyes. 

One can readily postulate that if this secondary suture functions for 
support of this area, the appearance of a second suture with a similar func- 
tion is possible. Indeed, this appears to be borne out in certain species of 
Culicoides, e.g., C. sublettei n. sp., and C. stellifer, in which there is a 
small interocular suture anterior to the transocular suture. 

The frontoclypeus surrounds the antennal bases in the region of the 
frons and is expanded anteriorly to form the convex clypeal region. Its 
strongly incurved border is joined with the labrum-epipharynx by two 
small, triangular sclerites designated by Peterson (1916) as tormae. An 
important taxonomic character involves a ratio of the head and proboscis. 
The length of the proboscis is measured from the distal end of the labrum- 
epipharynx to the anterior margin of the tormae. This length is divided into 
the length measured from the anterior margin of the tormae to the median 
hair socket. 

Jamnback (1965), in his paper on the Culicoides of New York, has de- 
fined this character as given here; however, he has inverted the fraction when 


948 Tue UNiversiry SCIENCE BULLETIN 


applying it to a description and said that a “proboscis-to-head ratio” of less 
than 0.65 should be indicative of a short proboscis; however, it should indi- 
cate a long proboscis. Therefore, no comparison could be made with regard 
to this character between eastern and southwestern representatives of 
several species. 

The biting apparatus includes the labrum-epipharynx, mandibles, maxillae, 
labium and the hypopharynx. The mandibular, hypopharyngeal and maxillary 
stylets are enclosed in a labial gutter, covered dorsally by the labrum; the suck- 
ing apparatus consists of cibarial and pharyngeal pumps (=pharynx and 
oesophageal pumps, respectively, of Jobling, 1928). The mandibles of the 
females are thin blades, finely toothed on the distal margins, the number of 


| 


mandibular teeth being of value in separation of closely related species. | 


Various authors have shown a device which interlocks the mandibles to 
give a scissor-like appearance. The endite lobe of the maxilla, usually ex- 
tended as a maxillary stylet, is believed to be the galea or fused galea and 


lacinia. On the basis of musculature, the action of the maxillae appears to be — 


protraction and retraction. The labium and hypopharynx are distally toothed. 


The mandibles of the females in the majority of blood-feeding species in 
the genus Culicoides are prominently denticulate. The females of other © 


species have weakly developed mouthparts and are apparently incapable of 


taking a blood meal. Jamnback (1965) has illustrated both hematophagous | 


and non-hematophagous types of mouthparts. 

The maxillary palps are five-segmented, the first segment short and in- 
completely sclerotized, the third segment large and more or less swollen. The 
third palpal segment bears on the mesal surface a specialized sensory organ 
which contains many small sensillae. Barth (1961) described the histological 
aspects of this organ in Forcipomyia, a genus related to Culicoides. 

The maxillary palps of females possess characters, such as the ratio of 
lengths of segments and the shapes of the third segment and sensory pit, 
which have been used by many workers for the delimitation of species. The 
palpal ratio (PR) is obtained by dividing the length of the third palpal seg- 


ment by its greatest breadth. This measurement is subject to variation, par-. 


ticularly in slide-mounted specimens, due to the orientation of the palpus on 
the slide; however, it is still of importance in the separation of some species. 
The relative lengths of the palpal segments is another useful character. In 
this paper the basal segment is not included in palpal measurements, since 
its degree of sclerotization appears to vary among individuals of the same 
species. 

The mouthparts of the males, although structurally similar to those of the 
females, are less well developed. They have no distal mandibular teeth; the 
mouthparts are sot fitted for piercing; and the maxillary palps are not as 
well developed as in females. 


Tue CuricoiweEs oF New Mexico 949 


The antennae provide some of the most useful characters for distinguish- 
ing species. They are divided into 15 units: a small ring-like scape, which is 
more or less hidden by the enlarged pedicel, and a flagellum composed of 13 
sub-segments. In various instances in the remainder of this paper the 
antennal sub-segments, or flagellomeres, will be referred to as “segments.” 
In the females the proximal eight flagellomeres are short, being slightly longer 
than wide, while the distal five are elongate, usually much longer than wide. 
Segment 3 is elongate and somewhat larger than 4-10. Segments 3-10 in the 
female possess long verticils. In males the pedicel is more enlarged than in 
females, and the transition in flagellar structure occurs between segments 12 
and 13. Segments 3-12 each have dense verticillate hairs, giving the male 
antenna a plumose appearance. 

The antennal segments contain various types of sensory organs, one of 
which is very important taxonomically. Jobling (1928) first demonstrated 
the presence of small pits on various segments and referred to them as 
olfactory pits. These pits, which are surrounded by a number of minute 
setulae, are found on the third segment of both sexes and various other 
segments in females and some males. First used taxonomically by Ortiz 
(1951), their pattern of distribution is more or less consistent for a species and, 
therefore, very useful in the separation of species and species groups. An- 
tennal sensillae (sensu Jobling) are usually evident under low magnification 
in slide-mounted preparations. A second type of sensilla on the antennal 
segments has been described by Campbell and Pelham-Clinton (1960). Ac- 
cording to these authors, small flask-shaped structures occur on various an- 
tennal segments and correspond to the sensilla coeloconica described by 
Snodgrass (1935). No taxonomic significance has yet been assigned to the 
sensilla coeloconica. The use of the term sensilla in this paper, unless other- 
wise stated, refers to the “olfactory pits” described by Jobling. 

Jamnback (1965) has shown some correlation between the number of 
antennal olfactory pits and the host preferences of the females. He has 
pointed out that females of ornithophilic species appear to have more olfac- 
tory pits than mammalophilic species preferring large mammals. 

The antennal ratio is obtained by dividing the combined lengths of the 
distal five segments by the combined lengths of the preceding eight. 

In addition to the antennal ratio, another antennal character is more or 
less constant for a species. This relationship involves the length of segment 
11 divided by the combined lengths of segments 9 and 10. 

The thorax is dorsally convex and protruding anteriorly, so as to cover 
the posterior region of the head slightly. The mesonotum possesses a pair of 
large and distinct sensory or humeral pits, which are located near the an- 
terior margins of the humeri and are possibly remnants of the pupal respira- 
tory horns. A pair of faint lines, the pseudosutural foveae, extend posteriorly 
from the pits to near the sides of the scutellum. The large, somewhat flattened, 


950 THe UNIversiry SCIENCE BULLETIN 


caudal portion of the mesonotum, the prescutellar depression, possesses a 
pair of sensory areas known as the prescutellar spots. The postscutellum is 
bare and arched. In many species the mesonotal disc is ornamented by a | 
distinct pattern, which along with various other thoracic markings, is of — 
use in species differentiation. 

The legs are slender; however, the femora occasionally may be slightly _ 
expanded, but are not spinous and bear no scales, as in many of the other 
ceratopogonid genera. The hind tibiae bear a comb of spines, which is | 
sometimes useful in differentiating species. The basal tarsomere is at least 
twice as long as the second, and the fourth tarsomere is shorter than the 
fifth. The shape of the fourth tarsomere is important in distinguishing the | 
subgenus Macfiella Fox and some other unrelated species. The pretarsal — 
claws are small, simple, and equal in both sexes and a very minute empodium | 
arises from the apical end of the unguitractor. | 

The wings possess two types of hairs: macrotrichia, which are long hairs i 
whose abundance may vary between species, and dense microtrichia, whose 
presence or scarcity at a particular location makes a pattern of light or dark 
spots, which are of great importance in the recognition of species. These pale | 
spots can be very stable in position and distinctiveness in some species, but | 
certain spots may be highly variable in others. | 

The costa extends approximately one-half the length of the wing, the 
proportion being a useful taxonomic character in some species. There are two 
radial cells which are usually of subequal length, the distal cell being broader 
than the slitlike proximal cell. The nomenclature of the wing veins used 
in this study follows the Tillyard modification of the Comstock-Needham — 
system, in which the anterior branches are Mi and Mb, and the posterior 
branches are Mz +4 and Cur (Fig. 3). 

The terms vannal vein and vannal cell used in this work are considered 
to be morphologically more correct than the terms anal vein and anal cell 
used by previous authors. The vannal portion of the wing refers to the fan- 
shaped area delimited basally by the third axillary sclerite and anteriorly by 
the vannal fold. Although the term anal vein is firmly entrenched in the 
taxonomic literature of this group, it is incorrect, according to current mor- 
phological thinking and is in need of change. 

The wings of the males are usually more elongate than those of females, 
and the pattern of light spots usually lacks the contrast seen in the female 
wing. The measurement of wing length was made from the basal arculus 
to the tip. This measurement appears to be more accurate than the practice 
of adding one-seventh to the wing length to account for that part of the 
wing basad to the basal arculus, utilized by some workers. 

The external features of the female abdomen are usually of little taxonomic 
importance and are easily discolored as a result of engorgement. The internal 
abdominal structures are, however, of paramount importance. The sclerotized 


Tue Cuticoiwes of New Mexico 951 


2nd RADIAL CELL 


Ist RADIAL CELL 


COSTA 
VA RADIUS 
Ss) ——— 


ee =e eee he 
ae 


Cu 


rom a a 2 
7 eee 7 


VANNAL CELL 


SPERMATHECA 


RUDIMENTARY SPERMATHECA 


NECK \ \ 


RING 7 


TELOMERE 
APICOLATERAL PROCESS 9th TERGUM 
AEDEAGUS = BasinerE 
Q | CLASPETTE 
Uf x ‘ee VENTRAL ROOT 

DORSAL ROOT: —— 


Po = oS MEMBRANE 
9th STERNUM =e 


Fic. 3. Wing of Culicoides. Fic. 4. Spermathecal system of Culicoides. Fic. 5. Male 
genitalia of Culicoides baueri Hoffman. 


parts of the spermathecal system (Fig. 4) are very significant, and females of 
many species cannot be accurately identified without reference to it. The 
spermathecae may or may not be sclerotized and may vary in number from 
one to three, the number being constant for a particular species. Those species 
having two functional spermathecae usually have a rudimentary sclerotized 
spermatheca, which is rarely expanded to resemble a small third functional 


952 Tue UNIvERsITY SCIENCE BULLETIN 


spermatheca. The spermathecal ducts may be sclerotized for some distance 
distally but are unsclerotized proximally near their connection to the bursa 
copulatrix. The junction, which forms the union of the spermathecal ducts 
and the bursa, may be marked by a small sclerotized ring. Length of the 
spermathecae was measured from the distal end to the proximal end of the 
sclerotized portion of the duct, and width was measured at the widest point. 

The abdomen of the male is considerably more slender than that of the 
female. The genitalia of the male (Fig. 5) are of primary importance in the 
recognition of species and higher taxa. In this study I have followed the re- 
vised interpretation of the male genitalia proposed by Snodgrass (1957, 
1959). This concept is in agreement with the earlier morphological ideas of 
Crampton (1942) and more recently the studies of Matsuda (1958). It is 
supported in the Diptera by the embryological studies of Christophers (1922), 
Christophers and Barraud (1926) and Abul-Nasr (1950), and in my opinion 
is more plausible than the earlier idea of appendicular origin of genitalia. 

Taxonomists of the Ceratopogonidae have long chosen to follow a 
terminology for the male genitalia that is, in many instances, without mean- 
ingful morphological basis. This condition, however, is not limited to the 
Ceratopogonidae. Crampton (1942) pointed out some of the confusion that 
exists regarding the nomenclature of various structures and made a plea 
for uniformity. This uniformity has been met in some groups by the use 
of taxonomically adequate, but morphologically inaccurate, terms. Although 
there is something to be said for uniformity of nomenclature, one should 
strive for morphological accuracy to facilitate uniformity in concepts of 
homology. 

The ninth abdominal segment of the male is in the form of an irregular 
scleroma composed of the fused tergum and sternum. The ninth tergum is 
a posteriorly tapering, platelike expansion, the posterior margin of which 
may or may not bear apicolateral processes. The ninth sternum is a narrow 
structure, usually emarginate on the posterior surface. The outermost genital 
claspers, the parameres, articulate on the ninth sternum and are divided into 
a proximal basimere and a distal telomere (basistyle and dististyle, respec- 
tively, of authors). The sufhx “-style,” used by many authors in the ter- 
minology for the outermost male genital claspers, invokes the old idea of 
the appendicular origin of the male genitalia and is not consistent with the 
terminology proposed in this paper. The term “paramere” has been affixed 
to many different structures of the genitalia, but in reality is the lateral 
genital clasper. Some workers have gone so far as to suggest abandoning this 
term because of the confusion as to its true identity. The basimere (Fig. 5), 
in the generalized condition, has two internal processes on the anterior sur- 
face, the inner ventral root and the outer dorsal root. The telomere generally 
has an expanded base and narrows distally and may have a curved-tipped 
apical spine. The claspettes (parameres of authors) are usually elongate with 


Tue Curicomwes oF New Mexico 953 


an expanded base and a simple, slender to greatly modified apex. In the 
subgenus Selfia Khalaf, the claspettes are fused into a platelike structure. The 
aedeagus is usually a V-or Y-shaped structure, less sclerotized distally, with a 
pair of basal arms. Its basal arch is usually curved and is connected to the 
ninth sternum by a membrane, which may or may not be spiculate. 


SYSTEMATICS 


The Ceratopogonidae show morphological affinity to the Chironomidae 
but differ in that they possess complete mouthparts, having developed 
mandibles in both sexes, and a branched Mi+2 wing vein, and a_post- 
scutellum lacking a median furrow or keel. The legs of ceratopogonids ap- 
pear stouter and the metathoracic pair is usually longer, as is the prothoracic 
pair of chironomids. Edwards (1939) is of the opinion that the ceratopogonids 
have much in common with the Simuliidae. 

The genus Culicoides can be differentiated from other ceratopogonids by 
the presence of two more or less equal radial wing cells, small and equal 
tarsal claws in both sexes, the absence of a hairy empodium, and large pro- 
nounced thoracic humeral pits. 

Root and Hoffman (1937) proposed the first supraspecific categories of 
Culicoides when they divided the North American species into two series on 
the basis of male genitalia and certain external characters. Edwards, in his 
monograph of the British species (1939), was able to substantiate such a 
division and placed the European species into two apparently natural groups 
on the basis of the male genitalia. 

Fox (1948) proposed the subgenus Hoffmania for 12 neotropical species, 
which he distinguished from Culicoides s. str. The known species of Hoff- 
mania were later reviewed by Ortiz (1950), who expanded the subgenus to 
include several additional species. 

Khalaf (1954) made the first major attempt to establish interrelationships 
and deduce phylogenies among species of Culicoides. Using primarily ex- 
ternal male genitalia and female spermathecae, he was able to divide the 
known world fauna into four subgenera: Culicoides s. str., Monoculicoides 
Khalaf, Oecacta Poey and Se/fia Khalaf, lumping the subgenus Hoffmania of 
Fox into the subgenus Oecacta Poey. Khalaf further subdivided these sub- 
genera into complexes and species groups, and although his concepts were 
fundamentally sound, he was handicapped in that he had to draw much of 
his information from the literature, which was, no doubt, vague in many 
instances. Many of the species descriptions in the literature made little refer- 
ence to the genitalia; therefore, the true relationship of these with other species 
could not be accurately determined. 

Vargas (1953) erected the subgenus Beltranmyia for the crepuscularts 
group, which Khalaf had included in Monoculicoides. Fox (1955), in his 


954 Tue UNiversiry SciENCE BULLETIN 


catalog of bloodsucking midges, arranged the species of Culicoides known to 
occur in the Americas into subgenera, introduced the subgenus Macfiella for 
the phlebotomus group and Avaritia for the obsoletus group, and augmented 
the subgenus Beltranmyia as Vargas had outlined it. In spite of these changes, 
the subgeneric classification of Fox is not entirely in agreement with the 
earlier attempt by Khalaf with respect to the content of the various subgenera. 

More recently Wirth and Hubert (1959) have proposed the subgenus 
Trithecoides for the Ethiopian and Oriental species whose females have three 
well-developed spermathecae and a long second radial cell. 

Vargas (1960), attempting to correlate genitalia and wing characters, 
proposed the following new subgenera for the species occurring in the 
Western Hemisphere: Anilomyia, Diphaomyia, Drymodesmia, Glaphiromyia 
and Mataemyia. In addition he resurrected the name Haematomyidium 
Goeldi 1905 for many species usually placed in the subgenus Oecacta. 

Incorporating two of the subgeneric names of Vargas (1960), i.e., Dry- 
modesmia for the copiosus group and Diphaomyia for the iriartet-baueri 
group, with previously known names, Wirth (1965) arranged the North 
American species into nine subgenera. In my opinion the classification by 
Wirth (1965), which differs from that of Vargas (1960) in the content of 
several of the subgenera, reflects more correctly in many instances the natural 
afhinities of the genus. It is the classification followed in this paper. 

Some subgenera, e.g., Oecacta, in the past have been considered as “catch- 
alls” for a rather heterogeneous accumulation of species. Attempts have 
recently been made to divide various subgenera into much smaller units. 
Although in some instances such divisions may be useful in pointing out 
interspecific relationships, there is little to be gained at present by assigning 
subgeneric names to a large number of species groups that would contain 
only a few species. As more information becomes available, much shifting 
of taxa may become necessary. It is only after more distributional data are 
available and extensive taxonomic studies, such as that of Wirth and Blan- 
ton (1959), have been made that more accurate conclusions regarding 
phylogenetic relationships can be drawn. 

Further studies, such as those of Jones (1961), are needed to correlate the 
classification of the immature stages with that of the adults. These studies 
should reveal interspecific relationships not evident in classifications based on 
adults. Several of the newer systematic techniques, e.g., numerical taxonomy, 
should show considerable promise at the subgeneric level. 

At the outset of this investigation nine species of Culicoides were reported 
in the literature as occurring in New Mexico (Wirth, 1952, 1965; Foote and 
Pratt, 1954). Extensive collecting has increased this number to 35. Nineteen 
species can be assigned to previously existing names, the remainder are con- 
sidered as new to science. Adequate material is available to describe four of 
these as new species. It is hoped that subsequent collection will allow descrip- 


Tue Cuticowes or New Mexico 955 


tion of the remaining species. One species, C. pamporkilus Macfie, previously 
reported from Mexico, Panama and Venezuela, was taken during this study 
at Ruidoso, Lincoln County, and is a new United States record. 

The Culicoides fauna of New Mexico can be divided into eight sub- 
genera. In the descriptive portion of this paper many species, particularly 
those placed in the subgenus Oecacta, are listed, when applicable, with the 
name of their respective species group. The following table gives the syste- 
matic arrangement of New Mexico species together with a summary of the 
mean values of a number of quantitative characters. The range of variation 
is given in the respective species descriptions. 


Taste 1. Systematic arrangement and summary of quantitative characters of 
New Mexico Culicoides. 


= ean 
= Bn: ao 
g@ wide = § 4 E 5 
= 00S, Bo 5° 552s 5 § Bo aes 35h 8 
g Se 28 G8 Bea, 88 S25 382 oat oe 
a SS S62 <f@ adhdn BE Age DSe Zen 2a 
Subgenus 
Avaritia 
obsoletus ....... gly) 0.60 1.14 0.72 Sole 2./4 MOD 2-165 
Subgenus 
Beltranmyia 
crepuscularis . 1.41 0.56 1.35 0.90 3-14(15) 2.27 1.05 10-16 4 
Subgenus 
Culicoides 
cockerellu _..... 1.84 0.60 1.07 0.71 SS ie 3.26 1.05 14-16 6 
9-15 
Subgenus 
Diphaomyia 
DBUUCTS eee 1ES35) 0.52 1.07 0.66 3, 5-10 2.17 10-12 4 
haematopotus 1.17 0.55 1.43 0.91 Shays 25) 2-1 4 
9-15 
Subgenus 
Drymodesmyia 
GHODG! ecient 1.34 0.51 1.07 0.69 3-9, (10) 233 Ot: 4 
11-15 
byerst 0.89 0.54 118 0.73 Ses ey G7, OE 10) 34 
9, 11-15 
Subgenus 
Monoculicoides 
varipennis 
SONOreNSIS ...... 1.55 0.55 0.86 0.59 3,4) 6) 2.38 IE WEIS “6-7, 
(6) (7), 8-10 
Subgenus 
Oecacta 
doeringae ...... 124, 0.58 1.63 ei BOs He 2.61 NEA ARIS; x 
9, 11, 13-15 
luciana N22 0.53 0.95 0.59 3-10 PEMA 1.05 13-16 4 
novamexicanus 1.23 0.56 1.02 0.58 3-5 aos ely? 1.04 14-16 4 
Mit =15 
oklahomensis .. 1.21 0.56 129 0.82 Beis 3.0 1.04 14 5 
9, 11-15 
palmerae ........ 1.48 0.56 1.28 0.77 3-15 2.26 1.02 14 4 
pampotkilus .. 1.52 0.54 155 0.89 3 5 2.0 132 iS. 


9, 11-15 


956 


THe Universiry ScrENCE BULLETIN 


Taste 1. Systematic arrangement and summary of quantitative characters of 


New Mexico Culicoides (concluded). 


G&G 
Eb ae Ry a S 4 
4a, 1,218. dp gc 2a 
st DD Ep SO oO Ono Si 90 So WO FOR ie 
z eG 8S se sah G8 22 SS Gbemecmeea 
a PS Of <a wchdn 4A aga DLeSe Zenza 
pecosensis ...... 122 0.54 125 0.81 327/,.), 2.48 0.93 17-18 5 
11-14 
VCCUCST eat ees 0.85 0.56 0.65 0.33 3, 8-10 Jel LG aa Gt 
Grellij cimee eG 0.54 0.96 0.62 3, 7-10 Del 222 
SLOLCI eee IG) 0.56 1.10 0.69 3-14(15) Pes Oey WARNS) 4 
SUDICTLCD ee 1ELS 0.55 1.07 0.67 3, 11-15 223i, O54 = Ose 
utahensis _...... IETS. 0.55 Hel 0.75 3-15 1.85 1.30 ler 4 
Subgenus 
Selfia 
brookmani .... 1.14 0.53 0.92 0.62 3-10 1.97 139 Hil 4 
hieroglyph cus 120 0.54 1.03 0.66 3,5-10 2.17 I24eenheremees 
(GIVES 27, 0.56 1.0 0.63 3-10 26 127 Wels aA> 
KEY TO FEMALES OF NEW MEXICO CULICOIDES 
1. Spermathecae unsclerotized; wings without pattern of light 
and dark spots... :2:230.2-.t.aedi eee 2 
Spermathecae sclerotized; wings with or without pattern of spots. ...... 3 
2..S€nsora oneantennal segments 3-10. 422s. =e brookmani| 1 |* 
Sensoria ‘on antennal, sepments.3, 9-10). 2. jamesi| 3 | 
SMe to5 5 pee a eee. ce OP an Ce On See Oe eee hieroglyphicus| 2 | 
3(1). Second radial cell with distal portion in a pale area (Fig. 18, 23) 
€yes CONCISUOUS: <4 tsps te ee A ee 4 
Second radial cell in a dark area; eyes may cr may not be contiguous... 5 
4(3). Wings poorly marked; sensoria on segments 3, 11-15; hind tibial 
comb with 5 spines; a small to medium sized species, 
Wile 0-1 Sema. -44i0 01. Bs aecents Oats ae, aan es see Owen obsoletus|5 | 
Wings well marked, with 3 transverse bands (Fig. 18) which may 
be reduced in some forms; sensoria irregularly distributed, usually 
3, 5, 7, 9, 11-15; hind tibial comb with 6 spines; a large species, 
Wie, 15622 Osmim sea ee A eee ee cockerelli| 4 | 
5(3). Onewspermathécas 2.0 a. ee ee 6 
Two spermathecaé;:. «2.22 8 
6(5). Wing pattern with irreguiar gray streaks (Fig. 33); hind tibial 
comb with 6-7 spines; spermatheca C-shaped. ..........-.-......-.- variupennis|7 | 
Wing pattern with definite pattern of spots; hind tibial comb with 
4, spines; spermatheca not as above: _2.1....c.c i] 
7(6). Antennal segments 9+-10 to 11 in ratio of 0.90; sensoria on 
segments 3-15> a large species, Wit) Ile) eae crepuscularis| 6 | 
Antennal segments 9+-10 to 11 in ratio of 0.30; sensoria on 
segments 3, 8-10; a'small species, Wil) 0!Siamms ee reevest| 17} 
8(5). Hind tibial comb with! 5-6:spines........22 2) 
Hind tibial comb with 4 spines: 2.....2..--. 11 


* Number in brackets refers to location in species descriptions. 


10(9). 


11(8). 


12(11). 


13(11). 


14(13). 


15(14). 


16(15). 


17(16). 


18(15). 


19(18). 


20(18). 


THe Cuticowrs oF New Mexico 957 


. Apices of veins My, My, M344 dark; cell R; with inverted 


U-shaped pale spot (Fig. 69); sensoria usually on segments 


By TI gael ARS ie Alaa AN AR AM Rests 2 oer het a ec Loh Eee stellifer| 13 | 
Apices of veins M;, Ms, M344 pale; cell R; not as above; sensoria 
Ontvarious proximal-sesments and om (i-loy se se eee 10 
Vein Cu, bordered by a pale spot for at least part of its 

Hee rm rt lama Gb ey SL (5: yates, et 2s NS Ae oklahomensis| 19 | 
Weim @unmdarka@E ig. 99)),'..22 3 eee Oe Seen ee pecosensis| 18] 
Waneawithout distal pale spots (Bigs) 775, $2)-ee = ee 12 
Wanesewathn distal pale spots..2 0 fi. 2 2 ee 13 
Wings completely lacking pattern of spots; sensoria on 

SECMMCMeSH SMa (Cls) an. oe eh ee ce heen oe nee Seca eee stonet| 14] 
Wings with very faint markings; sensoria on segments 

3, [Ugh ee ee enn eee nr ar sik ely tome to, sublettei| 15 | 
Spermathecae very unequal, lacking necks (Fig. 122); antennal 
segments 9-+-10 to 11 in ratio of at least 1.0; sensoria on segments 

3), Ds TPS i TR CIUDAD a Ne 3 [> ee en Benepe tees doeringae| 23 | 
Spermathecae equal or subsequal, necks usually present; segments 

9+-10 longer than 11, in ratio of less than 1.0; sensoria various. ............ 14 
C@elleReswith 3spale spots (Pigs 88). 2.2... pam potkilus| 16 | 
Gell Rs savatit 2 pale+ Spots, 52.26.20 aedels su nie te 15 


Pale Bands absent on femora; eyes separated by approximately 
diameter of an ommatidial facet; macrotrichia abundant over 
@MMelnic: SUITLAGE OlesW ANOS elec 2s cea aoe oe 16 
Some or all femora with pale bands; eyes usually separated less 
than diameter of ommatidial facet to contiguous; macrotrichia 


usually confined to distal and posterior regions of wing. -..............-.- 18 
Sensonia von seements 5-55 7,9, ll yi3-15. 2 ee novamexicanus| 22 | 
Semsomia Oi Seoments: S= 19% si. ccna nskscctsessecetecnscesse~sconscksteeeee ee ee 7! 
Proboscis long, HR 0.8-1.1; wing well marked; third palpal 


segment moderately swollen, PR 2.0-2.4; spermathecae without 
TE Kegupenene nana Vee to aris a ot aD ig eee ee palmerae|20| 
Proboscis of medium length, HR 1.30; wing poorly marked; third 
palpal segment greatly swollen, PR 1.85; spermathecae with 


TNC Ke oaN Re st fst ie OSG 2 tee 2d Le ee eee utahensis| 21 | 
Sensoria on segments 3, 5-10, 3, 7-10, or 3-10, never on distal 
fHRVEMSE RINE IMCS! ace cccb AN wo tc Ln deca eee 19 
Sensoria om scoments, 5, 5, /,95(10)| Ll-l>ion 3-15. Fe 20 
Spots in cells My, Ms and My, small, round, well removed from 


wing margin (Fig. 55) veins Mz +4 and Cu, bordered by pale 
area; spot in cell R; constricted mesally; sensoria on segments 3, 
G); (6), 7M aa aaasececcececeececcrassesessseteneeteeesbenssesesnessrersecerecsonssteaiseeeee bauert| 10] 
Spots 1 in cells M;, Ms and Mg large, those in My and My attaining 
wing margin (Fig. 66) veins M3+4 and Cu, not as above; spot in 
R; large, not constricted; sensoria on segments 3-10. ._.............- luglani|12| 
perenne segments 9-+-10 to 11 subequal, in ratio of 0.83-1.0; 
AR 1.25-1.57; pale spot in cell R; small at distal end of cell 
(LSI Sas (0) cal ee Reena me Pear eeP APR tS 7 coi, OT ee haematopotus| 11 | 
Segments 9+-10 to 11 unequal, in ratio of approximately 0.69-0.75; 
AR not more than 1.15; pale spot in cell R; mesally located. ............... 2] 


958 


21(20). 


5(1). 


6(5). 


7(5). 


BAY 
9(8). 


10(9). 


11(9). 


Tue UNIversity ScIENCE BULLETIN 


Proboscis very short, HR 1.5-1.7; pale spot on basal portion of 
vein M, absent (Fig. 49); a small species, WL 0.89 mm. ............ byerst| 9 | 
Proboscis of medium length, HR 1.0; mesally constricted pale 
spot in cell R;, attaining anterior wing margin; pale spots over 
basal portions of veins M; and My (Fig. 44); a medium sized 
species, Wi. 1.3) mamas o.oo e-.ceec cote caste eee sitiens| 8 | 


KEY TO MALES BASED UPON GENITALIA* 


. Claspettes fused for part or all of length (Figs. 12, 38) 2s ) 
Claspettes free and separate for entire length. 22 ee 5 
. Claspettes fused basally, apices free; aedeagus with bifid tip 
(Big 136 \)s ts a oS rae eee eee ee eee eri eee varupennis| 7 |** 
Claspettes completely-fused over entire length. 2 ee 3 
. Ninth sternum without mesal notch or lobes; telomere un- 
modified; boomerang-shaped sclerites present. ...........-.--.- brookmani|1 | 
Ninth sternum with prominent lobes or mesal notch; telomere 
modified apically; boomerang-shaped sclerites absent. —.......-.---------. 4 
. Ninth sternum with prominent caudo-median lobes; aedeagal arms 
widely <separated pit 8... <M pees tester eee hieroglyphicus| 2 | 


Ninth sternum with medio-posterior notch; aedeagal arms not 
widely separated; telomere distinctly modified with large foot- 
shaped apex (ENG 17) jee: oc ee jamesi|3 | 
Ninth tergum with apicolateral processes absent or very small 
(Figs. 22, 27); if small, medio-caudal margin very convex and 
extending posteriorly beyond tip of apicolateral processes; mesal 


notch of pesterior border of ninth tergum absent. -_ 6 
Apicolateral processes present and usually well developed; mesal 
motch-presemts 2 5.05 ak oot eee ae cee eee y) 


Ventral root very long and slender, much longer than dorsal root; 
ninth sternum with deep mesal notch (Fig. 27); apicolateral 
[EOGESSES ASC mb oes soc ces see ee assets cea on ee obsoletus|5| 
Dorsal and ventral roots subequal in length; ninth sternum with 
only slight emargination; ninth tergum with apicolateral process 
very small, medio-caudal margin very convex (Fig. 22). .... cockerellii| 4] 
Claspette with distinct apical and subapical spines (Figs. 64, 70, 


87, 122), ee Be Ie 8 
Claspetterwithisimple apex (Figs: 32) 045) 15) oe ee 13 
Ventral root with process on posterior margin (Figs. 65, 129). —......-.-- 9 
Ventral ‘root lackime process: sees sa ese tee ce ee ee 12 


Aedeagal arms with blade-like processes on posterior margin (Fig. 65). 10 
Aedeagal-anms without blade-like processes./2 2st eee eee 1] 
Claspette with distinct thumb-like lobe on lateral margin, apex 
greatly expanded with row of prominent spines. ........ haematopotus|11| 
Claspette lacking thumb-like lobe, apex not expanded. ............ bauert| 10] 
Claspette with small lateral lobe at approximately two-thirds of 
length; aedeagus in shape of inverted V, basal arch mesally 
notched, median posterior process short, conical. .............-.---- stellifer{ 13] 


* Male of reeves: unknown. 
** Number in brackets refers to location in species descriptions. 


12(8). 


13(7). 


14(13). 


15(14). 


16(15). 


17(14). 


18(17). 


19(18). 


20(19). 


21(19). 


Tue Cuuicoiwwres of New Mexico 959 


Claspette tapering to distal point bearing row of spines along 
margin; aedeagus with elongate, truncate median process, basal 
areiaroumdcd’..t2 <0 cee ce NN ee ees oe doeringae|23 | 
Claspette with fringe of spines, stem with prominent blade-like 
process on lateral margin (Fig. 69); median process of aedeagus 


with pointed lateral subapical processes. ...........00-.10-s00s000-00--00---- luglani| 12 | 
Claspette with 2-3 subapical spines, stem simple (Fig. 86); 

aedeagus truncate, lacking subapical projections. .................... sublettei[ 15 | 
Ventral root greatly reduced or absent (Fig. 33); membrane 
Spiculatese tes. 2 ee lh Se Ee ee eee crepuscularis| 6 | 
Ventral root present and distinct; membrane may or may not be 

S[SICUIE IC 2 te ee ee MMM ee Meme NN SNe poe 14 


Telomere bulbous basally, abruptly bent (Fig. 109); apicolateral 
processes distinct, of uniform width for most of length, much 
longer than wide; medio-caudal border of ninth tergum convex 
MeheGistiinct mesal moteh. =... ys 2 ee 15 
Telomere not abruptly bent; apicolateral processes usually  tri- 
angular; medio-caudal border of ninth tergum not distinctly convex. . 17 
Aedeagus with lateral shoulder-like processes on median posterior 
process (Fig. 112); claspette with tapering, usually recurved 


CDSN cea, ae Se eR I ere gc ene palmerae|20| 
Aedeagus lacking processes; apex of claspette usually not recurved. .. 16 
Claspette slender with contorted apex. .............-........ novamexicanus| 22 | 
Claspette stout with laterally directed apex. ......0.2.2---------- utahensis|21 | 
Median posterior process extended to slender point with subapical 

processes (Bic. NO5i)5 2 aoe eee oklahomensis| 19] 
Median process not as above, lacking subapical processes. -..........-....-. 18 


Claspettes with median portion of stems distinctly swollen, apex 
abruptly narrowed (Fig. 101); median posterior process elongate, 
parallel-sided truncate, about equal in length to basal 


GUMS umes ee neat ee ee oe ee pecosensis| 18] 
Claspettes tapering distally to simple pointed apex; median process 
noteciongateor parallel-sided. 12... :- 22. ee 19 
Apicolateral processes broad, very prominent, triangular (Figs. 79, 
91); claspette with distinct, anteriorly directed basal process. -............. 20 
Apicolateral processes slender, pointed; claspettes not as above. ........ 21 


Median posterior process slender with slight subapical swelling 

(Fig. 93); basal arch notched mesally; claspette slender, tapering 

to mesally recurved apex; membrane bare. ..............-.-------- pam potkilus| 16 | 
Median process broad, subtruncate without median swelling; 
claspette with rather stout, laterally turned sickle-shaped apex; 
apicolateral processes very large (Fig. 81); membrane 

SJOCCL DIET (oan eee ne ee eee I cea eee stone1| 14] 
Aedeagus very stout, conical, arms and basal arch short; claspettes 

with basal portion of stem swollen, knob-like, tapering to slender 
pointed, laterally directed tips; ninth sternum very narrow. .... sitiens[8 | 
Aedeagus with long slender arms, basal arch of medium length; _ 
claspettes tapering for entire length, apices entwined. ................ byersi|9 | 


960 Tue University ScIENCE BULLETIN 


Subgenus SELFIA Khalaf, 1954:38 
Type species Aieroglyphicus Malloch (orig. des.). 


This group of small to medium sized species contains females which have 
unsclerotized, therefore unapparent, spermathecae and plain wings, without 
a pattern of pale and dark spots. The antennal sensoria are located on various 
of the proximal 8 segments, depending on the species, but never on the 
distal 5. The hind tibial comb contains 4 or 5 spines. The claspettes of the 
male genitalia are fused into a platelike structure. The ventral root of the 
basimere is prominent; however, the dorsal root is often unapparent. Two 
lateral sclerites, often fused distally, usually comprise the adeagus. The un- 
marked wings of stoner James and bottimer: Wirth superficially resemble 
those of the Se/fia species; however, these species can be readily distinguished 
by the well sclerotized spermathecae and the distinctive male genitalia. 

Selfia is apparently restricted to the Nearctic region, as no neotropical 
forms have been described. This group reaches its highest development in the 
western portion of the United States, and only one species is reported east 
of the Mississippi. 

Three species, brookmani, hieroglyphicus and jamesi, were found in 
New Mexico. Unless otherwise stated, all distribution records of Selfia 
species in this paper refer to males which have characteristic genitalia. 


1. Culicoides (S.) brookmani Wirth 
(Figs. 6-9) 
Culicoides brookmani Wirth, 1952, Univ. Calif. Pub. Ent. 9:179 (male, female; California; ig. 
male genitalia, wing and palp). 
FEMALE—Length of wing 1.14 (1.04-1.27; n=4) mm. 

Head: Eyes separated less than diameter of a facet. Antenna with 
flagellomeres in proportion of 13:10:10:11:12:12:12:13:14:14:1515-21 ame 
tennal ratio 0.92 (0.82-0.97; n=4); distal sensory tufts present on segments 
3-10; segments 9-++10 to 11 in ratio of 0.62 (0.61-0.65; n=4). Distal four palpal 
segments in proportion of 15:25:8:12; third palpal segment short, very 
swollen, 1.97 (1.8-2.09; n=4) times as long as greatest breadth, with a 
large, very deep sensory pit; interior of pit lined with stalked sensillae (Fig. 
7). Ratio of length of head to proboscis 1.39 (1.38-1.42; n=4). Mandible 
with 11 teeth. 

Thorax: Legs unicolorous light brown; hind tibial comb with 4 or 5 
spines (Fig. 8). 

Wing: (Fig. 6). No pattern of pale and dark spots. Costa extending 0.53 
(0.51-0.56; n=4) of total length of wing. Macrotrichia sparse, occurring in 
small numbers on veins M1, Me, Ma and vannal cell. 

Abdomen: Spermathecae unsclerotized. 


Tue Cuticomwes or New Mexico 961 


Fics. 6-9. Culicoides brookmani Wirth. Female: 6, wing; 7, palpus; 8, tibial comb. Male: 
9, genitalia. 


MALE—(Fig. 9). Genitalia with ninth sternum broad, posterior margin in- 
definite; membrane spiculate. Ninth tergum tapering posteriorly, mesal 
notch slight, apicolateral processes rather small and tapering distally. Basi- 
mere with ventral root prominent, having a slightly curved apex; dorsal root 
evidently included in basal portion of basimere. Telomere broad basally, 
tapering to a slender tip. Aedeagus Y-shaped with widely separated basal 
arms; median posterior process conical, distally truncate; two distinct 
boomerang-shaped structures present, the posterior end of each forming an 
oblique angle with the anterior end. Occasionally these structures form the 
lateral margins of the aedeagus. Claspettes ventrally turned, invisible in 
most specimens or, at best, visible as narrow band. Jones and Wirth (1958) 
dissected the males of this species and found the claspettes fused into a plate 
with narrow, slightly curved, basal arms that join narrowly for approximately 
the middle one-third of their combined length. 


962 THe UNIversity SCIENCE BULLETIN 


DISTRIBUTION: Present records indicate that this species is probably 
restricted to the southwestern United States. It has previously been reported 
from Calif., Ariz., N. Mex. and Texas. New Mexico County Records: 
Catron, Grant. 

DISCUSSION: At present, alcoholic and slide-mounted females in the 
subgenus Se/fia are virtually inseparable. The descriptions of the females of 
the Selfia species in this paper are based on specimens from a locality where 
collection yielded large numbers of males of a single Se/fia species. An ex- 
tensive study is needed to determine characters useful in the separation of 
the species of this group. 

Slide-mounted specimens of brookmani tentatively can be separated from 
those of hieroglyphicus by the fact that Aieroglyphicus usually has sensory 
tufts on antennal segments 3, 5-10. In pointed specimens, the pattern of the 
mesonotal disc of hieroglyphicus is sufficient to distinguish it from brookmani, 
which has either an unadorned mesonotum or three longitudinal vittae, 
as reported by Jones and Wirth (1958) for Texas specimens. 

Although no good taxonomic character exists to distinguish slide-mounted 
and alcoholic specimens of brookmani from jamesi, separation can some- 
times be made on the basis of the large size of jamest, particularly its wing 
length. Pointed specimens can be distinguished on the basis of two sub- 
medial, longitudinal mesonotal stripes found in jamest. 


2. Culicoides (S.) hieroglyphicus Malloch 
(Figs. 10-13) 
Culicoides hieroglyphicus Malloch, 1915, Ill. State Lab. Nat. Hist. Bull. 10:297 (female; Arizona; 
fig. mesonotal disc). 
FEMALE—Length of wing 1.20 (1.12-1.23; n=8) mm. 

Head: Eye separation greater than diameter of a facet. Antenna with 
flagellomeres in proportion of 11:8:8:8:9:9:9:10:13:13:15 :16:24; antennal ratio 
1.03 (0.96-1.12; n=9); segments 9+10 to 11 in ratio of 0.67 (0.61-0.72; 
n=9); distal sensory tufts on segments 3, 5-10. Distal 4 palpal segments 
in proportion of 18:27:10:10; third palpal segment swollen, 2.11 (2.0-2.25; 
n=9) times as long as greatest breadth, with deep sensory pit (Fig. 11). 
Ratio of head length to proboscis 1.24 (1.10-1.46; n=8). Mandible with 11-13 
(n=9) teeth. 

Thorax: Legs yellowish brown; indistinct pale rings pre-apical on fore 
femora and subbasal on tibiae. Hind tibial comb with 4-5 spines. 

Wing: (Fig. 10). Wing without pattern of pale and dark spots. Costa 
extending 0.54 (0.53-0.55; n=8) of entire wing length. Macrotrichia abun- 
dant. 

Abdomen: Spermathecae unsclerotized. 


Tue Curicowses or New Mexico 963 


MALE—Genitalia (Figs. 12, 13) with ninth sternum divided, with elongate 
mesal lobes extending posteriorly to almost same length as aedeagus. Ninth 
tergum tapering; caudal portion laterally convex, mesal notch present in 
some specimens, absent in others; apicolateral processes small, very slender. 
Basimere short, tapering slightly; ventral root well developed, slender with 
apically expanded end; dorsal root apparently consolidated into base of 
basimere; many large spines on mesal surface of basimere. Telomere 
swollen basally, narrowing distally with more or less blunt foot-shaped apex. 
Aedeagus composed of 2 slender sclerites, the basal or arm portions widely 
separated proximally, converging mesally, then turned caudad to form 


Fics. 10-13. Culicoides hieroglyphicus Malloch. Female: 10, wing; 11, palpus. Male: 12, 
claspettes; 13, genitalia, claspettes removed. 


964 Tue University ScrENcE BULLETIN 


more or less parallel-sided median process. Claspettes fused to form modi- 
fed triangular structure with long, slender caudomesal extension. 

DISTRIBUTION: This species is a common western form occuring 
from South Dakota southward to Mexico and westward to California. New 
Mexico County Records: Bernalillo, Catron, Chaves, Colfax, DeBaca, Dona 
Ana, Eddy, Lincoln, Otero, Quay, Roosevelt, Sandoval, San Miguel, Santa 
Fe; andSierra: 

VARIATION: A large number of specimens collected near Glenwood, 
N. Mex., exhibited both male and female characteristics. These intersexes 
had normal male genitalia but possessed certain female head structures. The 
eyes were broadly contigucus, as one would expect in a male, but the third 
palpal segment was distinctly swollen with a deep sensory pit and many had 
denticulate mandibles. The most striking peculiarity were the antennae. 
These intersexes have the first eight antennal segments only slightly longer 
than wide and the transition between segment types occurs between segments 
10 and 11. In a few specimens segment 11 was shortened. The pedicel, which 
is enlarged in normal males, apparently due to the presence of Johnstons 
organ, is reduced to the femalelike structure. A slightly reduced type of 
verticillate hairs occurs on these forms. 

Downes (1958) has discussed a correlation between modification in head 
structures of males and mating behavior. Culicoides utahensis males, for in- 
stance, have femalelike antennae and palps, and this condition is apparently 
related to mating. Females of this species are often found in the ears of 
rabbits. The males, therefore, appear to be attracted to the rabbit, where 
they locate the females and mating occurs. A situation of this type is 
feasible for Aieroglyphicus since females of this species are known to feed 
in the ears of rabbits in California. I do not know whether the high in- 
cidence of intersexes in this population represents a behavioral modification 
associated with mating or an extremely high infestation of mermithial 
nematodes, which also is known to induce the formation of intersexes. 

FEEDING HABITS: Hoy (1966) has reported that in California this 
species attacks deer, feeding by day on the inner surfaces of the ears, and 
Jones (1965) found that it feeds on sheep in Colorado. These records, coupled 
with that of rabbits as hosts, indicate a rather broad host range for the 
species. 

DISCUSSION: The difficulty of separating the Se/fia species has been 
explained previously in the discussion of brookmant. The males of heiro- 
glyphicus are readily separated from other species by the distinctive genitalia; 
however, with the possible exception of the mesonotal pattern, no clear-cut 
character is now known to distinguish the females from those of other 
Selfia species. 


Tue Cuticomwes o— New Mexico 965 


3. Culicoides (S.) jamesi Fox 
(Figs. 14-17) 
Culico des jamest Fox, 1946, Ent. Soc. Wash. Proc. 48:244 (male, female; Montana; fig. wing, 
mesonotal disc, palpus and male genitalia). 
FEMALE—Length of wing 1.27 (1.20-1.39; n=15) mm. 

Head: Eye separation equal or subequal to diameter of a facet. An- 
tenna with flagellomeres in proportion of 14:11:10:10:11:12:12:13:15:17:17: 
18:25; antennal ratio 1.0 (0.94-1.13; n=16); distal sensory tufts present on 
segments 3-10; segments 9-+-10 to 11 in ratio of 0.63 (0.59-0.70; n=16). Distal 
4 palpal segments in proportion of 16:27:8:13; third segment long and 
swollen, 2.26 (2.08-2.5; n=15) times as long as greatest breadth, with a deep 
sensory pit with many stalked sensillae lining proximal end (Fig. 15). Ratio 


17 


Fics. 14-17. Culicoides jamesi Fox. Female: 14, wing; 15, palpus; 16, tibial comb. Male: 
17, genitalia. 


966 Tue UNiversiry SciENCE BULLETIN 


of length of head to proboscis 1.27 (1.20-1.39; n=15). Mandible with 11-13 
(n=16) teeth. 

Thorax: Legs brown. Hind tibial comb with 4 or 5 spines, in some 
specimens 4 on 1 tibia, 5 on the other (Fig. 16). 

Wing: (Fig. 14). Wing without pattern of light and dark spots. Macro- 
trichia present in distal and posterior portions of wing. Costa extending to 
0.56 (0.54-0.59; n=9) of distance to wing apex. 

Abdomen: Spermathecae unsclerotized. 

MALE—Genitalia (Fig. 17) with ninth sternum expanded mesally, very deep 
medial posterior cleft extending almost to base. Ninth tergum tapered with 
narrow, divergent apicolateral processes, mesal cleft poorly defined. Basimere 
long; ventral root small; dorsal root well developed and elongate; rugose 
membrane connecting basimere in area of dorsal root with lateral margin 
of the fused claspette. Telomere with large, distinctive, footlike apical ex- 
pansion; apex turned at right angle to remainder of telomere. Aedeagus 
very long and slender with high rounded basal arch; aedeagal arms long, 
extending anteriorly almost parallel, with bases turned antero-laterally. Clas- 
pettes fused to form a posteriorly pointed plate-like structure; anterior arms 
curved mesally and slightly laterad to give an almost semi-circular antero- 
medial emargination. 

DISTRIBUTION: Culicoides jamesi is a western species occurring from 
New Mexico to Montana, westward to California. New Mexico County 
Records: Catron, Lincoln, Grant and Sandoval. 

DISCUSSION: For distinguishing features see the discussion of brook- 


mani. 


Subgenus CULICOIDES Latreille 1809:251 
Type species Culex pulicaris Linneaus as Ceratopogon punctatus Meigen (monobasic). 
Large species with the apical portion of the second radial cell ending in 
a pale spot. Cell Ma is dark at the base of the mediocubital fork, and the 
apices of veins Mi, Mo, M344 and Cu are dark. 
The North American species of this subgenus belong to the cockerelli 
group. 


4. Culicoides (C.) cockerellii (Coquillett) 
(Figs. 18-22) 

Ceratopogon cockerelli Coquillett, 1901, U.S. Natl. Mus. Proc. 23:603. 
Culicoides cockerellii Coquillett; Kieffer, 1906, Gen. Insectorum, fasc. 42:54. 
FEMALE—Length of wing 1.84 (1.61-2.04; n= 5) mm. 

Head: Eyes contiguous, meeting only at a point. Antenna with flagel- 
lomeres in proportion of 16:13:15:15:15:15:15:15:21 :23:27:30:36; antennal 
ratio 1.07 (1.03-1.10; n=5); segments 9+10 to 11 in ratio of 0.71 (0.68-0.75; 


Tue Cutricomwss of New Mexico 967 


21 a 


19 22 


Fic. 18-22. Culicoides cockerelli (Coquillett). Female: 18, wing; 19, palpus; 20, tibial 
comb. Male: 21, claspettes; 22, genitalia, claspettes removed. 


n=5); distal sensory tufts on segments 3, 5, 7, 9-15, occasionally absent from 
10 or from 7, 9 and 10 (see discussion). Distal 4 palpal segments in propor- 
tion of 26:36:13:18; third palpal segment very slender and lacks a pit (Fig. 
19), 3.26 (2.7-3.8; n=5) times as long as greatest breadth, possessing numer- 
ous stalked sensillae. Ratio of head length to proboscis 1.05 (0.84-1.20; n=4). 
Mandible with 14-16 (n=5) teeth. 

Thorax: Legs dark brown. Hind tibial comb with 6 spines (Fig. 20). 

Wing: (Fig. 18). Distal half of second radial cell in a light area; wing 
with 3 transverse bands: first band lying between r-m crossvein and base of 
wing, extending posteriorly from costa to beyond vein Mi +2, interrupted by 
pale area which extends to slightly anterior of Cui, another dark spot over 
Cui, Cuz, and Ist V and extending slightly into vannal cell, majority of 
vannal cell in pale area; second dark band arising midway between r-m 
crossvein and distal end of second radial cell, extending to just behind Mi-M2 
fork, there interrupted by a light area, dark area in proximity of medio-cubital 
fork where it follows vein Cui to margin; third dark band distal to post- 


968 Tue University ScrENCE BULLETIN 


stigmatic spot, band very intense in anterior portion of wing. Macrotrichia 
long, rather dense in distal and posterior parts of wing. Costa extending 0.60 
(0.54-0.62; n=5) of entire wing length. 

Abdomen: Two oval spermathecae (crumpled in New Mexico specimens 
seen); ducts sclerotized at base; rudimentary spermatheca and ring present. 
MALE—Genitalia (Figs. 21, 22) with ninth sternum broad, with shallow 
caudomedian excavation. Ninth tergum tapering posteriorly, with distinct, 
rounded, caudal lobe; apicolateral processes very small. Basimere narrowing 
apically; roots small, subequal, tapering apically. Telomere large, gradually 
tapering to slightly expanded apex. Median posterior process of aedeagus 
broadly rounded apically, aedeagal arms slender, with slight swellings on 
lateral margins, basal portions turned laterally to give foot-shaped structure; 
lateral membraneous extensions extending from arms to lateral portion of 
median posterior process; basal arch rounded. Claspette with antero-laterally 
directed bases; stem swollen proximally, tapering to very slender anteriorly 
directed apex bearing fine fringe of subapical spines. 

DISTRIBUTION: This species has previously been recorded from 
Alaska southward to California and Colorado. This is the first record of 
it from New Mexico. New Mexico County Records: Poorly marked form— 
Colfax, San Juan and Lincoln; well-marked form—Taos. 

VARIATION: Wirth (1952), in discussing the cockerellit group in 
California, lists three varieties or forms in addition to cockerellit sensu stricto. 
Of the seven females of cockerellii taken during this investigation, four can 
readily be assigned to cockerellii s. str. as defined by Wirth, while the re- 
mainder had a reduced wing pattern and different sensorial configuration 
and closely agreed with the description of Wirth’s variety “C.” The cul- 
mination in reduction of wing pattern is exhibited by a specimen from San 
Juan River, San Juan Co., in which the wing pattern is reduced to the usual 
spot over the r-m crossvein, poststigmatic spot and two other very small round 
spots, one in the middle of cell Ms and one lying just below the Ist vannal 
vein in the vannal cell. The poorly marked form has a sensorial pattern of 
3, (5), 7, (9), (10), 11-15 as opposed to 3, 11-15 for the well-marked form. 
Since no males of cockerellit were taken during this study, information as to 
variation in the male genitalia is not available. 

DISCUSSION: Vargas (1960) proposed the new subgenus Anilomyza 
for the covagarciai group, in which he includes cockerellit and luteovenus 
Root and Hoffman, as well as the neotropical species rostratus Wirth and 
Blanton. The covagarciat group as originally outlined by Wirth and Blanton 
(1959) did not include /uteovenus, which was placed in the pwlicaris group. 
There appears to be little gained by dividing this rather homogeneous group 
of species, previously known as the subgenus Culicoides, into two subgenera. 


| 


Tue Cuticowes or New Mexico 969 


Subgenus AVARITIA Fox, 1955:218 
Type species Ceratopogon obsoletus Meigen (orig. des.). 


This subgenus is composed of a group of small species with poorly 
marked wings. The distal portion of the second radial cell is in a pale spot, 
and macrotrichia are usually sparse. Antennal sensoria are located on seg- 
ments 3 and 11-15, and the eyes of the females are contiguous. There are two 
subequal spermathecae with very short necks, and the hind tibial comb 
possesses 5 spines. The ninth tergum of the male genitalia usually lacks the 
mesal notch, the apicolateral processes are absent (except in chiopterus, in 
which they are poorly developed, blunt structures), the claspettes are sepa- 
rate and taper distally to a fine point which may or may not possess apical 
hairs, and the basimere has long, simple, subequal roots. 

The members of this subgenus are widespread, two species occurring 
in the Palearctic and Nearctic regions, while a third species is Nearctic and 
Neotropical. One species, obsoletus (Meigen) occurs in New Mexico. 


5. Culicoides (A.) obsoletus (Meigen) 
(Figs. 23-27) 
Ceratopogon obsoletus Meigen, 1818, Syst. Beschr. Gur. Zweifl. Ins., 1:76. 
Culicoides obsoletus (Meigen); Goetghebeur, 1921, Mem. Mus. Roy. Hist. Nat. Belg., 8:180. 
For additional synonomy see Fox (1955). 
FEMALE—Length of wing 1.19 (1.01-1.36; n=9) mm. 

Head: Eyes broadly contiguous. Antenna with flagellomeres in propor- 
won of |1-10:10:10:11:11:12:13:17:17:17:18:30; antennal ratio 114 (10-123; 
n=9); segments 9+10 to 11 in ratio of 0.72 (0.65-0.79; n=9); distal sensory 
tufts on segments 3, 11-15. Distal four palpal segments in proportion of 
17:18:8:10; third segment very slender, only slightly swollen (Fig. 24), 2.74 
(1.8-3.7; n=7) times as long as greatest breadth (the specimen measuring 1.8 
had an extremely short and slender third segment, hence the very small 
value); shallow sensory pore. Ratio of head length to proboscis 1.0 (0.87-1.24; 
n=6). Mandible with 12-16 (n=6) teeth. 

Thorax: Legs brown; lacking distinct banding pattern. Hind tibial comb 
with 5 spines, one nearest spur usually longest. 

Wing: (Fig. 23). Pattern indistinct and variable; apical half of second 
radial cell in a pale area; pale spot over r-m crossvein small, extending 
posteriorly beyond vein Mi+2, not attaining anterior wing margin; post- 
stigmatic spot rather elongate, extending to just anterior of vein M1; rather 
vague pale spots in distal portion of cells Rs, Mi and Mg, their position and 
intensity variable; pale spot in cell Mi extending to posterior wing margin; 
small spot in anterior portion of vannal cell; poorly defined pale area in 
medial portion of wing. Costa extending 0.60 (0.57-0.63; n=9) of entire 
wing length. Macrotrichia sparse. 


970 Tue UNiversiry ScriENCE BULLETIN 


SEE EES Ew 


Ge 5UBAM TS OTM SOR 


Yar Wea rv sone wy 


ET 


( 


24 26 


Fics. 23-27. Culicoides obsoletus (Meigen). Female: 23, wing; 24, palpus; 25, spermathecae. 
Male: 26, claspettes; 27, genitalia, claspettes removed. 


Abdomen: Two slightly unequal spermathecae (Fig. 25) 56 x 33 and 
46 x 40 », ducts sclerotized for a short distance; rudimentary spermatheca and 
ring present. 

MALE—Genitalia (Figs. 26, 27) with ninth sternum broad; caudomedian 
excavation narrow, deep, anteriorly pointed; posterior lips of excavation dis- 
tinctly lobe-like. Ninth tergum lacking apicolateral processes and mesal 
notch. Basimere with very long slender, pointed ventral roots; dorsal roots 
much shorter. Telomere only slightly swollen basally, mesal portion narrow, 
apex swollen, club-like. Aedeagus with short, sub-conical median posterior 
process; aedeagal arms with small, laterally directed bases, slightly sinuate 
posteriorly, abruptly bent mesally to form median process; basal arch very 
Jong, extending approximately 0.75 of entire length of aedeagus. Claspettes 


Tue Cuticomwes oF New Mexico 971 


with very stout basilateral arms; stems swollen proximally, tapering to very 
slender apex bearing microscopic fringe of spines. 

DISTRIBUTION: A very widespread species reported by Jamnback and 
Wirth (1960) as occurring in Eurasia, North Africa and North America. In 
North America odsoletus has been previously reported throughout the eastern 
United States and westward to include Colorado, Oklahoma and California. 
Canadian records include Alberta and British Columbia. New Mexico County 
Records: Lincoln, Otero. 

VARIATION: The specimens from New Mexico designated in this 
paper as obsoletus do not completely agree with the species descriptions from 
the eastern United States. The most obvious discrepancy occurs in the male 
genitalia. There are apical hairs on the claspettes, as in typical obsoletus, and 
the mesal notch on the ninth sternum agrees closely with that species. The 
aedeagus, however, has a shape closely approximating that of sanguisuga 
(Coquillett). Because of the paucity of male specimens of this form from 
New Mexico, I have refrained from describing it as new. The obsoletus 
complex in the western United States is very poorly understood and is in 
need of extensive study, such as given the eastern spcies by Jamnback and 
Wirth (1960). 

FEEDING HABITS: In many areas obsoletus constitutes a serious pest. 
Wirth (1952) has cited its annoyance in the mountains of the western states. 
Although various authors have described this species as pestiferous in the 
eastern states, Jamnback and Wirth (1960) state that many of these biting 
records actually refer to sanguisuga. 

During this investigation, this species was collected biting man in the 
mountains of southern New Mexico. It appears to be crepuscular in its 
activity, as all specimens were taken at dusk. 


Subgenus BELTRANMYIA Vargas, 1953:34 
Types species, crepuscularis Malloch (orig. des.). 


This subgenus contains species in which the females possess one sperma- 
theca and the antennal sensorial configuration varies from 3-15 to a very 
reduced condition in some species in which sensoria are present only on 
the third and various of the distal 5 segments. The wing has the second 
radial cell in a dark area, and the pattern of pale and dark spots may be 
either faint or well defined. The ventral root of the basimere in the male 
is either very reduced or absent. The claspettes are separate, tapering to very 
fine, pointed, unadorned apices. 

Numerous representatives of this subgenus are found in the eastern 
United States; however, the distributions of three continue into the western 
states: bermudensis Williams as far west as Texas, wisconsinensis Jones to 
California and Washington, and crepuscularis Malloch to California, 


972 THe UNIversiry SciENCE BULLETIN 


6. Culicoides (B.) crepuscularis Malloch 
(Figs. 28-33) 
Culicoides crepuscularis Malloch, 1915, Ill. State Lab. Nat. Hist. Bull. 10:303 (male, female; 

Illinois; fig. wing, mesonotal disc). ‘ 
FEMALE—Length of wing 1.41 (1.14-1.56; n=22) mm. 

Head: Eye separation about equal to diameter of a facet. Antenna with 
flagellomeres in proportion of 15:10:10:10:10:10:11 :13 :20:22:24:24:33; anten- 
nal ratio 1.35 (1.26-1.50; n=22); distal sensory tufts present on segments 
3-14 (15), occasionally absent from segments 10 and 11; segments 9+-10 to 
11 in ratio of 0.90 (0.79-1.05; n=12). Distal 4 palpal segments in proportion 
of 20:33:11:13; third palpal segment swollen, 2.27 (2.0-2.56; n=25) times as 
long as greatest breadth, sensory pit deep with numerous stalked sensillae 


located on inner surface of pit (Fig. 29). Ratio of head length to proboscis 
1.05 (0.95-1.14; n=23). Mandible with 14 (10-16; n=26) teeth. 


29 32 33 


Fics. 28-33. Culicoides crepuscularis Malloch. Female: 28, wing; 29, palpus; 30, tibial 
comb; 31, spermatheca. Male: 32, claspettes; 33, genitalia, claspettes removed. 


Tue Cuticowes or New Mexico 973 


Thorax: Legs brown; fore femora with pre-apical pale ring, all tibiae 
with subbasal pale rings. Hind tibial comb with 4 spines (Fig. 30). 

Wing: (Fig. 28). Second radial cell in a dark area; pale spot over r-m 
crossvein extending slightly posteriorly to vein Mi+»2 but not extending to 
anterior margin; remaining pale spots located as follows: immediately distal 
to second radial cell, in distal portion of cell Rs but not attaining wing mar- 
gin, in cells Mi, Mz, Ms; a double spot in distal portion of vannal cell, us- 
ually attaining the wing margin; two elongate spots in proximal portion of 
cells Mi and Me; medio-cubital fork in a dark area. Macrotrichia abundant 
over surface of the wing. Costa extending to 0.56 (0.51-0.64; n=22) of entire 
length of wing. 

Abdomen: One large spermatheca (Fig. 31) with neck sclerotized for a 

very short distance. Rudimentary spermatheca, without sclerotized ring, 
occasionally evident in slide-mounted preparations. 
MALE—Genitalia (Figs. 32, 33) with narrow ninth sternum, caudiomedial 
excavation, broad, deep; membrane spiculate. Ninth tergum tapering pos- 
teriorly, with prominent apicolateral processes. Basimere long and slender, 
dorsal root well developed, ventral root absent. Telomere expanded basally, 
tapering, with a slightly swollen, mesally tipped apex. Aedeagus with coni- 
cal, distally truncate, median posterior process; aedeagal arms long and 
slender with small laterally directed bases, basal arch rounded. Claspettes 
with pronounced foot-shaped basal expansions; mesal margins of stem 
straight, lateral margins swollen at approximate mid-point, stems sharply 
bent at apex to form simple, ventrolaterally directed tips. 

DISTRIBUTION: This species has been taken in most states from New 
England to California, Mexico into Canada. New Mexico County Records: 
Catron, Colfax, Dona Ana, Lea, Lincoln, Otero, Quay, Rio Arriba, Sandoval, 
San Juan, San Miguel and Socorro. 

VARIATION: Considerable variation exists in the wing pattern of 
this species at several localities. In Rio Arriba County, poorly marked forms 
were taken that exhibited reduction in number and size of the distal wing 
spots. Some of these specimens lacked the pale spots in cell M1, and the spots 
in cells Rs, Mz and Mg were greatly reduced. In other specimens pale spots 
were absent in these four cells. Dr. Willis Wirth (in litt.) informs me 
that this variation in wing markings as well as some additional characters 
appears to show regional trends. Together with other morphological and 
biological descrepancies reported by various authors, it suggests that there 
may be more than one species disguised under the name crepuscularts. 

Two intersexes were taken at light, along with numerous normal speci- 
mens, at the Bosque del Apache Refuge along the Rio Grande. 

DISCUSSION: Although the wing markings of crepuscularis super- 
ficially resemble the wing pattern of females of certain other New Mexico 


974 Tue University SciENCE BULLETIN 


species, the presence of one large spermatheca will readily distinguish this 
species. Only two other New Mexico species, reevesi and variipennis, have 
a single spermatheca. Separation of the females of these three species is ac- 
complished easily by the irregular wing markings, sensorial pattern and 
usual “C” shaped spermatheca of variipennis and the unusual shape of the 
antennal segments, sensorial pattern, etc., of reevest. 

This species is not closely allied to any of the other New Mexico species, 
but is very similar to several, particularly hollensis (Melander and Brues) 
and wisconsinensis Jones, which occur primarily in the eastern portion of 
the U.S. 

FEEDING HABITS: With the exception of a reference by Edmunds 
and Keener (1954), in which they state that C. crepuscularis is a very severe 
pest in Nebraska, the proponderence of evidence accumulated for this species 
indicates an ornithophilic behavior. This is supported by reports of Wil- 
liams (1955) and Snow (1955) that crepuscularis is active in considerable 
numbers in the forest canopy; and Messersmith (1965) has taken large 
numbers of engorged females of the species from chicken houses. 

The importance of crepuscularis in the interrelationship between various 
pathogens and their avian hosts has recently been stressed. Robinson (1961) 
cited the development of an avian filarial worm in this species after having 
fed on an infested starling, while Bennett (1961) used crepuscularis, along 
with other species, in trypanosome transmission experiments involving several 
species of birds. Fallis and Bennett (1961) reported developing odcysts and 
sporozoites of Haemoproteus in the gut wall of crepuscularis after the flies 
had fed on infected crows and purple finches. 


Subgenus MONOCULICOIDES Khalaf, 1954:39 


Type species Ceratopogon nubeculosus Meigen (orig. des.). 


Large species with second radial cell in dark area and with moderately 
abundant macrotrichia. The eyes are widely separated, and one C-shaped 
spermatheca is present. The ninth tergum of the male has well developed 
apicolateral processes, the claspettes are fused basally, with the apices ter- 
minating in slender separate points, and the basimere possesses a moderately 
long ventral root, but a very short dorsal root. 

Two species, gigas Root and Hoffman and varipennis (Coquillett), are 
included in the Nearctic fauna, although only variupennis is found within 
the boundaries of the United States. 


7. Culicoides (M.) variipennis sonorensis Wirth and Jones 
(Figs. 34-39) 


Ceratopogon varupennis Coquillett, 1901, U.S. Nat. Mus. Proc. 23:602. 
Culicoides variipennis; Kieffer, 1906, Gen. Insectorum, fasc. 42:55. 


Tue Cuuicowrs or New Mexico 975 


Culicoides varitpennis sonorensis Wirth and Jones, 1957, U. S$. D. A. Tech. Bull. no. 1170, p. 
18-20 (male, female; Arizona; fig. wing, spermatheca, mesonotal disc, palpus, male 
genitalia). 

Culicoides variipennis australis Wirth and Jones, 1957, U. S. D. A. Tech. Bull. no. 1170, p. 
15-17 (male, female; Louisiana; fig. wing, spermatheca, mesonotal disc, palpus, antenna). 
NEW SYNONOMY. 


FEMALE—Length of wing 1.55 (1.30-1.79; n=14) mm. 

Head: Eyes widely separated by much more than diameter of a facet. 
Antenna with flagellomeres in proportion of 13:13:13:13:13:13:13:14:16:17: 
18:18:26; antennal ratio 0.86 (0.79-0.90; n=13); segments 9+-10 to 11 in ratio 
of 0.59 (0.54-0.70; n=7); distal sensory tufts always on segments 3, 8-10, 


) 
Fics. 34-39. Culicoides variipennis sonorensis Wirth and Jones. Female: 34, wing; 35, 
palpus; 36, tibial comb; 37, spermatheca. Male: 38, claspettes; 39, genitalia, claspettes removed, 


976 Tue University ScrENcCE BULLETIN 


often on 5, 6 or 7, occasionally present on segment 4. Distal 4 palpal seg- 
ments in proportion of 25:35:10:15; third palpal segment very swollen (Fig. 
35), 2.38 (2.11-2.69; n=14) times as long as greatest breadth, with moder- 
ately deep sensory pore. Ratio of length of head to proboscis 1.10 (1.02-1.17; 
n=10). Mandible with 11-15 (n=14) teeth. 

Thorax: Legs brown; fore femora with median pale band; all femora 
with pre-apical pale ring, tibiae with subbasal pale rings. Hind tibial 
comb with 6-7 spines (Fig. 36). 

Wing: (Fig. 34). Second radial cell in a very dark spot; pattern of 
pale and dark spots irregular. Costa extending 0.55 (0.53-0.58; n=14) of 
entire wing length. Macrotrichia short, rather sparse. 

Abdomen: One large, mesally bent, C-shaped spermatheca (Fig. 37). 
MALE—Genitalia (Figs. 38, 39) with ninth sternum having little if any 
posterior excavation. Ninth tergum with prominent, triangular apicolateral 
processes. Basimere short, slightly tapered; ventral root small; dorsal root 
long, slightly curved laterally. Telomere curved, strongly swollen basally, 
abruptly narrowing approximately at mid-length, distal half gradually 
tapering to simple apex. Aedeagus elongate; median posterior process con- 
sisting of two narrow separate points with slight sub-apical expansion, lateral 
portion of median process membraneous; aedeagal arms short, gently 
curved laterally; basal arch short, rounded, extending only about 0.25 of 
entire length of aedeagus; membrane posterior to basal arch may or may 
not be spiculate. Claspettes fused basally, with slender, pointed, separate 
apices. 

DISTRIBUTION: This subspecies of a widely distributed form found 
in the United States, Canada, Mexico and Baja California is limited to the 
more arid portions of the southwestern United States and Mexico. Wirth 
and Jones (1957) reported sonorensis from Arizona, California, Nevada, New 
Mexico, Oklahoma, Utah, Washington and Mexico, and under the name 
australis it has been recorded in Kansas, Louisiana, Missouri, Oklahoma, 
South Carolina, Texas and Virginia. It is probably the most common 
species occurring in New Mexico. New Mexico County Records: Chaves, 
Curry, DeBaca, Dona Ana, Eddy, Lea, Lincoln, Quay and Roosevelt. 

DISCUSSION: Wirth and Jones (1957) proposed five subspecies for 
polytypic Culicoides variupennis. Consideration of the distributional data, 
ecological associations, as well as the critical examination of a number of 
specimens from various localities in New Mexico make evident the need 
for a discussion of the validity of certain aspects of this division. 

Many times when attempting to analyze a taxonomic problem, it is 
desirable to define the taxonomic category under investigation. The vari- 
pennis situation is complicated by the omission of, or reference to, such an 
explanation of what these writers consider to be a subspecies. If we take 
one of the more widely used definitions of subspecies, i.e., “an aggregate of 


Tue Cuticomwres oF New Mexico 977 


local populations of a species inhabiting a geographic subdivision of the 
range of the species and differing taxonomically from other populations of 
the species (Mayr 1963),” and apply it to the variipennis problem, several 
discrepancies become evident. 

There can be little doubt that variipennis sensu lato is divisible into two 
forms. On the basis of morphological and particularly biological evidence, 
it would appear that variipennis variipennis and vartipennis sonorensts are 
worthy of at least subspecific status. The former subspecies is restricted to 
the forested northern and eastern portions of North America and is not 
associated with saline environments. It is considered by Ross (1962) to be 
the oldest and most stable form. The sonorensis stock of the southwestern, 
arid portion of North America has been further divided into four sub- 
species on the basis of biological as well as morphological criteria. The 
australis form is typically found in saline environments and is characterized 
by Wirth and Jones as having long wings, more slender palps, higher num- 
ber of antennal sensoria and a bare aedeagus (see Table 2 for numerical 
values). The sonorensis form typically inhabits polluted waters and is 
distinguished by shorter wings, stouter palps, lower number of supernumerary 
antennal sensoria and a spinose aedeagus (Table 2). As will be shown be- 
low, this habitat isolation apparently breaks down in southeastern New 
Mexico, and populations occur that exhibit all morphological intergradation 
between sonorensis and australis. No definite statement can be made re- 
garding occidentalis, the west coast subspecies, or albertensis, which occurs 
in the Midwest, northward into Canada. These two forms lie outside the 
scope of this investigation; however, it should be noted that Wirth and 
Jones state that over much of its range the latter form cannot be distinguished 
from other forms of the sonorensis stock. Ross (1962) considers occidentalis 
to be the oldest and most stable form of the sonorensis stock; however, this 
is Open to serious question. 

A sample from the Pecos River, east of Roswell, provided an excellent 
opportunity to examine the extent of infraspecific variation in a given 
vartipennis population. The range of variation among 28 specimens ex- 
amined from this locality included individuals that could definitely be as- 
signed to the typical sonorensis form to those that exhibited all the char- 
acteristics of australis. Of 14 males, 4 had very spinose aedeagi of the 
sonorensis form, 8 possessed bare aedeagi of the azstralis form and 2 
specimens could be considered as intermediate in that they had only 5 
or 6 spines. Among the females (Figs. 40, 41), 5 specimens having short 
wings, swollen palpi and 1 or 2 additional sensoria could be assigned to 
the sonorensis form. Eight of the remaining females could be designated 
as australis, although some resembled sonorensis in 1 or 2 characters. One 
female appeared to be intermediate (shown in scatter diagram by an open 


978 Tue UNiversity SctENcE BULLETIN 


oO 
= a 
a 5 
a = 
=) 
i; 4 
= ef 
ar <q 
< a 
a fo) 
” 
z 
WW 
n” 
1213141516 17 18 12 13 14 15 16 17 18 
40 WING LENGTH 41 WING LENGTH 


Fics. 40-41. Scatter diagrams showing extent of intraspecific variation in a given population 
(Pecos River, east of Roswell, New Mexico) of Culicoides variipennis. Open square indicates 
apparent intermediate form between C. variipennis australis and C. variipennis sonorensis forms. 


rectangle, having a long wing (1.58 mm.), one supernumerary sensoria and 
a very swollen palp (P.R.=2.23). 

Genetically, a given local population can, if it occurs in a zone of inter- 
gradation, contain elements of more than one subspecies. If the zones of 
contact or overlap between subspecies have not been determined, however, or 
if the local population in question is remote from such a zone, the popula- 
uon should taxonomically be assigned to only one or another subspecies. 
That is, subspecies are generally thought of as being for the most part mutu- 
ally exclusive. 

The evidence described would imply that either avstralis is only an infra- 
subspecific form of sonorensis, or that the four characters Wirth and Jones 
used to differentiate supposed subspecific populations will not effectively do 
so. (A fifth character, the number of mandibular teeth of the females, is not 
included here because of the high degree of overlap that exists between the 
two forms). 

One remaining criterion for subspecific distinction of these two forms is 
habitat isolation. A large number of specimens were examined from a salt 
lake near Loving, New Mexico. The extreme salinity of this habitat was 
evidenced by a salt crust at various places along the margin. Potash was 
mined nearby. Over 50 specimens were examined from this locality, and 
only one form was present (Figs. 42, 43), but these were not australis, as one 
would expect, but sonorensis! All the males examined had spinose aedeagi, 
and the number of antennal sensoria and palpal ratio of the females were 
almost precisely those given by Wirth and Jones as mean values for 
sonorensis. The mean wing length of this series was somewhat greater 
than would be expected for a sonorensis population; however, the agreement 


Tue Curicomes oF New Mexico 979 


| 

2 

27 < 

2 = 7 i. 
a 2.5 =) 
° z 
Sx 
=) 6 e eee ° ° 
e 23 a 
4 2. < 
me Fin « 5 ee © © 8 00 ® 0 Ouro tee 
=e ° 
AL 2 

1.9 w 

” 4 e ee e Om0 @ OO e 
18 
17 
(:2een: Sinn 1S (Tce 1.8 200 13 I (5 Kea lke 18 

42 WING LENGTH 43 WING LENGTH 


Fics. 42-43. Scatter diagrams showing extent of intraspecific variation in a given popula- 
tion (Salt Lake, near Loving, New Mexico) of Culicoides varitpennis. Open circles indicate more 
than one individual exhibiting these character states. 


‘a the other characters would definitely suggest that this is a population of 
sonorensis. Here again the range of such characters as wing length and 
palpal ratio encompassed almost the entire range of variation Wirth and 
Jones established as 95° limits of both subspecies. 

If australis and sonorensis were actually subspecies, the long wing char- 
acteristic of the Loving population could be due to gene flow between them. 
Subspecific intergradation cannot account for the remaining ecological and 
morphological discrepancies, however, and the spinose aedeagi, low number 
of supernumerary sensoria and palpal ratio indicate that this population is 
definitely sonorensis. 

The wide variation of characters found in the Pecos River sample might 
be ascribed to a zone of contact or overlap between two phenotypically dis- 
tinct, allopatric populations. If these phenotypically different populations are 
considered as subspecies, i.c., australis and sonorensis, intergradation between 
them should not occur also at such distant localities as northern Oklahoma 
and southern Texas (cited by Wirth and Jones) and southern New Mexico, 
especially when both subspecies are shown to occur at various intermediate 
localities. The area enclosed by these sites of intergradation is almost half 
of the proposed range of australis. 

Rowley (1965), in his study of the bloodsucking midges in the Columbia 
Basin of Washington, has also reported two subspecies of vartipennis, 1.€., 
occidentalis and sonorensis, occurring in the same area. In the Columbia 
Basin Crab Creek flows near numerous alkaline potholes producing, with 
regard to salinity, two distinct larval habitats in very close proximity. Each 
of these habitats is theoretically better suited for a different subspecies. If these 
two forms were subspecies of one species, they would be expected to become 
merged where they occur sympatrically. Rowley reports that individuals from 


980 Tue UNiversiry ScrENcE BULLETIN 


this area, like those from the above mentioned localities in southeastern New 
Mexico, exhibit a great amount of variation in the characters used for sub- 
specific designation by Wirth and Jones. 

One possible explanation for the varitpennis problem is that these forms 
of the sonorensis stock are ecotypes produced by the response of the im- 
mature stages to their particular environment. The relatively large amount 
of morphological variation that exists among variipennis populations, such 
as at the Washington and New Mexico localities, could therefore be at- 
tributed to selection by the habitat. The designation of ecotypes rather than 
subspecies would also clarify the many populations of variipennis in the 
western United States that are unassignable to “subspecies.” 

The preponderance of evidence accumulated in this investigation relating 
to C. varipennis would imply that, on the basis of the characters presented 
by Wirth and Jones, only one subspecies, sonorensis, occurs in New Mexico 
and that azstralis is only an infrasubspecific form of sonorensis. Accordingly, 
I consider australis a synonym of sonorensis. 

FEEDING HABITS: Culicoides variipennis sonorensis is the most eco- 
nomically important species in the New Mexico fauna, due to its involve- 
ment with the transmission of bluetongue virus in sheep. Price and Hardy 
(1954) incriminated C. varipennis in the transmission of bluetongue virus 
when two sheep inoculated with macerated variipennis adults developed 
clinical bluetongue. Foster, Jones and McCrory (1963) obtained positive 
biological transmission of bluetongue in sheep when they allowed variipennis 
to feed on infected sheep and later on uninfected animals. 

As shown above, this species fulfills one of the requirements of a vector 
in that it is capable of biological transmission of the bluetogue virus. Later 
investigations by Jones (1961, 1965) have shown that sonorensis will naturally 
approach sheep in the field in sufficient numbers, and during a major out- 
break of bluetongue of sheep in Colorado, sonorensis was taken in large 
numbers by sheep baited traps. 


Taste 2. Summary of data regarding morphological variation of two populations 
of Culicoides vartipennis. 


Additional 


Subspecies of Wing Palpal Antennal 
Wirth and Jones Length Ratio Sensoria 
1957 or New 
Mexico locality Mean Range* Mean Range Mean Range Aedeagus 
(TIGERS. op ee 1.54 1.39-1.69 2.53 2.19-2.87 BAS 0.93-4.00 bare 
SONOTENSIS 2. ae 1.26 1.16-1.36 223 1.99-2.47 0.83 0-2.01 spinose 
Pecos River east 

of Roswell ...... 5) 1.30-1.79 2.38 2.11-2.69 2.0 1-4 bare to 


spinose 
Salt Lake near 
oving =e eee 1.46 1.21-1.70 2.42 1.86-2.66 0.8 0-3 spinose 
* Represents 95°, limits or Wirth and Jones (1957) for australis and sonorensis. 


Tue Cuticowes of New Mexico 981 


Although sonorensis feeds on man, Jones (1959) has shown that it prefers 
cattle, sheep and man in a descending sequence. Under laboratory conditions 
it will engorge on chickens. 


Subgenus DRYMODESMYIA Vargas, 1960:40 
Type species copiosus Root and Hoffman (orig. des.). 


This subgenus contains species with hairy wings and more or less dis- 
tinctly rounded pale spots. There is often a prominent pale spot lying on 
the basal portion of vein Mi and usually a pale spot over the median portion 
of vein Me. A distinct pale spot is usually present in cell Mi. The rm 
crossvein is in a pale spot. The third palpal segment of the females is 
usually swollen with a deep sensory pit. Two sclerotized spermathecae, a 
rudimentary spermatheca and ring are present. The hind tibial comb has 
four spines. The male genitalia has the claspettes swollen basally with the 
basal portion anterio-laterally directed and usually knobbed. The stems are 
slender, lack lateral lobes and possess slender, simple and often turned, twisted 
or contorted apices. The basimere has simple roots and the aedeagus is 
usually broad with a low basal arch and stout basal arms. 


8. Culicoides (Dr.) sitiens Wirth and Hubert 
(Figs. 44-48) 
Culicoides sitiens Wirth and Hubert, 1960, Ent. Soc. Amer. Ann. 53:652 (male, female; Calli- 
fornia; fig. wing, palpus, spermathecae, male genitalia). 
FEMALE—Length of wing 1.34 mm. 

Head: Eyes separated by about half the diameter of a facet; inferior 
transocular suture present. Antenna with flagellomeres in proportion of 
13:10:11 :12:13 13:13:13 :18:18 20:21:28; antennal ratio 1.07; distal sensory 
tufts present on segments 3-9, (10), 11-15; segments 9-+-10 to 11 in ratio of 
0.69. Third palpal segment greatly swollen, 2.33 times as long as greatest 
breadth, with a deep slender sensory organ, opening by a small rounded 
opening (Fig. 45). Ratio of head length to proboscis 1.0. Mandible with 
14 teeth. 

Thorax: Legs with broad pre-apical pale rings on fore and midfemora, all 
tibiae with subbasal pale rings. Hind tibial comb with 4 spines (Fig. 46). 

Wing: (Fig. 44). Second radial cell in a dark area; pale spot on r-m 
crossvein extending from costa posteriorly beyond vein Mi+2, uniting with 
elongate spot in extreme basal portion of cell Mz; poststigmatic spot curved 
proximally around second radial cell; mesally constricted pale spot located 
in median portion of cell Rs, extending from anterior wing margin to 
slightly before vein M1; pale spots in distal portions of cells Mi and Mz not 
attaining wing margin; spot in cell My broadly joining posterior margin of 
wing; 2 rounded spots in distal portion of vannal cell, posteriormost spot 


982 Tue University ScteNcE BULLETIN 


gore AC BN 3 3 
; bth ER ae TE meee 


Fics. 44-48. Culicoides sitiens Wirth and Hubert. Female: 44, wing; 45, palpus; 46, tibial 
comb. Male: 47, claspettes; 48, genitalia, claspettes removed. 


smallest; pale spots present on basal portions of veins Mi and Mb, spot over 
vein M: barely below posterior border of poststigmatic spot; a small spot 
resting on anterior surface of medio-cubital fork. Macrotrichia abundant over 
entire surface of wing. Costa extending to 0.51 of entire length of wing. 
Abdomen: Two unequal, subpyriform spermathecae present, both crum- 
pled in only female specimen seen, ducts absent; rudimentary spermatheca 
and ring present. 
MALE—Genitalia (Fig. 47, 48) with ninth sternum very narrow, with shallow 
caudomedian excavation. Ninth tergum with prominent, distally divergent, 


Tue Cuticorwres or New Mexico 983 


apicolateral processes. Basimere with very slender ventral root; dorsal root 
broad, truncate. Telomere tapering distally with a slightly curved, mesally 
tipped apex. Aedeagus stout, broad, conical, truncate median posterior 
process; aedeagal arms short with small upturned bases; basal arch gently 
curved, extending about 0.30 of total length of aedeagus. Claspettes with 
strong basolateral arm; stem swollen, diminishing basally to a very slender, 
laterally directed apex. 

DISTRIBUTION: This is a southwestern species previously reported 
from Arizona, California and Baja California. New Mexico County Record: 
Sandoval. 

DISCUSSION: This species has been described by Wirth and Hubert 
(1960) from the rotting parts of cacti, having been reared from Opuntia and 
Lophocereus. 


9. Culicoides (Dr.) byersi new species 
(Figs. 1, 49-54) 


FEMALE—Length of wing 0.89 (0.88-0.91; n=2) mm. 

Head: Eyes barely separated to contiguous. Antenna with flagellomeres 
in proportion of 10:8:9:9:9:9:9:10:15:15:17:17:21; antennal ratio 1.13 (1.09- 
1.15; n=3); segments 9++10 to 11 in ratio of 0.73 (0.72-0.75; n=3); distal 
sensory tufts present on segments 3, 5, 7, 9, 11-15. Third palpal segment short, 
very swollen (Fig. 50), 1.7 (1.6-1.8; n=3) times as long as greatest breadth, 
with deep sensory pore. Ratio of head length to proboscis 1.67 (1.57-1.77; 
n=2). Mandible with 9-11 very small teeth. 

Thorax: Legs dark brown; fore femora with pre-apical pale band, tibiae 
with subbasal pale band. Hind tibial comb with 4 spines (Fig. 51). 

Wing: (Fig. 49). Second radial cell in a dark area; small round spot 
over r-m crossvein, barely extending beyond vein Mi+2; rather long, narrow 
poststigmatic spot; remaining spots as follows: small, round spot in median 
portion of cell Rs; pale spot in distal portion of cell Mi not attaining wing 
margin; pale spots in distal portions of cells Mz and Ma may or may not 
extend to margin; a mesally constricted pale spot in distal portion of vannal 
cell attaining wing margin; a rather indefinite pale area anterior to medio- 
cubital fork. Costa extending 0.54 of entire wing length. Macrotrichia 
usually confined to distal and posterior wing areas and veins. 

Abdomen: Two subequal spermathecae (Fig. 52) 49 x 39 and 43 x 39 p, 
ducts sclerotized for a very short distance; rudimentary spermatheca present. 
MaLE—Genitalia (Figs. 53, 54) with ninth sternum with shallow caudo- 
median excavation. Ninth tergum with prominent, divergent apicolateral 
processes. Basimere with slender, simple, ventral root; dorsal root more or 
less truncate. Telomere tapering, with swollen, tipped apex. Aedeagus with 
truncate median posterior process, in allotype male slightly flared apically; 


984 THe University SctENcCE BULLETIN 


Fics. 49-54. Culicoides byerst, new species. Female: 49, wing; 50, palpus; 51, tibial comb; 
52, spermathecae. Male: 53, claspettes; 54, genitalia, claspettes removed. 


aedeagal arms arcuate, with knobbed, laterally directed bases, basal portion 
of stem with small lobe on lateral margin immediately distal to anterior bend 
of arms; basal arch curved. Claspette with stout basolateral arm; stem 
abruptly bent posteriorly, long, tapering to a very slender, contorted apex. 

DISTRIBUTION: New Mexico (Catron Co.) and Arizona. 

Specimens examined: Holotype female, Whitewater Forest Service Camp, 5 mi. E. of 
Glenwood, New Mexico, 24 June 1953, W. W. Wirth, at light. Allotype male, Glenwood Fish 
Hatchery, Glenwood, New Mexico, 2 Aug. 1965, W. R. Atchley, at light. Paratypes, | male, 1 
female, same data as holotype; 1 female, Whitewater Forest Service Camp, 1 Aug. 1965, W. R. 


Atchley, at light. Additional specimens examined: 1 female, Mormon Lake, Coconina Co., 
Arizona, 8-15 July 1956, F. W. and F. G. Werner. 


DISCUSSION: This species is very close to Ainmani Khalaf and 
borinqueni Fox, tree-hole breeders of the copiosus group. Culicoides byersi 
can be distinguished from Ainmani by the sensorial pattern and the presence 
of a single elongate pale spot immediately posterior to the second radial cell. 


Tue Cuticomwes or New Mexico 985 


Both hinmani and borinquent have two small spots immediately posterior to 
the second radial cell. The vannal cell of byers: has a large U-shaped spot 
broadly borderng the wing margin, whereas in Ainmani the vannal cell is 
provided with two or three pale spots, the distalmost spot at best narrowly 
attaining the wing margin. C. byersi can be distinguished from boringueni 
in that the combined length of antennal segments 9 and 10 is greater than 
11; and the antennal ratio of byersi is 1.13 as opposed to approximately 1.7 of 
borinquent. 

I am very happy to name this species for my Major Professor at the 
University of Kansas, Dr. George W. Byers. 


Subgenus DIPHAOMYIA Vargas, 1960:40 


Type species bauert Hoffman (orig. des.). 


Vargas (1960) erected this subgenus for the iriarter group as defined by 
Wirth and Blanton (1959). The species of this subgenus have the second 
radial cell in a dark area and a pale spot over vein My. The hind tibial comb 
has 4 spines. Two spermathecae are present, each with a long sclerotized 
neck. The ventral root of the basimere is long, slender and curved, with a 
prominent posterior process. The claspette is knobbed basally, with an 
apical fringe of spines. There are distinct bladelike processes located on the 
postero-lateral margins of the basal arms of the aedeagus. 

Two species of the subgenus, bawert and hematopotus, are found in the 
New Mexico fauna. 


10. Culicoides baueri Hoffman 
(Figs. 55-59) 
Culicoides bauerit Hoffman, 1925, Amer. Jour. Hyg. 5:297 (male, female; Maryland; fig. wing, 
mesonotal disc). 
FEMALE—Length of wing 1.35 (1.19-1.40; n=10) mm. 

Head: Eyes separated less than diameter of a facet. Antenna with flagel- 
lomeres in proportion of 11:10:9:10:11:10:10:12:15:15:18:19:30; antennal 
ratio 1.07 (0.98-1.28; n=9), segments 9+-10 to 11 in ratio of 0.66 (0.65-0.70; 
n=6); distal sensory tufts on segments 3, 5-10, occasionally present on seg- 
ment 4 or absent from segment 5 and/or 6. Distal 4 palpal segments in 
proportion of 17:24:7:10; third palpal segment swollen, 2.17 (1.9-2.5; n=12) 
times as long as greatest breadth (Fig. 56). Ratio of head length to proboscis 
1.24 (1.0-1.44; n=9). Mandible with 10-12 (n=9) teeth. 

Thorax: Legs dark brown; fore and middle femora with pre-apical pale 
rings, all tibiae with subbasal pale rings. Hind tibial comb with 4 spines. 

Wing: (Fig. 55). Second radial cell in a dark area; poststigmatic spot 
usually small, sometimes divided into 2 spots; pale spot in cell Rs with mesal 
constriction, rarely divided into 2 spots; pale spots in cells Mi, Mz, Ma and 


986 Tue UNIversITy ScIENCE BULLETIN 


58 


Fics. 55-59. Culicoides baueri Hoffman. Female: 55, wing; 56, palpus; 57, spermathecae. 
Male: 58, claspettes; 59, genitalia, claspettes removed. 


vannal cell, all well removed from wing margin; pale spot present over vein 
Me; medio-cubital fork bordered posteriorly by a rather long pale spot; a 
small spot just posterior to r-m crossvyein, separated from spot on r-m and 
spot forming posterior border of medio-cubital fork. Macrotrichia abundant. 
Costa extending 0.52 (0.48-0.55; n=10) of distance to wing apex. 
Abdomen: Two subequal spermathecae (Fig. 57) 49 x 66 and 42 x 66 p, 
necks well developed; rudimentary spermatheca and ring present. 
MALE—Genitalia (Figs. 58, 59) with ninth sternum broad, caudomedian 
excavation deep; membrane spiculate. Ninth tergum tapering posteriorly, 
with 2 pointed apicolateral processes and slight mesal notch. Basimere with 
long prominent roots; ventral root slender, tapering, almost touching op- 
posite root mesally, with boat-shaped hook; dorsal root elongate but stouter. 


THe Curicomwes oF New Mexico 987 


Telomere tapering distally, with knobbed apex. Aedeagus with median 
posterior process sub-truncate; aedeagal arms widely separated anteriorly, 
bases sharply curved laterad, converging posteriorly to form median process; 
arms more or less sinuate with blade-like processes on posterior margin of 
shoulders. Claspette stout; broad basally, distal one-third strongly recurved, 
apex with a fringe of long slender spines. 

DISTRIBUTION: Culicoides baueri is a widespread species that has been 
reported from Calif., Colo., Fla., Ga., Md., N.Y., Okla., Tenn., Texas, and 
Wis. New Mexico County Records: DeBaca, Catron, Lincoln, Otero, Quay 
and Sandoval. 

VARIATION: There exists some variation in the pattern of pale wing 
spots in New Mexico specimens. When compared with the well defined 
patterns of the low-land specimens, representatives from some of the moun- 
tainous areas, such as at Alto (Lincoln County) exhibit some reduction in 
spot definition. 

The distal pale spot in cell Rs may vary from large and mesally con- 
stricted to two small spots. 


11. Culicoides haematopotus Malloch 
(Figs. 60-65) 
Culicoides haematopotus Malloch, 1915, Ill. State Lab. Nat. Hist. Bull. 10:302 (male, female; 

Illinois; fig. wing, male antenna, male genitalia). 

FEMALE—Length of wing 1.17 (1.0-1.3; n=15) mm. 

Head: Eyes separated less than diameter of a facet. Antenna with flagel- 
lomeres in proportion of 12:8:8:8:8:8:8:9:19:19:21 :22:30; antennal ratio 1.43 
(1.25-1.57; n=13); distal sensory tufts on segments 3, 5, 7, 9-15, occasionally 
present on segment 8 and rarely on 6 and 8 together (see discussion on varia- 
tion for second form of haematopotus); segments 94-10 to 11 in ratio of 
0.91 (0.83-1.0; n=12). Distal 4 palpal segments in proportion of 15:25:8:11; 
third segment moderately swollen, 2.5 (2.3-2.7; n=14) times as long as 
greatest breadth, with a broad but moderately shallow sensory organ (Fig. 
61). Ratio of head length to proboscis 1.11 (0.97-1.25; n=11). Mandible 
with 12-15 (n=16) teeth. 

Thorax: Legs brown; fore and middle femora with pre-apical pale rings, 
all tibiae with sub-basal pale rings; knees dark. Hind tibial comb with 4 
spines, the spine nearest the spur usually longest (Fig. 62). 

Wing: (Fig. 60). Second radial cell in a dark area; spot over r-m cross- 
vein extending from anterior wing margin to just beyond vein M1+2, may 
be connected to a second small pale spot and apparently extending to im- 
mediately anterior of medio-cubital fork; poststigmatic spot present, usually 
connected to spot over vein Mz; distal spots in cells Rs, M1, Mz, Ma and 
vannal cell, all attaining wing margin; 2 spots in vannal cell sometimes con- 


988 THe UNIversITY SCIENCE BULLETIN 


61 


Fics. 60-65. Culicoides haematopotus Malloch. Female: 60, wing; 61, palpus; 62, tibial 
comb; 63, spermathecae. Male: 64, claspettes; 65, genitalia, claspettes removed. 


nected along posterior margin of wing to form a “C”; spots present over 
mesal portions of vein Mi and Mp; a small spot at base of medio-cubital fork. 
Macrotrichia present in distal and posterior portions of the wing. Costa 
extending to 0.55 (0.52-0.56; n=15) of distance to wing apex. 

Abdomen: Two unequal, oval spermathecae (Fig. 62) 82 x 43 and 66 x 
36 », with ducts sclerotized for a long distance; rudimentary spermatheca and 
heavily sclerotized ring present. 
MALE—Genitalia (Figs. 64, 65) with ninth sternum broadly, shallowly emargi- 
nate. Ninth tergum tapering posteriorly, with short, triangular apicolateral 
processes. Basimere tapering posteriorly; ventral root long, curved, with 
pointed apex, a large boot-shaped hook on posterior margin; dosal root broad, 
truncate. Telomere slender with slightly swollen apex. Aedeagus with very 
long, straight-sided median posterior process, with serrate, truncate apex; 
aedeagal arms turned laterad basally, curved postero-medially to form median 
posterior process; prominent triangular process on apical portion of aedeagal 
arms. Claspette with stout base; stem sinuate, with large, laterally directed 
thumblike lobe approximately at mid-length; stem narrowed immediately 


Tue Curicoirs of New Mexico 989 


above lobe, turned laterad to form greatly swollen apex bearing numerous 
spines, outermost spine largest. 

DISTRIBUTION: Culicoides haematopotus 1s a common species occur- 
ring over most of the United States. New Mexico County Records: Catron, 
DeBaca, Dona Ana, Eddy, Lincoln, Quay and Socorro. 

VARIATION: Two forms of this species exist in New Mexico. One 
has a relatively small wing and an antennal sensorial pattern of 3, 5, 7, 9-15, 
and is here considered typical haematopotus; a second form possesses a large 
wing and a sensorial pattern of 3, 10-15. This second form can be char- 
acterized as having a larger wing (WL=1.38 mm.), a slightly smaller an- 
tennal ratio (AR=1.37), a slightly more swollen third palpal segment 
(PR=2.3) and a slightly longer 11th antennal segment (9-+-10 to 11=0.86). 
In this form some of the distal wing spots are reduced in size and intensity. 
No perceptible difference has been found in the male genitalia of the two 
forms. Intermediates between these two forms exist at Glenwood, Catron 
County, New Mexico. 

The description by Jamnback (1965) of Aaematopotus as it occurs in 
New York differs markedly from that of specimens identified as this species 
in the southwestern United States. The differences between the eastern and 
western forms of this species and the presence of two forms or phases men- 
tioned above would suggest that Aaematopotus as we now understand it is 
a complex of very similar species. A critical study is needed of what is now 
understood to be Aadematopotus to determine the extent of morphological and 
biological variation and to determine if this variation is specific, sub-specific 
or infrasubspecific. 

An intersex of this species was collected at Glenwood, Catron Co., New 
Mexico. 

FEEDING HABITS: Several authors have reported this species as 
biting man, but only in small numbers. Others, such as Fallis and Bennett 
(1960, 1961) and Messersmith (1965), have presented evidence that haema- 
topotus may be ornithophilic. The studies by Snow (1955) on vertical dis- 
tribution of biting flies would seem to support this latter premise, since he 
found this species to be more common in the forest canopy than at ground 
level. 


Subgenus OECACTA Poey, 1851:238 (as genus) 
Type species furens Poey (monotype). 


A rather diverse assemblage of species including those forms with the 
second radial cell usually in a dark area and commonly with a distinct wing 
pattern of pale and dark spots. Occasionally, however, the wings are more 
or less unicolorous and lack a pattern of spots. The females have two sper- 
mathecae, rarely one. The claspettes are separate and may or may not be 


990) THe University ScrENCE BULLETIN 


modified apically. The basimere has well developed ventral roots, a boat- 
shaped hook may be present on the posterior margin and the dorsal root 
is usually long and simple. The apicolateral processes of the ninth tergum 
are usually prominent. 

This subgenus, as currently defined, contains a heterogeneous accumula- 
tion of species groups. There exists considerable disagreement among various 
workers as to the contents of Oecacta. Many of the species groups now 
“lumped” into Oecacta will no doubt later be elevated to subgeneric status 
when their natural affinities become better understood. 


Culicoides lima Group 


This group consists of species with large pale spots more or less filling 
the wing cells. A pale spot is found straddling vein M2 and occasionally a 
second spot over Mi. The sensorial pattern is 3, 7-10 or 3-10 and there are 
four tibial spines. The male genitalia may or may not possess a foot-shaped 
ventral root. The apex of the claspette is fringed and a lobe or blade-like 
process may be present. 

Culicoides luglant is provisionally referred to this group, but its affinities 
are still not clear. The /imaz group, as outlined by Wirth and Blanton (1958) 
for four Panama species, possess foot-shaped ventral roots and the median 
posterior process lacks lateral projections. Both lJuglani and tenuilobus, 
which it closely resembles, have simple ventral roots and the median process 
has prominent lateral projections. In addition, all the Panama representatives 
have a sensorial pattern of 3, 7-10 while /glani has 3-10. 


12. Culicoides luglani Jones and Wirth 
(Figs. 66-70) 

Culicoides luglani Jones and Wirth, 1958, J. Kansas Ent. Soc. 31:89-91 (male, female; Texas; fig. 

wing, mesonotal disc, spermathecae, palpus, male genitalia). 
FEMALE—Length of wing 1.22 (1.17-1.26; n=3) mm. 

Head: Eyes very narrowly separated. Antenna with flagellomeres in 
proportion of 10:10:10:11:11:11:12:12:15:15:18:18:21; antennal ratio 0.95 
0.89-0.99; n=3); distal sensory tufts present on segments 3-10, rarely absent 
from 5; antennal segments 9+10 to 11 in ratio of 0.59 (0.58-0.60; n=3). 
Distal 4 palpal segments in proportion of 15:28:8:10; third palpal segment 
greatly swollen, 2.12 (2.0-2.2; n=3) times as long as greatest breadth, with a 
deep sense organ with a rounded opening (Fig. 67). Ratio of length of 
head to proboscis 1.05 (1.01-1.09; n=3). Mandible with 13-16 (n=3) teethe 

Thorax: Femora with pre-apical pale rings, tibiae with subbasal pale 
rings. Hind tibial comb with 4 spines. 

Wing: (Fig. 66). Second radial cell in a dark area; small round pale 
spot over r-m crossvein, extending from anterior margin of wing to barely be- 


Tue Curicomwes or New Mexico 99] 


69 70 


Fics. 66-70. Culicoides luglani Jones and Wirth. Female: 66, wing; 67, palpus; 68, sperma- 
thecae. Male: 69, claspettes; 70, genitalia, claspettes removed. 


yond vein Mi-+2; poststigmatic spot narrow; a large pale spot in distal por- 
tion of cell Rs; spot in distal portion of cell Mi not attaining wing margin; 
pale spots in cells Mz and Mg both attaining wing margin; 2 small pale spots 
in vannal cell, neither reaching posterior edge of wing; a pale spot circum- 
scribing vein Mz extending anteriorly to posterior edge of vein Mi; a small 
spot in extreme proximal portion of cell M» lying just below r-m crossvein; 
medio-cubital fork in a dark area. Macrotrichia confined to distal and _pos- 
terior portions of wing. Costa extending to 0.53 (0.52-0.56; n=3) of distance 
to wing apex. 


992 Tue UNiversiry ScIENcCE BULLETIN 


Abdomen: Two slightly unequal spermathecae (Fig. 68) with long 

sclerotized necks, 73 x 50 and 66 x 46 ». Rudimentary spermatheca and 
ring present. 
MALE—(Figs. 69, 70). Ninth sternum with narrow, rather shallow emargina- 
tion. Ninth tergum strongly tapering, with small triangular apicolateral 
processes. Basimere tapering distally; roots prominent, ventral root with 
simple, expanded base, not tapering to point, almost touching mesally. 
Telomere swollen basally, then narrowing to slightly swollen tip. Median 
posterior process of aedeagus elongate with two rather obscure lateral lobes 
at apex; aedeagal arms long, slender and slightly bowed mesally, bases turned 
laterad. Lateral lobes of median process much stouter in New Mexico speci- 
mens than pictured by Wirth and Jones. Claspette with knobbed base; stem 
stout basally, tapering posteriorly; mesal margin of stem relatively straight, 
lateral margin with slight median swelling; distal portion of stem gently 
recurved, bearing several small, slender spines; a large, strong, laterally 
directed spine arising just before lateral curve of stem. 

DISTRIBUTION: Texas, Arizona, Mexico. New Mexico County Rec- 
ords: Catron, DeBaca, Grants and Lincoln. 

VARIATION: The New Mexico material of this species differs some- 
what from the original description. The proboscis of the New Mexico 
specimens is considerably shorter than that of Texas individuals. 

DISCUSSION: Females of this species superficially resemble members 
of the palmerae group, particularly with regard to the wing pattern. The 
sensorial pattern will readily separate /uglani from novamexicanus n. sp. 
With regard to palmerae and utahensis, separation can be made by the shape 
of the third palpal segment, the antennal ratio and eye separation. The 
genitalia of the males are quite distinct. 

Close examination of material identified as tenwilobus Wirth and Blan- 
ton from Panama in the USNM shows /uglani and tenuilobus not to be 
synonomous as previously thought (Wirth 1963). 


Culicoides furens Group 


The furens group is composed of an assemblage of species which often 
possess pale spots in the extreme distal portions of cells Rs and Mi. The 
sensorial pattern is 3, 8-10 and there are four or five tibial spines. The sper- 
mathecae have distinct selerotized necks. The ventral root of the male 
genitalia is foot-shaped and the claspettes have simple bases, a ventral lobe 
on the stem and an apical fringe of spines. 


13. Culicoides stellifer (Coquillett) 
(Figs. 71-76) 


Ceratopogon stellifer Coquillett, 1901, U.S. Nat. Mus. Proc. 23:604. 
Culicoides stellifer (Coquillett) ; Kieffer, 1906, Gen. Insectorum, fasc. 42:55, 


Tue Curicomwres oF New Mexico 993 


FEMALE—Length of wing 1.24 (1.18-1.32; n=12) mm. 

Head: Eyes separated less than diameter of a facet; inferior transocular 
suture always present. Antenna with flagellomeres in proportion of 10:10:10: 
10:11:11 :11:11:13:16:17:17:30; antennal ratio 0.96 (0.90-1.07; n=10); distal 
sensory tufts on segments 3, 7-10, occasionally absent from 7, rarely present 
on 5 and 6; segments 9-+10 to 11 in ratio of 0.62 (0.54-0.78; n=11). Distal 4 
palpal segments in proportion of 17:22:8:10; third palpal segment moderately 
swollen, 2.47 (2.25-2.87; n=11) times as long as greatest breadth (Fig. 72). 
Ratio of head length to proboscis 1.22 (1.11-1.28; n=8). Mandible with 
12-14 (n=10) teeth. 

Thorax: Legs dark brown; fore and middle femora with pre-apical pale 
ring; all tibiae with subbasal pale rings; hind tibiae with long, apical pale 
area indistinct on fore and middle tibiae. Hind tibial comb with 5 spines 
(Eig. 73): 

Wing: (Fig. 71). Second radial cell in a dark area; spot over r-m cross- 
vein extending from costa to vein M1+2; poststigmatic spot mesally con- 


76 


Fics. 71-76. Culicoides stellifer (Coquillett). Female: 71, wing; 72, palpus; 73, tibial comb; 
74, spermathecae. Male: 75, claspettes; 76, genitalia, claspettes removed. 


994 THe UNIversity SciENcCE BULLETIN 


stricted, basal portion of spot larger and extending proximally behind second 
radial cell; spot in cell Rs “C” shaped, following anterior and distal margins 
of the wing; 2 spots in distal portion of cell Mi, apical spot and spots in 
cells Mz and Mg attaining wing margin; 2 round spots in distal part of 
vannal cell; an elongate triangular-shaped pale area lying at basal end of 
cell Mi; a long pale area in basal portion of cell Mz beginning at arculus and 
extending well beyond medio-cubital fork. Macrotrichia rather sparse, oc- 
curring primarily on marginal part of wing. Costa extending 0.54 (0.50-0.58; 
n=12) of distance to wing apex. 

Abdomen: Two unequal spermathecae (Fig. 74) 69 x 40 and 60 x 34 p, 

(n=5); ducts well developed. 
MALE—(Figs. 75, 76). Niath sternum broad, with a narrow, deep caudo- 
medial excavation. Ninth tergum tapering, with long, slender, finger-like 
apicolateral processes. Basimere with prominent roots; ventral roots elongate, 
almost touching mesally, each with a distinct hook on posterior margin; 
dorsal root long and slender. Telomere tapering, with slightly swollen apex. 
Aedeagus in shape of V; median posterior process short, posteriorly rounded; 
aedeagal arms widely separated anteriorly, basal ends sharply curved laterad, 
converging posteriorly to form median process; basal arch rounded. Claspette 
with knobbed base; stems stout, sinuate; apical third strongly recurved and 
bearing spines; stem with slightly swollen lobe located immediately anterior 
to narrowed recurved apex. 

DISTRIBUTION: This is a common and very widespread species, hav- 
ing been recorded from the majority of states in the U.S. from Canada to 
Mexico, in addition to Trinidad and Venezuela. New Mexico County Rec- 
ords: DeBaca, Lincoln, Otero, San Miguel and Taos. 

VARIATION: A female specimen from Taos County exhibits an atypical 
wing pattern in that the “C” shaped spot in cell Rs lacks the heavier basal 
area and is, therefore, reduced to a curved line along the wing margin. 
Other specimens from Lincoln Co. have this spot divided into a mesally con- 
stricted proximal portion with its prolongation reduced to a small spot in 
the very distal part of the cell, reminiscent of the pattern in C. paraensis 
(Goeldi). Many specimens exhibit considerable variation in the number and 
position of the pale spots in cell Mi. The two distal pale spots in this cell 
may be separate, narrowly joined or completely fused. A female from Bonita 
Lake, Lincoln Co. has in cell M: a triangular spot in the proximal portion of 
the cell narrowly joined to the two completely fused distal spots. 

FEEDING HABITS: This species has been reported as biting man in 
Utah by Rees and Bullock (1954), while Malloch (1915), Hoffman (1925), 
Root and Hoffman (1937) and Murray (1957) have cited man-biting records 
in the eastern states. Snow et al. (1957) found that man-biting records be- 
lieved to pertain to stellifer in the Tennessee Valley were paraensis (Goeldi). 


Tue Curicomwes oF NEw Mexico 995 


Therefore, many eastern biting records for stellifer may be in error. Khalaf 
(1966) has given characters sufficient to distinguish these two species. 


Culicoides stone: Group 


The stoner group includes species with unmarked wings, a sensorial pat- 
tern of 3-14, 3, 7-14 or 3, 9-14, pale legs without a distinct banding pattern and 
the spermathecal system lacks a sclerotized ring. The male genitalia possesses 
very long, stout, diverging apicolateral processes and the mesal margin of the 
ninth tergum has a row of stout spines. The aedeagus has slender basal arms 
and the median posterior process is broadly rounded apically . The claspettes 
possess slender anterior processes on the basal knob. 

This group, as outlined above, includes C. stoner James and two ap- 
parently undescribed species in the USNM collections. The shape of the 
anterior process on the basal knob of the claspettes, the very prominent apico- 
lateral processes and the absence of a sclerotized ring closely ally the stonei 
group with the drguttatus and spinosus groups. 


14. Culicoides stonei James 
(Figs. 77-81) 
Culicoides stonei James, 1943, Pan-Pacific Ent. 19:149 (male, female; Colorado; fig. male 
genitalia). 
Culicoides weesei Khalaf, 1952, Jour. Kansas Ent. Soc. 25:65. 
FEMALE—Length of wing 1.17 (1.06-1.22; n=6) mm. 

Head: Eye separation approximately half the diameter of a facet. An- 
tenna with flagellomeres in proportion of 11:8:9:10:10:9 :10:10:14 :13:16:18:21; 
antennal ratic 1.10 (1.05-1.15; n=6); segments 9+10 to 11 in ratio 0.69 
(0.65-0.72; n=5). Distal sensory tufts on segments 3-14 (15), occasionally 
absent from segment 4. Distal 4 palpal segments in proportion of 20:24:10:8; 
third palpal segment swollen, 2.3 (2.12.4; n=6) times as long as greatest 
breadth, with shallow sensory pore, from which arise numerous. stalked 
sensillae (Fig. 78). Ratio of head length to proboscis 1.04 (1.01-1.09; n=5). 
Mandible with 12-13 teeth (n=7). 

Thorax: Legs pale, without apparent banding pattern. Hind tibial comb 
with 4 spines. 

Wing: (Fig. 77). Wings with no pattern of light or dark spots. Costa 
extending 0.56 (0.55-0.58; n=6) of entire length of wing. Macrotrichia rela- 
tively abundant. 

Abdomen: Two large, heavily sclerotized, unequal spermathecae (Fig. 
79); ducts well developed; rudimentary spermatheca present, ring absent. 
MALE—Genitalia (Figs. 80, 81) with ninth sternum possessing narrow deep 
caudomedial excavation; membrane spiculate. Ninth tergum with very 
prominent, broad, divergent apicolateral processes; posterior portion of ninth 


996 Tue University ScIENCE BULLETIN 


79 


78 80 


Fics. 77-81. Culicoides stonet James. Female: 77, wing; 78, palpus; 79, spermathecae. Male: 
80, claspettes; 81, genitalia, claspettes removed. 


tergum distinctly notched. Basimere with long, simple roots, dorsal root 
longer than ventral root. Aedeagus with very broad, sub-truncate median 
posterior process; aedeagal arms slender, straight, divergent, with slight 
postero-laterally directed bases. Claspette stout basally; basolateral process 
with tapering anterior extension; stems narrowing slightly to sickle-like apex. 

DISTRIBUTION: Culicoides stone: has been previously reported from 
the western half of the United States, Texas to South Dakota, westward to 
California. New Mexico County Records: Chaves, DeBaca, Eddy, Quay and 
Roosevelt. 

DISCUSSION: The unmarked wings of this species resemble those of 
bottimert Wirth, but these two species can be separated by sensorial pattern 
of the females and very distinctive male genitalia. In addition, the unmarked 


Tue Cuticomwes or New Mexico 997 


wings of stone: superficially resemble various species of the subgenus Selfia; 
however, the presence of well sclerotized spermathecae in stonez will easily 
differentiate females of these species. 


Culicoides spinosus Group 


The females of this group have poorly marked wings lacking distal pale 
spots, sensoria on segments 3, 11-15 (except wsimgerz), eyes narrowly sepa- 
rated to contiguous, and the ring is absent. (Examination of type material of 
usingert Wirth showed the ring not present as previously thought.) The male 
genitalia have prominent apicolateral processes and the claspettes, which bear 
marginal spines, have stems bent at almost right angles with the bases. 

The spinosus group, as outlined above, includes spinosus Root and Hoff- 
man, Jorsae Jamnback, wsingert Wirth and subletter n. sp. These species, in- 
cluded in the diguttatus group by many authors, appear to form a closely 
related group because of the great similarity of the male genitalia as well as 
the wing pattern, sensorial pattern, etc. 


15. Culicoides sublettei new species 
(Figs. 2, 82-87) 
FEMALE—Length of wing 1.19 (1.08-1.27; n=12) mm. 

Head: Eyes separated less than diameter of a facet; inferior interocular 
suture present (very faint in allotype female). Antnna with flagellomeres in 
proportion of 10:8:8:8:10:10:11:11:15:17:18:18:25; antennal ratio 1.07 (1.03- 
1.13; n=3); segments 9+-10 to 11 in ratio of 0.67 (0.65-0.72; n=4); distal 
sensory tufts on segments 3, 11-15, rarely absent from 11. Distal 4 palpal seg- 
ments in proportion of 17:23:12:8; third palpal segment swollen (Fig. 83), 
2.37 (2.08-2.60; n=10) times as long as greatest breadth, with moderately 
shallow, round sensory pore. Ratio of head length to proboscis 1.05 (0.90-1.12; 
n=9). Mandible with 14-16 (n=11) teeth. 

Thorax: Legs brown, without distinct banding on femora or tibiae. Hind 
tibial comb with 4 spines (Fig. 84). 

Wing: (Fig. 82). Wing with faint pattern; second radial cell in a dark 
area; small, round spot over r-m crossvein, extending barely beyond vein 
Mi +2; poststigmatic spot very small; pale spots in cell My and distal portion 
of vannal cell, both attaining wing margin; a very indistinct pale spot in 
distal portion of cell Mz. Costa extending 0.55 (0.53-0.59; n=12) of length 
of wing. Macrotrichia abundant over entire wing surface. 

Abdomen: Two dark equal spermathecae (Fig. 85) 66 x 43 » (slightly 
unequal in one paratype), ducts sclerotized for a short distance; rudimentary 
spermatheca present, ring absent. 

MALE—(Figs. 86, 87). Genitalia with ninth sternum having narrow, shal- 
low posterior emargination. Ninth tergum with broad, pointed apicolateral 
processes. Basimere with slender, simple roots. Telomere arcuate, stout 


998 Tue University Science BULLETIN 


SLL ADA LEIS SE NSS 


Se NE LPS. SSK: EE 


= 7 


83 86 


Fics. 82-87. Culicoides sublettei, new species. Female: 82, wing; 83, palpus; 84, tibial 
comb; 85, spermathecae. Male: 86, claspettes; 87, genitalia, claspettes removed. 


basally, tapering distally, with simple apex. Aedeagus with short, broad, 
truncate median posterior process; aedeagal arms widely separated anteriorly, 
with small, postero-laterally directed bases, arms converging posteriorly to 
form median process; basal arch gently rounded, extending approximately 
half length of entire aedeagus. Claspette with basal end L-shaped, anterior 
process pointed; stem slender, diminishing posteriorly to slightly curved, 
pointed apex, bearing two small but distinct sub-apical spines. 

DISTRIBUTION: New Mexico (DeBaca and Quay Counties) and 
Texas. 


Specimens examined: Holotype male, Ft. Sumner, DeBaca County, New Mexico, 8 June 
1964, at light, NMDPH. Allotype female, same data. Paratypes, 1 males, 12 females: 1 male, 
7 females, Ft. Sumner, 8 June 1964; 5 females, Ft. Sumner, 15-18 June 1964. Additional speci- 
mens examined: | female, Balmorhea State Park, Reeves Co., Texas, 24 Aug. 1964, F. S. 
Blanton. 


Tue Cuticomwes or New Mexico 999 


DISCUSSION: Culicoides sublette: n. sp. belongs to the spinosus group, 
being closest to spinosus Root and Hoffman and Joisae Jamnback, both of 
which occur in the eastern United States. C. subletter is very similar to the 
figures given by Jamnback for Jorsae; however, the females may be dis- 
tinguished from this species by the presence of distinctly denticulate mandi- 
bles. The more swollen third palpal segment as well as the longer proboscis 
in relation to the head length will separate swblettei females from those of 
spinosus. 

The male genitalia of sablettei have stouter and slightly bowed aedeagal 
arms with truncate tips, while in Joisae the arms are straight. In both 
eastern species the apex of the aedeagus is rounded or sharply narrowed. 

In the western states swblettei may be separated from uwsinger: Wirth by 
the sensorial pattern and the presence of five tibial spines in wsingert. Males 
of wsingeri have a conical median posterior process of the adeagus and about 
five spines on the claspettes. 

I am very happy to name this species for my good friend and former 
teacher, Dr. James E. Sublette, who first interested me in the study of 
Diptera. 


Culicoides daedalus Group 


The daedalus group, as outlined by Wirth and Blanton (1959), has 
females with very hairy wings. A pale spot straddles vein Mz and sensoria 
are found on segments 3, 11-15 or 3, 5, 7, 9, 11-15. The pit of the third palpal 
segment is broad and shallow or deep and opening by a small pore. The male 
genitalia possess slender, unmodified ventral roots, the claspettes are pointed 
apically with distinct anterior processes on the basal knobs, and the aedeagus 
is simple. 


16. Culicoides pampoikilus Macfie 
(Figs. 88-93) 
Culicoides pampork:lus Macfie, 1948, Ann. Trop. Med. Parasit., 42:79 (female; Chiapas, Mexico; 
Pao minics Ortiz, 1951, Nov. Cient. Mus. Hist. Nat. LaSalle, zool. ser., no. 5:7. 
FEMALE—Length of wing 1.52 mm. 

Head: Eyes contiguous. Antenna with flagellomeres in proportion of 
14:12:13 12:12:12 :13 :14 :23 :24:26:27:36; antennal ratio 1.55; segments 94-10 
to 11 in ratio of 0.89; distal sensory tufts present on segments 3, 5, 7, 9, 11- 
15, occasionally present on segments 4 and 8 (only specimen from New 
Mexico has sensoria on segment 4 on | antenna and 8 on other). Distal 4 
palpal segments in proportion of 15:30:8:8; third palpal segment very 
swollen, 2.0 times as long as greatest breadth, with large, deep sensory pore 
(Fig. 89). Ratio of head length to proboscis 1.32. Mandible with 15 very 
small teeth. 


1000 Tue University SctENcE BULLETIN 


89 


Fics. 88-93. Culicoides pampoikilus Macfie. Female: 88, wing; 89, palpus; 90, tibial comb; | 
91, spermathecae. Male: 92, claspettes; 93, genitalia, claspettes removed. | 


Thorax: Hind femora darker than middle and fore pair. Banding | 
pattern not evident on available female specimen, described by Wirth and _ 
Blanton (1959) as follows: “fore and mid femora with subapical and all — 
tibiae with subbasal narrow pale rings.” Hind tibial comb with 4 spines | 
(Fig. 90). 

Wing: (Fig. 88). Second radial cell in a dark area; pale spot on r-m 
crossvein, extending posteriorly beyond vein Mi +2; small, narrow, poststig- 


Tue Cutricoiwrs oF New Mexico 1001 


matic spot; 2 pale spots in cell Rs in addition to poststigmatic spot, a large 
spot located more or less in middle of cell Rs and a small, faint spot at 
distal margin; pale spots in cells Mi, Mz and Mg all reaching wing margin; 
two pale spots in distal portion of vannal cell, distalmost attaining wing 
margin; pale spot on vein Mz; small pale spot over medio-cubital fork; large 
pale spot in proximal portion of wing broadly overlapping veins Cui, Cuz 
and 1V. Costa extending 0.54 of wing length. Macrotrichia abundant on 
entire surface. 

Abdomen: Two small spermathecae (Fig. 91) with ducts sclerotized 

for a short distance (spermathecae crumpled in New Mexico specimen); 
rudimentary spermatheca slightly expanded. 
MALE—Male genitalia described from specimen from Cartago, Navarro, 
Costa Rica (Figs. 92, 93). Ninth sternum with slight caudomedian excava- 
tion. Ninth tergum with prominent triangular apicolateral processes. Basi- 
mere long, slender; dorsal and ventral roots elongate, slender, subequal. 
Telomere tapering, with distinctly hooked apex. Aedeagus with median 
posterior process slightly expanded mesally, truncate; aedeagal arms long, 
slender, slightly sinuate, with postero-laterally directed bases; basal arch 
pointed, extending approximately 0.35 of entire length of aedeagus. Basal 
end of claspette with slender, pointed process; stem rather stout basally, 
tapering distally, with simple, strongly sinuate apex. 

DISTRIBUTION: Culicoides pampotkilus is a tropical species, having 
previously been reported from Costa Rica, Mexico, Panama and Venezuela. 
Wirth and Blanton (1959) list pamporkilus as a rare species in the humid 
tropics of Panama. The following record of this species from the moun- 
tainous region of southern New Mexico constitutes a new United States 
record: Lincoln Co., Rudioso, Cedar Creek Canyon, elev. 7100 ft., 16 July 
1965, at light, W. R. Atchley. 

DISCUSSION: The single female specimen of this species was examined 
by Dr. Willis Wirth, who concluded that it was pampotkilus. The daedalus 
group, to which this species belongs, is most numerous in western Panama, 
according to Wirth and Blanton (1959). Culicoides pampotkilus is apparently 
not closely related to any other species in the New Mexico fauna. 


Culicoides leont Group 


Wirth and Blanton (1959) erected the /eon: group for a small assemblage 
of species possessing almost bare wings. The distal portion of the wings is 
marked by two transverse rows of pale spots: one in line with the poststig- 
matic spot, the other with the distal pale spot in cell Rs. The antennal seg- 
ments of the females are in 2 series: one from 3 to 10, the other from 11 to 
15. Each series gradually increases in length from the proximal segment 
so that segment 11 is shorter than 10. Sensoria are located on segments 3, 7 


1002 Tue University ScieNcE BULLETIN 


or 8 to 10, there are 4 tibial spines and one spermatheca. The male genitalia | 


have short claspette stems which may or may not possess mesal lobes. 


17. Culicoides reevesi Wirth 
(Figs. 94-98) 
Culicoides reevesi Wirth, 1952, Univ. Calif. Pub. Ent. 9:193 (female; Calif.; fig. wing). 


FEMALE—Length of wing 0.85 (0.82-0.88; n=2) mm. 


Head: Eyes very narrowly separated, considerably less than diameter of 


a facet. Antenna with flagellomeres in proportion of 10:8:8:11:12:12:10:13:8: | 


8:13:12:18; antennal ratio 0.65 (0.64-0.66; n=2); distal sensory tufts on seg- 
ments 3, 8-10; segments 9+-10 to 11 in ratio of 0.33 (0.31-0.34; n=2) (Fig. 98). 
Distal 4 palpal segments in proportion of 12:15:6:5; third segment only 


slightly swollen, 2.1 times as long as greatest breadth; sensory organ very 


deep (Fig. 95). Ratio of head length to proboscis 1.16. Mandible with 14-16 
(n==5) teeth: 

Thorax: Legs pale brown; fore and middle femora with subapical pale 
bands; all tibiae with subbasal pale bands, hind tibia with indistinct apical 
pale band. Hind tibial comb with 4 spines (Fig. 96). 

Wing: (Fig. 94). Second radial cell in a dark area; spot on r-m cross- 


vein extending from costa to vein Mi +2; poststigmatic spot large, extending — 
almost to vein Mi; round spot in cell Rs and spot in cell M: not attaining © 


wing margin; pale spots in cells Mz and Mg attaining wing margin; spot in 
vannal cell extending along wing margin for almost entire length of cell; 
small spot in distal portion of cell Mi on proximal portion of vein Mp»; 
elongate pale area extending proximally from approximately fork of Mi+2 
and medio-cubital fork to arculus. Macrotrichia sparse, occurring only in 
distalmost portions of cells Rs, Mi and Me. Costa extending 0.56 (0.54-0.56; 
n=3) of length of wing. 

Abdomen: One subspherical spermatheca (Fig. 97) 56 x 36 »; duct 
sclerotized for some distance. 

MALE—Unknown. 

DISTRIBUTION: This species has been reported from California, 
Arizona, and New Mexico. During this investigation it was taken at the 
following localities: Catron (5 mi. e. of Glenwood, 24 June 1953, W. W. 
Wirth, biting man), Grant (Cherry Creek near Pinos Altos, 22 June 1953, 
W. W. Wirth, biting man; Roberts Lake north of Silver City, 31 July 1965, 
W.R. Atchley, biting man) and Rio Arriba counties. 

DISCUSSION: Culicoides reevesi is not closely allied to any other New 
Mexico species, but it is nearest to a group of neotropical taxa centered 
around leon Barbosa. 

The presence of a single spermatheca and the unusual shape of antennal 
segments 11 and 12 will readily separate this species. 


Tue Curicores oF New Mexico 1003 


98 


Fics. 94-98. Culicoides reevest Wirth. Female: 94, wing; 95, palpus; 96, tibial comb; 97, 
spermatheca; 98, antennal segments 9-13. 


FEEDING HABITS: This is a man-biting species previously reported 
as attacking man in California. In New Mexico it has been collected biting 
man at three localities in the western portion of he state. My own records 


indicate that this species is probably crepuscular, attacking at approximately 
sunset. 


1004 THe UNIVERSITY SCIENCE BULLETIN 


Culicoides guttipennis Group 


This group has females characterized by having the anterior half of 
the mesonotal disc with light markings on a dark background (except 
ousairant, which has the anterior half dull), well marked wings with 
macrotrichia abundant over most of the wing and 5 tibial spines. The male 
genitalia have basimeres with simple roots and the apicolateral processes of 
the ninth tergum are well developed. The claspettes are usually swollen | 
mesally, tapering distally to more or less blunt points and lacking apical 
spines. 

Vargas (1960) proposed the subgenus Glaphiromyia for this group of 
species. 


18. Culicoides pecosensis Wirth 
(Figs. 99-104) 
Culicoides pecosensis Wirth, 1955, Wash. Acad. Sci. Jour. 45:358 (male, female; Texas; fig. wing, 
mesonotal disc, palp, male genitalia). 
FEMALE—Length of wing 1.22 (1.17-1.34; n=6) mm. 

Head: Eyes very narrowly separated to contiguous. Antenna with flagel- 
lomeres in proportion of 13:13:13:13:12:12:12:12:19:20:22:23:30; antennal 
ratio 1.25 (1.21-1.34; n=5); segments 9-++10 to 11 in ratio of 0.81 (0.79-0.87; 
n=6); distal sensory tufts more or less irregular in occurrence, usually on 
segments 3-7, 9, 11-14. Distal 4 palpal segments in proportion of 23:26:11:10; 
third palpal segment moderately swollen, 2.48 (2.33-2.60; n=7) times as long 
as greatest breadth, with shallow sensory pore (Fig. 100). Ratio of head 
length to proboscis 0.93 (0.84-1.0; n=6). Mandible with 17-18 (n=6) teeth. 

Thorax: Legs dark brown; femora with pre-apical pale bands; tibiae with 
subbasal pale bands, hind tibia with broad apical band. Hind tibial comb 
with 5 spines (Fig. 101). 

Wing: (Fig. 99). Second radial cell in a dark area; spot over r-m cross- 
vein large, extending from costa well beyond vein M1+2 where it joins an- 
other rather indefinite pale area in proximal portion of wing; poststigmatic 
spot small; cell Rs with an elongate, mesally constricted spot; small round 
spots in distal portions of cell Mi and M2; pale spot extending posteriorly in 
Mz from vein M344; none of preceding spots attaining wing margin; 2 
spots in distal portions on vannal cell; pale spot over vein Mi immediately 
posterior of poststigmatic spot; a double spot bordering each side of mesal 
portion of vein M2; a small spot over medio-cubital fork; apices of veins 
Mz, Mz and M344 pale. Costa extending 0.54 (0.53-0.55; n=6) of entire wing 
length. Macrotrichia abundant over entire surface of wing. 

Abdomen: Two slightly unequal spermathecae (Fig. 102) 57 x 42 and 
47 x 35 pw; ducts sclerotized for a very short distance. Rudimentary sperma- 
theca and ring present. 


Tue Curicomwes or New Mexico 1005 


/ 
Deer OS IE: TEESE DE SRY DE (GP OK SGA BEL SEE BS! eS RE SE ERR BE BG AS 
: af 


Fics. 99-104. Culicoides pecosensis Wirth. Female: 99, wing; 100, palpus; 101, tibial comb; 
102, spermathecae. Male: 103, claspettes; 104, genitalia, claspettes removed. 


MALE—Description of genitalia based on male paratype (Figs. 103, 104). 
Ninth sternum with broad, shallow excavation. Ninth tergum narrowing 
posteriorly, with long prominent, triangular, apicolateral processes. Basimere 
with simple, prominent roots. Telomere slightly bowed, tapering apically. 
Aedeagus with extended, truncate median posterior process; aedeagal arms 
slender, straight but slightly sinuate basally; basal arch not markedly rounded, 
extending about 0.40 of entire length of aedeagus. Claspettes with strong 


1006 THE UNIversITY SCIENCE BULLETIN 


postero-laterally directed bases; stems constricted immediately behind basal 
end, then greatly expanded, mesal portion of stem very stout, distal third 
of stem abruptly narrowed, apex contorted. 

DISTRIBUTION: Culicoides pecosensts has been previously known only 
from Texas. This is the first recording of this species from New Mexico 
(Catron Co.). 

DISCUSSION: Females of this species can be separated from the other 
members of the guttipennis group by the following characters: the anterior 
half of the mesonotum has a pattern of light and dark markings which are 
absent from ousairant Khalaf; a large spot on the r-m crossvein which in 
pecosensis extends well beyond vein Mi +2, barely extends beyond vein Mi+2 
in villosipennis Root and Hoffman; the pale spot which extends along vein 
Cu: posteriorly to the wing margin will distinguish pecosensis from gutti- 
pennis (Coquillett), arboricola Root and Hoffman and oklahomensis Khalaf. 
The male genitalia of pecosensis lack the subapical aedeagal processes of 
villosipennis and oklahomensis and the truncate aedeagus of pecosensis and 
ousairant, and are distinctly different from those of guttipennis. 


19. Culicoides oklahomensis Khalaf 
(Figs. 105-108) 


Culicoides villosipennis oklahomensis Khalaf, 1952, Ann. Ent. Soc. Amer. 45:355 (male; 

Oklahoma). 

Culicoides oklahomensis Khalaf; Jones and Wirth, 1958, Jour. Kansas Ent. Soc. 31:82-85. 
FEMALE—Length of wing 1.21 mm. 

Head: Eyes narrowly separated, much less than diameter of a facet. 
Antenna with flagellomeres in proportion of 13:12:13:13:13:13:13:13:23:23: 
26:28:33, antennal ratio 1.29; distal sensory tufts present on segments 3, 5, 7, 
9, 11-15; segments 9+10 to 11 in ratio of 0.82. Third palpal segment only 
slightly swollen (Fig. 106), 3.0 times as long as greatest breadth, with very 
shallow sensory pit. Ratio of length of head to proboscis 1.04. Mandible with 
14 teeth. 

Thorax: Fore and middle femora with broad, distinct pre-apical pale 
bands, hind femora with very faint pale pre-apical band, or band may be 
absent; all tibiae with subbasal pale bands, hind tibiae with large, distinc- 
tive pale bands. Hind tibial comb with 5 spines. 

Wing: (Fig. 105). Second radial cell in a dark area; light spot on r-m 
crossvein large, extending from costa to slightly beyond vein Mi +2; poststig- 
matic spot small; an elongate mesally constricted spot in cell Rs; pale spots in 
cells Mi and Mz not attaining wing margin; pale spot in cell My and 
distalmost spot in vannal cell attaining wing margin; pale spot on anterior 
surface of medio-cubital fork; pale spot crossing basal portion of vein Mi; 
pair of spots bordering vein Mg slightly basad of its midlength; apices of veins 


Tue Cuticomwes oF New Mexico 1007 


BC OP ie 
\ 


a a (AAS RERA: 
fohchh tL EE EE iy st 3 


Af tito rk 


BE 


107 


Fics. 105-108. Culicoides oklahomensis Khalaf. Female: 105, wing; 106, palpus. Male: 107, 


claspettes; 108, genitalia, claspettes removed. 


My, Me, M344 and Cui pale. Macrotrichia very abundant over entire surface 
of wing. Costa extending to 0.56 of distance to wing apex. 

Abdomen: Two subequal spermathecae with very short ducts, 53 x 36 
and 50 x 50 4; rudimentary spermatheca and ring present. 
MaLE—Genitalia (Figs. 107, 108) with ninth sternum narrow with very broad, 
deep caudomedian excavation. Ninth tergum tapering, with prominent 
blade-like apicolateral processes; median posterior portion distinctly notched. 
Basimere with simple, slender, well developed roots, dorsal root longer than 


1008 Tue UNIversity SctENcE BULLETIN 


ventral root. Telomere tapering to a slender, slightly curved, tipped point, 
Median posterior process of aedeagus broad basally, extending to very nar- 
row, pointed apex, bearing a pair of subapical filaments; aedeagal arms 
slender and arcuate, with small, knobbed bases; basal arch notched mesally, 
extending to approximately half of overall length of aedeagus. Claspette 
with expanded basal foot containing tapered anterior process; stem very 
swollen, sinuate, abruptly narrowing to slender, unmodified recurved apex. 

DISTRIBUTION: Culicoides oklahomensis, originally described as a 
subspecies of villosipennis, occurs in the southwestern part of the United 
States and in Mexico, Within the United States it has been reported from 
Oklahoma, Texas and Arizona. During this investigation oklahomensis was 
collected in Catron Co., New Mexico. 

DISCUSSION: Males of this species can be separated from most others 
of the guttpennis group by the presence of lateral projections of the 
aedeagus. The only other species of the guttipennis group with such pro- 
jections is the eastern villosipennis Root and Hoffman. The male genitalia of 
these two species can be distinguished by the shape of the claspettes. The 
females of oklahomensis can be recognized by the sensorial pattern, the 
large pale spot over the r-m crossvein and the pale border of vein Cun. 


Culicoides palmerae Group 


This group, composed of palmerae, utahensis, and novamexicanus, has 
male genitalia characterized by very prominent apicolateral processes of the 
ninth tergum, strongly bent distal two-thirds of the telomere, broad aedeagus, 
and the well developed basal portions of the claspettes and roots of the 
basimeres. The females have a swollen third palpal segment, widely sepa- 
rated eyes, banding pattern on the legs very faint or absent and wings with 
dense macrotrichia. 


20. Culicoides palmerae James 
(Figs. 109-112) 
Culicoides palmerae James 1943, Pan-Pacific Ent. 19:151 (male, female; Colorado; fig. male 
genitalia). 
FEMALE—Length of wing 1.48 (1.34-1.69; n=3) mm. 

Head: Eye separation approximately equal to diameter of a facet. An- 
tenna with flagellomeres in proportion of 12:9:9:9:10:10:11 :12:18 :18 :22:23 :33; 
antennal ratio 1.28 (1.18-1.43; n=3); distal sensory tufts on segments 3-15; 
segments 9-+-10 to 11 in ratio of 0.77 (0.75-0.78; n=2). Distal 4 palpal seg- 
ments in proportion of 20:26:11:12; third palpal segment swollen (Fig. 110), 
2.26 (2.07-2.45; n=3) times as long as greatest breadth, with broad, rather 
shallow sensory pore. Ratio of head length to proboscis 1.02 (0.89-1.15; 
n=4). Mandible with 14 teeth. 


Tue Cuticowes or New Mexico 1009 


110 11 


Fics. 109-112. Culicoides palmerae James. Female: 109, wing; 110, palpus. Male: 111, 
claspettes; 112, genitalia, claspettes removed. 


Thorax: Legs brown; knees dark; tibiae with faint subbasal pale rings. 
Hind tibial comb with 4 spines. 

Wing: (Fig. 109). Wing pattern indistinct; second radial cell in a dark 
area; pale spot over r-m crossvein, extending slightly posterior of vein Mi +2; 
poststigmatic spot small; an elongate, rather indistinct pale spot in distal por- 
tion of cell Rs; indistinct pale spots in cells Mi, Mz, Ms and vannal cell, all 
attaining wing margin; medio-cubital fork in dark area. Costa extending 
0.56 (0.55-0.57; n=3) of total length of wing. Macrotrichia abundant over 
entire surface of wing. 

Abdomen: Two elongate, subequal spermathecae 76 x 39 and 76 x 43 p; 
ducts not sclerotized; rudimentary spermatheca and ring present. 
MALE—Description of male genitalia (Figs. 111, 112) based on male paratype 
from Ft. Collins, Colo. Ninth sternum with slight caudomedian excavation; 
membrane bare. Ninth tergum narrowing slightly posteriorly; posterior 
border very convex, with distinct mesal notch; apicolateral processes long, 
prominent, truncate. Basimere tapering distally; roots long, slender, sub- 
equal. Telomere strongly bent at approximate basal third, abruptly narrowed 


1010 Tue UNiversiry ScrENCE BULLETIN 


at bend, of more or less uniform width to apex. Aedeagus with median 
posterior process broad, with subapical lateral shoulders bearing posterior 
spines, apex truncate; adedeagal arms widely divergent basally, with promi- 
nent, laterally directed basal ends, arms slightly arcuate mesally, converging 
posteriorly; basal arch rounded, extending to about midlength of aedeagus. 
Claspettes with swollen bases; stems swollen basally, tapering posteriorly, 
apex strongly bent laterad then cephalad; apex with a few microscopic hairs. 

DISTRIBUTION: Culicoides palmerae has previously been reported 
from the northwestern portions of the United States. Its range extends from 
Nebraska southwestward to California and northwestward to British Colum- 
bia. These first records from Sandoval and Taos counties in New Mexico 
constitute its southernmost known occurrence. 

DISCUSSION: This species is very similar to utahensis Fox but may be 
separated from it by a moderately swollen third palpal segment with a 
broad pit. The wing pattern of palmerae is distinctly marked, as opposed 
to the poorly marked wing of utahensis. The presence of lateral shoulders 
on the aedeagus and slender claspettes will readily distinguish the males. For 
separation from novamexicanus n. sp., see discussion of that species. 


21. Culicoides utahensis Fox 
(Figs. 113-115) 
Culicoides utahensis Fox, 1946, Ent. Soc. Wash. Proc. 48:246 (male; Utah; fig. male genitalia). 


FEMALE*—Length of wing 1.43 mm. 

Head: Eyes widely separated. Antenna with flagellomeres in proportion 
of 13:10:10:10:10:11 :11:12:18 :18 :20:21:25; antennal ratio 1.12; distal sensory 
tufts present on segments 3-15; segments 9+-10 to 11 in ratio of 0.75. Distal 
4 palpal segments in proportion of 12:26:9:12; third palpal segment greatly 
swollen (Fig. 113), 1.85 times as long as greatest breadth, with deep sensory 
pit. Ratio of head length to proboscis 1.30. Mandible with 14 teeth. 

Thorax: Legs dark brown. Hind tibial comb with 4 spines. 

Wing: The following description is taken from Wirth (1952): “Wings 
grayish hyaline, anterior radial cells in a very dark spot, distinct light costal 
spots at r-m crossvein and at apex of second anterior radial cell, other light 
spots variable and indistinct, in well marked specimens with large light 
spots at tips of cells Rs, Mi, Mz, Ma and anal cells, and large light areas at 
base of anal cell and in middle of wing between veins Mi+2 and Cui. Wings 
with abundant long yellowish macrotrichia over entire surface.” Costa ex- 
tending 0.55 of total length of wing. 


*No females of this species were collected during this study. The description was based on 
specimens collected from ears of rabbits at Battle Mountain, Nevada. 


THe Cuticomwes or New Mexico 1011 


114 


Fics. 113-115. Culicoides utahensis Fox. Female: 113, palpus. Male: 114, genitalia, claspettes 
removed; 115, claspettes. 


Abdomen: Two subequal spermathecae 59 x 49 and 66 x 43 »; ducts 

sclerotized for a very short distance. Rudimentary spermatheca and ring 
present. 
MALE—(Figs. 114, 115). Genitalia very large. Ninth sternum with slight 
caudomedian excavation, membrane bare. Ninth tergum only slightly taper- 
ing posteriorly; caudal portion convex, with prominent mesal notch; apico- 
lateral processes long and distinct, of uniform width for most of length. 
Basimere rather short, slightly tapering posteriorly with large distinct mesal 
shelf on basal half, shelf more or less flat posteriorly; short ventral root and 
long, slender, mesally blunt dorsal root. Telomere swollen basally, sharply 
bent at basal third, apical two-thirds narrowed, approximately uniform width 
for remainder of length. Aedeagus very stout; median posterior process 
broad, truncate, with slightly sinuate apex; aedeagal arms short, with laterally 
directed bases; basal arch rounded, extending only about 0.35 of length of 
aedeagus. Expanded base of claspette with curved anterior process; stem 
strongly swollen basally, tapering to laterally turned, lancolate apex. 


1012 Tue UNiversiry ScrENCE BULLETIN 


DISTRIBUTION: Previous records of utahensis indicate that it is 
limited to far western portions of the United States, having been reported 
from Utah, Nevada and California. The discovery of this species in the 
mountainous region of southern New Mexico (Lincoln Co., male) represents 
its easternmost record. 

DISCUSSION: The high incidence of males with female head parts 
in this species has prompted considerable discussion. Downes (1958) dis- 
cussed this phenomenon at some length and suggested the possibility of 
more than one species being included under the name wtahensis. Both sexes 
of this species have been taken from the ears of jackrabbits on numerous 
occasions, and the males taken in this fashion have antennae and palps of 
the female type. This modification is assumed to assist the males in locating 
the host of the female and thereby facilitate mating. Copulation is probably 
initiated by contact. The modification of the head structures of the males of 
this species may represent evolution toward parasitism, instead of the pre- 
datory behavior exhibited by most other species of Culicoides. 

So far as known all male representatives of this species from New Mexico 
have female head parts. 

See discussion of palmerae and novamexicanus n. sp. for characteristics 
sufficient to distinguish these species. 


22. Culicoides novamexicanus new species 
(Figs. 116-121) 


FEMALE—Length of wing 1.23 (1.13-1.36; n=3) mm. 

Head: Eyes separated about diameter of a facet. Antenna with flagel- 
lomeres in proportion of 13:10:10:10:10:10:11:12:15:16:18:18:25; antennal 
ratio 1.02 (0.98-1.05; n=3); distal sensory tufts on segments 3-5, 7, 9, 11, 
13-15, occasionally absent on segment 4 or present on 6 or 12; segments 
9+10 to 11 in ratio of 0.58 (0.55-0.60; n=3). Distal 4 palpal segments in 
proportion of 21:28:12:13; third palpal segment swollen (Fig. 114), 2.12 
(2.0-2.2; n=3) times as long as greatest breadth; sensory pit moderately 
deep. Ratio of head length to proboscis 1.04 (1.01-1.06; n= 3). Mandible 
with 14-16 (n=3) teeth. 

Thorax: Legs brown; tibiae with very faint, narrow subbasal pale rings. 
Hind tibial comb with 4 spines (Fig. 118). 

Wing: (Fig. 116). Second radial cell in a dark area; poststigmatic spot 
large, bounded posteriorly by vein M:; pale spot on r-m crossvein large, ex- 
tending from anterior wing margin well beyond vein Mi +2, connecting with 
large pale area in proximal portions of cells Mi and Mp; pale spot in cell 
Msg and distalmost spot in vannal cell broadly joining wing margin; 2 large 
pale spots in proximal portions of cells Mi and Me broadly encompassing 
basal half of vein Me; basal portion of vannal cell and extreme basal area 


Tue Cuticoiwes oF New Mexico 1013 


tia 121 


Fics. 116-121. Culicoides novamexicanus, new species. Female: 116, wing; 117, palpus; 118, 
tibial comb; 119, spermathecae. Male: 120, claspettes; 121, genitalia, claspettes removed. 


of cell Mz in pale area; mediocubital fork in dark area; distal ends of veins 
Mi, Me Ms+4 and Cur dark. Macrotrichia abundant over entire surface of 
wing. Costa extending 0.56 (0.55-0.57; n=3) of length of wing. 

Abdomen: Two subequal spermathecae (Fig. 119) 69 x 46 and 56 x 43 »; 
ducts sclerotized for a very short distance; rudimentary spermatheca present. 
MALE—(Figs. 120, 121). Genitalia with ninth sternum divided; narrow, 
shallow, pointed caudomedian emargination. Ninth tergum tapering; 


1014 Tue UNIversiIty SCIENCE BULLETIN 


postero-medial portion convex, with prominent notch; apicolateral processes 
distinctly long and fingerlike. Basimere with dorsal and ventral roots sub- 
equal in size, without processes; dorsal root more slender than stout ventral 
root. Telomere swollen basally, abruptly bent and narrowed at approximately 
one-third of length from base, continuing at more or less uniform width to 
apex. Aedeagus stout; median posterior process conical, truncate; aedeagal 
arms strong, diverging anteriorly, basal portion of each turned postero- 
laterally to give footlike appearance; basal arch rounded, extending to ap- 
proximately half length of aedeagus. Claspette with prominent foot-shaped 
base, with microscopic fringe of lateral spines on anterior “toe” portion; stem 
stout basally but tapering to slender, helical apex. 
DISTRIBUTION: Chaves County, New Mexico. 


Specimens examined: Holotype male, Pecos River, 7 mi. e. of Roswell, New Mexico, 14 June 
1965, W.R. Atchley, at light. Allotype female and 3 female paratypes, same data. 

DISCUSSION: Culicoides novamexicanus can be separated from pal- 
merae by the sensorial pattern, smaller antennal ratio, lack of lateral projec- 
tions on the aedeagus and the slender, simple claspettes; and from 
utahensis by the widely separated eyes, sensorial pattern, more extensive pale 
markings on the wings, the slender aedeagus and aedeagal arms and the 
characteristic claspettes. 


Culicoides piliferus Group 


The wings of the females of the piliferus group usually have pale spots 
straddling the midportions of veins Mi and Me; however, the pattern is 
sometimes very reduced. The sensoria are found on segments 3, 5, 7, 9, 
11-15 in the generalized condition, but may be reduced to the distal flagel- 
lomeres in some species. The eyes are contiguous to moderately separated 
and hind tibial comb possesses 4, occasionally 5, spines. The spermathecae 
are subequal to very unequal in size and usually lack necks. The ventral 
root of basimere is foot-shaped and the claspettes taper to a fine tip with a 
lateral fringe of spines. 


23. Culicoides doeringae new species 
(Figs. 122-129) 


FEMALE—Length of wing 1.24 (1.13-1.34; n=14) mm. 

Head: Eyes narrowly separated (Fig. 129); in holotype, separation ap- 
proximately equal to width of median hair socket. Antenna with flagel- 
lomeres in proportion of 13:8:8:8:8:8 :9:9:22:23:23:23:35; antennal ratio 1.63 
(1.48-1.73; n=14), segments 9+10 to 11 in ratio of 1.11 (1.0-1.27; n=14) 
(Fig. 126); distal sensory tufts on segments 3, 5, 7, 9, 11, 13-15, occasionally 
also on 12. Distal 4 palpal segments in proportion of 22:23:10:9; third seg- 
ment moderately broad (Fig. 123), 2.61 (2.3-2.87; n=12) times as long as 


Tue Cuticoies oF New Mexico 1015 


Fe rey aD 


a (x 
LEA EL TIERS a CG 
eer 


128 126 


Fics. 122-129. Culicoides doeringae, new species. Female: 122, wing; 123, palpus; 124, 
tbial comb; 125, spermathecae; 126, antennal segments 9-11; 127, interocular space. Male: 128. 
claspettes; 129, genitalia, claspettes removed. 


greatest breadth, with shallow sensory pit. Ratio of head length to proboscis 
1.12 (1.03-1.20; n=12). Mandible with 12-15 (n=12) teeth. 

Thorax: Legs brown; fore and middle femora with pre-apical pale bands, 
all tibiae with subbasal pale bands. Hind tibial comb with 4 spines (Fig. 
Vay 

Wing: (Fig. 122). Second radial cell in a dark area; small round spot 
over r-m crossvein barely extending beyond vein M1i+2; poststigmatic spot 
small; pale spot in distal portion of cell Rs; pale spots in distal portions of 
cells Mi, Mz and Mg all attaining wing margin; spot in distal portion of 
vannal cell variable, usually large, round anteriorly, mesally constricted in 
many specimens, with posterior portion either expanded or, as in holotype, 


1016 Tue UNIversITy SCIENCE BULLETIN 


extended to wing margin in same width as mesal constriction; pale spot 
on vein M: immediately behind poststigmatic spot; pale spot near mid- 
length of vein Me; a rather elongate pale area bounded posteriorly by medio- 
cubital fork extending anteriorly almost to veins Mi+2 and Mo. Costa ex- 
tending 0.58 (0.56-0.59; n=14) of wing length. Macrotrichia abundant over 
entire surface of wing. 

Abdomen: Two very unequal spermathecae (Fig. 125) 76 x 56 and 
56 x 39 mw, ducts unsclerotized; rudimentary spermatheca and ring present. 
MALE—Genitalia (Figs. 128, 129) with ninth sternum possessing deep caudo- 
median excavation. Ninth tergum with small, triangular apicolateral pro- 
cesses, mesal notch distinct. Basimere tapering slightly; ventral root tri- 
angular, with definite “heel-like” process on mesal margin; dorsal root long, 
slender. Telomere with swollen base, narrowing distally to mesally tipped 
apex. Aedeagus with long, parallel-sided, truncate median posterior process; 
aedeagal arms slightly knobbed basally, arcuate mesally; basal arch gently 
rounded, extending approximately 0.45 length of aedeagus. Claspettes dis- 
tinctly knobbed basally; stems sinuate, gradually diminishing toward tip, 
with a fringe of apical spines. 

DISTRIBUTION: New Mexico (Catron, Lincoln and Otero Counties) 
and Arizona. 

Specimens examined: Holotype female, Catron Co., Glenwood Fish Hatchery, 2 Aug. 1965, 
W. R. Atchley, at light. Allotype male, Lincoln Co., Alto Reservoir, 16 June 1965, W. R. Atchley, 
at light. Paratypes: 18 females, same data as holotype; 1 male, 2 females, same data as 
allotype; 4 males, 9 females, Lincoln Co., Ruidoso, Cedar Creek Canyon, 16 June 1965, W. R. 
Atchley, at light; 5 females, Otero Co., Silver Creek Canyon, 28 July 1965, W. R. Atchley, at 
light. Additional specimens examined: 2 females, Coconino Co., Oak Creek Canyon, Arizona, 22 
July 1959, C. W. O’Brien; 1 female, Cochise Co., Southwest Research Station, Arizona, 1 Aug. 
1964, D. R. Davis; 1 female, Cochise Co., Chiricahua Mts., 2 mi. n. of Paradise, Arizona, 6 July 
1964, D. R. Davis. 

DISCUSSION: Culicoides doeringae belongs to the piliferus group and 
is probably most closely related to the eastern piliferus Root and Hoffman. 
These two species may be differentiated by the slightly smaller antennal 
ratio, the greater ratio of antennal segments 9+-10 to 11 and the sensoria 
pattern 3, 5, 7, 9, 11, 13-15 in doeringae. 

Atchley and Wirth (1967) have given characters sufficient to distinguish 
doeringae from riggsi Khalaf, another southwestern species of the piliferus 
group. 

It gives me great pleasure to name this species for Dr. Kathleen C. 
Doering, Professor Emerita of insect morphology at the University of 
Kansas. 


LITERATURE CITED 


Aput-Nasr, S. E. 1950. Structure and development of the reproductive system of some species 
of Nematocera. Philos. Trans. Roy. Soc. London, Ser. B. 234:339-396. 

Axipa, K. 1960. Studies on the Lewcocytozoon found in the chicken. II. On the transmission 
of Leucocytozoon caulleryi by Culicoides arakawae. Jap. Jour. Vet. Sci. 22:309-317. 


Tue Cuticoiwes of New Mexico 1017 


Arnaup, P. 1956. The heleid genus Culicoides in Japan, Korea and Ryukyu Islands. Micro- 
entomology 21:84-207. 

Arcuey, W. R.,anp W. W. Wirtn. 1967. A redescription of Culicoides riggst Khalaf with 
notes on C. bickleyi Wirth and Hubert in Oklahoma. Jour. Kansas Ent. Soc. 40: (In 
Press). 

Barto, R. 1961. Das Sinnesorgan der Maxillarpalpen von Forcipomyia fuliginosa (Meigen). 
Ana. Acad. Bras. Ciencias. 33:100-105. 

Bennett, G. F. 1960. On some ornithophilic bloodsucking Diptera in Algonquin Park, Ontario. 

Canad. Jour. Zool. 38:377-389. 

. 1961. On the specificity and transmission of some Avian trypanosomes. Canad. Jour. 

Zool. 39:17-33. 

BENNETT, G. F., anp A. M. Fauuts. 1960. Blood parasites of birds in Algonquin Park, Canada, 
and a discussion of their transmission. Canad. Jour. Zool. 38:261-273. 

Buck ey, J. C. C. 1934. On the development in Culicoides furens Poey, of Filaria (Mansonella) 

ozzardi Manson, 1897. Jour. Helminth. 12:99-118. 

. 1938. On Culicoides as a vector of Onchocerca gibsoni (Cleland and Johnston, 1910). 

Jour. Helminth. 16:121-158. 

CAMPBELL, J. A., AND E. C. PELHAM-CLINTON. 1960. A taxonomic review of the British species 
of Culicoides Latreille (Dipt.: Ceratopogonidae). Proc. Roy. Soc. Edinburgh, Ser. B. 
67:181-302. 

Carter, H. F,, A. INGRAM anv J. W. S. Macrie. 1920. Observations on the Ceratopogonine 
midges of the Gold Coast with descriptions of new species. Ann. Trop. Med. Parasitol. 
14:187-210. 

Causey, O. R. 1938. Culicoides of Siam with descriptions of new species. Amer. Jour. Hyg. 
27:399-416. 

CHarDRoME, M., anp E. Peer. 1949. La repartition des filairs dans la region de Coquihatuille 
et la transmission de Dipetalonema streptocerca par Culicoides grahami. Ann. Soc. Belge 
Med. Trop. 29:99-119. 

CuristorHeRs, S. R. 1922. The development and structure of the abdominal segments and 
hypopygium in the male mosquito with observations on the terminal segments of the 
larva. Indian Jour. Med. Res. 10:530-572. 

CHRISTOPHERS, S.R., AND P. J. Barraup. 1926. The development of the male and female 
hypopygium of PAlebotomus. Indian Jour. Med. Res. 9:445-463. 

Crampton, G. C. 1942. The external morphology of the Diptera, p. 10-165. Im The Diptera 
or true flies of Connecticut. Guide to the Insects of Connecticut. Part VI. State Geol. 
and Nat. Hist. Sur., No. 64. 

Dampr, A, 1936. Los Ceratopogonidos o jejenes, como transmisores de filarias. Med. Rev. 
Mexicana. 16:227-233. 

Downes, J. A. 1958a. Assembly and mating in the biting Nematocera. Proc. X Int. Congr. 

Ent. 2:425-434. 

.1958b. The feeding habits of biting flies and their significance in classification. Ann. 

Rev. Ent. 3:249-266. 

.1958c. The genus Culicoides (Diptera: Ceratopogonidae) in Canada; an introductory 

review. Proc. Tenth Int. Congr. Ent., Montreal 3:801-808. 

Duke, R. O. L. 1954. The uptake of the microfilariae of Acanthocheilonema streptocerca by 
Culicoides grahamit and their subsequent development. Ann. Trop. Med. Parasitol. 
48:416-420. 

Du Torr, R. M. 1944. The transmission of Blue Tongue and Horse Sickness by Culicoides. 
Onderstepoort Jour. Vet. Res. 19:7-16. 

Epmunps, L. R., anp G. G. Keener, Jr. 1954. Observations of the biting habits of Culicoides 
crepuscularis, Mall. in western Nebraska, with notes on other species collected in light 
traps (Diptera: Heleidae). Mosquito News 14:82-83. 

Epwarps, F. W. 1926. On the British biting midges (Diptera: Ceratopogonidae). Trans. Ent. 

Soc. London 74:389-426. 

. 1939. Ceratopogonidae. In F. W. Edwards, H. Oldroyd, and J. Smart. British Blood- 

sucking Flies, p. 25-50. British Museum Natural History, London. 

Fauus, A. M., ano G. F. Bennetr. 1960. Description of Haemoproteus canochites n. sp. 

(Sporozoa: Haemoproteidae) and sporogony in Culicoides. Canad. Jour. Zool. 38:455- 
464. 

. 1961. Sporogony of Leucocytozoon and Haemoproteus in simulids and ceratopogonids 
and a revised classification of the Haemosporidiida. Canad. Jour. Zool. 30:217-228. 

Faris, A. M., AND D. M. Woop. 1957. Biting midges as intermediate hosts for Haemoproteus 
of ducks. Canad. Jour. Zool. 35:425-435. 


1018 Tue University ScrENcE BULLETIN 


FENNEMAN, N. M. 1931. Physiography of the Western United States. McGraw-Hill Co., New 
York. 

Frepter, O. G. H. 1951. The South African biting midges of the genus Culicoides. Onder- 
stepoort Jour. Vet. Res. 25:3-33. 

Foorr, R. H., ano H. D. Pratr. 1954. The Culicoides of the eastern United States. Publ. 
Health Monogr., No. 18:55 p. 

Foster, N. M., R. H. Jones anp B. R. McCrary. 1963. Preliminary investigations on insect 
transmission of bluetongue virus in sheep. Amer. Jour. Vet. Res. 24:1195-1200. 

Fox, I. 1946. Two new biting midges of Culicoides from western United States. Proc. Ent. Soc. 

Wash. 48:244-246. 

. 1948. Hoffmania, a new subgenus in Culicoides (Diptera: Ceratopogonidae). Proc. Biol. 

Soc. Wash. 61:21-28. 

. 1955. A catalogue of the bloodsucking midges of the Americas (Culicoides, Leptoconops 
and Lasiohelea) with keys to the subgenera and Nearctic species, a geographic index and 
bibliography. Jour. Agr. Univ. Puerto Rico 39:214-285. 

Gap, A. M. 1951. The head-capsule and mouthparts in the Ceratopogonidae. Bull. Soc. Fouad 
ler. d’Ent. 44:17-75. 

GarnuHAM, P. C. C., R. B. HeEIscH AND D. M. Minter. 1961. The vector of Hepatocystis 
(=Plasmodium) kochi, the successful conclusion of observations in many parts of 
tropical Africa. Trans. R. Soc. Trop. Med. Hyg. 55:497-502. 

Gisson, C. L., anp W. F. Ascorrt. 1952. The relation of Culicoides to the transmission of 
Onchocerca volvulus. Jour. Parasitol. 38: 315-320. 

GOETGHEBUER, M. 1920. Ceratopogoninae de la Belgique. Memoires du Musee d'Histoire 

Naturelle de Belgique 8 (fasc. 3) :1-116. 

. 1933-34. Heleidae (Ceratopogonidae). Im Erwin Linder, Die Fliegen der Palaearktischen 
Region. Lieferung 77 (pp. 1-48, Tafel I-VI, 1933); Lieferung 78 (pp. 49-94, Tafel VII- 
XII, 1934), Stuttgart. 

HENRARD, C., AND E. PEEL. 1949. Culicoides grahami Austen, vecteur de Dipetalonema strepto- 
cerca et non de Acanthocheilonema perstans. Ann. Soc. Belge Med. Trop. 29:127-143. 

HorrMan, W. A. 1925. A review of the species of Culicoides of North and Central America 
and the West Indies. Amer. Jour. Hyg. 5:274-301. 

Hopkins, C. A., AnD W. L. NicHoras. 1952. Culicoides austent, the vector of Acanthocheilonema 
perstans. Ann. Trop. Med. Parasitol. 46:165-172. 

Hoy, JAMes B. 1966. The behavior of members of the genus Symphoromyia attacking deer in 
northern California (Diptera Rhagionidae). Ph.D. Dissertation, University of Kansas 
162 p. 

INGRAM, A., AND J. W. S. Macrie. 1921. West African Ceratopogonidae. Ann. Trop. Med. 
Parasitol. 15:313-374. 

James, M. T. 1941. The genus Culicoides in northern Colorado. Pan-Pacific Ent. 19:148-153. 

Jamnsack, H. 1965. The Culicordes of New York State. New York State Museum Bull. 399: 
154) \p: 

, AND W. W. WirtH. 1960. The species of Culicoides related to obsoletus in eastern 
North America. Ann. Ent. Soc. Amer. 56:185-198. 

Jostinc, B. 1928. The structure of the head and mouth-parts of Culicoides pulicarts L. Bull. 
Ent. Res. 18:211-236. 

JoHaNNSEN, O. A. 1943. A generic synopsis of the Ceratopogonidae of the Americas, a bibliogra- 

phy, and a list of the North American species. Ann. Ent. Soc. Amer. 36:763-791. 

. 1952. Guide to the insects of Connecticut. Part VI. The Diptera or true flies of 

Connecticut. Fifth fascicle; midges and gnats. Part VI. Family Heleidae (=Cera- 

topogonidae). Conn. State Geol. and Nat. Hist. Survey Bull. 80:149-175, 232-250. 

Jones, R. H. 1955. The Culicoides of Wisconsin. Ph.D. Dissertation, University of Wisconsin. 

. 1959. Culicoides breeding in human sewage sites of dwellings in Texas. Mosquito News 
19:164-167. 

—. 1961. Some observations on biting flies attacking sheep. Mosquito News 21:113-115. 

—. 1961. Descriptions of pupae of thirteen North American species of Culicoides. Ann. 
Ent. Soc. Amer. 54:719-746. 

—. 1965. Epidemiological notes: Incidence of Culicoides varitpennis in an outbreak of 
bluetongue disease. Mosquito News 25:217-218. 

Jones, R. H., anp W. W. WirtH. 1958. New records, synonomy and species of Texas Culicoides 
(Dipt.: Heleidae). Jour. Kansas Ent. Soc. 31:81-91. 

Karstap, L. H., O. K. FLercuer, J. SPALATIN AND R. P. Hanson. 1957. Eastern equine en- 
cephalomyelitis virus isolated from three species of Diptera from Georgia. Science 
125:395-396. 


Tue Cuticomwrs oF New Mexico 1019 


Kertie, D. S. 1962. The Bionomics and Control of Culicoides and Leptoconops (Diptera, 
Ceratopogonidae=Heleidae). Ann. Rey. Ent. 7:401-418. 

Kuarar, K. 1952. The Culicoides of the Wichita Refuge, Oklahoma. Taxonomy and seasonal 

incidence. Ann. Ent. Soc. Amer. 45:348-358. 

. 1954. The speciation of the genus Culico:des (Diptera: Heleidae). Ann. Ent. Soc. 
Amer. 47:34-51. 

. 1957. Light trap survey of the Culicoides of Oklahoma. Am. Midland Naturalist 58: 
182-221. 

. 1966. Notes on the taxonomy of Culicoides from Louisiana. Jour. Kansas Ent. Soc. 
39:227-231. 

KierFer, J. J. 1906. Diptera, Fam. Chironomidae. Jn F. Wytsman, Gen. Insectorum, Fasc. 

42:78 p. 

—-—. 1925. Faune de France. #11. Diptéres (Nematoceres piqueurs): Chironomidae, 
Ceratopogonidae. Paul Lechevalier, Paris, 1-139. 

LaTrreILLE, P. A. 1809. Genera crustaceorum et insectorum secundum ordinem naturalem in 
familias disposita, inconibus exemplisque plurimis explicata. 4:399 p. 

Leg, D. J., anp E. J. Reve. 1953 (1952). Australian Ceratopogonidae. Part VI. Australian 
species of Culicoides. Proc. Linn. Soc. N. S. Wales 79:233-246. 

anpD A. L. Dyce. 1962. Sandflies as possible vectors of disease in domesticated 
“curiae in Australia. Proc. Linn. Soc. N. S. Wales 87:364-376. 

Lewis, B. F. 1959. Abundance and seasonal distribution of the common species of Ceratopogoni- 
dae occurring in the State of Connecticut. Canad. Ent. 91:15-28. 

Macrig, J. W. S. 1944. Ceratopogonidae collected in Trinidad from cacao flowers. Bull. Ent. 
Res. 35:297-300. 

Mattocn, J. R. 1915a. Some additional records of Chironomidae for Illinois and notes on 

other Illinois Diptera. Bull. Ill. State Lab. Nat. Hist. 11:306-363. 

.1915b. The Chironomidae, or midges, of Illinois with particular reference to the 
species occurring in the Illinois River. Bull. Illinois State Lab. Nat. Hist. 10:290-361. 

. 1917. A preliminary classification of Diptera, exclusive Pupipara, based on larval and 
pupal characters, with keys to imagines in certain families. Part I. Bull. Ill. State Lab. 
Nat. Hist. 12:161-409. 

Matsupa, R. 1958. On the origin of the external genitalia of insects. Ann. Ent. Soc. Amer. 

Dili8 4-942 

Mayer, K. 1933. Zue Imaginalbiologie der Ceratopogonidae. Deutsch. Ent. Zeit. 1933:56-63. 

Mayr, E. 1963. Animal species and evolution. Belknap Press, Cambridge 795 p. 

MecaHep, M. M. 1956. A culture method for Culicoides nubeculosus Meigen in the 
laboratory, with notes on the biology. Bull. Ent. Res. 47:107-114. 

MessERSMITH, D. H. 1965. Culicoides associated with poultry in Virginia. Mosquito News 
25:321-325. 

Morray, W. S. 1957. Investigations on the bionomics of Culicoides obsoletus (Mg.) and 
other biting midges at Mount Solon, Virginia. Mosquito News 17:77-81. 

Nicuoias, W. L., anp W. E. KersHaw. 1954. Studies on the intake of microfilariae by their 
insect vectors, their survival and their effect on the survival of their vectors. III. The 
intake of the microfilariae of Acanthocheilonema perstans by Culicoides austent and 
C. grahamu. Ann. Trop. Med. and Parasitol. 48: 201-206. 

Ortiz, I. 1950. Estudios en Culicoides. IV. Revision de las especies americanas del sub- 

genero Hoffmania Fox 1948, con la descripcion de los nuevos especies. Rev. Sanid. 

Asist. Soc. 15:437-460. 

. 1951. Estudios sobre Culicoides. VII. Culicoides dominicii n. sp. y clave para el 
reconocimiento de las especies venezolanas. Novy. Cient. Mus. Hist. Nat. La Salle, Zool. 
SerNo. >; 12 5p 

Peterson, A. 1916. The head capsule and mouth-parts of Diptera. Ill. Biol. Monog. 3:177- 

2 


Posnetre, A. F. 1944. Pollination of cacao in Trinidad. Trop. Agr. 21:115-118. 

Price, D. A., ano W. T. Harpy. 1954. Isolation of the bluetongue virus from Texas sheep— 
Culicoides shown to be the vector. Jour. Amer. Vet. Med. Assoc. 124:255-261. 

Rees, D. M., anp H. R. Buttock. 1954. Culicoid gnats of Salt Lake County, Utah. Utah 
Acad. Sciences, Arts and Letters 31:42-43. 

Reve, E. J. 1964. The problems of biting midges (Dipt.: Ceratopogonidae) in Queensland. 
Jour. Ent. Soc. Queensland 3:1-6. 

Riek, R. F. 1954. Studies on allergic dermatitis (Queensland itch) of the horse: the aetiology 
of the disease. Aust. Jour. Agric. Res. 5:109-129. 

Rosinson, E. J., Jr. 1955. Observations on the epizootiology of filarial infections in two species 
of the avian family Corvidae. Jour. Parasitol. 41:209-214. 


1020 Tue UNIversiry SCIENCE BULLETIN 


. 1961a. Early development of an avian filarial worm in the gnat, Culicoides crepuscularis. 
Jour. Parasitol. 47:898. 

. 1961b. Incidence of microfilariae in some Ohio birds and data on the habits of a 
possible vector. Jour. Parasitol. 47:441-444. 

Roor, F. M., anp W. R. HorrMan. 1937. The North American species of Culicoides. Amer. 
Jour. Hyg. 25:150-176. 

Ross, H. H. 1962. A synthesis of evolutionary theory. Prentice-Hall, Englewood Cliffs 387 p. 

Row.Ley, W. A. 1965. The occurrence and bionomics of bloodsucking midges in the central 
Columbia Basin. Ph.D. Dissertation, Washington State University 92 pp. 

SHarp, N. A. D. 1928. Filaria perstans; its development in Culicoides austeni. Trans. Roy. Soc. 
Trop. Med. Hyg. 12:371-396. 

Snoperass, R. E. 1935. Principles of insect morphology. McGraw-Hill Co., New York 667 p. 

. 1957. A revised interpretation of the external reproductive organs of male insects, 
Smithsonian Misc. Coll. 135:1-60. 

. 1959. The anatomical life of the mosquito. Smithsonian Misc. Coll. 139:1-87. 

Snow, W. E. 1955. Feeding activities of some blood sucking Diptera with reference to vertical 
distribution in bottomland forest. Ann. Ent. Soc. Amer. 48:512-521. 

, E. Pickarp anv J. B. Moore. 1957. The Heleidae of the Tennessee River Basin. Jour. 
Tenn. Acad. Sci. 32:18-36. 

StewArD, J. S. 1933. Onchocerca cervicalis (Railliet and Henry 1910) and its development in 
Culicoides nubeculosus Mg. Rep. Inst. Anim. Path., Cambridge Univ. 3:272-284. 

STonE, A., C. W., SaBrosky, W. W. WirtTu, R. H. Foore anp J. R. Coutson. 1965. A Catalog 
of the Diptera of America North of Mexico, U.S. Dept. Agric., U.S. Govt. Printing Office, 
Washington, 1696 p. 

Toxunaca, M. 1937. Sandflies (Ceratopogonidae, Diptera) from Japan. Tenthredo 1:233-237. 

Travis, B. V. 1949. Studies of mosquito and other biting insect problems in Alaska. Jour. Econ. 
Ent. 42451-4577. 

VarGas, L. 1953. Beltranmyia n. subg. de Culicoides. Rev. Inst. Salub. Enf. Trop. 13:33-36. 

. 1960. The subgenera of Culicoides of the Americas. Rev. Biol. Trop. 8:35-47. 

WarMkE, H. E. 1951. Studies on pollination of Hevea brasiliensis in Puerto Rico. Science 

113:646-648. 

. 1952. Studies on natural pollination of Hevea brasiliensis in Brazil. Science 116: 
474-475. 

Wixuiams, R. W. 1951. Observations on the bionomics of Culicoides tristriatulus Hoffman with 
notes on C. alaskensis Wirth and other species at Valdez, Alaska, summer 1949. Ann. 
Ent. Soc, Amer. 443173-183. 

. 1955. Studies on the Culicoides of Baker Co., Georgia. I. Preliminary survey and 
observations. Ann. Ent. Soc. Amer. 48:30-35. 

WInNNERTZ, J. 1852. Beitrage zur Kenntnis der Gattung Ceratopogon. Linn. Ent. 6:1-80. 

WirtH, W. W. 1952. Heleidae of California. Univ. Calif. Pub. Ent. 9:95-266. 

. 1955. Three new species of Culicoides from Texas. Jour. Wash. Acad. Sci. 45:355-359. 

———. 1956. The heleid midges involved in the pollination of rubber trees in America. Proc. 
Ent. Soc. Wash. 58:241-250. 

——.. 1963. New synonyms in American Culicoides. Proc. Ent. Soc. Wash. 65:68. 

———. 1965. Family Ceratopogonidae. In A. Stone, et al., A Catalog of the Diptera of 
America North of Mexico, p. 121-142. U.S. Dept. Agric., U.S. Govt. Printing Office, 
Washington. 

——, AND F. S. Branton. 1959, Biting midges of the genus Culicoides from Panama (Dipt:: 
Heleidae). Smithsonian Misc. Coll. 109:237-482. 

——., Anp L. J. Borrimer. 1956. A population study of the Culicoides midges of the Ed- 
wards plateau region of Texas. Mosquito News 16:256-266. 

——,anp A. A. Huserr. 1959. Trithecoides, a new subgenus of Culicoides (Dipt.: Cera- 
topogonidae). Pacific Insects 1:1-39. 

———. 1960. Ceratopogonidae reared from cacti with a review of the copiosus group of 
Culicoides. Ann. Ent. Soc. Amer. 53:639-658. 

———. 1962. The species of Culicoides related to piliferus Root and Hoffman in eastern North 

America. Ann. Ent. Soc. Amer. 55:182-195. 

, AND R. H. Jones. 1957. The North American subspecies of Culicoides vartipennis. U.S. 

Dept. Agr. Tech. Bull. 1170. 35 pp. 


1021 


ERRATUM 
In the paper by F. J. Rohlf and R. R. Sokal, “Coefficients of correlation 


and distance in numerical taxonomy,” University of Kansas Science Bulletin, 
vol. XLV, No. 1, pp. 3 to 27, Figures 9 and 10 were inadvertently inter- 
changed. Thus Figure 9 is above the legend of Figure 10, while Figure 10 
appears above the legend for Figure 9. 


1022 Tue University Science BULLETIN 
Contents of Volume XLVI 
NuMBER 
1. An Ecological Study of the Vegetation of Big Salt Marsh, Stafford 
Cotinty, Kansas: 5:22 a tes Pe oe eo ee ee Re ein eee ee Erwin Ungar, 
2. Fossorial Adaptation in the Bank Swallow, Riparia riparia 
(Linnaeus) ho aes ee ee Se PNP Re BAL Abbott S. Gaunt, 
3. A Revision of Micronecta of Australia and Melanesia (Heteroptera- 
Corixiclae)) arrest eee er eee se ee Ree Sere Se Ling-chu Chen, 
4. Microgeographic Variation and Covariation in Pemphigus 
POPUL GRSUCLS US Some wl SER Ree Rn tense LS Richard C. Rinkel, 
5. Geographic Variation of Pemphigus populi-transversus in Eastern North 
IAT EHI CAN Nett. PAE ee ee Pe Robert R. Sokal and Paul A. Thomas, 
6. New Asiatic and African Caecilians with Redescription of Certain Other 
SPECIES ee tie eect ett see eee na eA SS ke Edward H. Taylor, 
7. Catalogue of the Types in the Snow Entomological Museum. Part IV 
(Orthoptera); 22 ee George W. Byers and Lanny B. Carney, 
8. Catalogue of the Types in the Snow Entomological Museum. Part V 
(CAtcarinal) fis = Sos See ek ee es George W. Byers and Calvin L. Wong, 
9. The Life Cycle and Social Organization of Bees of the Genus Exoneura and Their 
Paras itenml it, q ttle Gg eee ee Ree re ee ROE See ae es as Charles D. Michener, 
10. The Life Cycle and Behavior of the Primitively Social Bee, Lasioglossum 
CDR Y TU TIA el alictiGac) meee aoe cee eee ee Suzanne W. T. Batra, 
11. A Report on the Occurrence and Distribution of Cladocera and Copepoda 
in Lewis and Clark Lake, South 
Wakotagee ee a wees. Jerry C. Tash, George A. Swanson, and Richard E. Seifert, 
12. Mosses of the Great Plains and Arkansas River Lowlands of 
Kansas BON eA 8 EE SE NR RIE Te UE ree Harold L. Smith, 
13. Some Intersectional Hybrids and Relationships in 
Fiaplopap pus <0 Seer. ae 28+ Pe es 03 8 Ne 2 be R. C. Jackson, 
14. Embryology of Haplopappus gracilis (Nutt.) Gray —....-..-------------.-- S. Venugopalan, 
15. Biorbia (Boraginaceae) in the Central U. S. Pliocence .....................--- Ronald H. Segal, 
16. Systematics of the genus Augochlorella (Hymenoptera, Halictidae) North 
OL MExi cog saree ee ees ees Seman He ne rere ok ee Ellen Ordway, 
17. The Binomics of Tenuipalpoides dorychaeta Pritchard and Baker (1955) 
(Acarina,, Trombidiformes; dletranychidac) meee a ee George Singer, 
18. A Revision of the Genus Exema of America, North of Mexico (Chrysomelidae, 
Coleoptera) pests. ae ees. Re ee Jay B. Karren, 
19. Pupation Site Differences in Drosophila melanogaster _..........--....... Robert R. Sokal, 
20. The Classification of the Diphaglassinae and North American Species of the Genus 
Caupolicana (Hymenoptera, Colletidae) -...............20..22.220200------ Charles D. Michener, 
21. A Revision of the Bee Genus Calliopsis and the Biology and Ecology of 
C. andreniformis (Hymenoptera, Andrenidae) ~............2.-20.-2.--22------- Alvin F. Shinn, 
22. The Culicoides of New Mexico (Diptera, Ceratopogonidae) ...... William R. Atchley, 
Erratuin.. 2:4 eect oe na ore ar oceectadesececsnecce=e: 


PAGE 


167 


201 


253 


303 


309 


317 


359 


425 


433 


475 
487 
495 


509 


625 


647 


HO! Library - Serials 


5 WHSE 0 


————, a : 
AED en cin Oe ee le 
aay ee a i 
= ee a eS ee 
ee 


enero ne 


we 
-_» na - 

ny rer tomer 

eww ——— 

- aon ee 


—— 5 
aot al 
el el 

Ae sec 3 O 


ahaha 


4 


Aggie 
PARLE A)