aa

HARVARD UNIVERSITY

et]

ASE

(f=
Jia

LIBRARY

OF THE

Museum of Comparative Zoology

t
A
‘
af

OL Us
py H fs \" }
ree { umal
all li A

Wa, at,
Asie fy afr

whitl

f.

el ‘f | . evi fon ' js.
‘ Ls i eG : Mls {. ‘
1, | ih ay ig
RS fu

1 av ag i o ae
: in ai:

bf Vie ; “i! ie " ¥ Uf. } } F .
‘ ‘ | icy it if rae N) as? WY Ba, he A
4 4 ata : F Hi i ts ah i! ee ah ‘ - f : ln iA
herd Mitts PaUR Gee ] heey
aa )

‘

Ay hi aie

2S
_

VARIATION AND SPECIATION

IN THE GRASSHOPPERS OF THE CONALCAEINI
(ORTHOPTERA: ACRIDIDAE: MELANOPLINAE):
THE LOWLAND FORMS OF WESTERN MEXICO,
THE GENUS BARYTETTIX

THEODORE J. COHN
DEPARTMENT OF ZOOLOGY
CALIFORNIA STATE UNIVERSITY, SAN DIEGO

AND

IRVING J. CANTRALL
MUSEUM OF ZOOLOGY
UNIVERSITY OF MICHIGAN

SAN DIEGO
SOCIETY OF NATURAL HISTORY

MEMOIR 6

1974

Frontispiece.

SAN DIEGO SOCIETY OF NATURAL HISTORY MEMOIRS
Memoir 6, pages 1-131
Issued February 26, 1974

Laboratory reared first generation Barytettix from field collected parents, all from Sinaloa, México.
Upper left. Barytettix poecilus (Hebard), adult male (length of body about 24 mm), 20 mi NW Mazatlan Cathedral.

Upper right. Barytettix nigrofasciatus Cohn and Cantrall , third instar nymph (length of body about 17 mm), 1.1 mi SW
San Ignacio Ferry.

Middle and lower left. Barytettix
Los Mochis turnoff.

Middle and lower right. Barytettix psolus Cohn and Cantrall ,
Culiacan Cathedral.

tridens Cohn and Cantrall , adult male (length of body about 27 mm), 16.4 mi N

adult female (length of body about 38 mm), 14 mi NW

PUBLISHED WITH FINANCIAL AID

FROM THE

W. W. WHITNEY PUBLICATIONS ENDOWMENT

AND BY DONATIONS FROM
DR. AND MRS. CARL L. HUBBS
THE AMERICAN FOUNDATION FOR OCEANOGRAPHY
SAN DIEGO STATE UNIVERSITY FOUNDATION

THE UNIVERSITY OF MICHIGAN MUSEUM OF ZOOLOGY

CONTENTS

INTRODUCTION

Acknowledgements .
Materials
Methods

THE MALE CONCEALED COPULATORY STRUCTURES

THE FEMALE INTERNAL COPULATORY ELEMENTS

Terminology
The Receptaculum Semin
The Taxonomic Value

THE SYSTEMATIC POSITION O OF THE CONALCAEA COMPLEX

CONALCAEINI NEW TRIBE .-

GENUS BARYTETTLX SCUDDER .

Diagnosis P
Barytettix and other Mclanopline Generali in Weer! Mexien
Generic Description .

THE SPECIES GROUPS OF BAR YTETTIX AND TABLES FOR THE IDENTIFICATION

OF SPECIES
THE CRASSUS GROUP
Barytettix poecilus (Hebard) .
GEOGRAPHIC VARIATION . :
HABITAT AND ASSOCIATED SPECIES
Barytettix crassus Scudder
Barytettix terminalis n. sp.
THE PSOLUS GROUP .
Barytettix contilus n. sp.
TAXONOMIC STATUS
Barytettix contilus contilus n. ssp:
Barytettix contilus tectatus n. ssp.
Barytettix contilus hiscatus n. ssp.
Barytettix contilus dicranatus n. ssp.
Barytettix contilus similis n. ssp.
Barytettix psolus n. sp.
GEOGRAPHIC VARIATION 4 :
HABITAT, ASSOCIATED SPECIES AND SEASONAL OCCURRENCE
Barytettix nigrofasciatus n. sp.
Barytettix paloviridis n. sp.
GEOGRAPHIC VARIATION .
The Guaymas Populations and Dwarfism .
The Guamuchil Populations and Primitive Aedeagal Characters
Leg Color and Interaction with poecilus .
Aedgal Character Reinforcement and Mechanical Reproductive Isolation
The Barrier Effect of Rivers
HABITAT AND ASSOCIATED SPECIES
DISTRIBUTION ae
THE HUMPHREYSII GROUP
Barytettix humphreysii (Thomas)
GEOGRAPHIC VARIATION AND SUBSPECIATION-
Barytettix humphreysii humphreysii (Thomas) .

GEOGRAPHIC VARIATION, BIOLOGICAL SIGNIFICANCE AND TAXONOMIC TREATMENT

HABITAT AND ASSOCIATED SPECIES
SEASONAL OCCURRENCE . dao SaWae
Barytettix humphreysii cochisei Gurney n. comb. .
GEOGRAPHIC VARIATION) 2
RELATIONSHIP BETWEEN cochisei and humphreysii
Barytettix tridens n. sp.

RELATIONSHIPS . . . .
GEOGRAPHIC VARIATION |.

SPECULATIONS ON THE PHYLOGENY OF THE SPECIES OF BARYTETTIX .

MECHANICAL ISOLATION :
INTERGROUP MECHANICAL ISOLATION in Bin tettix .
INTRAGROUP MECHANICAL ISOLATION in Barytettix .

The Humphreysii Group

The Crassus Group

The Psolus Group
SUMMARY . .

PROMISING PROBLEMS
Mechanical Isolation
Natural Hybridization
Stasipatric Speciation
Competitive Exclusion, Repieduceue Exclusion, and Pabitat Preference
Geographic Variation, Gene Flow and Selection F

THE DISTRIBUTION OF THE SPECIES OF BAR YTE TTIX

LITERATURE CITED

89
93
94
96
96
97
97
98
98
99
100
101
102
102
103
109

INTRODUCTION

Members of the genus Barvfettix are among
the commonest and most conspicuous grasshop-
pers of the northwestern coastal plain of mainland
México. Between the coast and the Sierra Madre
Occidental and from Arizona into the Trans-
Volcanic Belt, they occur in many natural habi-
tats as well as in most of the weedy environments
created by man. Although short-winged, they
are active and are strong jumpers. In view of this
behavior and apparently broad habitat tolerance,
one might expect extensive distributions and
little speciation or geographic variation in the
genus. The reverse is true. Eight species and at
least seven subspecies are found between Arizona
and Colima, and a ninth species occurs as an iso-
late in Baja California. Only three species have
broad distributions. These, together with the
other members of the genus, show complex pat-
terns of overlap, contiguity, and allopatry. Al-
most all display some degree of geographic vari-
ation, and in several forms this variation is
striking.

A genus such as Barytettix is ideal for the
study of a number of evolutionary problems.
Further, these grasshoppers are easy to observe
and collect in the field, and most, if not all, of
the species are easy to breed and to rear in the
laboratory. Obvious problems which might be
studied include the nature, development, and
maintenance of geographic variants, the nature
and extent of gene flow between populations and
between demes of populations, competitive and
reproductive exclusion, and the possibility of
Stasipatric speciation. Because of the variety of
patterns of allopatry and sympatry among closely
related species, the genus offers exceptionally
good material for the study of reproductive iso-
lation in animals which lack obvious precopula-
tory isolating mechanisms. Baryfettix has pro-
vided new evidence for the role of mechanical
isolation in reproduction (Cantrall and Cohn,
1972) and at present is being further studied
in laboratory crosses in an attempt to learn
the effectiveness of this mechanism in these
grasshoppers.

Our study is incomplete in many respects, but
since further refinement of the systematics and
biology of these insects will require much more

time and more intensive effort, we feel that we
have reached the point where it is important to
make known the basic taxonomic situation of
this group of common but largely unknown
grasshoppers. It is perhaps even more important
that we now outline those biological situations
found in the genus which appear to have a signif-
icant bearing on basic biological problems and
which new material and new approaches might
clarify.

In the near future we hope to complete our
studies on the tribe Conalcaeini described herein
and to publish descriptions of two new genera
and a review of Conalcaea. These genera along
with Barvtettix comprise the tribe as we now
understand it.

In 1951 Gurney published an account of the
Conalcaea Complex in which he brought together
widely scattered relevant information, reported
on type examination and identification, and
most ably summarized the existing knowledge of
Barytettix. His revision is the foundation upon
which we have based our study. Previous litera-
ture consisted largely of single descriptions, rec-
ords of distribution, and similar details.

ACKNOWLEDGEMENTS

A research project cannot be carried to frui-
tion without the aid of many people. Ours is no
exception.

N. D. Jago of the Centre for Overseas Pest
Research, London, England, loaned us specimens
from the Academy of Natural Sciences of Phila-
delphia while he was responsible for the orthop-
teroid collections of that institution. Ashley B.
Gurney prepared a drawing of the type of
Barytettix cochisei (Fig. 10D) for our use and
sent us material of Barytettix from the United
States National Museum. Edwin F. Cook, Uni-
versity of Minnesota, presented us with a speci-
men each of a male and a female topotype of
Barytettix crassus. Simon and Guadalupe Stopol
have most generously given of their time, and of
the hospitality of their home to provide Cohn
with a “home away from home” during his field
work in the vicinity of Culiacan, and Simon, in
particular, in giving unstintingly of his broad

10

knowledge of the land, the people, and the cus-
toms of México, has materially eased the way for
effective field work in a foreign country. George
E. Radwin, San Diego Museum of Natural His-
tory, and Vincent Roth, American Museum of
Natural History Southwestern Research Station,
have aided our cause through the gift of speci-
mens. Henry van der Schalie, The University of
Michigan Museum of Zoology, generously had
cross-sections of selected bursae prepared in his
microtechnique laboratories. James Mathewson
of the Chemistry Department, California State
University, San Diego, arranged for the use of
that department’s lyophilizer and gave gener-
ously of his time and expertise in the freeze-
drying of the 1970 collections.

Martha B. Lackey and John Tottenham have
borne up remarkably under our demands for the
many illustrations which they have prepared for
us. Mrs. Lackey prepared all of the plates and
the major portion of the illustrations. John
Tottenham rendered Fig. 7 A and B, Fig. 9 A-F,
and Fig. 13 A-F. The photographs for the
Frontispiece were taken by Dallas Clites of San
Diego. Ruth Fausett has painstakingly and with
much patience arranged, typed and retyped the
many revisions of the tables in the interest of
accuracy and clarity and sometimes according
to the whims of authors and editor. Final copy
of the tables was produced on an IBM Composer
typewriter at California State University, San
Diego. Mary Snider and Dorothy B. Hodgman of
Ann Arbor, Michigan, struggled with our much
amended, interlined, corrected, and often scrib-
bled early drafts and successfully transcribed
them to produce readable copy. Gwen Gardner
of the Word Processing Center, California State
University, San Diego, did the final typing of the
manuscript.

The 1970-71 rearing project in Tucson was
conducted in quarters supplied by the Cotton
Insects Biological Control Investigations of the
United States Department of Agriculture. D. I.
Bryan, R. E. Fye, and G. D. Butler of that organi-
zation, and W. E. Nutting of the University of
Arizona generously helped with many practical
problems. Richard V. Carr, then at the Univer-
sity of Arizona, helped set up the project and
assumed responsibility for it during the long
periods when Cohn was in México or Michigan.
It was largely through his efforts, later continued
by Steven Moore and Richard Sackett, that this
phase was so successful. In Culiacan, the grass-

T. J. COHN AND I. J. CANTRALL

hoppers were reared with care and affection by
Srta. Pauline Stopol aided by Dora Stopol, in
quarters generously and with much tolerance
supplied by Simon and Guadalupe Stopol.
Throughout the 1970-71 season, Jean W. Cohn
aided in all phases of the work, constructed all
the cages and solved many practical problems
when all else failed. Her moral support kept the
work going at high pitch, and her companionship
and physical support enabled the field and rear-
ing work to advance under the most difficult
circumstances.

Financial support from the Sigma Xi Research
Fund and from the American Philosophical
Society facilitated Cohn’s study of Neobarrettia
in 1958 and 1959 during which he also collected
Barytettix, and funds from National Science
Foundation Grant No. G 14012 to T. H. Hubbell,
Emeritus Director of The University of Michigan
Museum of Zoology, enabled us in 1961 to col-
lect along a transect of the Sierra Madre Occi-
dental as well as at various localities in Sinaloa,
Nayarit and central Jalisco.

To each of these individuals, and to those
responsible for the granting of financial aid, we
wish to extend our thanks, and to indicate grate-
ful appreciation for the many gestures which, in
large part, made this study possible.

MATERIALS

Our initial interest in the genus Barytettix was
stimulated by the incidental collection of mem-
bers of the distinctive Psolus Group. These were
taken by Cohn during the fall periods of 1958
and 1959 while he was doing field work on the
tettigoniid genus Neobarrettia. Since that time,
and as our studies of Barytettix developed, addi-
tional field work has seemed desirable in order
to work out some of the many details of distri-
bution and phyletic relationships which became
apparent. Cohn, once accompanied by Cantrall,
returned specifically to study Baryvtettix for
periods of from one to four weeks in five sepa-
rate years. Most work was done along the main
west coast Mexican Highway No. 15 during the
summers when rains make dirt side roads almost
impassable. A transect was made along the
Durango-Mazatlan Highway, and collections have
been made in the foothills farther north near
Cosala and Alamos. Terrain along roads leading
towards the base of the Sierra Madre Occidental
near San Ignacio, Tepuche, and Tezopaco has

CONALCAEINE GRASSHOPPERS: BARYTETTIX

been investigated, and, in the summer of 1970
and at the behest of Cohn, James Melli and John
Stallone worked along roads leading to the base
of the mountains near Badiraguato, as well as
east of Sanalona, and well into the mountains
on a road running east from Alamos.

A desire to further clarify population inter-
relationships resulted in Cohn returning to
México during the summer and fall of 1970, dur-
ing which time he carried out an extensive pro-
gram of collecting along Highway No. 15. Much
living material was returned to Culiacan, or to
Tucson, Arizona, where, with the exception of
crassus, nymphs of all known taxa of Barvfettix
were reared to maturity, and many infra- and
interspecific crosses were attempted. The clari-
fication of certain aspects of mechanical repro-
ductive isolation in Barytettix made observations
on the mating behavior of caged individuals
desirable, and, as a result, a number of pairs in
copula, both intra- and interspecific, were frozen
and then preserved in alcohol. Although some
of the results of Cohn’s 1970 activities are in-
cluded in this paper, results from crosses, obser-
vations, and study of dissections of copulating
pairs will be reported at a later date.

The bulk of the large collection of Barytettix
amassed by the above efforts is deposited in The
University of Michigan Museum of Zoology
Collections. Specimens discussed or listed in the
present study without an indication of ownership
are a part of this material. Representative speci-
mens from this assemblage, when available, will
be deposited in the major world Orthoptero-
logical Centers, and in the collections of the
Instituto de Biologia de la Universidad Nacional
de México, designated by the authorities of the
Mexican Government as a depository for samples
of specimens collected under permit (in Jitt., Dr.
Carlos Marquez Mayaudon, 1971).

During the course of our studies, it has been
necessary to examine a number of specimens
from the holdings of both the Academy of Nat-
ural Sciences of Philadelphia (ANSP), and of the
United States National Museum (USNM).

METHODS

MALE GENITALIA.— A comprehensive analysis
of the components of Barytettix would not have
been possible without a thorough study of the
male concealed genitalia and of the female com-
plementary structures. Museum specimens must

be softened before these organs can be moved or
dissected without destruction. Depending upon
the time element several methods have been em-
ployed. A few minutes in hot water, a little
longer in ammonia, perhaps twenty minutes sub-
merged in cold water, or sandwiched overnight
between several layers of water-saturated paper
hand toweling creates the desired amount of
relaxation.

After relaxation of a male, the specimen is
positioned! , and, using insect pins with the tip
bent at a right angle, the pallium is slipped back
and the subgenital plate lowered with one rotat-
ing motion. One bent insect pin 1s inserted into
the genital cavity anterior to the genital mass in
such a way that the tip of the pin is pointing
ventrally; a second bent pin is inserted behind
the mass and with the tip also pointing ventrally.
The tips of the pins are rotated upward and
toward each other simultaneously. These move-
ments roll the genitalia upward and out of the
genital cavity, and into position for examination.
Since most studies can be made from such prep-
arations the extruded genital mass was usually
left in place to dry. Occasionally it is desirable
to make a more thorough and detailed examina-
tion. In such cases, the membranes and muscles
holding the genital mass in place are cut and the
mass transferred to a macerating agent.

We have found the acid corrosive solution”
suggested by Mitchell and Cook (1952) most
useful for maceration. This fluid offers several
advantages in that the genital structures are not
discolored or distorted, heat is not necessary,
staining can be accomplished with ease, and the
genital mass can be left in the fluid for any
length of time without loss of fine structure.
Disadvantages are that the fluid does not seem
to work well on newly killed animals, nor on
those preserved in alcohol or formalin. By far
the best results are obtained with material which
has been dried and then relaxed. Tissue destruc-
tion requires five or more hours, preferably over-
night. Now and then, it is desirable or necessary
to place new solution on a specimen before the

IFor this purpose a sheet of pinning surface large enough
to cover the microscope stage is used. To this is glued a cork
stopper. The stopper should be wide enough to support the
grasshopper abdomen when the grasshopper is pinned in place
over the stopper. Lateral movement of the abdomen is pre-
vented by bracing with a pair of insect pins.

2Formula: 9 gm chloral hydrate, 6 cc glacial acetic acid,
9 ce distilled water. The fluid can be prepared in quantity
and keeps indefinitely. It is best not to reuse the solution.

12

desired results are obtained. The process can be
accelerated by stirring the solution from time to
time. As is the case with potassium hydroxide,
the membranous and sclerotized genital parts
must be brushed free of the softened tissue after
maceration. Membranes and feebly sclerotized
parts are quite transparent and are better seen
if the cleared genital mass is transferred to acid
fuchsin and stained for from five to ten seconds,
or up to a minute, depending on the size and
thickness of the mass and its membranes. The
mass is then washed with 70% ethanol, drained,
and stored in glycerin.

FEMALE GENITALIA.— A preparation of the
female copulatory structures can be made by
first cutting the inter-segmental membrane be-
tween the 8th and 9th abdominal segments of a
relaxed female. The tip of the abdomen is then
pulled caudally until the viscera can be cut ante-
rior to the spermatheca. The tergal sclerites
and the supra-anal plate are removed from the
part of the body which has been dissected away,
and are oriented and reattached to the top of the
female’s body by means of clear lacquer. The
remainder of the dissection is placed in acid cor-
rosive overnight. The dorsal ovipositor valves
are then removed, and a cut along the anterior
margin of the basivalvular sclerites is made to
separate the subgenital plate and ventral portions
of the genital cavity from the ventral ovipositor
valves and the dorsum of the genital cavity and
attached receptaculum seminis. The structures
are stained in acid fuchsin. All parts permanently
removed from the pinned specimen are stored in
glycerine.

FIELD PROCEDURES.— One of the more per-
plexing problems of the orthopterist is the eval-
uation of the significance of coloration as a
taxonomic tool. This is not helped by preserva-
tion which, more often than not, is so poor that
difficulty in color interpretation is experienced.
We have for many years attempted to improve
on this situation by experimenting with a num-
ber of techniques which might yield better study
material. The most practical method has turned
out to be the layering of freshly killed specimens
(often gutted) between sheets of cellucotton in a
cigar box, or by arranging the specimens on a
sheet of cotton which in turn is wrapped in paper
to produce a small flat packet. The packets are
placed in cigar boxes or other suitable contain-
ers. The containers, whether holding packets or

T. J. COHN AND I. J. CANTRALL

layered specimens, are tied under the hood of
an automobile and exposed to heat of the engine
as the vehicle is operated. Circulation of warm
air is essential to this method and to that end
the tops and bottoms of the containers have been
fitted with screen covers.

In 1970 Cohn vastly improved on the layer-
ing technique by placing freshly collected speci-
mens in a portable “‘Norcold”’ freezer operated
by power from the car battery. Such material
eventually reached the laboratory where it was
thawed, layered in cellucotton as described
above, refrozen, and then transferred to a large
lyophilizer operating at room temperature and
with a vacuum of a few tenths of a millimeter of
mercury. Complete dessication of five cigar
boxes at a time, each containing from 100 to
200 specimens, was accomplished in from three
to four days. Color preservation appears to have
been excellent, although there may have been
some slight fading of the greens. The only prob-
lem we have encountered to date with this
method is a slight difficulty in lifting up or
moving the male concealed genitalia. This ap-
pears to result from the excellent preservation of
the muscles which are rather elastic when re-
laxed and thus tend to draw the genitalia back
into the abdominal cavity. Application of
household ammonia to the parts facilitates the
dissection.

COLORATION.— Individuals of the genus Bary-
tettix, as they rest in the vegetation, are striking
insects. The colors are bright and the blotches
and stripes of the color patterns are sharply
defined. Red, orange, brown, yellow, gray,
olive-green, green, blue, and purple appear as
distinctive colors. However, upon examination
of series of specimens we have found that there
is considerable variation in the tone of a given
color, both within a population and between
populations. Tibial color may be polymorphic
in a given deme, and we have reason to believe
that large samples might well show continuous
variation from one extreme to another. For
these reasons we have deliberately avoided at-
tempting to indicate hue or shade. Our termi-
nology may thus seem rather broad and arbi-
trary, but there is functionally little or no
difficulty in recognizing the color indicated. We
have used the term bicolored with reference to
hind tibial color when the proximal and distal
ends of the tibia are of distinguishably different
colors. In all such cases, the transition between

CONALCAEFINE GRASSHOPPERS: BARYTETTIX

the two colors is gradual; in no case did we find
them sharply separated.

KEYS AND TABLES.— Although dichotomous
keys are probably necessary for the separation
of large numbers of taxa, and, in abbreviated
form, may have definite value in facilitating rapid
identification, they often have serious limita-
tions. From a strictly practical viewpoint, they
may be unreliable. An error in interpretation of
a single rubic, or the absence of a major character
in a couplet, or poor preservation of a single key
character in a given specimen may prevent iden-
tification. An undescribed species cannot be
classified with assurance, and may require a
major revision of the key in order to include it.
Most of these deficiencies are eliminated in a
table. As pointed out by Newell (1970), a table
includes each character for each taxon, and thus
can show similarities as well as differences. A
detailed and well-constructed table can be open-
ended, permitting the insertion of new species
with little or no change. Finally, and most im-
portant, a table lends itself to computerization
and, if complete, greatly facilitates data storage
and retrieval. For these reasons we have pre-
sented our keys in tabular form.

MEASUREMENTS.— All measurements are pre-
sented as ranges, or in bar histograms. Since
many of the specimens we have examined were
not collected randomly, and since all species of
Barytettix are readily recognizable on the basis
of morphological features, we feel that a detailed
statistical analysis of mensural characters would
be inappropriate at the present time.

Measurements were taken in the following
ways:

Body Length.— From the anterior margin of the
fastigium of the vertex to the distal edge of

13

the genicular lobe of the caudal femur with
the femur aligned parallel to the axis of the
body.

Length of Pronotum.— Taken along the median
carina.

Length of Metazona.— Taken along the median
carina.

Length of Caudal Femur.— From the anterior
edge of the upper basal lobe to the distal edge
of the genicular lobe.

Length of Cephalic Femur.— The length of a line
along the posterior surface of the appendage
from the distal edge of the genicular lobe to
the dorsal junction with the trochanter.

Length of Male Supra-Anal Plate.— Maximum
possible length.

Length of Tegmina.— Length of a line running
from the tip to the junction of the ventral
(costal) margin of the tegmina with the caudal
margin of the pronotum. Owing to the posi-
tioning of the pronotum as a result of pinning,
a certain amount of inaccuracy exists in this
measurement.

Width of Caudal Femur.— Maximum possible
measurement.

Width of Tegmina.— Maximum possible measure-
ment.

Width of Male Supra-Anal Plate.— Maximum pos-
sible measurement.

Aedeagal Measurements.— These are indicated
and discussed under the section dealing with
the terminology of the male genitalia (page 14 ).

DRAWINGS.— All stippled illustrations of aed-
eagal structures have been drawn from cleared
material, or from phallic masses which have been
moistened with household ammonia until all
membranes have swollen to the condition found
in fresh specimens.

THE MALE CONCEALED COPULATORY
STRUCTURES

During the past three decades much progress
has been made in the evaluation of the taxonomic
and biological significance of the concealed geni-
talia of male Acridoidea. A steady increase in
the use of the diversity in these structures as
differentiating characters has resulted in the
publication of several efforts to clarify termi-
nology in the hope that a better understanding
of homology and relationships might be obtained.

The approach was begun by Roberts (1941) and
was improved upon and greatly expanded by
Dirsh (1956). Both of these projects have been
reevaluated recently by Kevan, Akbar, and Chang
(1969) who not only ably summarized the work
of Roberts and Dirsh, but attempted a further
refinement by blending in the results of an ex-
tensive literature survey.

The concealed genitalia of male Barytettix

can be categorized into three groups based on the
size and shape of the aedeagal parts. Figure |
shows the relationships of these aedeagal ele-
ments, one to another within a given group, and
indicates homologies between the groups. Al-
though our nomenclature largely follows that
set forth by Kevan, et al (1969), we have devi-
ated to some extent in the interests of simplicity
and clarity, or to indicate relationships more or
less peculiar to Baryvrettix. Further, in the course
of this study, we have had occasion to measure
certain of the aedeagal elements. A number of
terms, aedeagal measurements, and intrageneric
relationships of these parts are presented and
discussed or defined below.

The epiphallus shows little interspecific varia-
tion in Barytettix. In structure it is typically
melanoploid in that a well-defined and undivided
bridge is present, as are lophi and ancorae. Oval
sclerites are lacking.

The dorsal aedeagal valves and sclerites become
variously involved with the sheath (SH in Fig. 1 A
and B) to such an extent that it is often difficult
to recognize the point of contact between the
elements. In the interests of simplicity we are
referring to these fused structures as the dorsal
valves (DV in Fig. 1). In Barytettix the dorsal
valves are fused proximally. The distal tips, here-
in referred to as the free lobes of the dorsal
valves, may each have an essentially straight
medial margin, or the margin may be curved,
feebly sinuous, notched, or the notch may be
deep enough to produce an aciculate process (see
Figs. 10; 14 A-F).

The sheath is well-developed in this genus of
grasshoppers. Not only does it contribute much
to the size and shape of the dorsal valves, but it
extends around the lateral and ventral portions
of the base of the aedeagus. There it forms a
collar (SH in Fig. | A and B) or it may be en-
larged and elongated to produce a flattened
(VLSH in Fig. | E and F) or a finger-like protru-
sion (VLSH in Fig. | Cand D). We are referring
to the latter two developments of the sheath as
the ventral lobes of the sheath (VLSH). In the
two species groups where the ventral develop-
ment of the sheath is most pronounced, each
protrusion is reinforced by a sclerotized plate,
the process of the ramus of the cingulum (PRC
in Fig. | C-E). In the Psolus Group this process
can be observed readily even in dried specimens.
The homologue in dry material of the Hum-
phreysii Group is obscured by the more heavily

T. J. COHN AND I. J. CANTRALL

sclerotized lobes of the sheath, but is easily seen

in cleared genitalic preparations.

The ventral lobes of the ectophallic membrane
are well-developed. Each lobe bears a distinctive,
rather large sclerotized plate (SVL in Fig. 1).
These plates are elongate obovate in shape, are
usually decidedly granular papillose, and lie on
the under side of the genital mass just below and
lateral to the ventral cleft of the phallotreme.

The ramus of the cingulum is broad in the
Crassus Group (RC in Fig. | A and B). In the
Psolus and Humphreysii Groups, it is narrower
and bears a process (PRC in Fig. | C-F) which
functions as a support for the ventral develop-
ment of the sheath (VLSH in Fig. 1 C-F). In all
groups, the ectophallic membrane extends be-
yond the ends of each ramus to form a ventral
lobe of the ramus of the cingulum (VLRC in
Fig. 1). Each of these lobes bears a more or less
strengthened disc, bordered by a membranous
margin. The margin folds back to become con-
tinuous with the endophallic membrane of the
phallotreme to produce a ventral cleft (VC in
Fig. 1).

In order to interpret the significance of the
variation of aedeagal parts, a number of measure-
ments were made of the dorsal and ventral valves,
of certain relationships between these valves, and
of their relation to the point of origin of the
ventral lobes of the sheath. In the Psolus and
Humphreysii Groups which have well-developed
ventral lobes of the sheath, a distinct angle or
notch occurs on the lateral side of the aedeagus
at the base of the visible portion of the ventral
valve. This notch (point X in Fig. | C and E) has
been used as a reference point. In the Crassus
Group the notch is absent and we have used for
the purposes of measurement a morphologically
equivalent point (X in Fig. 1 A).

Measurements taken were:

Length of Aedeagus.— The length of that portion
of the ventral valve not covered by the sheath,
the distance from X to the distal tip of the
valve, taken in lateral or dorso-lateral view.

Width of Aedeagus.— Ratios of the length to the
width of the aedeagus were made for each of
the three groups of species. Owing to the
different shapes of the internal genitalia, it
was necessary to select different reference
points for aedeagal width in the groups, as
follows: Psolus Group, maximum dorsal
width of the dorsal valves; Crassus Group,
maximum dorsal width of the dorsal valves;

CONALCAEINE GRASSHOPPERS: BARYTETTIX

Humphreysiti Group, width of the ventral
lobes of the sheath, taken in ventral view.
Length of Exposed Portion of Ventral Valves.—
The distance from the distal tip of the dorsal
valves to the distal tip of the ventral valves,

taken in dorsal view.

Length of Free Lobes of Dorsal Valves.— The
distance from the distal terminus of the fusion
of the dorsal valves to the tips of the valves,
taken in dorsal view.

POECILUS B

PALOVIRIDIS

H. HUMPHREYS 11

Figure 1. Semi-diagramatic drawings of the three types of
concealed male genitalia found in the species groups of Barytettix.
Scale below each figure equals 1 mm. A, C, E. Lateral views.
B, D, F. Ventral views.

ANB: poecilus. México: Sinaloa, El Venadillo
Cap: paloviridis. México: Sinaloa, 39.9 mi SE Culiacan
ke humphreysii humphreysii. México: Sonora, 17.7 mi
N Imuris
Symbols used in figures:
BF — basal fold
DV — dorsal valve
PRC — process of ramus of cingulum
RC — ramus of cingulum
SH — sheath
SVL — sclerite of ventral lobe of ectophallic
membrane
vc — ventral cleft
VLRC — ventral lobe of ramus of cingulum
VLSH — ventral lobe of sheath
VV — ventral valve
x — the notch

THE FEMALE INTERNAL COPULATORY
ELEMENTS

TERMINOLOGY.— Kevan, Akbar, and Chang
(1969) presented a lengthy and useful summary
of the literature on concealed copulatory struc-
tures in acridoid insects and, in an attempt to
homologize the various copulatory elements in
both sexes, have proposed a number of changes
in terminology. We follow the terminology for
the female elements essentially as proposed by
Gregory (1965) [see Figures 2 and 16], because
his study dealt with a single species of the Acridi-
dae and, in a sense, his terminology is therefore
more relevant to our work. Further, we are at

variance with Kevan, Akbar, and Chang with re-
gard to the terms “‘proximal”’ and ‘“‘distal,” pre-
ferring to consider “‘proximal” as pertaining to
the point of origin or base of the receptaculum
seminis, rather than from the viewpoint of an
anterior-posterior relation. Since the receptacu-
lum seminis is essentially a diverticulum we con-
sider the spermatheca to be distal. Thus, several
terms proposed by Kevan, Akbar, and Chang are
not appropriate to our use. We agree with these
authors, though, in that less confusion will exist
if the term “‘spermatheca’’ is limited in use. We

16

use their proposed term ‘“‘receptaculum seminis”’
for the entire organ generally referred to as the
“spermatheca.”

In the receptaculum seminis of females of
Barytettix there is a variously elongated enlarge-
ment (BC in Figs. 2 and 16) of the spermathecal
duct between the aperture of the duct and the
orifice of the thick tube’. The aedeagus is fitted
into this chamber during copulation. This being
true, we refer to this enlarged portion of the
receptaculum seminis as the “bursa copulatrix,”’
or, sometimes, simply as the “‘bursa.”’

THE RECEPTACULUM SEMINIS.— Our dissec-
tions were based upon females selected from
localities which we considered critical, and, even
though we have not been able to determine the
limits of variation, we have little reason to be-
lieve that the results of our studies will be greatly
modified by an examination of a larger number
of specimens. This conclusion is based upon our
experiences with intrademe variation in the acri-
doids and upon the findings of Slifer and Ran-
dell. Slifer (1940a:3), investigating the distal
end of the receptaculum seminis, and Randell
(1963:248), working with the female copulatory
armature, both found the intra-specific range of
variation to be minor in the acridids which they
studied.

Other than the comments of Slifer (1940b:
209), who figured and described the condition
found in several species of acridids, there is little
comparative information in the literature regard-
ing the functional shape and structure of the
proximal portions of the spermathecal duct.
Berlese (1882) and Fedorov (1927) noted dila-
tion in this portion in some species. Slifer
(1940b:209) pointed out that in those acridids
“. .. Which possess a spermatheca with both an
apical and a preapical diverticulum usually have
the proximal end of the duct conspicuously en-

3Kevan, Akbar, and Chang (1969:248) refer to this region
as the “valve of spermathecal duct” and define their term as
. a small, thickened, often locally sclerotized area (or areas)
within the orifice of the spermathecal duct ... .” Gregory
(1965) stated that in Locusta “ . close to the spermathecal
aperture the ventral wall bears a triangular, sclerotized plate
(tsp). In some species of Barytettix this region of the recepta-
culum seminis is variously sclerotized at the distal end. Our
interpretation of the function of these hardened areas is that
they serve to more effectively bring the distal portions of the
aedeagus into precise apposition with the orifice of the thick
tube, thus producing a continuous channel from the male
spermatophore sac to the female spermatheca. Our dissections
indicate that these plates also serve as the terminal point of
aedeagal insertion. How these sclerotized areas could function
as a valve capable of closing off the entrance to the thick tube
is not clear.

T. J. COHN AND I. J. CANTRALL

larged, grooved above and its walls much thick-
ened.” Packard (1878) termed this region of the
duct the “bursa copulatrix,’” and Paoli (1937)
and Gregory (1965:40) both refer to this portion
as the “thick tube.”

The chief differences between the internal
genitalia of the females of Barytettix lie in the
shape and diameter of the thick tube, and in
modifications of the bursa copulatrix (BC in
Figs. 2 and 16). The more distal parts of the
receptaculum seminis, the constricted tube, the
vestibule, and the diverticulae, are typically
melanoploid and are rather similar from taxon
to taxon (Figs. 2 and 16). In all taxa, an elon-
gate shallow concavity exists in the roof of the
genital cavity (GC in Fig. 2). This concavity is
indicated by dashed lines in Figures 2 C, 17 D-F,
and 18 G-L. It extends from the opening to the
exterior between the ventral ovipositor valves (X
in Fig. 2 C), to near the center of the area en-
compassed by the basivalvular sclerites (ABS in
Figs. 2B and 16 A). Here the concavity forms
an opening (the spermathecal aperture of Greg-
ory, 1965) into the bursa extending above and
distally. The channel is open ventrally, has a
broad, gently arched roof, and bears a shallow
reinforced concavity on each lateral surface. The
amount of reinforcement of the lateral surfaces
is interspecifically variable. The positioning of
the genital components in the cleared tips of the
abdomens of a mating pair of Barytettix suggests
that the channel serves to orient the aedeagus
during insertion into the bursa.

A cleared receptaculum seminis is flexible,
resilient, and cartilaginous-like in texture. When
any part is displaced, it will spring back into
position when the cause of the tension is re-
moved. This resiliency is much more noticeable
in the bursa and proximal portions of the thick
tube than in the more flexible remainder of the
organ.

The walls of the bursa are capable of consid-
erable expansion or stretching. This is most
evident in the lateral and dorsal surfaces of the
lumen which may be folded or pleated (note
cross-sections of bursa in Fig. 18 L). A bursa
which is minimally reinforced with sclerotin is
rather feebly pleated and probably is subjected
to minimal stretching and stress during mating
(Figs. 16 A and E, 17). Similarly, in those spe-
cies where a little more than minimal stress ap-
pears to be involved, the reinforced areas are
smooth, or folded only laterally. However, per-

CONALCAEINE GRASSHOPPERS: BARYTETTIX

tinent areas are pleated in bursae which must
accommodate for an aedeagus of greater diame-
ter, as well as for the stress of aedeagal insertion
(Figs. 16 C, 18 L). Smooth and pleated areas of
the lateral and dorsal walls, and, additionally
in some species, a reinforcement in the distal
portions of the ventral wall, all may occur in the
same bursa. These reinforced areas lend strength
and rigidity to the walls which undergo the
greatest stress as the aedeagus is inserted, and
serve to aid in aligning the aedeagus correctly.
The pleatings and resiliency of the bursa walls
permit the bursa to expand to receive the aedea-
gus and to ensure a close “hand-in-glove’’ fit,
thus further ensuring proper alignment of the
orifice of the phallotreme and the opening of the
thick tube.

B ABS

PALOVIRIDIS

C PALOVIRIDIS

TAXONOMIC VALUE.— In certain groups of the
Acridoidea the structure of the elements of the
male concealed genitalia are sufficiently varied
so that these parts are used taxonomically at the
specific and intraspecific levels. In these same
groups, the female internal genitalic elements
have been considered to be less diverse, and, as
pointed out in the lengthy summary of Kevan,
Akbar, and Chang (1969), less reliance has been
placed on them as taxonomic tools. Few studies
have involved the entire receptaculum seminis,
and we are not aware of any study of this organ
on a comparative specific basis. As a result, we
have been quite surprised to find that in Bary-
tettix the receptaculum seminis affords diagnostic
features of unusual value at several taxonomic
levels. Furthermore, we feel the pattern of vari-
ation of female genitalia in this genus forms an
integral part of the primary evidence for genitalic
reproductive isolation, a theory discussed in a
later section (page 93 ).

MEASUREMENTS.— In our study we have indi-
cated relative size of the bursa copulatrix by
using a length to width ratio based upon the dis-
tance from the distal tip to the distal edge of the

Figure 2. Female internal copulatory elements in Barytettix.
Scale below each figure equals 1 mm.

A. humphreysti humphreysii. Semi-diagramatic, parasagittal sec-
tion of ventral portion of tip of female abdomen. México:
Sonora, Copete Mine, 30 mi E Carbo (ANSP)

B. paloviridis. _Receptaculum seminis and ventral ovipositor
valves, dorsal view. Topotype. México: Sinaloa, 2.5 mi NW
bridge over Rio Culiacan at Culiacan. I. J. Cantrall and T. J.
Cohn, 1961 No. 56

C. paloviridis, Bursa copulatrix, dorsal view (same data as for B.)

Symbols used in figures:

ABS — anterior basivalvular sclerite

AD — apical diverticulum

BC — bursa copulatrix

CKG — Comstock-Kellogg Gland

Cr — constricted tube

DPBTkT — dilation proximal bend of thick tube
DS — distal sclerite of bursa copulatrix
EG — egg guide

GC — genital chamber

OTkT — orifice of thick tube

PBTkT -— proximal bend of thick tube

PD — preapical diverticulum

Pe — pleats

PS — proximal sclerite of bursa copulatrix
SA — spermathecal aperture

SGP — subgenital plate

SLPD — secondary lobe preapical diverticulum
se — thin tube

Vv — vagina

VOV — ventral ovipositor valve

VST — vestibule

xX — distal edge of membrane connecting ventral

ovipositor valves

opening from the genital cavity into the bursa
(dashed transverse line in Figs. 2C, 17 D-F,

I. J. COHN AND I. J, CANTRALL

18 G-L), and a maximum width between these
points of reference.

THE SYSTEMATIC POSITION OF THE
CONALCAEA COMPLEX

In recent years much progress has been made
in the stabilization of higher categories in the
Acridoidea. Roberts (1941), Dirsh (1956, 1961),
and Uvarov (1966a) have contributed and char-
acterized classifications at the subfamilial and
familial levels, and Dirsh (1966) established
several superfamilial taxa. Rehn (Rehn and
Randell, 1963) has presented a lengthy, although
somewhat biased, account of the historical devel-
opment along these lines, particularly as it per-
tains to a single supertribe, the Melanoplini.

Much of the evidence for the association of
various genera into higher categories has been
derived from the study of the internal genitalia
of the male. Although Uvarov (1966b) gave no
specific reasons for his action, we postulate that
his establishment of the subfamily Melanoplinae
is based largely on the presence of a sigmoid
flexure between the aedeagus and the endophallic
plates in Catantops (Catantopinae) as compared
to a constricted point of articulation between
these structures in Melanoplus (Melanoplinae).
This difference between Catantops and Melano-
plus was shown by Roberts (1941:222-223).

Rehn and Randell (1963:7) gave a number of

. chief characters...’ which they used to
characterize their supertribe Melanoplini (prob-
ably = Melanoplinae of Uvarov, 1966b). All but
one of these features are so variable that it is not
possible to utilize them singly in the assignment
of specimens to a supertribe. One feature, a con-
stricted point of articulation between the aedea-
gus and the endophallic plates, is unique. On the
basis of this genitalic characteristic, Barytettix

“ee

and Conalcaea were considered components of
the Melanoplini.

Similarly, characteristics used by Rehn and
Randell (1963:8-10) to differentiate tribes are
so variable and so overlapping that we are unable
to associate any species of Barytettix with a
specific tribe. Fortunately Rehn and Randell
presented ina list (1963:10-12), their assignment
of genera to various tribal groupings. Sin¢ge a
cluster of characters appears to be needed to
associate a given genus with a given tribe or
generic group, these authors, in their generic
assignments, undoubtedly depended heavily
upon character similarities, as well as subjective
evaluation grounded upon the historical categori-
zation of a taxon, and upon Rehn’s vast experi-
ence with the Orthoptera. Their assignment of
numerous genera as being ‘‘atypical’ suggests
that their characterizations are not refined to
the point where specific characteristics are
unique for a given category. Numerous authors
(vide supra) have pointed out the generic hetero-
geneity of the Catantopinae and of the Melano-
plinae. The classification of Rehn and Randell
was, in their words, “... admittedly, a tentative
one,...”’ (1963:6), and, on this basis, they must
be credited with making some progress. Never-
theless, much heterogeneity is still apparent in
their hierarchy. This is evident in their placing
Barytettix and Conalcaea as “‘atypical’? compo-
nents of the tribe Melanoplini (=Melanoplinae
of Uvarov, 1966b).

Barytettix, Conalcaea and two undescribed
genera have a number of characteristics in com-
mon. The four genera form a rather cohesive
unit which we propose as the tribe Conalcaeini.

Conalcaeini new tribe

In western México, in the northern portions
the area occupied by the Conalcaea Complex,
the taxa assignable to the genus Conalcaea are
montane, whereas those assignable to the genus
Barytettix occupy the lowlands. Southwardly
both of these genera overlap undescribed generic
counterparts. All of these we consider as belong-
ing to the tribe Conalcaeini. The species of

these genera share numerous characteristics listed
below.

1. A rather stocky or robust body.

A pronotum gradually widening caudad
and with the posterior margin broadly
emarginate.
3. Tegmina

bo

lateral, elongate-lobate, with

CONALCAEINE GRASSHOPPERS: BARYTETTIX

rounded apices (Frontispiece).

4. Male supra-anal plate trigonal, with rounded
apex and usually with a single tooth or
sharp projection dorsally on each medio-
lateral surface (Figs. 8 A-B).

5. Furculae represented by short, rounded
lobes (Figs. 8 A-B).

6. Cerus shape based on a short falcate pattern

with the disto-dorsal margin variously en-

larged, distally and usually dorso-distally

incurved (Fig. 6).

Subgenital plate conical (Figs. 8 H, 15 H).

8. Sheath of the phallus complex and involved
in the formation of the dorsal valve of the
aedeagus.

9. Presence of striking, much enlarged, ap-
pressed and usually sclerotized ventral lobes

~

19

of the ramus of the cingulum (VLRC in
Fac) 1):

10. Presence of an exceptionally well-developed
pair of sclerotized areas in the ventral mem-
brane of the phallic mass (SLV in Fig. 1).
These obovate areas lie in the ectophallic
membrane on each side and below the
phallic mass, the ventral portions near one
another, thus forming, in caudal view, a
broad broken U-shaped reinforcement.
They are variously sclerotized and papillose.
Their function is unknown, although we
surmise that they serve to aid in holding the
male genital mass in position during copula-
tion. We have found them in all of the spe-
cies of the Conalcaeini.

Genus Barytettix Scudder

Barytettix Scudder, 1897, Proc. Amer. Acad. Arts and Sci.
32:197, 204 (January). Type species: Barytettix crassus
Scudder, 1897, by designation of Scudder, Proc. U. S. Nat. Mus.
20:27 (December 1897).

DIAGNOSIS.— The genus Barytettix is related
most closely to Conalcaea. The features differ-
entiating the two genera are listed in Table 1.

BARYTETTIX AND OTHER MELANOPLINE GE-
NERA IN WESTERN MEXICO.— Within the range of
Barvtettix there are a number of described and
undescribed short-winged grasshoppers which
might possibly be confused with Barytettix.
Several of these are quite distinctive and, in the
field, offer little difficulty in recognition. Others
approach Barytettix in one characteristic or
another.

As noted above, species of Conalcaea are
readily distinguished from those of Barytettix.
The remaining taxa of Conalcaeini are as yet un-
described, but can be recognized from Bary tettix
by having the habitus and color pattern of a
Conalcaea, or in lacking the distinctive Barytettix
black and yellow blotch on the lateral lobes of
the pronotum. In these forms the disto-lateral
ends of the lophi smoothly join the posterior
projections of the lateral plates of the epiphallus
without creating the notch characteristic of
Barytettix and Conalcaea. Further, the sheath
caudad of the ventral aedeagal valves is com-
plexly convoluted.

Other than the taxa of the Conalcaeini there
are few short-winged species which bear even a

superficial resemblance to Baryfettix except for
the genus Sinaloa. The general body habitus,
shape of the pronotum, and shape of the tegmina
of species of this genus are similar to those char-
acteristics in Barytettix. Even color and color
pattern tend to resemble those of green individ-
uals of Barytettix. This is particularly true of
Sinaloa nitida (Scudder) where the resemblance
is striking. However, the mainland species of
Sinaloa before ust are readily separated from the
species of Baryvtettix on the basis of several char-
acteristics. A fastigium less protruding before
the eyes than in Barytettix and a pronotum
without humeral sinuses give specimens of Sinaloa
a rather characteristic appearance. Elongate
finger-like furculae, absence of spinules on the
medio-lateral surfaces of the supra-anal plate,
and a subgenital plate with a short blunt cone
serve to differentiate the males of these Sinaloa.
The presence of a proximal node on each side of
the sub-obsolete longitudinal furrow of the
supra-anal plate of the males of S. behrensii and
nitida is also distinctive. The females of Sinaloa
lack the deep sinus found laterally on each side
of the subgenital plate of female Barytettix.
The species of Sinaloa lack a mid-dorsal longi-
tudinal light line on the abdominal segments,
and the black spot on the lateral lobes of the

4sinaloa nitida Scudder, behrensii Scudder, and pulchella
Hebard. We have not seen peninsulae (Scudder) from Territorio
de Baja California. The species is known from the female type
described as a Barytettix. It was transferred to Sinaloa by
Hebard (1917: 262).

20

T. J. COHN AND I. J. CANTRALL

TABLE 1

DIAGNOSTIC COMPARISON OF BARYTETTIX AND CONALCAEA

Bary tettix

Conalcaea

Colors bright and contrasting (Frontispiece): ground
color green, blue, gray or brown, usually with much
yellow; hind tibiae bright yellow, red, blue, or
intermediates, or combinations.

Antennae orange.

Disc of pronotum with yellow dorso-lateral stripes;
black blotch on lateral lobes usually broken with
yellow lines (except in southern nigrofasciatus and
southern psolus which have a continuous dark band
across entire lateral lobe) and usually restricted to
pro- and mesozona.

Metanotum and abdomen usually with a yellow
medio-longitudinal band (absent in terminalis and
some southern poecilus).

Tegmina usually unicolorous, rarely darker below, then
never sharply divided in color.

Hind femora with thin, black, dark green, or gray
lines on ventral outer and dorsal carinae (except in
terminalis and some poecilus).

Nymphal pronotum (Frontispiece) (nymphs of
contilus contilus and c. hiscatus not available) with a
broad median whitish or yellow stripe bordered by
broad dark or usually velvety black stripes (except
some individuals and population variants in tridens
and humphreysii), stripe sometimes extending to
occiput and usually only as far as second or third
abdominal segment. (This pattern is unique in
Mexican Melanoplinae known to us.)

Pronotum transversely convex, without lateral
carinae; metazona finely and densely punctate.

Dorsal aedeagal valve in side view, linear, without a
decided lobe (Figs. 1 A, C, E).

Lowland

Dull brown or gray, never green or blue; with little or
no yellow; hind tibiae pale orange or dull yellowish.

Antennae brown.

Disc of pronotum brown; lateral lobes with a usually
continuous dark band from cephalic to caudal
margin.

When present, yellow medio-longitudinal band
restricted to abdomen.

Tegmina almost always sharply divided into a lighter
upper half and a lower darker (usually blackish) half.

Hind femoral carinae brown and not darker than the
rest of femur.

Nymphal pronotum brown, or with narrow median
stripe, or with a somewhat broader stripe continuing
the length of the abdomen; sometimes bordered
laterally with dilute brown stripes.

Dorsum of pronotum flatter, with well-defined, but
not sharp lateral carinae; metazona less densely
punctate, with punctations ill-defined and often
sub-obsolete.

Dorsal aedeagal valve with a lateral, lobe-like,
decurved expansion covering part of lateral edge of
ventral valve.

Montane

CONALCAEINE GRASSHOPPERS: BARYTETTIX

pronotum is an unbroken stripe either running
the length of the pronotum with yellow above
and below, or terminating at the principal sulcus
and bounded by yellow. This black or dark bar
is of solid color except in females of nitida where
flecks of yellow may occur in the black.

In the geographic area under consideration,
the tyro may possibly experience some difficulty
in separating the genera Philocleon and Oedo-
merus from Barytettix. O. corallipes Bruner has
been reported from southwestern Arizona, Guay-
mas, Sonora, and Territorio de Baja California.
The species is mottled sombre brown in appear-
ance, and is without the conspicuous orange of
the antennae, the black and yellow markings of
the lateral lobes of the pronotum and the dark-
ened upper half of the pagina characteristic of
Barytettix. The dorsal surfaces of the caudal
femora are bifasciate and the outer carinae of the
caudal tibiae bear 6-7 spines instead of 8-10.
The supra-anal plate of the male is without
spinules, and the furculae are absent. The female
subgenital plate is laterally so deeply notched as
to appear W-shaped in ventral view. Philocleon
nigrovittatus (Stal) occurs in both Jalisco and
Colima. The dorsum of the pronotum of the
species of Barytettix in this part of México is
green, bordered laterally with a thin yellow line.
The pro- and mesozona of this structure in
P. nigrovittatus is longitudinally multi-striped
with black and lighter colors; the metazona is
marked with brick red.

Barytettix may be differentiated from any
other short-winged melanopline genus occurring
within its range by the conical subgenital plate
and the presence of spinules on the supra-anal
plate of the male, the broadly emarginate caudal
margin of the pronotum, and the elongate-lobate
tegmina.

GENERIC DESCRIPTION. — BODY robust, usually large;
length from fastigium to tip of caudal femora: male, 20.4-39.2
mm, female, 25.3-45.2 mm. HEAD normal, not exserted; vertex
feebly elevated above level of eyes; fastigium of vertex weakly
sulcate, produced before the eyes, rounding smoothly to the
facial carinae; facial carinae percurrent to clypeus, interspace
sulcate throughout; antennae filiform, extending to about the
middle of the tegmina. PRONOTUM gradually widening caudad;
convex in cross-section, without lateral carinae; median carina
weak to obsolete, intersected by the principal sulcus only; ante-
rior margin weakly convex-truncate; posterior margin distinctly
broadly emarginate; punctate dorsally and laterally, with the
metazona decidedly and contrastingly so. TEGMINA (Frontis-
piece) lateral, elongate-lobate, with rounded apices; reaching
about to or slightly beyond the base of the second abdominal
tergite. TYMPANUM present. No known stridulatory mecha-
nisms. MESOSTERNAL lobes rounded, well-separated. FE-
MORA of cephalic and middle legs moderately bullate. CAUDAL

21

FEMORA incrassate to moderately incrassate, lower lobe shorter
than upper, slightly exceeding tip of abdomen in male. BRUN-
NER’S ORGAN present. CAUDAL TIBIAE terete in cross-section,
without external distal spine, outer margin bearing 9 (8-10) evenly
spaced spines. SECOND TARSAL joint shorter than first.
PROSTERNAL SPINE well-developed; conical; bluntly acute;
straight or feebly retrorse. TERMINALIA, MALE~Supra-anal
plate (Figs. 8 A-B) trigonal; as long as wide to wider than long;
usually with mid-distal lateral distinct spinules; usually with
shoulders near tip; apex rounded, often with distal surface
raised and forming a plate-like extension to the edges of the
longitudinal furrow. Furculae reduced to short, rounded lobes.
Cerci (Fig. 6) falcate with disto-dorsal margin variously enlarged,
distally and usually dorso-distally incurved. Subgenital plate
(Figs. 8H, 15 H) distinctly evenly conical. TERMINALIA,
FEMALE-Lateral margins of subgenital plate evenly rounded.
Ovipositor valves normal, well-sclerotized; dorsal margins of the
upper valves and ventral margins of the ventral valves dentate.
CONCEALED GENITALIA, MALE (Fig. 1)—Epiphallus bridge-
shaped; ancorae well-developed, tooth-shaped; lophi strongly
developed, arched, disto-lateral ends separated from the posterior
projections of the lateral plates by a distinct sinus?; oval scler-
ites absent. Sclerite of the ventral lobe of the ectophallic mem-
brane highly developed into a distinct, strongly sclerotized area.
Ventral lobes of the ramus of the cingulum strongly and dis-
tinctly developed, appressed and usually sclerotized. Sheath
fused with the dorsal aedeagal valves, in most taxa completely
covering them; ventrally converging medio-distally from the tips
of the rami of the cingulum toward and under the ventral side of
the ventral aedeagal valves so as to form a pair of valve-like vari-
ously sclerotized, flattened, collar-like or finger-like projections
(the basal lobe of Gurney, 1951); these simple, not convoluted
or folded in a complex manner. INTERNAL GENITALIA,
FEMALE (Figs. 2 and 16)—Comstock-Kellogg glands present.
Spermatheca with both apical and preapical diverticulae, the
preapical bearing a secondary lobe. Bursa copulatrix variously
developed from a small, simple, unpigmented, sack-like structure
to elongate, variously sclerotized and pigmented chamber.
COLORATION (Frontispiece)—Ground color yellow, yellow-
brown, brown, gray, blackish, blue or green; with yellow and
black stripes and markings. Antennae orange (probably always
so in life), rarely brownish. Head usually with yellow post-ocular
stripes. Pronotum usually with dorso-lateral yellow stripes run-
ning the entire length, or terminating at principal sulcus; remain-
der of dorsum monochromatic or with an obscure median yellow
stripe margined with dark brown or black; lateral lobes with a
black spot extending over at least the dorsal portion of the pro-
and mesozona, bordered with yellow, black spot partially or
completely broken on prozona by a ventro-cephalic diagonal
yellow line, and often partially or completely broken on meso-
zona by a horizontal yellow line. Abdomen usually with a mid-
dorsal yellow stripe narrowly margined with black, and usually
proximally with a dorso-lateral broken yellow line or band,
bordered laterad by narrow black blotches; caudal segments
usually tinged with color of caudal tibiae (probably always so in
life). Tegmina with yellow or yellow-brown veins, membrane
black, rarely precostal and costal areas totally black. Metepister-
num yellow or yellow-brown; meso- and metepimera green,
black, or yellow margined with black. Caudal femora with
carinae black, or a shade of brown or green darker than lighter
portions of the femur; genicular areas usually tinged with the
color of the caudal tibiae, with black or dark lunae; dorsal half
of pagina usually darkened, often contrasting strongly with a
lighter colored ventral half. Caudal tibiae brightly colored,
yellow, orange, pink, red, reddish-purple, purplish-blue, or blue;
spines black tipped. Tarsi concolorous with distal portion of
tibiae. Nymphs usually and distinctively marked with three
broad stripes on dorsum of pronotum: the median yellowish,

SThis notch, characteristic of the epiphallus, is shared by
the species of Conalcaea.

te
Nm

the laterals velvety black. CHROMOSOMES-—Prof. B. John,
(in litt.) Department of Zoology, The University of Southamp-
ton, who examined the chromosomes of four or five individuals
each of humphreysii humphreysii, tridens, paloviridis, psolus,
contilus dicranatus, contilus similis, nigrofasciatus, poecilus, and

T. J. COHN AND I. J. CANTRALL

terminalis, indicated that “*... all of them had the same chromo-
some number, 2n = 23 male (XO), ...” and that “... all hada
marked pattern of chiasma localization with chiasmata predom-
inantly occurring at the distal ends of the chromosomes... at
least in males.”

THE SPECIES GROUPS OF BARYTETTIX
AND TABLES FOR THE
IDENTIFICATION OF SPECIES

It would be easy for us to state our conclu-
sions on the grouping of species within Barytettix
and to justify these by citing carefully selected
data. This is the traditional method, but it has
serious drawbacks. It tends to obscure many
data, and it prevents the independent assessment
of the reliability of those groupings and the pos-
sibility of alternative ones. The most objective
method would be some sort of numerical tax-
onomic treatment. In this genus, however, we
have encountered much difficulty in coding
character states and in handling the very exten-
sive geographic variation in many of the species.
Furthermore, the coding itself makes rapid ex-
amination and assessment of the data almost im-
possible. We have therefore chosen a compro-
mise method. In Tables 2 and 3 (p.24 ) we list
most of the morphological and color character-
istics of all the species, and present below a dis-
cussion of the correlations seen in the tables and
the results of grouping the species in different
ways. We have not included certain characteris-
tics present in only one species, others which are
intraspecifically highly variable, a few which
require detailed description, and those in which
the differences between species are too slight to
describe adequately in abbreviated form.

These tables can also be used as tabular keys
for identification. Our reasons for substituting
tables for keys are given in the Methods section.
Any character or combination of characters may
be used in any sequence for identification, but
the characters are so arranged that the first three
and the tenth (phallotreme), used together, will
serve to identify any male of the genus. In addi-
tion, we have used bold-face type to designate
distinctive and easily recognizable conditions.
Without dissection most females cannot be iden-
tified to species, but the possibilities can be
limited to two or three by color characteristics.

The most obvious correlation in the tables is
that of the characters of poecilus, crassus and
terminalis. Of the 12 morphological character-

istics shared by these species, three of the ae-
deagus and four of the bursa are distinctive.
These species are similar in almost all other
genitalic and color characteristics, although the
same features may also be found in other species.
Terminalis differs from the other two species in
five characters. In each case, the condition in
terminalis resembles no other species, or it differs
only slightly from conditions in crassus or poect-
lus. The three listed differences between crassus
and poecilus also are small differences in degree
and not of kind. Other differences among these
species listed in Table 4, p. 26, are either unique
or slight. There is little question that poecilus,
crassus and terminalis represent a cohesive unit
which we name the Crassus Group.

Of the remaining taxa, contilus, psolus, nigro-
fasciatus, and paloviridis share an unusually elon-
gate aedeagus and associated female bursa, strap-
shaped ventral valves, distinctive ventral lobes of
the sheath, and a ventro-caudal orientation of the
proximal bend of the thick tube. Among these
species there is a mosaic of similarities in several
other distinctive features (see also Table 6, p. 34,
and Table 7, p.38). Thus, psolus, nigrofasciatus
and paloviridis share distinctively long and
dilated thick tubes, a similarly very elongate
aedeagus, and an almost invisible phallotreme,
whereas contilus, nigrofasciatus and psolus share
a distinctive color pattern, semicircular dorsal
valves (in cross-section), as well as two distinc-
tive features of the ventral lobes of the sheath
(at least as variants in psolus). Psolus and nigro-
fasciatus also share with one of the contilus sub-
species (fectatus) almost identical sclerotizations
of the bursa. Both contilus and psolus have
lateral apices of the dorsal valves, and nigrofasci-
atus, some paloviridis, and contilus tectatus share
a distinctive hump on the dorsal valves.

Paloviridis and contilus show specific similari-
ties to other species. The color and pattern of
paloviridis is identical to those of poecilus and
terminalis (except for the lack of an abdominal
stripe in the latter), and to the green geographical

CONALCAEINE GRASSHOPPERS: BARYTETTIX

variant in h. humphreysii. Paloviridis shares no
other characters with poecilus and terminalis,
and very few with h. humphreysii. In the case
of the listed similarity to h. humphreysii the
excision of the dorsal valves is between acute
median teeth in humphreysii, but between the
two simple free lobes in paloviridis. Otherwise
those two species are similar in only two general
features, the dorsal position of the dorsal valves
(asin the Crassus Group also), and their narrowly
rounded apices (in some of the geographical var-
iants in humphreysii which usually also have
median teeth, unlike paloviridis). Contilus is
similar to h. cochisei in the broadly open phal-
lotreme with more or less parallel dorso-lateral
margins, and in the short proximal bends of the
thick tube of the female. However, the ventral
valves are otherwise rather differently shaped
and the thick tube has a different orientation in
the two species. The similarities which palovi-
ridis and contilus bear to the Crassus Group and
to h. humphreysii and h. cochisei do not out-
weigh the more numerous and distinctive simi-
larities they bear to psolus and nigrofasciatus.
We therefore designate contilus, psolus, nigro-
fasciatus, and paloviridis as members of a second
species group, the Psolus Group.

The remaining taxa, tridens, h. humphreysii,
and fh. cochisei, together share no highly distinc-
tive characteristics, but are generally similar in
almost all features. Furthermore, tridens is al-
most identical to the southernmost and adjacent
h. humphreysii populations in most characters
except the ventral lobes of the sheath, details of
the ventral valves, and the color of the northern
colonies. Even in these characters, there is gen-
eral similarity to at least some humphreysii
colonies, and little similarity to any other species.
Tridens and southern humphreysii also share the
distinctive strongly incurved and distally greatly
expanded cercus. H. cochisei is almost identical
to the adjacent northern h. humphreysii popula-
tions in cercus, ventral lobes of the sheath, thick
tube of the female, ground color, color pattern,
geographic trends in color and pattern, and in
some genitalic features (Table 15, p.79). The
humphreysii populations south of Hermosillo
have ventral valves which are generally similar to
those of the adjacent cochisei populations (Figs.
11B and G). Where cochisei differs strongly
from humphreysii in the dorsal valves, it shows
no obvious similarity to any other species. Nor
is the more elongate aedeagus or bursa of cochisei

more similar to contilus except in the broadly
visible lateral walls of the phallotreme. As noted
above, the ventral valves which contain the phal-
lotreme are of a different shape in cochisei, and
in this they are more similar to humphreysii.
That subspecies also has a broadly open phallo-
treme, at least in the northern populations, but
it is obscured by the more closely set dorsal
valves. Thus, tridens, humphreysii and cochisei
are much more closely similar to one another
than any are to other species, thus qualifying
them for designation as the third species group
which we name the Humphreysii Group.

The decision to treat these clusters of species
as groups rather than distinct genera was based
on a number of considerations. A genus is an
arbitrary unit in nomenclature and can reflect
either a typological approach or one of biological
continuity and relationship. All the species here
considered have geographic cohesiveness in dis-
tribution, are all readily recognizable as being
very similar in size, body contour, wing shape,
wing length, and color pattern, as well as several
features of the male external copulatory arma-
ture. The elements of the concealed male and
the internal female genitalia which seem to differ
so strongly one from another can all be readily
homologized and are without doubt derivatives
one from another, part by part. We feel that an
appreciation of these phylogenetic relationships
would be seriously obscured by the recognition
of additional generic entities.

THE CRASSUS GROUP

Occupying the southern portion of the range
of the genus are a group of populations which
are similar in all bursal, and almost all aedeagal
and color features. Several of the shared aedeagal
and bursal characteristics are distinctively differ-
ent from all other Barytettix species. These
southern populations thus form a morphologi-
cally and geographically cohesive group within
the genus.

The group comprises a widespread form in
Sinaloa and Nayarit, poecilus, and two geograph-
ical isolates, crassus in southern Baja California,
and a new species, terminalis, in coastal Jalisco
near Colima. Although these might well be
treated as the subspecies of a polytypic species,
we think that the differences between them are
numerous enough, and some are distinctive
enough, to indicate that they represent the com-

ponents of a biological superspecies. Further-
more, we find no tendency toward the distinc-
tive features of either crassus or terminalis in the
widespread and geographically variable poecilus,
suggesting a longer rather than a shorter period
of isolation.

We consider it best to treat the marked geo-
graphic variation in poecilus descriptively rather
than using it to erect subspecies. One of four
geographically variable characters in poecilus is
discordant with the other three. Of the remain-
ing three which are concordant, one exhibits
additional variation not shown by the others,
and two show different degrees of intrapopula-
tion variation. We believe that designation of
subspecies might obscure this biologically signifi-
cant pattern.

The three species share the characteristics
listed below which are unique within the genus.

1. Dorsal aedeagal valves (Figs. 7 B, D, F; 8 C):
strongly fluted longitudinally, enlarged disto-
laterally (when pressed flat), disto-lateral por-
tions up-turned, free lobes short, making
valves appear merely notched medio-distad.

2. Ventral aedeagal valves with a usually well-
developed, oblique ridge on ventral surface
(Figs. 8 D-E).

3. Ventral valves strongly curved downward or
outward (Figs. 7 B, D, F).

4. Ventral lobes of aedeagal sheath short, form-
ing only a brief collar beneath ventral valves
Gica):

5. Ramus of cingulum well-developed disto-
ventrad, thus reducing lobes of sheath to a
collar (Fig. 7).

6. Bursa copulatrix short, the width about two
times the length (Figs. 17 A, D).

7. Thick tube of receptaculum seminis arising
dorso-laterally, and proceeding dorso-cephalad,
wide at origin and immediately beyond, prox-
imal bends much longer than bursa (Figs.
17 A, D).

The following characteristics, shared by all
three species of the group, are also occasionally
found in the Humphreysui or Psolus groups.

1. Ground color always bright green (possibly
olive or brownish in crassus) (Frontispiece).

2. Dorso-lateral yellow pronotal stripes extremely
narrow (possibly absent in some individuals,
but this may be the result of poor preserva-
tion) (Frontispiece).

T.J. COHN AND I. J. CANTRALL

3. Bursa copulatrix triangular,
tized (Figs. 17 A, D).

weakly sclero-

Table 4 summarizes the differences between
the three species of the Crassus Group. Charac-
teristics in bold-face type are more distinctive
and easily recognized, and may serve as key
characteristics. To distinguish any of the three
species from all other species of the genus on
the basis of color alone, see Table 3, p. 25, and
RablessespeZor

Barytettix poecilus (Hebard)
Frontispiece, Figs. 1 A-B;6 S-X; 7 A-B; 8 A-E;
16 A-B; 17 A-D

Conalcaea poecila Hebard, 1925, Trans. Amer. Entomol. Soc.
51:290-292, Pl. VII, Fig. 7, Pl. VIII, Fig. 4. [Holotype, ¢,
México, Sinaloa, Venvidio; Hebard Collection in the Academy of
Natural Sciences of Philadelphia, Type No. 983.]

Cantrall has determined that Venvidio as re-
corded by Hebard is a misspelled transliteration
of Venadillo, a small village 6 mi N Mazatlan
Cathedral, and not near Los Mochis as surmised
by Gurney (1951:302). This determination was
based on a study of the correspondence of the
collector, J. A. Kusche, the localities of collec-
tions made within a few days of dates when he
was known to have been at El Venadillo, and the
distribution of species collected by Kusche at
El Venadillo.

DIAGNOSIS.— Males of poecilus may be dis-
tinguished from all other Barytettix by the com-
bination of an invariably green ground color,
down-curved ventral aedeagal valves (Fig. 7 A),
and a moderately deep, V-shaped excision of
the dorsal aedeagal valves (Figs. 7 B, 8 C). Fe-
males may be distinguished from all species other
than crassus and terminalis by the combination
of a green ground color, short bursa copulatrix
and thick, dorso-lateral thick tube. From termi-
nalis, both males and females differ in the dark
carinae of the femora, the frequent presence of a
dark paginal stripe, and a yellow mid-dorsal ab-
dominal stripe. From crassus, apparently only
those females lacking paginal or abdominal
stripes, or possessing blue tibiae may be distin-
guished, although the uniformity of alternative
color characteristics of crassus is as yet unknown
because of poor preservation of material before
us and the small number of localities from which
the species is known. See Table 4 for other
comparisons of the species of this group.

TABLE

2

GENITALIC CHARACTERISTICS IN BARYTETTIX SPECIES
(Numbers and letters in parentheses refer to figures; bold-face type indicates easily recognizable, “key” characteristics.)

= T F
poecilus Crassus: terminalis tridens h. humphreysii h. cochisei contilus psolus nigrofasciatus iridi.
paloviridis
| (p.24) (p.30) (p. 31) | (p.84) (p. 66 ) (p.77) al (p.37 ) (p.44) (p.47) (p.51)
MALE GENITALIA
AEDEAGUS
GENERAL SHAPE [ Wider than long (p.14 and Figs. 1 B, 1 F) it Much longer than wide (p.14 and Fig. 1 D)
Length/width 0.76 — 0.88 0.67 — 1.0 0.75 — 0.90 0.48 — 0.63 0.57 — 0.84 0.74 — 0.87 1.6—-1.9 2.6 — 3.0 js PG] 22 — 79
VENTRAL LOBES OF SHEATH [ REDUCED TO NARROW COLLAR ————__ ] | Laterally narrow [—————_ Truncate or medially mee PRODUCED, NARROW, APPROXIMATE
(1 A-B) lobed, U-excised lobed, V-excised (1 C-D)
(12 O-S) (1 E—-F; 12 A—N)
DORSAL VALVES
Median excision V-shaped, moderately Very broadly V- Narrowly V- or U- Narrowly U-shaped, Variable: V- to Broadly U- or V- Variable: narrow U- Broad distad, narrow V-shaped, deep Narrow U-shaped,
between free lobes deep (7 B; 8 C) shaped, shallow shaped, moderately deep or shallow narrowly U-shaped; shaped, deep shaped proximad, U-shaped proximad; (14 H) deep (14 F)
(7 D) deep (7 F) (10 S—T) deep (10 F—R) (10 A-E) variable distad; deep (14 G)
deep (14 A—E)
Projection on mesal [ None ]| Aciculate or acute None, obtuse, or None or obtuse; None, lobate or [ None
margin (7B, D, F; 8 C) (10 S—T) acute; near middle near base acute produced; at (14 F-H)
(10 F—R) (10 A—E) middle (14 A—E)
Apex or disto-lateral Broadly rounded or Broadly rounded- Broadly rounded; Elongate-acute; Broadly rounded to Narrowly acute or Narrow or broad; Broadly rounded; Narrowly rounded; Narrowly rounded;
projection rectangulate; truncate; dorsal dorsal (7 E—F) dorso-lateral elongate-acute; narrowly rounded; lateral lateral dorso-lateral dorsal
dorsal (7 A—B) (7 C-D) (10 S—T) dorsal to dorso- dorso-lateral (13 E; 14 A—B) (13 C; 14 G) (14 H; 15 EB) (13 A; 14 F; 15 A—D)
lateral (10 F—R) (10 A-E)
Cross-section [—-—— W-SHAPED ABOVE (dry preparation) -———— -] | Slightly convex Moderately to Slightly convex Semi-circular above, Semi-circular above Semi-circular above, Strongly convex
(medio-distal) above to turned-up slightly convex above sometimes slightly often flared distad above, flaring
edges, or concave above flared distad
VENTRAL VALVES:
Attitude [——————__ DOWN-CURVED —————__] OUT-CURVED Broadly incurved Broadly incurved [ Straight or slightly incurved
(7 A, C) (7 F) (11 N—P) (Tezopaco-traight) (11 A—B; 14 I—P; 15 F—G)
(11 E—M)
General shape (— Gently tapering distad t—_—__________ } | Moderately tapering Moderate to strongly Strongly tapering [ Strap-shaped
(7 B, D, F; 8 D-E) distad (11 N—P) tapering distad, some distad, strong (and often gently tapering distad)
with strong proximal proximal shoulders (14 I-P; 15 F—G)
shoulders (11 E—M) (11 A—B)
Apices [— = Blunt Briefly acute Sharp-acute or ACICULATE Acute or narrowly [ Briefly acute, or narrowly blunt
(7 B, D; 8 D-E) (7 F) ACICULATE (Tezopaco-sharp-acute) blunt (11 A—B) (14 I-P)
(9 F; 11 N—P) (11 E—M)
Visibility of lateral walls [| —————————__Invisible (obscured by dorsal valves) mae Broadly to narrowly visible —-——— ] | [ ——————— BROADLY VISIBLE ]| Narrowly visible Narrowly visible Invisible or
f phallotreme (7 B, D, F, ventral view) but usually narrowly (8 F; 14 A—E) distad (14 G; 15 R) (14 H) narrowly visible
(from above) obscured by dorsal valves (8 G) distad (14 F; 15 P-Q)
Oblique ridge on ventral Weakly to not projecting | Strongly, roundly Angulately projecting [ = Absent
urface proximad laterad (7 B, 8 D—E) | projecting laterad (7 D) || laterad (7 F) (but swellings and excavations may be present)
CERCUS
Distal enlargement Slight (6 S—X) Weak (6 R) Moderate (6 Q) [ ————._ Moderate to strong (6 B~P) ———— ]} | Moderate (6 A) Weak to moderate [ Weak or not at all
(6 FF—JJ) (6 Y—EE)
Incurvature [ ~ Weak Moderate [—————— Moderate to strong ——————— ] | Moderate [ Weak
(15 H) (8 H) (15 H)
FEMALE GENITALIA
BURSA COPULATRIX
General Length [ Short =: mie Very long
(Length of bursa/ 0.04 0.04 0.04 0.08 0.05 0.11 0.27 — 0.35 (— 0.42?) 0.50 0.54 | 0.42
length of pronotum)
Apex (dorsal view) [ - - Triangular (17 A) — } | [_— Truncate (17 E) — — ] | Triangular (17 F) [ ——————_ Rounded (18 G—J) WI Rounded-triangular (18 K—L)
THICK TUBE
Origin [ Dorso-lateral (17 A, D) ] | [— Sub -disto-dorsal (17 B, E) - ] | Disto-dorsal (17 C, F) Ventral (18 A—C) [ ——————_ Disto-ventral (18 D-E) | Disto-dorsal (18 F)
Direction from origin [— ——— DORSO-LATERO-CEPHALAD — Ht DORSO-CEPHALAD If VENTRO-CAUDAD
(17 A, D) (17 B-C, E-F) (18 A-F)
Length of proximal bends {| ———. - — 3 times length of short bursa Vt - About 1/2 length of short bursa ] | 1/3 length of elongate [ 2/3 length of elongate bursa
(17 A, D) (17 B-C, E-F) bursa (18 A-C) (18 DF, J-L)
Dilation of proximal bends [ Uniformly wide WI — Uniformly narrow ] |Uniformly narrow or one | [ A large dilation in each bend
mp) (17 E-F) small dilation (18 A-C) (18 D-F, J-L)

oe |

CONALCAEINE GRASSHOPPERS: BARYTETTIX

2S
TABLE 3
COLOR CHARACTERISTICS IN BARYTETTIY SPECIES
* 3 g Sy E ES
te eares , aS Sk yaiss serge
S 2 = S = 9 = = S 5
Sac aees a pene s 2 8 §
a S = = = = S g = i
GROUND COLOR
Grayiraw-a bene xX Xx :
Brownish, or yellow brown .............,.... F : F S X xX X X xX R!
Green, olive or yellow-green... _ . ; Beka X XxX? xX ; S S xX
Blue, or yellow and blue ..._ | 2 Paty aes x £ i N
PRONOTUM, LATERAL LOBES
Dorso-longitudinal black bar
Continuous and equal across lobes or inconspicuously
narrowed caudad and cephalad... x 4 X
Conspicuously weak, or narrow on metazona xX xX NR
AIDsen tonne taz Ona ime neweiayuen ene tania ana x X X X xX X X
Conspicuously broken on prozona ....... . X Xx Xx xX Xx X xX X R > 4
Ventral quarter
MGUORE Gus Hota dia Sait te ol eee ee : xX xX X
Not yellow, but same as ground color ...... _ - xX x X xX Xx xX xX
Ventral carina
Blackie psec Pa VM: nurse awd x Send ele Xx Xe a RZ
Not black, but same as ground color......... xX x xX X xX XxX XxX
HIND FEMUR, DORSAL STRIPE OF PAGINA
Absent or inconspicuous ............. el hee S : x So? S R x
Green, brown, blue, or weakly defined... ... . : x X ‘ xX X x S x
Jet black and sharply defined ........,.,...... ; X xX xe N
HIND TIBIAL COLOR?
Unicolored ae mn meaner Pen ees, ee ee ee X X X xX N N xX xX S X
Conspicuously,bicolored) ............... R X S S X xX N
VWENOWP Gioicr oh ore OAS Deen ae ae ae oe N RN
ORME oA Suoloue a ane b Aa ee ee N RN XxX S
Reda arma nt heir mest hate cee S X X x S X
WHS > 0-016 Scho hG Bc oe eee ne ee ee N ».¢ xX S X xX N X
EUrp emp me ae gis Cosy aces chee ee, We N S XxX xX N RS
BCP EEE STE cee i a er Ae, a ee S X x N
ABDOMEN, MEDIO-LONGITUDINAL YELLOW STRIPE
Present, broad and usually black bordered ........ . N xX X X xX X X X xX
Absent, or present as a thin whitish line ....... ~~ Ss R xX
X the commonest condition in many or all colonies of the species.
R rare, found as individual variants or in a few scattered colonies.
N the common condition found in northem colonies only.
S the common condition found in southern colonies only,
1 only in some individuals near Guaymas and Altata.
2 only in a few brown individuals, black on the ventral, but not on the outer surface.
3 the geography of tibial color is often helpful in making probable identifications, see Tables 5 (poecilus), 10 — 11 (Psolus Group), 14 (humphreysii), 15 — 16 (cochisei)

However, extension to areas, especially in the mountains, not treated in this paper is questionable. When using these tables for specimens found between Guaymas and
Mazatlan, the urea of overlap of many species, it should be noted that paloviridis is wine-legged, except for a few individuals with purple legs south of Culiacan, and that

poecilus is always purple-legged in that region.
see Frontispiece.

*

TABLE 4
DISTRIBUTION OF CHARACTERISTICS IN THE CRASSUS GROUP

T. J. COHN AND I. J. CANTRALL

(Bold-face type indicates easily recognizable, “key” characteristics.)

Character

poecilus

Crassus

terminalis

Supra-anal plate: lateral
processes (Figs. 8 A—B)

Cercus: dorsal margin
(Figs. 6 Q-X)

Cercus: disto-dorsal
portion (Figs. 6 Q-X)

Dorsal valves: median
notch (Figs. 7 B, D, F;
8 C)

Ventral aedeagal valves:
lateral margins (Figs. 7 B,
D, F;8 D-E)

Ventral aedeagal valves
(Figs. 7 A, C, E-F)

Ventral aedeagal valves:
ventral ridge (Figs. 7 B,
D, F;8 D-E)

Ramus of the cingulum:
dorso-lateral surface
(Figs. 7 A-F)

Ramus of the cingulum:
ventro-medial margins
(Figs. 7 B, D, F)

Hind femur: dorsal dark
stripe on pagina
(Frontispiece)

Hind femur: carinae
(Frontispiece)

Dorsal abdominal yellow

stripe (Frontispiece)

Hind tibia: color
(Frontispiece)

None or fine spines to
wide-based teeth

Barely concave

Barely enlarged

Moderately deep,
V-shaped

Smoothly convex or
straight

Strongly down-
curved

Oblique to almost
longitudinal

Smooth

Weakly divergent

Present in north,
absent in south

Black or dark green

Often absent, usually
inconspicuous when
present

Purple, reddish or
blue

Small spines or
tubercles
Weakly concave

Weakly enlarged

Shallow, very broadly
V-shaped

Obtuse-angulate,
distally almost

straight

Strongly down-
curved

Almost transverse

Smooth

Moderately divergent

Present

Gray-green

Apparently well-
developed

Red

Weak swellings to short,
small teeth
Moderately concave

Moderately enlarged

Moderately deep,
narrowly V- or
U-shaped

Ob tuse-angulate,
distally concave

Strongly out-curved

Oblique

With a large knob

Strongly divergent

Absent

Usually green, if darker,
then only distad

Barely indicated

Purplish-red

CONALCAEINE GRASSHOPPERS: BARYTETTIX

Color comparisons of all species of the genus
may be found in Table 3, p.25. Some specimens
of poecilus are indistinguishable from some green
humphreysii and paloviridis on the basis of color
alone. However, green humphreysii occurs far to
the north, and the red and purple-legged palo-
viridis are also allopatric to the poecilus which
they resemble (see Table 5, p. 28, and the dis-
cussion of geographic variation under paloviridis,
p.53).

Within the Crassus Group, poecilus closely
resembles crassus in most of the characteristics
by which it differs from terminalis. These com-
parisons are summarized in Table 4, and their
phylogenetic implications are discussed below in
a section dealing with relationships among all
Barytettix species.

SPECIES DESCRIPTION.— SUPRA-ANAL PLATE (Fig.
8 A): medio-lateral processes absent or varying from swellings
through small, narrow spines to large, wide-based teeth. CERCUS
(Figs. 6 S-X): weakly incurved; dorsal margin weakly concave;
disto-dorsal portion slightly enlarged; disto-ventral tooth not at
all, to moderately produced, rectangulate to acute. AEDEAGUS
(Figs. 7 A-B): short (0.05-0.06 times length of pronotum), wider
than long (length 0.76-0.92 times greatest width). DORSAL
VALVES (Fig. 8C): dorsally with a deep longitudinal furrow
on either side of mid-line, medial portion strongly convex, lateral
portions strongly bent upward, so that valves appear W-shaped in
distal view; free lobes short, broad, apices blunt rectangulate,
medial margins forming a moderately deep V; lateral margins
straight or weakly concave in proximal half, then flaring and
convex to apices, distal expanded portions if flattened would
appear intermediate in length and width between crassus and
terminalis. VENTRAL VALVES: in caudal view (Figs. 8 D-E),
usually gently tapering from near base, lateral margins straight
and parallel or convex; medial margins straight and weakly diver-
gent or parallel; apices rounded, truncate or briefly blunt acute;
ventral surface of each valve with a prominent, oblique, curved
ridge which rounds into outer margin near middle, sometimes
forming convex bulging portion of that margin; surface cupped
between ridges, becoming flat distad; in lateral view (Fig. 7 A),
evenly bent ventrad; in distal view, each valve with a longitudinal
ridge on dorso-medial margin largely obscured by dorsal valves.
RAMUS OF CINGULUM (Figs. 7 A-B): ventro-distal portion
broad, strongly developed and projecting distad to level of junc-
tion of dorsal valves with ventral lobes of sheath, thus reducing
ventral lobes to a narrow collar below ventral valves; dorso-
laterally smooth, ventro-laterally rough and variably rugose; in
caudal view (Fig. 7 B), ventro-lateral margins sub-parallel, medial
margins weakly divergent. VENTRAL LOBES OF SHEATH
(Figs. 7 A-B): forming a narrow collar below ventral valves;
lateral portions usually extending distad of ramus and visible in
side view; ventral portions with a brief flange curled dorsad
toward ventral valves. BURSA COPULATRIX AND THICK
TUBE (Figs. 17 A, D): very short and broad (width about twice
the length), triangular, proximally with a moderately well-
developed sclerite on each side, distally pleated, and strongly
sclerotized at or possibly within orifice of thick tube; thick tube
arising dorso-laterally and proceding dorso-cephalad, wide at
origin and sclerotized there, moderately dilated immediately
beyond origin, thence gradually narrowing, length between
proximal bends moderate but much greater than bursa. COL-
ORATION (Frontispiece): ground color green with a small
amount of yellow. Head without yellow post-ocular stripes.
Pronotum with dorso-lateral yellow stripes very narrow and not

27

extending onto metazona, apparently often absent (possibly the
result of poor preservation); black spot of lateral lobes not ex-
tending onto metazona, usually completely broken by cephalic
diagonal yellow line, caudal horizontal line usually absent,
when present weakly defined, spot narrowly bordered ventrad
by yellow, ventral third of lobe and ventral carina green. Epimera
partly black in north, green in south. Abdominal mid-dorsal
yellow stripe often absent, especially in south, when present
usually inconspicuous (possibly the result of poor preservation);
tip of abdomen often weakly tinged with color of hind tibia.
Hind femur with pagina in south unicolored green, in north with
a dorsal green or brown stripe and a ventral yellow or yellowish-
green stripe; carinae dark green or black; geniculae usually tinged
with color of hind tibia and with black lunae. Hind tibia purple,
reddish, or blue. MEASUREMENTS: those of the series studied
are summarized in Figures 19 and 20.

GEOGRAPHIC VARIATION.— The geographic
variation displayed by this species in four char-
acters is summarized in Table 5. Only critical
localities are listed. Others from which material
has been examined have been indicated as lines
connecting listed localities. Transitions between
the variant characteristics are relatively abrupt
geographically, with little intermediacy or mix-
ture of characteristics (except in the transition
between small and absent supra-anal plate proc-
esses). Exceptions are indicated in the table by
parenthetical data, the fraction representing the
proportion of that particular population showing
the deviant character. The difference between
the extreme conditions of the ventral valves is
obvious (see Figs. 8 D and E), but individual var-
iation and the technical difficulty of comparing
specimens makes difficult the determination of
the nature and position of the break. In this
character, there may be a more gradual transi-
tion, and more mixture of characteristics in the
El Venadillo-Villa Union region than is indicated
in the table. Southeast of the Villa Union popu-
lations, there is little variation in the ventral
valves.

The strong break in two and possibly three
characters at the Rio Presidio suggests that large
rivers play a major barrier role in the dispersal of
Barytettix species, as is also suggested in the var-
ilation in paloviridis and the distribution of
psolus. In the region of the Rio Presidio collec-
tions where vegetation, climate and other Or-
thoptera are relatively uniform, it is difficult to
envision strongly different selective factors on
either side of the river which would account for
the difference observed. There is evidence that
the three characters are independent of one
another in the few individuals which show dif-
ferent combinations of these characters, and
there is no evidence to indicate that the charac-
ters arose together. Thus, the break in all three

TABLE 5

T. J. COHN AND I. J. CANTRALL

GEOGRAPHIC VARIATION IN BARYTETTIX POECILUS

Ventral Valve Supra-anal
Localities Ridge Processes
(distances in miles) Tibial Color Paginal Stripe (Figs. 8 D-E) (Figs. 8 A-B)
66 SE Culiacan PURPLE DARK LOW, STRONGLY WIDE TEETH
(70 N. Mazatlan) sometimes CURVED
reddish
distad or
bluish
proximad
El Venadillo
(5 N Mazatlan)
1.3-2.6 NW Villa DARK LOW, STRONGLY WIDE TEETH
Union (1/11 light)* CURVED (1/6 small,
thin)*
Rio Presidio at Villa Union
2 E -2.4 NE Villa LIGHT OR NONE HIGH, WEAKLY SMALL, THIN
ee (4/15 light)* CURVED
15.1 SW Santa Lucia (1/7 reddish)*
(NE Concordia)
|
26 NW Escuinapa (1/5 light)*
5 N Acaponeta PURPLE
1 W Acaponeta, 2 S REDDISH
Acaponeta
Rio Acaponeta — ==
3 SE Acaponeta SMALL;
REDUCED;
OR ABSENT
Tepic (2/2 purple*
12 S. Jalisco)
Plan de Barrancas BLUE OR LIGHT OR NONE HIGH, WEAKLY USUALLY
(E Ixtlan del Rio) PURPLISH CURVED ABSENT

*Fractions indicate proportion of population with indicated characteristic.

CONALCAEINE GRASSHOPPERS: BARYTETTIX

at the river probably means that there has been
little communication across it in recent times.
Each character has a wide distribution northwest
or southeast, and must therefore have been in
existence fora considerable period of time. That
there is some crossing of the river is suggested
by the presence of a very few individuals with
some eastern characteristics (but not in com-
bination) on the west side, and the possible in-
fluence of western paginal color on the east side
populations. The continuity of the fourth char-
acter, tibial color, across the river is probably to
be explained by the possession of this color by
the ancestral populations which either crossed
the river, or whose distribution was split by it.
Farther south tibial color shifts from purple to
red on the west side of the Rio Acaponeta and
within seven miles of that river. Either the red
characteristics has just crossed the river north-
westward, or the red of these populations has
not yet been replaced by purple moving south-
eastward. The blue-legged populations near
Ixtlan del Rio represent a geographically isolated
sample, but is probably connected to the coastal
plain populations by colonies in the valley of the
Rio Grande de Santiago where we have not
collected.

Larger samples near both the Rio Presidio and
the Rio Acaponeta, and samples farther upstream
might provide more definitive evidence on the
problem. It might also be instructive to continue
to sample these populations over a period of
years and thus possibly determine the rate of
movement of the characteristics.

HABITAT AND ASSOCIATED SPECIES.— This is a
common species in weedy, bushy habitats along
the road in thorn forest and tropical deciduous
forest zones. It has been found within thorn
forest at several localities, in thin weedy and
bushy vegetation. It has also been found in open
oak woodland at 2950 feet in Nayarit, in rather
thin weedy, grassy vegetation with scattered
bushes. However, farther north on the Mazatlan-
Durango Highway, poecilus apparently does not
extend into the oak zone, which begins at about
3500 ft., although the lower edge of this zone
has been sampled at only a few localities.

The northernmost colonies of poecilus are
sympatric with the southernmost colonies of
tridens, paloviridis and the three known colonies
of contilus similis. Between 66 and 77 mi S of
Culiacdn, various combinations of poecilus and
two of the other species occur (see Table 19).

29

Poecilus is also sympatric with nigrofasciatus at
most of the known localities for the latter be-
tween Santa Lucia and Concordia and at San
Ignacio.

The distributional pattern of poecilus, palo-
viridis, and psolus in the region of the northern
limits of poecilus invites some interesting ques-
tions concerning the factors limiting the distribu-
tion of that species. Because of the rarity of
tridens in this region, it is ignored in the follow-
ing discussion. Along the west coast highway,
poecilus and paloviridis overlap for about 10
miles. Immediately north of this zone we have
many collections of paloviridis at short intervals
but neither poecilus nor psolus. East of this
zone, we have collections from two roads. That
leading to Cosala leaves the west coast highway
in the middle of the overlap zone. Fourteen
miles northeast of this junction, and about 600
feet higher, our first collection contains only
paloviridis and psolus, both of which were com-
mon. At several localities in the vicinity of
Cosala, 14-25 miles beyond the first and at about
1500 feet, only psolus was found in abundance.
The habitat at the first locality was rich thorn
forest or transitional to tropical dediduous forest,
and that at the Cosala localities was transitional
or rich tropical deciduous forest. On this road
we have another set of collections west of the
poecilus-paloviridis coast highway overlap zone
on the road leading to San Ignacio. Along this
road, which leaves the coast highway fourteen
miles south of the overlap zone, the two species
are geographically sympatric but occur in appar-
ently slightly different habitats. Thus, palovi-
ridis was found to be common, and poecilus
absent at 4.6 mi E Coyotitan, whereas poecilus
occurred within a mile to the east in a slightly
bushier habitat. At 1.1 mi SW San Ignacio Ferry,
paloviridis was found only in the weedier parts
of a cut-over roadside area, whereas poecilus
occurred generally in more bushy areas as well
as in the weedy patches. In view of the abun-
dance of paloviridis in many habitats north of
the overlap zone, the similar abundance of
poecilus south of the zone, and the occurrence
of poecilus up to 3500 feet about 50 miles to
the south on the Mazatlan-Durango Highway, the
restriction of poecilus northward and upward in
the overlap zone may thus be the result of inter-
action with its congeners. This area is relatively
little disturbed by man and poecilus is generally
common here. It would be an ideal area to study

30

the possibility of competitive exclusion in these
herbivorous species.

SEASONAL OCCURRENCE.~— Earliest records for
adults of this species are 26 and 27 July at 4.3
and 11.5 mi NE Concordia. Many of these indi-
viduals were teneral, and late instar nymphs were
abundant. Our latest record is 28 November at
1.1 mi SW San Ignacio Ferry. The species was
common then, suggesting a longer life, but no
nymphs were seen. Cantrall failed to find any
Barytettix in early April when he spent two
weeks collecting for Orthoptera at several low-
land localities between San Blas and Mazatlan.
The above evidence suggests that the eggs hatch
in the summer, perhaps by the first summer rains
as we have suggested for paloviridis on the basis
of other evidence, and that the adults die in early
or mid-winter.

DISTRIBUTION.— The northwesternmost rec-
ord for poecilus is 66 mi SE Culiacan, Sinaloa,
the southeasternmost, 5.4 mi E bridge at Plan de
Barrancas, Jalisco.

RECORDS.— Specimens examined: 23200 166 $$, and
many juveniles reared to maturity. Other than material listed in
Table 19, we have examined specimens from the following local-
ities, all in Mexico: NAYARIT: 5 miN Acaponeta on Huajicori
Mine Road (1), 1 mi W Acaponeta (1), 2 mi S Acaponeta, at
Hwy. No. 15 river bridge (4), 3.1 mi SE Acaponeta (4), 4.9 mi
SE Acaponeta on Hwy. No. 15 (2), 8.6 mi SE Acaponeta (7),
9.6 mi SSE Acaponeta (6), 25 mi SE Acaponeta on Hwy. No. 15
(6), 31 rd. mi NE Tepic on Hwy. No. 15 (5), 2.2 mi NE Rio
Santiago (6), 17.9 mi NW Tepic on Hwy. No. 15 (7), 15 mi N
Tepic (AMNH) (2), 14.4 mi NW Tepic (14), 12.7 rd. mi S
Jalisco on Compostela Road (2), 4 mi N Compostela (3);
JALISCO: 20 mi N La Quemada (AMNH) (1), 5.4 mi E bridge
at Plan de Barrancas (6).

Barytettix crassus Scudder
Figs. 6 R, 7 C-D

Barytettix crassus Scudder, 1897, Proc. U.S. Nat. Mus. 20:27-28,
Pl. Il, Fig. 10. [Holotype, @%, San José del Cabo, Lower Califor-
nia. [Territorio de Baja California]; Hebard Collection in the
Academy of Natural Sciences of Philadelphia. |

DIAGNOSIS.— Males of this form may be dis-
tinguished from all other species of Barytettix by
the briefly and broadly notched dorsal aedeagal
valves (Fig. 7 D) or by the strongly and broadly
produced disto-ventral tooth and the weakly en-
larged disto-dorsal portion of the cercus (Fig.
6R). In other species the cercus sometimes
approaches the condition found in crassus, but
the cercus is either strongly incurved and twisted
with the disto-dorsal portion distinctly expanded,
or the insect has a brown ground color, or an
elongate aedeagus indicated in undissected speci-

T. J. COHN AND I. J. CANTRALL

mens by an enlarged pallium. Female crassus can
be distinguished only by a combination of color
and receptaculum seminis characteristics. The
species is apparently always green, and may be
distinguished from paloyviridis and green hum-
phreysii by the short bursa copulatrix and broad
dorso-lateral thick tube. From terminalis it may
be distinguished by the dark carinae and paginal
stripe of the hind femora and by the mid-dorsal
abdominal yellow stripe, and from poecilus by
the combination of those two characteristics and
the red hind tibiae (Table 4, p.26). The uni-
formity of the color characteristics for crassus is
as yet unknown because of poor preservation
and the small number of localities from which
we have material. Our preliminary measurements
based on larger series do not bear out Gurney’s
(1951) suggestion that the hind femora of crassus
are more robust than those of poecilus.

Color comparisons of all Barytettix species
may be found in Table 3, p.25, and the geog-
raphy of the color characteristics in poecilus may
be found in Table 5, p.28. Crassus probably
cannot be recognized by color alone from most
paioviridis or from some humphreysii.

Within the Crassus Group, this species is simi-
lar to poecilus in most of the characters by which
it differs from terminalis. These comparisons are
summarized in Table 4, p. 26, and the phylo-
genetic implications are discussed below in a sec-
tion dealing with the relationships of all species
of Barvtettix.

SPECIES DESCRIPTION.— SUPRA-ANAL_ PLATE:
medio-lateral processes either small spines or tubercles. CERCUS
(Fig. 6 R): weakly incurved; dorsal margin weakly concave;
disto-dorsal portion weakly enlarged; disto-ventral tooth strongly
and broadly produced, acute. AEDEAGUS (Figs. 7 C-D): short
(0.05-0.08 times length of pronotum), wider than long (length
0.67-1.0 times greatest width). DORSAL VALVES (Fig. 7 D):
dorsally with a deep longitudinal furrow on either side of mid-
line, medial portion strongly convex, lateral portions strongly
bent upward, so that valves appear W-shaped in distal view; free
lobes short, moderately broad, apices truncate, blunt rectangu-
late or rounded, medial margins forming a brief narrow medial U,
then flaring to form a wide V, the entire medial excision shallow,
lateral margins barely flaring in dried material, if flattened, distal
expanded portion would appear narrower and shorter (more dis-
tad) than in poecilus and terminalis. VENTRAL VALVES: in
caudal view (Fig. 7 D), strongly tapering; lateral margins obtuse-
angulate in proximal quarter, thence converging but slightly con-
cave to near apices where margins converge sharply to short,
medially directed angulate apices; ventral surface of each valve
with a strong, almost transverse ridge in proximal quarter which
forms the obtuse angle of lateral margins, surface distad of ridge
slightly cupped; in lateral view (Fig. 7 C), gently curved ventrad;
in distal view, each valve with a longitudinal ridge on dorso-
medial margin largely obscured by dorsal valves. RAMUS OF
CINGULUM (Figs. 7 C-D): ventro-distal portion broad, strongly
developed and projecting distad to level of junction of dorsal

CONALCAEINE GRASSHOPPERS: BARYTETTIX

valves with ventral lobes of sheath, thus reducing ventral lobes to
a narrow collar below ventral valves; dorso-laterally smooth,
ventro-laterally with a few strong tubercles or rugose; in caudal
view (Fig. 7 D), ventro-lateral margins sub-parallel, medial mar-
gins moderately divergent. VENTRAL LOBES OF SHEATH
(Figs. 7 C-D): forming a narrow collar beneath ventral valves,
lateral portion extending distad of ramus and visible in side
view, ventral portion in distal view with a brief flange curled
dorsad toward ventral valves. BURSA COPULATRIX AND
THICK TUBE (similar to Figs. 17 A, D): short and broad (width
about twice the length), triangular, moderately pleated, prox-
imally weakly sclerotized laterad, region of orifice of thick tube
weakly sclerotized; thick tube arising dorso-laterally and pro-
ceeding dorso-cephalad, wide at origin and moderately dilated
immediately beyond origin, thence gradually narrowing, length
between proximal bends moderate but much greater than bursa.
COLORATION (based on dried alcoholics from San José del
Cabo, and discolored specimens from San Bartolo and Puerta
Azul Mesa): ground color possibly greenish or olive green (best
preserved specimens show a tinge of green, others are brown),
probably with a small amount of yellow. Head apparently with
narrow post-ocular yellow stripes. Pronotum apparently with
dorso-lateral yellow stripes narrow, not extending onto meta-
zona; black spot on lateral lobes usually restricted to pro- and
mesozona, rarely extending briefly onto metazona, sometimes
extending to caudal margin as a weak wash, spot usually entire,
rarely weakly cut by narrow, irregular cephalic diagonal light
line, that line usually well-developed only ventrad, making black
spot appear narrow cephalad, caudal horizontal yellow line
usually absent, rarely barely and irregularly indicated by spots;
black spot bordered ventrad by yellow, ventral quarter of lobe
and ventral carina concolorous with metazona, probably greenish.
Epimera almost entirely black. Abdomen apparently with mid-
dorsal stripe well-developed, but poorly preserved in most speci-
mens; black dorso-lateral markings well-developed, yellow dorso-
lateral markings barely indicated, but probably poorly preserved;
supra-anal and subgenital plates tinged with reddish. Hind femur
with ventral half of pagina yellow, dorsal half greenish or
brownish; carinae darker gray-green; geniculae tinged with red-
dish, with brownish or reddish lunae. Hind tibia unicolored red,
apparently often fading to pinkish, MEASUREMENTS (in mm):
those of the series studies are summarized in Figures 19 and 20.

HABITAT, ASSOCIATED SPECIES, AND SEASON-
AL OCCURRENCE.— We have no direct informa-
tion about this species other than the locality
and date on the pin labels from two small collec-
tions and a questionable third. The description
and pictures of the vegetation at San José del
Cabo and other lowland localities south of La
Paz in Nelson (1921) suggest thorn forest, much
like that around Mazatlan. Nelson indicates that
Tropical Arid vegetation occupies the lowlands
in this region and suggests that the tropical ele-
ments originated on the coastal plains and foot-
hill slopes on the mainland directly opposite. We
suspect that the collections were made in the
lowlands because it seems likely that the extra
effort of getting into the mountains would have
been noted on the label. No other species of
Barytettix are known from Baja California, but
little collecting has been done there. The dates
of collection range from 1 to 12 November at
San Bartolo and Puerta Azul Mesa.

31

DISTRIBUTION.— Crassus is known from the tip
of Baja California (female allotype of Melano-
plus nitidus Scudder tentatively identified as
Barytettix crassus) northward to San Bartolo
(about 40 mi S La Paz). We have been unable
to locate Puerta Azul Mesa, but suspect that it
is close to San Bartolo from which specimens
are labelled as having been taken the day after
the Puerta Azul Mesa collection.

RECORDS.— Specimens examined: 9c, 12 9 9 3 juve-
niles. TERR. BAJA CALIF.: San José del Cabo (15) (UMMZ,
ANSP®), San Bartolo (7) (USNM), Puerta Azul Mesa (2) (or
Mesa Pta. Azul) (USNM).

Barytettix terminalis’ n. sp.
Figs. 6 Q; 7 E-F

HOLOTYPE.— o&, México, Jalisco, 16.2 mi NE
Barra de Navidad (on Hwy. 80), 2 September
1968 (T. J. Cohn No. 33); University of Michi-
gan Museum of Zoology.

DIAGNOSIS.— Terminalis may be distinguished
from all other species of the genus by the
strongly but briefly out-curved ventral aedeagal
valves (Fig. 7 F), the dorso-lateral knob on the
ramus of the cingulum (Fig. 7 F), and the con-
colorous or narrowly and inconspicuously black-
ened hind femoral carinae. Additional differ-
ences between terminalis and its relatives may be
found in Table 4, p.26 , and the color compari-
sons with all Barytettix species are summarized
inabless sips251

Within the Crassus Group, terminalis stands
apart from the other two members in the above
three characteristics. In other characters it is
about as similar to poecilus as it is to crassus.
These comparisons are summarized in Table 4,
p.26 , and their phylogenetic implications are
discussed below in the section on the relation-
ships of all Barytettix species.

SPECIES DESCRIPTION (in variable characters, condi-
tion in holotype indicated by asterisk preceding that condition), —
SUPRA-ANAL PLATE: medio-lateral processes varying from

6 All the material of this species at the Academy of Natural
Sciences of Philadelphia except the type ( 3d6, 699, 3 juvs.), as
well as a pair from the University of Minnesota, now in the
University of Michigan Museum of Zoology, bear locality labels
similar in size, paper, type and arrangement of words, but which
differ slightly in spelling and punctuation. Only the type label
bears the collector's name, G. Eisen. All have been dried from
liquid and have the same appearance. They are all probably
part of the original series of which Scudder apparently saw
only the one male which he described.

7In reference to the southernmost position of this species
in the distribution of the genus and to the fact that it was
discovered in the terminal stages of this study.

32

swellings to *short, small teeth. CERCUS (Fig. 6 Q): moder-
ately incurved and sometimes slightly twisted; dorsal margin
moderately concave; disto-dorsal portion moderately enlarged
and elongated; disto-ventral tooth *weakly to moderately pro-
duced, acute. AEDEAGUS (Figs. 7 E-F): short (0.07-0.08 times
length of pronotum), wider than long (length 0.75-0.90 times
greatest width). DORSAL VALVES (Fig. 7 F): dorsally with a
deep longitudinal fold or furrow on either side of mid-line,
medial portion strongly convex, lateral portions strongly bent
upward, so that valves appear W-shaped in distal view; free lobes
very short, broad, apices more or less truncate, medial margins
forming a moderately deep, very narrow *V or U, lateral margins
diverging from near base, weakly angulate near middle, thence
rounding to medial margins, distal expanded portion if flattened
would appear generally longer and wider than in crassus and
poecilus. VENTRAL VALVES: in caudal view (Fig. 7 F), lateral
margins strongly rectangulate in proximal quarter, thence strong-
ly concave to short, outward-twisted, blunt acute apices; ventral
surface of each valve with a strong oblique ridge on proximal
quarter which becomes almost flange-like laterad and forms the
angle on lateral margin of valve; surface between and distad of
ridge tilted dorsad laterally to near apex where it becomes more
or less flat; in lateral view (Fig. 7 E) more or less straight; in dis-
tal view, each valve with a longitudinal ridge on dorso-medial
margin largely obscured by dorsal valves.) RAMUS OF CINGU-
LUM (Figs. 7 E-F): ventro-distal portion broad, strongly devel-
oped, and projecting distad to level of junction of dorsal valves
with ventral lobes of sheath, thus reducing ventral lobes to a
narrow collar below ventral valves; dorso-lateral portion with a
prominent knob, ventro-lateral portion rugose, in caudal view
(Fig. 7 F), ventro-lateral and medial margins strongly divergent.
VENTRAL LOBES OF SHEATH (Figs. 7 E-F): forming a
narrow collar beneath ventral valves, lateral portion usually *not
extending distad of ramus and thus usually invisible in side view.
BURSA COPULATRIX AND THICK TUBE (similar to Figs.
17 A, D): short and broad (width 1.3 to about 2 times length),
triangular, proximally with a well-developed sclerite on either
side, distally moderately pleated, region of orifice of thick tube
strongly sclerotized; thick tube arising dorso-laterally and pro-
ceeding dorso-cephalad, wide at origin and moderately dilated
immediately beyond origin, thence gradually narrowing, length
between proximal bends moderate but much greater than bursa.
COLORATION: ground color green with almost no yellow.
Head with narrow post-ocular stripes. Pronotum with dorso-
lateral yellow stripes narrow, not extending onto metazona,
black spot of lateral lobes not extending onto metazona, com-
pletely broken by cephalic diagonal yellow line which is narrow,
caudal horizontal line barely indicated and broken in females,
absent in males, spot narrowly bordered ventrad by yellow,
ventral third of lobe and ventral carina green. Epimera varying
from entirely green, through *green margined with black, to
largely black. Mid-dorsal abdominal stripe *barely indicated,
narrow and not margined in a few specimens; dorso-lateral
yellow markings narrow and inconspicuous, absent in several
specimens; dorso-lateral dark margins narrow; tip of abdomen
tinged with reddish. Hind femur with pagina unicolored green,
carinae usually green, sometimes *narrowly and inconspicuously
black, geniculae tinged with color of hind tibia and with black
lunae. Hind tibia purplish-red (more purple than southern popu-
lations of poecilus). MEASUREMENTS (in mm): holotype
male: length of body, 25.61; length of pronotum, 4.69; length
of tegmen, 3.77; length of fore femur, 4.73; length of hind
femur, 13.53; greatest width of hind femur, 3.75. Measurements
of the series studied are summarized in Figures 19 and 20.

PARATYPES.— Other than the holotype, all
adult specimens examined in this study are desig-
nated as paratypes. Male and female paratypes
are deposited in the University of Michigan
Museum of Zoology, the Academy of Natural

T. J. COHN AND I. J. CANTRALL

Sciences of Philadelphia, the United States Na-
tional Museum, and the Instituto de Biologia,
Universidad Nacional de México.

HABITAT, ASSOCIATED SPECIES, SEASONAL OC-
CURRENCE, AND DISTRIBUTION.— This species
is known from only two localities a few miles
apart in the coastal range of low mountains in
southwestern Jalisco. The vegetation in this
region is characterized as Tropical Subdeciduous
Forest by Rzedowski and McVaugh (1966). The
specimens collected at the type locality were
found at the mouth of a canyon with tall trees,
many vines and heavy undergrowth. They were
most common in the narrow roadside strip of
lush but low weeds and bushes and some were
found in the edge of the forest. At the second
locality, 2.1 miles farther inland, the forest was
lower, but collections were made in the more
open cleared roadside areas with scattered spiny
bushes and variably dense weeds. A number of
other localities have been investigated in Colima
and western Jalisco and southwestern Nayarit at
high as well as low elevations. Neither terminalis
nor any other species of Barytettix has been
found, although poecilus has been found at
higher elevations near Compostela and Tepic,
Nayarit, and is common north of Tepic on the
coastal plain. At the type locality, 16.2 miN
Barra de Navidad, terminalis was found with a
similarly large species of a related new genus.
The earliest record for this species was 31 August
and the latest 6 October. On the latter date only
one male was found and females were not com-
mon, although other orthopterans were abundant.

RECORDS.— Specimens examined: 9 & S18 2 Y, and
several juveniles reared to maturity. JALISCO: 18.3 miN Barra
de Navidad (on Hwy 80) (12); 16.2 mi N Barra de Navidad (on
Hwy 80) (15).

THE PSOLUS GROUP

Most of the Barytettix populations in the mid-
dle portions of the range of the genus possess a
highly distinctive, elongate aedeagus and recepta-
culum seminis, and share several distinctive de-
tails of these structures. We have grouped these
populations together as a species group, consist-
ing of four new taxa, paloviridis, psolus, con-
tilus and nigrofasciatus. Our reasons for dividing
the group into four species are discussed in detail
below.

The distinctive features shared by all four
forms of the group follow:

CONALCAEINE GRASSHOPPERS: BARYTETTIX

AEDEAGUS (Figs. 1 C-D)

1. Shape elongate (at least 0.19 times length of
pronotum), projecting dorsally from genital
cavity into a sac-like enlargement of the pal-
lium (Fig. 15 H), narrow (length at least 1.6
times width, compared to sub-quadrate in
other species groups).

. Ventral valves more or less straight in lateral
and caudal views, strap-shaped in caudal view,
lateral margins subparallel (sometimes slightly
broader across middle), apices in caudal view
acute or rounded acute (never aciculate or
down- or out-curved).

bo

3. Ramus of cingulum with ventro-distal portion
narrow, and ending considerably proximad of
junction of dorsal valves with ventral lobes of
sheath, thus exposing a lengthy portion of
ventral lobes (as in Humphreysii Group).
Ventral lobes of sheath elongated distad of
junction with dorsal valves as a pair of parallel,
approximate, finger-like projections, these
often projecting along their entire length at an
angle from ventral valves, thus giving the
aedeagus a fork-shaped appearance in lateral
view; apices in distal view irregularly oval, the
two parallel or forming an acute angle (flat-
tened and forming an obtuse angle only in
paloviridis from Guamuchil).

. Proximal portion of ventral lobes of sheath
usually keel-shaped, and always provided with
an elongate sclerite on either side (the process
of the ramus of the cingulum) which is readily
visible in dried preparations.

BURSA COPULATRIX (Figs. 16 C, 18)
Elongate and narrow (length at least 1.9 times
width, compared to length equal to or shorter
than width in other species groups), strongly
sclerotized; in cross-section concave dorsally,
convex ventrally; sclerotization divided into
proximal and distal sclerites by a colorless
flexible area, proximal sclerite strongly pleated
dorso-laterally and appearing weak or color-
less along mid-line.

THICK TUBE (Figs. 16 C, 18)

Originating distally or ventrally, and proceed-
ing immediately caudo-ventrad, proximal
bends usually half the length of bursa, and
each bend usually provided with a bladder-like
dilation (in contilus, proximal bends about
one-fourth the length of bursa and usually
without dilations).

ss

nN

The following distinctive color pattern is found

in the three brown species of the group (psolus,
contilus and nigrofasciatus) (Frontispiece):

1. Pronotum with dorso-lateral yellow stripes
usually reaching caudal margin.

2. Lateral lobes of pronotum with dorso-
longitudinal dark band extending to caudal
margin, often lighter and narrower on
metazona.

. Ventral third of lateral lobes yellow, with
black ventral carina.

4. Pagina with sharply defined, jet black stripe
on dorsal half (often lighter and poorly de-
fined in nigrofasciatus).

ios)

In the field, males of this group may be read-
ily identified by the sac-like enlargement of the
pallium (Fig. 15 H), and by the forked appear-
ance of the extruded aedeagus as seen from the
side (Figs. 13 A, C, E). Females of psolus, con-
tilus and nigrofasciatus may be distinguished by
the color pattern described above. Paloviridis is
similar in color and pattern to the Crassus Group
and to the green populations of humphreysii (see
Table 3, p. 25, for a comparison of the color and
color pattern of all species of the genus).

The systematic status of Barytettix popula-
tions with these characteristics poses theoretical
and practical problems, and the solutions which
we offer here require comment on our method-
ology. These populations fall into two color
types (Table 6), both of which have the black
and yellow spots and maculations found in all
species of the genus. In one type, the ground
color of the body is green, and in color and pat-
tern this type is identical to the green color types
found in the Humphreysii and Crassus Groups.
In the other type, the ground color is yellow-
brown with special black and yellow markings
(listed above) not found in other species of the
genus. These two color types overlap broadly in
Sinaloa, and individuals of both are commonly
found in the same habitat if not in the same
bush. Although there is considerable genitalic
variation within each color type, in the many
localities where the two are sympatric, all indi-
viduals possessing any one genitalic type (each
type always differing in several characters from
any other) are invariably of the same color type.
Nor does the green type ever have any of the
special markings of the brown type within this
area of sympatry. It is therefore evident that in
each locality the two color types are reproduc-
tively isolated and are acting as good species. We

peisip sui

payioo)
JOAQU * fayjeird JSOW]Y

suoyrndod wou nos
pure utoy wou ul yUasald

SARA

jeayuan yore Jo UrBIeUL
cIpatl Apau Pur *]PS1O
: C

T. J. COHN AND I. J. CANTRALL

ystdand
10 dUIM-Ppol palojootuy)

YSTUMOLG
10 Uaaid YAP “}UasgV

(A[uo aoryins [RAUAA UO

OrTq Al
PUNT Ud UMOIG) Waasy

1O[OO

(UMOIG

joo punold Udy
UMOIg) Uaatd 4

(sewAensy Irau
{juo yYysY Asaa ing
juasaid) juasgr Aypensy)

(eyeiy
pur sewAensy Iau ATUO
UMOIG) Uaads ATTENS/

siplaaopod

34

peysip BURL] 10 Saayea
Jesjuaa 0} passauddy

SOAJPA [PA}UDA OJUT passasdutt
aq 0} Buueaddy

SOATBA [P1]UDA
0) passaidde 10 Suuely

pay oo}
J9AQU ‘PRISIP JUASIAAIp
‘pewxold jajjeieg

y}00} 10
aqoy & YVIM 410 SuldiaAIp
uay) ‘pewrxold jayjeieg

payioo) 1aAdu “aseq
WO) BUISIIAIP A[UIAM

(snjD}Ia) 9 UI

}UASI1g yuasqy yuasaid) juasqe Ajyensp

(smjuoo UL suoNseatold [esiop
UPA INQ) [eSIOP-O1a}P[ oO pee] —— —- ~-

SUOTRUIQUIOD 10
‘ang 10 ajdind ‘pai
‘parojooig 10 pasojootuy)

auim-anyq to ajdind-an|q
patojoorg to Sysijdand
1o pal parojootuy

pai-anyq 0}
ajdind-anyq parojoorg 10
‘pal 10 anyq parojootuyy

poutjap ATyeam
pue Aes 10 UMO1g 10

= pourop Ajdaeys pue xorg —————

yorlq Ajayajdwo) —

MOTPIA ==

— Lys] pue MOLeU A1aA SALUTZALOS ING “}Uaseld =

— ysIuMoIg — ]

apnimy

SUISJEW 1QUU]

(S{ pue ¢[ “s8ty) uonsod

pasny uo dwiny

saoide jo UONISOg

(pl 314)
SHAIVA 1VSUOd

snoVvdadayv

Vidi GNIH

4d MLS YUVA TVNIOVd

PULIDD [CLJUIA

puny) [esuaA,

vuozrjaul
uo adiys 41eq

WNLONOUd
dO SA8OT TWHALV 1

(sdurew Morjad
pue Yoe]q JO aasnjoxa)

YO1O) GNNOUD
(adatdsyuo14) YOTOD

SNIDLISD{OABIU snosd SNJUUOI

JayIVIEY)

(sonstiajorreyo ,,Aox,, “a[qeztusoda1 ATIsee So} ROTpUul ad A} 208j-pjog)

dNOUD SATOSd AHL NI SOLLSIYALOVUAVHS JO NOLLNATELSIG
9 ATAVL

35

[PS1Op-0}SIG

SNONUNUOD pur [eIPayy

passaidde uayjo

({lyonturny rau Ajuo aygur
ur qe Bunaatu) jayeied Aqyensn
‘passaudwios Apsuons 0) punoy

XdAUOd ATYLAM
Ajaivs yYysens Aypensy

AVIARIUOD [PISIP-OIPaUl
(JIM SaLUTALUOS LO So;pprlu
IEIU PR1I}P] Ud|JOMS 10 “7L]4

a[ppltu JO prwrxoid 10 rau
a}RUIRIeUID APYBIM 10 X9AUOD
Ajpeoig pure Ajuaaq

PEISIP a[QISIA
A|MOLICU 10 d[GISIAU]

(uroRIIN AN
Il Q¢ JO YOU AyUO sonfea

£70 — 610) 8¢'0 — 61:0

CONALCAEINE GRASSHOPPERS: BARYTETTIX

SNONUTJUOD pur [PUTXOId

$______ [P41] UdA-O7SIG

PoplAIp puke [eUUIXxolg

[our ur ye sunoatu Ayjensn pure

XdAUOD APIYBIIS
67 1YsIRAs JsoWpy

Wl

XaAU0d
A[proig pue A[uaaq

S[GISIA ATMOLIEN

oc0

<= uly) ALA O1 [LAG

WysIeNs Ajoues
*xaauod APYsys Aypensy

A}IAROUOD
[EISIP-OIPAUL B ULTAL

(Saaea

[es1op JO saoide je)

J[PY [eIstp ul ayrursiewa
Ayatg ing Ajounsiq

PEISIP a[QISIA A[MOLICN,

[R1]UaA

(SMJDJI0] 9

UL UONRIIp [fetus

duU0 YIM 10) UBLO

Wody SuimoueU ATJenpery

(9110}.99] “9 UL paptatp

pur jeunxoid JO) snonunuos

Pur [risipqns 40 [vysig

[sur ur ye
Sunaow ‘ulyy Aloug

prsyuan
uaq suuvadde 10 xaauoy

DALIUOD ATL BUI
Ajjensn *‘ayqrur,

(Q[Pplul yo prutxoid
Sulpuayxe uoneursirwa

PROIg *SNDJI01
J UT) X9AUOD A[prolg

a14istA A[peorg

€c0~ 610

ILUO Jo uontI07

adrys

481 NOIHL

UONEZNOIA]IS [LISIG

MIMLV TdOO VSund

(g{ 514)
SINIWAS WA1NDV1dd 994

(¢] ‘3I4)
MOLA [PISIP
tsaatde yo adrys

(S{ pur ge] sary)
MOIA OPIS UL UIP
[esuop jo adrys

HIV4AHS
dO AMOT IVYLNAA

OIU|ANS [PIU A,

SUIDIBUL [CIAIeT]

(AVQIA [ESIOP)
awanoleyd
JO SPM [e1OIR]

tunjouoid jo ysuay
DA[PA JO YSUIT

(pI S14)
SHATVA TVULNIA

36

have found four genitalic characteristics which
correlate with the two color types: the position
of the dorsal aedeagal valve apices, the shape of
the ventral lobes of the aedeagal sheath in side
view, the shape of the ventral lobes in distal view,
and the position of the orifice of the thick tube
of the receptaculum seminis (Table 6). Thus the
color types appear to represent two separate lines
of evolution.

Within the green color type we find complex
geographic variation in many different characters
and with much discordance. Part of this pattern
of variation may be ascribed to reproductive
interaction (but not interbreeding) with a closely
related species, and the pattern suggests repro-
ductive continuity throughout and beyond the
region of sympatry. Designation of subspecies
would serve no particular purpose and would
obscure this situation as well as several puzzling
discordant characters. The populations near
Guamiuchil display a few distinctive aedeagal
characters, one of which is not uniform in its
occurrence there. Again subspecific designation
appears to be of no particular value. The nor-
thernmost green populations differ in several
characters from the rest of the species, but are
geographically widely separated from the nearest
green populations to the south. We therefore
cannot determine their status. They may merely
represent the end of several clines. Finally, a few
characters are largely discordant with others in
the range of the species and the pattern can only
be described. Thus it appears to be reasonable to
describe the green color type as a single species
(paloviridis) and to discuss each character and
population variant separately.

The brown color type also displays consider-
able geographic variation but of a different kind.
Genitalic differences are strongly concordant and
sharply circumscribed geographically, but within
any one genitalic type there is usually much geo-
graphic variation in tibial color. Two levels of
similarity can be discerned. The northern and
southernmost populations, although differing
from each other in several details of genitalic
structure, both differ from the middle popula-
tions in having a long aedeagus with the phallo-
treme walls invisible or only narrowly visible
from above, a longer bursa copulatrix, and long
dilated proximal bends of the thick tube (Table
6). The middle populations all share a shorter

T. J. COHN AND I. J. CANTRALL

aedeagus with phallotreme walls broadly visible
from above, a shorter bursa, and short, undilated
bends of the thick tube, but they differ among
themselves in genitalic details (Table 6). Each of
the genitalic types of these middle short aedeagus
populations is circumscribed geographically and,
with one exception, each is morphologically
quite uniform within its area. Only one genitalic
type (fectatus) shows some similarity to the
northern and southern long aedeagus popula-
tions, but is separated at least from its southern
relative by three different middle aedeagal types.
Despite variation of tibial color in some of these
middle populations, it seems useful to designate
each as a subspecies of a short aedeagus species
(contilus) to emphasize their similarities. To the
north as well as to the south, this species ap-
proaches the long aedeagus populations within a
few miles. We have no evidence of introgression
in these areas, and conclude therefore that the
short aedeagus type is probably reproductively
isolated from the two long aedeagus types. De-
spite their overall similarities, each of the latter
is uniform in its genitalic differences from the
other, as well as from the short aedeagus type.
We are therefore designating the northern and
southern long aedeagus types as two species,
respectively psolus and nigrofasciatus.

The relationship between psolus and nigro-
fasciatus poses special problems. A gap of 35
miles separates their presently known ranges.
Their genitalic differences appear to offer no
barrier to the insemination of the females of
either species by males of the other species (see
discussion in the sections, Mechanical Isolation,
and Promising Problems). On the other hand,
there is no genitalic evidence from the geograph-
ically closest collections of the two forms that
they ever have been in contact, although certain
color characters are more similar there. The dif-
ferences between them are sufficiently great and
uniform to suggest that postzygotic incompatibil-
ities already present might form the basis for the
rapid attainment of complete reproductive isola-
tion should the two forms meet. This may
already have happened in the zone where no col-
lections have yet been made.

The four species of the Psolus Group are com-
pared and contrasted in Table 6. The character-
istics in bold face type are the most distinctive
and most easily used and thus may serve as key
characters.

CONALCAEINE GRASSHOPPERS: BARYTETTIX

Barytettix contilus* n. sp.
Figs. 6 FF-JJ; 13 E-F; 14 A-E, I-M, Q-U;
1S FJ; 18 A-C, G-I

HOLOTYPE.— 6, México, Sinaloa, 3 mi NE
Tepuche (13.5 mi N Culiacan), 6 September
1966 (T. J. Cohn No. 40); University of Michi-
gan Museum of Zoology.

DIAGNOSIS.— Males of this species may be dis-
tinguished from all other Barytettix species by
the combination of a relatively elongate aedeagus,
strap-shaped ventral valves without prominent
shoulders (Figs. 14I-M), and a_phallotreme
broadly but briefly exposed distally as seen from
above (in contrast to the condition in cochisei in
which the phallotreme is lengthily exposed by
the widely separated dorsal valves, compare Figs.
14 A-E and 8 F). The elongate aedeagus is indi-
cated in undissected specimens by a sac-like en-
largement of the pallium, as in all members of
the Psolus Group (similar to Fig. 15 H). Females
of contilus may be distinguished from all others
by the combination of an elongate bursa copula-
trix and by small, undilated proximal bends of
the thick tube (one small dilation in c. tectatus)
(Figs. 18 A-C, G-I). Also distinctive in the bursa
is the strong distal pleating and the distal posi-
tion of the distal sclerite close to the orifice of
the thick tube (neither condition holds for
c. tectatus, but that subspecies is unique in the
genus in having a subdistal ventral origin of the
thick tube). Further comparison is made in
Table 6, p.34. Color comparison of all species
of the genus may be found in Table 3, p. 25.
Contilus may be distinguished by color from
psolus and nigrofasciatus only by the geography
of various of the color features. This is discussed
under geographic variation of psolus. Tibial
color comparisons are made in Table 11, p.49,
under that species.

Within the Psolus Group, contilus, psolus and
nigrofasciatus form a distinctive subgroup in
femoral, pronotal and ground color characteris-
tics, in the shape of the ventral lobes of the
aedeagal sheath in side view, the position of the
dorsal aedeagal valves, and the ventral or ventro-
distal origin of the thick tube of the receptacu-
lum seminis (Table 6, p. 34). The apices of the
ventral lobes of the sheath in distal view are

8From the Greek, kontilus, diminutive of long pole, in
facetious allusion to this species’ elongate aedeagus which is,
however, shorter than in other members of the Psolus Group.

w
~

almost identical to those in nigrofasciatus and to
individual variants in psolus. Of all the subspe-
cies of contilus, tectatus shows the greatest simi-
larities to other species of the group. The dorsal
valves of this subspecies are similar to those in
nigrofasciatus in their median hump, flaring lat-
eral portions, and the general shape of their
inner margins. Tectatus displays more detailed
similarities to psolus, as summarized in Table 7.
Contilus shares no distinctive characteristics with
paloviridis, the fourth member of the Psolus
Group. Comparisons of these features are sum-
marized in Table 6 and their phylogenetic impli-
cations are discussed in the section on the rela-
tionships of all Barytettix species.

SPECIES DESCRIPTION (in variable characters, condi-
tion in holotype indicated by asterisk preceding that condition). —
CERCUS (Figs. 6 FF-JJ): dorsal margin almost straight to
*moderately concave; disto-dorsal portion not at all, to *mod-
erately enlarged; disto-ventral tooth moderately produced, acute.
AEDEAGUS (Figs. 13 E-F, 14 A-E, I-M, Q-U): moderately elon-
gate (0.19-0.22 times length of pronotum, 0.21 in holotype),
and narrow (length 1.6-1.9 times mid-dorsal width, 1.8 in type).
DORSAL VALVES: in cephalic view (Figs. 14 A-E), *rounded
from side to side, or roof-like, or flaring laterally; free lobes with
median margins *parallel or weakly diverging proximad, thus
forming a narrow V or *U there, then either diverging smoothly
to latero-ventral acute apices, or each lobe with a dorso-medial
*angle, lobe, tooth or prong, in addition to ventro-lateral acute
apical angulation; fused portion sometimes with a broad, low,
median hump, or *with low humps on free lobes; lateral margins
variable, *convex, sinuate or straight; in lateral view (Fig. 13 E),
ventro-lateral margin variable, straight, *sinuate or convex, con-
cealing or *exposing ventral valve. VENTRAL VALVES: in
cephalic view (Figs. 14 A-E), phallotreme widely open dorsally
beyond dorsal valves, lateral walls visible from above and form-
ing dorsally a strong *median or lateral ridge on each valve; in
lateral view (Fig. 13 E), straight or *with apices curving weakly
dorsad, valves thick *at or immediately proximad of apices of
dorsal valves, the valves then immediately becoming thin, or
“narrowing less rapidly to acute apices but appearing blunter;
ventro-lateral margin obscured or *exposed by dorsal valves; in
caudal view (Figs. 14 I-M), medial margins *separate from near
base or only in distal third, lateral margins variously straight,
concave, or *convex, ventral surface variously concave, *flat, or
keeled. RAMUS OF CINGULUM (similar to Figs. 1 C-D):
ventro-distal portion narrow, ending considerably proximad of
junction of dorsal valves with ventral lobes of sheath, thus expos-
ing a lengthy portion of ventral lobes. VENTRAL LOBES OF
SHEATH (Figs. 13 E-F; 14 Q-U): lengthily developed proximad
of junction with dorsal valves and partly sclerotized there, proxi-
mal portion weakly or not at all flanged; greatly elongated distad
of junction into two approximate finger-like structures which
often project along their entire length at a variable angle from
ventral valves; in lateral view (Fig. 13 E), apical third of lobes
usually *strongly oblique-truncate (appearing bent slightly ven-
trad), sometimes apical third rounded or rounded-truncate,
ventro-proximal portion with an arcuate enlargement provided
with an elongate sclerite (process of ramus of cingulum) on either
side; in distal view (Figs. 15 I-J), thick, apices briefly compressed,
but depressed and twisted outward so that rims form a V, attin-
gent or *separated below, never appressed, ventro-proximal en-
larged portion somewhat keel-shaped. BURSA COPULATRIX
AND THICK TUBE (Figs. 18 A-C, G-I): bursa moderately elon-
gate (length 1.9-2.8 (-3.1?) times width), strongly sclerotized; in
cross section, concave dorsally, convex ventrally; sclerotization

pezjuds-ojsiq
yuasqy
psang fo yssua] €/T

esing sso1oe snonunuos Ajuaeddy
xade je poled
pajyuousid Aqeysiq

T. J. COHN AND I. J. CANTRALL

OE
€€0-610

adOgD WOAf A]GISIA A]PDOAT

aavouod ATTY} SUS]
Warf ‘popaody APFRIM saultpaulog

sey jeurxoid
UT QARDUOD IO *xaAUod “JaTfered-qng

ON
XOAUOD IO JUdBIOAIG

paonpoid Ap YsTs seo]
1B SABMTR So}R[NSUB a]NoP 0} BsN1qO

pesquaa-0}SIpqng
ROM
psang fo yjsua] €/T

papiuip djjuasnddpy
xadp Jp pajvajd JON
pajuawsid A]]DUIXOAd

ORT
610

adogn wodf apqisia A]poo1g

pvysIp aavouos ApyRams
SOUITJOULOS * pRIPIU pa]aay

Jf]Dy [v1SIp 40 ajppnu ul aavIUoD

ON
Jajjpavd-qng

paonpoad JOU ‘ajnjnsuv asniqC

peijuda-0}SIq
BU0IIS

esing Jo ysug ¢/Z

papuip Apuasnddy
xadbp Jp pajpajd JON
pajuawusid A]]DWIXOL

DAO GE WOIJ d[GISIA A[MOLIPN
peisip aavouod

‘pewrxold ep} 10 pajaay
fJDY [PISIP Ul aaDIUQD

SoA

jajjpand-qns

paonpoad JOU ‘ajnjnsUv asnlqC

uBO

uoneyiq

spueg [eurxold Jo yysueT]
(81 3d) dan MOIWL

d}La[Os [PISIG
(gL 314) XIMLV1INdOO VSUNnd

SNOVAGAV JO HLGIM/HLONAT

WONLONOUd AO HLONAT
/AANTVA IVULNAA JO HLONAT

atuarjoTeyd

dOBfINS [PIJUIA,

SUISIPUL [PIO eT]
(pl 3d) SHATVA TVU.LNAA

SOATPA [P1JUDA OJ UT UOTSserdUT]

SUIBIVUW [PII}ET

SUIZIeUI 1QUU]

(pI 31d) SAATVA T¥SYOd

soradsqns s7j1JU09 13y1O

SNJDJIA] SNPUOI

snjosd

Jajobleyy)

SHIOAdSANS SATLLNOD ‘¢ UAHLO AGNV SNTOSd “F OL SALVLOAL SQATILNOD XILLALAYVA AO SALLTAV TINIS

L ATdVL

CONALCAEINE GRASSHOPPERS: BARYTETTIX

divided into proximal and distal portions by a colorless flexible
area, proximal sclerotization strongly pleated dorso-laterally and
appearing weak or colorless along mid-line, distal sclerotization
usually short, apical, variously pigmented, strongly pleated distad
and usually extending across bursa (in c. tectatus, distal scleroti-
zation long, interrupted along mid-line, pigmented proximad); in
lateral view (Figs. 18 A-C), roof of bursa more or less straight
almost to apex; thick tube arising ventrally, narrow at origin (but
wider than in Humphreysii Group), thence narrowing (except in
c. tectatus which has a slight proximal dilation), proceeding
ventro-caudad to first bend, proximal bends about one-third the
length of bursa. COLORATION: ground color light to blackish
brown (moderately dark in holotype) with varying amounts of
yellow. Head with distinct, usually broad yellow post-ocular
stripes (narrow in holotype). Pronotum often *entirely brown
dorsally but usually with narrow to broad yellow dorso-lateral
stripes on pro- and mesozona and often extending to caudal
margin; dorso-longitudinal black band of lateral lobes usually
lighter and narrower on metazona (sometimes barely visible
there), and usually almost broken by a conspicuous diagonal
yellow line on prozona (in San Ignacio colonies of similis, this
band often complete across entire lobe and almost uniform in
width), horizontal yellow line in mesozonal portion of band
rarely present, then weakly indicated; ventral half of lateral lobes
largely yellow (at least on prozona); ventral carina largely black.
Abdomen with mid-dorsal stripe often present but usually weakly
defined (absent in holotype); dorso-lateral yellow markings often
absent or *obscured, narrow when present; dorso-lateral black
spots present, but usually not prominent; tip of abdomen often
tinged with color of hind tibia. Hind femur with pagina yellow
in ventral half, black in dorsal half, the black stripe sharply de-
fined; geniculae largely black, often tinged with color of base of
tibia. Caudal tibia *unicolored, or bicolored, *red, purple or
blue. MEASUREMENTS (in mm): holotype male, see under
contilus contilus, Measurements of the series studied are sum-
marized in Figures 19 and 20.

PARATYPES.— Other than holotypes and juve-
niles, all specimens examined in this study are
designated as paratypes. Male and female para-
types of all subspecies are deposited in the Uni-
versity of Michigan Museum of Zoology, and of
all except c. hiscatus in the Academy of Natural
Sciences of Philadelphia, the United Station Na-
tional Museum, the British Museum (Natural His-
tory), and the Instituto de Biologia, Universidad
Nacional de México.

TAXONOMIC STATUS.— We have erected this
species to include five allopatric populations in
the Culiacan region, all of which possess the dis-
tinctive feature of a widely open phallotreme.
This characteristic is the result of the lateral or
medial position of the dorsal edge of the phallo-
treme wall making the lateral and ventral walls
broadly visible from above. The five populations
also share a relatively short aedeagus and bursa
copulatrix, shorter than other members of the
Psolus Group, thicker and more twisted apices of
the ventral lobe of the sheath, and short, undi-
lated proximal bends of the thick tube. On the
other hand, the populations differ markedly
among themselves in the shape of the dorsal
aedeagal valves, in details of the ventral valves

39

and the bursa, and in hind tibial color. With one
exception (the dorsal valves in the Tepuche
colony), each population is relatively uniform in
all aedeagal characteristics. Each population is
presently known from a single, or a few adjacent
localities, and separated from the next by less
than 21 miles. We have grouped these popula-
tions into a single species to emphasize the
shared distinctive open phallotreme, and short
aedeagus and bursa, and we have designated each
population as a subspecies to emphasize their
allopatry and the uniformity within the popula-
tions of the morphological differences between
them, although we have no evidence of reproduc-
tive compatibility or isolation. The differences
between the subspecies are summarized in
Table 8.

HABITAT, ASSOCIATED SPECIES, AND SEASON-
AL OCCURRENCE.— Each of the subspecies of
contilus has been found in moderately tall, but
thinned thorn forest with a moderate growth of
bushes in the shade and a good growth of weeds
in the sun. In all but one instance (see under
tectatus), each subspecies has been found with
paloviridis in the same habitat and often the
same bush, and the southernmost subspecies
(similis) with poecilus as well. Our earliest rec-
ord for the species is 28 July (similis) when most
adults were teneral and late instar nymphs com-
mon, and the latest, 28 November (similis) when
both males and females were present. Other
early and late records are 30 July (tectatus), 13
November (contilus) and 16 November (tectatus).

DISTRIBUTION AND RECORDS.— This species ranges
from 3.1 mi NE Tepuche (13.6 mi N Culiacan) to 2 mi N San
Ignacio Ferry, Sinaloa. Material examined: 1050%c%, 103 2 9,
and numerous juveniles reared to maturity. For a complete list
of records see Table 19, p. 104.

Barytettix contilus contilus n. ssp.
Figs. 6 FF; 14 A, I, Q;151;18 B, H

HOLOTYPE.— % México, Sinaloa, 3 mi NE
Tepuche (13.5 mi N Culiacan), 6 September
1966 (T. J. Cohn No. 40); University of Michi-
gan Museum of Zoology.

DIAGNOSIS.— See Table 8 in the discussion of
the species.

SUBSPECIES DESCRIPTION (in variable characters, con-
dition in holotype indicated by asterisk preceding that condi-
tion).— CERCUS (Fig. 6 FF): dorsal margin weakly concave,
disto-dorsal portion weakly enlarged, disto-ventral tooth moder-
ately produced. AEDEAGUS (Figs. 14 A, I, Q): narrow in com-
parison with other contilus subspecies (length 1.8 times median
width). DORSAL VALVES: in cephalic view (Fig. 14 A),
rounded from side to side, appressed to ventral valves, some-
times *with a hump on each free lobe, medial margins of free

. J. COHN AND I. J. CANTRALL

40

ystdind 10 pay

porepip you
*[es1U9A-O}SIP UBUC,

A][PISIP
pairajd Ajayerapow
pure pouawsig

pewixoid
JOLY ‘aawouos
Apsuolys ¢/7 [BISIC,

XOAUOD ATAIENUIS

peuitxoid
BUIBIDAUOD *[L1AIE]
‘durys 10 JUNTg

qed 10
[Je ‘pasodxa Ajaieg

a}yeNUIg

auON

passoidde ‘papunoy

paonpoid Apysrys
‘arejnsur asniqg

pai 10 ulm
‘ysydind ‘ysinig

pareyip you

*[P11UDA-O1SIP UBUO

Ajyeastipqns payeayd
A[BUOS Puke PayUdWdtg

A]feqSIp JBL “ayppru ut
aaeouod ‘Aypeuutxod

poaay Apron
ATTeaSIP

xaauod ‘AT[etuIxXold
AARIUOD ATPROIG

pewrxoid surdiaarp
‘jeipow ‘dirys

pasodxq

DABIUOD APYRIAY

PEISIP ‘OM
Aqjeistp ydaoxe

possoidde ‘papunoy

suondatoid
AYT-1OSUL MOLIEN,

ysintd,

paieiip jou
‘[e}UAA-O}SIP UBUC

Ay[eastpqns payeayd

A[suolys pue pay uawisig

AqyeysIp aaeouo0d
Apyrom ‘Alpeuxod

pojeay APYRAA,

latezed

pewixoid Sursiaauod
*[esaiey uN

(Al Sty)
pasodxa A[proig
1ysieNg

dUON

Ayjesquoa Suuey

‘Ayjesiop papunoy

paonpoid
Apysrs ‘ayepnsueryoy

ysinig
parelip

[B11UaA-O]SIPGNS UBUO

Ayfeisip pareayd you
“Aypeutxoid paiuawalg

Aleasip
SABDUOD IO ‘XO9AUOD

Vey la[ppru ul papaay

dARIUO)
pewrxoid
japjeszed uayy “Buls1aauo09

*(parraaja urmsseU [es9,e]
aitjua) [eiaqey UN g

poainosqg,

1YBNS

upIpatu
‘juautwiold ‘apsurg

SuUURL pur oyxIOOY

paonpoid jou ‘ayejnsue
asniqo A][proig

ped

pareyip jou
“[P1]U9A-O1SIP UIBUCQ

Ayyeqsipqns
payeatd Afsuoss
pure payuawsig

QABIUOD
APEaM 10 1k] 4

XOAUOD APYBAM

pewtxoid dursiaatp
‘yerpatu ‘dieys

pasodxq

alenuls
peasip

OM) SAWTIWOg

possaidde *papunoy

ayepnBuejoai 10 aynoy

YOTOD TVIGIL

(FD 'D-V 81
“'s8t4) aqma YOY

(rd
‘O-V BI “sdty)
aquaas [eisiq

XIYLV1NdOO VSand

(WIFI “S8t4)
SOBJINS [PIUIA,

(WI Pl “set4)
uIdiew [esaiey]

(AV pl S8l4)
aspu jesiog

(¢/T fewrxoid)
AMAIA APIS Ul
UIdIBUW [PLa1e 7]

SAATVA TVULNIA

AMAIA APIS Ul
uIgIeW [eLUaA,

diuny [esi0g

apis 0}
apts tuoly adeys

(AV Pl S8ty)
uone[nsue
UIBIelU [eIPay|

SAATWA TVSYOd

symuls

SNIDUDAIIP

SNIDISTY

SNJDJIII

snyuuod

JaJIBIPY)

SaIDddSANS SATIINOD XILLFLAYVA AO NOSTAVdWOO
8 ATaVL

CONALCAEINE GRASSHOPPERS: BARYTETTIX

lobes parallel proximad, forming a narrow U, then each margin
with a rounded rectangulate or *acute angulate tooth, this usu-
ally *shorter, but sometimes as long as lateral apical tooth, never
finger-like, the excision between dorsal and lateral teeth always
shallow, lateral margins of valves diverging slightly distad to near
apex, then weakly converging to apex; in side view, ventro-lateral
margin weakly sinuate, exposing lateral portion of ventral valve,
apex broadly rounded, hardly produced. VENTRAL VALVES:
in cephalic view (Fig. 14 A), dorsal ridge usually sharp, arising
at apex of valve or *near inner margin slightly behind apex, pro-
ceeding proximad along mid-line, the ridges usually diverging
slightly before disappearing under dorsal valves; in side view,
ventro-lateral margins weakly and broadly sinuate; in caudal view
(Fig. 14 I), lateral margins sub-parallel to near middle, then con-
verging gradually to *narrowly blunt or sharp acute apices
(longer than those in c. similis), ventral surface *almost flat or
weakly longitudinally concave. BURSA COPULATRIX (Figs.
18 B, H): distal sclerotization short, apical, pigmented subdis-
tally across bursa, and strongly pleated; thick tube arising ven-
trally, immediately proximad of apex, without dilations. COL-
ORATION: in dried specimens ground color varying from light
to *dark brown, dull colored above. Hind tibia *entirely bright
ted in dark specimens, lighter in specimens with lighter ground
color, almost orange in one light individual (probably faded or
poorly preserved). MEASUREMENTS (in mm): male holotype:
length of body, 32.13; length of pronotum, 6.87; length of teg-
men, 5.20; length of fore femur, 6.25; length of hind femur,
16.91; maximum width hind femur, 4.45. Measurements of the
series studied are summarized in Figures 19 and 20.

DISTRIBUTION.— Known only from a few lo-
calities about 3 mi N of Tepuche (about 13 mi N
Culiacan).

RECORDS.— Specimens examined: 240°, 26 2 2, several
Juveniles reared to maturity. All records are listed in Table 19,
p. 104

Barytettix contilus tectatus? n. ssp.
Figs. 6 GG; 14 B,J, R:18C, I

HOLOTYPE.— 07, México, Sinaloa, 9 mi SE Cul-
iacan, 16 November 1958 (T. J. Cohn No. 288):
University of Michigan Museum of Zoology.

DIAGNOSIS.— See Table 8 in the discussion of
the species, and Table 7 comparing this subspe-
cies with B. psolus.

SUBSPECIES DESCRIPTION (in variable characters, con-
dition in holotype indicated by asterisk preceding that condi-
tion). CERCUS (Fig. 6 GG): dorsal margin barely or *not at
all concave, disto-dorsal portion not enlarged, disto-ventral tooth
moderately produced. AEDEAGUS (Figs. 14 B, J, R): broad in
comparison with other contilus subspecies (length 1.6 times
median width). DORSAL VALVES: in cephalic view (Fig. 14 B),
toof-like, with a broad median hump proximad of free lobes,
medial margins of free lobes parallel or *subparallel a variable
distance proximad, forming a narrow U, then diverging evenly to
lateral apices, angle of medial margin broadly obtuse, lateral mar-
gins of valves almost straight or *weakly flared in middle; in side
view, ventro-lateral margin almost straight, completely obscuring
all but apices of lateral margins of ventral valve, apex of dorsal
valve blunt acute, lying on dorso-lateral surface of ventral valve;
in caudal view (Fig. 14 J), extending far laterad of ventral valves.
VENTRAL VALVES: in cephalic view (Fig. 14 B), lateral mar-

9From the Latin, tectum, roof, in allusion to the roof-like
dorsal valves of the aedeagus.

41

gins strongly elevated proximad of apex, strongly converging
proximad as high, blunt ridges, then proceeding parallel, but not
touching until covered by dorsal valves, parallel portions some-
times *covered by dorsal valves; in lateral view, ventro-lateral
margin invisible except for upturned apical portion, medial mar-
gin visible, broadly convex; in caudal view (Fig. 14 J), lateral
margins broadly and weakly concave to near apex, then converg-
ing to short, sharp acute apices, ventral surface keeled in middle
(broadly V-shaped in cross section), variously flat, *convex or
concave distally. BURSA COPULATRIX (Figs. 18 C, I): distal
sclerotization long, unpigmented along mid-line, pigmented prox-
imad, thickly sclerotized to apex, not pleated distad; thick tube
arising subdistally (more proximad than in other contilus subspe-
cies) and with a moderate dilation in proximal bend region.
COLORATION: ground color light to blackish brown (moder-
ately dark in holotype), with prominent post-ocular and dorsal
pronotal yellow stripes. Hind tibia dull blue, often *with a
slight purplish cast distad. MEASUREMENTS (in mm): holo-
type male: length of body, 34.26; length of pronotum, 6.93;
length of tegmen, 5.19; length of fore femur, 6.22; length of
hind femur, 17.99; maximum width of hind femur, 4.87. Meas-
urements of the series studied are summarized in Figures 19
and 20.

HABITAT AND ASSOCIATED SPECIES.— We have
collected this subspecies only at the base and
summit of Cerro Tule, seven to nine miles south
of Culiacan. In both places it occurs in heavy
weeds and ecotonal plant growth at the edge of
the thorn forest, and more sparsely within the
forest proper. It is found alone at the top, but
occurs sympatrically with paloviridis at the bot-
tom of the mountain. Cerro Tule appears to be
clothed with a rich thorn forest, broken only by
the long narrow road to a micro-wave station at
its summit. The forest has probably been little

disturbed except for limited lumbering.

The presence of tectatus at the summit and
the conspicuous absence there of the widespread
and common paloviridis suggests that tectatus is
more a denizen of bushy perennials and low
growth of forest reproduction than of the more
herbaceous and grassy environs to which palovi-
ridis is partial. This may explain the geographic
restriction of fectatus and the other four contilus
subspecies. These forms may be unable to pene-
trate very far into open and disturbed habitats
because of environmental factors operating on
nymphs and adults, or species requirements for
oviposition, or exclusion from these environs by
interaction with paloviridis. We have not investi-
gated any of these factors. Owing to the prob-
able sparsity of tectatus throughout the forest, it
may be difficult to demonstrate the manner of
its penetration to the summit. We suspect that
movement was made possible by the creation of
localized ecotonal conditions resulting from
major disturbances such as fire or wind.

DISTRIBUTION.— Known only from Cerro

Tule, 7-9 mi SE Culiacan. Adjacent wooded
hills have been inadequately sampled, and only
the edges of flat-land thorn forest have been in-
vestigated between Cerro Tule and 23 mi SE
Culiacan where dicranatus has been found. We
also have a few specimens labelled as from 15 mi
N Culiacan collected by Cohn and E. R. Tink-
ham, which locality we believe to be in error.
The specimens so labelled do not correspond
with the numbers and description of the Bary-
tettix from this locality in Cohn’s field notebook.
Collections made immediately previous to this
one were just south of the Rio Culiacan in the
known range of fectatus, and specimens and field
notes for this locality again do not correspond.
We have failed in several attempts to locate a
tectatus colony 15 miles north of Culiacan or at
other nearby localities.

RECORDS.— Material examined: 12 367 9 9, and anum-

ber of juveniles reared to maturity. Localities are listed in Table
19, p. 104.

Barytettix contilus hiscatus\” n. ssp.

Figs. 6 HH; 13 E-F;14 C, K,$;15 J

HOLOTYPE.— co, México, Sinaloa, 19.9 mi SE
Culiacan on Hwy. 15, 280 ft., 31 August 1961
(I. J. Cantrall and T. J. Cohn No. 59); University
of Michigan Museum of Zoology.

DIAGNOSIS.—
the subspecies.

See Table 8 in the discussion of

SUBSPECIES DESCRIPTION.— CERCUS (Fig. 6 HH):
dorsal margin weakly concave, disto-dorsal portion weakly en-
larged, disto-ventral tooth moderately produced. AEDEAGUS
(Figs. 13 E-F; 14C, K, S): medially narrow in comparison with
other contilus subspecies (length 1.8 times median width). DOR-
SAL VALVES: in cephalic view (Fig. 14 C), rounded from side
to side dorsally, then flaring, without a dorsal hump, medial
margins of free lobes parallel proximad, forming a long narrow U,
then each margin with a slightly produced, rounded, rectangulate
lobe before diverging to lateral apices, lateral margins sub-parallel
in proximal half, then weakly convex in distal half; in side view
(Fig. 13 E), ventro-lateral margin almost straight, broadly expos-
ing lateral margin of ventral valve, apex blunt acute. VENTRAL
VALVES: in cephalic view (Fig. 14 C), dorsal ridge strong, blunt,
arising near, but not on lateral margin of each valve, the ridges
converging slightly proximad, but remaining lateral until ob-
scured by dorsal valves; in side view (Fig. 13 E), ventro-lateral mar-
gin almost straight; in caudal view (Fig. 14 K), lateral margins
sub-parallel to near apex, then briefly rounded mediad to short,
broad, blunt-acute apices; ventral surface weakly concave distad,
weakly keeled proximad (broadly V-shaped in cross-section).
BURSA COPULATRIX as inc. contilus (see Figs. 18 B, H): dis-
tal sclerotization short, apical, pigmented subdistally across
bursa, and strongly pleated; thick tube arising ventrally, immedi-
ately proximad of apex, without dilations. COLORATION:
ground color dark brown with prominent or moderately devel-

10From the Latin, hisco, open, in allusion to the dorsally
open phallotreme.

T. J. COHN AND I. J. CANTRALL

oped post-ocular and dorsal pronotal yellow stripes. Hind tibia
dull blue with a faint purplish cast distad. MEASUREMENTS
(in mm): holotype male: length of body, 34.58; length of pro-
notum, 7.44; length of tegmen, 5.11; length of fore femur, 6.23;
length of hind femur, 18.24; maximum width of hind femur,
4.63. Measurements of the series studied are summarized in Fig-
ures 19 and 20.

DISTRIBUTION AND RECORDS.~— Specimens ex-
amined: 1 6, 1 9. SINALOA: 19.9 mi SE Culia-
can on Hwy. 15. A careful search was made for
this subspecies in 1970, and although paloviridis
was common throughout the region, and other
contilus subspecies were common elsewhere, /his-
catus was not found. Our field notes indicate
that the pair was collected in copula at the edge
of dense thorn forest; now the area is mostly
cleared. The subspecies probably still occurs in
thorn forest which we have not investigated east
and west of the main road. The type locality is
only about three miles northwest of the northern-
most colony of dicranatus with no obvious bar-
rier separating them.

Barytertix contilus discranatus '' n. ssp.
Figs. 6 11; 14 D, L, T

HOLOTYPE.— co, México, Sinaloa, 30 mi S Cul-
iacan on Hwy. 15, 30 August 1965 (T. J. Cohn
No. 85); University of Michigan Museum of
Zoology.

DIAGNOSIS.— See Table 8 in the discussion of

the species.

SUBSPECIES DESCRIPTION (in variable characters,
condition in holotype is indicated by asterisk preceding that
condition).— CERCUS (Fig. 611): dorsal margin weakly to
*moderately concave, disto-dorsal portion weakly to *moder-
ately enlarged, disto-ventral tooth moderately produced.
AEDEAGUS (Figs. 14D, L, T): broad in comparison with
other contilus subspecies (length 1.6 times median width). DOR-
SAL VALVES: in cephalic view (Fig. 14 D), rounded from side
to side, strongly appressed to ventral valves, free lobes each with
a hump, medial margins of lobes lengthily parallel or converging
slightly distad, apices of lobes deeply excised producing two sub-
equal, tapering, blunt, finger-like projections, one dorsal, the
other ventro-lateral, lateral margins flaring slightly and irregularly
distad to ventro-lateral apices; in side view, ventro-lateral margin
broadly and weakly concave, exposing lateral margin of ventral
valve. VENTRAL VALVES: in cephalic view (Fig. 14 D), dorsal
ridges sharp, arising near inner margin of each valve near apex,
diverging slightly proximad, and proceeding along mid-line of
each valve until obscured by dorsal valves; in side view, ventro-
lateral margin broadly sinuate; in caudal view (Fig. 14 L), lateral
margins broadly concave in proximal half, thence convex, some-
times angulately so, distal third gradually converging to moder-
ately long, sharp acute apices, medial margins *gradually diverg-
ing from middle, or distal third concave, ventral surface *moder-

1lProm the Greek, dikranon, pitchfork, in allusion to the
toothed dorsal aedeagal valves.

CONALCAEINE GRASSHOPPERS: BARYTETTIX

ately longitudinally concave in middle, sometimes strongly so,
with lateral margins more strongly bent ventrad than medial
margins, surface more or less flat distad, and rounded from side
to side or *weakly keeled proximad (broadly V-shaped in cross
section). BURSA COPULATRIX: as inc. contilus (see Figs. 18
B, H): distal sclerotization short, apical, pigmented subdistally
across bursa, and strongly pleated; thick tube arising ventrally,
immediately proximad of apex, without dilations. COLORA-
TION: ground color blackish brown, with moderately developed
post-ocular and dorsal pronotal yellow stripes. Hind tibia bluish
purple, *purple, wine-red, or red, or various combinations of
those colors except red. MEASUREMENTS (in mm): holotype
male: length of body, 33.61; length of pronotum, 6.72; length
of tegmen, 5.08; length of fore femur, 6.14; length of hind fe-
mur, 18.5; maximum width of hind femur, 4.84. Measurements
of the series studied are summarized in Figures 19 and 20.

GEOGRAPHIC VARIATIONS.~— In the colony of
this subspecies closest to c. hiscatus, only four
miles away, we find no tendencies toward ae-
deagal conditions in that subspecies. However,
in leg color, the adjacent dicranatus colony not
only averages bluer, but several specimens have
tibial color almost indistinguishable from that of
hiscatus (Table 9). Variation is continuous in
tibial color and the tabulation is but a crude esti-
mate of the trends in the two colonies. The
Sanalona colony has bright red tibiae, much as
in c. contilus found about 13 miles to the north.

DISTRIBUTION.— This subspecies is known
from near Sanalona (east of Culiacdn) and from
23-30 mi S Culiacan. South of the southernmost
locality, suitable habitats have been investigated
without finding dicranatus, although paloviridis
has been found commonly. The northwestern
limits and geographic relationship with the near-
by hiscatus have not been accurately determined,
nor has the region east of the main highway been
investigated except on the Sanalona and Tepuche
roads.

ASSOCIATED SPECIES — As with all other con-
tilus subspecies, dicranatus is found in the same
habitat as paloviridis. Cohn’s impression while
collecting the two was that there was a slight
difference in habitat preference. This would be
interesting to investigate further in view of the
marked shift in the genitalic variation of the
paloviridis colony which is sympatric with dicran-
atus. This is the only paloviridis colony showing
such a shift south of the Rio Culiacan, although
paloviridis occurs with each of the other subspe-
cies of contilus. This subject is fully discussed
under paloviridis.

RECORDS.— Specimens examined: 340°d', 22 99, and
several juveniles reared to maturity. Localities are listed in Table
19, p.104.

A F ae gee)
Barytettix contilus similis '~ n. ssp.
Figs. 6 JJ; 14 E,M, U; 18 A,G

HOLOTYPE.— 3, México, Sinaloa, 66 mi SE
Culiacan, 6 November 1958 (T. J. Cohn No. 258):
University of Michigan Museum of Zoology.

DIAGNOSIS.— See Table 8 in the discussion of
the species.

SUBSPECIES DESCRIPTION (in variable characters,
condition in holotype indicated by asterisk preceding that condi-
tion).— CERUS (Fig. 6 JJ): dorsal margin barely to *weakly
concave; disto-dorsal portion *weakly, sometimes moderately en-
larged; disto-ventral tooth moderately produced. AEDEAGUS
(Figs. 14 E, M, U): medially narrow in comparison with other
contilus subspecies (length 1.9 times median width). DORSAL
VALVES: in cephalic view (Fig. 14 E), rounded from side to
side, appressed to ventral valves, without dorsal hump, medial
margins parallel proximad for varying distance, forming a narrow
U, each margin then forming a *gently rounded or truncate
obtuse angle before continuing to lateral apex, lateral margins
broadly convex in distal half, weakly divergent in proximal half;
in side view, ventro-lateral margin broadly and weakly sinuate,
barely exposing *part or all of lateral margin of ventral valve.
VENTRAL VALVES: in cephalic view (Fig. 14 E), dorsal ridge
*blunt, sometimes sharp, arising *on or near lateral margin sub-
apically, converging slightly proximad and proceeding along mid-
line of each valve until obscured by dorsal valves; in side view,
ventro-lateral margins sinuate; in caudal view (Fig. 14 M), lateral
margins straight or *weakly diverging to near middle, then mar-
gins converging gradually and sinuately to long acute apices
(shorter than in c. contilus), ventral surface strongly longitudi-
nally concave in distal two-thirds, becoming flatter proximad.
BURSA COPULATRIX (Figs. 18 A, G): distal sclerotization
short, apical, pigmented distally across bursa and around thick
tube origin, moderately pleated (less so than in c. contilus); thick
tube arising ventrally immediately proximad of apex, without
dilations. COLORATION: — ground color moderately dark
brown, post-ocular yellow stripes wide, dorsum of pronotum and
abdomen *dull brown in northern colony, or usually with bright,
wide yellow lateral stripes in Elota colony. Hind tibia com-
pletely bright *red or orange-red in northern and San Ignacio
colonies, reddish purple with bluish at proximal end and on
genicular lobes in Elota colony. MEASUREMENTS (in mm):
holotype male: length of body, 31.42; length of pronotum, 6.17;
length of tegmen, 5.24; length of fore femur, 5.63; length of
hind femur, 16.76; maximum width hind femur, 4.24. Measure-
ments of the series studied are summarized in Figures 19 and 20.

GEOGRAPHIC VARIATION.— The two northern
colonies of this subspecies, one collected in
November, the other in August, are strikingly
different from one another in the amount of
yellow on the body and the color of the tibia
(see above description). The duller body color is
similar to that of c. contilus, also collected in
November. However, earlier collections of that
subspecies are similarly dull, and aging can hardly
explain the difference in tibial color. The two
northern colonies are only about eight miles

12From the Latin, similis, like or resembling, in allusion to
the similarity of this subspecies to c. contilus, but separated
from it by three other contilus subspecies.

44

T. J. COHN AND I. J. CANTRALL

TABLE? 9
TIBIAL COLOR VARIATION IN BARYTETTIX CONTILUS DICRANATUS AND B. C. HISCATUS

Wine Purplish Reddish Bluish
Localities Red Red Purple Purple Purple Blue
hiscatus = = = = 5) ="
19.9 mi SE Culiacan
dicranatus - - 6 9 8 1
23.9 mi SE Culiacan
dicranatus 1 l 7 3 ~ -

30 mi SE Culiacan

apart; the precise distance is not known because
of the discrepancies in speedometer readings
upon which a comparison of the distances is
based.

ASSOCIATED SPECIES.— At all localities poe-
cilus and paloviridis were present. Poecilus was
apparently common, and paloviridis rare at the
northernmost locality; the reverse was true at
the Elota locality.

DISTRIBUTION.— Known from a few localities
from 64 to 73 mi SE Culiacan, and from San
Ignacio about 90 mi SE Culiacan. Thorough in-
vestigation of thorn forest elsewhere in this
region has not been made, although we have
numerous collections from ruderal habitats con-
taining only paloviridis or poecilus.

RECORDS.— Specimens examined: 35 &&, 21 29, anda

number of juveniles reared to maturity. Localities are listed in
Table 19, p.104.

Barytettix psolus '* n. sp.
Frontispiece; Figs. 6 EE; 13 C-D; 14 G, O, W;
IS/OSRE13 DJ

HOLOTYPE.— 6, México, Sinaloa, 6 mi W Jests
Maria (19 mi N Culiacan on Presa Humaya
[= Presa Adolpho Lopez Mateos] Road), 11 No-
vember 1958 (T.J. Cohn No. 279); University of
Michigan Museum of Zoology.

DIAGNOSIS.— Psolus is the only species of
Barytettix in which the dorsal aedeagal valves
are impressed into the surface of the ventral
valves, usually creating a low ridge just distad of

13From the Greek, psolos, circumcized, in allusion to the
seemingly cut-back dorsal valves and strangulate ventral valves
of the aedeagus.

the end of the dorsal valves. The brown ground
color, pronotal and femoral color pattern, elon-
gate aedeagus, and elongate bursa copulatrix dis-
tinguish psolus from all but nigrofasciatus (Table
6, p.34). The elongate aedeagus is indicated by
the sac-like enlargement of the pallium, as in all
members of the Psolus Group (similar to Fig.
15 H). Psolus differs from nigrofasciatus in the
male by the shape of the dorsal valves (compare
Figs. 13 C and 14 G with 15 E and 14 H) and by
the sculptured ventral surface of the ventral
valves (compare 13 D and 140 with 14 P) and
in the female by the straight disto-dorsal margin
and longitudinally divided distal sclerite of the
bursa (Fig. 18 D, J). Further comparisons are
made in Table 6.

Color comparison of all species of the genus
may be found in Table 3, p.25.

Psolus may be distinguished by color from
contilus and nigrofasciatus only by the geography
of various of the color features. This is discussed
below under geographic variation, and _ tibial
color comparisons are made in Table 11. In pro-
notal and femoral color pattern, psolus and con-
tilus are identical. Psolus is different in both of
these characters from nigrofasciatus throughout
most of the range of the former, but, except in
tibial color, Cosala specimens of psolus are al-
most identical to San Ignacio individuals of
nigrofasciatus.

Within the Psolus Group as well as within the
genus, psolus, contilus and nigrofasciatus form a
distinctive subgroup in pronotal, femoral and
ground color characteristics, in the shape of the
ventral lobe of the aedeagal sheath in side view,

CONALCAEINE GRASSHOPPERS: BARYTETTIX

the position of the dorsal aedeagal valves, and in
the disto-ventral or ventral origin of the thick
tube of the receptaculum seminis (Table 6, p.34).
Psolus is more similar to nigrofasciatus than it is
to contilus in the longer aedeagus, longer bursa,
dilations of the thick tube, more proximal posi-
tion of the distal sclerite of the bursa, disto-
ventral origin of the thick tube and almost invi-
sible phallotreme walls in dorsal view. The first
three characteristics are also found in paloviridis.
In most southern populations of paloviridis which
are allopatric to psolus the slight exposure of the
phallotreme (compare Figs. 15 Q and R) and the
sculpturing of the ventral surface of the ventral
valves are remarkably similar to psolus. The
bursa of contilus tectatus is similar to that of
psolus (Table 7) except in the position of the
origin of the thick tube (Figs. 18 J and I). These
comparisons are summarized in Table 6, p.34 ),
and their phylogenetic implications are discussed
in a later section on the relationships of all Bary-
tettix species.

SPECIES DESCRIPTION (in variable characters, condi-
tion in holotype is indicated by asterisk preceding that condi-
tion).— CERCUS (Fig. 6 EE): dorsal margin *almost straight to
moderately concave; disto-dorsal portion not at all, to *slightly
enlarged; disto-ventral tooth *moderately to strongly produced,
acute. AEDEAGUS (Figs. 13 C-D; 14 G, O, W;15 O, R): elon-
gate (0.23-0.26 times length of pronotum), and very narrow
(length 2.6-3.0 times mid-dorsal width). DORSAL VALVES:
in cephalic view (Figs. 14 G, 15 R), usually smoothly rounded
from side to side, more or less appressed to ventral valves and
impressed into their surface distally, medial margins of free lobes
proximally parallel and forming a narrow U, then diverging
smoothly at obtuse but not produced angles to latero-ventral
rounded apices, medial margins rarely arcuately diverging from
base, proximal fused portion usually *medially flat, sometimes
broadly and weakly swollen, never with a distinct hump, lateral
margins of valves sub-parallel; in lateral view (Fig. 13 C), ventro-
lateral margin straight for most of length, often *curving slightly
dorsad near apex, usually covering lateral margin of ventral
valve. VENTRAL VALVES: in cephalic view (Figs. 14 G, 15 R),
with dorsal valves appearing impressed into surface, sometimes
with a low swelling just distad of apices of dorsal valves, visible
surface convex, lateral walls of phallotreme narrowly visible at
least distally; in lateral view (Fig. 13 C), lateral margin usually
covered by dorsal valve, but much of ventral surface usually
visible, ventro-lateral margin straight in proximal half, dis-
tinctly but shallowly concave at and just beyond apex of dor-
sal valve, then slightly convex to apex, valves in this view thick
at apex of dorsal valve and tapering rapidly distad; in caudal
view (Figs. 13 D, 14 O), medial margins touching in proximal
third, then diverging gradually and becoming *sub-parallel or
converging slightly distad, lateral margins sinuate, but always
with a distinct shallow concavity at distal third just proximad of
apices of dorsal valves, ventral surface *weakly keeled, convex
across both valves, or flat in proximal third, with a lateral longi-
tudinal swelling near middle third, and usually distinctly concave
in distal quarter, this portion of valve appearing slightly twisted
inward. RAMUS OF CINGULUM (similar to Figs. 1 C-D):
ventro-distal portion narrow, ending considerably proximad of
junction of dorsal valves with ventral lobes of sheath, thus expos-
ing a lengthy portion of ventral lobes. VENTRAL LOBES OF

45

SHEATH (Figs. 13 C-D, 14 W): lengthily developed proximad
of junction with dorsal valves and partly sclerotized there, proxi-
mal portion slightly or not at all flanged; greatly elongated distad
of junction into two approximate, finger-like structures which
often project along their entire length at a variable angle from
ventral valves, rarely appressed to valves; in lateral view (Fig.
13 C), dorsal margin usually slightly convex at least distad, rarely
straight, apices variously round or *truncate, sometimes almost
squarely truncate, ventro-proximal portion with a large arcuate
enlargement provided with an elongate sclerite (process of ramus
of cingulum) on either side; in distal view (Fig. 15 O), apices
*more or less oval, often depressed and twisted, sometimes with
narrow outer rims forming a V, touching or open below as in
contilus, never appressed, ventro-proximal enlarged portion some-
what keel-shaped, often narrow. BURSA COPULATRIX AND
THICK TUBE (Figs. 18 D, J): bursa elongate (length 3.5 times
width), strongly sclerotized; in cross section concave dorsally,
convex ventrally; sclerotization divided into proximal and distal
sclerites by a colorless flexible area, proximal sclerite strongly
pleated dorso-laterally, and appearing weak and colorless along
mid-line, distal sclerite pigmented and pleated proximally, inter-
rupted along midline by a narrow, colorless area, thick sclerotiza-
tion extending to apex of bursa, but not pigmented distally; in
lateral view (Fig. 18 D), roof of bursa more or less straight to
apex; thick tube arising disto-ventrad and proceeding ventro-
caudad to first bend, proximal bends about two-thirds the length
of bursa, each with a relatively small bladder-like dilation.
COLORATION (Frontispiece): ground color light to dark
brown (moderate in holotype), blackish in south, with a variable
amount of yellow. Head with *narrow to wide yellow post-
ocular stripes. Pronotum with dorso-lateral yellow stripes wide,
*narrow or absent, when present sometimes not extending onto
metazona or lighter there; dorso-longitudinal black band of
lateral lobes usually lighter and narrower on metazona (some-
times barely visible there), and usually almost broken by a con-
spicuous diagonal yellow line on prozona (in Cosala specimens,
this band often complete across entire lobe and almost uniform
in width), horizontal yellow line in band in mesozona rarely
present; ventral half of lateral lobes yellow, ventral carina largely
black. Abdomen with mid-dorsal yellow stripe well-defined in
light specimens, *absent in dark ones; dorso-lateral yellow spots
wide and prominent in light specimens, weak or *absent in
darker specimens; dorso-lateral black spots prominent; tip of
abdomen usually concolorous with the rest, sometimes tinged
with color of proximal portion of hind tibiae. Hind femur with
pagina yellow in ventral half, jet black in dorsal half, the black
stripe sharply defined; geniculae with *black or brownish lunae,
and blue ventral half (reddish or orange at Cosala). Hind tibia
proximally blue (reddish at Cosala), distally bluish with a slight
purplish cast in north, grading southward through blue-purple,
purple, *wine, to bright red and orange-red, thus *conspicuously
bicolored in middle populations but color grading imperceptibly
from proximal to distal end of tibia, and unicolored in northern-
most and southernmost colonies. MEASUREMENTS (in mm):
holotype male: length of body, 33.08; length of pronotum,
6.87; length of tegmen, 4.91; length of fore femur, 5.80; length
of hind femur, 17.24; greatest width of hind femur, 4.66. Meas-
urements of the series studied are summarized in Figures 19
and 20.

PARATYPES.— Other than the holotype, all
adult specimens examined in this study are desig-
nated as paratypes. Male and female paratypes
are deposited in the University of Michigan Mu-
seum of Zoology, the Academy of Natural Scien-
ces of Philadelphia, the United States National
Museum, the British Museum (Natural History),
and the Instituto de Biologia, Universidad Naci-
onal de México.

46

GEOGRAPHIC VARIATION.— This species dis-
plays marked geographic variation in tibial color
which is summarized in Table 10. North of Culi-
acan we have two sets of collections, one almost
forming a transect north to Badiraguato, and
another forming a northwest transect to Guamt-
chil, which lies almost due west of Badiraguato
and is separated from it by a small range of moun-
tains. The northwesternmost samples have en-
tirely blue tibiae (under strong light the distal
portions are slightly purplish). The distal por-
tion of the tibiae becomes progressively redder
southeastward, and the samples contain progres-
sively fewer individuals with all blue tibiae, and
more with purple or wine distally. This variation
appears to be clinal geographically, and continu-
ous within any one population. Close to Culia-
can almost all individuals have the tibiae blue
proximally and wine distally. In the northern
transect, on the other hand, all individuals have
the blue-wine tibiae found only much closer to
Culiacan in the northwestern transect. It is inter-
esting to note in this regard that some of the
northern colonies must come very close to, or
are sympatric with c. contilus. These are the
colonies with tibial color most similar to the
orange-red tibiae of c. contilus, which also has a
pronotal and femoral color pattern identical to
that of psolus. These nearby psolus colonies
show no aedeagal similarities to confilus.

The southeasternmost colonies of psolus in
the vicinity of Cosala are separated by a wide gap
from those north of Culiacan, but their orange-
red tibiae may be a continuation of the trend
seen in the northwest transect.

The Cosala colonies are distinctive in two
other color characters. Individuals there are
much darker than those to the north, and some
appear even blackish in life. In addition, some
have a continuous black band across the lateral
lobes of the pronotum and are thus almost indis-
tinguishable from nigrofasciatus. The nearest
colony of that species, 35 miles to the southeast
at San Ignacio, is very similar in pronotal and
femoral color, but has purplish tibiae. Not far
east of the Cosala psolus colonies, contilus sim-
ilis is found with similar tibial and femoral color
but with the dark bar on the lateral pronotal
lobes much lighter on the metazona. The distri-
bution of tibial color of psolus and its relatives is
summarized in Table 11.

In aedeagal characters, there appears to be
more variation within the Cosala colonies than

T. J. COHN AND I. J. CANTRALL

elsewhere. It is difficult to accurately gauge the
extent of this variation because most of the spec-
imens from Cosala are teneral. This situation
might offer additional evidence for the theory of
mechanical reproductive isolation developed
under paloviridis and in a later section of this
paper. The psolus colonies close to Cosala ap-
pear to be above the altitudinal range of palovi-
ridis, and their aedeagal variability might be the
result of reduced selective pressure for uniform-
ity in the absence of its close relative. The two
species do overlap at lower elevations (19.5 mi
SW Cosala), and a mismated pair has been col-
lected there. Unfortunately we have only two
males from that locality, and although they dif-
fer somewhat from one another, many more are
needed to determine the extent of the aedeagal
variability here as a check against that in colonies
closer to Cosala.

HABITAT, ASSOCIATED SPECIES, AND SEASON-
AL OCCURRENCE.— This species has been found
in the weedy-bushy ruderal habitats typical of
many Barytettix species, but usually at the edge
of thorn forest. It has also been collected in
bushy overgrown fields, in thinned thorn forest
among thin weeds and bushes, and, at its south-
ernmost locality, in thinned but uncultivated
tropical deciduous forest. It is apparently absent
from more extensively disturbed habitats, such
as in the weedy-bushy edges of fields in exten-
sively cultivated areas.

Paloviridis has been found at almost all psolus
localities. At 14 mi NW Culiacan, there is some
indication that psolus and paloviridis have slightly
different habitat preferences or are partially ex-
cluding each other from parts of the habitat.
There, psolus was found only within and at the
edge of thinned thorn forest, while paloviridis
was less common there but abundant in the more
weedy open areas. A difference in habitat may
also be indicated by the abundance of psolus in
contrast to the conspicuous absence of paloviridis
in tall but somewhat thinned thorn forest at 6 mi
W Jesus Maria, and in low, dense thorn forest
near Badiraguato. This pattern is not borne out,
however, near Culiacan. Psolus has not been
found at the paloviridis type locality, 2.5 mi NW
Bridge over Rio Culiacan at Culiacan, in heavily
thinned thorn forest with abundant weeds and
bushes but also with thinner shady areas; yet
psolus is not uncommon with paloviridis on the
embankment of the southern end of the railroad
bridge at Culiacan only three miles away in an

CONALCAEINE GRASSHOPPERS: BARYTETTIX

area of heavy weeds and low bushes largely de-
void of trees.

South of Culiacan there appears to be a sharp
difference in altitudinal range between the two
species. On the Cosala road, about 69 mi SE
Culiacan, psolus is common at elevations from
1600 feet around Cosala, to 700 feet at 19.5 mi
SW Cosala. Paloviridis occurs no higher than the
latter locality, where it is sympatric with psolus,
and is common in the nearby lowlands where
psolus is absent. This is the only region in the
range of psolus where elevations above that of
the coastal plain have been investigated.

In the lowlands between the Cosala road and
Culiacan, psolus is apparently completely absent
along the main highway, an area which has been
intensively searched for Barytettix on several
occasions. This situation is unexpected in view
of the abundance of psolus in lowlands immedi-
ately north of Culiacan and in the city itself. It
is possible that the closely related contilus occu-
pies the habitat required by psolus in the area
south of Culiacan as well as that north of Te-
puche. Nowhere does psolus occur sympatrically
with contilus. The known southernmost colonies
of psolus around Cosala are found within about
10 miles of contilus similis. The Culiacan colony
of psolus is located only about seven miles from
the nearest contilus tectatus colony. In the
Tepuche region, psolus has been found only a
mile from the c. contilus colony, and a large
psolus colony was found only five miles to the
southwest. Both species occur on the east side
of the Rio Humaya and there are no obvious
barriers between them. Sufficient collecting was
done at the c. contilus locality to make reason-
ably certain that psolus would have been found
had it been present, but inadequate collecting has
been done between that locality and the nearest
psolus colony. All three contilus subspecies are
found within and at the edge of thorn forest, the
habitat preferred by psolus. This situation is
suggestive of competitive or reproductive exclu-
sion, and may explain the absence of psolus
from the lowlands south of Culiacan.

Tridens has also been found at a number of
psolus localities, where the latter has usually
been more abundant.

Psolus is known from near sea level on the
coastal plain to 1600 feet near Cosala, the high-
est elevation investigated within its range.

Our earliest record for the species is 24 August
at Cosala at which time nymphs were abundant.

47

Late instar nymphs were common on 31 July at
14 mi NW Culiacan, but no adults were found
during an intensive search. The latest record is
13 November at Jesus Maria when the species
was common. No collecting for psolus has been
done earlier or later than the above records.

DISTRIBUTION.— The northwesternmost rec-
ord for psolus is 62 mi NW Culiacan, the north-
easternmost, 5.3 mi NW Badiraguato, and the
southernmost, 19.5 mi SW Cosala, all in Sinaloa.
The species is apparently absent in the lowlands
between the last locality and Culiacan. Inland
the foothills and mountains have been investi-
gated only between Sanalona and Los Mayos
(east of Culiacan) where only contilus dicranatus
and paloviridis were found.

RECORDS.— Specimens examined: 131 && 113 ? 9, and

several juveniles reared to maturity. All localities are listed in
Table 19, p. 104.

Barytettix nigrofasciatus'* n. sp.

Figs. 6 DD; 14 H, P, X; 15 E,K; 18 E, K

HOLOTYPE.— o, México, Sinaloa, 8.3 mi SW
Santa Lucia [on Durango-Mazatlian Highway],
2740 ft., 27 August 1961 (I. J. Cantrall and T. J.
Cohn No. 52); University of Michigan Museum
of Zoology.

DIAGNOSIS.— This species may be distinguished
from other Bary tettix species with the exception
of psolus by the combination of brown ground
color, special pronotal markings (Table 6, p.34),
an elongate aedeagus with the walls of the phal-
lotreme almost invisible in dorsal view (Fig.
14 H), and two bladder-like dilations in the thick
tube of the receptaculum seminis (Figs. 18 E, K).
The elongate aedeagus is indicated in undissected
specimens by a sac-like enlargement of the pal-
lium (similar to Fig. 15 H), as in all members of
the Psolus Group. From psolus, this species may
be distinguished by the conspicuous dorso-medial
bulge on the dorsal aedeagal valves (Fig. 15 E),
and by the undivided distal sclerite and declin-
ing distal third of the roof of the bursa copulatrix
(Figs. 18 E, K). The combination of a complete
dark bar of relatively uniform width across the
lateral lobes of the pronotum, and a gray or
poorly defined paginal stripe, both found near
Concordia (but not at San Ignacio), is unique in
the genus. Further comparisons are made in
Table 6.

14From the Latin, nigro, black, + fascia, ribbon or band,
in reference to the broad dark band on the lateral lobes of the
pronotum.

48

T. J. COHN AND I. J. CANTRALL

TABLE 10

GEOGRAPHIC VARIATION IN THE DISTAL COLOR OF THE HIND TIBIAE IN BARYTETTIX PSOLUS

Western Populations

Eastern Populations

Blue Purple Wine Red Blue Purple Wine Red
62, 60 mi NW Culiacan 10 _ ~ - 5.4 mi N Badiraguato — — 9 _
50 mi NW Culiacan 26 4 — —
42 mi NW Culiacan 2 - ~ —
6 mi W Jesus Maria = 9:2. cg
(23 mi N Culiacan)
20.7 mi NW Culiacan il 5) 2 -
1 mi E San Rafael -- = 16 --
(13 mi N Culiacdn)
15.4-15.2 mi NW 9 17 3) -
Culiacan
*2 mi SW Tepuche — — 7 =
(11 mi N Culiacan)
2 mi W Agua Caliente _ — 7 -
(11 mi N Culiacan)
E] Bario = - 2 --
(7.5 mi N Culiacan)
*4.7 mi SW Tepuche — — = -
(5.8 mi N Culiacan)
*3 mi NE Culiacan - - 12 =
2 mi NW Rio Culiacan — 1 8 —
6.6 mi NE-19.5 mi SW — = - 46
Cosala
*Localities east of the Rio Humaya.
Color comparisons of all species of the genus 11, p.49.

may be found in Table 3, p.25. Nigrofasciatus
may be distinguished by color from contilus and
psolus only by the geography of the various color
features, with the exception of the unusual weak
paginal stripe in the Concordia colonies. This is
discussed under geographic variation of psolus
and tibial color comparisons are made in Table

Within the Psolus Group as well as within the
genus, nigrofasciatus, psolus and contilus form a
distinctive subgroup in pronotal, femoral and
ground color, in the shape of the ventral lobe of
the aedeagal sheath in side view, the position of
the dorsal aedeagal valves, and in the ventral or
disto-ventral origin of the thick tube of the re-

CONALCAEINE GRASSHOPPERS: BARYTETTIX 49

TABLE 11
GEOGRAPHIC VARIATION IN THE DISTAL COLOR OF THE HIND TIBIAE
IN THE BROWN MEMBERS OF THE PSOLUS GROUP
AND IN OTHER SPECIES FOUND WITH THEM

Localities ! psolus 2 contilus nigrofasciatus Other Species 3
6242 mi NW Culiacan Blue, purple — — paloviridis wine
5 mi N Badiraguato Wine - _ -

23-13 mi N Culiacan Wine Red, orange-red = tridens blue
(c. contilus)
21-15 mi NW Culiacan Blue, purple, — ~ paloviridis wine
wine tridens blue
9-3 mi N Culiacan Wine — — paloviridis wine
2 mi NW Culiacan Wine, purple - - paloviridis wine
7-9 mi SE Culiacan — Bluish paloviridis wine
(c. tectatus) tridens blue
20 mi SE Culiacan - Bluish = paloviridis wine
(c. hiscatus)
24-30 mi SE Culiacan — Bluish-purple, ~ paloviridis wine
purple, wine tridens blue
(c. dicranatus)
66-74 mi SE Culiacan = Orange-red, red, = paloviridis wine,
purplish purple
(c. similis) poecilus purple
tridens blue
7 mi NE to 20 mi SW Red - - paloviridis wine
Cosala (about 50 mi tridens blue
SE Culiacan) (Foot-
hills above 700 ft.)
2 mi N to 2 mi SW - Orange-red Purplish* paloviridis wine
San Ignacio (c. similis) poecilus purple
Concordia-Santa Lucia - = Orange poecilus purple

I Mileage rounded to nearest mile.

2Except between 62 and 15 mi NW Culiacdn where many specimens have entirely blue tibiae, psolus is conspicuously
bi-colored north of Culiacan, and in this region is always blue proximad whereas paloviridis is wine proximad.

SP alaviridis: poecilus and tridens are always unicolored in this entire region. The first two species are always bright green,
the third is brown but always with a light brown ventral carina of the lateral lobes of the pronotum (as opposed to the
black ventral carina in all brown members of the Psolus Group).

4 Usually bicolored, darker proximad.

50

ceptaculum seminis. Nigrofasciatus resembles
psolus more than it does contilus in the similarly
elongate aedeagus and bursa, almost invisible
walls of the phallotreme, dilations of the thick
tube, more proximal position of the distal scler-
ite of the bursa, and a similar shape of the distal
portion of the bursa. The first three character-
istics are also found in paloviridis. On the other
hand, the apices of the ventral aedeagal lobes in
distal and often in side view in nigrofasciatus are
almost identical to those in contilus and differ-
ent from the general condition in psolus. Fur-
ther, the nigrofasciatus bursa and dorsal aedeagal
valves are similar to those of contilus tectatus.
These comparisons are summarized in Table 6,
p. 34 , and their phylogenetic implications are
discussed below in a section on the relationships
of the members of the genus Bary tettix.

SPECIES DESCRIPTION (in variable characters, condi-
tion in holotype indicated by asterisk preceding that condi-
tion).— CERCUS (Fig. 6 DD): dorsal margin *straight to barely
concave; disto-dorsal portion *not at all or barely enlarged; disto-
ventral tooth moderately produced, acute. AEDEAGUS (Figs.
14 H, P, X; 15 E): very elongate (0.3 times length of pronotum),
very narrow (length 2.6-*2.7 times mid-dorsal width). DORSAL
VALVES: in cephalic view (Fig. 14 H), proximal half rounded
from side to side and appressed to ventral valves, distal half flar-
ing laterad, medial margins of free lobes evenly diverging and
forming a V, lateral margins more or less straight, subparallel,
apices narrowly rounded and lying on dorso-lateral surface of
ventral valves, a prominent medio-dorsal hump present near
origin of free lobes; in lateral view (Fig. 15 E), ventral margin
straight to weakly sinuate, exposing at least distal half of lateral
edge of ventral valve. VENTRAL VALVES: in cephalic view
(Fig. 14 H), visible surface more or less flat and sloping laterad,
usually with a weak longitudinal ridge near medial margin,
apices of valves slightly twisted, lateral walls of phallotreme nar-
trowly visible from above and weakly convex distad; in lateral
view (Fig. 15 H), ventro-lateral margin straight or *barely curved
ventrad, very briefly and weakly convex distad, valve in this view
moderately thick at apex of dorsal valve and tapering evenly to
apex; in caudal view (Fig. 14 P), medial margins touching or
slightly separated in proximal third of half, then very gradually
diverging and becoming sub-parallel or *slightly convergent dis-
tad, lateral margins broadly and weakly concave in proximal two-
thirds, then gently convex to apices, ventral surface flat, but
distal third of valve gently twisted inward. RAMUS OF CINGU-
LUM (similar to Figs. 1 C-D): ventro-distal portion narrow, end-
ing considerably proximad of junction of dorsal valves with ven-
tral lobes of sheath, thus exposing a lengthy portion of ventral
lobes. VENTRAL LOBES OF SHEATH (Figs. 14 X, 15 E):
lengthily developed proximad of junction with dorsal valves and
partly sclerotized there, with a weak to moderate proximal
flange; greatly elongated distad of junction into two approxi-
mate, finger-like structures which often project along their entire
length at a variable angle from ventral valves, or appear lipped in
side view; in lateral view (Fig. 15 E), dorsal margin almost straight
to *somewhat convex, apices variously rounded or truncate,
sometimes *squarely truncate, ventro-proximal portion with a
large arcuate enlargement provided with an elongate sclerite
(process of ramus of cingulum) on either side; in distal view (Fig.
15 K), apices more or less oval, or thick and trigonal, or *de-
pressed and twisted so that the narrow rims form a V which is
*open or closed below, apices never appressed, ventro-proximal

T. J. COHN AND I. J. CANTRALL

enlarged portion somewhat keel-shaped. BURSA COPULATRIX
AND THICK TUBE (Figs. 18 E, K): bursa elongate (length 3.5
times width), strongly sclerotized; in cross section, concave dor-
sally, convex ventrally; sclerotization divided into proximal and
distal sclerites by a colorless flexible area, proximal sclerite
strongly pleated dorso-laterally and appearing weak and colorless
along mid-line, distal sclerite pigmented and pleated proximally,
unbroken across bursa, pigmented area rounded distally, distal
third of bursa colorless and apparently only weakly sclerotized;
in lateral view (Fig. 18 E), roof of bursa gently declining in distal
third; thick tube arising disto-ventrad and proceeding ventro-
caudad to first bend, proximal bends about two-thirds the length
of bursa, each with a large bladder-like dilation. COLORATION
(one female noted below is of a peculiar coloration and is not
included in this description): ground color light to *dark brown,
never blackish, usually with much yellow. Head with narrow to
broad yellow post-ocular stripes (moderate in holotype). Pro-
notum dorso-medially brown, median carina lighter; dorso-lateral
yellow stripes broad and always extending to caudal margin;
lateral lobes with dorso-longitudinal black band extending to
caudal and cephalic margins, slightly narrower on prozona, usu-
ally slightly lighter but broader on metazona (in lightest speci-
men band is narrower there), no cephalic diagonal or caudad
horizontal yellow lines in black band (cephalic line complete in
one specimen), the band therefore appearing continuous and of
more or less even width in most specimens; ventral half of lobes
yellow, ventral carina largely black. Abdomen with mid-dorsal
yellow stripe always present, but *brownish and not well-defined
in dark specimens, lateral yellow and black spots present but
variable in intensity (weak in holotype); tip of abdomen con-
colorous with rest of abdomen, in some females vaguely marked
with orange below. Hind femur with pagina yellow in ventral
half, *gray, brown or blackish in dorsal half, this dark stripe
often weakly defined or *light and inconspicuous; lunae of
geniculae black and brown, ventral half of geniculae usually
*orange or blue, sometimes light brown; tibia *unicolored
orange (Concordia-Santa Lucia), or bicolored, bluish proximad,
purple or wine distad (San Ignacio) MEASUREMENTS (in
mm): holotype male: length of body, 32.10; length of pro-
notum, 6.28; length of tegmen, 4.78; length of fore femur, 5.63;
length of hind femur, 17.19; maximum width of hind femur,
4.29. Measurements of the series studied are summarized in
Figures 19 and 20.

PARATYPES.— Other than the holotype, all
adult specimens examined in this study are des-
ignated as paratypes. Male and female paratypes
are deposited in the University of Michigan Mu-
seum of Zoology, the Academy of Natural Sci-
ences of Philadelphia, the United States National
Museum, the British Museum (Natural History),
and the Instituto de Biologia of the Universidad
Nacional de México.

GEOGRAPHIC VARIATION.— Specimens from
the only two regions from which this species is
known differ significantly in color pattern. In
the south, near Concordia, individuals possess
orange-red legs, weak and poorly defined grayish
paginal stripes, and a complete dorso-longitudinal
dark bar across the lateral lobes of the pronotum.
Fifty miles to the north, specimens taken near
San Ignacio possess purplish and often bicolored
tibiae, a jet black and well-defined paginal stripe,
and a sometimes interrupted pronotal dark bar.
It is interesting to note that where nigrofasciatus

CONALCAEINE GRASSHOPPERS: BARYTETTIX

occurs within a few miles of its close relative,
contilus similis at San Ignacio, the leg color is
different from both this contilus as well as from
the not far distant psolus near Cosala. On the
other hand, the leg color is similar to another
relative, paloviridis, with which nigrofasciatus is
sympatric at San Ignacio. However, paloviridis
has a different ground color and pronotal and
femoral color pattern.

INDIVIDUAL VARIATION.— A_ single female
from 3.6 mi NE Santa Lucia, at a higher eleva-
tion and different habitat than our other collec-
tions, is distinctively different in color and color
pattern from other females. Although the bursa
and the tibial color are identical to other females,
this specimen completely lacks all the dark and
light markings characteristic of nigrofasciatus
as well as the genus. The specimen is badly
rotted, but preservation of at least some of the
darker markings would have been expected.
Instead, only a vague indication of the dark bar
on the lateral lobes of the pronotum is visible.

In males, the apex of the ventral lobe of the
sheath is variable. In some it is distinctively dif-
ferent (squarely truncate in side view) from
psolus and contilus, but in others it is indistin-
guishable from them. In Concordia-Santa Lucia
specimens there is considerable variation in the
paginal stripe which ranges from almost jet black
to a grayish wash.

HABITAT, ASSOCIATED SPECIES AND SEASON-
AL OCCURRENCE.— All but one specimen were
collected along the roadside in thin weeds and
bushes or in cut-over heavier growth next to
thorn forest. It ranges from the edge of the
foothills of the Sierra Madre at 400 feet, well
into the mountains at 2740 feet along the
Mazatlan-Durango Highway. The single exception
is the female described above which was found
at 4880 feet at 3.6 mi NE Santa Lucia not far
east of the next easternmost locality along the
same highway. This specimen was found in oak
woods (pines were nearby) among acacias and
much succulent undergrowth on a rocky slope
with an undescribed relative of Conalcaea. At
three of the other four nigrofasciatus localities,
poecilus was also common. The fourth locality,
8.3 mi SW Santa Lucia, was not carefully
searched, but we suspect poecilus was also pres-
ent because it occurs nearby at higher and lower
elevations. Nigrofasciatus has not been found at
higher elevations in the oak and pine forests at

51

several localities sampled farther east along this
road, nor does it occur on the coastal plain
which has been intensively investigated to the
north and south, and where poecilus is abundant.
Nigrofasciatus also occurs with poecilus as well
as paloviridis in the low foothills of the Sierra
Madre at 1.1 mi SW San Ignacio Ferry. Here in
the previously cleared area next to the road,
there was a good growth of weeds and bushes
next to tall but thinned thorn forest. It was the
commonest species here, but surprisingly was
absent only a few miles to the north across the
Rio Piaxtla where contilus similis was the com-
monest species along with poecilus and palovi-
ridis in a similar habitat but one with more trees
and bushes. We have investigated the foothills
of the Sierra Madre elsewhere only around Co-
sala: no nigrofasciatus was found at a number
of localities investigated between 700 and 1600
feet, although psolus was common, and palovi-
ridis and tridens were also present at some
localities.

Our earliest record for this species is 26 July
at 11.5 mi NE Concordia. Only a few adults
were found here and at nearby localities during
intensive collecting. Most of the adults were
teneral and several late instar nymphs were
found. Our latest record is 28 November at 1.1
mi SE San Ignacio Ferry, when males and females
were common, but less so than in August, and
no nymphs were found.

DISTRIBUTION.— Known only from localities
between Concordia and Santa Lucia, and at San
Ignacio, 50 miles to the north, all in Sinaloa.

RECORDS.— Specimens examined: 24 & 6 28 2 2, and

several juveniles reared to maturity. For localities see Table 19,
p.104.

Barytettix paloviridis Den sp.
Figs. 1 C-D; 2 B-C; 6 Y-CC; 13 A-B; 14 F, N, V;
15 A-D, F-H, L-N, P-Q; 16 C-D; 18 F, L

HOLOTYPE.— <& México, Sinaloa, 2.5 mi NW
of Bridge in Culiacan [over the Rio Culiacan],
29-30 August 1961 (I. J. Cantrall and T. J. Cohn
No. 56); University of Michigan Museum of
Zoology.

DIAGNOSIS.— This species may be distin-
guished from all other Barytettix by the combi-
nation of an elongate aedeagus, an elongate bursa

1SFrom the Latin, palus, pole or stake, + viridis, green, in
allusion to the long aedeagus and green body color, and to our
facetious reference to this species as “green flagpole” in the
early stages of this study.

n
io]

copulatrix and a usually green ground color.
The elongate aedeagus is indicated in undissected
specimens by a sac-like enlargement of the pal-
lium, as in all members of the Psolus Group (Fig.
15 H). A few brown individuals are found near
Guaymas and Altata, and may be distinguished
from the three brown species of the Psolus
Group by the lack of distinctive pronotal mark-
ings (see Table 6, p. 34), and from brown mem-
bers of the Humphreysii Group by the elongate
aedeagus or bursa. Further comparisons are
made in Table 6.

Color comparison of all species of the genus
may be found in Table 3, p.25, and the geo-
graphical distribution of tibial color in Sinaloa
may be found under psolus in Table 11, p.49
On the basis of color alone, green paloviridis is
indistinguishable from green humphreysii with
reddish legs occurring in southern Sonora, brown
paloviridis is indistinguishable from humphreysii
in the Guaymas area, and purple-legged palovi-
ridis cannot be distinguished from poecilus south
of Culiacan (see under geographic variation of
paloviridis, p.53 ). Poecilus is red-legged as in
paloviridis but only far south of the range of
paloviridis. Male paloviridis can always be recog-
nized from humphreysii and poecilus by the en-
larged pallium.

Compared with other members of the Psolus
Group, paloviridis stands apart in color, color
pattern, dorsal aedeagal valves, ventral lobes of
the aedeagal sheath, disto-dorsal origin of the
thick tube of the receptaculum seminis, and
more distal position of the distal sclerite of the
bursa copulatrix (Table 6). Surprisingly the
ventral valves of the aedeagus of paloviridis pop-
ulations south of the area occupied sympatric-
ally with psolus are similar to psolus in several
details (see under the discussion of geographic
variation in paloviridis). North and south of the
area of sympatry, the apices of the ventral ae-
deagal lobes of paloviridis tend to be less com-
pressed and appressed and thus are similar to the
general condition found in psolus (Figs. 15 M-O).
The dorsal valves of paloviridis are more similar
in general shape to nigrofasciatus (Figs. 14 F, H;
15 A-E), and northern and southern colonies
possess a median hump on the valves somewhat
as in that species (Figs. 15 C, E). Paloviridis,
psolus and nigrofasciatus all have similarly elon-
gate aedeagi, bursae, and proximal bends of the
thick tube, similar dilations of the thick tube,
and a generally similar, almost invisible phallo-

T. J. COHN AND I. J. CANTRALL

treme, in contrast to the conditions in contilus.
Comparisons are summarized in Table 6, and
phylogenetic implications will be discussed in a
later section dealing with the relationships of all
Bary tettix.

SPECIES DESCRIPTION (in variable characters, condi-

tion in holotype is indicated by asterisk preceding that condi-
tion).— CERCUS (Figs. 6 Y-CC): dorsal margin *almost straight
to weakly concave; disto-dorsal portion not at all to slightly en-
larged (but entire cercus may be very broad); disto-ventral tooth
weakly to *moderately produced, acute. AEDEAGUS (Figs. 1
C-D, 13 A-B): elongate (0.19-0.38 times length of pronotum,

width, 2.6 in holotype). DORSAL VALVES: in cephalic view
(Figs. 14 F; 15 F-G, P-Q), convex dorsally but flaring laterally
for proximal two-thirds, becoming more or less flat distad,
medial margins of free lobes diverging only slightly distad, form-
ing a narrow V, lateral margins weakly convex, valves tapering
to short (Guaymas colonies) or *long, rounded apices which lie
along medial margins of dorsal surface of ventral valves; fused
portion usually with two low, longitudinal ridges (barely indi-
cated in holotype), ending in one or two humps distad (northern
and southern populations), or *without humps (middle popula-
tions); in lateral view (Figs. 13 A, 15 A-D), ventral margin con-
vex or sinuate (Guaymas colonies), or more or less straight or
*weakly sinuate, usually exposing ventral valve for most of its
length. VENTRAL VALVES: in cephalic view (Figs. 14 F,
15 P-Q), *almost completely obscured by dorsal valves, or lenth-
ily exposed, visible surface *more or less flat, or sloping laterad,
or weakly convex, lateral walls of phallotreme *invisible (middle
colonies, Fig. 15 P), or barely visible apically where walls flare
slightly (Fig. 15 Q); in lateral view (Figs. 13 A, 15 A-D), latero-
ventral margin straight (Fig. 13 A), weakly convex (Fig. 15 A)
or *with distal third gently curved (middle colonies) (Fig. 15 D),
sometimes *with a strong median convexity (Figs. 15 B, D), usu-
ally completely exposed by dorsal valve, distal quarter *uni-
formly thin (middle colonies), or distal sixth tapering rapidly to
apex (compare Figs. 15 D and 13 A); in caudal view (Figs. 13 B,
14.N, 15 F-G), touching to or beyond middle, then medial
margins diverging usually more strongly than in psolus, but be-
coming parallel distad, often *slightly convergent near apices, or
divergent to apices (Guaymas colonies), lateral margins more or
less straight and parallel (Guaymas and Guamichil colonies,
Figs. 15 F-G), or *parallel in proximal half, weakly emarginate
near middle, and weakly convex or *sinuate in distal half (Figs.
13 B, 14.N), rarely distinctly emarginate in distal half, ventral
surface with distal half or third convex and appearing half
twisted outward (Guaymas and Guamichil colonies), or *more
or less flat and broad (middle colonies, Fig. 14 N) or somewhat
flat, narrow, twisted slightly inward and often slightly concave
(southern colonies, Fig. 13 B), proximal half or two-thirds of
surface *more or less flat, or weakly or moderately concave
longitudinally, often with a *lateral swelling near middle which
is sometimes *prominent (Fig. 14 N) and sometimes continued
to base as a ridge. RAMUS OF CINGULUM (Figs. 1 C-D):
ventro-distal portion narrow, ending considerably proximad of
junction of dorsal valves with ventral lobes of sheath, thus ex-
posing a lengthy portion of ventral lobes. VENTRAL LOBES
OF SHEATH (Figs. 13 A-B; 15 A-D): lengthily developed prox-
imad of junction of dorsal valves and partly sclerotized there,
proximal portion strongly but briefly reflexed or flanged; greatly
elongated distad of junction into two approximate, finger-like
structures which usually project along their entire length at a
variable angle from ventral valves (appressed to valves only in
Guamichil samples); in lateral view (Figs. 13 A, 15 A-D), dorsal
margin usually straight, rarely *weakly convex (Fig. 15 D),
apices usually roundly truncate, disto-dorsal angle usually rela-
tively sharp, disto-ventral angle usually rounded, ventro-proximal
portion weakly enlarged, *straight or weakly arcuate below,

CONALCAEINE GRASSHOPPERS: BARYTETTIX

provided with an elongate sclerite (process of ramus of cingulum)
on either side; in distal view (Figs. 15 L-N), apices usually some-
what to strongly compressed (moderately compressed in holo-
type), parallel, often appressed, rarely round (apices meeting
at a broad angle only in Guamuchil colonies, Fig. 15 L, but
lengthily flat, unlike contilus and its relatives), proximal en-
largement variable, often *keel-shaped, often somewhat rounded
(flat only in Guamitchil colonies). BURSA COPULATRIX AND
THICK TUBE (Figs. 2 B-C; 15 C; 18 F,L): bursa elongate
(length 3.0 times width), strongly sclerotized; in cross section
(Fig. 18 L), concave dorsally, convex ventrally; sclerotization
divided into proximal and distal sclerites by a colorless flexible
area, proximal sclerite strongly pleated dorso-laterally and ap-
pearing weak and colorless along mid-line, distal sclerite proxi-
mally uncolored but pleated, distal portion pigmented, extend-
ing across bursa, truncate or notched distally, distal fifth of
bursa only weakly sclerotized; in lateral view (Fig. 18 F), roof of
bursa more or less concave in distal third; thick tube arising
disto-dorsad, gently recurved and proceeding ventro-caudad to
first bend, proximal bends about two-thirds the length of bursa,
each with a very large bladder-like dilation. COLORATION:
ground color green, abdomen and caudal femur often yellowish
(ground color brown only in individuals from Guaymas region
and Altata). Head with yellow post-ocular stripes *sub-obsolete
or missing. Pronotum sometimes with a *dorsal, narrow, dark,
median stripe on pro- and mesozona; dorso-lateral yellow stripe
narrow and restricted to pro- and mesozona (broad and extend-
ing to caudal margin only in some individuals near Guaymas),
sometimes absent (probably resulting from poor preservation);
black spot of lateral lobes restricted to pro- and mesozona (ex-
tending briefly onto metazona and continued as a dark wash to
caudal margin only in some individuals near Guaymas), spot
usually completely divided by cephalic diagonal yellow line
(barely capped by black in holotype), caudal horizontal yellow
line in spot often *incomplete, sometimes absent, spot narrowly
bordered ventrad by yellow, ventral quarter and ventral carina
green (ventral surface of carina black in a few individuals in
Guaymas colonies). Abdomen always with *well-defined mid-
dorsal yellow stripe, but sometimes obsolete on metanotum and
first abdominal segment; dorso-lateral yellow and black spots
usually present, yellow spots often obscured apparently by poor
preservation; supra-anal plate, cerci, and dorsum of subgenital
plate usually *reddish (purplish in individuals with purple tibiae),
probably always so in life, but fading in poorly preserved speci-
mens (in Guaymas colonies, tip of abdomen yellowish or with
weak trace of red in green individuals, or light brown in brown
individuals). Hind femur with pagina *green to yellow, *uni-
colorous, or with a variously developed green or brown dorsal
stripe, darkest in southern colonies; carinae *dark green to black;
geniculae reddish (purplish in individuals with purple tibiae) or
in Guaymas colonies, yellowish brown, lunae *black or brown.
Caudal tibia dull *red-wine over most of range, purplish in some
southern individuals, *unicolored, or weakly and inconspicu-
ously bicolored. MEASUREMENTS (in mm): holotype male:
length of body, 31.76; length of pronotum, 6.55; length of teg-
men, 5.55; length of fore femur, 5.79; length of hind femur,
16.63; maximum width of hind femur, 4.54. Measurements of
the series studied are summarized in Figures 19 and 20.

PARATYPES.— All adult specimens examined
in this study except the holotype, are designated
as paratypes. Male and female paratypes are de-
posited in the University of Michigan Museum of
Zoology, the Academy of Natural Sciences of
Philadelphia, the United States National Museum,
the British Museum (Natural History), and the
Instituto de Biologia, Universidad Nacional de
México.

n
Ww

GEOGRAPHIC VARIATION.— Paloviridis displays
considerable geographic variation, a situation not
unexpected in view of its extensive range through
arid and subhumid habitats and its sympatry
with several other species of Baryvtettix. The
extreme northern populations around Guaymas,
and those west of Guamuchil in the middle of
the range, each show distinctive features sharply
different from the other populations of the spe-
cies. The significance of these variants is un-
known, although each of these populations occu-
pies a habitat somewhat different from the rest
of the species. In the more southerly popula-
tions, several variant characteristics are found
which resemble nearby or sympatric populations
of the distantly related poecilus. Finally, and of
particular interest, the pattern of variation in
certain other characters over the whole range of
the species strongly suggests reproductive inter-
action with closely related species. This pattern
has been used as evidence of mechanical repro-
ductive isolation (Cantrall and Cohn, 1972). Each
of these four situations is discussed in detail
below.

The Guaymas Populations and Dwarfism.—
The northernmost colonies of paloviridis (18 mi
N and 8 to 21 mi SE Guaymas) are characterized
by their small size, dull green or brown ground
color, and by distinctive ventral aedeagal valves.
These are thick and short, with acute apices,
parallel lateral margins, evenly divergent inner
margins, and strongly tilted, convex ventral sur-
faces of the distal halves of the valves (Figs. 15
A-B). This construction is somewhat similar
only to the paloviridis west of Guamuchil and
near Altata, both localities far to the south.
However, those individuals have longer, more
slender and thinner valves with sinuate lateral
margins (Figs. 15 C, F, G). In other aedeagal
characters, the Guaymas colonies generally match
those of adjacent colonies, whereas the Altata
and Guamuchil colonies are similar to southern
paloviridis, or possess their own unique features
(Guamuchil ventral lobes, see below).

The Guaymas colonies, sampled in three dif-
ferent years, average smaller than all other palo-
viridis, and include the smallest individuals in the
species (Table 12). Only the colony at Altata,
far to the south, approaches the size conditions
of the Guaymas populations. Most other sam-
ples do not overlap the Guaymas range and
average well over 6 mm in pronotal length. Inter-

54

T. J. COHN AND I. J. CANTRALL

TABLE 12
DWARF SIZE POPULATIONS IN BARYTETTIX PALOVIRIDIS

Locality and date
of collection

Guaymas (1966, 1968, 1970)

40 mi S Navajoa (160 mi
SE Guaymas) (1968)

12.2 mi N Los Mochis (206
mi SE Guaymas) (1968)

10.9 mi W Guamuchil (272
mi SE Guaymas) (1968)

Altata — 15 mi SW Navolato
(350 mi SE Guaymas) (1968)

estingly, the few other colonies which do over-
lap the range, and average below 6 mm, are also
the geographically closest ones (Navojoa-
Guamuchil) (see Table 19, p. 104). All of these
were collected in 1968. In other samples col-
lected near the Navojoa, Los Mochis and Guamu-
chil localities in preceding years, the minimum
and average sizes are much larger, thus indicating
size variation from year to year.

The dull green color of the Guaymas individ-
uals is found nowhere else within the range of
the species, and the brownish individuals are
similar only to some members of the Altata
colony. In the two darkest brown Guaymas indi-
viduals, the black spot of the lateral pronotal
lobes extends as a dark wash to the caudal mar-
gin of the pronotum. They also possess a very
narrow black line on the ventral surface of the
ventral carina of the lateral lobes. These color
characteristics seem to be a development toward,
or a reduction from the condition in psolus and
its brown relatives, but there are no other simi-
larities to those species in color or aedeagal
structures.

The size and color of the Guaymas popula-
tions are matched by the nearby humphrevsii
colony at 22.7 mi SE Guaymas. This leads us to
suspect that small size and brown or dull green
color are responses to environmental conditions
of a flat area almost at sea level where the soil is

Number of Male pronotal length (mm)
specimens Range Mean
22 4.2-6.0 Sel
6 5.1-6.2 5.8
i 5.1-6.0 5.4
16 5.5-6.5 59
15 4.8-5.5 5.4

probably saline. The only other collections of
Barytettix made at this elevation have been
among the dunes near Altata where paloviridis
is again small and often brown. Although the
flats near Guaymas seem very different from the
dunes at Altata, there are probably several envi-
ronmental factors common to the two areas
which are different from nearby higher and often
hilly localities where both species are larger and
greener. It should be noted, however, that the
northernmost of the Guaymas colonies (18 mi N
Guaymas) is located inland and at a somewhat
higher elevation, but possesses the same charac-
teristics as those on the coast (except for brown
individuals), whereas in a nearby humphreysii
colony found just behind the beach sand 4 mi W
Guaymas, all individuals are large and bright
blue-green.

The Guamuchil Populations and Primitive Ae-
deagal Characters.— The two colonies of palovi-
ridis found west of Guamuchil possess distinctive
ventral lobes of the aedeagal sheath. When seen
in distal view, the valves usually form an open V
(Fig. 15 L) in contrast to the parallel and often
appressed lobes in all other colonies of this spe-
cies (Figs. 15 M-N). This feature is somewhat
variable in our material: in 15 individuals, the
lobes form an open V as described and figured;
in four, the lobes form a much narrower V (one
of them closely resembling the condition in

CONALCAEINE GRASSHOPPERS: BAR YTETTIX

nigrofasciatus, Fig. 15 K); and in one individual,
the lobes are fully as twisted toward one another
as in the rest of the species. This last individual
has the apices of the lobes rounded in distal
view, a relatively uncommon condition in palo-
viridis. In the nearest paloviridis colony to the
south, 5 mi SE Guamuchil, one male out of the
six collected shows a strong tendency to the
open V condition, and a second has a narrower
V condition. Several males in this colony also
have the apices rounded. There is a similar but
weaker tendency toward the Guamuchil condi-
tion in the next northern colony at 12.2 mi N
Los Mochis.

The ventral lobes lie flat against the ventral
valves in most of the Guamuchil specimens. In
this position they strongly resemble the ventral
lobes in humphreysii (compare Fig. 15 G with
Figs. 9B and D), and thus may represent the
ancestral condition in the Psolus Group. How-
ever, the lobes in the Guamuchil colonies are al-
ready more elongate than in humphreysii and
when they project away from the ventral valves
they do so along their entire length as is typical
in the Psolus Group, rather than briefly or only
at the tip as in humphreysii.

The Guamuchil colonies also possess what at
first glance appears to be unusual ventral valves.
They are narrow and have diverging inner mar-
gins, acute apices, and strongly convex ventro-
distal surfaces (Figs. 15 F-G). However, this
type of valve is almost identical to that in the
Altata colonies 80 miles to the south, which dif-
fer only in having the valves slightly less acute
and less divergent. Nor are the Guamuchil valves
very different from those in the next northern
colonies (between Los Mochis and Navojoa)
where they are somewhat wider, less acute,
slightly incurved, and sometimes flatter ventrally.
Still farther north, the Guaymas populations dis-
play such features of the Guamuchil type as the
diverging inner margins, convex ventro-distal sur-
face, and acute apices, but the valves there are
much thicker, shorter (Figs. 15 A-B), more diver-
gent, and have straight rather than sinuate lateral
margins. Thus the Guamuchil valves are mark-
edly different only from the very broad, blunt,
incurved and ventrally flat valves possessed by
the populations north of Culiacan (Figs. 14 F, N;
15 D). Even in this case, the geographically in-
termediate colony at 5 mi SE Guamuchil (62 mi
NW Culiacan) is intermediate in valve morphol-
ogy. Southeast of Culiacan, the valves are again

55

somewhat similar to those at Guamuchil (Figs.
13 A-B). There may thus be an irregular north-
south cline in valve characteristics interrupted
by the strikingly different valves in the middle
between Guamuchil and Culiacan. There is good
evidence that the unusual type found between
those two cities is a response to special circum-
stances in that region, a topic which will be dis-
cussed below.

It is tempting to speculate on the possibility
that the Guamuchil colonies have retained sev-
eral of the more primitive aedeagal features of
the species. Under this hypothesis it is difficult
to explain why the populations to the north and
south both possess the same derivative condi-
tions. Possibly the Guamuchil colonies represent
a recent intrusion of amore eastern, foothill pop-
ulation. Much more collecting needs to be done
east of the main highway as well as along the
coast in this region before the problem can be
adequately explored.

In view of the mosaic of aedeagal similarities
and intermediates, we do not think that any pur-
pose would be served in designating the Guamu-
chil population as a subspecies.

Leg Color and Interaction with Poecilus.— Cer-
tain color characteristics of southern populations
of paloviridis are of interest because of their re-
semblence to the unrelated poecilus which occurs
in the same region. All paloviridis populations
between Culiacan and Mazatlan possess tibiae
which are distinctly but slightly more purple
than those in the north. Most individuals have
wine-red tibiae, but purple tibiae are found as
individual variants in several populations from
37 to 60 mi SE Culiacan, north of the range of
poecilus. The purple of these individuals is in-
distinguishable from the purple found uniformly
in all northern populations of poecilus. How-
ever, in the zone of the overlap of the two spe-
cies between 66 mi SE Culiacan and 20 mi N
Mazatlan, all paloviridis individuals possess red-
dish or wine-colored tibiae. On the other hand,
it is in this general area that paloviridis possesses
a dark paginal stripe on the hind femur as is
found in all northern colonies of poecilus. The
presence of the stripe is variable and it is usually
weak in colonies of paloviridis between Culiacan
and 30 mi SE Culiacan, but in all colonies farther
south it is always present and is as dark as in
poecilus. This pattern of geographic variation
may also be the result of interaction with psolus
and contilus (see discussion below). Thus, there

56

are paloviridis individuals very close to the zone
of sympatry with poecilus which are indistin-
guishable from the latter species in external ap-
pearance. No evidence of interbreeding has been
found in the genitalic characters of these popula-
tions, thus eliminating the possibility of the ori-
gin of these poecilus characteristics in paloviridis
through introgression. The tibial color differ-
ences in the sympatric populations may be a case
of character divergence resulting from copulation
errors, even in the absence of hybridization.
However, initial cage studies suggest that males
do not differentiate between differently colored
females. Furthermore, the colonies between
Navojoa and Los Mochis also have wine or pur-
plish tibiae in an area of overlap with similarly
colored, green humphreysii. North and south of
this area paloviridis usually has redder tibiae.
This problem is discussed further in the section
on Promising Problems.

Aedeagal Character Reinforcement and Me-
chanical Reproductive Isolation.— The pattern
of geographic variation of most biological signifi-
cance is that of the marked shift in the variation
of many characters at the Rio Culiacan and Rio
San Lorenzo. This is coupled in a few of the
same characters with a similarity of populations
south of Culiacan with those northwest of
Guamuchil but separated by distinctive popula-
tions in between. This pattern of variation en-
compasses all geographically variable characters
studied with the exception of features unique to
the Guaymas and Guamuchil colonies, and tibial
color.

The geographic variation in eight aedeagal
characters over the whole range of the species is
illustrated in Figures 3 and 4, which should be
used in conjunction with the analysis presented
below. The localities in Figure 3 have been ar-
ranged from northwest to southeast following
the main west coast highway along which most
of the collections have been made. In Figure 4
we have grouped localities to the east and west
of the main highway in the vicinity of Culiacan,
and have repeated the histograms for localities
immediately north and south of Culiacan. These
eastern and western populations vary somewhat
from the above pattern and are discussed sepa-
rately. To shorten both figures, some samples
have been grouped where differences between
them seemed minimal. No attempt has been
made to make the grouped samples equivalent
or equidistant. In most of the qualitative char-

T. J. COHN AND L. J. CANTRALL

acters the intermediate conditions were subjec-
tively determined and should not be considered
definitive. Bars have been blackened to empha-
size the pattern. Where a histogram includes
three conditions, the leftmost bar has been
blackened; if there were four or five conditions
graphed, the leftmost two bars were blackened.

Parenthetically it should be noted that the
Guaymas and Guamuchil populations closely
follow the pattern of adjacent ones in at least
five of the eight characters graphed. This sug-
gests that they have not been isolated from the
rest of the species and thus are not worthy of
subspecific designation.

Every character illustrated in Figure 3 shows
a marked shift in variation at the Rio Culiacan
or within six miles south of it. This shift often
occurs Over a distance of 3.5 miles in large col-
lections on either side of the river and as close
to it as the habitat permits. The condition in
two of the characters (length of exposed ventral
valve and depth of apex of ventral valve) south
of the Rio Culiacan, and particularly south of
the Rio San Lorenzo is similar to that northwest
of Guamuchil. Two other characters (dorsal
valve hump, and ventral valve width) show the
same pattern of similarity, except for the differ-
ent Guaymas populations. Between 23.9 and 30
mi SE Culiacan, there is a clear reversion in all
but two characters (dorsal valve hump, ventral
valve width) to the condition of the colonies
north of Culiacan. These six characters shift
even more sharply across the Rio San Lorenzo
in a distance of less than three miles.

It is precisely in the region between Culiacan
and Guamuchil that the closely related psolus
occurs and is sharply limited to the south by the
Rio Culiacan (it is present south of the river only
at the south end of the railroad bridge where it
may be a recent arrival). Between 23.9 and 30
mi SE Culiacan another related species, contilus
dicranatus, is found. Both psolus and c. dicrana-
tus occur commonly in the same general habitat
with paloviridis, and are sometimes found roost-
ing at night in the same bush with paloviridis.
We thus suspect that the distinctive paloviridis
characteristics in these areas of sympatry and the
sharp shift in variation immediately beyond them
are the result of interaction of paloviridis with
its congeners.

In the area of sympatry, the condition of all
but two of the characters (dorsal valve hump,
and width of ventral valves) in paloviridis is very

CONALCAEINE GRASSHOPPERS: BARYTETTIX

n

1

L-EXPOSED! DEPTH APEX HUMP ON CURVATURE MAX. WIDTH — APPRESSION EXPOSURE OF THICKNESS OF PRESENCE OF
V.VALVE V.VALVE D.VALVE V. VALVE _V.VALVES ——OF_ APICES OF PHALLOTREME APICES OF OTHER
T.V.VALVE (SIDE VIEW) (SIDE VIEW) —_L.V.VALVE V. LOBES Vv. LOBES SPECIES
(DISTAL VIEW)
wy
rr 2 ta
wa wa
< = z < B <
SxXSan 3 Sz aa Ais as a
§ 2 Ezo od35 Ze a> 3 z
Tupi 28s Sz pratt ae Bien zac
Airs. ann aes 22 2328 ae 235 Ses
Siehasrties Bee 235 BE hid nas nae
LOCALITIES z 3
(DISTANCE
IN MILES)
GUAYMAS REGION
(362-323*) __ 0s i i a] i a
NAVOJOA-LOS MOCHIS |
es le LL dk ff 2 ew
W. GUAMOCHIL ( b
(76-69) al IL L. al. | a
=
62-42 NW. CULIACAN | ' PSOLUS
| J = Sat
30-15 NW. CULIACAN gl | | = LL @ eb | JU ees
n av 27 35 28 35
CULIACANCITO- is 6 § L j #
2 NW. CULIACAN PSOLUS
(13-2*) |
e ere
fo Culiacan
28 30 29 32 30
ees J 4 J | fh / L b
_o_ I
6-9 SE. CULIACAN | | alll
i | zal ae
17 SE. CULIACAN —=_1 =_ 15} I at] —) my =o
19.9 SE. CULIACAN =) | 8) al walls SU) a sia
CONTILUS
24.9 SE. CULIACAN ss! a3 1) a) —_=_— at a. =—-— DICRANATUS
: F | — Mm q CONTILUS
30 SE. CULIACAN ci fa 5 | i. i DICRANATUS
2
Rio San Lorenzo
—
42.7- 43 SE
obs fale STH] aE a5) = 5) 8) sal
26 20
Nn im - |
35.2- 39.9 SE y
CULIACAN go
Li is om!) —met_|
> (nal
nat rach ia a rH ay
CULIACAN lal i
¥ = — J PSOLUS
19.5 SW. COSALA Je) ial —D) 5) se os lik. eau
(60*)
4 |
64.6- 2U N. MAZATLAN
cutie) al | iJ ot ac J

Figure 3. Aedeagal variation in Bary tettix paloviridis throughout its range. Bars have been blackened only to emphasize the pattern:
the black bars generally represent conditions of the populations immediately north of Culiacan, and the white bars, conditions found
south of Culiacdan, In histograms with three conditions, only the left bar has been blackened; in histograms with four or five condi-

tions, the left two bars have been blackened.

*miles from Culiacan
the greater the ratio, the more valve is exposed (Fig. 15 Q in contrast to Fig. 15 P)
“The greater the ratio the wider the valves
Scale: height of bar represents numbers of specimens measured or scored. The lowest bar represents one specimen, the highest un-
broken bar, 25. A number over a broken bar indicates numbers over 25.

58

different from psolus and c. dicranatus, whereas
to the south and often to the north of that area
the paloviridis conditions are similar to those of
the other two species (Figs. 15 P-R). The degree
of exposure of the ventral valves and of the phal-
lotreme can be seen to be directly associated
with the difference between paloviridis females
and psolus and c. dicranatus females in the posi-
tion of the aperture of the thick tube of the re-
ceptaculum seminis (Figs. 18 B, H, D, J, F, L).
The aedeagal condition of paloviridis where it is
sympatric with those species would interfere
with the entry of the spermatophore tube into
the more ventral opening of the thick tube of
either psolus or c. dicranatus. The function of
the other aedeagal structures is not so obvious,
but each could well be associated with the proper
seating of the aedeagus in the bursa of the fe-
male. Thus conditions in the paloviridis aedeagus
which are different from those of psolus or con-
tilus might prevent its proper seating in the bursa
of those species and thus prevent insemination.
The function of these aedeagal structures and
their role in interspecific crosses are discussed
in greater detail in the section on Mechanical
Isolation.

A ninth character which we have not included
in Figure 3 because of difficulty in precision of
analysis is the sculpturing of the ventral surface
of the ventral valves. This character appears to
follow the same pattern described above, at least
southeast of Guamuchil. In the area of sym-
patry with psolus and c. dicranatus the valve
surface is relatively smooth and more or less flat
from side to side, but south of this area it is ex-
cavate distad and longitudinally ridged mesad as
in psolus (see under Species Description). There
seems to be an area of transition and variability
south of Culiacan, but in the southern end of the
range of the species the valves in ventral view
are generally similar to those of psolus (except
for their only slightly sinuate lateral margins and
the strong mesal and proximal excavation of the
ventral surface). It is difficult to envision a func-
tion for such sculpturing or the lack thereof. A
detailed study of the interior surface of the fe-
male bursa may reveal such a function. North-
west of Guamtchil, and near Altata, the ventral
surface of the valve is usually convex, a condi-
tion not found in any of the more southerly
populations.

There exist several apparent exceptions to this
pattern of variation and similarity which might

T. J. COHN AND L. J. CONTRALL

be used to test our hypothesis of reinforcement
of aedeagal characters in response to the pres-
ence of closely related species. One exception
to the pattern of a strong shift in variation at the
Rio Culiacan can be seen in the characteristics of
the populations east of the Rio Humaya, illus-
trated in Figure 4. These populations lie to the
east of the main northwest-southeast transect,
and their variation can be compared in the fig-
ure with that in nearby or adjacent populations
in the main transect.

The east side Rio Humaya populations (NE
Culiacan) agree closely with those on the west
side (NW Culiacan) in six characters, all of which
show a strong shift at the Rio Culiacan, but they
are more similar to populations southeast of
Culiacan in two others, the dorsal valve hump
and the ventral valve width. These are the same
two characters which fail to change between
23.9 and 30 mi SE Culiacan where other charac-
ters revert to conditions found northwest of
Culiacan (Fig. 3). Thus the colonies east of the
Humaya are almost identical in all features to
those at 23.9-30 mi SE Culiacan. It is precisely
in these two areas that restricted colonies of the
related species contilus are found, c. dicranatus
in the southern area and the nominate subspe-
cies east of the Humaya. Thus, the difference in
the characteristics of the paloviridis colonies on
either side of the Humaya is probably the result
of a different response to the presence of psolus
than to the presence of contilus. However,
psolus is also found on the east side of the
Humaya, south of Tepuche. It is unfortunate
that our samples are too small to determine
reliably whether paloviridis south of Tepuche
and sympatric with psolus is different from sam-
ples north of Tepuche where it is sympatric
with contilus. The trend in variation is in the
expected direction in each area. However, the
ventral valve width in the south Tepuche sample
(sympatric with psolus) is even more different
from northwest Culiacan paloviridis also sym-
patric with psolus, than from the north Tepuche
sample allopatric to psolus. It is possible that all
paloviridis populations east of the Humaya are
more strongly influenced by gene flow from
those south of the Rio Culiacan. The eastern ex-
tension of the Rio Culiacan — the Rio Sana-
lona — is smaller, and there is much less inten-
sive agriculture east of the city. In contrast, the
northwest Culiacan populations are separated
from those south of the Rio Culiacan by the

CONALCAEINE GRASSHOPPERS: BARYTETTIX

, EXPOSED? JEPTH \ PEX HUMP ON

V ALVE - VALVE V.VALVE

|
THIN

INT

THICK

(SIDE VIEW) (SIDE VIEW)

INVISIBLE

PSOLUS

LOCALITIES
(DISTANCE
IN MILES)
32
CULIACANCITO- _ e
2 NW. CULIACAN
(13-2)
WEST SIDE RIO HUMAYA
13-3. conmachn fj | z. _=_,
y
Rio Humaya
EAST SIDE RIO HUMAYA S| E 7]
E. TECORITO-
N. TEPUCHE (13*) = | = ant
S. TEPUCHE-
3 NE. CULIACAN
(9-3*) | om |
4 . a
Rio Culiacan
28 30 29° 32
CULIACAN-7 E., 1.5 SE. N .

PSOLUS

tr 2

6-9 SE. CULIACAN

>

:

ALTATA REGION

=
|

(35 WSW. CULIACAN)

=

]

-
:

=—

Figure 4. Aedeagal variation in Barytettix paloviridis in the Culiacan region of Sinaloa. For explanation see legend for Fig. 3, p. 57.

larger river, the city itself, and the intensively
farmed, irrigated land to the west and southwest.
It is also possible that psolus may have only re-
cently arrived on the east side of the Humaya
(as suggested by Cohn, 1965, for the katydid
Neobarrettia hakippah) or that it is present there
in small numbers.

Another apparent exception to the general
pattern is the lack of change in those southern
paloviridis populations which are in contact with
other contilus subspecies, tectatus, hiscatus and
similis (at 7-9, 19.9, and 66-74 mi SE Culiacan
respectively) or with southernmost psolus (at
19.5 mi SW Cosala) (Fig. 3). This seems to con-
tradict our hypothesis of the cause of the marked
shift in the characteristics of paloviridis where it
is sympatric with c. dicranatus or with psolus.
However, the bursae of similis and tectatus
(Figs. 18 A, G, C, I) are somewhat different from
that of dicranatus (almost identical to Figs. 18 B,
H) in the position of the aperture of the thick
tube, and it seems reasonable that neither of
these two could be inseminated by any palovi-

ridis. This point is further discussed in the sec-
tion on Mechanical Isolation. As for hiscatus,
that subspecies appears to be rare, at least along
the main highway where we have done all of our
sampling within its apparent range. The area
was specifically searched for hiscatus in 1970 by
Cohn who found abundant paloviridis but only
one questionable female hiscatus. It is possible
that any influence of hiscatus on paloviridis is
swamped in this area. The effect of the nomi-
nate subspecies on paloviridis is almost exactly
the same as that of dicranatus and has been dis-
cussed above.

We have no such easy interpretation for the
lack of change in paloviridis where it is sympatric
with the southernmost colony of psolus. The
ranges of the two may not overlap much beyond
the point where they were found together at
19.5 mi SW Cosala. The road was sampled at
several points between that locality and 5.6 mi
NE Cosala at higher elevations. Psolus was gen-
erally common and paloviridis absent. On the
other hand, paloviridis was abundant and psolus

60

absent at lower elevations along the main high-
way only about 10 miles to the west, where the
highway has been intensively worked for many
miles to the north and south. Thus any ten-
dency toward a change in characteristics in the
sympatric paloviridis populations may be quickly
swamped. It is also possible that one or the
other species has only recently arrived in the
area, for it is near or at the edge of the range of
both. That there is selective pressure for diver-
gence of one from the other at this locality is
indicated by our observation here of one of the
few cases of interspecific copulation (¢ psolus X
2 paloviridis ).

The marked shift in the southernmost palovi-
ridis population toward a closed phallotreme is
puzzling, but may be the result of interaction
with nearby nigrofasciatus, the bursa of which is
very similar to that of psolus.

Another exception is seen in the Altata popu-
lation (Fig. 4) which is found just behind the
coast west-southwest of Culiacan, but north of
the Rio Culican. Despite its geographic position
this population is similar to those southeast of
Culiacan in four characters (depth and curvature
of ventral valves, appression and thickness of
ventral lobes), shows strong southeast influence
in three characters (exposure of ventral valves,
dorsal valve hump, ventral valve width) and is
similar to the northwest Culiacan populations in
only one character (phallotreme). The region is
clearly deltaic and is drained by many streams.
The Rio Culiacan flows west of Culiacan to Nav-
olato where it makes a sharp bend to the south
and empties to the sea well south of Altata. It is
interesting to note that another river which
drains the Navolato area flows northwest to
enter the sea north of Altata. This is called the
Rio Viejo, and we suspect that this may have
been an older drainage channel for the Rio Cul-
iacan. Therefore, it is likely that the Altata pop-
ulation may have been in freer contact with
southern colonies in the recent past. Further-
more, psolus has not been found in the area (but
it has not been searched carefully), and there
may thus be little selective pressure for the main-
tenance of northwest Culiacan characteristics.
In the one character displaying the strongest
northern influence (exposure of the phallotreme)
there is more variation in the Altata population
than is found in any of the Culiacan-Guamuchil
colonies.

T. J. COHN AND L. J. CONTRALL

Two final exceptions are more difficult to
interpret. The pattern of similarity where palo-
viridis and psolus are allopatric and of difference
where they are sympatric is reversed in two char-
acters. The dorsal valve hump is absent and the
width of the ventral valve is greater where the
species are sympatric. Although it is not at all
clear to us why these two features should be
similar to the condition found in psolus, the
change occurs concomitantly with the change in
the other six features of the aedeagus at the Rio
Culiacan. It seems certain that the morphology
of the aedeagus has much to do with the proper
placement of the spermatophore tube. Our pres-
ent lack of information regarding the function of
the hump and the relative width of the ventral
valves, as they relate to the bursa copulatrix,
prevents an explanation as to how the shape of
these two features can contribute to mechanical
isolation. We anticipate that a study of the re-
lationship of the male and female copulatory
structures in interspecific mating pairs will go far
in elucidating this problem. Presently, we do
know that the ventral valves are shaped quite
differently where the two species are sympatric
although the measurements of width are similar.
In psolus the widest part of the valves is at the
distal quarter, with the narrowest portion imme-
diately proximad, whereas in paloviridis the
widest portion is at the distal third or more
proximad, and the valves narrow more gradually
proximad. This difference may create a differ-
ence in the seating of the aedeagus in the bursa.

A non-genitalic character which seems to fol-
low the same pattern of variation and similarity
is femoral color. This character has not been in-
cluded in the figures because of difficulty in pre-
cisely differentiating the conditions. South of
the Rio San Lorenzo (32 mi SE Culiacan, see
Fig. 3), all paloviridis possess a well-marked dark
paginal stripe of dark green, brownish or gray-
black. As in the genitalic characters there is a
very sharp break across the Rio San Lorenzo; all
specimens immediately south of the river (32.5-
33 mi SE Culiacan) have a dark stripe, while all
specimens immediately north of the river (30-
23.9 mi SE Culiacan) in the range of contilus
dicranatus lack the stripe. Farther north, in the
range of psolus, occasional paloviridis specimens
have a well-marked stripe, some have a weak
stripe, but most have none at all. Both contilus
and psolus always have a sharply defined black
paginal stripe. In paloviridis, the absence of such

CONALCAEINE GRASSHOPPERS: BARYTETTIX

a stripe north of the Rio San Lorenzo and its
presence south of that river might be another in-
stance of character reinforcement similar to that
in genitalia discussed above, although our pre-
liminary findings suggest that some species of
Barytettix, at least, do not depend on color or
color pattern to identify their mates. As in the
genitalic pattern, there is no reinforcement in
paginal striping where paloviridis is sympatric
with contilus similis at 66 and 74 mi SE Culia-
can, or at 19.5 mi SW Cosala where it is sym-
patric with psolus.

The Barrier Effect of Rivers.— These data on
variation in aedeagal and femoral color charac-
teristics also demonstrate how effectively rivers
act as barriers to the distribution of Barytettix
species. We have hypothesized above that the
conditions at 30 mi SE Culiacan in six of the
eight characters graphed in Figure 3 are the re-
sult of interaction with c. dicranatus. That sub-
species has not been found south of the Rio San
Lorenzo, where each of the six characters shows
a marked shift in condition. Not only does the
colony on the south bank (32.7-33 mi SE Culia-
can) possess the alternative condition in each to
those on the north bank, but the characters are
not even more variable than in the colonies to
the south. There seems to be minimal gene flow
across this river. This is not true in the palovi-
ridis colonies north of the northernmost dicran-
atus at 23.9 mi SE Culiacan. No river intervenes
between the paloviridis colony at that locality
and others farther north. Nor has dicranatus
ever been found north of that point despite in-
tensive survey work in the area. The influence
of dicranatus can be seen extending at least four
and possibly as far as 6.9 miles to the north. In
one character (exposure of ventral valves) a re-
markably smooth cline is shown, despite the
inadequacy of the sample sizes. It is possible
that swamping might be more effective in the
south because the populations there are of larger
extent. The nearest psolus is only about 24
miles north of the dicranatus colony, and may
thus have some influence on populations south
of Culiacan. This is suggested by the variability
of paloviridis in the city itself just south of the
river. However, at 6-9 mi S Culiacan, at least
four of the characters are uniformly different
from those in the colonies sympatric with either
psolus or dicranatus. In each of these there is a
cline southward.

61

A similar shift in seven of the eight characters
occurs at the Rio Culiacan west of the Rio
Humaya, over a distance of at most 3.5 miles
(Fig. 4). In this case the variation is greater, sug-
gesting greater gene flow. This variation may be
the result of the presence of pso/us on the south
bank of the river at one locality, where we sus-
pect it is a recent arrival.

HABITAT AND ASSOCIATED SPECIES.~— Paloviri-
dis appears to occupy a somewhat different hab-
itat north of Los Mochis than south of that city.
In the north, the coastal plain vegetation is
largely desert or thorn scrub, and here paloviridis
is found only in areas of denser vegetation usu-
ally along water courses. It occurs in the weedy
and bushy edge of these habitats which resemble
thorn forest but which are usually lower and
more open. Where it is found in the vicinity of
Guayimas, paloviridis is the only Barytettix pres-
ent although the widespread and common hum-
phreysii is found in numbers in both more open
and more densely vegetated habitats nearby. The
two species are found together only at Estacion
Luis, 40 mi SE Navojoa, at the edge of what ap-
pears to be a patch of thorn forest. To the south,
paloviridis is again found alone at 12.2 mi N Los
Mochis in a densely bushy arroyo. This locality
is well within the range of tridens which occurs,
like its close relative humphreysii, in more open
and more dense vegetation to the north and
south of the paloviridis locality. The exclusive
occurrence of paloviridis in regions where hum-
phreysii or tridens are common and widespread
is puzzling and suggests some kind of interaction
between them or special habitat requirements of
each.

South of Los Mochis, paloviridis is abundant
in almost all weedy bushy ruderal habitats, in
weedy or overgrown fields, and at the edge of
thorn forest. In the more open and disturbed
habitats, especially along the main highway, it is
almost continuously distributed and is often the
only Barytettix species present. It is absent from
the most intensively cultivated areas near Los
Mochis, possibly because even the roadsides had
once been scraped clear, or because of extensive
spraying.

Although paloviridis has been found within
thorn forest (but always near the edge) as at 3.1
mi N Tepuche, some of our data suggest that this
species is poorly adapted to that habitat, or is at
a disadvantage there relative to its congeners.

62

First, it was conspicuously absent from the thorn
forest at two localities where other Barytettix
species were common (5.3 mi NW Badiraguato
and 6 mi W Jesus Maria). Second, at 51.7 mi
SE Culiacan, it was abundant in a bushy area ad-
jacent to thorn forest, but absent from the thorn
forest itself, where tridens was moderately com-
mon. Finally, it was absent from the open weedy
summit of Cerro Tule, 7 mi SE Culiacan, where
contilus tectatus was moderately common, al-
though it was abundant at the base of the moun-
tain in weedy fields at the edge of the thorn
forest. This particular situation is further dis-
cussed under fectatus, where we hypothesize
that tectatus is better at penetrating the thorn
forest, or may actually exclude paloviridis from
that habitat.

Paloviridis has been found with psolus, nigro-
fasciatus, and each of the subspecies of contilus,
often in the same bush, although as indicated
above this may represent an overlap rather than
identity of habitat preferences. These species
are common in and at the edge of thorn forest
and only rarely occur in the more open habitats
occupied by paloviridis. At the southern end of
the range of both species, paloviridis altitudinally
overlaps psolus only slightly at the 700 foot
level at 19.5 mi SW Cosala. Paloviridis is com-
mon at lower elevations to the south and west,
and psolus at higher elevations to the north in
the vicinity of Cosala.

Between Culiacdn and Mazatlan, paloviridis is
also sometimes found with poecilus. The narrow
zone of overlap (10 miles) along the main high-
way and the abundance and continuous distribu-
tion of each species north and south of this zone
suggest some form of interaction, possibly com-
petitive exclusion, although the habitat of the
two may differ slightly. This is more fully dis-
cussed under poecilus.

Paloviridis occurs with tridens at a number of
the localities (Table 19, p. 104). The collections
have all been made at the edge of the thorn for-
est where paloviridis is the more abundant, al-
though at one locality, 51.7 mi SE Culiacan,
tridens penetrated farther into the thorn forest
than paloviridis.

The altitudinal range of this species has not
been adequately determined. At least at the
southern end of its range it was not found over
700 feet in a transect up to 1500 feet in the vi-
cinity of Cosala. No other mountains have been
investigated in its main range (north to Los

T. J. COHN ANDI. J. CANTRALL

Mochis). It has not been found in the foothills
near Tezopaco nor well into the Sierra Madre
near Alamos, but in this region it is rare even on
the coastal plain.

SEASONAL OCCURRENCE.~— Our earliest record
for paloviridis is 3 August at 21 mi SE Guaymas.
At Culiacan in the same year (1970), the species
did not mature until the middle of August in fields
under observation since the end of July. South
of that city we have several records of teneral
adults in the middle of August. In 1964 Cohn
worked briefly in the Culiacan-Mazatlan region
in late July without finding any Barytettix. The
rains at least in Culiacan were late that year and
the first shower had occurred only about a week
before. A similar brief investigation of a few
Barytettix localities in mid-June 1971 before the
summer rains similarly yielded no Barytettix. It
seems likely, therefore, that the eggs of this as
well as other species in the Sinaloa-Nayarit re-
gion hatch with the first summer rains in late
June or early July and that the animals mature a
month or two later.

Our latest record for the species is 28 Novem-
ber at 1.1 mi SW San Ignacio Ferry and 64 mi
SE Culiacan, and 1 December at 21.2 mi N Los
Mochis turnoff. Although the species was not
common at either locality, the abundance of
other species of Barvtettix there suggests a some-
what longer life. No collecting for Barytettix in
the range of paloviridis has been done in the
winter or spring.

DISTRIBUTION.— The northernmost record for
paloviridis is 68 mi S Hermosillo, Sonora, and
the southernmost, 20 mi N old Mazatlan Airport,
Sinaloa. North of Los Mochis the species ap-
parently occurs in scattered colonies and has not
been found east of the main highway on two
roads investigated (Alamos and Tezopaco re-
gions). South of Guamuchil it is almost continu-
ously distributed along the main highway to 63
mi NW old Mazatlan Airport. Along the high-
ways south of this locality we have only two
males from 30 and 20 mi N Mazatlan, although
the species is fairly common farther inland on
the San Ignacio road, about 49 mi N Mazatlan.
Within its main range south of Guamuchil, palo-
viridis occurs as far east as we have investigated
in the vicinity of Tepuche, Sanalona and San
Ignacio, but all of these collections have been in
the coastal plain or low foothills. It is absent
from around Badiraguato. In the only area with-
in its main range for which we have a transect

CONALCAEINE GRASSHOPPERS: BARYTETTIX

into the mountains, it was found no farther
inland than 19.5 mi SW Cosala. The limitation
may have been altitudinal or from interaction
with psolus which occurs commonly from this
point inland and at higher elevations.
RECORDS.— Specimens examined: 512c°d, 899 9, 2 ju-

veniles and many more reared to maturity. A complete list of
the records of this species will be found in Table 19, p.104.

THE HUMPHREYSII GROUP

The remaining Barytettix populations are
much more diverse from one another than are
the members of the Psolus or Crassus Groups.
Some are the exclusive occupants of the north-
ern end of the generic area, and the rest extend
far into the middle of that range where they are
sympatric with members of the Psolus Group.
Although these populations do not share among
themselves any single outstanding characteristic,
they lack any of the more distinctive features
of either the Crassus of the Psolus Groups. Nev-
ertheless, the cercus, aedeagus, bursa, and thick
tube are constructed on the same general plan
which is somewhat different from that found in
the other species groups. Relationship can also
be demonstrated between the most highly diver-
gent elements by hybrid or intermediate popula-
tions. We are therefore designating these taxa as
members of a third species complex, the Hum-
phreysii Group.

Two species have been previously recognized
in this complex, humphreysii and cochisei, both
from Arizona and now known from Sonora. We
now have evidence of hybridization of the two
forms from one area in Arizona and from an-
other about two hundred miles to the south in
Sonora. Therefore, cochisei is here reduced to
subspecific status, even though critical problems
remain concerning its biological status. The
nominate subspecies displays an extraordinary
amount of variation throughout its range in
Sonora. Much of this variation is discordant,
suggesting differential gene flow and selection
rather than past or present restriction of gene
flow between populations, and thus not worthy
of subspecific recognition. In southern Sonora,
however, a series of morphological and color
changes takes place concordantly. Although
intermediate conditions or tendencies toward
southern conditions are found just north of the
point of change of several of these characters,
the gap between the two morphotypes has now
been narrowed to 3.4 miles, and neither hybrid

63

populations nor evidence of introgression has
been found. We are therefore recognizing the
southern populations as a separate, new species,
tridens.

The three taxa share the features listed below.
These are not restricted to the Humphreysii
Group, but their combination is unique.

AEDEAGUS

1. Short and broad (as in the Crassus Group, but
slightly larger) (Fig. 1 E).

2. Dorsal valves deeply cleft along the midline
(except ina few variant populations of tridens,
in which excisions laterad of the aciculate
median teeth are deeper than that between the
teeth) (as in the Psolus Group) (Figs. 8 F, G;
10 A-T).

3. Dorsal valves more or less flat (weakly convex
in northern humphreysii), smooth from side
to side, or very weakly fluted.

4. Ventral valves considerably wider across the
middle or base than near the apex (somewhat
as in the Crassus Group) (Figs. 1 F, 11 A-P).

5. Ventral valves in side view straight or slightly
up-turned near apex (as in the Psolus Group),
never down- or out-turned (Figs. 9 A, C, E).

6. Ventral lobes of the sheath lengthily exposed
beyond the ramus of the cingulum (as in the
Psolus Group), never collar-like (Figs. 1 E-F);
the process of the ramus of the cingulum ob-
scured by the sheath (Fig. 1 E).

CERCUS

1. Strongly incurved and twisted, so that the dis-
tal portion lies almost flat over the supra-anal
plate (less so in northern humphreysii) (ex-
treme condition seen in Fig. 8 H).

. Disto-dorsal portion strongly produced (less
so in northern humphreysii, but always more
than in terminalis, crassus, and contilus) (Figs.
6 A-P).

BURSA COPULATRIX (Figs. 16 E; 17 E, F)
Short (but not as short as in the Crassus
Group).

THICK TUBE (Figs. 16 E; 17 B, C, E, F)

Arising dorsally and proceeding dorso-
cephalad, thin at and beyond origin, proximal
bends shorter than bursa and not dilated (all
but the first as in contilus).

COLOR (Frontispiece)

1. Pronotum with metazona, ventral carina, and
ventral third of lateral lobes of pronotum
same as ground color, without black markings.

2. Abdominal mid-dorsal longitudinal stripe and

i)

64

dorso-lateral yellow and black spots always
distinct.

The features listed below are found only in the
Humphreysii Group, but not in all populations:

1. Ventral aedeagal valves aciculate or lengthily
acute (except h. cochisei, Tezopaco h. hum-
phreysii, and some tridens) (Figs. 9 B, D, F;
11 E-H).

2. Ventral valves with the medial margins con-
cave at least in the distal half, and the lateral
margins usually convex, thus the valves dis-
tinctly, and often strongly, incurved (except
h. cochisei and Tezopaco h. humphreysii)
(Figs. 11 E-H, J-P).

3. Ventral lobes of the sheath (always lengthily
exposed) in ventral view truncate or only
broadly and briefly produced, never with
approximate finger-like projections (except in
tridens which possesses finger-like, but widely
separated projections) (Figs. 12 A-S).

4. Apices of the ventral lobes of the sheath in
distal view flattened, and more or less straight
across (except in ¢ridens and Tezopaco h. hum-
phreysii), never oval and parallel.

Table 13 under the diagnosis of humphreysti
will serve to differentiate the two species of this
group.

Barytettix humphreysii (Thomas)

Included in this taxon as a subspecies is
cochisei Gurney, for reasons discussed under

Geographic Variation and more extensively under

cochisei.

DIAGNOSIS.— Because of the extensive geo-
graphic variation in this species, no one charac-
ter (with the possible exception of the visibility
of the ventral valves), or even a combination of
characters will serve to distinguish humphreysii
from tridens. However, a number of character-
istics are restricted to one or the other species,
even though they are not found throughout the
species in which they are present. On this basis,
Table 13 may be used to distinguish all males,
but only some females of the two species. Note
particularly that the middle column has been in-
cluded for completeness, but cannot be used for
identification.

Color comparison of all species of the genus
may be found in Table 3, p.25. Geographic
variation in color and other characters in h. hum-
phreysii may be found in Table 14, p. 68, and in

T. J. COHN AND I. J. CONTRALL

h. cochisei in Table 15, p.79. Brown specimens
of humphreysii may be readily recognized from
the brown species of the Psolus Group by the
lack of the distinctive pronotal markings of the
latter. The brown variants of paloviridis in the
Guaymas area are indistinguishable in color from
the nearby brown humphreysii, and the green
humphrevsii in southern Sonora are similarly
indistinguishable from sympatric green palovi-
ridis. In the latter region, however, in contrast
to the reddish or reddish-purple unicolored tibiae
of paloviridis, humphreysii often has some blue
on the tibiae which are often distinctly bicolored.
Green humphreysii is indistinguishable in all
color characters except tibial color from poecilus
and crassus which are found far to the south. In
these cases where color identification fails, male
humphreysii may be distinguished from members
of other species groups by the twisted, expanded
cercus, and from paloviridis and other members
of the Psolus Group by the absence of an en-
larged pallium.

SPECIES DESCRIPTION.— CERCUS (Figs. 6 A-K): in-
curved, twisted, dorsal margin concave, disto-dorsal portion en-
larged, each character moderately so in north, strongly so in
south; when strongly twisted and incurved, disto-dorsal portion
lies horizontally in normal resting position (similar to Fig. 8 H);
disto-ventral portion acute angulate, often moderately produced,
rarely rectangulate. AEDEAGUS (Figs. 1 E-F; 9 A-D): short
(length 0.05-0.12 times length of pronotum) (or moderately
elongate, 0.11-0.14 times length of pronotum in h. cochiset),
wider than long (length 0.57-0.69 times width, 0.82-0.84 only
at 48 mi S Hermosillo and 74.4 mi S Navojoa; 0.74-0.87 in
h. cochisei). DORSAL VALVES (Figs. 8 F-G, 10 A-R): mod-
erately convex from side to side, to more or less flat and some-
times weakly fluted longitudinally; free lobes broad and diver-
gent, or parallel, or narrow and convergent (narrow, parallel and
widely separated in h. cochisei), medial margins smooth or with
small to large obtuse teeth, or large acute teeth, margins forming
a V or a narrow U between teeth and either strongly divergent
or convergent distad of teeth (margins forming a broad U or V
in h. cochisei); lateral margins weakly to strongly convex, often
divergent, apices broad and blunt to narrowly acute. ,. VENTRAL
VALVES: in cephalic view (Figs. 8 F-G), dorso-lateral portions
in north more widely separated than ventral portions, thus
widely exposing ventral wall of phallotreme, in south dorsal and
ventral portions about equally separated; in caudal view (Figs.
°B, D; 11 A-M), lateral margins always converging distad, con-
vex in proximal half, convex, straight or weakly concave in distal
half (margins conspicuously concave beyond base in h. cochiset),
medial margins concave distad, and strongly (in north) or weakly
(in south) convex proximad (straight and divergent in Tezopaco
specimen, straight and almost parallel or slightly divergent near
apex in h. cochisei), apices sharp acute to acuminate (usually
narrowly rounded in h. cochisei), ventral surface longitudinally
concave (Fig. 9 B) in north and south, or abruptly thickened
near middle in Tezopaco-Obregén-Navojoa region, medially
strongly excavate longitudinally below ventral phallotreme wall
in Esperanza-Tezopaco specimens (Figs. 11 I-J); in side view,
valve straight and moderately thick in north (Fig. 9 A), elsewhere
variously straight or sinuate, thin or thickened at shoulder (Fig.
9D) or near middle. RAMUS OF CINGULUM (Figs. 1 E-F):
ventro-distal portion narrow, ending considerably proximad of

CONALCAEINE GRASSHOPPERS: BARYTETTIX

COMPARISON OF BARYTETTIX HUMPHREYSIT AND BARYTETTIX TRIDENS

Character

TABLE 13

Conditions restricted to
humphreysii

Conditions common to
both humphreysii and
tridens

Conditions restricted to
tridens

GROUND COLOR
TIBIAL COLOR
TIBIAL COLORATION

DORSAL VALVES

Attitude of free
lobes

Dorsal surface

Apices

Median teeth

Plane of apices and

median teeth

Inner margins of
latero-apical
teeth

VENTRAL VALVES

In dorsal view

Apices

VENTRAL LOBES OF SHEATH
Apices

Median excision

CERCUS

Disto-dorsal portion

Disto-ventral tooth

Attitude

BURSA COPULATRIX
Distal end

Green; gray; blue-green
Yellow; orange; red

Bicolored

Broad and divergent or
parallel (Figs. 10 A-Q)

Strongly and evenly
convex

Blunt (Figs. 10 F-Q)

Rounded; obtuse; absent
(Figs. 10 A-L)

Both on same plane

Forming a broad U
(cochisei) or U at
least distad of median
teeth (Figs. 10 A-Q)

Apices and usually lateral
margins exposed by
dorsal valves

(Figs. 8 F, G)

Lengthily aciculate

(Figs. 11 E-H, K-M) or
briefly rounded (cochisei)
(Figs. 11 A-B)

Truncate or broadly
rounded (Figs. 12 A-F,
K, M, N)

Slight and shallow
(Figs. 12 A-L)

Moderately expanded
(Figs. 6 A-H)

Produced (Figs. 6 A-I, K)

Weakly curved and
twisted

Triangular (cochise/)
(Fig. 17 F)

Brown
Purple

Unicolored

Narrow and convergent
(Figs. 10 R-T)

Flat; weakly and irregularly
convex; weakly fluted;
somewhat concave

Acute (Figs. 10 R-T)

Acute (Figs. 10 M-O,
Q-S)

Apices raised, median
teeth depressed

Convergent (Figs. 10
R-T)

Acute or briefly
aciculate (Figs. 11 1, J,
N-P)

Narrow but not elongate
(Figs. 12 G-I, S)

Moderate (Figs. 12
M-N, QO)

Strongly expanded
(Figs. 6 I-P)

Rectangulate (Figs. 6
J, M)

Strongly curved and
twisted (Fig. 8 H)

Truncate
(Fig. 17 E)

Blue (Frontispiece )

Blue (Frontispiece)

Aciculate (Figs. 10 T)

Usually completely
obscured by dorsal valves
(Fig. 9 F, ventral view)

Appearing finger-like,
lateral (Figs. 12 O-R)

Wide and deep (Figs. 12
OF PSRs)

Obtuse or rounded (Figs.
6 L, N-P)

65

66

junction of dorsal valves with ventral lobes of sheath, thus ex-
posing a lengthy portion of ventral lobes.) VENTRAL LOBES
OF SHEATH: lengthily developed proximad of junction with
dorsal valves but without strong sclerotization there (Figs. 1 E-F,
9 A-D); flat from side to side to swollen along midline; proximal
portion strongly flanged (briefly reflexed) in north and south or
flat in middle part of range; distad of junction with dorsal
valves (Figs. 9 A-D, 12 A-N), not at all to moderately elongated;
apices truncate, or rounded, or briefly and angulately produced
in middle of each apex (Obreg6n region, Figs. 12 H-I), or mod-
erately produced into blunt, widely separated broad processes (in
south, Fig. 12 N) which are never finger-like or appressed to each
other; apical portions of lobes appressed to valves in north, or
projecting like lips at variable angles from valves. BURSA CO-
PULATRIX AND THICK TUBE (Figs. 16 E; 17 B, C, E, F):
bursa wider than long (length 0.66 times width) (or about as
wide as long in hf. cochisei, length 1.1 times width), pleated;
proximally with a moderately or well-developed sclerite on either
side; apex truncate (triangular in h. cochisei); thick tube undi-
lated, arising subdistally from dorsum of bursa (or distally in
h. cochisei), and proceeding dorso-cephalad, length between
proximal bends much shorter than bursa. COLORATION:
eround color green, blue-green, gray, or brown, with varying
amounts of yellow, abdomen usually more yellow. Head with
yellow post-ocular stripes narrow in north, wide in south, Pro-
notum with dorso-lateral yellow stripe narrow in north, gener-
ally broad in south, rarely extending weakly onto metazona;
black spot on lateral lobe restricted to pro- and mesozona, usu-
ally completely cut by cephalic diagonal yellow stripe, caudal
horizontal stripe absent to complete, spot narrowly to broadly
bordered ventrad by yellow, ventral third of lobe and ventral
carina concolorous with metazona (yellowish apparently only in
poorly preserved specimens when ground color is yellow brown).
Abdomen with mid-dorsal yellow stripe usually sharply defined,
sub-obsolete only in south; yellow and black dorso-lateral mark-
ings usually well-developed; supra-anal plate, cerci, and subgeni-
tal plate tinged or strongly colored with color of base of hind
tibia. Hind femur with ventral half of pagina yellow or green,
dorsal half darker green, blue-green, brown or blackish, some-
times entire pagina green; geniculae usually tinged with color of
base of hind tibia, with black lunae. Hind tibia unicolored yel-
low, orange, light red, wine, or purple; or bicolored, bluish or
purplish proximad, gradually becoming purple or reddish distad.
MEASUREMENTS (in mm): those of the series studied are
summarized in Figures 19 and 20.

GEOGRAPHIC VARIATION AND SUBSPECIA-
TION.— This species displays a complex pattern
of variation in which two types appear to be dis-
tinguishable. All specimens from southeastern-
most Arizona and from a small area south of
Hermosillo in Sonora possess a number of dis-
tinctive aedeagal and bursal features. These
characteristics are concordant in marginal popu-
lations of this morphotype which are found
within 10 to 20 miles of other humphreysii
morphotypes, with the exception of two hybrid
populations near Bisbee, Arizona and two south
of Hermosillo, Sonora. The hybrid zone is very
narrow, and individuals only two to six miles
away appear to be unaffected by the hybridiza-
tion. One of the hybrid populations is consider-
ably variable. We interpret these data on con-
cordance and hybridization as evidence for
recent contact between two previously isolated

T. J. COHN AND I. J. CANTRALL

populations, thus satisfying our definition of
subspecies. The southeastern Arizona form was
described by Gurney (1951) as a full species,
cochisei, which we here reduce to subspecific
status and include within it the Hermosillo pop-
ulations. Certain data are difficult to interpret
under this concept and the possibility that
cochisei and humphreysii represent stasipatric
species in the sense of White et al (1967) and
Key (1968) are discussed under cochisei, where
a complete description and discussion of the
hybrid populations will also be found.

The remaining geographical variants of this
species do not display this degree of concordance
and many show no morphologically intermediate
transitions to other variants. Accordingly, these
are not named, but are described and discussed
under the nominate subspecies.

DISTRIBUTION.— This species occupies most
of the southeastern quarter of Arizona and ex-
tends southward to northernmost Sinaloa in
México. Eastward it is known from southwes-
tern New Mexico near the Arizona border, and
northwesternmost Chihuahua, but farther south
in México it has not been found east of the
Sierra Madre Oriental.

Barytettix humphreysii humphreysii
(Thomas)
Figs. 1 E-F; 2 A; 6 B-K; 8 G;9 A-D; 10 F-R;
11 E-Met2€-N: 16 7E-F2 17 BSE

Pezotettix humphreysii Thomas, 1875, Rep. Geogr. Geol. Expl.
Surv. West of the 100th Meridian in charge of G. M. Wheeler, 5
(Zoology): 890-892, Pl. XLV, Figs. 1-2 [In: Rep. Geogr. Surv.
West of the 100th Meridian]. [Typeso*, 9, Southern Arizona,
1874; apparently lost. ]

Conalcaea neomexicana Scudder, 1897, Proc. U. S. Nat. Mus.
20:24, 26, Pl. 2, Fig. 9. [Type, o, New Mexico, Grant Co.,
Silver City; Academy of Natural Sciences of Philadelphia. ]
{Synonymy from Gurney, 1951.]

Barytettix borealis Caudell, 1908, Proc. Entomol. Soc. Washing-
ton 9:69-70. [Types, 2%, 1 2, Arizona, Pima Co., base Santa
Catalina Mts.; lectotype in United States National Museum.]
{Synonymy from Gurney, 1951.]

The types of this subspecies probably should
be in the United States National Museum where
other Thomas types are now located, including
that of Pedioscirtetes nevadensis which was de-
scribed in the same paper as B. humphreysi.
However, a search for the types was made by
Ashley B. Gurney of that institution during the
preparation of his revision of the Conalcaea com-
plex, and he reports that the types are “appar-
ently not in existence” (1951:299). The original

CONALCAEINE GRASSHOPPERS: BARYTETTIX

description and the figures accompanying it
clearly indicate a species of Barytettix!®, but
these are not precise enough for us to determine
to which species of the genus the name applies.
Thomas indicated that the specimens were col-
lected on the 1874 expedition in ‘southern
Arizona,” a region where only h. humphreysii
and h. cochisei occur. We have been unable to
locate an account of the route of that expedi-
tion, and cannot determine whether it passed
through the range of only one or both of the
subspecies. We therefore follow Gurney’s (1951)
assignment of the name humphreysii to the more
widespread of the two forms in southern Arizona.

DIAGNOSIS.— This subspecies may be distin-
guished from h. cochisei in the male by the con-
vex outer margins of the ventral aedeagal valves,
usually concave inner margins, and often acicu-
late apices (Figs. 11 E-M), and in the female by
the truncate distal end of the bursa copulatrix,
and the distinctly sub-distal origin of the thick
tube (Figs. 17 B, E). South of Hermosillo, ad-
jacent to the range of cochisei, the lateral mar-
gins of the ventral aedeagal valves may be
slightly concave somewhat as in cochisei (Fig.
11G). The dorsal aedeagal valves, although
quite variable, are usually distinctively different
from those found in cochisei, Their free lobes
are usually broad and blunt (except in southern
Sonora) and their medial margins are often pro-
vided with prominent teeth near the middle
(Figs. 10 F-R), and distal to the teeth are never
parallel, being either divergent or rarely
convergent.

Body and tibial color, color pattern, and
shape of cercus are identical in the two subspe-
cies in the region of contact in the Douglas and
Hermosillo areas, but elsewhere h. humphreysii
may be distinctively different. The ground color
may be bluish-green or green, the tibiae bluish-
purple, the cercus may be strongly incurved and
twisted, the disto-dorsal portion greatly expanded
and the disto-ventral tooth rectangulate or but
little produced. In addition, certain geographic
variants in the ventral aedeagal valves and in the
ventral lobes of the sheath differ from cochisei.
The ventral surface of the valve may be strongly
swollen in the middle, and it may be sharply ex-
cavate proximally or near the medial margin,

16Fmarginate caudal margin of pronotum; abbreviate, nar-
row, spatulate tegmen; falcate and distally enlarged male cercus;
conical male subgenital plate; interrupted black spot on lateral
lobes of pronotum; robust form.

67

and the lateral margin may be uniformly thick or
greatly swollen near the middle, all in contrast
to the smooth but longitudinally concave ventral
surface and distally thin, proximally moderately
swollen lateral margin of the valve in cochisei.
Other variants in these characters are similar to
cochisei. In the ventral lobe, humphreysii has a
strongly developed proximal flange in Arizona
and northernmost Sonora where it is in contact
with cochisei, which lacks the flange or has it °
only weakly to moderately developed. The
flange is weak or absent in central Sonoran hum-
phreysii (in the region where cochisei is also
found), but it is well-developed again in southern
Sonora. The distribution of the geographic vari-
ants in this subspecies is summarized in Table 14,
and color comparisons with all species may be
found in Table 3, p.25.

SUBSPECIES DESCRIPTION.— Characteristics of
h. humphreysii are those given in the preceding
species description where all characteristics
found exclusively in the subspecies h. cochisei
are enclosed within parentheses.

GEOGRAPHIC VARIATION.—This subspecies dis-
plays strong geographic variation in almost every
character studied. Table 14 is an attempt to
summarize these data visually in order to dem-
onstrate the degree and location of concordance
among the variants.

Most of the localities listed in that table repre-
sent a series of closely spaced collections along
the main west coast highway between Nogales
and Los Mochis. Between Guaymas and Los
Mochis (SE Navojoa) the highway runs roughly
through the middle of the coastal plain. Mate-
rial from transects and localities to the east of
this road is listed in indented portions at the
appropriate geographical position in the list.
The variation in each character is indicated by
words, abbreviations in capital letters usually of
the first letter of the condition, or by letters re-
ferring to figures. In the latter case, conditions
which we consider to be individual variants of
the same basic structural type are listed in paren-
theses. All such variants are listed for the first
locality at which the basic type occurs, even
though the variants may not occur there, and
they are not repeated for subsequent localities
where they do occur. This has been done to
make the discontinuities clearly visible in the
table. Solid horizontal lines in the character col-
umns represent major discontinuities, dashed

4
=
<
o ———
= uy M Od sewAendy as 772 — LI
S (sewxend 4S 17 — 8
we sipiaia ond)
A “pautiayuy sewdeny
4 UML YsInid, Mb ‘Abg eUIperES
Zz “poulsa} Uy]
5} HEM paddry “Uy (W-1) ¥ nN on sewAeNd N Tb — S
O
ie (sewAeny N gi
as sipiaidojnd)
WPM ~ padi] zapynoys x
“duos ‘pouazuy “TY 7B YOTYL fC d ‘MU O[[IsoWaH § 8b — Ob
ae ar “pouuayuy Japynoys pani
peoig suons d Vel ye YOU L (H) 9 JO 314 (H) 9 d peolg (e) O[[ISOWMAH S 6E
(S OPI — O[ftsouLtoH
19814002 “Y)
ae aos ~~ Spauuayuy ~uraew a a in eee ||) we oe
*pauwiia}u] pauls9}Uy veld Buoye ANY d xXaAU0d) a YW “pawlio}uy fay: | O[ISOWAH N IL
= a5 MoU wen “Ig (6) OjfsowsaH N 1
|
powsajuy f Y prog (eo) OjfisowsaH NTE
aiger =e ~ MOLEN Od. Buy eyUeS § OS
qu sunw] MS P€
UO'a sunw] N 19 — $1
A sunw] N L°L1 — Seleson
(uordal
“STW PONYyIeNH-adJatq)
WcOvk PUOZUY Usa}seayInosg
(sayedoN,
| uriew (amet Y SO) — [PyOg — 098A)
suo7 GUuoNS (ado vel Suoye YOY, (pd (-9) 4 xXaAuoy (4-0) 4 A Buns MOUPN od euozuy [eue)
UIPIAN adue|y are surddry ssouyory IL 34 Ol 34 a0e}INS 9 34 aeiqry ads saduis JOJO) (says Ur saoueysIp)
JPWIXOIg saoidy jeoidy adeys adeys yesioq adeys puly yeureg yejOuolg punoly Aypeo07
adeys
yyRays JO saqo7T [eIWUdA SOATEA [PIIUIA, SaajeA [eSIOd
snovaddav Snows YOToOD

68

HISKAMHdUWAH USXATMHdWAH XILLALAYVA NI NOILVIAVA OIHdVaDOD
vl ATaVL

69

CONALCAEINE GRASSHOPPERS: BARYTETTIX

pany (Oa OL1aD)
duoT 3uons N paddrq uly Ww Y JO 184 (1) M YEON, peoig uaa ROLOARN AS bbl
(uog uolorysq)
N un Ww Yu M@ my PofOatN AS 81S
2 “Tpourajupe ln a Fie ee aoe, (sinq uoreysq)
y Spry) ‘paw x Oo q d-d ‘M-d ‘A eofOAeN AS Ob
“pauajzuy
| “UL Mod ‘A rofoatn AS 9¢
|
“pawsaju]
duos Sypnp “pay d-€ ‘M-d ‘M eofoaeN AS IZ — OC
~ yeay
‘suns YOY) “Paw x d-d ‘d°M rofoarn AS POL
powa}uy ae nS
yoru) “paw 1'f M~d ‘A bofoatn AS 6'bl
| | *pauiiayuy
8u07 (wy “yO “paw (Dx f Mod ‘M HEI prog Us915) Pofoatn AS POL
3u07 Ww) uly ( f A Buons MOLIEN o€ sowrly 4 €€ — ol
| Ves |
| (o0) du sowrtly 4 8
| uly. M d MOLIEN Od sowtly § [ ‘MNZ~ |
“pauiiajuy dM Surg peoug O sowrly M £1
suo] (WX YOY) “pay x ft é “pawiia}uy: prog o‘o"d voforeN 4 OL
ee een EO, *pawsajuy | “pawa}uy a > ae
“pawwsayuy “Zu0NS “WY ‘yop ‘paw y I M Buns proig uaa vofoatN MAN 89
|
HOM
yea H yo “paw M~d ‘MX “Buons PofOAtN MN 77
“pauiia}uy POM
“yoryy “paw M-d ‘A “Buns uoHIGO AS [7
yoru
“pawiaquy (‘DH Aleipaw (Dx M~d (suo peolg Users) uoda1qO AS OL
yom
“pauusazuy 95 AUeIpaW I iM EMOLEN ood oovdozaL
yoryy
“paulajuy H ATeIpaW ri M-d prog (@) ezuriodsy AN
Poxiw pax pax pexiw (O‘d‘N)O M peoig VERO) uosAIGO AN OL
oe ee =. es ee ee |
proig a *pauuajyuy 5 Y I A woRqdO MNM ST
“pawisajuy i
“uty 2) Mod “A sewAend AS ge — LE
“pauiayuy Mod uaa sewAeny qS ¢¢

70

lines minor or questionable discontinuities. The
variation in each character is discussed below,
and this is followed by a general discussion of
the significance of the concordance or discord-
ance seen among all characters.

Ground Color. This was the most difficult
character to assess because of individual variation
and differences in preservation. The few notes
we have on living individuals from southeastern
Arizona and northern Sonora indicate that the
ground color is pearly gray rather than brown.
However, most dried material is brownish and
even a large series of freeze-dried specimens con-
tains brownish individuals among the gray ones.
We have designated this color type as brown-
gray (BG) in the table. Specimens between 32
mi N Hermosillo and Guaymas, as well as two
others near Santa Ana but not separately listed,
appear to be greenish and may be transitional
between the brown-gray to the north and the
brighter green to the south. They have been des-
ignated as olive (O) in the table. Unfortunately,
most of these specimens also show an increase in
yellow, and the green sometimes appears to be a
bluish pigment, making precise determination of
ground color difficult. The Saladita specimens
are definitely bluish (but not as bright as tridens
found much farther south). If this color is wide-
spread in the area (our samples here are inade-
quate), then specimens to the north which we
have called olive may be transitional to this color
type. Southeast of Guaymas a pocket of small
brown specimens occurs. It is interesting to note
that in this area some of the paloviridis are also
brown and small, although elsewhere that spe-
cies is almost universally green and large. South-
ward, the shift to green is abrupt, and all speci-
mens from the coastal plain south of 33 mi SE
Guaymas are bright green. Inland and at higher
elevations, specimens become browner. How-
ever, those from 10 mi E Navojoa are clearly
more brown than specimens farther inland at 13
mi W Alamos. The Navojoa sample was collected
from the summit of a small mountain and very
late in the season. Either factor might explain
the reversal in a clear west-east trend toward
brown. It is perhaps significant that individuals
from the southernmost collection of humphreysii
are just as green as those to the north, although
they are only about three miles from the nearest
blue tridens.

Pronotal Stripes.— The width of the stripes isa
subjective estimate. The difference between the

T. J. COHN AND IL. J. CANTRALL

narrow northern and wide southern types is ob-
vious. The sample termed intermediate at 11 mi
N Hermosillo is hard to assess because the ground
color is considerably lighter than the adjacent
populations. As in ground color there is a trend
toward the northern condition inland and at
higher altitudes.

Paginal Stripe.— The estimate of the degree of
intensity of color of this stripe is subject to the
same problems of preservation and individual
variation as that of ground color. In Arizona,
northern Sonora, and around Alamos the stripe
is almost always present and dark. South of
Guaymas the stripe is usually lighter, and from
10.4 mi SE Navojoa to the southernmost colony
the stripe is usually absent or very weak.

Tibial Color.— This character seems to be
much less affected by preservation than other
color characters. We suspect that all tibial color
will show complete integradation, and our des-
ignations are arbitrary points on a continuum,
although northern Sonoran series seem to be read-
ily divisible into discrete red, orange, and yellow
catagories as in cochisei. Specimens have not
been compared to a color standard, but rather to
adjacent populations. In Arizona, red-legged indi-
viduals occur almost exclusively in the southeast-
ern corner adjacent to the range of the primarily
red-legged cochisei, although a few pale red or
orange-legged individuals occur west of the
Pajarito Mts. Not all the southeastern specimens
are red-legged, and some closest to cochisei (Bis-
bee, Portal) are yellow or pale orange. On the
other hand, the hybrids between humphreysii
and cochisei in this region are red. The shift to
red may come at the Mexican border in the Naco
region as evidenced by the all red-legged popula-
tions at that town and near Cananea. However,
130 miles south of the border in the mountains
30 mi E Carbo (not listed in Table 14), a small
collection of females contains yellow, orange,
and red-legged specimens. This may be an alti-
tudinal effect, because the transition on the main
highway to the west occurs only 43 miles south
of the border at Imuris. South of Imuris all spec-
imens in numerous collections possess red or
darker tibiae. In the Hermosillo region, tibial
color shifts to purple or wine. It is tempting to
ascribe this shift in humphreysii to introgression
from cochisei which is largely bicolored purple-
wine in this region as are the hybrids between
the two subspecies. However, purple or wine
colored tibiae are found continuously from

CONALCAEINE GRASSHOPPERS: BARYTETTIX

Hermosillo to the southern end of the range of
humphreysii which is far beyond any indication
of morphological influence of cochisei on hum-
phreysii (see discussion below under Ventral
Valves). The irregular variation south of Cd.
Obregon is hard to assess. The irregularity may
be the result of sampling error; most of our
series in this region are small and several color
types are present. It is possible that there is here
a trend toward the blue tibiae of tridens found
farther south. Such a trend is clearly reversed in
the southernmost humphreysii populations at
51.8 and 74.4 (Cerro Prieto) mi SE Navojoa,
where all 23 specimens are wine-legged. Thus,
the humphrevsii colonies closest to tridens are
more different from that species than are col-
onies farther north. This may be a case of char-
acter reinforcement, although the two species
are not known to be sympatric at the present
time (the closest known colonies are only 3.4
miles apart).

In the table, the following abbreviations for
tibial color have been used: Y, yellow; O, orange;
R, red; W, wine; P, purple; P-W, purple proximad,
wine distad; B-P, blue proximad, purple distad.

Cercus.— Toward the south the cercus be-
comes more incurved, the disto-dorsal portion
higher, and the disto-ventral tooth shorter and
more rectangulate, as illustrated in Figures 6 B-K.
South of Hermosillo there appears to be a minor
change to a more expanded disto-dorsal portion
and a more down-turned tooth (G) compared
with those to the north (B-F). Between Obre-
gon and Navojoa (and inland to Tezopaco) the
disto-dorsal portion is distinctly higher (1), and
the distal margin becomes almost vertical. South
and east of Navojoa there is a more distinct
break. Here, the disto-dorsal portion is greatly
expanded, and the disto-ventral tooth reduced,
and often rectangulate (J). Southward from
Navojoa, the disto-dorsal portion is somewhat
less expanded, and the tooth often slightly pro-
duced (K). At 51.8 mi SE Navojoa and at Cerro
Prieto, 22 miles farther south, the disto-dorsal
portion expands more gradually and thus appears
broader, and the tooth is more rectangulate. This
condition is very similar to that in the nearby
tridens colonies and is indicated in the table by
(L) although that figure is one of a tridens speci-
men. There is little difference between these
southern humphreysii and tridens. The cerci of
the latter are slightly more incurved, the distal

71

margin often somewhat rounded, and the tooth
often obtuse.

Dorsal Valves of Aedeagus: Convexity.— The
valves of northern specimens are uniformly con-
vex, even distally, and sometimes appear to be
weakly inflated; these are given the designation
Convex. South of Hermosillo the valves are
somewhat flatter and each free lobe usually has a
weak longitudinal concavity distad; the edges
are therefore slightly turned up. In the Guay-
mas region and southward each free lobe is more
or less flat or slightly fluted, and the two lobes
together often form a flat surface. The Hermo-
sillo and Guaymas types are too variable to be
distinguished, and they are grouped together
under the designation Flat or Fluted. The transi-
tion from the northern to the Hermosillo type is
abrupt.

Dorsal Valves: Shape.— Despite considerable
individual variation, five basic types may be dis-
tinguished, as illustrated in Figure 10. Type F
of Table 14 (Fig. 10 F) is characterized by broad,
blunt-tipped free lobes, the axes of which appear
to be divergent. The median notch proximad of
the teeth is almost always V- rather than U-
shaped. There is much variation in the median
teeth. They are usually blunt (1), sometimes
acute (F), and occasionally entirely absent (G).
Type J has unusually prominent median teeth,
and the free lobes distal to the teeth are some-
what concave on their inner margins, and thus
rather narrow; these characteristics may be the
result of introgression from nearby h. cochisei.
Type K has the free lobes still appearing diver-
gent, but the lateral margins have a prominent
bulge and the median notch proximad of the
teeth is usually U- rather than V-shaped. In type
O the axes of the free lobes have shifted to paral-
lel or slightly convergent. The apices given the
appearance of being smaller, probably the result
of a distal displacement of the median teeth.
The latter teeth are variable, but often constant
within any one locality; thus the Tezopaco and
Esperanza specimens are characterized by broad,
prominent teeth with broad, ragged lateral mar-
gins (N and less frequently as in O), whereas the
teeth in the Alamos population are almost al-
ways greatly reduced and often absent altogether
(P). The final type (R) is very similar to that
found in tridens, characterized by long, narrow,
incurved lateral teeth, and acute, almost approx-
imate median teeth.

If the series is examined for overall pattern,
there appears to be a trend southward in a grad-
ual shift of the axis of the free lobes from diver-
gent to convergent accompanied by a narrowing
of the lateral apices and median teeth. This
trend continues almost without a break across
the tridens species boundary. The only differ-
ence between the adjacent colonies of the two
species is the final development of aciculate and
possibly fused median teeth. Curiously, the
more distant middle populations of tridens have
acute median teeth (S) much like those of south-
ern humphreysii (R), whereas the northern and
southern populations of tridens possess the most
extreme development of these teeth (Fig. 10 T).

The changes between the major types dis-
cussed above are relatively abrupt, and nowhere
do we have obvious intermediates or populations
polymorphic for two of the basic types. Al
though individual variation may cloud our iden-
tification, it seems clear that the transition zones
must be quite narrow, or are of short duration
before one or the other type is established in the
population.

Ventral Valves: Shape.— Several distinctive
variants occur in this structure, but, except in
the north, the location of the boundaries, if any
sharp ones exist, have not been adequately iden-
tified in this study. Between the northern type
(Figs. 11 E-F, 9 B) and that found south of Her-
mosillo (Fig. 11 G) there is a very sharp break.
The valves in colonies closest to cochisei south of
Hermosillo appear to be precisely intermediate
between the two subspecies (compare Figs. 11 B
and F with G), and may well be of hybrid origin.
The characteristics of these valves are uniform
from 39 to 48 mi S Hermosillo, and the general
type persists as far south as 10 mi N Obregon
(Figs. 11 H,9 D). Both the Hermosillo type and
the type found farther south occur at that local-
ity. In the southern type the valves appear more
strongly bent inward, and the inner margins
more strongly but more briefly concave (Fig.
11K). Between Alamos and 40 mi SE Navojoa
the outer margin makes an almost angulate bend.
At 51.8 and 74.4 (Cerro Prieto) mi SE Navojoa,
both margins are more gently curved and the
valve appears longer (Fig. 11 M). In addition to
these types, two others exist inland. At Tezo-
paco the valves are distinctive in their barely
curved outer and straight inner margins (Fig.
111), but this may be a modification of the
Hermosillo-Guaymas type. East of Alamos the

T. J. COHN AND I. J. CANTRALL

apical half of the valve is much narrower and
straighter, and the proximal shoulder of the
inner margin more prominent. Although this
type is not illustrated, we have designated it as
(K) in the table.

Ventral Valves: Thickness.— Various portions
of the ventral valves are thickened in different
parts of the range of h. humphreysii, but only
near the southern end are they conspicuously
and broadly tnickened. In the north the entire
length of the lateral margin appears somewhat
thickened (Fig. 9 A), which is partly the result of
a slight curling of the edge of the valve. South of
Hermosillo only a small portion of the margin at
the shoulder is thickened (somewhat as in Fig.
9C). From the Guaymas region to 25 mi WNW
Obregon the valves are usually thin, but several
specimens show some thickening at the shoulder.
From Tezopaco (NE Obregon) to the Navojoa
region the valves are conspicuously thickened in
much of their visible distal portions, as indicated
in Figures 11 I and J, and distad of the dashed
lines in Figures 11 K and L. The swollen por-
tion extends from near the medial margin to the
lateral margin, and is conspicuous in disto-lateral
view. The apices in this view taper rapidly to
aciculate points. The thickest valves are found
at Tezopaco (Fig. 111) where the swelling is
accompanied by a striking medial excavation.
Proximally the valves become abruptly thinner,
forming an almost rectangulate ridge. Farther
south specimens from 27.8 mi NE Esperanza are
intermediate between the Tezopaco condition
and the less swollen, less excavate type found
south of Obregon. A small collection from 10
mi NW Obregon contains specimens with thin,
intermediate and thick valves. Eastward from
Navojoa the valves become thinner, and they
have more or less returned to the thin northern
condition at Alamos and in the mountains to the
east. Thick valves persist but with some varia-
tion to 40 mi SE Navojoa, and become thin
again in the southernmost colonies at 51.8 and
74.4 (Cerro Prieto) mi SE Navojoa.

Ventral Lobes of Sheath: Apical Lipping.— The
apices of the ventral lobes may be closely
appressed to the ventral valves, or the distal
third to eighth of the lobes may project away
from the valves at an angle often as much as
90°. Although the distal portion of the lobe is
somewhat membranous, and our scoring has
been done entirely on dried specimens, the sharp
difference between northern flat and southern

CONALCAEINE GRASSHOPPERS: BARYTETTIX

lipped types makes us believe that the differ-
ences are not artifacts of preservation. North
of Hermosillo the lobes are usually appressed to
the valves (designated in the table as Flat), and a
small proportion of lobes is variously lipped but
rarely strongly so. Between 11 miN and 48 miS
Hermosillo, the valves are mixed and often inter-
mediate (with a somewhat swollen rim as in Fig.
9D), and from the Guaymas region southward
the valves are strongly lipped, with rare excep-
tions only northwest of Obregon.

Ventral Lobes of Sheath: Shape of Apices.—
Our scoring of this character may have been
influenced by differential preservation, and by
the visual effect of the degree of lipping. Never-
theless a north to south trend is discernible
from the more or less truncate apices (forming
a straight line or broad V, as in Figs. 12 C and D,
rarely with margins as rounded as in E) in the
north, through broadly rounded apices (Figs.
9D and 12F) to briefly produced angulate
apices (Figs. 12 G-J) and finally to more pro-
duced but narrowly rounded apices (Figs. 12
K-M) in the south. In the two southernmost
colonies the roundly produced apices are more
lateral (Fig. 12 N) and approach the condition in
the nearby tridens (Fig. 12 O). Zones of mixed
or intermediate valves are found at 10 mi NW
Obregon, possibly north of Hermosillo, and just
north and south of Navojoa.

Ventral Lobes of Sheath: Proximal Flange.—
The sclerotized proximal portions of the lobes
are obviously curled or strongly curved in the
northern and the southern portions of the range
of the subspecies, and flat or weakly curved in
the middle of the range. Zones of intermediate
or mixed conditions are found south of Her-
mosillo where the adjacent h. cochisei colonies
have weak or flat lobes, and in the Navojoa
region. As in the other characters, conditions
which might be ascribed to cochisei influence
(flat lobes) extend far to the south of the
cochisei colonies at Hermosillo, although we
do not know the range of cochisei east of the
main coast highway. The proximal portion of
the ventral lobe is sometimes obscured by mem-
brane. We have often softened this membrane
with ammonia, and then probed with a pin to de-
termine the degree of curvature of the sclerite.

Ventral Lobes of Sheath: Width.— This was a
subjective estimate of the length to width ratio.
The lobes of northern and southern specimens
appear to be longer than wide (despite the

73

proximal flanging of the lobe), whereas those in
the Hermosillo-Guaymas region appear to be
quadrate or wider than long. Apparent inter-
mediate conditions are found north of Hermo-
sillo, and between Tezopaco and Navojoa.

The Chihuahua Colony at Casas Grandes.—
Because of the isolated position of this colony
far to the east of the main distribution of the
subspecies, we have not included it in Table 14.
It possesses a curious combination of northern
and southern characteristics which are listed,
and the implications of this combination dis-
cussed, under Distribution.

Biological Significance and Taxonomic Treat-
ment.— We have long pondered and argued the
usefulness of designating various segments of
h. humphreysii as additional subspecies based on
this pattern of geographic variation. However,
if such designation is to have meaning beyond
mere description, it must be based on the bio-
logical significance of the variants.

North of Hermosillo only tibial color shows a
distinct pattern of geographic variation, red
tibiae replacing yellow at Imuris in northern
Sonora. Because this is not accompanied by
shifts in other characteristics in this area, we do
not consider the populations worthy of subspe-
cific designation. The shift to red tibiae may
occur much farther south in the mountains as
indicated by the mixed population east of Carbo,
although this is the only collection of Barytettix
in the mountains north of Alamos. The red
tibiae found in the southeasternmost corner of
Arizona may be the result of introgression from
the adjacent red-legged h. cochisei (see discus-
sion of a possibly hybrid colony at Don Luis,
near Bisbee, Arizona, under h. cochisei).

At Hermosillo there is a striking shift in many
characteristics. Much of this may be the result
of introgression and hybridization with h. cochi-
sei, which interrupts the range of h. humphreysii
for about 25 miles south of that city. The two
subspecies are now hybridizing freely at 30 and
36.8 mi S Hermosillo where the colonies are
highly variable and contain parental types and
intermediates (see discussion under h. cochisei).
For 11 miles south of these hybrid colonies,
most characteristics of h. humphreysii are rela-
tively uniform. Some are intermediate between
the two subspecies (ventral valves, Fig. 11 G and
compare with 11 B and 11 F, the parental types;
possibly dorsal valves, Fig. 10 J, compare with
10 E and 101). Other characteristics are similar

74

to the conditions in Hermosillo h. cochisei (tibial
color, ventral lobe width, proximal flange, and
possibly apices). Some of the latter shift back
toward normal h. humphreysii conditions not
far to the south. Such a situation would justify
subspecific treatment, except for the irregular
and non-concordant distribution of these char-
acteristics in the area. Thus, the dorsal valves
change well north of Hermosillo; the proximal
flange of the ventral lobes is variable in the first
colonies south of Hermosillo, then becomes even
more like cochisei south of the point where the
ventral valves have changed. The general condi-
tion of the Hermosillo ventral valves persists far
to the south, although a minor change occurs at
Guaymas. Thus any boundaries drawn would be

arbitrary, and the establishment of zones of

introgression would apply to some characters
and not to others. Perhaps we are now seeing
the results of a long history of hybridization and
introgression, with cochisei having been elimi-
nated in the south but leaving evidence of its
former presence there. Interaction between the
two may also be taking place to the east where
we have no collections. The present evidence
points to relatively free gene flow between the
two subspecies in the Hermosillo region, but at
different rates for different characteristics. More
intensive study in this area would be rewarding.
Other characteristics which shift near Hermosillo
cannot readily be ascribed to interaction with
h. cochisei. In both subspecies, ground color,
pronotal stripes, bicoloration of the hind tibiae,
and twisting and distal expansion of the cerci
shift in this region but persist to the southern
end of the range of h. humphreysii. A similar
pattern is seen in the convexity of the dorsal
valves in the latter subspecies. The shifts in these
characters are also non-concordant, and no
boundaries can be established which apply to
more than a few of them.

There is greater concordance in the shift in
characteristics northwest of Obregon. Cercus
and dorsal valve shape change here, and charac-
teristics of ventral valve shape and thickness, and
ventral lobe width and apical shape are either
intermediate or mixed. A meaningful subspecific
boundary might be established here were it not
for a series of characteristics which pass across
this point unchanged, and which sometimes
change not far away (ground color, pronotal
stripes, possibly tibial color, and flanging and
lipping of ventral lobes). More adequate collec-

T. J. COHN AND I. J. CONTRALL

tions from this area might clarify the situation.

A similar area of concordant shifts occurs in
the short distance between Estaciones Luis and
Don, 40 and 53.5 mi SE Navojoa respectively.
Marked changes occur in tibial color, cercal
shape, dorsal valve shape, ventral valve shape
and thickness, and ventral lobe apical shape.
The change in tibial color might be ascribed to a
reinforcement phenomenon involving the nearby
tridens, because it represents a reversal in a
southward trend toward the bluish color found
in that species. Most of the changes, however,
are toward conditions in ¢ridens, suggesting in-
trogression. On the other hand, some charac-
teristics of these humphreysii colonies which
differ from fridens persist across the point of
change of other characteristics, suggesting free
gene flow to the north. We believe that desig-
nating these southern populations may further
obscure an already puzzling situation, and sug-
gest an explanation where none now seems en-
tirely satisfactory.

At Alamos and Tezopaco, conditions of vari-
ous characters are highly distinctive. At Tezo-
paco there is a remarkable development of the
ventral valve (Fig. 111) involving considerable
thickening, deep excavation, and almost straight
margins. The valves of a sample at 24.7 mi NE
Esperanza are almost exactly intermediate be-
tween the Tezopaco valves and those of the
coastal plain populations. Other characteristics
are either not much different from nearby popu-
lations, or are matched by variants in those
populations (as in the shape of the apices of the
ventral lobes, Figs. 12 G and I). Similarly the
distinctive cerci of the Alamos populations ex-
tend to the coastal plain where probable inter-
mediates exist, and the narrowed distal portions
of the ventral valves most likely represent the
end of a cline. Of greater interest is the apparent
development of northern color characteristics in
these populations, suggesting a connection be-
tween these and northern populations in the
mountains. Thus designation of these popula-
tions as a separate subspecies would probably
require the arbitrary division of clines, and would
obscure a possible northern contact and genetic
continuity in the mountains.

HABITAT AND ASSOCIATED SPECIES.— Humph-
reysti occurs through a wide altitudinal range
and in a variety of habitats. Our highest alti-
tudinal record is 6250 feet above Carr Canyon,
Huachuca Mts., Arizona, and our lowest is near

CONALCAEINE GRASSHOPPERS: BARYTETTIX

sea level southeast of Guaymas, Sonora. At high
altitudes, humphreysii has been found well into
the oak zone, and at 33 mi E Alamos it occurred
in the lower edge of pine-oak woodland. Al-
though detailed descriptions of the habitat at
these higher elevations are lacking, there are in-
dications that humphreysii occurs in the more
open woodlands, in savannah-like formations (as
near Ruby, Arizona) or rocky areas where desert
or chapparal bushes intrude (above Carr Canyon,
Arizona). It is conspicuously absent from moist
stream bottoms in the mountains where the trees
form an almost complete canopy (4.5 mi SW
Portal, Arizona). Below the oak zone, this spe-
cies is found in a variety of desert habitats, in-
cluding low bush mesquite, desert grassland, and
creosote bush. We have many collections along
the main highway from Nogales to Guaymas.
The species is generally common in depressions
with weedy growth, but we have not checked
adjacent drier areas lacking weeds. It is appar-
ently not present in patches of almost pure grass
south of Hermosillo, and may also be absent
from savannah-like areas north of that city. The

species is also common in thorn scrub south of

Guaymas and was abundant in the edge of thorn
forest-tropical deciduous forest between Navojoa

and Alamos, and near Estacion Luis south of

Navojoa.

There appears to be no difference in habitat
between humphreysii and cochisei in the Hermo-
sillo area, and little if any difference in south-
eastern Arizona. In the latter region, Aumph-
reysii appears to occupy mountain slopes and
cochisei the valleys (Fig. 5, p. 81), but this is
probably the result of inadequate data on cochisei
and the paucity of collections of humphreysii in
the vicinity. Not far to the east and south
humphreysii has been found at lower elevations
in desert and desert grassland habitats. Similarly,
tridens replaces humphreysii in northern Sinaloa
where the two occupy similar habitats, across a
gap of only a few miles of desert vegetation in
which either could apparently exist.

The northern portion of the range of palo-
viridis extends deep into that of humphreysii, yet
the two are sympatric in the same habitat at
only one locality, Estacion Luis, 40 mi SE Navo-
joa. Here, along a stream, the vegetation was
dense and weedy, and trees formed an almost
closed canopy much like thorn forest, the typical
habitat of paloviridis. In a similar but drier and
more open habitat 18 mi N Guaymas, paloviridis

75

has been found alone in several different years,
and once in considerable numbers. This is the
northernmost colony of paloviridis in a habitat
which appears to be more suitable for humph-
reysti. The nearest humphreysii records are 20
miles north and 13 miles south, but intervening
areas have not been investigated. South of Guay-
mas, paloviridis has again been found alone at
four localities between 8 and 21 miles southeast
of that city, and humphreysii alone only 1.6
miles farther south. (A questionable record of
humphreysii at 17 mi SE Guaymas lies within
the range of this paloviridis colony.) In this area,
near sea level, the two species were found sepa-
rately in a habitat of scattered low but often
dense bushes, mesquite trees, saltbush, and road-
side weeds and low bushes. It is interesting to
note that humphreysii 1s very small and brownish
at 22.7 mi SE Guaymas just as are some palo-
viridis between 8 and 21 miSE Guaymas, whereas
both species are green or greenish and larger to
the north and the south.

Barytettix humphreysii appears to be a low
altitude replacement for the closely related genus
Conalcaea, at least in Arizona. Three of the four
northernmost localities for humphreysii (Clark-
dale, Phoenix and Globe, from Gurney, 1951) lie
below 4000 ft. and the species probably occurs
in desert or subdesert habitats there. No Conal-
caea have been reported from the Clarkdale
locality, but to the west, Conalcaea huachucana
covoterae has been reported as common at Pres-
cott at around 5400 ft. (Gurney, 1951), and to
the east, Conalcaea cantralli has been found
above Pine and north of Payson between 4600
and 7200 ft. Not far south of Globe covoterae
has been collected at about 5-6000 ft. in the
Pinal Mts. Surprisingly, humphreysii has been
collected twice at or near Prescott, once appar-
ently with coyoterae by Poling and Kusche in
1917. These specimens are depauperate in size
and should have been collected at over 5000 ft.
in pine-oak woodland or grassland if they are
labelled properly. However, there is a possibility
that material labelled Prescott came from a wide
radius of that city and was actually collected in
valleys and canyons at considerably lower eleva-
tions. Another questionable record of sympatric
occurrence of these two species is Bruner’s
(1908) record of coyvoterae (as neomexicana,
see Hebard, 1935) from Phoenix where humph-
reysii has also been found. This record, from
1100 ft., is far lower in elevation than any other

76

for the genus Conalcaea, and if the specimens
were not mislabelled, they almost certainly came
from nearby mountains. In southeastern Arizona
the distributions of B. humphreysii and C. h.
huachucana are coextensive. However, at all
localities for which we have good data, huach-
ucana occupies higher elevations and humphreysti
lower elevations with some overlap in the mid-
dle. Thus in a transect on the road from the
desert floor to 7500 ft. above Carr Canyon,
Cohn found only huachucana at 7500 and 6700
ft., both Auwachucana and humphreysii at 6250
and 5800 ft., and only humphreysii at nearby
localities on the desert grassland floor (Fry and
Sonoita) at 4700 ft. In the Santa Rita Mts. we
have large series of huachucana alone at over
6000 ft. in the upper part of Madera Canyon and
large series of humphreysii alone below 5000 ft.
in the lower part. In the Santa Catalina Mts. we
have a large series of huachucana alone from
Summerhaven at 8000 ft., a small series of
humphreysii alone from 4-6000 ft., and a large
series of humphreysii alone from Sabino Canyon
at the base of the mountains at 3000 ft. Simi-
larly, huachucana has been found at higher eleva-
tions in the Chiricahua Mts., overlaps humph-
reysii in Cave Creek Canyon at 5360 ft. (South-
western Research Station), and is absent from
the mouth of the canyon at 4900 ft. near Portal
where humphreysii is abundant. The difference
in habitat of the two species is strikingly ap-
parent in a side canyon (South Fork of Cave
Creek) at 5280 ft., 4.5 mi SW Portal. Here in
the moist shady creek bottom with rich leaf
litter huachucana was common and humphreysii
conspicuously absent, although the latter was
present only a few miles up the main canyon and
abundant 3 miles down the canyon at its mouth.
We have no evidence of competitive exclusion
between these two species, but Madera Canyon
would be an ideal locality to study such a possi-
bility because of its width and ease of access.

Much material of Barytettix humphreysii and
of various species of Conalcaea from Arizona
has accumulated at the Academy of Natural
Sciences of Philadelphia. We have only super-
ficially examined these series, and they may hold
the clues to some of the habitat and distribu-
tional problems in that state.

SEASONAL OCCURRENCE.— Throughout — its
range hMumphreysii is common in late summer
and fall. We suspect that the eggs, at least in
México, are stimulated to complete their devel-

T. J. COHN AND I. J. CONTRALL

opment by the first summer rains, which begin
abruptly at the end of June, and that they hatch
shortly thereafter. In 1970, only middle instar
nymphs were present in the Imuris region, Son-
ora, on 22 July, and near Portal, Arizona, on
18 July. By 3 August a few adults (some teneral)
were found among the nymphs at Imuris. In
1968, at 39 mi S Hermosillo, late instar nymphs
were abundant and adults few and teneral as late
as 21 August. Local residents informed us that
the rains were late that year. The earliest sum-
mer record for adults is 17 July reported by Ball
et al (1942) for Arizona. Our latest record is 2
December, 2 mi W Alamos, Sonora, when the
species was common. We know of no winter
collecting specifically for Barytettix, and thus do
not know how much later into the winter the
adults persist, or whether they overwinter. Ball
et al(1942) state that eggs overwinter in Arizona,
but there is evidence of overwintering adults or a
spring generation in the Baboquivari Mts. Sev-
eral males and females were collected there by
O. C. Poling in April 1924. To confirm these
records, we have examined the correspondence
between Poling and T. H. Hubbell who bought
and labelled this material. Poling reported
abundant Orthoptera on | December 1923, and
on 4 April 1924 wrote that he had been accumu-
lating Orthoptera ever since December. This is
an area of some winter rainfall (Dunbier, 1968),
and Poling reported short, mild winters. It is
therefore possible that the April adults were pro-
duced from fall or winter eggs, hatched by
winter rains, the nymphs developing during a
warm winter and early spring. Unfortunately,
there is no information on the diapause require-
ments of the species which might limit such a
probability. To the south winter rains decrease
and a spring generation is increasingly improbable.

DISTRIBUTION.— This subspecies is presently
known from the southeastern quarter of Arizona
and adjacent portions of New Mexico and Chi-
huahua, through Sonora to northernmost Sina-
loa. Limiting records are as follows: northern,
Verde Valley, directly north of Clarkdale, Yava-
pai Co., Arizona (from Gurney, 1951); western,
Alamo Canyon, Ajo Mts., Pima Co., Arizona
(from Gurney, 1951); southern, Cerro Prieto,
23.3 mi N Los Mochis, Sinaloa; eastern, 1.8 mi
W Casas Grandes, Chihuahua. It is absent from
the southeastern corner of Arizona, where /.
cochisei is found (see Fig. 5, p. 81, on which
all records of both subspecies and the hybrids

CONALCAEINE GRASSHOPPERS: BARYTETTIX

between them are indicated). A questionable
record from Douglas is discussed under cochisei.
The eastern slopes of the Sierra Madre Occt-
dental have been inadequately investigated, and
the western slopes have been penetrated at only
three places: east of Alamos, northeast of Cd.
Obregon, and near Cananea, all in Sonora.

The range of h. humphreysii is interrupted by
h. cochisei between Hermosillo and 14.6 miles
south of that city. The only two males in a col-
lection at 30 miles south of Hermosillo are
hybrids, and a larger series at 36.8 mi S Hermo-
sillo contains hybrids and specimens almost but
not quite typical of the next h. humphrevysii
colony at 39 mi S Hermosillo (see Table 17, p.
82). The region east of this hybrid zone has not
been investigated. The range of this subspecies is
apparently also interrupted by colonies of palo-
viridis in the Guaymas region. This is more fully
discussed under Habitat.

The most puzzling feature in the distribution
pattern of h. humphreysii is the presence of a
colony near Casas Grandes in northwestern Chi-
huahua. Not only is this the only record of a
Barytettix east of the Sierra Madre Occidental,
but its presence there suggests that h. humph-
reysii completely surrounds h. cochisei. A num-
ber of the characteristics of this colony more
closely match those of colonies south of Hermo-
sillo than the features of nearby Arizona humph-
reysii (dorsal and ventral aedeagal valves, ventral
lobe flange and width, see Table 14 under S
Hermosillo). This suggests to us that the Casas
Grandes colony is a recent invader from south
of the range of cochisei. The few Arizona char-
acteristics of this colony (all color features, lip-
ping and apical shape of ventral lobes, see Table
14 under Central Arizona) might be the result
of even more recent contact and intergradation
with the Arizona populations through New Mex-
ico. No collecting has been done in either south-
western New Mexico, or in the Sierra Madre
Occidental east of Hermosillo, where clues to
this invasion might be found. The river valleys
which dissect much of the northern Sierra Madre
Occidental might have served as migration routes
for humphreysii. Alternatively, cochisei might
have originated as a northern isolate, and the
characteristics of the Casas Grandes colony
could represent the original combination of
features in humphreysii south of cochisei in the
Hermosillo-Guaymas region. However, the par-
allel north-south geographic variation of the two

77

subspecies makes it more likely that cochisei
was a midwestern isolate and that humphreysii
has only recently migrated around it to the east.

RECORDS.— Specimens examined: 392 7, 3672 2, 12
juveniles and several reared to maturity. MEXICO: A complete
list of Mexican records of this subspecies will be found in Table
19, p.104, and Table 17, p.82.; see also Fig. 5, p. 81. Hybrids
between fh. humphreysti and h. cochisei are listed under h.
cochisei, p. 84, and in Table 17, p. 82. ARIZONA: The follow-
ing Arizona material has been studied by us. (We have not in-
cluded ANSP material cursorily studied, or records from Gurney
(1951).) YAVAPAI CO., Granite Peak, Prescott (ANSP) (1),
Prescott (ANSP) (1); GILA CO,, Globe (1); Pima Co., 15-20 mi
S Summerhaven, Sta. Catalina Mts., 4000-6000 ft. (AMNH,
UMMZ) (5), 22 mi SW Redington, 3500-4000 ft., (AMNH,
UMMZ) (2), Sabino Canyon, Sta. Catalina Mts., 2600-3500 ft.
(AMNH, UMMZ) (29), Baboquivari Mts., 4000 ft. (10), Babo-
quivari Mts., 5000 ft. (17), Sycamore Canyon, Baboquivari Mts.,
3700-4200 ft. (S), El Mirador Ranch, 4 mi SW Sasabe, Babo-
quivari Mts., 3900 ft. (AMNH, UMMZ) (8), mouth Madera Can-
yon, Sta. Rita Mts., 4500 ft. (AMNH, UMMZ) (5), Lower Madera
Canyon, Sta. Rita Mts., 4500-5000 ft. (AMNH, UMMZ) (24), 11
mi N Sonoita, 4700 ft. (AMNH, UMMZ) (2); SANTA CRUZ
CO., 4 mi E Arivaca (7), | mi E Ruby (1), Yanks Springs, 4 mi
SE Ruby, Pajarito Mts., 4000 ft. (1), 21 mi SE Ruby (AMNH,
UMMZ) (6), 61 Ranch, 15 mi SE Ruby, 4000 ft. (AMNH,
UMMZ) (3), Nogales (S), Florida Canyon, 9 mi E Nogales, 3800
ft. (AMNH, UMMZ) (5); COCHISE CO., 8 mi S Fry, 4725 ft. (5),
Huachuca Mts., above Carr Canyon, 4 rd mi W Fletcher’s Round-
up, 6250 ft. (1), 3.5 rd mi W Fletcher’s Roundup, 6025 ft.
(AMNH, UMMZ) (4), 3 rd mi W Fletcher’s Roundup, 5800 ft.
(AMNH, UMMZ) (18), Montezuma Pass, Huachuca Mts., 6500 ft.
(1), 5 mi E Montezuma Pass, Huachuca Mts., 5000 ft. (AMNH,
UMMZ) (3), 3 mi N Bisbee, Mule Mts., 5500 ft. (AMNH, UMMZ)
(2), 7 mi N Gleeson (1), 2 mi W Chiricahua Nat. Mon. (4), 5 mi E
Cochise Stronghold (5S), Paradise graveyard, Silver Creek Road, 5
mi NW Portal (ANSP) (16), Southwestern Research Station, Cave
Creek Canyon, Chiricahua Mts., 5400-5500 ft. (ANSP) (3), 1.3
mi SW Portal, 4900 ft. (16).

Barytettix humphreysii cochisei
Gurney n. comb.
Figs. 6 A;8 F;10 A-E; 11 A-B; 12 A-B;17C, F
Barytettix cochisei Gurney, 1951, Proc. U.S. Nat. Mus. 101:292-
296; Figs. 58 b, d, f: 64 a, b, i, m; 65; P1, 10, Fig. 2. [Holotype

o Arizona, Cochise County, Douglas; United States National
Museum Type No. 59155.]

DIAGNOSIS.— Males of this subspecies may be
distinguished from the nominate form, as well as
from all other members of the genus, by the wide
U-shaped excision of the dorsal aedeagal valves
and the narrow, parallel free lobes thus produced
(Figs. 8 F, 10 A-E), and by the straight or slightly
diverging medial margins, blunt apices, strong
proximal shoulders and conspicously concave
lateral margins of the ventral aedeagal valves
(Figs. 8 F, 11 A-B). The receptaculum seminis
of the female (Figs. 17 C, F) is also distinctive
in the following characteristics: the bursa is
short, weakly sclerotized, and triangular at its
apex; the thick tube arises disto-dorsally and

78

proceeds dorso-cephalad, and is narrow along
its entire length. In Arizona the dorsal aedeagal
valves of cochisei often have a shallow U-shaped
medial notch at the base of the excision, and
thus may also have small teeth there. In humph-
reysii these teeth, when present, are almost al-
ways near the middle of the medial margins of
the free lobes. The proximal flange on the ven-
tral lobes of the aedeagal sheath is moderate,
weak, or absent in cochisei, whereas the flange is
very strongly developed in all Arizona and nor-
thernmost Sonoran humphreysii. South of Her-
mosillo this flange is usually weak or absent in
both subspecies.

In color characteristics cochisei is identical to
nearby humphreysii colonies in both the Douglas
and Hermosillo regions. Elsewhere, humphreysii
is often distinctively different in ground and
tibial color (see Table 14, p.68). The only other
brown species of Barytettix are members of the
Psolus Group in which the distinctive pronotal
color characteristics (Table 6, p. 34) are quite
different from cochisei. Brown individuals of
paloviridis in the Guaymas area are also similar
to cochisei, but have dull reddish tibiae, whereas
the tibiae in cochisei in this region always have
some purple. Males of cochisei may always be
distinguished from paloviridis by the more ex-
panded and twisted cerci, and the lack of an
enlarged pallium in the former species (Fig. 15 H).

SUBSPECIES DESCRIPTION.— CERCUS (Fig. 6 A):
moderately incurved and twisted; dorsal margin moderately con-
cave; disto-dorsal portion moderately enlarged; disto-ventral
tooth moderately to strongly produced (but less so than crassus),
sometimes weakly so, acute angulate.e AEDEAGUS (Fig. 8 F):
moderately elongate (0.11-0.14 times length of pronotum),
slightly wider than long (length 0.74-0.87 times width), DORSAL
VALVES (Figs. 10 A-E): convex from side to side; free lobes
narrow, parallel, medial margins smooth or with basal teeth, usu-
ally more or less parallel in distal two-thirds, sometimes slightly
divergent, lateral margins weakly convex to almost straight,
apices narrow, blunt acute. WENTRAL VALVES: in cepahalic
view (Fig. 8 F), dorso-lateral portions widely separated, straight
or slightly concave, exposing most of lateral and ventral wall of
phallotreme; in caudal view (Figs. 11 A-B), always with a promi-
nent shoulder proximad, lateral margins conspicously concave
distad, medial margins straight and slightly divergent, in south
distal eighth usually convex, apices narrowly rounded, rarely
briefly sharp acute, ventral surface longitudinally concave; in side
view, almost straight in north to slightly sinuate in south, thin
distad, slightly thicker proximad at shoulder. VENTRAL LOBES
OF SHEATH (Figs. 12 A-B): strongly swollen along midline in
proximal two-thirds in north, usually more or less flat in south
(but whole valve slightly convex from side to side), sometimes
weakly swollen along midline; proximal portion flat or moder-
ately flanged; distad of junction of lobes with dorsal valves not
at all to slightly produced; apices broadly rounded, truncate,
straight across, or forming a broad V; apical portion usually lip-
ped or swollen, rarely flat. BURSA COPULATRIX AND THICK
TUBE (Figs. 17 C, F): about as long as wide (length 1.1 times

T. J. COHN AND I. J. CANTRALL

width), apex triangular, thick tube arising disto-dorsally. COLOR-
ATION: ground color gray or gray-brown (rarely greenish) with
little yellow in north (Gurney, 1951, reports “‘variable intensity
of olivacous green”) (usually fading to brown in most dried spe-
cimens), gray-green with much yellow in south. Head with
yellow post-ocular stripes narrow in north, usually wide in south.
Pronotum with dorso-lateral yellow stripes narrow in north, wide
in south, sometimes extending caudad briefly onto metazona;
pronotal color and black spot on lateral lobes as in species de-
scription. Hind femur with pagina always with a dorsal blackish,
brownish or gray dorsal stripe, and a yellow ventral stripe. Hind
tibia red, orange, or sometimes yellow (Rodeo), red or sometimes
orange (Douglas region), or wine or purplish (Hermosillo region);
rarely bicolored in north (one specimen, possibly poorly pre-
served), sometimes bicolored, darker proximad, in south. MEAS-
UREMENTS (in mm): those of the series studied are summarized
in Figures 19 and 20.

GEOGRAPHIC VARIATION.— The strong north-
south variation displayed by this subspecies in
many of its characters is summarized in Table 15.
We have as yet no material between the Hermo-
sillo region and southeastern Arizona, and thus
cannot determine whether this variation is clinal,
or where the breaks occur if it is not clinal. A
striking aspect of this variation is the parallelism
with that in the nominate subspecies (indicated
in the table by asterisks), some of which may be
ascribed to introgression from either subspecies
(see under Geographic Variation in h. humph-
reysii).

Within the southeastern Arizona colonies (in-
cluding those from nearby New Mexico and
Sonora as above) additional variation is apparent
but we do not as yet have large enough series
from south of Rodeo to be certain of the pat-
tern. Color variation of the hind tibia is sum-
marized in Table 16. Moderately developed or
weak proximal flanges of the ventral lobes of the
aedeagal sheath are found in equal numbers in
Rodeo males, whereas specimens from the Doug-
las area usually have weak flanges or none at all.
In the dorsal aedeagal valves all Rodeo males
have prominent median teeth separated by a
moderately deep notch (Fig. 10 A), and the free
lobes taper weakly if at all and are apically blunt.
Specimens from the San Bernardino Ranch have
virtually no median teeth, and the lobes are
more tapering and more acute (as in Fig. 10 C).
The lobes in Perilla Mts. and Agua Prieta males
are similarly shaped, but are variable and have
small teeth (Fig. 10 B). Paratype males from
Douglas are quite variable in the shape of the
free lobes and the nature of the median teeth,
which more or less encompass the variation in
the other colonies (Figs. 10 C, D), and have vary-
ing combinations of these characteristics.

CONALCAEINE GRASSHOPPERS: BARYTETTIX

Some of the characteristics of the Rodeo col-
ony might be ascribed to introgression from h.
humphreysii, the nearest known colony of which
is located at Portal, only 10 miles to the west
(Fig. 5). The yellow and orange tibiae of the Ro-

TABLE 15

79

deo colony are unique in cochisei, but yellow
tibiae are the rule in Arizonah. humphreysii. The
blunt apices and more prominent median teeth
of the dorsal valves, and the flange of the ven-
tral lobes may also represent h. humphreysii

GEOGRAPHIC VARIATION IN BARYTETTIX HUMPHREYSII COCHISEI

Ghnrneis SOUTHEASTERN ARIZONA WEST-CENTRAL SONORA
Byes (Agua Prieta-Rodeo Region) (Hermosillo Region)
CERCUS *Less incurved and twisted *More incurved and twisted
AEDEAGUS
Shape Long and narrow Short and broad

Dorsal valves

Ventral valves

Ventral lobe of sheath

COLOR
Ground color

Pronotal dorso-longitudinal
yellow stripes

Hind tibial color

Hind tibial color pattern

Free lobes usually narrow, apices
usually more acute, base of median
excision usually with a brief notch

Longer and lengthily blunt acute
*Lateral margins less concave
Medial margins straight to apices
*Long and narrow (lobes together
quadrate)

*Distal margins usually truncate

*Swollen medially

*Gray or brownish

*Narrow and inconspicuous

*Red, orange, yellow

*Unicolored (rarely bicolored)

Free lobes wide, apices more
blunt, base of median excision
usually smooth

Shorter and less acute
*Lateral margins more concave

Medial margins convex near
apices

*Short and broad (lobes
together much broader than long)

*Distal margins together
forming an open V

*Flat from side to side

*Gray-green or olive

*Wide and bright

*Purplish

*Usually bicolored

*Indicates parallel variation of humphreysii humphreysii

80 T. J. COHN AND I. J. CANTRALL
TABLE 16
TIBIAL COLOR VARIATION IN BARYTETTIX HUMPHREYSII COCHISEI
Tent Male Female
raat Red Orange Yellow Red Orange Yellow
1.3 mi E Rodeo 11 11 3 14 4 1
2 mi SW Chiricahua 3 0 0 2 0) 6)
San Bernardino Ranch;
S end Perilla Mts;
7-10 mi SE Agua Prieta 11 0 0 21 0 0
Douglas 3 1+(1) (1) 4 0) 0

(_) Indicates an individual with each hind tibia a different color

influences as might be the rare occurrence of
acute apices and only barely concave outer mar-
gins of the ventral valves. Interestingly, there is
only the vaguest hint of cochisei influence on the
Portal colony of h. humphreysii. One specimen of
the eleven males studied from that colony pos-
sesses strong proximal shoulders of the ventral
valves, but is otherwise typical of humphreysii.
None of the Rodeo specimens is intermediate
between the two subspecies, and there is little of
the variation (except in tibial color) which
characterizes recent hybridization. Further-
more it is possible that the yellow tibiae in the
Rodeo cochisei represent a trend parallel to that
inh. humphreysii which varies from yellow in the
north, through red, to purple in the south. This
and other parallel trends found in the two sub-
species are indicated in Table 15.

HYBRID POPULATIONS.— Our decision to re-
duce cochisei to subspecific status is based ex-
clusively on our interpretation that two widely
separated groups of populations are of hybrid
origin. Each group, one in Arizona and the other
in west-central Sonora, lies between popula-
tions of typical cochisei and typical humphreysii
(see Figure 5 and Table 17).

In a small collection from Don Luis (3 mi S
Bisbee, Arizona, including one specimen from
nearby Osborn or Bisbee Jct.) six of the seven
males have the ventral aedeagal valves slightly
incurved distally, and in all, the apices are sharp-

ly acute (Fig. 11 C). No cochisei from this re-
gion (Agua Prieta-Rodeo) show any tendency
toward this condition, which is somewhat simil-
ar to that in nearby humphreysii. In addition,
two of the Don Luis males show a straightening
of the concavity of the lateral margin of the
valves. There is more variation in this character
in nearby cochisei populations, but no single
cochisei exactly matches this condition, which
appears to be intermediate between the typical
cochisei and humphreysii conditions. In their
dorsal valves, the Don Luis specimens display a
certain amount of variation which is hard to as-
sess. All have more or less prominent, but basal
medial teeth, and in all, the free lobes are
relatively narrow. In several specimens, how-
ever, the lobes are somewhat broader and blunt-
er than are found in cochisei, and in all Don Luis
specimens the medial margins are slightly di-
vergent beyond the teeth. All are conditions
somewhat similar to those in humphreysii. These
characteristics of the dorsal and ventral valves
suggest the influence of humphrevsii, but in each
character the condition is closer to cochisei than
to humphrevsii. Unfortunately, we have no series
of humphreysii within about 20 miles of this
locality against which we might check these
characters, although a single male six miles to
the northwest is typical of the humphreysii in the
region. It should be noted that in the ventral
valves of the Don Luis specimens, there is sur-
prisingly little variation in the hybrid character-

CONALCAEINE GRASSHOPPERS: BARYTETTIX

*PEARCE

Dragoon

Mts. 1e

“Whetstone

"GLEESON

y. FORT

HUACHUCA
+

+

Mule

2
2Carr Canyon 1@ Mts
ae = BISBEE

[eo LUIS.

'@ossorn

Huachuca

1 1@
Mis. @ Montezuma Pass

Sierra de |
San José

eCANANEA

Sierra

de fos Ajos

ne

Sierra
Manzanal

Figure 5.

6@
11@ PORTAL

3
SW Research Sta

Chiricahua

Mts.

cy CHIRICAHUA

Perilla
Mts

San
Bernardino

oy
“A DOUGLAS SAsiRench

FaGa PRIETA

7

LEGEND

@ humphreysii
A cochisei
@ hybrid

like cochisei

O hybrid

like humpreysii

OVER 5,000 FEET
5 10
MILES

Distribution of Barytettix humphreysii humphreysii and h. cochisei in southeastern Arizona, adjacent New Mexico and

Sonora. Large arrows indicate h. humphreysti collections at localities lying outside the limits of the map. Numbers accompanying

symbols indicate number of male specimens from each locality.

istics. There is also some evidence of hybridity in
a small collection about 10 miles south of Don
Luis in the Sierra de San José near Naco. The
four males from this locality are very similar to
humphreysii from near Cananea (about 15 to 25
miles to the southwest). However, the ventral
valves in one male are apically acute rather than
acuminate and they lack the distinctive angula-
tion in the medial margin (similar to Fig. 11 D)
found in all the Cananea specimens. None of the
Naco males have the medial teeth of the dorsal
valves as prominent as those in most of the

Cananea specimens, but there is considerable
variation in the latter.

In the second group of presumed hybrid
populations (see Table 17), that at 0.7 mi S Los
Pocitos (36.8 mi S Hermosillo Cathedral) dis-
plays much more variation. The condition of
both the dorsal and ventral valves in the eight
males from this locality range from almost but
not quite typical of the nearest huwmphreysii
population, 1.9 miles to the south, to almost
precisely intermediate. In two of the Los Pocitos
males, the dorsal valves are almost indistin-

T. J. COHN AND I. J. CANTRALL

TABLE 17

DISTRIBUTION OF BARYTETTIX

HUMPHREYSIT HUMPHREYSII

AND H. COCHISEI IN THE HERMOSILLO REGION

cochisei ( 2.7 mi S Hermosillo Cathedral) ! 34.7 mi N Los Pocitos 2 Orn (62
cochisei 7.0 mi S Hermosillo Cathedral 30.4 mi N Los Pocitos 6c

cochisei 10.4 mi S Hermosillo (26.7 mi N Los Pocitos)! os api)
cochisei 14.6 mi S Hermosillo Cathedral 22.6 mi N Los Pocitos 1A?
HYBRID 30.0 mi S Hermosillo Cathedral 7.5 mi N Los Pocitos Dae | Dae
HY BRID (36.8 mi S Hermosillo Cathedral) | 0.7 mi N Los Pocitos Sto eae
humphreysii 39.0 mi S Hermosillo Cathedral 1.2 mi S Los Pocitos shot il ©
humphreysii 40.8 mi S Hermosillo Cathedral 3.3 mi S Los Pocitos 4c 109
humphreysii 43.0 mi S Hermosillo Cathedral 5.6 mi S Los Pocitos eh 9)
humphreysii 48.0 mi S Hermosillo 11.0 mi S Los Pocitos ily; 3

I Distances in parentheses indicate calculated mileage data.

guishable from those of the nearby /humph-
reysii, and the ventral valves differ from those
humphreysii only in the shorter and less acu-
minate apices. In the remaining six males, the
ventral valves (Fig. 11 D) are intermediate be-
tween the condition in the nearest cochisei
population, 22 miles to the north (Fig. 11 B),
and the previously mentioned humphreysii pop-
ulation (similar to Fig. 11 G). The inner
margins of the ventral valves in the Pocitos
males are divergent from near the base and not
or barely incurved, the outer margins are con-
vex or barely concave, the apices are lengthily
blunt acute, and the shoulders in side view are
slightly thickened or thin (as in cochisei). The
dorsal valves always have the inner margins of
the free lobes divergent as in humphreysii, but
the median teeth vary from almost non-existent
(almost as in cochisei) through prominent but
short (a condition found in neither form in the
area) to prominent and typical of humphreysii.
At 7.5 mi N Los Pocitos the only two males in
the collection also appear to be hybrid in having

the lateral margins of the ventral aedeagal valves
almost straight or only weakly concave, the
apices sharp but not acuminate, and the medial
margins straight to the apices rather than con-
vex distally or incurved. The dorsal aedeagal
valves have short, inconspicuous teeth, and in
one the medial margins are divergent. The two
specimens are much more similar to cochisei
than to humphrevsii in the sum of their
characteristics.

RELATIONSHIP BETWEEN COCHISEI AND HUM-
PHREYSII.— The data discussed above indicate
that these two forms are hybridizing, that
backcrossing is taking place, and that hybrids of
all types are surviving in nature. The definition
of subspecies is therefore satisfied, and we are
treating cochisei and humphreysii as subspecies.

These same data, however, indicate that a
more complicated situation may obtain. The
most obvious problem in treating the two forms
as simple subspecies is the interruption of the
range of humphreysii by the pocket of cochisei

CONALCAEINE GRASSHOPPERS: BARYTETTIX

south of Hermosillo. It seems likely that hum-
phreysii occupied this pocket until recently in
view of its abundance to the north and south in
similar habitats. If this is true, then cochisei must
be invading the range of humphreysii and re-
placing it, indicating that while the two forms
may be reproductively compatible in the short
run, one or both of the parental forms may be
superior to the hybrids in the long run, thus the
two may be effectively isolated reproductively
and represent separate species. This may also be
an explanation for the narrow hybrid zones in
Arizona and near Hermosillo. In both areas
where hybrids are known, the distance between
the two parental forms cannot be more than
about 25 miles, and further collecting in the gaps
will probably reduce this distance (see Fig. 5, p.
81, for Arizona, and Table 17 for Hermosillo,
and note the gaps between the parental forms
and the hybrids). Furthermore, in the Portal-
Rodeo area the two forms are separated by
only 10 miles and neither population shows
much morphological influence on the other.
There appear to be no major physical or
vegetational barriers to the movement of these
grasshoppers in these areas, and one would ex-
pect relatively free gene flow. Finally, the
uniformity of the shape of the ventral aedeagal
valves in the hybrid Don Luis-Oborn popula-
tion suggests some sort of stabilization of hy-
bridization, rather than the free gene flow ex-
pected if there were complete compatibility of
the two forms. But under this concept, that the
two represent species, one would expect parent-
al types to outnumber hybrids wherever hy-
bridization is occurring and one species to be re-
placing the other. This is not the case at two of
the hybrid localities where we have the largest,
though still inadequate series. At Don Luis-
Osborn all eight males are hybrid, and at 30 and
36.8 miS Hermosillo, 10 of the 12 males are hy-
brid, and the two of parental type are humph-
reysii, not the presumably replacing cochisei. At
the third locality, the Sierra de San José, only
one male is hybrid (although all may show some
effect of hybridization) and again the parental
types are humphreysii. Further data from south
of Hermosillo indicate cochisei influence in
humphreysii populations considerably south of
the zone of active hybridization at least along
the main highway (see Table 14 and the discus-
sion under Geographic Variation of h. humph-
reysii). These data, and those on humphreysii

83

parental types in hybrid colonies, might be con-
sidered to indicate a long history of hybrid-
ization during which cochisei was slowly eli-
minated from all but a pocket of the region
south of Hermosillo. On the other hand, some of
these data might be explained on the basis of
recency of contact between subspecies. But it
seems inherently improbable that the two forms
have come into recent contact at the same time
at two points 175 miles apart and 2700 feet dif-
ferent in elevation. Nor does this hypothesis ex-
plain the uniformity of the Don Luis-Osborn hy-
brids. Recent contact between reproductively
compatible subspecies would be expected to lead
to highly variable characters.

Although much more distributional data are
necessary, the present information suggests broad
contiguity between the two forms and narrow
zones of intergradation. This situation may
thus be an example of stasipatric speciation,
a theory offered by White, Blackith, Blackith,
and Cheney (1967) and modified by Key (1968).
Cytological and breeding data, and more exten-
sive information on natural hybridization will be
necessary to test this hypothesis. The 1970 lab-
oratory work indicates these forms will cross
readily and both adults and nymphs are easy to
rear. The area in Arizona where the two are con-
tiguous is easily accessible and apparently not
seriously affected by man. In Sonora, the areas
of contiguity near Hermosillo and possibly in the
Cananea-Agua Prieta region are also easily
accessible.

The recently discovered hybrids between
cochisei and humphreysii bring into question the
efficacy of mechanical genitalic isolation in these
two forms, in spite of major differences in their
aedeagi and bursae. We have too few collec-
tions as yet to determine whether the hybridiza-
tion is extensive in nature, or the result of single
accidents followed by extensive backcrossing.
This problem is discussed further in a later sec-
tion on Mechanical Isolation.

HABITAT AND SEASONAL OCCURRENCE.— _ AIl
known localities for h. cochisei in Arizona lie
in valleys or the low foothills of mountains
between 3780 and 4650 feet, in areas mapped
as desert or desert grassland. Cochisei has not
been recorded as occurring in mountains, but
this may be the result of a lack of collections
in the area. At the Rodeo locality it was
common in thin weeds in a cemetery and among
low bush mesquite on the sandy outwash slopes

84

of the Peloncillo Mts. This habitat did not ap-
pear grossly different from that occupied by h.
humphreysii only 10 miles across the valley at
Portal (see Fig. 5), although it was drier and the
vegetation sparser. Southeast of Agua Prieta
cochisei was found in foothills covered with
Baccharis grass and acacia (V. Roth). It is per-
haps significant that the localities for the north-
ern hybrids, Don Luis, Osborn and the Sierra de
San José, at 4700 to 5150 feet, are all above the
known upper altitudinal limits for the sub-
species. The habitat at Don Luis included short
grass, creosote bush and ocotillo (from Gurney,
1951, probably from the field notes of Rehn and
Hebard, the collectors). South of Hermosillo,
cochisei is common in roadside bushes and
weeds, especially pigweed, in rich desert or
thorn scrub with tall bushes or trees of mes-
quite, paloverde, creosote, etc. At the south
edge of Hermosillo it was found in dense bushes
along a small water course. The hybrid colonies
near Los Pocitos were found in the same type of
scrub habitat with a good growth of weeds and
bushes near a water course. No obvious dif-
ference was noted between this habitat and that
of the nearest h. humphreysii colony four miles to
the south for which we have adequate field
notes. In this area, only weedier habitats were
investigated. Here cochisei was often abundant,
especially in pigweed, a situation identical to
collections and abundance of humphreysii both
to the north and to the south.

Our earliest record for adult h. cochisei is 6
August at Rodeo, and the latest, 19 October at
36.8 mi S Hermosillo (hybrids) when they were
common. Abundant late instar nymphs, but no
adults were found as late as 3 August at 14.6 mi
S Hermosillo.

DISTRIBUTION AND ASSOCIATED SPECIES.—
Cochisei is known only from a small area
centering on the southeastern corner of Arizona
(see Fig. 5), and an even smaller area just
south of Hermosillo (see Table 17). Little
collecting has been done east of the limiting
records in New Mexico, and probably no
orthopteran collecting has been done south of
the Agua Prieta record (except at Casas Grandes,
Chihuahua, where h. humphreysii has been
found) or east of Hermosillo. In Arizona
the southern end of the Sulphur Springs Valley
(north of Douglas), the Mule Mts. and the
Chiricahua Mts. have been inadequately in-
vestigated except in the Portal area. From the

T. J. COHN AND I. J. CONTRALL

Huachuca Mts. and to the west we have reason-
ably adequate samples which contain only h.
humphrevsii. One specimen of h. cochisei from
Madera Canyon, Santa Rita Mts., collected by
E.R. Tinkham, we believe to be mislabelled. We
have not found either h. cochisei or hybrids in
several other collections containing h. humph-
reysii or Conalcaea made at various points in or
near this canyon. It is suggestive, however, that
this Madera Canyon h. cochisei shows slight hy-
brid characteristics in the acute apices and bare-
ly incurved medial margins of the ventral
aedeagal valves.

Nowhere is h. cochisei sympatric with h.
humphreysii, except for the improbable Madera
Canyon record discussed above, and another at
Douglas. A single male labelled as having been
collected at the latter locality by W.W. Jones is
typical of h. humphrevsii, although the type
series of h. cochisei, also collected by Jones,
shows little or no influence of h. humphreysii
characteristics. Because most of Jones’ insect
material is labelled Douglas, it is possible that it
came from a wide area around Douglas, and that
this specimen was actually collected at one of
the nearby humphreysii localities.

All males from Don Luis and Osborn are hy-
brid and one of the males in an otherwise h.
humphreysii series from Naco shows some hybrid
characteristics. The series from 30 and 36.8 miS
Hermosillo are mostly hybrids with a few h.
humphreysii.

RECORDS.—Specimens examined: 5966, 6199, 1 juvenile
and some reared to maturity. N. MEX.: Hidalgo Co., 1.3 mi
E Rodeo (29) (ANSP, UMMZ). ARIZ.: Cochise Co., 2 mi
SW Chiricahua, San Bernardino Valley (25 mi NE Douglas)
(3) (ANSP); San Bernardino Ranch (17 mi E Douglas) (S)
(ANSP); Perilla Mts. (8 mi W Douglas) (3) (ANSP); Douglas
(8) (USNM, ANSP, UMMZ). SONORA: records are listed in
Table 17, p. 82.

Hybrids (females not examined; entire series listed even
though some males may resemble the parental types): ARIZ.:
Cochise Co., Osborn (Bisbee Jct.) (1) (ANSP); Don Luis (7),
(ANSP). SONORA: Sierra de San José, near Naco (9); 30 mi
S Hermosillo Cathedral, T.J. Cohn 1970 No. 32 (4); 36.8 mi S
Hermosillo Cathedral, T.J. Cohn 1970 No. 68 (12).

Arizona and northeastern Sonoran records of h. cochisei,
h. humphreysii, and hybrids between them are mapped in
Figure 5, p. 81.

Barytettix tridens ETS
Figs. 6 L-P; 8 H; 9 E-F; 10 S-T; 11 N-P; 12 O-S

HOLOTYPE.— *, México, Sinaloa, 17.1 mi N

17From the Latin, tridens, the three pronged spear of Neptune,
in allusion to the very similar appearance of the dorsal valves of
the male aedeagus.

CONALCAEINE GRASSHOPPERS: BARYTETTIX

Los Mochis turnoff on Highway 15 (6.2 mi S
Cerro Prieto), 17 October 1970 (T. J. and J. W.
Cohn No. 63); University of Michigan Museum
of Zoology.

We have selected the holotype of this species
from a locality only three miles south of its
northernmost limit for several reasons. First, it
is in the range of the most distinctive geo-
graphic variant of the species. Second, the spe-
cies is relatively common there. Third, the local-
ity lies in a hill mass not likely to be seriously
disturbed by agriculture or housing.

DIAGNOSIS.— Males of tridens may be dis-
tinguished from all other species of the genus,
except the southern Sonoran populations of /huz-
phreysii, by the strongly twisted and incurved
cercus with an expanded disto-dorsal portion
(Figs. 8 H, 6 L-P), by the narrow, incurved apices
of the dorsal aedeagal valves which are always
provided with acute or aciculate median teeth
(Figs. 10 S-T), and by the widely separated
finger-like structures forming the distal half of
the ventral lobes of the sheath (Figs. 12 O-S).
The brown or blue ground color of tridens will
distinguish the species from southern Sonoran
colonies of humphreysii with similar genitalic
structures but which are always green with red-
dish tibiae. Female tridens may be differentiated
from all Baryvtettix species except humphreysii
by the blue or brown ground color, blue or
brown ventral third and ventral carina of the
lateral lobes of the pronotum, the metazona
without dark stripes, and the very short bursa
copulatrix and undilated thick tube. From brown
humphreysii females, which have a similar bursa
and thick tube, females of tridens may be recog-
nized by their uniformly blue or deep purplish-
blue tibiae. For identification by color alone,
see Table 3, p.25 , comparing all Barytettix
species and Table 11, p.49 (under Geographic
Variation in psolus), dealing with the geography
of tibial color of all Barytettix species in the
Guamichil-Mazatlin region. The combination of
blue tibiae and pronotal markings described
above is unique in the genus.

Tridens may also be distinguished from hum-
phreysii on the basis of a number of character-
istics restricted to one or the other species, but
not found throughout that species. These are set
forth in Table 13, p.65 , under the humphreysii
diagnosis.

As indicated in the group analysis, p.63 , and
further discussed under relationships below,

85

p.86 , tridens appears to be a southern derivative
of humphreysii, with most of its characteristics
representing relatively small modifications of
those found in humphreysii.

SPECIES DESCRIPTION (in variable characters, condi-
tion in holotype indicated by asterisk preceding that condition).—
CERCUS (Figs. 6 L-P, 8 H): strongly incurved and twisted so
that disto-dorsal portion lies horizontally when in normal rest-
ing position (Fig. 8 H); dorsal margin deeply concave; disto-dorsal
portion usually strongly enlarged, sometimes *forming a rounded
flap, rarely only moderately enlarged; disto-ventral portion *rec-
tangulate and *sharp, to obtuse and rounded, rarely barely pro-
duced. AEDEAGUS (Figs. 9 E-F): short (length 0.04-0.06 times
length of pronotum), wider than long (length 0.48-0.63 times
width). DORSAL VALVES (Figs. 10 S-T): weakly convex proxi-
mad, lateral teeth slightly turned up distad, median teeth flat or
*depressed; free lobes narrow, convergent, apices narrow, *blunt
or acute angulate, lateral margins *strongly and smoothly convex
to apices, or strongly convex in proximal two-thirds thence weakly
concave or straight (but converging) to apices, each medial margin
provided at base with either a flat acute tooth much shorter than
lateral tooth, or a *depressed, strongly compressed and therefore
acuminate tooth almost as long as lateral tooth, the two median
teeth briefly separated but parallel, or *appearing partly fused,
medial margins of lateral teeth smoothly concave to apices, rarely
apices slightly curved laterad. VENTRAL VALVES: usually in-
visible in cephalic view; in caudal view (Figs. 11 N-P), medial
margins usually *evenly sinuate, concave distad, weakly convex
proximad, sometimes concave from base, lateral margins always
convex, apices acute or *acuminate; distal quarter twisted; ventral
surface sometimes *more or less flat, but usually with either a
broad longitudinal concavity appearing to result from warping of
valve, or a longitudinal excavation on medial margin and valve
convex laterad; valve *thin in north, thick in lateral half in south.
RAMUS OF CINGULUM (similar to Fig. 1 E-F): ventro-distal
portion narrow, ending considerably proximad of junction of
dorsal valves with ventral lobes of sheath, thus exposing a lengthy
portion of ventral lobes. VENTRAL LOBES OF SHEATH (Figs.
12 OS): lengthily developed proximad of junction with dorsal
valves but without strong sclerotization there, proximal portion
strongly flanged (briefly reflexed); proximal third beyond flange
flat from side to side to strongly swollen along midline; medial
margins divergent from near base or *in distal half, the lobes thus
appearing finger-like, widely separated, and slightly to *lengthily
produced distad of junction with dorsal valves; lobes usually ap-
pressed to ventral valves, sometimes *projecting from them at an
angle. BURSA COPULATRIX AND THICK TUBE (similar to
Figs. 16 E, 17 B-F): bursa slightly wider than long (length 0.77
times width), weakly sclerotized and pleated, apex truncate; thick
tube undilated, arising subdistally from dorsum of bursa and pro-
ceeding dorso-cephalad, length between proximal bends much
shorter than bursa. COLORATION (Frontispiece): ground color
*blue or brown, with varying amounts of yellow. Head with yel-
low post-ocular stripes always present, usually *wide. Pronotum
with dorso-lateral yellow stripes broad and extending briefly onto
metazona, sometimes extending as a weak yellow wash to dorso-
caudal margin of lateral lobes; black spot on lateral lobes re-
stricted to pro- and mesozona, usually completely cut by cephalic
yellow stripe which is always well-developed, caudal horizontal
yellow stripe in spot usually incomplete or absent, sometimes
*broad and complete, spot narrowly to *broadly bordered ven-
trad by yellow, ventral quarter of lobe and ventral carina usually
concolorous with metazona (brown or *blue), yellow only ina few
southern specimens which appear to have faded generally to
yellowish, never bright yellow. Abdomen with mid-dorsal stripe
usually *sharply defined but sub-obsolete in apparently poorly
preserved specimens; lateral yellow and black spots usually *well-
developed, weak in apparently poorly preserved specimens, supra-

86

anal plate, cerci. subgenital plate, and often subterminal segments
tinged with blue. Hind femur with pagina bright yellow in ventral
half, and usually with a *blue or jet-black, sharply defined stripe
in dorsal half, often weak, narrow, incomplete or absent; carinae
blue or black (both in holotype); geniculae blue with black lunae.
Hind tibia unicolored blue or *deep blue-purple (often deep
purple, or *purplish distad when seen under strong light). MEAS-
UREMENTS (in mm): holotype male: length of body, 36.7;
length of pronotum, 8.1; length of tegmen, 5.3; length of fore
femur, 6.9; length of hind femur, 19.0; maximum width of hind
femur, 5.7. Measurements of the series studied are summarized
in Figures 19 and 20.

PARATYPES.— Other than the holotype and
juveniles, all specimens examined in this study
are designated as paratypes. Male and female
paratypes are deposited in the University of
Michigan Museum of Zoology, the Academy of
Natural Sciences of Philadelphia, the United
States National Museum, the British Museum
(Natural History), and the Instituto de Biologia
of the Universidad Nacional de México.

RELATIONSHIPS.— Tridens appears to be a
southern ‘isolate of h. humphreysii. The two
forms are identical or very similar in five char-
acteristics found nowhere else in the genus:
short, truncate female bursa; sub-disto-dorsal
thick tube; incurved, aciculate ventral aedeagal
valves; dorsal valves with incurved and up-
turned lateral teeth; and greatly expanded disto-
dorsal portion of the cercus. In five characters
the condition in fridens represents an extreme
development of a north-south trend in hum-
phreysii (see discussion under Geographic Varia-
tion in that species) and each is unique in the
genus: strongly incurved cercus; obtuse or
rounded disto-ventral portion of the cercus; aci-
culate median teeth of the dorsal valves (north-
ern and southernmost tridens); narrow ventral
valves obscured by dorsal valves in dorsal view;
and narrow, lateral and elongate ventral lobes of
the sheath and their wide distal separation. The
two also share a strongly developed proximal
flange on the ventral lobes, a characteristic
found in southern and northern humphreysii but
lacking in the middle Sonoran colonies of that
species. Finally, the two are identical in two
characteristics more widespread in the genus, the
color pattern and the narrow thick tube. Only
one characteristic found throughout the range of
tridens, the shorter ventral valves, is not fore-
shadowed in the nearest humphreysii colony. All
other differences between them are found only
as distinctive geographic variants; the alterna-
tive conditions in one species are identical to
those in the other species (Table 13, p. 65).

T. J. COHN AND I. J. CONTRALL

That humphreysii and tridens represent differ-
ent species is suggested by the concordant
change in a series of characters across the 3.6
mile gap now separating the two forms. The two
closest colonies differ in the following eight
characters: red tibiae versus blue; uniformly
green paginae versus blue and yellow striped;
green ground color versus blue; longer, less in-
curved cerci versus shorter and more incurved;
acute angulate disto-ventral cercal tooth versus
obtuse- or rectangulate; acute median and lat-
eral teeth of the dorsal aedeagal valves versus
aciculate median and blunt lateral teeth; ventral
valves longer than dorsal valves versus ventral
valves shorter and completely covered by dorsal
valves. None of our material shows evidence of
intergradation in any of these characters with
the possible exception of the first two. The
northernmost ¢tridens colonies possess deep
purple-blue tibiae, which might represent a com-
bination of the blue of more southern tridens
and the red of southern humphreysii. The purple-
blue seems to be much too dark for this, and the
purple-blue colonies extend far beyond the pos-
sible influence of Aumphreysii to Guamuchil.
Perhaps more significant is the lack or lightness
of the blue stripe in some individuals from north
of Los Mochis, in colonies closest to the un-
striped humphreysii. Nearby tridens from the El
Fuerte region and from the next southern collec-
tions at Guamiichil uniformly possess well-
marked dark stripes. On the other hand, many
specimens from south of Guamtchil possess a
weak stripe or lack it altogether. In the absence
of other evidence of hybridization in the north-
ern colonies, it seems improbable that reduction
or loss of stripes is the result of introgression
from humphreysit.

The distance between the northernmost
tridens and the southernmost humphreysii col-
onies was reduced to 3.6 miles in 1970. Unfor-
tunately, our collections in the area were not
large enough for us to be certain that the forms
do not actually overlap there, and we have done
no collecting in the gap. The habitats of the two
closest colonies are different, but there is no ob-
vious barrier between them. Thus the two forms
could at least be in contact in the gap. The evi-
dence points either to reproductive isolation or
to the forms having arrived only recently in the
area and not yet having made contact. The
former seems more reasonable and we are there-

CONALCAEINE GRASSHOPPERS: BARYTETTIX

fore considering tridens and humphrevsii as dis-
tinct species.

GEOGRAPHIC VARIATION.— As might be ex-
pected of a relative of the highly variable /uw-
phreysii, this species displays marked north-
south geographic variation, summarized in
Table 18. South of Jess Maria we have no more
than three males from any one locality so that
some of the differences between samples there
may be the result of individual variation. Al-
though there seems to be considerable discord-
ance and reversals, especially in the south, a shift
in most characters takes place between Guamt-
chil and Jestis Maria across a gap of about 40
miles. A single brown female from 5 mi SE
Guamuchil (60 mi NW Culiacdn) may indicate a
much more abrupt change. A similar change
south of Guamtichil occurs in paloviridis, cor-
responding to the northern limit of the range of
psolus.

The most interesting aspect of the variation of
this species is the greater difference between
tridens and humphreysii where they are in prob-
able contact north of Los Mochis, than where
they are far separated. Thus the striking simi-
larity of the dorsal valves of southernmost /um-
phreysii (Fig. 10 R) and tridens in the Culiacan
region (Fig. 10 S, others are even more similar)
contrasts with the marked difference in tridens
north of Los Mochis. Similarly middle and
northern humphreysii and southern tridens are
brown, whereas adjacent humphreysii and _ tri-
dens are green and blue respectively. It is tempt-
ing to ascribe these differences to character rein-
forcement in the area where the two species are
in contact. However, it is not certain that the
species are sympatric at the present time or that
they were sympatric in the past. They are found
close together in a flat area (now being exten-
sively prepared for irrigation) close to sea level
in northern Sinaloa. There are extensive la-
goons (“esteros”’) nearby along the coast, and a
very slight flooding of these might easily have
isolated the ancestral species. Or the two may
have been isolated by the Rfo Fuerte, although
tridens now occurs north of the river. Not far in-
land of the area of contact the river runs almost
north and south, separated from the coast by a
low range of hills. In the recent past the Fuerte
may well have drained eastward into the la-
goons south of El Carrizo, thus separating the
two species.

87

The ventral valves are deeply excavate in the
Culiacé4n region in a manner closely similar to
that in Tezopaco humphreysii (Fig. 111). In both
cases the excavation is beneath and separated
from the ventral walls of the phallotreme, and its
function is obscure.

Tibial color has not been tabulated because of
difficulty of analysis and the possibility that
some of the differences are the result of varia-
tion in preservation. Nevertheless, there seems
to be a tendency for the tibiae to be deep pur-
plish blue in the north and lighter blue in the
south. All the material from north of Los
Mochis is purplish blue, but that from 6 mi NE
El Fuerte (actually farther north than the Los
Mochis collections but also farther inland) are
dark blue with a few slightly purplish. From 14
mi S El Fuerte to 1.4 mi SE Los Mochis, the
series are mixed purplish blue and dark blue. At
10.9 mi W Guamuchil the entire series is a
darker purplish blue than those to the north.
From that point south the tibiae are more bluish
and usually lighter. The very small samples from
15.2 mi NW Culiacdn are slightly darker, and
those from 7.2 mi SE Culiacdn and from east of
Sanalona (east of Culiacdn) are darker yet.
Under the microscope most specimens have a
slight purplish cast at least distally.

The variation in paginal striping is harder to
assess because of the probably greater effect of
different means of preservation, and possibly be-
cause of a more complex pattern of geographic
and individual variation. As indicated above,
specimens from north of Los Mochis are vari-
able and range from having almost no dark
stripes to well-marked ones. All other material
from Guamtchil to northeast of El Fuerte is uni-
formly dark striped. The single female from 60
mi NW Culiacén and all the males from Jestis
Maria (19 mi N Culiacdn) lack stripes, but three
females from the latter locality and those from
San Rafael (13 mi N Culiacan) are variable but
relatively lightly striped. Specimens from 15 mi
NW and 7.2 mi SE Culiacdn, and from east of
Sanalona (east of Culiacdn) have dark or jet
black stripes. All specimens south of that point
are light or only partly striped, with the excep-
tion of the two specimens from 6.6 mi NE
Cosalé. Many of the stripeless individuals come
from poorly preserved earlier collections and
may have lost the stripes, but those from north
of Los Mochis and two weakly striped ones from

“]Jo] 1L Sat PLoO] 0} Jayar sIOquINN y

4
—
z
5 a} RAROXO aynoe
S Ppaepua}xo JON A[peolg ayR[NOIYy juasolg ‘yioys AlIaA, aynoe SUOT uMOIg uroryng qS €L
g d}PAROXO
Z (S61)
= Ayperoiq 10 Lee)
a popuarxe Ayjoug ¥(9°9) A[daeq aye[NoIoy }UdSaIg ande ‘WIOYS —- MJR[NOIOe opi UMOIg MS S’61 - AN 9°9
oS)
5 popua}xe Aya eIOpop] YOOWS ayeynoIoy jUISa1g aynoe ‘WOYS aye[NoIy uMOIg uroryng qs c€
eH
d}BABOXA aynor
papuaxe ApRay Aldooaq ay eNO }UISIIg ‘yroys AI9A 9} B[NOTOR OPT AA uMOIg euojeurg q IW “pl gI-[
a} RALOXA aynoe aynoe = ayeynotoek apIM
popuarxe Ape Ajdaaq dieys ‘10YS }UasaI ‘yroys AIaA «10: 9B BUOT uMOIg uroelng qS TL
a} RAROXO
popusxe ATjolg Ajdoaq aynoe JOYS juasqy aynoe ‘IOUS andy uMOIg uroelyng N TSI
(papue}xa aynoe 3uo] (ueoelng N
A]}Ba15-0}11099 J ) a} BALOXA SOUITJOWIOS €1-61) OWIOIOL MA I
popuorxe ATo]e1opOyy Ajdaoq ‘aynoe WOYS JUasald aynoe ‘SuOT andy UuMOIg euRI snsof A 9
a} BIPWI9} UI ayepnoroe yuasqe
popuaixa Ape auosyJOOUIg =O}: BNP YIOYS Aqyens¢, qunyq “WOUS aye[NoIy ang pryonureny A 601
ayepnoroe
Ajaied ‘oynoe sTHDOW SOT AS t'l
popuarxe ApRory yoourg = Joys 0} SUOT UISI1g yunTq ‘WOYS aye[NoDy ang ‘sony IAS vl - AN 9
aynoe A[MOIeU jjoun}
popuajxa Apealn YyJOOuwlS 0} oJeNIIY JUISalg quny{q *}10Ys 9} R[NITIOY oni sIuSOW SOT N T'LI-6'61
SOATPA [ESIOP IM goRjing sooidy a3[ng 190} [e19}eT 190} ULIPap| 10jo) (SoyIU UT sa0ueysIp)
uonount puoAaq [e1oyeT punoly SOT}T[LIOT
saqo7T fenuoa, SATE A [PIU SOATEA [PSIOG
SNICIML XILLALAYVA NI NOILVINVA OIHdVaDOND
foe)
ioe)

8l ATaVL

CONALCAEINE GRASSHOPPERS: BARYTETTIX

the south were freeze-dried and have excellently
preserved color patterns.

HABITAT, ASSOCIATED SPECIES, AND SEASON-
AL OCCURRENCE.— This species occupies the
usual broad range of habitats as do most other
species, but it seems to be absent from weedy
fields and from wide weedy areas along the road-
side. It was not very common at any locality
south of Culiacdn, where it has always been
found with other species. It was common at
10.9 mi W Guamitchil in the edge of thorn forest
with paloviridis, at 6 mi NE and 14 mi S El
Fuerte in the edge of thorn forest, and was
common in thorn scrub in the hills 17 mi N
Los Mochis. It has been found in desert-like
thorn scrub north of Los Mochis through
heavier thorn forest where it seems to occur
sparingly. All thorn forest investigated has been

89

thinned or disturbed. It has been found with
paloviridis, psolus, contilus tectatus, contilus
dicranatus, and poecilus.

Our earliest record for tridens is 22 August
(10.9 mi W Guamichil) and our latest 28
November (64 mi S Culiacin) and 1 December
(18 mi N Los Mochis). At the latter locality the
species was common. No special collecting for
Bary tettix has been done between December and
July.

DISTRIBUTION AND RECORDS.— The north-
westernmost record for tridens is 19.9 mi N
Los Mochis, the southeasternmost, 63 mi NW
[old]Mazatlan Airport, both in Sinaloa. Speci-
mens examined: 72 6d, 51 22, 2 juveniles and
several reared to maturity. Localities are listed
in Table 19, p.104.

SPECULATIONS ON THE PHYLOGENY
OF THE SPECIES OF BARYTETTIX

A phylogeny of modern species can only be
based on the determination of primitive and
derivative conditions. This has proven difficult
in the genus Barytettix for three reasons. First,
we have been unable to identify with assurance
genera clearly related to the Conalcaeini. Sec-
ond, many of the usable characters in the genus
are relatively simple structures, making con-
vergence difficult to determine. Finally, in the
case of the elongate aedeagus in the Psolus
Group, we are not sure how many of the dis-
tinctive conditions in other aedeagal and associ-
ated bursal structures are the direct results of
elongation and not independent of it.

Despite these problems a reasonable case can
be made for the determination of primitive and
derivative conditions for three characters. The
aedeagus and bursa of all members of the Psolus
Group are much longer than almost any other
melanopline known to us (an exception is the
unrelated Proctolabus), and thus must be the
derivative condition relative to the shorter
aedeagus and bursa in the Crassus and Hum-
phreysii Groups. Similarly, and probably asso-
ciated with this character, are the elongated,
approximate ventral lobes of the sheath in the
Psolus Group. In this structure, the opposite
condition found in the Crassus Group (lobes re-
duced to a narrow collar beneath the ventral
valves) is very similar to that in Conalcaea, two

undescribed genera in the Conalcaeini, and sev-
eral other west coast Mexican melanoplines as
well. The ventral lobes, as well as the structure of
the ventral portion of the ramus of the cin-
gulum, are almost identical to that in Conalcaea
miguelitana, which is widespread on the Central
Plateau of México and is also the southernmost
species of its genus. On this basis it is reason-
able to conclude that the condition of the lobes
in the Crassus Group is most primitive, that in
the Psolus Group the most derivative, and that
in the Humphreysii Group intermediate. In the
third character the situation is not as clear. The
cercus is almost straight and weakly to moder-
ately expanded in all species but humphreysii and
tridens, in which it is greatly expanded distally
and strongly incurved. These latter conditions
are not often found in the melanoplines and are
probably derivative. However, the cerci are just
as expanded (but not incurved) in two northern
species of Conalcaea. The southern Arizonan
and Mexican members of the genus (several of
which are undescribed), and the two unde-
scribed genera of Conalcaeini, have the more
common unexpanded straight cerci. We there-
fore conclude that the tridens-humphreysii type
is derivative.

If these inferences are correct, then the Cras-
sus Group has retained the greatest number of
primitive features. It, however, has a green
color, a condition rare among melanoplines and

90

thus possibly derivative. It would be interesting
if the color of crassus is brown, making that geo-
graphically isolated species the most primitive of
the group. We are not certain of the color in that
species because of the poor state of preservation
of the available specimens. Otherwise it is im-
possible to determine which of the three species
comes closest to resembling the common ances-
tor. The curvature of the aedeagus (either down-
or out-curved) is unique and other differences
are slight.

The isolation of crassus in the southern tip of
Baja California raises a number of intriguing
problems. It may be a relict of a much wider dis-
tribution extending into Arizona and around the
head of the Gulf of California, and was forced to
retreat southward with increasing aridity in the
north during the middle and later Cenozoic. If
so, then this goes far to explain how cochisei, the
most primitive member of the group presum-
ably derived from the Crassus Group (discussed
below), has become so widely separated from its
ancestral group. However, crassus must have be-
come isolated, under this hypothesis, in the mid-
dle Cenozoic. If so, then it has differentiated re-
markably little from its mainland relatives,
poecilus and terminalis, while its derivative
groups have differentiated enormously in the
same amount of time. Alternatively, crassus may
have been isolated by the geological events
which separated at least the southern end of Baja
California from the mainland and created the
southern Gulf of California. Larson, Menard
and Smith (1968) suggest that this may have oc-
curred only within the last 4 million years, which
again seems too long for the relatively minor dif-
ferentiation seen today in the grasshoppers.
More recent studies by Gastil, Phillips, Rod-
rigues-Torres and Bonneau (1972), seem to indi-
cate an earlier geological separation; the prob-
lem is apparently under intensive study by geol-
ogists. This hypothesis does not help to explain
the geographic origin of the Humphreysii
Group, but it does offer hope for dating rates of
morphological change in the Crassus Group. In
this regard, it is interesting to note that termu-
nalis is more different from either poecilus or
crassus than poecilus is from crassus. We do not
yet have adequate collections from southern
Jalisco to determine the probable geographical
factors isolating terminalis from poecilus. If such
can be determined and their origin dated, then a
comparison of rates of morphological differ-

T. J. COHN AND I. J. CANTRALL

entiation in terminalis and crassus can be made.
Crassus could have gotten to Baja California by
long-distance passive dispersal over water, an
event which cannot be dated.

Despite its far northern position, the Hum-
phreysii Group was probably derived from a
crassus-poecilus-like ancestor. It is clearly inter-
mediate between the Crassus and Psolus Groups
in the length of its aedeagus and the develop-
ment of the ventral lobes of the sheath, and it is
difficult to envision its derivation from a Psolus
Group ancestor. The evolution of the ancestor
of the Humphreysii Group would have involved
the greater development of the ventral lobes of
the sheath (by elongation of the sheath or re-
treat of the ramus of the cingulum or both) and
the greater development of the free lobes of the
dorsal valves (by elongation of the lobes or the
deepening of the excision between them or
both), and the straightening of the ventral
valves. The most primitive member of the Hum-
phreysii Group appears to be cochisei. Unlike its
relatives, humphreysii and tridens, it shares with
the Crassus and Psolus Groups parallel, blunt
ventral valves, and relatively simple cerci. In
addition, the disto-dorsal orifice of the thick
tube of the receptaculum seminis is more simi-
lar in position to that of the Crassus Group than
is the sub-distal position in humphreysii and
tridens. The brown color may also be primitive.
Only in its broadly excised dorsal valves does
cochisei. differ markedly from what might be
considered an ancestral aedeagal condition. In
this character it shows no similarity to other
species, and is probably best regarded as a more
recent development. The cercus in cochisei is
also expanded moderately, a condition fore-
shadowing the development of a greatly en-
larged disto-dorsal lobe in humphreysii and
tridens. Humphreysii may have developed as a
western or lowland isolate from ancestral cochi-
sei, and there developed its distinctive incurved,
aciculate ventral valves, and sub-distal orifice of
the thick tube. As it moved south, it developed a
green color, and more expanded and incurved
cerci. (derivative conditions), and prominent
median teeth on the dorsal valves. Tridens
appears to represent the southern extension of
these characteristics, often in more extreme
form. The situation is not that simple, however,
because the northern colonies of tridens possess
a more extreme (derivative) condition in at least
color and dorsal valves than do the southern

CONALCAEINE GRASSHOPPERS: BARYTETTIX

populations. A few possible explanations of this
situation come to mind. In regard to the color,
the ancestral, widely distributed Aumphreysii
may have been brown, as are northern /uwi-
phreysii and southern tridens. The middle
populations subsequently developed a green
color, and after separation, northern tridens fur-
ther developed this into blue. We know nothing
of the relationships of these pigments and it
would be interesting to see if one were derivable
from the other. As for the dorsal valves, the un-
usual aciculate median processes in northern
tridens could represent a reinforcement phe-
nomenon. Unfortunately, this condition ex-
tends farsouth of any humphrevsii influence and
we are not sure if the ranges of the two species
actually overlap at the present time. It is pos-
sible that they have done so in the past, but this
seems to us to be less likely. This condition of
the dorsal valves may have no more special
significance than any other population variant.

It is interesting to note that the proximal por-
tion of the thick tube is dilated in the Crassus
Group and thin in the Humphreysii Group. We
suspect that the former condition is a derivative
one because the thick tube is thin in most other
melanoplines and that it represents an intra-
group development after the origin of the ances-
tor of cochisei. This in turn suggests that the dif-
ferentiation of the members of the Crassus
Group is relatively recent, as is suggested by the
small differences among them. It is also inter-
esting to speculate on the origin of the sub-distal
aperture of the thick tube in humphreysii. This
combined with the distinctive aedeagal shape ap-
pears to form a good means of mechanical iso-
lation from cochisei. However, there is no evi-
dence that these two species overlap broadly
now, and further, the mechanism seems to break
down where the two are presently in contact.
This is discussed further in the section on Me-
chanical Isolation, and in the Promising Prob-
lems section.

We have argued above that certain features of
the Crassus Group are primitive because they
are shared with other related genera. It is al-
ways possible that those genera are derivative
from the Crassus Group, and that all may be de-
rived from another Barytettix stock. Thus, the
massive ramus of the cingulum of the Crassus
Group and the other genera of the Conalcaeini
may represent a derivative condition. A narrow
ramus which does not extend to the ventral mid-

9]

line is the commonest type in other melano-
plines and in the few genera which might be re-
lated to the Conalcaeini. This type of ramus is
seen in the Humphreysii Group, the most primi-
tive member of which, cochisei, might thus be
the least changed descendant of the ancestor of
the entire genus. This hypothesis better fits the
apparently small differences among the species
of the Crassus Group, suggesting recent diver-
gence, and the probably derivative condition of
the dilated thick tube of the female. It also ac-
cords better with the probability that brown
ground color, as in cochisei, is more primitive
than green, as in the Crassus Group. However, it
offers no help in the problem of the very wide
separation of the Crassus Group from cochisei
whose ancestor gave rise to it, ex hypothesi, and
does not accord well with the probability that
the sheath in the Humphreysii Group is deriva-
tive and that in the Crassus Group primitive.

The Psolus Group is probably derived from
the Humphreysii Group. The two groups share
the derivative condition of well-developed ven-
tral lobes of the sheath, and the greatly elon-
gated lobes of the Psolus Group are clearly the
most extreme derivative condition. No member
of the Psolus Group shares any derivative char-
acter of humphrevsii or tridens, with the possible
exception of paloviridis which has a green
ground color as do some of the populations of
humphreysii. In some humphrevsii the ventral
lobes of the sheath are slightly produced in the
middle or more produced laterally and thus
might be considered to be a tendency toward the
greatly elongated lobes in all members of the
Psolus Group. The similarity seems to be small,
and we do not consider it significant. On the
other hand, the ventral valves in cochisei aver-
age longer than in humphreysii, they are straight
and blunt acute as in the members of the Psolus
Group, and the phallotreme is broadly open
dorsally, much as it is in contilus. That species
also has a shorter aedeagus than any member of
the Psolus Group, a primitive condition. We
therefore consider cochisei and contilus to be the
links between the two groups. Another possible
derivation may again be through paloviridis if
bursal characteristics are considered. The disto-
dorsal aperture of the thick tube of that species
occupies the same position as it does in cochisei.
It is easy to envision the development of the
bursa in the ancestor of the Psolus Group start-
ing in a cochisei-like form with the elongation of

the bursa and aedeagus followed by a ventral-
caudal bending of the thick tube, and being com-
pleted with the ventral and caudal displacement
of the orifice of the thick tube. In such a se-
quence paloviridis holds an intermediate posi-
tion except for its greatly elongate aedeagus and
bursa. One of the geographic variants of palo-
viridis (at Guamtchil) also possesses flattened
ventral lobes of the sheath which could easily be
the primitive condition in the entire group.
However, derivation of other members of the
Psolus Group from paloviridis requires con-
vergent development of the primitive brown
color, shortening of the aedeagus and the open-
ing again of the phallotreme in contilus.

The evolution of the contilus ancestor from a
cochisei-like form must have involved the elon-
gation of the dorsal and ventral valves, the ven-
tral lobes of the sheath, and the bursa of the fe-
male, as well as the narrowing of the ventral
valves, and the development of the distinctive
pronotal markings and black femoral stripe. It is
no easier to envision ancestral paloviridis ac-
complishing these changes than ancestral
contilus, except that the development of the
color pattern typical of other members of the
Psolus Group would have been postponed. On
the other hand paloviridis must have developed a
green ground color, which would subsequently
have had to revert to brown at the time it gave
rise to the other members of the group.

If contilus is accepted as the least changed
from the ancestor of the Psolus Group, then the
other three members could have developed from
it by the elongation of the aedeagus and closure
of the phallotreme. The striking similarity of the
bursa of contilus tectatus to that of psolus (see
Table 7, p.38 ) suggests an ancestral-descendant
relationship. |The only slightly less similar
bursa of nigrofasciatus, and the greater simi-
larity of its dorsal valves to those of tectatus,
suggest a similar relationship. The present dis-
tribution of the three species gives evidence of an
original inland (foothills)-coastal split of the an-
cestral contilus population. The inland popula-
tions developed elongate aedeagi and bursae and
closed phallotremes, which was followed by
north-south separation leading to the develop-
ment of modern psolus and nigrofasciatus. Each
of those species has retained different character-
istics of the ancestral contilus tectatus. We sus-
pect that paloviridis evolved from the psolus-
nigrofasciatus stock. There seems little question

T. J. COHN AND I. J. CANTRALL

that the characteristics of the populations north
of Culiacin where it overlaps with psolus are
derivative and were developed as a barrier to
insemination of psolus females (see under Geo-
graphic Variation of paloviridis and the section
on Mechanical Isolation). The sculpturing of the
ventral valves of paloviridis populations south of
Culiacdn is very much like that of psolus and
may indicate a derivation from that species. The
dorsal valves in this region (but also in the
Guamuchil-Navojoa region to the north) are
quite similar to those in nigrofasciatus, suggest-
ing a relationship with that species. It is in this
southern area also that the hind femora have
dark stripes as in the other three members of the
Psolus Group. Both psolus and nigrofasciatus are
found at higher altitudes than paloviridis and at
least nigrofasciatus farther inland as well. Thus
ancestral paloviridis may have developed from a
reinvasion of the southern coastal plain from a
psolus-nigro fasciatus stock. It may have become
so successful there that it has replaced contilus
and psolus in all but thorn forest habitats where
it seems to be at a disadvantage (see under Habi-
tat of paloviridis, psolus, and particularly con-
tilus tectatus). Another suggestion of its south-
ern Origin is its sympatry with psolus north of
Culiacdn, as opposed to its possible replace-
ment of lowland psolus in the Cosala region, and
its infrequent occurrence north of Guamtchil. It
may be limited in its southern distribution by
poecilus which is also abundant in the lowlands
south of the Cosala road. Here it may be out-
competed by an equally well-adapted lowland
species. It is also possible that poecilus has been
spreading northward and invading the range of
paloviridis and replacing it, possibly through
competitive or reproductive exclusion. South of
the Cosal4 road our records of paloviridis are
quite infrequent, and may represent relict popu-
lations. To better hypothesize on the origin of
paloviridis we need better data on primitive and
derivative similarities, and more distributional
data from the foothills of the Sierra Madre.

The success of paloviridis in the lowlands may
have affected the variation in the bursae of other
members of the Psolus Group. If the bursa of
paloviridis is considered primitive, then selective
pressures caused by that species for reproduc-
tive isolation might have been responsible for
the ventral and subdistal orifice of the thick tube
in psolus and contilus, although the orifice is also
ventral in nigrofasciatus which is largely allo-

CONALCAEINE GRASSHOPPERS: BARYTETTIX

patric to paloviridis.

If contilus developed in the lowlands where it
is now found, then the reinvasion of the low-
lands by paloviridis and its adaptation to more
open environments may be responsible for the
present isolation of the subspecies of contilus.
Populations of contilus formerly occupying mar-
ginal open habitats may have been eliminated by
paloviridis. _ However, the destruction of the
thorn forest habitat of contilus by man is more
likely to have caused both the restriction of con-
tilus and the spread of paloviridis. This destruc-
tion seems to have been too recent a phenom-
enon to explain the differences among the con-
tilus populations. For the original isolation,
therefore, one is forced to consider paloviridis
competition or geographic barriers. Rivers ap-
pear to be the only obvious barriers in the lat-
ter category. Two of the subspecies come close
to, but do not cross rivers (dicranatus at the Rio
San Lorenzo and contilus at the Rio Humaya)
but the ranges of most end far from any major
rivers. The present distribution of the contilus
subspecies is undoubtedly the result of a com-
bination of the several factors discussed above,
and we cannot now reconstruct the original iso-
lating factor.

A major stumbling block in any phylogeny of
species of Barytettix is the distribution of color
and color patterns among the species. We have
suggested above that brown may be the primi-
tive color in Barytettix because it is the com-
monest color among North American melano-
plines even in humid areas with greater quan-

MECHANICAL

The great interspecific variability of male
genitalia in the Melanoplinae has been known
since Hubbell (1932) first called attention to it,
and since that time this variability has been in-
creasingly important as a taxonomic character.
Concomitantly, the biological significance of
genitalic variability has probably been assumed
to be associated with reproductive isolation.
However, evidence for this role has been meag-
er or lacking. It has been known for some time
that male genitalic structures in the Oedipodinae
and Gomphocerinae show very little interspeci-
fic or even intergeneric variability (Gurney,
1940; Barnum, 1959; Uvarov, 1966:388). Otte
(1969) demonstrated that, at least in the North

93

tities of green vegetation. However, this as-
sumption required the independent develop-
ment of the derivative green color three times no
matter which of the above phylogenetic alterna-
tives is chosen (Crassus Group, middle popula-
tions of humphreysii, and paloviridis). Further-
more, the species or populations with green
ground color are identical in color pattern (with
a few small differences in terminalis). If for the
sake of argument the assumption is made that
green is primitive, other problems are en-
countered. Although the Crassus Group retains
the primitive color (assuming that it represents
the descendants of the ancestral species of
Bary tettix), cochisei and contilus, considered on
morphological grounds to be the primitive mem-
bers of their respective groups, not only possess
the derivative color, but also give rise to species
with the primitive green color, humphreysii
(southern populations) and paloviridis. The
green humphreysii populations are unlikely
candidates for the ancestor of the Humphreysii
Group. They do not retain any particularly
primitive features, and a number of reversals
would have to be assumed if humphreysii were to
have given rise to cochisei (reduced cerci, blunt,
straight ventral valves). Nor is it easier to derive
paloviridis rather than contilus from the Humph-
reysii or Crassus Groups, and several reversals
would also have to have taken place if the other
Psolus Group members were to have been de-
rived from paloviridis (short aedeagus and bursa
and an open phallotreme in contilus, as well as

independent development of brown ground
color).

ISOLATION

American fauna, stridulation in the

Gomphocerinae, and wing display and acous-
tical behavior in the Oedipodinae serve as
reproductive isolating mechanisms. In those
species of the Melanoplinae which he studied, he
found virtually no pre-copulatory behavioral
mechanisms of these sorts which might serve
that purpose. Thus there appears to be a posi-
tive correlation between genitalic variability and
degree of development of acoustical behavior or
display mechanisms. Considering that the num-
ber of mistakes leading to interspecific copu-
lation depends upon the development or excel-
lence of pre-copulatory recognition mechan-
isms, it may be assumed that errors in recog-

94

nition would be more common among melano-
pline individuals than between members of
species of the Gomphocerinae and Oedipod-
inae. During the course of the field work carried
out for this study, Cohn found in nature three
examples of mismatings in Barytettix. He
estimates that these represent between three and
four percent of the copulations which he ob-
served in the field. Further, in caged individuals
he found that the males and females of different
Barytettix taxa readily accept each other. Such
actual or potential mistakes in interspecific
copulation indicate that there could be selective
pressure on genitalia to serve as isolating mech-
anisms.

We first become interested in the problem of
mechanical isolation in Barvtettix because we
were intrigued by the enormous differences in
the length of the male aedeagus of members of
the Psolus Group compared with members of
the Humphreysii and Crassus Groups. The ques-
tion of how, in areas of sympatry, the female was
structurally adapted to receive such different
aedeagi under conditions of copulatory error led
Cantrall to study the receptaculum seminis. He
found that the bursa copulatrix was very dif-
ferently developed in the different species
groups. Differences in size, shape, sclerotiza-
tion, and location of the aperture to the thick
tube were striking and, in each group, cor-
related with the basic structure of the terminus
of the male genital mass. Although less pro-
nounced within each group, differences in de-
gree of sclerotization, location of sclerites, and
position of the aperture of the thick tube cor-
related with the shape of the aedeagus and, in
particular, with the opening of the phallotreme.
Details and further consideration of these dif-
ferences are presented by Cantrall and Cohn
(1972)18.

Based on a comparison of the male aedeagus
and the female bursa copulatrix and thick tube,
we envision several ways in which interspecific

181y 1970, at the time of preparation of the manuscript
for this 1973 paper, the present study had not progressed far
enough to enable the naming of the populations discussed
therein. These taxa are now recognizable as:

Cantrall and Cohn, 1972 Present Study
Group A Humphreysii Group
Group B Psolus Group
Species 1, Figs. 3, 4, 8, 13 B. paloviridis
Species 2, Figs. 9, 14 B. psolus
Species 3, Figs. 10, 15 B. contilus contilus
Species 4, Fig. 16 B. nigrofasciatus
Group C Crassus Group

T. J. COHN AND I. J. CANTRALL

insemination could be mechanically prevented in
Barytettix. At least three possibilities are sup-
ported by evidence from our studies. First, it
seems clear that a functional mating is depend-
ent upon the fit, one with the other, of the
aedeagus and the bursa of a mating pair. Any
discrepancy or lack of accommodation in size
and shape of an aedeagus and a bursa is impor-
tant because of the need for the male to elab-
orate and place in position in the female a
spermatophore tube. Gregory (1965) pointed
out that the spermatophore tube is developed
from a fluid produced in the interior of the male
concealed genitalia. This fluid is forced through
the phallotreme and on into the spermathecal
duct of the female. As the fluid passes along
these channels, the outer portions of the cylind-
er of fluid change to form a spermatophore
tube which carries any remaining fluid until the
fluid is exhausted by conversion to sperma-
tophore tube. The male reproductive cells are
then forced through the tube into the diverticu-
lae of the female. Second, the amount of secre-
tion necessary for the production of a sperma-
tophore tube must be dependent upon the area
of the inner surface of the spermathecal duct.
Too large an amount could possibly fill the
diverticulae of the receptaculum seminis thus ex-
cluding spermatozoa, and too little could cause
the length of the spermatophore tube to fall
short of that needed to reach the diverticulae.
Thus a successful mating would be dependent
upon the correct amount of secretion in relation
to the inner surface area of the spermathecal
duct. Third, the need for an alignment of the
opening of the phallotreme with the orifice of
the thick tube seems obvious. If these two open-
ings are not aligned properly the sperma-
tophore tube cannot be threaded into the
spermathecal duct. The locations of the opening
of the phallotreme and of the orifice of the thick
tube relative to the internal genitalia of each of
the members of a mating pair must certainly be
of primary importance to a successful mating.
These three aspects of mechanical isolation are
more amply discussed below as they pertain to
the possible evolutionary development of the
taxa of the genus Barytettix.

INTERGROUP MECHANICAL ISOLATION

Owing to differences in size and shape of the
aedeagi and bursae, and of the location of the
opening of the phallotreme and of the thick tube

CONALCAEINE GRASSHOPPERS: BARYTETTIX

it is difficult to envision a successful mating
between a female of the Psolus Group and a male
of either the Humphreysii or Crassus Groups. The
aedeagus of the male is approximately one-fifth
the length of the bursa of a female of paloviridis,
psolus, or nigrofasciatus and if inserted into the
bursa of one of these species (see particularly
Fig. 16; compare Figs. 7, 9, 13, 14 A-H, 17, 18),
would leave the distal end of the bursa unoc-
cupied. Even though the aedeagus would fit
better in the bursa of a female contilus, the dis-
crepancy in lengths would still result in a sac-like
space in the bursa. In the absence of a close fit
between the opening of the phallotreme and the
aperture of the thick tube, the mucilaginous
secretion elaborated by the male would pour
into this space, preventing the formation of a
spermatophore tube capable of conducting
spermatodesmes. In a reciprocal mating, the
elongate aedeagus of a Psolus Group male (Figs.
13; 14 A-H) would have to penetrate the short
bursa of a female of the Humphreysii or Cras-
sus Group (Fig. 16) and hold it in position long
enough to produce an effective spermatophore
tube. Thus a Psolus Group male would not be
able to develop and maintain a proper position
to successfully mate with a female from one of
the other two groups.

In either of the above crosses a further
complication is the location of the aperture of
the thick tube as it relates to the distal opening
of the phallotreme. In the Crassus and
Humphreysii Groups (Figs. 7B,D,F and 9 B,
D, F, respectively) the phallotreme is so formed
distally that a developing spermatophore tube is
directed disto-dorsally in the bursa (the aedeag-
us is inserted upside down). The aperture of the
thick tube in the females of these two groups is
located in the disto-dorsal part of the bursa (Fig.
17 A-F). In the Psolus Group the spermato-
phore tube is directed disto-ventrally in the bur-
sa except in males of paloviridis from the region
between the Rio Culiacén and the Rio San Lor-
enzo where the opening of the phallotreme is es-
sentially distal. The aperture of the thick tube in
the bursa of females of this group is distal,
disto-ventral, or ventral. It is evident that a
proper alignment between the distal opening of
the phallotreme and the opening to the thick
tube is not readily obtained in matings between
members of the Psolus Group and members of
either the Crassus or Humphreysii Groups.

The bursae in members of the Humphreysii

and Crassus Groups are not so obviously
different in size and shape from one another
(Figs. 16 B, E; 17). A major difference is the
large dilation found in the proximal end of the
thick tube in females of the Crassus Group
(Figs. 16 B; 17 A, D). The same area in females
of the Humphreysii Group is in no way en-
larged (Figs. 16 E; 17 B-C, E-F). We do not yet
know the precise manner in which the sperma-
tophore tube is formed in dilations of the thick
tube, nor do we know how much spermato-
phore tube producing fluid a given male can pro-
duce during one copulation. If, and we suspect
such to be the case, a male is capable of pro-
ducing only the proper amount of fluid to de-
velop a tube of the correct length in a female of
his own species, then, depending upon the mem-
bers of a mismating, a male might produce too
much fluid which could fill and block off the
spermatheca, or too little and fail to develop a
tube of adequate length to reach the sperma-
theca.

There is some difference in the size of the
aedeagi between members of the Humphreysii
and Crassus Groups. Not only are the aedeagi in
the Humphreysii Group often twice as long as
those found in the Crassus Group, but the wid-
est part is usually almost twice as wide (Figs. 16
B, F). These differences might well have little ef-
fect because of the elasticity of the bursal walls
which are thinner and less sclerotized than those
found in the Psolus Group. On the other hand,
the aperture to the thick tube is decidedly dor-
sal in the bursae of females of the Crassus Group
(Fig. 17 D). Further, thickenings exist at this
point which we believe have to do with the
positioning of the valves of the aedeagus during
matings. The valves are bent (Figs. 7 A, C, E) as
though they might fit up into the reinforced
aperture of the thick tube (Figs. 17 A, D) where
they could lock into place ensuring a precise
channel through which the spermatophore tube
easily could pass. The relationship of the
terminus of the aedeagus with the opening to the
thick tube is quite different in taxa of the Hum-
phreysii Group (Figs. 9; 16 A-B, E-F; 17 B-C, E-
F) where the aperture of the thick tube is more
distal and is unreinforced, and the valves of the
aedeagi are not down-curved.

Although the question of mechanical repro-
ductive isolation between members of the Cras-
sus and Humphreysii Groups may well be a
moot one because the ranges of the two groups

96

are not known to overlap and they may very well
have never been in contact, we believe that
genitalic differences between the groups are of
sufficient magnitude to preclude crossbreeding
were any of the taxa sympatric. The differences
in genitalia between members of the Psolus
Group and taxa of the Crassus and Humphrey-
sii Groups are much more pronounced. These
differences clearly seem to be adequate to func-
tion as mechanical reproductive — isolating
mechanisms for we have no evidence of any
intergroup hybridization. Thus it is evident that
the three species groups of the genus Bary tettix
have diverged far enough so that mechanical iso-
lation without doubt can function as an inter-
group prezygotic isolating mechanism.

INTRAGROUP MECHANICAL ISOLATION

The Humphreysii Group.— Although far more
variation is found in the elements of the male
genitalia in the Humphreysii Group as com-
pared to the situation in the Crassus Group, the
characteristics of the female bursa copulatrix
and the thick tube are less well-developed. There
are distinct differences between humphreysii
humphreysii and humphreysii cochisei in the
shape of the distal portions of the bursa and in
the location of the aperture of the thick tube,
and smaller but detectable differences in these
features between fridens and humphreysii humph-
reysii. Tridens and humphreysii humphreysii are,
as presently known, allopatric. In the absence
of hybrids or sympatric populations we are un-
able to judge whether or not the two forms are
mechanically isolated. During mating the acicu-
late processes on the median margin of the dor-
sal valves in tridens (Fig. 9 F) can be expected to
deflect dorsally in the bursa a spermatophore
tube elaborated by the male. (During copula-
tion the aedeagus is upside-down in the bursa.)
This would serve to align the tube with the
slightly more proximal aperture to the thick tube
found in tridens. The more distal aperture to the
thick tube and the more distal opening of the
phallotreme in humphreysii are so located that a
spermatophore tube would be directed a little
more distally and hybridization between
humphreysii humphreysii and tridens con-
ceivably could be prevented. On the other hand,
these differences are minimal and we do have
hybrids between humphreysii humphreysii and
humphreysii cochisei where differences in the
terminus of the aedeagus and features of the

T. J. COHN AND I. J. CANTRALL

bursa are much greater. Thus it is possible that
the slight differences between /umphreysii
humphreysii and tridens are not great enough to
prevent hybridization. We have no information
on the detailed and precise relationships between
the size and shape of the aedeagus, as
determined by the morphology of the various
aedeagal parts, and the size, shape, degree of
sclerotization and location of the sclerites of the
walls of the bursa, and the position of the aper-
ture of the thick tube. Neither do we know what
effect a mating or several matings may have up-
on the bursa. What appear to be real and
meaningful differences could be the results of
stretching and possible distortion from such
matings. Dissection of the coupled genitalia of
mating pairs should ultimately reveal the fate
of the spermatophore tube in matings between
individuals of different but closely related taxa.
Cantrall is presently working with such material.

The ventral valves of the aedeagus of humph-
reysii cochisei (Figs. 11 A, B) are long and slend-
er, and are close enough to each other to cause
the spermatophore tube to issue distally from
the phallotreme and prevent it from turning dor-
sad in the bursa during mating. The orifice of the
thick tube is located in the proper position to re-
ceive a tube oriented in such a direction (Fig. 17
F). On the other hand, in humphreysii humph-
reysii the ventral valves are of such shape and
are oriented in such a manner that the spermato-
phore tube would need to protrude dorsally (in
the bursa) in order to make proper entry into the
thick tube. Since we have evidence of hybridi-
zation between humphreysii humphreysii and
humphreysii cochisei, the differences in size,
shape, and location of the male and female
genitalic counterparts seem not to be adequate
to produce a completely functional isolating
mechanism. However, the number of hybrids
which we have found is limited in number, and
the areas of hybridization are few and very nar-
row. The hybrids may have resulted from a
chance pairing of two individuals which varied
enough from the norm in genitalic structure to
enable a transfer of male reproductive cells. If
this is what happened, then it is likely that very
strong selection for mechanical isolation must be
underway between these forms, and_ rein-
forcement of this nature could eventually result
in the development of completely functional
reproductive isolation.

CONALCAEINE GRASSHOPPERS: BARYTETTIX

The Crassus Group.— The species of the
Crassus Group all have rather similar genitalia.
Differences in the bursae found in the three taxa
are minor and probably do not represent func-
tional reproductive isolating mechanisms. The
species are completely allopatric and there is no
apparent evidence of demes of any of the
populations having come into contact since iso-
lation. Selection pressure for the evolution of
mechanical isolation seems not to have been
operative. There is, however, clear evidence in
the male aedeagal parts (Fig. 7) of divergence
under conditions of isolation. Should demes of
differing taxa come into contact in the future,
selection pressure would develop for complete
mechanical isolation based on differences al-
ready present.

The Psolus Group.— Although the close rela-
tionship of the members of the Crassus Group
foreshadows the development of mechanical
isolation as a reproductive isolating mechanism,
it is within the Psolus Group that evidence of
such developments are most interesting. Here,
character reinforcement, as a mechanism for the
development of mechanical isolation, is clearly
discernible in populations of paloviridis which
are sympatric with those of psolus or with con-
tilus dicranatus.

In the area of overlap of the ranges of
paloviridis and psolus the dorsal valves in palo-
viridis almost completely cover the apices of the
ventral valves (Fig. 15 P), quite different from
the position of these valves in psolus (Fig. 15 R).
Outside the area of sympatry the ventral valves
of paloviridis are more extensively exposed, and
very similar to psolus (compare Fig. 15 Q with
R). In these same illustrations a similar pattern
is shown in the exposure of the distal opening of
the phallotreme. Precisely this same develop-
ment of the dorsal valves is found where palo-
viridis overlaps the range of contilus dicranatus.
The aedeagal valves of this subspecies of contilus
are basically similar to those of psolus (Fig. 14
D). Both characters must affect the position or
attitude of the spermatophore tube as it emerges
from between the dorsal and ventral valves. The
tube must find its way into a disto-dorsal aperture
of the thick tube in paloviridis (Figs. 18 F, L) and
a disto-ventral aperture in psolus (Figs. 18 D, J)
and dicranatus (similar to Figs. 18 B, H). In the
males of the latter two species the spermato-
phore tube is able to turn ventrad in the bursa

97

(the aedeagus is inserted upside down into the fe-
male) before the apex of the ventral valves be-
cause the dorsal valves end much before the apex
and the phallotreme is open ventrally in the bur-
sa. In sympatric paloviridis the extension of the
dorsal valves to the apex of the ventral valves
forces the spermatophore tube to emerge
straight distally, probably beyond the opening of
the ventrally located aperture of the thick tube in
psolus and dicranatus. In this the action of the
dorsal valves is aided by the sharp, inward
pointing dorso-lateral walls of the phallotreme.
In the same manner, the curvature of the ven-
tral valves (Fig. 15D) also must force the
spermatophore tube to emerge disto-dorsally in
the bursa and prevent its entry into the ventral
aperture in psolus and dicranatus. Outside the
area of sympatry with psolus and dicranatus the
characteristics of each of these structures shift
to become more like the condition found in
those species. Variation in other aedeagal fea-
tures in paloviridis which undergo character rein-
forcement in areas of sympatry with psolus and
dicranatus has been discussed in detail above on
pages 56-59 and graphed in Figs. 3 and 4. At the
moment we can offer little regarding an inter-
pretation of the functions of these features other
than to indicate our conviction that they have to
do with the proper seating of the aedeagus in the
bursa, thus ensuring an alignment of the open-
ing of the phallotreme with the aperture of the
thick tube.

Mismatings between paloviridis and psolus
have been observed in the field, yet among the
many specimens examined we have not found a
single hybrid. In the absence of hybridization we
conclude that the differences in genitalic struc-
ture account for the failure of mismatings and
that these differences represent a mechanical
prezygotic isolating mechanism.

The aedeagus and bursa of nigrofasciatus are
as long as in psolus and paloviridis (Figs. 14 F-H,
N-P; 15 Q; 18 D-F, J-L) and seemingly would
make hybridization easy between these species.
Nigrofasciatus is allopatric with psolus. The
opening of the phallotreme, the size and shape of
the bursa, the location of the aperture of the
thick tube, and the pleating and reinforcement
of the walls of the bursa are all quite similar be-
tween the two species (Figs. 18 D-E, J-K), and
are probably not sufficiently different to pre-
vent hybridization if the two forms were symp-
atric. On the other hand, nigrofasciatus and palo-

viridis have been taken together at 1.1 mi SW
San Ignacio Ferry in Sinaloa. Although 36
specimens of the two species were taken, none
show any evidence of hybridization. Since the
aedeagi of these species are rather similar
(compare Figs. 14H and 15 Q) the burden of
mechanical isolation must fall upon the bursae.
These organs are quite different in the two taxa.
Not only is the aperture to the thick tube disto-
ventral in nigrofasciatus as compared to disto-
dorsal in paloviridis (Figs. 18 E-F), but the
pleating and reinforcement of the bursal walls is
much stronger in paloviridis than in nigro-
fasciatus (stippled areas in Figs. 18 K-L). These
differences apparently are of sufficient magni-
tude to serve as a prezygotic isolating
mechanism.

In the subspecies of contilus selective forces
other than those of reproductive isolation must
have been responsible for the distinct but uni-
form differences in the aedeagi (Figs. 14 A-E, I-
M). Although these populations are completely
allopatric, mechanical isolation has not
developed to the point where it is functional, for
Cohn in 1970 was able to obtain nymphs from
crosses involving c. fectatus (Figs. 14 B, J) X c.
dicranatus (Figs. 14D, L), and c. similis (Figs. 14
E,M) X © dicranatus. Compared with other
members of the Psolus Group, the subspecies of
contilus have a shorter and wider aedeagus (Figs.
14 A-H), the orifice of the thick tube of the
receptaculum seminis is ventral, and the thick
tube is relatively free of dilation (Fig. 18). It is
only inc. tectatus that dilation approaching that
found in the remaining species of the Psolus
Group is present. It is not surprising, therefore, to
find that intersubspecific crosses are possible.

Cohn also crossed c. dicranatus and c. tectatus
with both psolus and nigrofasciatus. As in the
crosses above, nymphs were obtained. It is sur-
prising that nymphs were obtained at all, for the
aedeagi found in psolus and nigrofasciatus are
longer and narrower than those of the sub-
species of contilus (Figs. 13 and 14). The bursae
of the females reflect in size and shape this same

T. J. COHN AND I. J. CANTRALL

relationship, and, except for a small dilation in c.
tectatus, the thick tube dilations in psolus and
nigrofasciatus are not found in contilus (Fig. 18).
All of the members of the above crosses are allo-
patric in distribution; the crosses were made in
the laboratory and may be considered stress
matings.

SUMMARY

Gross interspecific genitalic differentiation has
taken place in orthoptera having no or feebly-
developed stridulation, wing display, or acousti-
cal behavior. Genitalic divergence in Barytettix
is strong. Since mismating is not rare in
sympatric populations in this genus, it is reason-
able to conclude that there is direct selective
action on the genitalia as reproductive isolating
mechanisms. That this has occurred is indicated
by character reinforcement in the genitalia of
paloviridis where it is sympatric with psolus or
contilus dicranatus. Further, strong bursal dif-
ferentiation between paloviridis and nigro-
fasciatus is evident in a sympatric population
of these species. In both of these cases exten-
sive collecting has failed to turn up a single
hybrid specimen, although mismatings have been
observed between paloviridis and psolus. Our
only hybrids are from areas of subspecific inter-
gradation. In such cases (Aumphreysii hum-
phreysii X humphreysii cochisei) the zones of
intergradation are narrow, indicating the pos-
sibility that genitalic differentiation has de-
veloped to a point where it is becoming func-
tional as an isolating mechanism.

In populations of allopatric taxa such as the
subspecies of contilus, selective action on
genitalia seems not to have been adequate to
prevent crossing, at least under stress matings
in the laboratory. We suggest that under con-
ditions of hybridization in nature resulting from
range extension of one or more of these sub-
species, character reinforcement with resulting
accelerated genitalic differentiation and repro-
ductive isolation will obtain.

PROMISING PROBLEMS

Any intensive taxonomic study will uncover
situations of general biological interest. Most of-
ten such situations are buried in the body of the
resulting paper and readily available only to

taxonomists, and then often only to specialists in
a particular group. In this study we have un-
covered many problems which seem to have
broad evolutionary implications, partly because

CONALCAEINE GRASSHOPPERS: BARYTETTIX

we were looking for them, and partly because we
have had the time to return to the field to obtain
more data on them. In this section we draw to-
gether and summarize the problems we have
found in Baryrettix, the discussions of which are
scattered through the body of this paper. We do
so in the hope of encouraging others to work on
them. For the solution of many of these prob-
lems, Bary tettix offers ideal material. The species
are usually large, conspicuous and common, and
are thus easy to collect. The 1970 work indicates
that all taxa can be reared from medium-sized
nymphs with little mortality (at least at high
temperatures), that all lay eggs readily, and that
many eggs hatch without special treatment.
Many of the problems reported below are under
study by one or the other of us, but there is am-
ple room for others to join us.

Mechanical Isolation.— This is probably the
most interesting problem in the genus. Our evi-
dence is indirect that the major prezygotic
isolating mechanism in Barytettix is genitalic.
That there is no pre-copulatory behavioral
mechanism still needs to be demonstrated. Al-
though we have three records of mismated
couples in copula in the field, we have made no
study of mating behavior under natural condi-
tions. Cohn conducted two types of cage experi-
ments in 1970. Many combinations of males of
one species and females of another always re-
sulted in some copulation. The frequency and
length of such copulations, which were not re-
corded, are probably the critical factors in the ef-
ficacy of mechanical isolation. This could readi-
ly be compared with conspecific copulations in
adjacent cages under almost identical condi-
tions. A second set of experiments involved six
males of one species in a cage with six females of
their own and six females of another species. At
least some of the species appear to copulate at
random with the females. The cages were small
(slightly less than a cubic foot) and possibly
caused the animals trouble with any behavioral
mechanism, and no attempt was made to elimi-
nate the possibility of pheromonal attraction.
The species of Barytettix are easy to handle and
observe, and a series of simple experiments
could readily be devised to demonstrate the
nature of the isolating mechanism.

The only other barrier to insemination be-
tween different species is in the misfitting of the
genitalia. Our best evidence for this is the aedea-
gal reinforcement phenomenon in_ paloviridis

99

where it is sympatric with psolus and contilus di-
cranatus in the Culiacan region (see under palo-
viridis Geographic Variation). Further evidence
might come from a study of the variability of the
aedeagus in psolus and paloviridis near Cosala
where the two species are largely altitudinally al-
lopatric. Although no reinforcement was seen in
either species in the narrow zone of sympatry
(see Fig. 3), there appeared to be greater aedea-
gal variability in the small series of allopatric
psolus (see under psolus Geographic Variation),
suggesting a release from strong selective pres-
sure. One of the miscopulating pairs seen in the
field was taken in this region. Larger series,
more precise determination of range limits, and
more frequent collections near those limits in
both regions would help to establish this phe-
nomenon more firmly. Less direct evidence for
mechanical isolation derives from our compari-
son of male and female structures in sympatric
species. However, all of our ideas concerning the
function of the different aedeagal and bursal
parts are based on inference (see preceeding sec-
tion on Mechanical Isolation). Dissection of
homo- and heterospecific copulating pairs pre-
served at different times after the initiation of
copulation should demonstrate not only the pre-
cise function of each of the parts, but also the
significance of aedeagal and bursal differences in
preventing insemination. In this, acridologists
may have an advantage because the male
produces an easily visible, firm, proteinaceous
spermatophore tube which must thread into the
female spermatheca, a process which apparent-
ly takes place over at least an hour’s time. Can-
trall has begun a study of abdomens from copu-
lating pairs, but it is not yet clear whether the
material was adequately preserved. The abdo-
mens were cut from copulating pairs which had
been slowly frozen, or from pairs in the field af-
ter they had ascended the vegetation in the even-
ing and had become quiescent. In both cases, the
abdomens were preserved in alcohol. It may be
easier to section preserved copulating abdo-
mens, although the Highly sclerotized aedeagal
parts may make this difficult. Randell (pers.
comm.) may have developed a technique to
overcome this difficulty. Gregory (1965:35) re-
ports that abdomens removed from living Lo-
custa migratoria continue to function and can be
dissected and observed while still alive. This may
be a most useful technique in a study of
mechanical isolation.

100

In regard to this problem, it will be of particu-
lar interest to know if a copulating male is able
to detect a misfit of genitalia, a blockage of the
spermatophore tube or an inability to produce
enough material for the spermatophore tube (as
in the case of a species with a short aedeagus
copulating with a female possessing a long bur-
sa, or a male of a species in which the female has
an undilated thick tube copulating with a fe-
male possessing a dilated thick tube, or vice ver-
sa). A male which was able to detect such prob-
lems would save much effort. This was sug-
gested by the results of some but not all of the
1970 interspecific laboratory crosses.

If genitalic isolation is proven to be the major
mechanism isolating the species of Barvtettix
(and by inference other melanoplines as well),
then it would be most interesting to compare the
reproductive effort and risks in these species
with others in which song and wing display are
the primary isolating factors. Superficially it
would appear that copulation with any female of
the right size and wing shape would be very
wasteful of time and energy compared to identi-
fication of the right female by some behavioral
mechanism prior to copulation. This may not be
true, especially in the case of the singing or-
thoptera in which the males spend long periods
of time stridulating. Isolating mechanisms serve
two functions in this case, one to attract mates
and the other to prevent mismating. It may be
that efficient identification is as important as
isolation in small dispersed populations, where-
as in larger denser populations discrimination is
less important and mechanical isolation just as
effective. Other factors determining the effi-
ciency of these isolating mechanisms may be
density of vegetation, wandering habits, and the
abundance of other species. We have made no
attempt to correlate any of these factors with the
type of isolating mechanism.

Also involved in the problem of the nature
and effectiveness of the isolating mechanism are
the unusually bright colors in Barytettix and the
often radically different colors and patterns in
the different species. In several cases we have
tried to correlate differences in color with sym-
patry and allopatry. In some cases there is a
striking color and pattern difference between
sympatric species, as in the sympatry of paloviri-
dis with every other member of the Psolus
Group. Contilus, psolus and nigrofasciatus are al-
most completely allopatric and very similarly

T. J. COHN AND I. J. CANTRALL

colored. However, where paloviridis is sympat-
ric with green humphreysii the two cannot be
told apart by color or color pattern, and where
paloviridis is sympatric with poecilus the species
differ only in tibial color. It is interesting to note
that just beyond the zone of overlap of the latter
two species, paloviridis sometimes has purple ti-
biae just as in poecilus. But among other sym-
patric species there seem to be no clear cases of
tibial color being more different there than
where the species are allopatric. Tibial color is
immensely and_ strikingly variable between
populations as well as between species. It is
possible that the bright colors serve merely to at-
tract males to their congeners, and that the ac-
tual color makes little difference. This hypothe-
sis agrees with our preliminary findings that
some species of caged Baryvtettix do not dif-
ferentiate between differently colored females.
Further cage experiments involving models or
painted Baryrettix might be rewarding.

Finally, the nature of the reproductive isolat-
ing or identifying mechanisms may affect the
way in which one species is able to penetrate the
range of a close relative and thus the nature of
the boundary between them. Occasional indivi-
duals of a species with only mechanical isola-
tion which invade the range of a relative may
spend so much time mating or being mated with
the wrong species as to make penetration possi-
ble only by large numbers. Even then, the less
common species may be quickly eliminated in
the zone of overlap by the same mechanism. One
might therefore expect related species with only
mechanical isolating mechanisms to have conti-
guous or parapatric ranges (a situation which
appears to be common in the Humphreysii and
Psolus Groups). Penetration of the range of a
relative may be much easier for species with be-
havioral precopulatory isolating mechanisms.
Penetrants of that type will actively seek out an
appropriate mate and will ignore and be ig-
nored by the other species. Such species may
show more overlapping distributions.

Natural Hybridization.— Only one case of
hybridization, as opposed to gradual intergra-
dation, has been found in Barytettix, that be-
tween humphreysii humphreysii and h. cochisei.
These two taxa were reduced to subspecific sta-
tus because of our interpretation of two widely
separated groups of specimens as hybrids. There
appears to be no satisfactory explanation for this
hybridization, and the problem is discussed ex-

CONALCAEINE GRASSHOPPERS: BARYTETTIX

tensively under cochisei. Four intriguing prob-
lems are associated with this hybridization.

First, how was insemination accomplished?
From a comparison of the male aedeagus and
the female bursa it would appear that males of
either could not inseminate females of the other
species. The elongate ventral valves of cochisei
would cover the subdistal orifice of the thick
tube of humphreysii, whereas the subdistally
issuing spermatophore tube in /fumphreysii
would have difficulty forcing its way past the
ventral valves into the distal orifice in cochisei
(see under Mechanical Isolation). Was the
original insemination an accident brought about
by a strong difference in size, or by a misshapen
aedeagus or bursa? And is subsequent
hybridization possible because of the inter-
mediacy of the hybrid genitalia? We have not
studied hybrid bursae, and this may hold a clue
to the physical mechanism of hybridization.

Second, if the hybridization is not accidental,
how have the restricted populations of cochisei
south of Hermosillo survived without merging
with humphreysii, or how were they able to in-
trude into the range of humphrevsii originally? In
this area we have collected only along the main
highway; further collections inland may reveal a
somewhat different picture. It is hard to believe
that the widespread humphreyvsii is at a compe-
titive disadvantage here. However, there is a
marked break in the variation of humphreysii
populations across the area occupied by cochi-
sei. This may indicate that there was a gap in the
distribution of humphreysii there, or that a
marked change in some environmental factors
occurs in the area, allowing cochisei to compete
with the edge populations of humphreysii which
were not so well adapted to the conditions in the
gap. But there is also some evidence that cochi-
sel was present, but has been eliminated farther
south (see discussion under both subspecies).

Third, how is the apparently contiguous dis-
tribution of the two taxa maintained in Arizona
where there are few obvious barriers? There
should either be more evidence of hybridiza-
tion or examples of sympatry. Our collections in
this area are very poor, and transects across the
range of the two forms would be very enlighten-
ing. Cohn is planning such studies but there is
‘much room for more field work.

Fourth, could a hybrid habitat created by
overgrazing or other form of agriculture explain

101

the hybridization between humphreysii and
cochisei? We know little of the ecological re-
quirements of the two. If there was a sharp
ecological separation of the two, then a disrup-
tion of the habitat where they were contiguous
or sympatric may have allowed any accidents in
mechanical isolation to survive and contribute
further to the breakdown of the isolation be-
tween the two taxa. A detailed study of ecologi-
cal differences between the two species might be
relatively easy in this part of Arizona.

Stasipatric Speciation.— Part of White’s rea-
son for believing in the widespread occurrence of
this form of speciation is the existence of many
cases of contiguity or parapatry of closely re-
lated species and subspecies rather than their al-
lopatry or sympatry. In Barytettix there is one
relatively clear case of parapatry of subspecies,
that of h. humphreysii and h. cochisei in Arizo-
na, discussed above. Additional collections have
narrowed the gap between the two forms to five
to 20 miles at several points along the probable
line of contiguity of about 40 miles. No sympa-
tric populations or hybrid swarms have been
found. Investigation of the area of contact of the
two subspecies in the Hermosillo area might al-
so be rewarding. There however, our collections
come only from along the main highway, and
the two forms are clearly hybridizing (see under
h. cochisei). Both areas are of easy access and
probably have not been drastically affected by
man’s activities. We have produced hybrids in
the laboratory, but these have not yet matured,
and we thus have no data on chromosomal com-
patibility.

Other cases of suspected contiguity exist
among the brown members of the Psolus Group.
Contilus, psolus and nigrofasciatus are all al-
lopatric, as are the five subspecies of contilus.
However, the ranges of several forms come with-
in a few miles of others especially in the region
around Culiacan (see Table 19). This is the area
of easiest access in the range of these species, but
the region is rapidly being cleared of thorn for-
est which appears to be their favored habitat.
We have no information concerning the breadth
of contact between the various forms because
our collecting has been restricted to only a few
roads. Light plane survey of the road and trail
network and of the status of the thorn forest
would be easy and would greatly facilitate in-
tensive collecting.

102

Competitive Exclusion, Reproductive Exclu-
sion, and Habitat Preference.— A number of
striking cases of the interruption of the range
of one species by another, and the apparent re-
placement of one by another might be explained
readily by one of these phenomena (for the ef-
fect of reproductive exclusion see the preceding
discussion in this section on Mechanical Isola-
tion). Although we have not specifically sought
explanations of such distributional situations,
their investigation might be particularly in-
teresting in this genus.

Within the range of the common h. hum-
phreysti, small areas occupied exclusively by
paloviridis are found at 18 mi N Guaymas and
just south of that city. Humphreysii occupies a
number of habitats north and south of Guay-
mas and is common there (see Table 19). It is
thus hard to envision habitat preferences which
would exclude humphreysii from the areas now
occupied by paloviridis. However, in Sinaloa
there appears to be a distinct difference in habi-
tat preference between paloviridis and tridens,
the southern representative and close relative of
humphreysii. Whereas in the north paloviridis
usually occupies only the more moist, bushy
habitats, and humphreysii the more open, as well
as the denser habitats, the reverse is true in Sina-
loa. There, paloviridis is the occupant of open
weedy environments, and fridens is the occu-
pant of denser vegetation and thorn forest. Be-
cause tridens is not common south of Gua-
mutichil, it seems likely that this distributional
situation in Sinaloa is the result of habitat pre-
ference rather than species interaction.

In a similar manner, psolus and contilus usual-
ly occupy thorn forest or ecotonal habitats and
do not venture far into the more open areas oc-
cupied by paloviridis. In several instances, palo-
viridis was conspicuously absent from habitats
occupied by psolus and contilus tectatus al-
though it was common nearby. Because all three
species are often abundant, these situations
might be the result of some sort of species in-
teraction. Particularly noteworthy is the ab-
sence of paloviridis from the weedy summit of
Cerro Tule occupied by contilus tectatus.
Whether paloviridis, which is common at the
base, cannot penetrate the thorn forest clothing
the slopes of the mountain because of physical
factors or is being excluded by tectatus cannot be
determined from the information at hand. The
area is ideal for intensive investigation.

T. J. COHN AND I. J. CANTRALL

On the Cosala road, paloviridis is common in
the lowlands but is replaced by psolus at around
the 700 ft. level; psolus becomes equally abun-
dant at higher elevations in and at the edge of
thorn forest. The similarity of the habitat of the
two species at the edge of the thorn forest leads
to the suspicion of interaction between the two.

It is also in this area that paloviridis is abrupt-
ly replaced by poecilus along the main highway
(see Table 19). Both species are abundant in
roadside weedy, brushy habitats, one to the
north and the other to the south of a 10 mile
overlap zone. Farther south but several miles in-
land on the San Ignacio Road, both species are
found in a somewhat checkerboard distribu-
tion, but are also found together at a few
localities.

Near San Ignacio nigrofasciatus and contilus
similis are apparently separated by the Rfo
Piaxtla, but we have not investigated the area
thoroughly. The physical barrier of the river and
recent arrival in the area of one or both may ex-
plain the replacement of one species by the
other.

Geographic Variation, Gene Flow and Selec-
tion.— Although all Barvtettix species display
some geographic variation, that in h. hum-
Phreysii and paloviridis is unusually extensive
and distinctive. A study of gene flow and selec-
tion would be facilitated by these factors as well
as by the abundance and almost continuous dis-
tribution of the two species, and by the well-
marked geographic limits of the variants.

Much of the geographic variation in paloviri-
dis is associated with the development of mechan-
ical reproductive isolation resulting from inter-
action with psolus and contilus dicranatus in
Sinaloa (see under Geographic Variation in
paloviridis, and in the discussion under Mechani-
cal Isolation). Studies of the variation in palo-
viridis characters involved in mechanical isola-
tion near the edge of the range of psolus and di-
cranatus (Fig. 3, p. 57, and Fig. 4, p. 59) might
yield much information on the swamping phe-
nomenon and the strength of selection. Of par-
ticular interest is the striking shift in variation at
the Rio San Lorenzo. Paloviridis occurs com-
monly on both sides, but contilus dicranatus on-
ly on the north side.

Almost every character in fh. humphreysii,
from color to concealed genitalia, shows strik-
ing geographic variation (see Table 14, p.68).
The significance of these variants is unknown.

CONALCAEINE GRASSHOPPERS: BARYTETTIX

Many of the changes are abrupt and occur in
areas without apparent barriers. Near Hermo-
sillo, at least some of the characteristics in
humphreysii appear to be the result of introgres-
sion from cochisei. Near Cd. Obreg6n a series of
concordant changes may be the result of the bar-
rier effect of the Rio Yaqui. More collecting
needs to be done in this area. Some of the dis-
tinctive changes in southernmost Sonora and
northernmost Sinaloa seem to be related to the
presence of nearby fridens. Several of the
humphreysii characteristics here closely resem-
ble those of tridens, while others are more dif-
ferent from ¢ridens than are the characteristics of
humphreysii colonies farther north. These situa-
tions are discussed in detail under h.
humphreysii.

Many of the geographically variable features in
Barytettix appear only in the adult. Under these
circumstances there is the possibility of deter-
mining the degree of selective pressures and whe-
ther such pressures are sexual or resulting from

103

other environmental factors. The genetic consti-
tution of populations polymorphic for the
character under study (such as tibial color in the
colonies just north of Imuris) may be deter-
mined by rearing a random sample of nymphs to
maturity in the laboratory and scoring the adult
characteristics. These figures may then be com-
pared with those derived from a random sample
collected from the same population § several
weeks or months after maturity. If sexual selec-
tion were operating, a difference might be ex-
pected between the frequency of characteristics
of the adult population of one year and the pre-
selection frequency (determined by rearing
nymphs) of the next year. This situation is prob-
ably not as simple as envisioned here, but cer-
tainly deserves investigation.

The list of promising problems in this genus,
as in any well-studied group of organisms, is
endless. We hope that our presentation of sev-
eral of these problems will encourage others to
take advantage of this most interesting genus.

THE DISTRIBUTION OF THE SPECIES
OF BARYTETTIX

The vast majority of the specimens used in this
study were collected along the route of Mexican
Highway Number 15, the main west coast high-
way between the Arizona border at Nogales, and
Tepic, Nayarit. Because of the resulting linear dis-
tribution we have presented localities in tabular
form (Table 19) to illustrate the distribution of
the species and subspecies, and to demonstrate
more readily problems of overlapping ranges,
range limits and gaps in ranges. By utilizing such
a list we have been able to include the numbers
of specimens collected, which should help in
evaluating distributional data relative to allo-
patric and sympatric species, gaps in ranges, ex-
clusion of one species by another, and similar
problems.

We have listed only the records for Sonora,
Sinaloa and Chihuahua. The area of contiguity
and hybridization of humphreysii and cochisei in
Arizona is mapped in Figure 5, and that in

Sonora is treated in Table 17, p.82. For areas
where only one species occurs, as in Arizona
(humphreysii), Nayarit and central Jalisco
(poecilus), southwestern Jalisco (terminalis), and
Baja California (crassus), the presence of the
species is indicated in the table by an X, and
the records are listed in the species treatment.

Indented localities are those to the east or
west of the main highway, and are arranged
from north to south and east to west. We have
indicated collections containing only females by
using the female symbol. All other collections
contain at least one male.

We hope that this method of presentation will
make the many significant distributional details
and problems in Barytettix easier to recognize
and evaluate, and will enable others who may be
interested in working on these problems to com-
plement our collections more effectively.

104

TABLE 19

T. J. COHN AND I. J. CANTRALL

DISTRIBUTION OF THE SPECIES OF BARYTETTIX

i 7 — =
| CRASSUS | PSOLUS GROUP HUMPHREYSII
| GROUP GROUP
| | ; Coane
| elzgzgi, #8/838

g S = S Ss. 8 ss 2 Se = =a
tO eS So es0 a ee eS = 8S
<7 lS See | See Se ee ees
LOCALITIES (distance in miles) {Sa 2] Ss 9S SS Se SE eee
+ + = +
ARIZONA — NEW MEXICO
Southcentral Arizona and Silver City, New Mexico . . Se
Southeasternmost Arizona: see Fig.5,p. . ...--. 1 aXe Xs
CHIHUAHUA
Te SiWiGasasiGrandes: ee 2 cee seeps see cites ys wie | 18
|
SONORA |
F7-NOSEAguabrictal 22sec oo | 5 oY
Sierra de San José near Naco (hybrids, see text) Ved)
4sroadimi SE Cananeat sy e.c2 ce. une soe ete | lez
NORSKGananeaiers os a eaeeseee heavens Geen ee | | 27
17.7 N Imuris (31.5 S Nogales, Arizona) ......--- | | 11
AASNIBITUTISS poe Posts cdttcn de Gat ecuatsas esse ties | 8
NOsSsNGlimuriSpeenetehe acc saeaceeas 2 Anche ci anrasestoene | | 55
GMIENGlinnuni sme eee ees sen oak caer atoms Died 18
SAS WillinplttSigcksoeenes = aeaer aye) can cea aed eed Cais | 10
U5 SWa Mag dal em alate yacusmeespaus acces cies tuo eteegereys = 14
Bi DsWa Sani tay AN Alvcns ome gers on pope actsblesers apes | Ss
1D'SW Magdalena: 0: 564m c hese eaeee ens | 2068
LALGiSiSantaeAna(e6:S: lan O)).2)2 sce = Silence Loe ors
TS TES Santa ANA ae cise <n aeee cs octet: tis ae | 32.
DIAS) San taw Anal oangawey A ovens saluoteiann she centres | | 21
39 SSW Magdalena (28S Santa Ana).........-- | | 6
75 N Hermosillo (Conejos Pass) (32 S Santa Ana) | 4
AOi7iSuSantavAmal west s. 2 Aestees oe e ccas Ghsasgs ay eeewe 5
50 S Santa Ana (11.3 N El Oasis) (56 N Hermosillo) | 11
BOINibenmosillomte reece concrete eh mekoesyaes- heme 3
Copete Mine, 30 E Carbo (ANSP) ......... yr
14:2 NiHennosillo;Cathedrall ©. 75 ieee .s. coats ces | 16
Hermosillo to 48 S (11 S Los Pocitos) (35.5 N | |
Guaymas Cathedral): see Table 5,p. ....---- | |X X
18 N Guaymas Cathedral, Puente Noche Buena | |
(28.5 S Los Pocitos, 65.9 S Hermosillo Cathedral) . | | 26 :
5 NW Guaymas (1 SE jct. San Carlos Road)... .... | | 3
AROUNi Guayimas acres ciery a chews ces mene casts oh ecm | 4
Saladita:Bay/(4)'W\Guaymas)!; 2.2 3..205-% 0. -.-- - : 8
8 SE Guaymas (6.4 NW Cruz de Piedra) ......... | 4
10 SE Guaymas (2.6 NW Cruz de Piedra) ........ 4

CONALCAEINE GRASSHOPPERS: BARYTETTIX 105
TABLE 19 — DISTRIBUTION OF THE SPECIES OF BARYTETTIX
rp
CRASSUS PSOLUS GROUP HUMPHREYSII|
GROUP GROUP
ee i see a ae
= eS = oe Ss) eS Se Ss £ 3s 2
BO ES 8S bs Secs 8 Se eas
LOCALITIES (distance in miles) 3 Se | SS So So 8 SS |e es
16 SE Guaymas (2.5 SE Cruz de Piedra) ......... 4
{17 SE Guaymas, 3 SE Cruz de Piedra — data
possibly in error, see T. J. Cohn Field
INGOs, IGE ING; OA) 36560 sce mecsonoar . 14)
21 SE Guaymas (7—7.3 SE Cruz de Piedra) ....... 5 ,
22.7 SE Guaymas (8.6 SE Cruz de Piedra) ........ 10
33) GiB. Cini cy cues boot oe aD been ooo 2
SPS EMG UAVIM ASN nar, «MMe n nccueh ne est chete Sites coe 11
38.4 SE Guaymas (24.3 SE Cruz de Piedra) ....... 4
25 WNW Cd. Obregon (53 SE Guaymas) ......... 1
10 NW Cd. Obregon (3.5 W Esperanza).......... 5
Tezopaco (40 road mi NE Esperanza)....... 2
24.7 NE Esperanza (Tezopaco Rd.) ........ 10
OS Cal Oonayehit soe ctcmeadoes Aco mee ee onc 2
22 NW Navojoa (2.9 SE Fundicidn) ............ 14
ZIESE Obregoni(20:SiNWiNavojoa)) 5-5. ....5-.-- 7
GESINWaNay.0] Odeemre etemonetos ca tnccs 0s ieee ete ees 4
33.5 road mi NE Alamos Cemetery ........ VP with
27.4 road mi NE Alamos Cemetery ........ 7
19.0 road mi NE Alamos Cemetery ........ 3
8.3 road mi NE Alamos Cemetery ......... 5
IGSi(SE?)WAlamosiCemeteny 0. 46 sas ase 10
IBIN\WirAllarn OS) ariscicy sis) cusuny= 6 So ieee | 14
DINWeAVamosiCathedralleacn ce ee ee 10
IS aWeAlam osm mcrs cuter aie cee heen ean: 13
10 E Navojoa (23 W Alamos) (Cerro Prieto) . . 36
NOFES EA NAV OjOde yrse alerts sue ccot Aon sec aseaer tut 9
14.9 SE Navojoa (1.5 SE Bacabachi) ........... 15
16.4 SE Navojoa (2.9 SE Bacabachi) ........... if
20.2—21 SE Navojoa (7.4—7.8 SE Bacabachi) ..... 19
36.4 SE Navojoa (23.6 SE Bacabachi)
(B2 INVER MEOM ILS) seseeannouassasuer 17
40 SE Navojoa (0.5 SE Estacion Luis) (12.1 NW
Estacion Don) (males only, 13 females not
ielematial WO FOXES) oo oagadsonanocadaoeac 6 4
49.3 SE Navojoa (4.2 NW Estacién Don) ........ 6 ;
51.8 SE Navojoa (1.7 NW Estacion Don) ........ 10
SINALOA
ArASte staclOmey Onunrina cee -eees euclens srqs etre alee , (aie
29.2 N Los Mochis turnoff (16.8 S Estacion Don) 2

106

TABLE 19 — DISTRIBUTION OF THE SPECIES OF BARYTETTIX

T. J. COHN AND I. J. CANTRALL

= a 5
CRASSUS PSOLUS GROUP HUMPHREYSI],
GROUP GROUP
i) “ a = = “ 2 3s
» if 8 8 & sz Sissi Se
a = = & 8s 8 § cS See : = i}
Se SS eS BS SE eee
SoS oS eS SS: 6 fas Se I aS
LOCALITIES (distance in miles) S88) 6 8 SS sy Se See cee
Cerro Prieto, 23.3 N Los Mochis tumoff (21.3 S |
Estacion: Don) Aeets cnc ocean sc woes ees es we : 13
2.1 S Cerro Prieto, 21.2 N Los Mochis tumoff ... . 4 :
3.4 S Cerro Prieto, 19.9 N Los Mochis tumoff ... . 7392 2
3.8 S Cerro Prieto, 19.4 N Los Mochis turnoff ... . 5 é
4.3 S Cerro Prieto, 19.0 N Los Mochis turnoff .... 7
5.3 S Cerro Prieto, 18 N Los Mochis tumoff ..... 3
6 S Cerro Prieto, 17.8 N Los Mochis turnoff .... . 7
6.2 S Cerro Prieto, 17.1 N Los Mochis turnoff .... ; 17
12.2 N Los Mochis, 33.6 S Estacién Don ....... 12 :
6 NE El Fuerte (52 NE Los Mochis turnoff) . 17
14 S El Fuerte (32 NE Los Mochis turnoff) . . 9
4 SEAcos Mochissturotf 22024. .eee ss one . 3
13.5 W Guamuchil (50 SE Los Mochis turnoff) : 5 -
OLS WeGuamichil geese eee) ay sepedaps o sellete cs @ Gro) ee 13 17
SH OMWAGUAMUCHlMts ewe ata Ler ticesnces coe ers iene ‘ 8
62 NW Culiacdn Plaza (5 SE Guamuchil) ........ 5) 17 ;
GOPNWaCuliacanyPlazayerers ers a aictesee ene erence pene 5 1 12
57 N (NW) Culiacdn on Hwy. 15 (ANSP) ....... 2 é
50 NW Culiacdn Plaza (3.3 NW Tereros) ........ ail 4
SOPNWECuliacantseenee 26 fees eee ioe tae 4 4
5.3 NW Badiraguato (45 WNW Culiacdn) .... 9 :
AUN WACUliacanmenn teers ete wars eee case ds on sere 2 3
30 N (NW) Culiacdn on Hwy. 15 (ANSP) ....... : 4
ZOWBNWaCulaCan ye cure Mecetn check be esis Non Sick 14 5 é
1'5:2—15).4 NW) Culiacan: Plaza: i. 5: as. eause oe 29 5 3
2.5 NW Rio Culiacdn Bridge at Culiacdn (Hwy. 15) . 43
1—2 E Culiacancito (13 WNW Culiacdn) ........ 3 5
2NW,)Rio.Guliacdniat Culiacdn® o2)..c14. = ts se 9 6
ZSANWiGuliacantblazas ssa ey eci ce ee ere ee : 4 ;
6 W Jesus Marta (19 WNW Culiacdn)....... 17 12
Tecorito—1 W (=La Reforma?)
(15 N Culiacan, W side Rio Humaya) : 2 6 3
1 E Tecorito (E side Rio Humaya) ........ 4 : 4 :
1 E San Rafael (W side Rio Humaya) ...... 17) 12 2
3—3.1 NE Tepuche (13 N Culiacdn)
(EjsidepRiotHumaya) erases acess ore tee 34 Y)
PD INA(NE?,)lepuche its spspescuntte c ses ets 11 a> oy 6
DEINE Lepuch emer eae eye eae ne 12. 19
DAS Wilcpuche tgey-cats. esi eer eee 6 1
2 W Agua Caliente (9 N Culiacdn)
(W side Rio Humaya) ............... a 1
|

CONALCAEINE GRASSHOPPERS: BARYTETTIX

TABLE 19 — DISTRIBUTION OF THE SPECIES OF BARYTETTIX

107

LOCALITIES (distance in miles)

CRASSUS
GROUP

Crassus
poecilus
terminalis

c. contilus
c. tectatus

PSOLUS GROUP

c. hiscatus
c. dicranatus
c. similis

psolus

nigro fasciatus

paloviridis

re

IHUMPHREYSII
GROUP

h. humphreysii
h. cochisei

tridens

Palos Blancos—1 S (8 N Culiacan)
(EfsidesRiogHumaya) ieee esse:

El Bario (7.5 N Culiacdn) (W side Rio Humaya)

4.7 SW Tepuche (E side Rio Humaya) .....
3.9 N Culiacan (E side Rio Humaya) ......
By NER Culacdng-eaenci Ce poner reeks ccee eee
3 N Culiacdn (W side Rio Humaya) .......
Culiacdn (S end RR. Bridge over Rio Culiacdn)....
Culiacany (BE Cathedrall)iqees) nies os cece cee
DAWA GULIACAN Meer R tp sears et sonore Ne
15 SW Navolato (27 WSW Culiacdn) .......
Altata, and 1 NE Altata (31 WSW Culiacan)
1.1 W Los Mayos (19.7 road mi ENE Sanalona)
DAWaOSsMayOSpa ny chats een Oe ee oot
Nic4SE; Sanalona\(80°E Culiacin)) 2.2.5.2. .
SPE TOA alON ace wyewe 2. La Pewees aie heen ee
feEi@uliacdny(SanalonaRds)ie 2-.0es sche ae
EOE OE ACUIIACAN MEPE, holies ices Ages hae Son ious
GESEICuliacanPenatris trea sie els oe Aes ee
CerromulesSummiti(@/sSEiGuliacan)) 5. «eels
7.2—7.3 SE Culiacan Plaza (Cerro Tule Rd.) .....
S) 3) Cuilineein GS oclos ee o he Goon ee etn o eee
ISSSEACuliacaniern wey eve 5 se eec) eas a cusses
Hes ba Guliacdnimr warns Cea er eS nein
IPO Se Cuitkigih. 5 odes ee sloicee oie eee
DS SESERGuliacanper seeder sco Arye gece cr cuageenia
30 SE Culiacdn (2.4 N Rio San Lorenzo) .......
32.7—33 SE Culiacdn (S bank Rio San Lorenzo) ..
35.5 SE Culiacdn (3.2 S Rio San Lorenzo) ......
SiS osSEiCuliacanemae tec er eee Ore ss ee

60 SE Culiacan (27 SE Rio San Lorenzo) .......
64 SE Culiacan Cathedral (3.1 SE El Espinal) ae
CORSE Culiacangae eee ei ate cee ee:
6.6 E Cosald (Nuestra Senora Mine Rd.)
(60 air mi SE Culiacdn), 1600 ft. .......
5.6 E Cosald (Nuestra Senora Mine Rd.),
INSKOLO) RIES ices oto Grete EEN oon or NOR Nene
DAs SWiCosalanyl SOO sitar eee peice
OMY (CORAEL UO Th Ses bens cansooeae

|

2762 .

all

nN

nn’

NwWwnn'

N-*

12
15

w
PANDO ANKHK HNN

="

—
(ony Ly (eo Ty TS) 2

= — —Vas
NaANWe CO f*

—_
~~)

Ie
I) [eo Te

b

me UO
NNAND W

N¢

Nr

12

108

TABLE 19 — DISTRIBUTION OF THE SPECIES OF BAR YTETTIX

T. J. COHN AND I. J. CANTRALL

—,
CRASSUS PSOLUS GROUP UMPHREYSII
GROUP GROUP
, g z
SS fle 8 FF Se 5 S15 oe
48 £ | 8 8 8 3 8 3 ess |e come
LOCALITIES (distance in miles) See NCS eee Gi MSG at See Salles ems
=
NOESISW, Cosalas 700i rcs asc soe coe wee : ff 6 : 3
69.4 SE Culiacan (Cosald Rd. turnoff), 300 ft ..... 33 10
DZ INWERIG Blotan ae sk se ssctenie sakeile se hae a 9 : :
4.3 NW Rio Elota (40 SE Rio San Lorenzo) ...... 11 : 3 if
SF OUNGRIOUE] Otay cevanote ino Sy Btn eran aaa 12 . 25 5)
63 NW (old) Mazatlan Airport (1.5 NW Rio Elota),
(ESET Guliacam) | oeacecrenare: seereecktePaee arse are 9 2
58 NW (old) Mazatlan Airport ............... 4
84 SE Culiacdn (52 N [old] Mazatlan Airport)... . . 28 : :
2INiSamilonactOug 5. cute tne leeeis ve sue ae ete 3 11 pees)
1.1 SW San Ignacio Ferry (17.4 NE Coyotitan) 13 28 «8
SINE Coyotitanes 22a. tes nance 3
5 NE Coyotitan (49 N [old] Mazatldn Airport) 16
48 N (old) Mazatlan Airport ................. 18
38.8 N (old) Mazatlan Airport ............... 18
29 N(NW) (old) Mazatlan Airport ............ 6 :
31 N Mazatlan on Hwy. 1S(ANSP) ............ : 1
20 N (old) Mazatlan Airport ................. q 1
14 N (old) Mazatlan Airport ................. 13
El Venadillo (4 N [old] Mazatldn Airport) ....... 52
1.3—2.6 NW Villa Union (W side Rio Presidio) ... . . 5
3.6 NE Santa Lucia (Durango Rd.)
(SONNE Concordia) e422. she eee : NS 2
4’SWeSantaubucia waren m nc ety eae eres baton 2 :
SrsiS Was antaeucial a speten sce ayeteuseene oa : 1
I DESENEConcordiatescs-y-4 ce stgs es ciceciel 2 ae 4 3
IMESSNE} Concordia @ a-aseseeeie eon 8 13
474-4 6INE\Concordiay ios seieecs = aes cee 2 5
BINE*Concordiat meant tos ee cea eee tse 6
GAAS WAC ONCOLdIats ae cun neonate atone 3
2—2.4 E Villa Union (Durango Rd.) ........ 15
ZOINWEEScuinapat- bese eiewag siecle susie term ate 5
ZOISESEscuinapay ri .ba «ta.-soies oie. «cesta coenshecie a 3
NAYARIT (See under species account)
Northwestern and Westcentral Nayarit .......... X
JALISCO (See under species accounts)
Westcentralivaliscoms acraieirsieeie oes se sraes oe re %
SouthwesternmostJaliscoi 2 4-1. ss ee xX

TERRITORIO DE BAJA CALIFORNIA (See
under species account)

Southemmost Baja California ................

CONALCAEINE GRASSHOPPERS: BARYTETTIX

EMEBRATURE Gl tED

Bal ES DY EWR: Tinkham, Ro Flock andiC, 1.
Vorhies

1942. The grasshoppers and other Orthop-
tera of Arizona. Univ. Arizona Agr.
Exp. Sta. Tech. Bull. No. 92:254-
SWS:
Barnum, A. H.
19S! The phallic complex in the Oedipo-
dinae (Orthoptera: Acrididae).
Ph.D. thesis, Iowa State College.
University Microfilms, Inc., Ann
Arbor. 220 p.
Berlese, A.
1882. Ricerche sugli organi genitali degli
Ortotteri (Mantidae, Locustidae,
Gryllidae, Gryllotalpidae, Truxali-
dae, Acrydiidae). Atti della R.
Accad. dei Lincei. Serie terza.
Memorie della classe scienze fisiche,
matematiche e naturali, 11:259-
298:
Bruner, L.
1908. Insecta. Orthoptera. Vol. 11. The

Acrididae. Biologia Centrali-Ameri-
cana. 1902-1909, p. 25-412.
Cantrall, I. J. and T. J. Cohn
2, Melanoploid genitalia and mechani-
cal isolation. Proc. Intl. Study
Conf. Current and Future Problems
of Acridology, London, 1970, p. 35-
44.
Caudell, A. N.
1908. A New Barytettix from Arizona.
Proc. Entomol. Soc. Washington
9:69-71.
Cohns i Je
1965. The arid-land katydids of the North
American genus Neobarrettia (Or-
thoptera: Tettigoniidae): their
systematics and a reconstruction of
their history. Misc. Publ. Mus. Zool.
Univ. Michigan, No. 126:1-179.
Dirsh, V. M.
1956. The phallic complex in Acridoidea
(Orthoptera) in relation to taxono-
my. Trans. Royal Entomol. Soc.
London 108:223-356.
A preliminary revision of the famil-
ies and sub-families of Acridoidea

IDI

109
(Orthoptera: Insecta). Bull. Brit.
Mus. (Nat. Hist.), Entomol. 10 (9):
349-419.

Dunbier, R.

1968. The sonoran desert. Its geography,
economy and people. Univ. Arizona
Press, Tucson. xili+ 426 p.
Fedorov, S. M.

OD ais Studies in the copulation and ovi-
position of Anacridum aegyptium L.
(Orthoptera: Acrididae). Trans.
Entomol. Soc. London 75:53-61.

Gastil, G., R. P. Phillips, R. Rodriquez-Torres,
and M. Bonneau

ODD? The reconstruction of Mesozoic
California. Proc. 24th Intl. Geol.
Congr. (Montreal). Section 3:217-
2a

Gregory, G. E.

1965. The formation and fate of the
spermatophore in the African migra-
tory locust, Locusta migratoria
migratorioides Reiche and Fair-
maire. Trans. Royal Entomol. Soc.
London 117 (2):33-66.

Gurney, A. B.

1940. A revision of the grasshoppers of

the genus Orphulella Giglio-Tos,
from America north of Mexico
(Orthoptera: Acrididae). Entomo-
logica Amer. 20:85-157.
Revision of the North American
grasshoppers of the Conalcaea Com-
plext Proc: Usiss Nat. Mus. NO:
275-304.

LOSE

Hebard, M.

OMe Notes on Mexican Melanopli (Or-
thoptera: Acrididae). Proc. Acad.
Nat. Sci. Philadelphia 69:251-275.
Dermaptera and Orthoptera from
the State of Sinaloa, Mexico, Part
II, Saltatorial Orthoptera. Trans.
Amer. Entomol. Soc. 51:265-309.
Studies on the Orthoptera of Ari-
zona, Part II, A list of the Dermap-
tera and Orthoptera with new rec-
ords and corrections of the litera-
ture subsequent to 1909. Trans.
Amer. Entomol. Soc. 61:269-316.

1925:

I9352

110

Hubbell, T. H.

1932. A revision of the Puer Group of the
North American genus Melanoplus,
with remarks on the taxonomic
value of the concealed male geni-
talia in the Cyrtacanthacridinae

(Orthoptera: Acrididae). Misc.
Publ. Mus. Zool. Univ. Michigan,
No. 23:1-64.

Kevan, D. K. McE., Syed S. Akbar, and Yu-Cheng
Chan

1969. The concealed copulatory structures
of the Pyrgomorphidae (Orthop-
tera: Acridoidea). Part I. General
introduction. Eos 44:165-266.

Keys Kk. EH. 1.

1968. The concept of stasipatric specia-

tion. Systematic Zool. 17:14-22.

Larson, R. L., H. W. Menard, and S. M.
Smith

1968. Gulf of California: a result of

ocean-floor spreading and transform
faulting. Science 161:781-784.

Mitchell, R. D. and D. R. Cook

952% The preservation and mounting of
water-mites. Turtox News 30 (9):
4 p.
Nelson, E. W.
1921. Lower California and its natural
resources. Mem. Nat. Acad. Sci.
16 (First Mem.): 1-194.
Newell, I. M.
1970. Construction and use of tabular
keys. Pacific Insects 12 (1):25-37.
Otte; D:
1970. A comparative study of communica-
tive behavior in grasshoppers. Misc.
Publ. Mus. Zool. Univ. Michigan,
No. 141:1-168.
Packard, A. S.
1878. Chapter IX. Anatomy and embryol-

ogy. In: C. V. Riley, A. S. Packard,
and C. Thomas. First annual report
of the United States Entomological
Commission for the Year 1877 re-
lating to the Rocky Mountain Lo-
cust, the best methods of preventing
its injuries and of guarding against
its invasions, in pursuance of an
appropriation made by Congress for
this purpose. Washington, D. C.,
ps 257-279:

T. J. COHN AND I. J. CANTRALL

Paoli, G.
NOS Osservazioni su alcune particolarita
di structtura e funzione dell ‘ap-
parato genitale femminile di Doci-
ostaurus maroccanus Thnb. (Or-
thop., Acrididae). Redia 23:17-26.
Randell, R. L.
1963. On the presence of concealed geni-
talic structures in female Caelifera
(Insecta: Orthoptera). Trans. Amer.
Entomol. Soc. 88:247-260.

Rehn, J. A. G. and R. L. Randell
1963. A preliminary analysis of the lines
of the supertribe Melanoplini (Or-
thoptera: Acrididae: Cyrtacan-
thacridinae). Proc. Acad. Nat. Sci.
Philadelphia 115 (1):1-32.

Rzedowski, J. and R. McVaugh

1966. La vegetacion de Nueva Galicia.
Contr. Univ. Michigan Herb. 9 (1):
1-123.
Roberts, H. R.
1941. A comparative study of the sub-

families of the Acrididae (Orthop-
tera) primarily on the basis of their
phallic structures. Proc. Acad. Nat.
Sci. Philadelphia 93:201-246.
Scudder, S. H.
1897a. The genera of North American
Melanopli. Proc. Amer. Acad. Arts
and Sciences 32 (9):195-206.
Revision of the Orthopteran group
Melanopli (Acridiidae), with special
reference to North American forms.
Proc. U. S. Nat. Mus. 20:1-421.

Slittersy 2s):

1940a. Variations in the spermatheca of
two species of grasshoppers (Or-
thoptera: Acrididae). Entomol.
News. 51:1-3.
The internal genitalia of female
Ommexechinae and Cyrtacantha-
cridinae (Orthoptera: Acrididae).
J. Morph. 67 (2):199-239.

1897b.

1940b.

Thomas, C.

87S: Report upon the collections of
Orthoptera made in portions of
Nevada, Utah, California, Colorado,
New Mexico and Arizona, during
the years 1871, 1872, 1873, and
1874. Rep. Geogr. Geol. Surv.
West of the 100th Meridian in charge

CONALCAEINE GRASSHOPPERS: BARYTETTIX

of G. M. Wheeler, 5 (Zoology):
843-908 [In: Rep. Geogr. Surv.
West of the 100th Meridian].
Washington, D.C.

Uvarov, B. P.

1966a.

Grasshoppers and locusts. A hand-
book of general acridology. Vol. 1,
Anatomy, physiology, development,
phase polymorphism, introduction
to taxonomy. Cambridge Univ.
Press, London. 481 p.

1966b.

111

Hibernation of active stages of Acri-
doidea in temperate climates. Atti
della Accademia Gioenia di Scienze
Naturali in Catania, Serie Sesta,
18:175-189.

White, M. J. D., R. E. Blackith, R. M. Blackith,
and J. Cheney

1967.

Cytogenetics of the Viatica Group
of Morabine grasshoppers. 1. The
“Coastal” species. Australian J.
Zool. 15:263-302.

112

T. J. COHN AND I. J. CANTRALL

Figure 6. Cercus in the species of the genus Baryrettix. Scale below each figure equals 1 mm.

humphreysii cochiset

A. Arizona:

Cochise Co., Douglas, W. W. Jones, 1944 (Paratype)

humphreysii humphreysti

BE Arizona:
CG: Arizona:
D: México:
B; México:
F. México:
G. México:
Jae México:
ig México:
ie México:
K. México:
tridens
ee México:
M. México:
N. México:
O. México:
Pe México:
terminalis
Q. México:
crassus
R. México:
poecilus
S: México:
‘Ti México:
U. México:
V. México:
W. México:
xe México:
paloviridis
V- México:
(hs México:
AA. México:
BB. México:
CC. México:
nigrofasciatus
DD. México:
psolus
EE. México:
contilus contilus
FF. México:

contilus tectatus
GG. México:
contilus hiscatus
HH. México:
contilus dicranatus
Il. México:
continus similis
Ue México:

Cochise Co., 7 mi N Gleeson, T. J. Cohn, 1951

Pima Co., mouth Madera Canyon, T. J. Cohn, 1957 No. 84

Sonora, 17.7 mi N Imuris, T. J. Cohn, 1965 No. 107

Sonora, 14 mi N Imuris, T. J. Cohn, 1965 No. 106

Sonora, 3.4 mi SW Imuris, T. J. Cohn, 1965 No. 104

Sonora, 5 mi NW Guaymas, T. J. Cohn, 1966 No. 26

Sonora, 33 mi SE Guaymas, T. J. Cohn and E. R. Tinkham, 1957 No. 90
Sonora, 24.7 mi NE Esperanza on Tezopaco Rd., T. J. Cohn, 1966 No. 29
Sonora, 1 mi NW Alamos, T. J. Cohn, 1959 No. 236

Sonora, 21 mi SE Navojoa, T. J. Cohn, 1966 No. 32

6 mi NE El Fuerte, T. J. Cohn, 1966 No. 49

17.8 miN Los Mochis, T. J. Cohn, 1965 No. 93

Sinaloa, 6 mi W Jesus Marta, T. J. Cohn, 1958 No. 279

Sinaloa, 15 mi N Culiacan, T. J. Cohn and E. R. Tinkham, 1957 No. 106
Sinaloa, 63 mi NW (old) Mazatlan Airport, T. J. Cohn, 1958 No. 257

Sinaloa,
Sinaloa,

Jalisco, 16.2 mi NE Barra de Navidad, T. J. Cohn, 1968 No. 33

Territorio de Baja California, San José del Cabo, ANSP (Topotype)

Sinaloa, 66 mi SE Culiacan, T. J. Cohn, 1958 No. 258

Sinaloa, El Venadillo, T. J. Cohn, 1959 No, 214

Sinaloa, El Venadillo, I. J. Cantrall and T. J. Cohn, 1961 No. 55

Sinaloa, 6.7 mi NE Concordia, I. J. Cantrall and T. J. Cohn, 1961 No. 61B

Nayarit, 1 mi W Acaponeta, T. J. Cohn, 1958 No. 303

Nayarit, 8.5 mi SE Acaponeta, I. J. Cantrall and T. J. Cohn, 1961 No. 64

Sonora, 68 mi S Hermosillo, T. J. Cohn, 1966 No. 24

Sinaloa, 29.2 mi N Los Mochis (turn off), T. J. Cohn, 1966 No. 35

Sinaloa, 42 mi NW Culiacan, T. J. Cohn and E. R. Tinkham, 1957 No. 98

Sinaloa, | mi W Tecorito, T. J. Cohn, 1959 No. 221

Sinaloa, 2.5 mi NW Culiacan, I. J. Cantrall and T. J. Cohn, 1961 No. 56 (Holotype)
Sinaloa, 8.3 mi SW Santa Lucia, I. J. Cantrall and T. J. Cohn, 1961 No. 52 (Holotype)
Sinaloa, 6 mi W Jesus Marta, T. J. Cohn, 1958 No. 279 (Holotype)

Sinaloa, 3 mi NE Tepuche, T. J. Cohn, 1966 No. 40 (Holotype)

Sinaloa, 9 mi SE Culiacan, T. J. Cohn, 1958 No. 288 (Holotype)

Sinaloa, 19.9 mi SE Culiacan, I. J. Cantrall and T. J. Cohn, 1961 No. 59 (Holotype)
Sinaloa, 30 mi S Culiacan, T. J. Cohn, 1965 No. 85 (Holotype)

Sinaloa, 66 mi SE Culiacan, T. J. Cohn, 1958 No. 258 (Holotype)

CONALCAEINE GRASSHOPPERS: BARYTETTIX

H. COCHISE!

(—)

A

H. HUMPHREYSI1

SL Lt

|

H, HUMPHREYSI1 TRIDENS

\6
<

aoe
ogee

|

K ———

oO

TRIDENS CRASSUS

POECILUS

i=)

PALOVIRIDIS

113

CONTILUS

¢. CONTILUS

C. HISCATUS

C. DICRANATUS

114 T. J. COHN AND I. J. CANTRALL

POECILUS CRASSUS TERMINALIS

AA
F

a

g

igure 7; Concealed male genitalia in the Crassus Group of the genus Barytettix. Scale below each figure equals 0.5 mm.
E: lateral views. B,D, F: ventral views.

C,

B. poecilus. México: Sinaloa, El Venadillo, I. J. Cantrall and T. J. Cohn, 1961 No. 55

aD: crassus. México: Territorio de Baja California, San José del Cabo (ANSP) (Topotype). In glycerine
F. terminalis. México: Jalisco, 16.2 mi NE Barra de Navidad, T. J. Cohn, 1968 No. 33 (Holotype)

CONALCAEINE GRASSHOPPERS: BARYTETTIX 115

POECILUS
POECILUS

POECILUS
POECILUS POECILUS

TRIDENS
H. COCHISEI ZY

H. HUMPHREYSII

Figure 8. Genitalic structures and variation in poecilus and the Humphreysii Group of the genus Barytettix. Scale below
each figure equals 0.5 mm.

A-E. poecilus
Male supra-anal plate. México: Sinaloa, 2.4 mi NE Villa Union, T. J. Cohn, 1965 No. 82
Male supra-anal plate. México: Sinaloa, 1.3-2.6 mi NW Villa Union, T. J. Cohn, 1965 No. 81
Dorsal valve of aedeagus, dorsal view. México: Sinaloa, El Venadillo, I. J. Cantrall and T. J. Cohn, 1961 No.55
Ventral valve of aedeagus, ventral view. México: Sinaloa, 66 mi SE Culiacan, T. J. Cohn, 1961 No. 258
Ventral valve of aedeagus, ventral view. México: Jalisco, 5.4 mi E Plan de Barrancas, T. J. Cohn, 1965 No. 73
F-H. Humphreysii Group
F. humphreysii cochisei. Dorsal view of aedeagus. Arizona: Cochise Co., Douglas, W. W. Jones, 1944 (Paratype)
G h. humphreysti. Dorsal view of aedeagus. México: Sonora, 37.2 mi W Santa Ana, T. J. Cohn, 1966 No. 51
H. tridens. Dorso-lateral view of tip of abdomen. México: Sinaloa, 1 miW Tecorito, T. J. Cohn, 1959 No. 221

moOp>

116 T. J. COHN AND I. J. CANTRALL

Figure 9.

Concealed male genitalia in the Humphreysii Group of the genus Barytettix. Scale below each figure equals 0.5
mm. Note:

B, D, and F do not include the structures associated with the ramus of the cingulum (as in A, C, and E) and thus

are comparable only with the distal portions of Figs. 7 B, D, F and Figs. 13 B, D, F. Lateral views of genital mass: A, C, E. Ventral
views of aedeagus: B, D, F.

A, B.
CD:
EES

>

humphreysii humphreysii. México: Sonora, 17.7 mi N Imuris, T. J. Cohn, 1965 No. 107
humphreysii humphreysii. México: Sonora,36 mi SE Guaymas, T. J. Cohn and E. R. Tinkham, 1957 No. Wid
tridens. México: Sinaloa, 63 mi NW (old) Mazatlan Airport, T. J. Cohn, 1958 No. 257

CONALCAEINE GRASSHOPPERS: BARYTETTIX

H. HUMPHREYSII \

H. HUMPHREYSII

TRIDENS

117

118

Figure 10.

T. J. COHN AND I. J. CANTRALL

Dorsal valve in dorsal view in the Humphreysii Group of the genus Barytettix. Scale below each figure equals

0.5 mm except for D.

humphreysii cochiset
New Mexico: Hidalgo Co., 1.5 mi E Rodeo, N. D. Jago, 1968 (ANSP)

A.
B.
Cc:
D.
Es

México:

Sonora, 7-10 mi SE Agua Prieta, V. Roth, T. J. Cohn and J. W. Cohn, 1970 No. 23

Arizona: Cochise Co., Douglas, W. W. Jones, 1944 (Paratype)
Arizona: Cochise Co., Douglas, W. W. Jones, 1944 (Holotype, USNM)

México:

Sonora, 10.4 mi S Hermosillo, T. J. Cohn, 1965 No. 101

humphreysti humphreysii
Arizona: Cochise Co., 7 mi N Gleeson, T. J. Cohn, 1951

EF;

POROZEM AST EO

trid

®

ns

iva

México:
México:
México:
México:
México:
México:
México:
México:
México:
México:
México:
México:

México:
México:

Sonora, 17.7 mi N Imuris, T. J. Cohn, 1965 No. 107

Sonora, 3.4 mi N Imuris, T. J. Cohn, 1965 No. 104

Sonora, 5.5 mi SW Magdalena, T. J. Cohn, 1966 No. 103

Sonora, 14 mi N Hermosillo, T. J. Cohn, 1966 No. 21

Sonora, 4.2 mi N Guaymas, T. J. Cohn, 1959 No. 237

Sonora, 33 mi SE Guaymas, T. J. Cohn and E. R. Tinkham, 1957 No. 90
Sonora, 37 mi SE Guaymas, T. J. Cohn and E. R. Tinkham, 1957 No. 111
Sonora, Tezopaco, F. Pacheco M., 1956

Sonora, 24.7 mi NE Esperanza on Tezopaco Rd., T. J. Cohn, 1966 No. 29
Sonora, 1 mi NW Alamos, T. J. Cohn, 1959 No. 236

Sonora, 21 mi SE Navojoa, T. J. Cohn, 1966 No. 32

Sonora, 51.8 mi SE Navojoa, T. J. Cohn, 1966 No. 34

Sinaloa, 6 mi W Jesus Marta, T. J. Cohn, 1958 No. 279
Sinaloa, 63 mi NW (old) Mazatlan Airport, T. J. Cohn, 1958 No. 257

CONALCAEINE GRASSHOPPERS: BARYTETTIX 119

H. COCHISE! H. HUMPHREYSI1 H. HUMPHREYS11 H. HUMPHREYSII

Ww? LEN

Ov" ‘
(Omnawen
o

as
=

ie)

TRIDENS

["\

N
O- VW

120

T. J. COHN AND I. J. CANTRALL

Figure 11. Ventral valves in ventral view in the Humphreysii Group of the genus Barytettix. Scale below each figure equals

0.5 mm.

humphreysti cochisei

A.

B.

Arizona: Cochise Co., Douglas, W. W. Jones, 1944 (Paratype)

México: Sonora, 10.4 miS Hermosillo, T. J. Cohn, 1965, No. 101

cochisei-humphreysii hybrids

Cc.

D.
humphreysti humphreysit

EB.

a
&
S

Sr Ain ee els GOT

a

VOZ

Arizona: Cochise Co., Don Luis, Rehn and Hebard, 1922 (cochisei paratype) (USNM)
México: Sonora, 0.7 mi N Los Pocitos (36.8 mi S Hermosillo Cathedral), T. J. and J. W. Cohn, 1970 No. 68

Arizona: Santa Cruz Co., 4 miE Arivaca, T. J. Cohn, 1951
Arizona: Pima Co., mouth Madera Canyon, T. J. Cohn and E. R. Tinkham, 1957 No. 84

México:
México:
México:
México:
México:
México:
México:

México:
México:
México:

Sonora, 48 mi S Hermosillo, T. J. Cohn, 1965 No. 99

Sonora, 4.2 mi N Guaymas, T. J. Cohn, 1959 No. 237

Sonora, Tezopaco, F. Pacheco M., 1956

Sonora, 24.7 mi NE Esperanza on Tezopaco Rd., T. J. Cohn, 1966 No. 29
Sonora, 10 mi S Cd. Obregon, F. Orozco, 1957

Sonora, 1 mi NW Alamos, T. J. Cohn, 1959 No. 236

Sonora, 51.8 mi SE Navojoa, T. J. Cohn, 1966 No. 34

Sinaloa, 6 mi NE El Fuerte, T. J. Cohn, 1966 No. 49
Sinaloa, 17.8 mi N Los Mochis, T. J. Cohn, 1965 No. 93
Sinaloa, 6 mi W Jesus Marta, T. J. Cohn, 1958 No. 279

CONALCAEINE GRASSHOPPERS: BARYTETTIX 121

H. COCHISEI H. HUMPHREYSII H. HUMPHREYS] H. HUMPHREYSII

TRIDENS
ee
B (—

COCHISE! - HUMPHREYS11
HYBRIDS

AeA ee

to
tN

T. J. COHN AND I. J. CANTRALL

Figure 12. Ventral lobes of the sheath in ventral view in the Humphreysii Group of the genus Barytettix. Scale below each
figure equals 0.5 mm.

humphreysti cochisei
A. Arizona: Cochise Co., Douglas, W. W. Jones, 1944 (Paratype)
B. México: Sonora, 10.4 miS Hermosillo, T. J. Cohn, 1965 No. 101
humphreysii humphreysti

ZErASO OMA

2
=.
ss
3

2 ROTO

México:
México:
México:
México:
México:

Sonora,
Sonora,
Sonora,
Sonora,
Sonora,
Sonora,
Sonora,
Sonora,
Sonora,
Sonora,
Sonora,

Sinaloa
Sinaloa
Sinaloa
Sinaloa
Sinaloa

Arizona: Pima Co., mouth Madera Canyon, T. J. Cohn and E. R. Tinkham, 1957 No. 84
México:
México:
México:
México:
México:
Mexico:
México:
México:
Mexico:
México:
México:

14 mi N Imuris, T. J. Cohn, 1965 No. 106

37.2 mi W Santa Ana, T. J. Cohn, 1966 No. 51

4.2 mi N Guaymas, T. J. Cohn, 1959 No. 237

Tezopaco, F. Pacheco M., 1956

24.7 mi NE Esperanza on Tezopaco Rd., T. J. Cohn, 1966 No. 29
10 mi S Cd. Obregon, F. Orozco, 1957

6.8 mi NW Navojoa, T. J. Cohn, 1965 No. 94

1 mi NW Alamos, T. J. Cohn, 1959 No. 236

21 mi SSE Navojoa, T. J. Cohn and E. R. Tinkham, 1957 No. 95
21 mi SE Navojoa, T. J. Cohn, 1966 No. 32

$1.8 mi SE Navojoa, T. J. Cohn, 1966 No. 34

,6 mi NE El Fuerte, T. J. Cohn, 1966 No. 49

, 17.8 miN Los Mochis, T. J. Cohn, 1965 No. 93
, 15.2 mi NW Culiacan, T. J. Cohn, 1966 No. 43
, 6 mi W Jestis Maria, T. J. Cohn, 1958 No. 279

, 33 mi SE Culiacan, T. J. Cohn, 1958 No. 260

CONALCAEINE GRASSHOPPERS: BARYTETTIX 123

TRIDENS

ARN,
MOONY)

ery n w aS |

ae

124 T. J. COHN AND I. J. CANTRALL

PSOLUS C. HISCATUS

PALOVIRIDIS

Figure 13. Concealed male genitalia in the Psolus Group of the genus Barytettix. Scale below each figure equals 1 mm. A, C, E:

lateral views. B, D, F: ventral views.

paloviridis. México: Sinaloa, 39.9 mi SE Culiacan, I. J. Cantrall and T. J. Cohn, 1961 No. 60

psolus. México: Sinaloa, 15 miN Culiacan, E. R. Tinkham and T. J. Cohn, 1957 No. 106
Sinaloa, 19.9 mi SE Culiacdn, I. J. Cantrall and T. J. Cohn, 1961 No. 59 (Holotype)

A, B.
Ci D:
EOF.

>

contilus hiscatus. México:

CONALCAEINE GRASSHOPPERS: BARYTETTIX 1125

Cc. CONTILUS Cc. TECTATUS Cc. HISCATUS Cc. DICRANATUS Cc. SIMILIS PALOVIRIDIS PSOLUS

NIGROFASCIATUS

Figure 14. Aedeagal structures in the Psolus Group of the genus Barytettix. Scale below each figure equals 0.5 mm. A-H. Dorsal

views of aedeagus. I-P: ventral views of aedeagus. Q-X: ventral lobes of sheath, ventral views (as if lobes are closely appressed to
ventral valves).

contilus contilus. México: Sinaloa, 3 mi NE Tepuche, T. J. Cohn, 1966 No. 40 (Holotype)

contilus tectatus. México: Sinaloa, 9 mi SE Culiacan, T. J. Cohn, 1958 No. 288 (Holotype)

contilus hiscatus. México: Sinaloa, 19.9 mi SE Culiacan, I. J. Cantrall and T. J. Cohn, 1961 No. 59 (Holotype)
contilus dicranatus. México: Sinaloa, 30 mi S Culiacan, T. J. Cohn, 1965 No. 85 (Holotype)

contilus similis. México: Sinaloa, 66 mi SE Culiacan, T. J. Cohn, 1958 No. 258 (Holotype)

paloviridis. México: Sinaloa, 2.5 mi NW Culiacan, I. J. Cantrall and T. J. Cohn, 1961 No. 56 (Holotype)
psolus. México: Sinaloa, 6 mi W Jestis Marfa, T. J. Cohn, 1958 No. 279 (Holotype)

nigrofasciatus. México: Sinaloa, 8.3 mi SW Santa Lucia, I. J. Cantrall and T. J. Cohn, 1961 No. §2 (Holotype)

or
ZO

TAMmMUOw>
OZEIRIA

ro:
ae ee

126 T. J. COHN AND I. J. CANTRALL

PALOVIRIDIS
* PALOVIRIDIS PALOVIRIDIS

Bay | PALOVIRIDIS PALOVIRIDIS

PALOVIRIDIS
PALOVIRIDIS Q} (}
L M 5

PALOVIRIDIS PALOVIRIDIS PSOLUS

oO.) O/|

N ———— 9 —____. P

NIGROFASCIATUS

PALOVIRIDIS

Figure 15. Aedeagal details and tip of abdomen in the Psolus Group of the genus Barytettix. Scale below each figure equals 0.5 mm,

A-E. Side views of aedeagus
A. paloviridis. México: Sonora, 18.2 mi N Guaymas, T. J. Cohn, 1968 No. 63
B. paloviridis. México: Sonora, 68 miS Hermosillo, T. J. Cohn, 1966 No. 24
(ee paloviridis. México: Sinaloa, 3.6 mi W Guamtchil, T. J. Cohn, 1965 No. 90
D. paloviridis. México: Sinaloa, 2.5 mi NW Culiacan, I. J. Cantrall and T. J. Cohn, 1961 No. 56 (Holotype)
E. nigrofasciatus. México: Sinaloa, 8.3 mi NW Santa Lucia, I. J. Cantrall and T. J. Cohn, 1961 No. 52 (Holotype)
F-G. Ventral views of aedeagus in paloviridis
Bs México: Sinaloa, 10.9 mi W Guamichil, T. J. Cohn, 1968 No. 11
G. México: Sinaloa, 3.6 mi W. Guamtchil, T. J. Cohn, 1965 No. 90
He Dorso-lateral view of tip of abdomen in paloviridis
H. México: Sinaloa, 17 mi SE Culiacan, T. J. Cohn, 1958 No. 294
1-O. Distal views of ventral loves of the sheath
l. contilus contilus. México: Sinaloa, 3 mi N Tepuche, T. J. Cohn, 1966 No. 40
Ue contilus hiscatus. México: Sinaloa, 19.9 mi SE Culiacan, T. J. Cohn, 1961 No. 59 (Holotype)
K. nigrofasciatus. México: Sinaloa, 8.3 mi SE Santa Lucia, I. J. Cantrall and T. J. Cohn, 1961 No. 52 (Holotype)
L. paloviridis. México: Sinaloa, 10.9 miW Guamtchil, T. J. Cohn, 1968 No. 11
M. _ paloviridis. México: Sinaloa, 17 mi SE Culiacan, T. J. Cohn, 1958 No. 294
N. paloviridis. México: Sinaloa, El Bario, T. J. Cohn, 1959 No. 231
O. psolus. México: Sinaloa, El Bario, T. J. Cohn, 1959 No. 231
P-R. Dorsal views of tip of aedeagus
Pe paloviridis. México: Sinaloa, 30 mi S Culiacan, T. J. Cohn, 1965 No. 85
Q: paloviridis. México: Sinaloa, 33 mi SE Culiacan, T. J. Cohn, 1958 No. 260
R. psolus. México: Sinaloa, 3 mi NW Culiacan, T. J. Cohn, 1968 No. 267

CONALCAEINE GRASSHOPPERS: BARYTETTIX 127

WL 7 =>
Ww f / CN
Cy )
( (ae.

H. HUMPHREYS11

Figure 16. Ovipositor and receptaculum seminis, and aedeagus of representatives of the groups of Barytettix. A, C, E: female
structures in dorsal view. Scale below each figure equals 1 mm. B, D, F: male aedeagus in ventral view (as they would be inserted
into the female) drawn to the same scale as that of the female structures.

ee

F

Symbols

ABS

poecilus. México: Sinaloa, 84 mi SE Culiacan, T. J. Cohn, 1958 No. 244

poecilus. México: Sinaloa, Fl Venadillo, 4 mi N (old) Mazatlin Airport, I. J. Cantrall and T. J. Cohn, 1961 No. 55
(Topotype)

paloviridis. México: Sinaloa, 2.5 mi NW bridge over Rio Culiacan at Culiacan, I. J. Cantrall and T. J. Cohn, 1961
No. 56 (Topotype)

paloviridis. México: Sinaloa, 39.9 mi SE Culiacan, I. J. Cantrall and T. J. Cohn, 1961 No. 60

humphreysii humphreysii. Arizona: Pima Co., Sabino Canyon, Santa Catalina Mts., M. Cazier, P. Boone and T. J. Cohn,
1950

humphreysii humphreysii. México: Sonora, 17.7 mi N Imuris, T. J. Cohn, 1965 No. 107

used in figures:

— anterior basivalvular sclerite
— apical diverticulum

— bursa copulatrix

— Comstock-Kellogg Gland

— constricted tube

— egg guide

— preapical diverticulum

— secondary lobe preapical diverticulum
— thin tube

— ventral ovipositor valve

— vestibule

128 T. J. COHN AND I. J. CANTRALL

POECILUS H, COCHISE!
A H, HUMPHREYSI1

' i
H
\ ; + eee
POECILUS H \ -
D ——

H. HUMPHREYSII H. COCHISE!

Figure 17. Bursa copulatrix and thick tube of receptaculum seminis in the Crassus and Humphreysii Groups of the genus Barytettix.
Scale below each figure equals 0.5 mm. A-C: side views, left side of figures is dorsal, D-F: dorsal views.

A; -D: poecilus. México: Sinaloa, El Venadillo, T. J. Cohn, 1959 No. 214

Baye humphreysii humphreysii. Arizona: Pima Co., Sabino Canyon, M. Cazier, P. Boone, and T. J. Cohn, 1950

Cok: humphreysii cochisei. Arizona: Cochise Co., Douglas, W. W. Jones, 1932

Dashed lines encompass shallow concavity in the roof of the genital chamber (see Fig. 2, p. 17, dorsal portion of the genital chamber).

CONALCAEINE GRASSHOPPERS: BARYTETTIX 129

C. SIMILIS C. CONTILUS C. TECTATUS PSOLUS NIGROFASCIATUS PALOVIRIDIS

2

z\

OO gs Te

Figure 18. Bursa copulatrix and thick tube of receptaculum seminis in the Psolus Group of the genus Barytettix. Scale below
each figure equals 1 mm. A-F: side views, left side of the figure is dorsal. G-L: dorsal views.

A,G. contilus similis. México: Sinaloa, 66 mi SE Culiacan, T. J. Cohn, 1958 No. 258

B, H. contilus contilus. México: Sinaloa, 3.1 mi NE Tepuche, T. J. Cohn, 1966 No. 39

CT: contilus tectatus. México: Sinaloa, 7.2 mi SE Culiacan (Cerro Tule Rd.,), T. J. Cohn, 1966 No. 38
Daye psolus. México: Sinaloa, 15.2 mi NW Culiacan, T. J. Cohn, 1966 No. 43

E, K. nigrofasciatus. México: Sinaloa, 3.6 mi NE Santa Lucia, I. J. Cantrall and T. J. Cohn, 1961 No. 50
lanl be paloviridis, México: Sinaloa, 2.5 mi NW Culiacan, I. J. Cantrall and T. J. Cohn, 1961 No. 56

Dashed lines encompass shallow concavity in the roof of the genital chamber (see Fig. 2, p. 17, dorsal portion of the genital chamber).

T. J. COHN AND I. J. CANTRALL

130
SPECIES LENGTH OF LENGTH OF LENGTH - OF LENGTH OF WIDTH OF
BODY PRONOTUM CEPHALIC 1 EMUR CAUDAL FEMUR CAUDAL FEMUR
MALES
Felon alee! CO ee a) fa oe Fa | (era re oe oes |
20 25 30 35 4O 45 SO S45 6:7) 8: (9! 110 Site SAGE.
POECILUS —_—_ — Fe
CRASSUS — — —
TERMINALIS =_= _- —_—
PALOVIRIDIS —— ee —— _
PSOLUS —_ — _
NIGROFASC IATUS — = oI
C. CONTILUS = —_ ——
C. TECTATUS a = —
C. HISCATUS = = -
C. DICRANATUS > f= ce
C. SIMILIS =| raat oe
H. HUMPHREYSII ayo bar ra er re a
H. COCHISEI —) ee a
TRIDENS ———— aor a=
1
FEMALES

POECILUS

CRASSUS

TERMINALIS
PALOVIRIDIS
PSOLUS
NIGROFASCIATUS
. CONTILUS

- TECTATUS

. HISCATUS

DICRANATUS

- SIMILIS

HUMPHREYSII

COCHISEI

TRIDENS

20 25 3B 35 40 45 S50
t

4 i

—e

1

J

Figure 19. Summary of measurements of the species of Barytettix.

Ranges of measurements of males (above) and females (below).

CONALCAEINE GRASSHOPPERS: BARYTETTIX 131

eerereS WIDTH OF LENGTH OF LENGTH OF WIDTH OF
TEGMEN TEGMEN SUPRAANAL PLATE SUPRAANAL PLATE
MALES
f mal T T 46 7 SU ALL Lime! er eg en r i T T T= 1
TOMS e2s0n2esnse0Ns. 2345678 1.0 1.5 2.0 2.5 3,0 1,5 2,0 2.5 3.0 3.5 4.0
|
POECILUS — — — —
CRASSUS at i — ———
TERMINALIS - - 4 =
PALOVIRIDIS on — ba tt
PSOLUS _ — —— —
NIGROFASCIATUS = « mall Le
C. CONTILUS - - - —
C. TECTATUS - —_ — =
. HISCATUS C » L 4
. DICRANATUS — — - —
. SIMILIS —_ = — =
. HUMPHREYSII —— — —— Ee
H. COCHISEI - —_ — mes
TRIDENS —- — — —-
FEMALES
POECILUS oa pe
CRASSUS — —_
TERMINALIS — =
PALOVIRIDIS ———ta i
PSOLUS —_————_ a
NIGROFASCIATUS _—_ ak
C. CONTILUS — -
C. TECTATUS - -
C. HISCATUS . .
C. DICRANATUS S a
C. SIMILIS : -
H. HUMPHREYSII ——_ os
H. COCHISEI ~ -
TRIDENS — -
1.01.5 2.0 2.5 3.0 3.5 2345678 1.@ NS Bs) 255 Saw) 1.5 2.0 2.5 3.0 3.5 4.0
oe Pot tt | ee L 1 —t. 1 4

Figure 20. Summary of measurements of the species of Barytettix. Ranges of measurements of males (above) and females (below).

irs .
r eee i!
re eee aie
iis dine oe aa hates i
Stes rae eee gow

Base

are eee SN

7 creat meig, 3:

1 we

ead ied Se ea lince a ete — CE A IREY Ne ree” areas
‘ilar, SS a ~

a

BOUND SEP 1974

i a ee [* “ ‘s
bi el \s "

iT i

han) an uy b
iat

= : — = “a — —
= - - 3 : F
: . ; > : aK = — vas. >
: : ¥ : eS PE ea :
5 Sepa —_—~—_ a
ne a, A

| i : y!
Wis

[gs

Gtdulai a i

i f

ay

ya Cau | a a

4 mr

3 i edn Aa | 39
} ] i 1 ui Vv [i
Aah Haae a ' | OP) © cna ned 1 tik ; i.
qe fh it ‘ Bia fiipatl, Wa alo
Fe Hae rc i anh joe i f Rarihe
Se bu alae Ce Rau ee i
lt en re AA ie ‘dle die dD Ri i ou
ff ry itd, hi Hy ey On sa ref
ail | Hi. i a nie ie
tai a Ae ih Aa e
ie i oh ‘ ‘ AAG Sta eran ORL
a ig if re" Ul Ae | i i My A HS Xi , mi ie |
iN ! \ hie mn sh i id i)! gl
Beles
cee Mi ci tee
I Taian
if i bia os f i:
(at ane es

i
A a the

Pye

Te sf

Ue

q
‘

Et NN
Seay,

SORES EASES

poe
payee a see

ee