m i. .
# *'■ iii-'
«: S*' •■
9'"»:ri'v*
:«>; >
S !>'
.*,>-»'
.*,»■..
r »■ p. »■ I
5?- »>
J». *. B
If fc
*\ *
VIM* I ^ **••»'■
* 9 f- ■
>- *
tt i^
*'«^-
»■' *
* V
|
\ |
|
r |
P P^f. ^- ¥ k
k-' V
w- ,^..- -ai> ,«.
H^-^o
HARVARD UNIVERSITY
Library of the
Museum of
Comparative Zoology
bulletin OF THE
Museum of Comparative Zoology ^
^-
Volume 142 1971
HARVARD UNIVERSITY
CAMBRIDGE, MASSACHUSETTS 02138 U.S.A.
CONTENTS
No. 1. The Systematics, Distribution, and Zoogeography of the Marine Hatchetfishes (Family Stenioptychidae ) . By Ronald Clay Baird. October, 1971 1
No. 2. The Carabid Beetles of New Guinea. Part IV. General Considera- tions; Analysis and History of Fauna; Taxonomic Supplement. By P. J. Darlington, Jr. October, 1971 129
No. 3. A New Rhinoceros from the Late Miocene of Loperot, Turkana
District, Kenya. By D. A. Hooijer. October, 1971 . 339
No. 4. Osteology of the Malaysian Phallostethoid Fish Ceratostethus hicornis, With a Discussion of the Evolution of Remarkable Struc- tural Novelties in its Jaws and External Genitalia. By Tyson R. Roberts. December, 1971 393
No, 5. Revision of North American Ciidae ( Coleoptera ) . By John F.
Lawrence. December, 1971 . 419
^^^^^^^^l^^-^;!?!l<jriP^Kf- '' ■' '
^fflff^^^L
Museum of
Comparative
Zoology
S^^itisatf)^ •id^fj?«»'.:<j^"?^' lAIi^ ^- ^ j,ii5 ri.-:-:^^^
The Systematics, Distribution, and
Zoogeography of the Marine Hatchetfishes
(Family Sternoptychidae)
RONALP CLAY BAiRD
HARVARD UNIVERSITY VOLUME 142, NUMBER 1
CAMBRIDGE, MASSACHUSETTS, U.S.A. 18 OCTOBER 1971
PUBLICATIONS ISSUED
OR DISTRIBUTED BY THE
MUSEUM OF COMPARATIVE ZOOLOGY
HARVARD UNIVERSITY
Bulletin 1863-
Breviora 1952-
Memoirs 1864-1938
JoHNSONiA, Department of Mollusks, 1941-
OccAsiONAL Papers on Mollusks, 1945-
Other Publications.
Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth.
Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth.
Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth.
Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposimn on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth.
Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. )
Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae ( Mollusca: Bivalvia ) . $8.00 cloth.
Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth.
Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. )
Publications of the Boston Society of Natural History.
Authors preparing manuscripts for the Btilletiii of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, PubHcations Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A.
© The President and Fellows of Harvard College 1971.
THE SYSTEMATICS, DISTRIBUTION, AND ZOOGEOGRAPHY OF THE MARINE HATCHETFISHES (FAMILY STERNOPTYCHIDAE)
RONALD CLAY BAIRD'
TABLE OF CONTENTS
Introduction 2
Methods 3
Osteological Character Complexes 7
Osteological Conckisions 15
The Fossil Record 22
Systematics 29
Family Stemoptychidae 29
Key to the genera of Stemoptychidae 29
Genus Argyropelectis 31
Key to the species of Argijropelecus 32
Argyropelectis affinis 34
Argyropeleciis gigas 38
Argyropdecus hemigymniis 42
Argyropeleciis aculeatus 48
Argyropelecus olfersi 52
Argyropeleciis sladeni 56
Argyropelecus lynchus -. 63
C^enus Sternopiyx 67
Key to the species of Sternoptyx 68
Stcrnoptyx ohscura 69
Sternoptyx pseudobscura 72
Sternoptyx diaphana 75
Genus Polyipnus 79
Key to the species of Polyipnus ^ 82
Polyipnus tridentifer 86
Polyipnus sterope 88
Polyipnus spinosus 89
Polyipnus oluolus 90
Polyipnus indicus 91
Polyipnus niittingi 92
Polyipnus laternatus 92
Polyipnus oniphus 94
Polyipnus aquavitus 96
Polyipnus iinispinus 97
Polyipnus triphanos 97
^ Museum of Comparative Zoology, Hai"vard Uni- versity, and Marine Science Institute, University of South Florida, St. Petersburg, Fla. 33701.
Polyipnus polli 98
Polyipnus asteroides 99
Polyipnus matsuharal 101
Polyipnus ruggcri 102
Polyipnus kiiviensis 103
Polyipnus meteori 104
Discussion and Conclusions 104
Patterns of Distrilaution 104
Geographic Variation 106
Distributional Factors 107
Zoogeographic Regions 110
Acknowledgments 1 14
Summary 115
Appendix A ^ 115
Appendix B 117
Literature Cited 123
ABSTRACT
The systematic history of tlie Stemoptychidae has been one of instability in higher classification. A study of comparative osteology indicates that die hatchetfishes are closely related to the Gonosto- matidae but differ from them in certain significant aspects. The Stemoptychidae are therefore given familial rank.
Fossil evidence indicates that the family prob- ably arose during the early Tertiary and reached its present evolutionary grade by the middle Miocene. Three phylogenetically divergent genera are recog- nized, these being Argyropelecus, Polyipnus, and Sternoptyx, with seven, seventeen, and three species respectively. Many species exhibit geographical variation and morphologically distinct populations were defined in some instances.
The genera differ broadly in habitat as well as morphology. Argyropelecus is a high seas pelagic genus limited to the upper 600 m. Sternoptyx shows a similar pattern horizontally but inhabits the 500 to 1500 m depth zone. Polyipnus occurs only in close association with land, exhibiting a chstribu- tion and speciation pattern similar to many tropical shore species. Argyropelecus and Sternoptyx spe-
Bull. Mus. Conip. Zool., 142(1): 1-128, October, 1971
Bulletin Museum of Comparafive Zoology, Vol. 142, No. 1
cies are seemingly restricted to waters with similar hydrographie and biological properties. Certain species assemblages are used to define zoogeo- graphically distinct areas of the world's oceans.
INTRODUCTION
The systematic histoiy of marine hatchet- iishes begins with Hermann's ( 1781 ) description of a photophore-bearing fish he called Sternoptijx diaphana (from the Greek words "sternon" (chest) and "ptyx" ( plate ) ) and from which the family derives its name. Hermann called at- tention to the extraordinarily deep and compressed body shape and thus estab- lished one of the principal descriptive characteristics of the group. The genus Arp,\iropclcctis was described by Cocco in 1829 and both genera then appeared in the classic work of Cuvier and Valen- ciennes (1849).
Giinther (1864: 384) placed the above genera in the family Sternoptychidae and included also other midwater genera (presently placed in the family Gono- stomatidae) using such characteristics as photophorcs and gill structures. In addi- tion, Argyropeleciis and Sternoptijx were given subfamilial rank characterized by the presence of a spinous dorsal blade.
Gill (1884), while recognizing that the congener of the Gonostomatine fishes was allied to the Sternoptychidae, nevertheless restricted the family to include only Sternoptijx and Argyropeleciis. He recog- nized too, a degree of difference between these genera and gave them subfamilial I'ank.
Giinther (1887) added the newly de- scribed genus FoJijipnus to his family Sternoptychidae, which still included the present gonostomatid genera. Goode and Bean (1896) followed Gill in recognizing three families from Giinther's one: Gono- stomatidae, Maurolicidae, and Sternopty- chidae; and in addition, placed the genus Pohjipniis with the Sternoptychidae. Gar- man (1899), citing the similarity of the larval forms of Goode and Bean's three
families, returned to Giinther's original scheme. Brauer (1906: 101) later con- tinued to recognize Giinther's classification.
Regan ( 1923 ) attempted to clarify the earlier confusion by examining osteological characters, thereby giving more explicit definitions to the taxa. This resulted in assignment of the genera Sternoptyx, Argy- ropeleciis, and Folyipmis to the family Sternoptychidae, while other related gen- era were placed in the family Gonostomati- dae. Such basic differences as the absence of a basi- and alisphenoid bone in the former family were cited as justification for this split. Regan's classification was later accepted by Norman (1930, 1944) and Berg (1940).
Regan's work did not resolve the prob- lem of family relationship and taxon rank, however. While generally recognizing Regan's classification. Fowler ( 1936 ) gave the Maurolicidae familial rank and fur- ther complicated the issue by including the genus Valenciennelliis with Regan's Sternoptychidae. Gregory and Conrad (1936) included Maiiroliciis in the family Sternoptychidae, acknowledging the primi- tiveness of this genus, as well as its role as a possible congener of Regan's Sternoptychidae. They cited the deep, compressed body form as an evolutionary trend in the family. Smith (1953) essen- tially returned to Goode and Bean's old classification while Hubbs (1953), refer- ring to the connecting links in the evo- lution of the Sternoptychidae from the Gonostomatidae, recommended a revival of Giinther's classification, thus reducing Regan's Sternoptychidae to subfamilial rank. While Rechnitzer and Bolilke (1958) and Ebeling ( 1962 ) have accepted Hubbs' l^roposal, most modern authors recognize Regan's classification (Schultz, 1961; Mor- row, 1964; Backus et al., 1965; Berry and Perkins, 1965). However, Greenwood et al. (1966) indicate that the Sternoptychidae are a specialized offshoot of the Gono- stomatidae and. although still recognizing the former as a separate family, they
Marine Hatchetfishes • Baird
suspect that further morpliological study will support the earher conckisions ot Hubl)s.
Historically then, there has been a failure to achieve a stable classification of the Stemoptychidae. The numerous reasons for this failure may be attributed primarily to the use of superficial or highly variable character complexes, the lack of detailed morphological studies using osteological or other acceptable criteria, and subjective conceptual differences concerning the fam- ily rank.
The first consideration of a systematic study of the Stemoptychidae must include an attempt to clarify some of the historical confusion. Accordingly, a comparative study of primarily osteological character complexes was undertaken with the follow- ing objectives: to help elucidate the family question; to provide characters for explicit definitions of the taxa; and to comment on generic relationships and evolutionary trends among the genera and species com- plexes. The character complexes cited were sufficiently numerous and function- ally distinct to reasonably satisfy the initial objectives. The gonostomatid genera Maii- rolicus and Valenciennelhis were chosen for comparison with the Stemoptychidae as they are thought, classically, to be most closely related to them, and because any other choice would have to involve a de- tailed study of the Gonostomatidae.
The use of osteological characters and character complexes as the primaiy criteria in a systematic study invcjlves the follow- ing concepts: 1) The skeletal system is a major constituent of the functional mor- phology of an indi\ idual and should reflect its general evolutionary history. 2) As selection acts on a particular morphological regicjn, it alters the osteology of that region. Both between and within regions, osteological characters may be independent with regard to rate and direction of evo- lution. 3) The skeletal system is not strictly a single one with a limited function and morphology. Rather, it may be thought of
as a series of semi-independent systems or "functional units," each reflecting the functional requirements of that particular unit. 4) An osteological study results in a composite of individual character com- plexes, some of which may be primitive, others advanced, but which reflect the evo- lution and specialization within i^hyletic lines. 5) Osteological characters have been shown to be as consistent as other characters in reflecting phylogeny and evolution. We know more about osteology and its limitations. 6) Falcon tological evi- dence is primarily osteological.
Fossil evidence was also considered and a detailed study was made of the fossil record to provide additional information on the evolutionary history and relation- ships ascertained from the osteological results. After using these in resolvhig the family question and in presenting an evo- lutionaiy history, the various higher taxa were defined and a revision of the re- spective genera undertaken.
The widespread occurrence and ease of capture of the Stemoptychidae make them ideally suited for studies involving popu- lation structure, speciation, and distri- bution in the midwater or mesopelagic environment. Several recent studies ( Haff- ner, 1952; Ebeling, 1962; Nafpaktitis, 196S) have indicated some of the distributional patterns of certain midwater fishes and the possible factors involved therein. This study attempts to examine some of these factors with regard to present sternopty- chid distributions.
METHODS
Material. Because of the vast amount of material examined a detailed list of specimens and stations is not included in this work. Appendix A lists the institutions, vessels, and respective cruises from \\'hich material was obtained. A detailed listing of material examined is on permanent file in the Museum of Comparative Zoology, Harvard University.
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Collecting:, and sampling techniques. In a study such as this one, involving material from so many cruises employing a wide variety of gear and using various fishing philosophies, the sources of sampling bias are too numerous to list. However, some of the major problems can be discussed.
Horizontally, there is a marked differ- ence in the amount of sampling between areas. A few areas have been adequately sampled (California, North Atlantic) while others have not been sampled at all. The Pacific in general, the South Atlantic, and the Indian Ocean — especially the southern and eastern portion — are markedly under- sampled. The "pseudopelagic" or near- shore midwater environment has not been sampled in most parts of the world. In most cases sampling was seldom extensive enough to appreciate any micro-distri- butional features or seasonal variation (see Pearcy, 1964).
In addition to differences in collecting gear, there were significant differences in fishing philosophy. Some ciTiises were faunal surveys with many oblique tows to numerous depths. Other cruises were in- terested in sampling only over a certain depth range (e. g., upper 500 m), while still others sampled particular environ- ments or collected in sound-scattering lay- ers. The majority of cruises were diurnally biased, collecting primarily in the upper 200 m at night and much deeper during the day. The upper 500 m was much more extensively sampled than deeper waters, especially at depths below 1000 m.
A wide variety of fishing gear was em- ployed. The gear most frequently used was the lO' Isaacs-Kidd midwater trawl. Many other types of trawls, ring nets, plankton tows, and even dip nets, provided material. Depth determination and data recording varied widely. For example, it was often impossible to tell if a certain sample was a horizontal or oblique tow, or whether the depth recorded was calcu- lated by triangulation or determined elec- tronically with automatic depth recorders.
There is an abimdance of literature on the problems encountered in sampling mid- water organisms from behavioral responses to gear characteristics and performance. For a comprehensive discussion of the problem see Suzuki (1961), Aron (1962), and Harrisson (1967).
Hatchetfishcs are easily caught by slow moving towed nets. There is some corre- lation between size of tow and size of individuals taken. Plankton tows take primarily very small individuals, while 10' Isaac-Kidd trawls take larger specimens. In general the 10' IKMT appears to imder- sample the large individuals, although it does on occasion catch the very largest individuals of a species. Comparisons with catches by the huge Engalls trawl in the northeast Atlantic show that there are more of the larger individuals present than IKMT samples indicate. In the case of ArgyropeJcciLs glgas, the largest specimens ever recorded were taken in numbers by this trawl. With the exception of Argij- ropelcciis glgas, hatchetfishcs are small sized and are adequately sampled, except for the largest sizes, by the standard IKMT. Indications are that more work with large midwater trawls, especially those that operate at depths greater than 500 m, will add a new dimension to the "lilliputian" midwater fauna (see Harrisson, 1967: 104).
Measurements and counts. The methods of measurement usually used were those described by Hubbs and Lagler (1947: 13), although the peculiar morphology of the Stcrnoptychidae necessitated several ad- justments. In addition, measurements were adjusted so that in some cases reference points are somewhat different between the genera. Measurements of standard length (SL) and body depth (RD) were made with needle point dividers to the nearest uhole millimeter. Other measurements were taken with vernier calipers, and were determined to the nearest tenth of one millimeter.
Characters chosen for measurement were those which appeared to have systematic
Marine Hatchetfishes • Baird
importance, or could be directly or in- directh' tied to ecological considerations. The following measurements were taken; Standard length — measured from the end of the snout to the farthest extension of the well-marked caudal peduncle (in Ster- nopfyx the peduncle asymmetrical, the low'er lobe extends farthest posteriorly). Bodv depth — in Arg,ijropeleciis and Polyipnus measured from the origin of the dorsal blade to the most ventral extension of body margin, excluding ventral keel scales; in Sfcrnoptyx measured from the end of the dorsal fin and essentially a tinink measure- ment. Dorsal blade — height measured from dorsal body margin to greatest extension of major element in the blade along blade axis (in Stcrnopfyx there is only one ele- ment). Jaw length — measured from the point of the retroarticular to the anterior- most extension of the lower jaw. Jaw width — measured in the lateral plane between the left and right lower jaw articulations. Caudal peduncle — a depth measurement across the narrowest dorsal- ventral axis of the caudal peduncle. Ab- dominal length — used only in Sternoptyx, measured from the dorsalmost point of the supra-anal photophore to the posterior- most extension of the caudal peduncle. Supra-abdominal photophore — a Ster- noptyx character measured from the dorsal- most point of the supra-anal photophore to the dorsal body margin normal to the midabdominal axis. Dorsal fin length — in Sternoptyx measured from the origin of \he anterionnost fin ray to the origin of the posteriomiost fin ray. Orbital diameter — in Polyipnus measured along the anterior- posterior axis. Post-temporal spine length — in Polyipnus measured from the \entral origin of the spine to its tip. Head length — in Polyipnus measured from the end of the snout to the posterior opercular mar- gin. Photophore measurements — measured from the farthest extension of the dark pigmented pliotophorc margins.
The following counts were made. Gill raker number: the number of gill rakers
on the first branchial arch of the left side; only clearly defined rakers were counted. Caudal, median, and pectoral ray counts were as per Hubbs and Lagler (1947). Vertebral counts were made from fossils, X-ray photographs, or cleared and stained specimens. Vertebral counts included all separate vertebrae, except the urostylar element(s); vertebral counts for fossil ma- terial included only those elements pos- terior to the major element in the dorsal blade.
Keys and key characters. Because of the damaged condition of many specimens in midwater collections, keys include several characters to aid in identification. Care must be taken when making measurements on, or using key characters with, damasfed specimens. Keys were constructed for adults and late juveniles only, and are roughly limited to individuals greater than 20-25 mm in standard length. Photophore complement, especially in the anal series of Polyipnus, is complete only in the adults. Most of the key characters are discussed in the descriptions; however, several of the more common ones are expanded as fol- lows. The post-temporal spine in certain species of Polyipnus bears small basal spines on its ventral-lateral surface; dorsal, postabdominal, and preopercular spines are often worn or broken, especially in larger individuals. Subcaudal spines ap- pear late in ontogen\^ and are always small. Spinose borders of the preopercle and ventral keel scales are obvious and well developed. Canine teeth may be missing or broken, but when present they are conspiciously longer than other teeth. Teeth present on the midline of the pos- terior vomerine shaft in certain species of Polyipnus are difficult to see in small speci- mens. Caudal ray pigment is often reduced by loss or abrasion of the caudal fin. Pig- ment characteristics used are dark mela- nistic areas which appear stable in most common preser\'ati\'es if the specimens are undamaged. Preopercle spine character- istics in the Argyropelecus lychnus complex
6
BuUctin MuscKin of Comparative Zoology, Vol. 142, No. 1
are sometimes variable, and occasionally borderline cases occur. While keys were constructed for indi\'idual identification, population and distribution data should always be checked.
Photophorcs (Figs. 17 and 18). Photo- pliore nomenclature was adopted from Schultz (1961). The photophore groups are as follows: preorbital (PO) — a single photophore located anterior to the eye (ventrally located in Sternoptyx); post- orbital (PTO) — a single photophore just posterior to the eye; preopercular (PRO) — a single large photophore located at the ventral margin of the opercular region; subopercular (SO) — a single small photo- phore at the posterior ventral margin of the opercular region; suprapectoral (SP) — a series of three photophores (two in Arii,y- ropelecus) in the region above the pectoral fin; branchiostegal (BR) — a cluster or group of photophores located in the bran- chial region; isthmus (I) — a group of five to six photophores along the anterior ven- tral body margin below the preopercular complex; abdominal (AB) — a large group of 10-12 photophores along the ventral abdominal body margin; preanal (PAN) — a group of three to five photophores lo- cated in the region just above and posterior to the pelvic fins; anal (AN) — a variably numbered group of photophores located along the ventral body margin in the region of the anal fin; subcaudal (SC) — a group of four photophores along the ven- tral body margin in the region of the caudal peduncle; these usually fonn a single close-packed cluster but may be separated in certain species of Arfiyropcle- ciis; supra-abdominal (SAB) — (absent in Sternoptyx) a series of three (Polyipmis) or six {Ariiyropelecus) photophores above the abdominal series along the lateral body margin; supra-anal (SAN) — (absent in Argyropelecus) a single photophore in Sternoptyx which is anterior to and raised above the anal group; a series of three photophorcs in Polyipntis anterior to and usually raised above the anal group; in
certain species the three supra-anal photo- phores are anterior to but are essentially continuous with the anal series; lateral (L) — a single photophore in the midlateral region of the trunk found only in Polyipnus.
Photophore number and position are re- markably constant in the Sternoptychidae. However, rare individuals do have photo- phores in somewhat abnormal positions or occur with an abnormal number in any group. The number is constant in most photophore groups throughout a genus, although the resultant pattern may be somewhat different owing to differences in body form or photophore location. No sexual dimorphism in photophore number or pattern was observed.
CAearinLi and staining. For the osteo- logical study, a series of specimens of each species examined was cleared and stained using a slightly modified trypsin digestion technique described by Taylor (1967). This method gave excellent results even on specimens preserved for long periods of time. In addition, the method is consider- ably more rapid than other techniques. Distorted specimens often gave good re- sults since they were partially relaxed by the digestion process.
Analysis and presentation of data. Because of the magnitude of material ex- amined, computer techniques were em ployed extensively. Programs (primarily in the Fortran IV language for use with the IBM 7094 at the Harvard University Computation Center) were designed to plot and analyze the data. Four types of data cards were punched and then cross- indexed by cruise and station number. One card contained station location and deptli plus hydrographic and time data where available. The catch card incorporated the total catch, the size breakdown of the catch, and other data such as maximum size or size of gravid females. Mor]:)ho- metric and meristic data cards completed the raw data input.
Plorizontal distributions were computer
Marine Hatchetfishes • Baird
plotted and broken down into three arbi- trary, relative abundance categories which were indicated by separate symbols. These plots formed the basis for the distributional charts on each species. Plots of juveniles and gra\'id females did not differ signifi- cantly from overall plots, so they were not included in the data presented.
Depth data were subjected to two sepa- rate analyses. The first was a tabular breakdown of all depth data taken pri- marily by IKMT and in which depth was determined by pressure depth recorders in most cases. Depth figures represent only maximum net depths and in many in- stances probably represent oblique tows, although where this was definitely in- dicated oblique tows were excluded. The results are listed in Appendix B. This method was particularly helpful in ap- preciating sampling bias. For the second analysis, only known horizontal tows were used and a plot of the rate of catch in fish-per-hour against depth was made. Only rates greater than one fish/hour (one-half fish/hour in certain species) were plotted. A much finer definition of the depth range of each species was thus obtained, although the sampling bias can- not be fully ascertained.
Where hydrographic data were available, temperature-salinity plots were made for each species and compared with known water mass T-S envelopes. These plots formed part of the data for Table 24.
Morphometric and meristic data were analyzed using standard statistical meth- ods. All proportional data were computer plotted against standard length, and re- gression statistics were computed by the least square method. Only adults or late juveniles were used, and relationships were linear in all cases. Variabilitv was quite low in most instances, and as long as stratified samples were taken (covering most of the length range of a species), excellent repeatable regression lines were obtained. Confidence limits decreased with sample size to about 20 individuals, beyond
which little reduction could be obtained. Stratified samples as small as eight indi- viduals were adequate to establish good regression lines, which were consistent with larger samples in most cases. In many areas, sample sizes were inadequate and the population parameters presented must be verified further with more sam- pling. Slope differences were tested statisti- cally and are presented in the various tables. Positional differences between populations could often be detected al- though the slopes were not statistically different. These, when noted, were plotted (Fig. 23). Meristic data were plotted and a difference of two standard errors on either side of the mean formed the basis for statistical comparison. Dorsal blade height in Argtjropelcciis hemigijmmis was plotted in the same manner as meristic data. The slope of blade height to standard length was very low (.008-.02), so that comparisons between individuals over a small length range (22-28 mm) were con- sidered equivalent.
Oeeanographic data were obtained from various standard sources (Fuglister, 1960; Sverdrup et al., 1960; Muromtsev, 1963; Schroeder, 1963) and from cruise reports.
OSTEOLOGICAL CHARACTER COMPLEXES Cauda] skeleton (Fig. 1). There is a definite similarity among the caudal skele- tons examined. Features in common in- clude: somewhat flattened neural and haemal spines; three characteristic hypural or hypural-like elements in the ventral caudal lobe (definitions and abbreviations of bones follow Norden ( 1961 ) ; see Weitz- man ( 1967a ) concerning definition of a hypural element); often one or more post- terminal vertebrae; one or two free epurals (except Stemoptyx); and a caudal fin ray count of 10 + 9, with a varying number of dorsal and ventral procurrent rays.
There is considerable \'ariation in the degree of fusion of hypural elements. With the exception of the Argijvopelecus af finis and Polyipnus spinosus species complexes.
8 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
f\JS 1
HYP
HS
NS 2
A
B
NS 1
NS2
Figure 1. Caudal Skeleton: A. Maurolicus muelleri; B. Argyropelecus affinis; C. Polyipnus asteroides; D. Sternoptyx pseudobscura. Abbreviations: EP r; epural; HS i^ haemal spine; HYP =: hypurals; NS z^ neural spines; and UN =: uroneurol.
hypurals 1-2 and 3-4 are always fused. In some cases, there is complete (Sternoptyx) or almost complete {ValencienneUtis) fusion of hypural elements.
The following are the important evo-
lutionary features. The caudal skeleton of Maurolicus appears primitive and is similar to the caudal skeleton of Vificiiiuerria as illustrated by Weitzman (1967b). The three sternoptychid genera are character-
Marine Hatchetfishes • Baird 9
ized by a modification of the first neural spine into a short triangulate, vertical blade. The second neural spine often sup- ports the first. In marked contrast, the gonostomatid genera examined (also Vin- ciguerria, see Weitzman, 1967b) show little modification in this area, and the first neural spine is elongate and forms an integral part of the upper caudal lobe. Stemoptyx shows a high degree of special- ization with considerable reduction or fusion of elements. Fohjipnus shares with MaiiroVicus (and Vinciguenia) the lack of fusion in hypurals 5 and 6. In some respects Fohjipnus resembles the gonostomatid gen- era examined in size and shape of the uroneurals although, in general, it appears similar to Argijropeleciis.
Axial skeleton (Figs. 8-11). While there is a similarity in structure and appearance of the vertebral centra in all genera ex- amined, there are differences in neural and haemal spine pattern and structure. Pos- teriorly, the haemal and neural spine ar- rangement is symmetrical in all cases. In Maiirolictis and Valenciennellus both spines are relatively long, unflattened, and ta]:)ering. The sternoptychid genera show a definite broadened and flattened con- dition particularly evident at the distal end. FoJijipnus and Argijropeleciis are alike in this respect. Stemoptyx, with considerable elongation of the posterior neural and haemal spines, reflects an independent and highly modified condition. Vincigucrria (Ahlstrom and Counts, 1958) appears more similar to Fohjipnus than either of the gonostomatids examined.
Anteriorly, the symmetrical pattern of haemal and neural spines continues in Valenciennellus with no marked transi- tional region. However, in Maurolicus and the Sternoptychidae, there is an area of transitional vertebrae which is peculiar. There is a reduced, although fully formed, plural rib-bearing member followed by a number of characteristic haemal spines which may or may not be arched. This series of spines carries at least one pair
Figure 2. Post-temporal and supracleithrum: A. Argyropele- cus aculeatus; B. Maurolicus muelleri; C. Valenciennellus tripunctulatus; D. Polyipnus asteroides; E. Sfernopfyx pseud- obscura. Abbreviations: PT ^ post-temporal; SCL =: su- pracleithrum.
of greatly reduced or vestigial ribs. In Maurolicus, the first arched haemal spine is somewhat flattened distally. Polyipnus and Argyropelecus show a marked broad- ening of the distal end of these anterior haemal spines, with an increase in length proceeding posteriorly. Stemoptyx has a shortened flat first haemal spine; however, the posterior spines are elongate and not characteristically flattened.
The neural spines posteriorly are long, thin, and tapering in Maurolicus and Valenciennellus. As before, the stenio- ptychids show a broadened pattern vmlike the above gonostomatids. Fohjipnus and Argyropelecus are remarkably similar in this region.
10 Bulletin Museum of Comparative Zoohg^ij, Vol. 142, No. 1
The articulation of ribs is similar in all species, as well as the presence of reduced or vestigial pleural ribs, probably an in- dication that many more centra were rib- bearing in more primitive forms (see Weitzman, 1967b: 518). Maurolicus has a higher number of pleural and reduced pleural ribs than the stemoptychids. The latter are quite distinctly separated from Maurolicus and Valenciennellus by the presence of six or seven greatly broadened and lengthened pleural ribs which form a heavy protective cage around the now expanded visceral cavity. The number of abdominal vertebrae (the first caudal vertebra is defined as the anteriormost vertebra with a complete haemal arch) is relatively constant in the Sternoptycliidae at about eleven (one specimen of Polyipmis asteroides had ten; Kotthaus (1967) re- ports twelve for P. meteori). Sample sizes were small, however.
Dorsal blade (Figs. 8-11). Weitzman (1967b) reported that the anteriormost pterygiophore of the dorsal fin consists of at least two fused pterygiophores in Vinciguerria. This same characteristic is found in Maurolicus which, in addition, has a number of pterygiophores that do not bear fin rays anterior to the fused one. The stemoptychids have this same basic feature, but have further modified it into essentially a "spinous dorsal." In Polyipnus, the fused pterygiophore is extended above the dorsal body surface and is spinose at the distal end. The anteriormost pterygio- phores are enlarged, and closer together and more extensively allied to the support- ing neural spines than they are in Mauro- licus. These anterior pterygiophores be- come even more enlarged and closely allied, extend further above the dorsal surface, and with the fused pterygiophore form an extensive, sharp, dorsal blade in Arg,yropelecus. Sternoptyx retains the Maurolicus configuration anteriorly, but the fused pterygiophore becomes consider- ably extended and modified into a large dorsal spine.
Pelvic fi,irdle (Figs. 8-11). In Maurolicus (also Vinciiiuerria, Weitzman, 1967b), the basipterygia are located even with or be- low the ventral margin of the pleural ribs. The paired basipterygia lie almost hori- zontally above the ventral body surface and are not closely joined to any rib ele- ment. With the broadening and deepening of the anterior thoracic region in the stemoptychids, the pelvic girdle has be- come a major structural element for the midregion of the trunk. Polyipmis exhibits a more intermediate condition than Ster- noptyx and Argyropelecus. In the fonner, the basipterygia are oriented at approxi- mately 45° to the ventral body surface and are located between the posterior- most large pleural ribs. There is now a relatively long ventral extension which ends in a spine protruding below the ventral body surface. The pattern becomes more pronounced in Argyropelecus. In this instance, the basipterygia are closely allied to each other and to the posterior- most large pleural rib. In some cases, the basipterygia are fused (A. hemif.iymnus) and the last pleural rib may become further enlarged for support (A. aculeatus). The ventral spiny process has also become more pronounced. Sternoptyx exhibits essen- tially the same evolutionary trend as Argyropelecus, with the fused basipterygia extending dorsally for a considerable length along the pleural ribs.
Pectoral girdle (Figs. 8-11). The pec- toral elements, their general location and shape, are similar in all genera examined and include a well-developed mesocoracoid (see Weitzman, 1967b: 519). Polyipnus and Argyropelecus have an extended pos- terior flange of the cleithrum which pro- tects and strengthens the pectoral area. The ventral margin of this flange has a characteristic spinose edge. The flange is noticeably reduced in Sternoptyx.
A forked post-temporal and well-de- veloped supracleithrum are present in all genera (Fig. 2). Polyipnus and Argyro- pelecus are unique in that these two bones
Marine Hatchetfishes • Baird 11
lOP
Figure 3. Opercular Series: A. Maurolicus muelleri; B. Polyipnus asteroides; C. Sfernopfyx pseudobscura; D. Argy- rope/ecus hemigymnus. Abbreviations: lOP = interopercle; OP ^ opercle; POP =^ preopercle; SOP == subopercle.
12 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
AL
PAL
MET-
B
■PAL
MES
PAL
ECT
Figure 4. Suspensorium: A. Polyipnus osfero/'des; B. Maurolicus muelleri; C. Sternoptyx pseudobscuro; D. Argyrope/ecus hemigymnus. Abbreviations: ECT ^ ectopterygoid; HY := hyomandibular; IH z= interhyal; MES ^ mesopterygoid; MET =: metapterygoid; PAL zz palatine; Q == quadrate; SY = sympiectic.
are fused. The post-temporal half of this process extends posteriorly above the dorsal body margin and bears spines. In certain species of Pohjipntis these spines may become quite elaborate. In Sternoptyx, which has no such fusion, the post- temporal is forked and enlarged, and the
\\'holc structure reflects a different evo- lutionar)- development.
Opercular series (Fig. 3). There is a classic opercular series present in the genera examined, with an interopercle be- low the ventral margin of the prcopercle. There appears to be an evolutionary trend
Marine Hatchetfishes • Baird 13
from Maurolicus through Polyipmts to Argyropelectis. In Polyipmis, the inter- opercle is similar in shape to Maurolicus, but somewhat less broad. The preopercle has developed a ventral spine. A reduction in the anterior process of the interopercle, which now covers only the posterior ven- tral margin of the preopercle, may be observed in ArgyropeJecus. The pre- opercle, while similar in form to Polyipmis, has a lateral spine in addition to the ven- tral. Sternoptyx is somewhat indepen- dently modified with elongation and re- shaping of the opercle and preopercle. The interopercle is similar to Argyropclecus, and the preopercle has a single ventral spine.
Upper jaw. The upper jaw, considering its close relation to feeding ecology, is somewhat similar in Maurolicus, Polyipnus, and Argyropclecus. There are two char- acteristically shaped supramaxillae, a well- developed toothed maxilla and premaxilla. The premaxillae have short ascending proc- esses (as does Vinciguerria). The maxilla, included in the gape to a small degree, is markedly broadened posteriorly in Polyip- mis, and the whole jaw apparatus reflects a peculiar method of feeding. Sternoptyx is quite different. In this instance the maxilla is heavily toothed and the major upper jaw bone in the gape. Tlie pre- maxilla is small, although toothed, and has no ascending process. The second supra- maxilla has been lost.
Suspensorium (Fig. 4). There appears to be a general evolutionary trend in the Sternoptychidae in which the suspensorium migrates from behind and slightly below the posterior orbital region, ventrally and anteriorly to a point directly below the anterior half of the orbit. This trend can be seen by examining the ratio of quadrate length to hyomandibular length: Mauroli- cus, 1:1.25; Polyipmis, 1:1.5; Argyro- pclecus, 1:2.5; Sternoptyx, 1:7.4. The metapterygoid bone is proportionately smaller in Maurolicus and Polyipnus, and the mesopterygoid is greatly enlarged in
D
Figure 5. Urohyal: A. Sternoptyx pseudohscura; B. hAaur- olicus muelleri; C. Argyropelecus sladeni; D. Polyipnus asteroides.
the latter. This again reflects the peculiar jaw morphology in this genus.
Hyoid (Fig. 5). The most notable hyoid featiu-e is the gradual reduction of the platelike posterior extension of the urohyal in the sternoptychids. Polyipnus illustrates an intermediate condition, while Sternoptyx and Argyropelecus show complete reduc- tion to a Y-shaped bone.
Chondrocranium. The curvature of the parasphenoid exhibits a continuous gra- dation from a nearly horizontal position in Maurolicus to the extreme right-angled bone in Sternoptyx. The presence and de- gree of ossification of the basisphenoid is variable. It is well developed and has two centers of ossification in Valenciennellus. Only the dorsal ossification remains in Maurolicus, while the bone is absent in Argyropelecus. Polyipnus and Sternoptyx have well-developed basisphenoid bones.
The neurocranium (Fig. 6) is generally conservative when viewed as a whole. The shape, relative size, and location of the bones are similar in all genera examined. Tlie neurocranium resembles Vinciguerria (Weitzman, 1967b), especially in the gen- eral shape and location of the sphenotics, pterotics, and epiotics. Important features
14 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
soc
Figure 6. Neurocranium (lateral view); A. Polyipnus asfero/des; B. Argyropelecus hemigymnus (frontals removed); C. Maurolicus pennanfi; D. Sfernopfyx pseudobscura. Abbreviations: BOC ^:z basioccipifal; BS = baslsphenold; EOC = exoccipital; EPO := epiotic; F rz frontal; HY =: hyomandibular; PA =^ parietal; PRO ^ prootic; PS ::= pterosphenoid; PTO =: pterotic; SOC = supraoccipital; SPO =: sphenotic.
are: the epiotics meet below the supra- occipital in sternoptychids, while there is no tendency in this direction in Maurolicus and Valenciennellus (Fig. 7); the presence of well-developed parietals with dorso- lateral ridges in sternoptychids, but not in other genera examined; the presence of
a well-developed alisphenoid (pterosphe- noid) bone in all genera; and tlie pro- gressive tendency for the neurocranial axis, as measured along the frontal, to assume a more vertical configuration from Polyipmis to Sternoptyx. There is considerable de- velopment of the otic region in Polyipnus
Marine Hatchetfishes • Baird 15
SBQ
Figure 7. Epiotic-Supraoccipital Relationship (posterior view): A. Maurolicus muelleri; B. Polyipnus asteroides; C. S/ernoptyx pseuc/ofascuro; D. Argyropelecus bemigymnus. Abbreviations: EPO =^ epiotic; SOC = supraoccipital.
which contains pecuHarly shaped and very large otoHths (Kotthaus, 1967).
Abdominal keel scales (Figs. 8-11). Polyipnus and Argijropelecus have de- veloped ossified plates (modified scales) which form a keel and serve to give struc- ture to the abdominal region and associ- ated photophore groups. Several plates appear posterior to the pelvic fins; most are anterior to them. The plate size, niun- ber, and distance between plates is less well developed in Polyipnus. Stemoptyx seems also to have a keel-like structure, but this is not ossified. The gonostomatids examined have little keel development and no ossification in this region.
Anal pterygiophores (Figs. 8-11). In stemoptychids, the anal pterygiophores show a characteristic gap. Several pterygio- phores are associated with and between the same haemal spine forming a circular gap. In this respect the P. spinosiis species complex is the least well developed. In the gonostomatids examined there is one pterygiophore for each haemal spine with no gap. The anteriormost anal pterygio- phore possesses flangelike processes pro- jecting laterally in Polyipnus, Stemoptyx, and Maurolicus. The former two have, in addition, pronounced ventral processes lacking in Maurolicus. Argyropelecus has
no processes, although the anal pterygio- phores are enlarged.
Photophores. The glandular nature and pattern of photophores seem to indicate some relationship among all genera studied. The trend appears to be from a condition of an essentially unbroken row of photo- phores on the ventral body surface (Maurolicus) to one in which this row is broken both horizontally and vertically (stemoptychids). As before, Polyipnus is intermediate in this respect.
OSTEOLOGICAL CONCLUSIONS
The osteological results lead to the fol- lowing conclusions. The present definitions of the family (e. g., Regan, 1923; Schultz, 1961; and Morrow, 1964) and included genera are inadequate, often seriously in error, and require revision. The Stemo- ptychidae appear to be derived from some antecedent of the primitive genus Mauro- licus. The genera Stemoptyx, Argyro- pelecus, and Polyipnus form a separate taxon. Each of these genera has probably been distinct for a long period, as each shows a great deal of divergence and in- dependent evolution.
From the evidence above there is little doubt that the tsvo maurolicid genera and the Sternoptychidae are closely related. The traditional differences such as absence of a mesocoracoid and alisphenoid (ptero- sphenoid), curved parasphenoid, and even the particulars of the dorsal blade have been found to be, wholly or in part, similarities rather than differences. Basic differences do exist, however, and in gen- eral follow from Hennann's original char- acterization of the Sternoptychidae as fish having a deep, highly compressed body form. It is this striking evolutionary pat- tern that gives rise to many of the follow- ing character complexes which separate the present Sternoptychidae from those gonostomatids examined.
1. Modification of the first neural spine, appearing as a short, triangulate, vertical blade with further modification of the
16 Bulletin Mtiseum of Comparative Zoology, Vol. 142, No. 1
o
oo
CO
X
U
<
I U
Marine Hatchetfishes • Baird
17
6
\
o
CM
Jf%,
CN CN
CO
X
o
^0
U
<
X
o
a. _>,
o
18 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
\
V
CO
o
U
<
X
u >
-Q O
a o
c
Marine Hatchetfishes • Baird
19
O
S3
CO
D
z
z z <
C/5
Z to
ID
a o
^ii£:
20
Bulletin Museum of Compavatwe Zoology, Vol. 142, No. 1
second neural spine to serve as a support- ing element. (In the P. spinosus species complex, the second neural spine resembles the first.)
2. Characteristic broadening and flat- tening of the haemal and neural spines in the posterior caudal region.
3. The presence of six or seven large, heavy, pleural ribs with relatively few reduced or vestigial ribs. This includes a low number (10-12) of abdominal verte- brae.
4. Development of the dorsal pterygio- phore system into a "blade" or spine.
5. A vertically oriented pelvic girdle, the basipterygia bearing spines, sometimes fused, and closely allied to the heavy pleural ribs.
6. A preopercle with a well-developed ventral spine.
7. A heavy, forked, post-temporal which is fused to the supracleithrum in Argijro- pelecus and Polyipnus, formmg a spiny extension dorsally.
8. A progressive migration forward of the suspensorium.
9. Reduction of the bony extension of the urohyal.
10. Epiotics meeting below the supra- occipital and the presence of well-de- veloped, ridged parietals.
11. Presence of a well-developed ab- dominal keel-like structure which is ossi- fied in Argijropelecus and Polyipnus.
12. Presence of a circular gap in the anal pterygiophore series, these pteiyg- iophores being enlarged.
13. Presence of ventral processes on the anteriormost anal pterygiophore in Ster- uoptyx and Polyipnus.
14. Marked similarity of photophore pattern and number.
Some of these character complexes are not radically different from the gonosto- matids examined, and there is a degree of convergence and parallel evolution which is difficult to appraise. Taken as a whole, however, they strongly suggest that the stemoptychid genera have reached a com-
mon evolutionary grade, typified by their peculiar body form, and by which they differ from the more generalized and primitive maurolicid gonostomatids.
While acknowledging that the Sterno- ptychidae are a specialized offshoot of maurolicid or premaurolicid stock, for the following reasons I do not feel justified in combining the Gonostomatidae and Ster- noptychidae as some have suggested.
The present family Gonostomatidae is an unwieldy one which involves many diverse types and requires extensive revision (Weitzman, personal conversation). The problem of gaps, their size and importance, cannot be adequately answered without further stud>- within the Gonostomatidae. Osteologically, the Sternoptychidae have reached an evolutionary grade peculiar to themselves and one quite distinct in several major ways from the gonostomatids ex- amined. Using for a guideline the family concept as it is generally employed by Mayr, Linsley, and Usinger (1953), it ap- pears that the Sternoptychidae do have an ecological, or at least adaptive, distinctness.
The adaptive distinctness concerns the peculiar body shape and its possible func- tional significance. Tliere are at least two major adaptive features involved. The first deals with the ideas and evidence presented by Denton and Nicol ( 1965 ) and Nicol (1967) on the relationship between silvery color and body shape in teleost fishes. The midwater environment is one in which the distribution of daylight is in- dependent of the altitude of the sun and cloudiness of the sky, and light distribution is essentially symmetrical about a vertical to the surface. Furthermore, the Sterno- ptychidae have brilliant, silvery sides. All fish species with these features so far examined (Denton and Nicol, 1965; Nicol, 1967) have layers of reflecting platelets \\'hich are oriented to make the fish as invisible as possible. It may be assumed that the same is true with hatchetfish. Tliere is a change in reflectivity with body rotation in the several Stemoptychid
Marine Hatchetfishes • Baird 21
species examined. A silveiy fish which is flattened laterally, having very little in- clined ventral surface will approach the ideal in camouflaging (see Denton and Nicol, 1965: 717). The Sternoptychidae could thus serve as a living model for such a body fonn.
The second adaptive feature concerning body shape is the development of heavy structural ossifications and spines, espe- cially the dorsal "blade." Spines have developed in fish, presumably, for pro- tection. The sternopt)'chids have several extensive spine complexes: post-abdominal, post-temporal, preopercle, and dorsal. The spines are rigidly braced and the whole body strongly ossified, resulting in a compact rigid body shape. A spinous dor- sal has developed somewhat analogous to that of the higher Perciform fishes. This, coupled with the expanded abdominal region, results in a high length-to-depth ratio ( Table 1 ) .
In an environment populated by a host of predators, many with special adap- tations for ingesting large prey items, an increase in the length-to-depth ratio of a prey should be advantageous. A predator nomially capable of swallowing Valen- ciennelhis would require an approximate threefold increase in mouth diameter in order to accommodate A. hemigijmnus of the same length (Table 1). Ossification also takes place quite early. Juveniles or prejuveniles of about 10 mm have well- developed spines and are ossified.
Pliylogenetic relationships. The question of a monophyletic origin of the hatchetfish is unanswerable. The three genera show a great deal of divergence and independent evolution even within genera. Using the character complexes examined, some com- ments about generic relationships can be made, however.
The family appears primitive and prob- ably originated from a i^remaurolicid an- cestor, possibly something between the very early Vinciguerria and Maiiroliciis. Most of the characters examined could
have been derived from a form somewhat intermediate to the above genera.
The genus Sternoptijx seems to have di- verged quite early from the line or lines leading to Pohjipniis and Argyropelecus. It then continued to evolve independently, resulting in the present highly specialized form. In almost every case, Sternoptijx shows marked differences. The presence of a basisphenoid, the characteristically shaped, enlarged, first anal pterygiophore; the simple anterior, dorsal pterygiophores; possibly the meeting of the parietals, and the unfused post-temporal and supra- cleithrum all appear primitive. These char- acters are also shared with Polyipnus with the exception of the unfused post-temporals and meeting parietals. The presence of a small premaxilla and large maxilla as the major jaw bone in the gape are generally regarded as primitive. However, jaws and dentition have varied considerably in gonostomatids (Grey, 1964), and this may be a secondary phenomenon. The dis- appearance of the anterior pedicels of the premaxilla and loss of the second supra- maxilla can be explained in the same way, especially since the orbital region seems to have undergone considerable expansion. The resemblance of the urohyal to Argy- ropelecus may again be the result of parallel or convergent evolution involving feeding ecology which is similar in these genera.
Evolution from a premaurolicid ancestor can be traced somewhat more directly in the case of Polyipnus and Argyropelecus. Polyipnus and Argyropelecus share several character complexes: the characteristic blade-shaped, caudal haemal spines; the presence of the double pterygiophore as the major element in the "blade"; the presence of ossified, bony keel j)lates; the fusion of the post-temporal and supra- cleithrum; and separation of parietals by the supraoccipital (known to be variable in the Gonostomatidae). Polyipnus ap- pears intermediate between Maurolicus and Argyropelecus in several characters:
22 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Table 1. Body depth and standard length MEAsuIlE^'rENTS.
Family
Species
SL ( mm )
Maximum Body Depth* ( mm )
Gonostomatidae
Stemoptychidae
Danaphos aculatus Valenciennellus tripunctatiis
Argtjropelecus hemigymus Argyropelecus hemigymmis
38.0 26.0
38.0 26.0
9.2
5.4
24.0 16.9
* Includes dorsal blade.
the axial skeleton in general; anterior dorsal pterygiophore development; pelvic girdle modification; evolution of the cleithrum, first branchiostegal rays, and urohyal; the opercular series, especially the preopercle and interopercle; suspensorium develop- ment; parasphenoid curvature, and pro- gressive deepening and shortening of the body with reduction in a long unbroken series of ventral photophores. PoJyipmis has characters that are not shared with Argyropelecus in addition to those which are shared with Sternoptyx. These include: hypurals 5 and 6 unfused (3 and 4 also in the P. spinosus complex); small, relatively unmodified dentition; and, a urostylar element with several unfused post-terminal centra.
Polyipmis also has several highly special- ized characters: the peculiar jaw morphol- ogy; a greatly enlarged otic region with characteristically shaped, large otoliths (Kotthaus, 1967); and, the peculiar de- velopment of the cleithrum (pectoral shield ) .
The divergence of Polyipnus and Argy- ropelecus has involved the continued evo- lution of many intermediate characters mentioned above. Other major develop- ments in Argyropelecus are: fusion of hypurals 5 and 6 and the post-tenninal centra; the presence of seven rather than six heavy pleural ribs; development of a lateral preopercular spine (one species of Polyipnus has this); development of a fanglike dentition; loss of the basisphenoid; loss of the flangelike process on the first anal pterygiophore; and, the development of telescopic eyes.
Because the number of character com- plexes examined was limited, the suggested phylogeny is only a tentative one. The family consists of three divergent, inde- pendently specialized genera. Polyipnus ajDpears the most primitive, Sternoptyx the most highly specialized and the most diffi- cult to place, while Argyropelecus falls somewhere in between.
THE FOSSIL RECORD
The earliest reported stemoptychid fossil is from the Eocene of the Dabakhan beds of Georgia, USSR ( Daniltshenko, 1962). Tlie fossil, Polyipnoidcs levis, is not well preserved and many important characters cannot be appraised. It does have long pleural ribs and a characteristic broadening of the body anteriorly. Tlie post-temporal, however, is unlike any modem stemopty- chid. The dorsal "blade" or pterygiophore development is absent and the jaws seem more gonostomatidlikc, although this is difficult to determine with certainty. The neural and haemal spines show little char- acteristic flattening, and the frontals do not exhibit the heavy development char- acteristic of the hatchetfish. Consequently, it appears that while this fossil could be a proto-stemoptychid fish, I cannot accu- rately place it with the present Stemopty- chidae or Gonostomatidae.
Polyipnus sobnioviensis was reported from the Jaslo shales of Poland (Jerz- maiiska, 1960; Jerzmanska and Jucha, 1963) and dates as late Eocene-early Oligocene. Enlarged pleural ribs with a general broad- ening of the body anteriorly are present in this species. Pterygiophore development
Marine Hatchetfishes • Baird 23
anterior to the dorsal fin rays is definite, above the dorsal fin origin, low; last pleural
and there is evidence of very slight, dorsal rib only slightly reinforced; anterior haemal
blade development. Some photophore spines not greatly flattened, postabdominal
groups conform roughly to modem Poly- spines short, symmetrical, not markedly
ipmis, although the fossil supra-abdominal curved; transitional vertebrae two in num-
group is more numerous. The cleithrum ber; anal pterygiophores relatively simple,
displays the marked ventral curve typical not markedly broadened at distal end; anal
of the Stenioptychidae and the maurolicid- pterygiophore gap contains t\vo haemal
gonostomatids. There is, however, little spines; number of anal pterygiophores be-
flattening of haemal and neural spines. The fore gap, seven, after gap, four to five;
pelvic girdle, while partially vertical, is hypurals 1 and 2 separate; number of
still below the rib line, and the body shape, vertebrae from posterior margin of dorsal
while somewhat broad, is more similar to blade to last neural spine, 29. the maurolicid gonostomatids. Tliere is no
spine on the preopercle, the orbit shows no Fossil B (Figure 13)
great expansion, and there are no signs of . , r,. -i t-. i * i
keel plates. Tliis fish, while it has some Description. Similar to Fossil A above;
sternoptychid characters, appears essen- f ^™ber of vertebrae from dorsal blade to
tially to be maurolicid-gonostomatid. Con- ^^'^ "^^^^1 spine 30; hypurals 1 and 2
sequently, its place in the genus PoZy^-pm/s separate; postabdommal spines simple,
is questionable, although it may be near symmetrical; posteriormost pleural ribs not
the basal stock which gave rise to modem greatly enlarged, hatchetfishes.
Pauca (1931) described Stemoptyx Fossil C (Figure 14) prisca from the lower Oligocene deposits Description. SL 60 mm, body depth 40 of Piatra Neamt. The presence of a well- mm; number of vertebrae from posterior developed dorsal "blade," heavy cleithmm, dorsal blade to last neural spine 26, pos- and pleural rib characteristics place it in sibly 27; both abdominal and trunk regions the genus Argyropelecus. If the dating is greatly broadened; hypurals 1 and 2 fused; correct, it represents the earliest known anterior haemal spines broad, flat, blade- fossil of this genus. lilce; distal end of anal pterygiophores
By Ohgocene, and certainly by Miocene broad, gap well developed, circular, and
times, several examples of the genus includes two neural spines.
Argyropelecus were evident in Tethys Fossils A and B are indistinguishable in
deposits of Europe (Arambourg, 1929; both key characters and meristics from the
Daniltshenko, 1960), and in various de- modem species A. affinis (Fig. 15) and
posits of California (David, 1943). All of can be assigned to this species complex,
these fossils clearly represent members of Fossil A seems broader than the modern
the above genus, and A. logearti (Aram- form, but the fossil appears distorted ven-
bourg, 1929) appears to be closely related trally and there are no other obvious dif-
to the modern A. hemigymniis. ferences.
In the present study, three remarkable Fossil C is a member of the A. lychnus
fossils from Miocene deposits in California complex (Figs. 11 and 16). Osteologically
were examined and compared with modern there is little difference between A. olfersi
relatives. and A. lychnus. However, the fossil has
a relatively low dorsal blade, measured
Fossil A (Figure 12) from the origin of the dorsal fin rays, a
Description. SL 50 mm, body depth 26 characteristic of A. lychnus (Fig. 16). The
mm; dorsal blade from its extension second transitional vertebra of Fossil C
24 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
a U
J2 O
u
o
CN
<
Marine Hatchetfishes • Baud 25
1
' ' V
■1
, \
V
s
-%
'"> r^
^ 1 V
I ■!>*-
¥
o U
CD
>
o
I
u
o
2 <
26 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
a U
o en _o
o
o u
o
u <
u
Marine Hatchetfishes • Baud 27
•o
o
•f
CM
CO
X
/ -
|
£3iW |
SHI ^ ^^^pJMP^rA. |
|
L |
|
|
i |
b^^v |
Art
o
V3
U
<
X
u
ID
a o
>..
28
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
^0
CM
1/1
3
Z
o
z <
c -c
a o
Marine Hatchetfishes • Baird 29
Key to the Genera of Sternoptychidae
la. Abdominal photophores 12; telescopic, dorsally oriented eyes; several dorsal pterygio-
phores fomi extensive blade anterior to dorsal rays genus Argijropelecus {-p. 31).
b. Abdominal photophores 10; eyes normal; dorsal blade consisting of only one or two
spines from a single or two fused pterygiophores 2
2a. Anal photophores 3; no supra-abdominal photophores; single large dorsal spine with anterior serrate extension; first anal pterygiophore greatly enlarged, supporting tri- angulate transparent membrane above anal fin rays genus Sternoptyx (p. 67).
1). Anal photophores 6 or greater; 3 supra-abdominal and a lateral photophore; dorsal blade reduced; no large transparent membrane above anal fin rays genus Poltjipmis (p. 79).
has a pair of vestigial ribs whereas the one time also. Many modern gonostomatid
modem A. hjchnus examined does not. genera were present during that time, and
One of the specimens of A. olfersi ex- Miocene faunas have distinctly modem
amined has small vestigial ribs on this resemblances (David, 1943; Grey, 1964;
vertebra; otherwise, all of the fossil char- Crane, 1966; Daniltshenko, 1960). The
acters and vertebral counts are identical to salmonoid-derived midwater fauna appears
these modern species. Fossil C is probably to have replaced earlier forms during the
A. hjchnus or at least its immediate prede- early Tertiary, and it remains the dominant
cessor. element today.
By mid-Miocene times evolution within
the genus Ar^ijropehciis was essentially SYSTEMATICS complete and species distributions show
modem characteristics. This genus with '^^'^''y STERNOPTYCHIDAE
its many specializations must have origi- ^/P® Genus: Siernopiyx Hermann 1781
nated by the late Eocene at the latest and Diagnosis. Neural spine of first pre-
possibly as far back as the Paleocene or terminal vertebra vertically oriented,
late Cretaceous. broadened, with triangulate paddle shape,
Durmg or prior to the early Cretaceous, no fin rays attached; second preterminal some members of the early salmonoid vertebra modified for support of first (ex- fishes began to adapt to a deep water eept F. splnosus complex); basiptervgia envn-onment. After the basic adaptations verticallv oriented, spine bearing, con- to this environment were acquired (at tained dorsally within, and closely joined latest mid-to-late Cretaceous) there was to the ventral margin of the posterior considerable stomiatoid radiation which pleural ribs; pelvic fin ravs xertically continued into the late Eocene to early oriented; six to se\'en i^leural ribs enlarged Ohgocene. This radiation led to many to form an expanded rib cage; epiotics diverse forms, of which the maurolicid- meet below supraoccipital; parietals well gonostomatids ^^'ere one. Within the latter, developed, bearing dorsolateral ridges; one an ancestor, possibly resembling P. sohnlo- or more dorsal pterN giophores enlarged to viensis, gave rise to a form or series of form blade or spinelike extension anterior forms with many features of the modem to dorsal rays; anal ptervgiophores form genus Pohjlpmis. From this basic stock the characteristic gap below anal photophore modern genera evolved, conceivably quite group; preopercle bearing well-developed rapidly. By the Miocene, evolution was ventral spine.
practically complete in the specialized Description. Bright silverv colored,
Argyiopclcciis and possibly the other small fishes; standard length usually less
genera as well. The stomiatoid-gonosto- than 90 mm; body deep" strongly com-
matid radiations of the early tertiary show pressed; bony scalelike plates forni keel
evidence of being fairly complete by that below ventral photophore groups (except
30 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
PTO
PRO
Figure 17. Photophore and spine characteristics: top — Po/yipnus; bottom — Argyropelecus. Abbreviations — photophores:
see p. 6; spines: DB = dorsal blade; PAS ^ postcbdominal spine; PTS z= post-temporal spine.
Sternoptyx); 10-12 abdominal, and four muscular stomach, five or more plyroic
subcaudal photophores always present (see caecae, and short straight intestine; eyes
Figs. 17 and 18 for photophore and spine large, well developed; gape vertical; adi-
characteristics ) ; nasal lamallae well de- pose fin usually present; scales thin or
veloped; digestive tract simple, with thick absent except along ventral surface; swim
Marine Hatchetfishes • Baiid 31
Figure 18. Photophore characteristics: genus Sternoptyx. Abbreviations: see p. 6.
bladder present ( see above for osteological description ) .
Genus ArgyropeJecus Cocco, 1829
Argijropelecus Cocco, 1829: 146 (type species: ArgyropeJecus hemigijmnus Cocco, 1829, bv monotypy ) .
Fleurothysls Lowe, 1843: 64 (type species: Sterno- ptyx olfersi Ciivier, 1843, by original designa- tion ) .
Sternoptychides Ogill^y, 1888: 1313 (type spe- cies: Sternoptychides amahUis Ogilby, 1888, by monotypy ) .
Diapiosis. Twelve abdominal, six supra- abdominal and two suprapectoral photo- phores; eyes telescopic, dorsally oriented; frontal ridges compressed dorsally above eyes; basisphenoid absent; several teeth directed anteriorly on posterior maxillary margin; dorsal "blade" consisting of several broadened pterygiophores anterior to dor- sal rays; seven enlarged pleural ribs.
Description. Photophorcs: PO 1; PTO 1; BR 6; I 6; AB 12; PRO 1; so 1; SP 2; SAB
6; PAN 4; AN 6; SC 4 (for anatomical details see Brauer, 1908; Bassot, 1966).
Spines: Post-temporals extended pos- teriorly to form a small spine; preopercle bears one ventrally and one posteriorly directed spine; retroarticular bears ven- trally directed spine; basipterygia extended ventrally bearing one or two postabdominal spines; cleithrum extends ventrally form- ing preabdominal spine; spiny scales present in adults of some species below subcaudal and preanal photophorcs.
Eyes: Large, well developed, telescopic, lens dorsally oriented, fitting into dorsal grooves in the frontal bone.
Gill Rakers: Total 15-24; rakers well developed with rough toothlike surface; epi- and ceratobranchials bear well-de- veloped spines on internal surface.
Jaws and Dentition: Jaws somewhat vertically oriented; premaxilla well de- veloped, toothed, and majoi- upper jaw bone in gape; maxilla also somewhat in-
32 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
eluded in gape, toothed, the posteriormost teeth eurved markedly forward; lower jaw sturdy, heavily toothed, oecasionally with large eanines; dentition eonsisting of multi- rowed single cusped, curved caninelike teeth; palatine teeth present, often well developed; epibranchial of third and fourth arch extends ventrally and laterally to form toothed plates.
Meristies: Vertebrae 35-40; C. 9+10; D. 8-10; A. 6-8 + 5-6.
Color: Bright silvery in life, quickly lost in formalin preservative; dark pigmenta- tion often striated posteriorly; stable for long periods in preservation.
Internal Anatomy: Relatively thin-walled swim bladder (see Marshall, 1960) and gas gland well developed; digestive tract simple, consisting of heavily pigmented, double compartmented stomach; the an- terior internal lining very thick walled and covered with rasping tubercles; posterior lining thin and distensible, five to seven thick-fingered pyloric caecae, large liver, and a short straight intestine; caelomie cavity lined with heavily pigmented mem- brane; gonads when mature fill the dorsal and lateral posterior half of the body cavity; nephritic tissue moderately well developed.
Species complexes. There has been sub- stantial radiation within the genus and even to some extent within the species complexes. The A. af finis complex appears to be the most primiti\'e and other forms can be derived from it. Primitive char- acters of this complex include: three hypural elements in lower caudal lobe; posterior ventral photophores in an almost unbroken series; glandular photophore ar-
rangement simple, the posterior photo- phores not joined in glandular clusters; little reinforcement of posteriormost pleural rib; body not markedly deepened ante- riorly; basipterygia lacking support ami for keel plates; generally unspecialized axial skeleton, including lack of marked broad- ening of anterior haemal spines; vertebral number 38-40.
The more advanced members of the genus are characterized by two hypural elements in the lower caudal lobe; pos- terior photophores joined into distinct glandular clusters; a general deepening of the anterior body region with subsequent reduction in vertebral number; increased complexity of structural ossification espe- cially in the axial skeleton, including a marked reinforcement of the last large pleural rib, and a keel supporting ex- tension on the left basipterygia.
A. he mi [1,1/771)1 lis appears slightly more primitive than the A. h/climis complex and is highly specialized. Important characters include: primitive transitional vertebrae (like A. af finis):, dwarfism (maximum length 38-40 mm); fused basipterygia forming single postabdominal spine; 38 vertebrae; epiotics with dorsal extensions (Fig. 7); peculiar dorsal blade shape, often \\'ith supplementary spiny spurs on the major element.
The most specious group, the A. hjchmis complex, shows a high degree of structural ossification in the axial skeleton, including the dorsal and anal pterygiophore systems; some species have developed long, fang- like canines in the lower jaw; there is a marked deepening of the anterior body region with reduction in vertebral number.
Key to the Species of Argijwpelectis
la. Snpra-abdominal, pieanal, anal, and subcaudal photophores in a nearly continuous straight line; subcaudal photophores separate, an appreciable gap between each photo- phore; .3 separate hypural elements in lower caudal lobe A. affinis complex 3
b. Supra-al)dominal, preanal, anal, and subcaudal photophores not in a continuous straight line; subcaudal photophore group in a single glandular cluster, no appreciable gap be- tween each photophore; two separate hypural elements in lower caudal lobe 2
A single posteriorly directed, serrate postabdominal spine; gap between anal and sub- caudal photophore groups greater than 2.2 times gap between preanal and anal groups; dorsal rays 8 A. heiiiigyiuiius (id. 42).
'Sa
Marine Hatchetfishes • Baird
33
b. Two separate postabdominal spines; anal subcaudal gap less than 2.0 times anal-pre-
anal gap; dorsal rays 9 A. lijchntis complex 4
3a. Dorsal blade low, its height less than one-third its length; body margin not markedly raised posterior to dorsal blade; ventral keel scales do not extend far below abdominal photophores; no laterally directed sphenotic spine near dorsal, posterior edge of orbit A. affinis (p. 34 ) .
b. Dorsal blade high, its height greater than one-third its length; body margin markedly raised posterior to dorsal blade; ventral keel scales extend well below alsdominal photo- phores forming flaplike process; prominent laterally directed sphenotic spine near dorsal, posterior edge of orbit A. gigas (p. 38).
a.
b.
4a. Posterior postabdominal spine directed posterioventrally and markedly larger tlian the anterior; anterior margin of posteriormost abdominal keel scale slants markedly forward; standard length less than 3.4 times body depth; pair of enlarged canine teeth present in
lower jaw; subcaudal spines present A. aculeatus (p. 48).
h. Both postabdominal spines of about equal length and size; anterior margin of posterior- most abdominal keel scale almost vertical; SL greater tlian 3.5 times body depth; pair of enlarged canines may or may not be present in lower jaw; subcaudal spines may or may not be present 5
5a. Pair of enlarged canine teeth in lower jaw; sharply pointed anteriormost postabdominal spine curved markedly and evenly forward; upper preopercular spine short, not extend- ing much beyond posterior border of preopercle; outermost tips of dorsal and ventral- most caudal rays streaked with dark pigment (fish greater than 35 mm) - — -
A. olfersi (p. 52).
b. No pair of enlarged canines in lower jaw; anteriormost postabdominal spine squared or blunt (except very small individuals), not curving evenly foi^ward; upper preopercular spine extends well beyond posterior border of preopercle; no pigment on outermost caudal rays 6
34 BiiUetin Museum of Comparative Zoology, Vol. 142, No. 1
fia. Dorsal blade low, blade height less than 2.5 mm for SL 25-50 mm, less than 3.1 mm for SL 50-70 mm; body narrow (see regression, body depth, Table 11); no spines on scales below subcaudal photophores; dark well-developed pigment spots fonn line along posterior midline ( especially in smaller sizes ) ; upper preopercle spine usually ciuved
dorsally, never ventrally — . A. .sladeni (p. 56).
b. Dorsal blade high, blade height greater tlian 2.7 mm for SL 25-50 mm, greater than 3.8 mm for SL greater than 50 mm; body robust (see regression, body depth, Table 14); spiny scales present below subcaudal photophores (fish greater than 30 mm SL); pigment spots minute along posterior midline; upper preopercle spine usually ciuved ventrally, never dorsally A. hjchnus (p. 63).
a.
Argyropelecus affinis Garman Figure 19
Argyropelecus affinis Garman, 1899: 237 (holo- type USNM 44593; tropical North Adantic; not seen); Brauer, 1901: 120; 1906: 103 (fig. larvae); Regan, 1908: 218; Barnard, 1925: 153; Norman, 1930: 301 (fig.); Jespersen, 1934: 15 (fig.); Fowler, 1936: 221; Beebe, 1937: 201; Parr, 1937: 49; Norman, 1939: 19; Nybehn, 1948: 23; Misra, 1952: 367; Smith, 1953: 102; Haig, 1955: 321; Fowler, 1956: 67; Schultz, 1961: 597 (fig.): Bahamonde, 1963: 83; Blache, 1964: 71 (fig.); Schultz, 1964: 241 (fig.); Backus et al., 1965: 142; Bussing, 1965: 185; Bright and Paquegnat, 1969: 27.
Argyropelecus pacificus Schultz, 1961: 599 (fig.); 1964: 241; Berry and Perkins, 1965: 625; Lavenberg and Ebeling, 1967: 185.
Species distinction. Differs from A. gigas
(in addition to key characters) by its nar- rower body depth and trunk (see regres- sion, body depth. Tables 2 and 3); less distinct trunk striations; relatively longer teeth in lower jaw; less well-developed post-temporal spines; smooth dorsal body surface; and less well-developed neuro- cranial crests (frontals, sphcnotics, and parietals ) .
Description. D. 9; A. 12-13; P. (10) 11; total gill rakers 18-22; vertebrae 38-39 (40).
Medium size species rarely exceeding 70 mm SL; body more evenly tapered than others in genus; body depth at end of dorsal greater than 3.5 times into SL; caudal peduncle long and narrow, its depth
Marine Hatchetfishes • Baud 35
/-':'' -^''iV" ■•^'■. /
:■::::■:■[: y^-''4>U
i ^ '0.1 J
Figure 19. Argyropelecus altinis; R/V CHAIN, Cruise 60; Station 1257; SL 51 mm.
less than length of siibcaudal photophore opercle spine directed latero-anteriorly;
group; dorsal spine low, its height less ventral preopercle spine long, curved an-
than one-third its length; post-temporal tcriorly; jaws large; teeth short, recurved,
spine short; postabdominal spines of equal better developed in lower jaw; gill rakers
size, with no marked curving; dorsal pre- long, closely set; in preservative, trunk
Table 2. Regression statistics for various populations of A. affinis.
Character
Indian Ocean (5°-12°N, 160°-168°E) Body depth Jaw length
Gtilf of Guinea
Body depth Jaw lengtli Jaw width
NW Atlantic (30°-33°N, 73°-78°W) Body depth Jaw length Jaw width
S£ Pacific (Chile) Body depth Jaw length
NE Pacific (California) Body depth Jaw lengtli
Regression
-1.83 -0.78
-2.44 -0.27 -0.23
2.59 -0.78 -1.35
0.55 0.64
-1.02 0.27
|
0.49 |
± |
.146 |
|
0.25 |
± |
.065 |
|
N |
— |
11 |
|
0.46 |
± |
.121 |
|
0.22 |
±: |
.060 |
|
0.12 |
±2 |
.059 |
|
N |
= |
10 |
|
0.37 |
±2 |
.176 |
|
0.24 |
^ |
.107 |
|
0.14 |
^ |
.087 |
|
N |
|
7 |
|
0.42 |
+ |
.086 |
|
0.21 |
^ |
.047 |
|
N |
— |
13 |
|
0.49 |
^ |
.062 |
|
0.22 |
± |
.029 |
|
N |
= |
19 |
36
Bulletin Mitseuni of Comparative Zoology, Vol. 142, No. 1
o
-o > -a
(U
E
o
□
o
c
o
o
X
Marine Hatchetfishes
Baird
37
D
E P
H
A •
B ▲
C O
D •
FISH HR
Figure 21. Diurnal vertical distribution of A. alfinis determined by rate of capture with depth during the day (D) and night (N). A = Pacific (California); B = Pacific (Chile]; C = Gulf of Guinea; D = Gulf of Mexico and Caribbean.
region exhibits cross pigment striations with well-defined, midlateral line.
Distribution. Horizontal distribution (Fig. 20): Taken abundantly in the Gulf of Guinea, off California, Chile, and in the northern Indian Ocean; moderate catches are recorded from the northern Gulf of Mexico and the coast of Venezuela in the Caribbean; smaller catches w^iich may rep- resent possible populations are recorded southeast of Hawaii, south of Java, and off the southeast coast of the United States;
scattered samples representing this species appear in the Bay of Bengal, Gulf of Aden, tropical Atlantic, and off the southeast coast of Brazil. (Additional records: At- lantic, occasional catches between Azores and Madeii-a; Pacific, moderate catches near coast of northern Peru. )
Vertical distribution (Fig. 21): Appears concentrated between 350 m and 600 m by day with the highest concentrations in the vicinity of 400 m; by night the dis- tribution is somewhat more shallow, major
38
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
—I—
18
I
iO
I
22
Figure 22. Geographic variation in gill raker number in A. affinis. A := Gulf of Guinea; B ^ Caribbean; C ;= Indian Ocean; D = Pacific (Chile); E = Pacific (Cali- fornia); F = NW Atlantic (NW Atlantic pocket). Numbers refer to sample size.
concentrations occnning from 170 m to 400 m. With the possible exception of the Gulf of Guinea, there are no indications of marked geographic variation in depth dis- tribution, although Appendix C indicates slightly shallower daytime depths off Cali- fornia than in the tropical Atlantic.
Geo<irap]uc variation. Five separate populations could be recognized and are identified and statistically defined in Figures 22 and 23 and Table 2. Regression sample sizes are small in the Atlantic. Figure 23 indicates positional variation in body depth even though there is no sig- nificant difference in slope.
Argyropelecus gigas Norman
Figure 24
Argyropelecus gigas Norman, 1930: 302 ( Iiolo- type BMNH 1.12.329; Gulf of Guinea; not seen); Jespersen, 1934: 15 (fig.); Fowler, 1936: 1208; Parr, 1937: 49; Maul, 1949a: 17 (fig.); 1949b: 13; Koefoed, 1961: 3; Schultz, 1961: 600 (fig.); 1964: 241 (fig.); Blache, 1964: 71 (fig.); Backus et al, 1965: 129; Bright and Paquegnat, 1969: 28.
3 0
B D
2D
B Q
SL
Figure 23. Geographic variation in the regression of body depth (BD) on standard length (SL) in A. affinis. A (California); B = Pacific (Chile); C = NW Atlantic; D = Gulf of Guinea; E = Indian Ocean.
=: Pacific
Marine Hatchetfishes • Baud 39
E E
O
u
<
X U
D
0)
a. o
CN
40 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
E
o
en
o
o
<.
'o
"5
o
X
Marine Hatchetfishes • Baud
41
|
N |
|||
|
Zoo |
|||
|
400 |
• |
||
|
600 |
< Oir 0 • |
• |
|
|
800 |
0 |
||
|
looo |
• |
||
|
ISoo |
«* |
aO
B •
D •
10
20
30
FISH
H R
Figure 26. Diurnal vertical distribution of A. gigas determined by rate of capture with depth during the day (D) and night (N). A z= Pacific (Chile); B = Gulf of Gulnear C = Gulf of Mexico; D = NE Atlantic.
Argtjropeleciis affinis: Jespersen, 1915: 6; Roule and Angel, 1933: 46- Bnen, 1935: 52; Nybelin, 1948: 23; Dollfus, 1955: 1.
Species distinction. See A. affinis (p. 34).
Description. D. 9 (10); A. 12-13; P. 10-11; total gill rakers 18-21; vertebrae 38-39.
Giant species often exceedino; 110 mm SL; trunk tiiangiilate, body depth at end of dorsal less than 3.3 times into SL; caudal
peduncle deep, its depth nearly equal to subcaudal photophore length; dorsal spine high, its height greater than one-third its length; post-temporal spine prominent; postabdominal spines symmetrical; pre- opercle spines as in A. affinis; jaws large; teeth small, recurved, a pair of larger canine teeth in upper jaw ( premaxilla ) ; gill rakers well developed; parietals, post- temporal, frontals and sphenotics with prominent spines or flanges; in preserva-
42 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
A
B C D
I 8
20
22
Figure 27. Geographic variation in gill raker number in A. g/gos. A ^ NE Atlantic; B =:: Gulf of Guinea; C ^ Indian Ocean; D rr Pacific (Cfiile). Numbers refer to sample size.
tive pigment in small spot.s along midline; very heavy mncoid secretion often present. Distribution. Horizontal distribntion (Fig. 25): Althongh occurring in all oceans except the North Pacific, this species ap- pears quite restricted locally. It is taken in abundance in the Gulf of Guinea, in the eastern North Atlantic off North Africa and southern Spain, in the northern Gulf
of Mexico, and off Chile in the South Pacific; small catches of this species are recorded along longitude 67°E from the equator to 40°S in the Indian Ocean, across the South Pacific between 35° S and 50° S from Chile to New Zealand in the South Pacific, and between New York and Cape Cod in the North Atlantic. Additional records: 36°35'S, 95°28'E.
Vertical distribution (Fig. 26): Con- centrated between 400 m and 600 m; no indication of diurnal vertical movement or marked geographic variation in depth dis- tribution.
Geographic variation. While having the most disjunct horizontal distribution in the family, this species shows the least vari- ability in the characters measured (Fig. 27, Table 3). Sample sizes in most cases were quite small, and increased sampling and the use of other characters may re- sult in better population definition.
Argyropelecus hemigymnus Cocco Figure 28
Ar<iyropclecii.s hcmigijtnmis Cocco, 1829: 146 (holotype unknown; Mediterranean, Messina); Alcock, 1896: 331; Jordan and Evermann, 1896:
Table 3. Regression statistics for various populations of A. gigas.
Regression
Character
Gulf of Guinea
Body depth Jaw length Jaw width
NE Atlantic (37°N, lO-^W) Body depth Jaw lengtli Jaw width
Indkin Ocean (0°-40°S, 167°E)
Body depth Jaw length
SE Pacific (Chile) Body depth Jaw lengtli
0.65
0.71
-0.33
-4.42
0.23
-1.36
-0.60 0.39
-2.21 -0.71
0.50 0.21 0.14
N
0.58 0.22 0.16
N
0.54 0.22
N:
0.54 0.24:
N:
.131 .055 .040 10
.104 .044 .029 14
.332 .137 5
.155 .068 9
Marine Hatchetfishes • Baird
43
■igure 28. Argyropelecus hemlgymnus; R/V CHAIN, Cruise 60; Station 1299; SL 30 mm.
604; Handrick, 1901: system, light organs ) ;
1 ( anatomy, nervous Collett, 1903: 108;
Ledenfeld, 1905: 170 (light organs); Braner,
Regan, 1908: 218;
( lai^vae, fig. ) ; Ziig-
and Byrne, 1913: 21
1906: 106 (larvae, fig.) Ehrenbaum, 1909: 357 mayer, 1911: 52; Holt (larvae, fig.); Jespersen, 1915: 6; Jespersen and Taning, 1919: 220 (lai-vae, eye muscles): Nusbaum-Hilarowicz, 1923: 10 (anatomy): Barnard, 1925: 153; Jespersen and Taning, 1926: 59; Sanzo, 1928: 50 (eggs, larvae), Norman, 1930: 301; Borodin, 1931: 44 (eggs, larvae); Jespersen, 1934: 15 (larvae, fig.): Buen, 1935: 52; Fowler, 1936: 1208; Beelie. 1937: 201; Parr, 1937: 49 (spines); Norman, 1937: 82; 1939: 19; Nybehn, 1948: 23; Maul, 1949b: 13; Misra, 1952: 367; Smith, 1953: 102; Kotthaus and Krefft, 1957: 3; Peres, 1958: 4 (bathyscaphe); Koefoed, 1961: 5; Schultz, 1961: 601; 1964: 241; Blache, 1964: 71; Backus et al., 1965: 139; Kotthaus, 1967: 22 (photo, otoliths); Bright and Paquegnat, 1969: 28.
Argijropelccus d'uiviUi Valenciennes, in Cuviei and Valenciennes, 1849: 405; Goode and Bean, 1896: 127.
\rp,\)vovelccus intermedins Clarke, 1878: 248; Schultz, 1961: 587; 1964: 241; Blache, 1964: 71; Berry and Perkins, 1965: 625; Kotthaus, 1967: 11 (photo.); Lavenberg and Ebeling, 1967: 185.
Argyropelectis heathi Gilbert, 1905: 601; Fowler, 1949: 42; Haig, 1955: 321. '
Species distinction. Differs from all other species in genus by its narrow trunk, single postabdominal spine, small size, minute teeth, presence of only eight dorsal and eleven anal rays.
Description. D. 8; A. 11; P. 10-11; total gill rakers (18) 19-23 (24); vertebrae (36) 37-38.
Dwarf species rarely exceeding 38 mm SL; trunk very long and narrow; its depth at origin of anal photophores three or more times into greatest body depth, subcaudal photophores well separated from anals; dorsal spine medium-to-high, its height often exceeds its length; post-temporal spines well developed; postabdominal spines fused to form a single spine com- plex; lower preopercle spine directed ven- trally, the upper posterio-dorsally; jaws medium; teeth small to minute; gill rakers long and numerous; in preservative ab- dominal region dark, trunk pigmentless
44 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Table 4. Comparisons of regression statistics for three populations of A. hemigymnus. Re- gression A refers to pigment form a (see text); Regression B refers to pigment form B. NS
INDICATES no SIGNIFICANT DIFFERENCE BETWEEN THE SLOPES OF THE TWO REGRESSIONS INDICATED.
|
Regression A |
- |
Regression B |
- |
|||||
|
Character |
A |
B |
A |
B |
Slope-T-Test |
|||
|
Southern Ocean |
(Pacific) |
|||||||
|
Body depth |
-0.61 |
0.53 ± |
.041 |
0.45 |
0.48 ± |
.063 |
NS |
|
|
Caudal peduncle |
width |
0.52 |
0.07 ± |
.099 |
0.59 |
0.07 ± |
.016 |
NS |
|
Jaw length |
0.09 |
0.23 ± |
.026 |
0.68 |
0.21 ± |
.036 |
NS |
|
|
Jaw width |
-0.36 N = |
0.14 ± :41 |
.020 |
0.01 N = |
0.13 ± :49 |
.041 |
NS |
|
|
Gulf of Mexico |
||||||||
|
Body depth |
-0.95 |
0.54 ± |
.086 |
1.62 |
0.45 ± |
.161 |
NS |
|
|
Caudal peduncle |
width |
0.30 |
0.09 ± |
.020 |
0.34 |
0.09 ± |
.037 |
NS |
|
Jaw length |
0.15 |
0.23 ± |
.050 |
-0.15 |
0.26 ± |
.108 |
NS |
|
|
Jaw width |
-0.33 N = |
0.14 ± :17 |
.038 |
0.88 N = |
0.09 ± :10 |
.053 |
NS |
|
|
N Pacific (California) |
||||||||
|
Body depth |
0.38 |
0.51 ± |
.121 |
0.43 |
0.49 ± |
.176 |
NS |
|
|
Dorsal blade |
0.56 |
0.08 ± |
.031 |
0.72 |
0.07 ± |
.063 |
NS |
|
|
Jaw length |
0.60 |
0.21 ± |
.047 |
-0.58 |
0.24 ± |
.103 |
NS |
|
|
Jaw width |
0.30 N = |
0.11 ± : 14 |
.042 |
-0.36 N = |
0.14 ± :8 |
.060 |
NS |
except in definite patches along midline and above anal and subcaudal photophore groups.
Pigment forms. Two pigment forms designated form "A" and form "B" occur over much of the species range. Form A is characterized by distinct and clearly de- fined body pigmentation, while in form B the body pigmentation is quite diffuse. This pigment difference is not a function
of size or sex, is intermediate in few in- dividuals, and both forms do occur in the same catch. A morphometric analysis of three sympatric populations in several characters and meristics (Table 4; Figs. 33 and 34), plus measurements from one or the other pigment forms from other areas (Table 6) failed to show any sig- nificant difference between sympatric populations. In addition, there was no
Table 5. Diurnal capture comparisons of the two pigment forms of A. hemigymnus from
VARIOUS AREAS FROM DEPTHS OF 0 M TO 1000 M. # = TOTAL NUMBER OF HAULS; # POS. = NUMBER OF POSITIVE HAULS; # 20+ = NUMBER OF POSITFV'E HAULS CONTAINING 20 OR MORE INDIVIDUALS.
|
Pigment Form |
Night |
Day |
|||||
|
Locale |
# |
# pos. |
# 20-f- |
# |
# pos. |
# 20+ |
|
|
NE Atlantic |
A |
41 |
9 |
2 |
47 |
20 |
14 |
|
(20-35°N, 0-30°W) |
B |
41 |
25 |
9 |
47 |
7 |
1 |
|
N Atlantic |
A |
129 |
15 |
6 |
102 |
43 |
22 |
|
(37-45°N, .30-70°W) |
B |
129 |
27 |
12 |
102 |
6 |
2 |
|
Gulf of Mexico |
A |
45* |
3 |
1 |
35 |
19 |
6 |
|
and Caribbean |
B |
45* |
4 |
0 |
35 |
0 |
0 |
|
Southern Ocean |
A |
58 |
12 |
3 |
25 |
19 |
3 1 |
|
B |
58 |
32 |
10 |
25 |
3 |
\ |
* 18 of these hauls were less than 200 ni.
Marine Hatchetfishes • Baird
45
o
_Q
E
o
CD
O U
<
E o
E 'a
c
o
O
X
46 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
consistent sorting out over the range of variability of any one form in any char- acter. Analysis of diurnal depth distri- butions revealed a marked sorting out of pigment types with form A most numerous during the day while form B was pre- dominantly caught at night (Table 5). Distributions from which Table 5 was com- piled were chosen from areas where this species appears to occur throughout the horizontal sampling space. It is concluded that these two forms represent pigment states of the same species whose states can be varied individuallv.
Recently Badcock ( 1969 ) reported diel color variation in several mesopelagic fishes (including A. hemigymnus) and at- tributed it to a correlation with ambient light conditions. This appears to be the case in Argyropelecus hemigymnus. There are two anomalies, however, which raise some interesting points. First is the occur- rence, occasionally in great numbers, of the day form at night and the night form during the day. Second is the relatively rare occurrence of form B in the tropics and its apparent absence from the Medi- terranean (over 300 specimens examined).
Tropical submergence is present in this species (see below) and may help explain the rareness of form B in the tropics. Nevertheless, there is considerable com- plexity in the depth distribution of this species and if the pigment change is sen- sitive to small differences in ambient light. it may be used as an indication of seasonal or geographical changes in depth distri- bution, changes in sea water turbidity, or of other correlated information.
Distribution. Horizontal distribution ( Figs. 29 and 30 ) : Occurs in the South Atlantic around the Falkland Islands and abundantly off the southeast coast of Brazil; a scattering of catches along lati- tude 35° S to the Cape of Good Hope sug- gests a broad distribution across the South Atlantic; occurs in small catches along the southwest African coast, appears absent in the Gulf of Guinea, but occurs in the
western tropical Atlantic; is taken in moder- ate numbers in the Caribbean and Gulf of Mexico and abundantly in the western Atlantic; is abundant across the North Atlantic and the eastern North Atlantic as far south as the Cape Verdes Islands; it represents the only species of this fam- ily in the Mediterranean, where it oc- curs abundantly in the western basin; scattered moderate-to-small catches are present from 5°N to I2°S latitude in the central Indian Ocean, and another popu- lation is scattered from 20° S to 40° S with several small catches reported from the southeastern and southwestern Indian Ocean; a single catch off the Philippines, another at 42°N, I69°E, and small catches from the Banda Sea and near Hawaii represent this species in the west and cen- tral Pacific; large populations occur off California and Chile; it is taken abun- dantly across the Southern Ocean from 35°- 55°S latitude from Chile to New Zealand; taken in small numbers in the Tasman Sea and off Sidney, Australia.
Vertical distribution (Figs. 31 and 32): Occurs from 200 m to 700 m by day with the greatest concentration l)etween 350- 550 m; occurs from 100 m to 650 m by night with concentrations between 150-380 m; tropical submergence indicated in the Gulf and Caribbean by examining number of catches above 200 m (Appendix B) com- pared with the North Atlantic; by day it appears to concentrate at about 550 m in the Sargasso Sea (Dr. James Craddock, WHOI, personal conversation).
Geographic variation. At least seven dif- ferent populations could be discerned and are identified and statistically defined for a number of characters in Tables 4 and 6 and Figures 33 and 34. Small samples from the central Pacific and Cape Verdes Islands may indicate separate populations also. Broad variations in slope between several populations were noted and these ' were tested for statistical significance (Table 7) indicating considerable world- wide variability and distinct population
Marine Hatchetfishes • Baird 47
s
o
E
E >^
i
(U
c
o
o
I
48 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
aO
B D C O □ • E <r F •
F ISH H R
Figure 31. Diurnal vertical distribution of A. hemigymnus, pigment form A, determined by rate of capture with depth during the day (D) and night (N). A ^ Pacific (California); B = Southern Ocean; C = SW Atlantic; D = Gulf of Mexico and Caribbean; E ^ N Atlantic; F == NE Atlantic.
characteristics; dorsal blade height and gill raker number differences (Figs. 33 and 34) further emphasized the disthictness of populations in this species.
Argyropelecus aculeatus Valenciennes Figure 35
Argyropelecus aculeatus Valenciennes, in Cuvier and Valenciennes, 1849: 406 (holotype MNHNP 1817; Azores; not seen); Giintlier, 1864: 384; Sauvage, 1891: 483; CoUett, 1903:
108; Brauer, 1906: 110; Regan, 1908: 218; Jespeisen, 1915: 11; Norman, 1930: 301; Borodin, 1931: 68; Jespersen, 1934: 15; Beebe, 1937: 201: Bertin, 1940: 314 (holotype); Maul, 1949a: 17; Misra, 1952: 367; Bigelow and Schroeder, 1953: 149; Koefoed, 1961: 7; Schultz, 1961: 607; 1964: 241; Backus et al., 1965: 139; Kamohara and Yamakawa, 1965: 22; Bright and Paquegnat, 1969: 29.
Argyropelecus olfersi: Goode and Bean, 1896: 127; Jordan and Evermann, 1896: 604 (?); Rivero, 1934: 31; 1936: 56; Cer%ig6n, 1964: 1.
Argyropelecus ( Sternoptychides ) amabilis Ogilby
Marine Hatchetfishes • Baird
49
2oo.
400
600
800 1000
□ 1500
E P
T
D
□ °So
D
D
\0
20
30
|
0 |
n |
D |
N |
|||||||||
|
200. |
it |
■it |
||||||||||
|
400. |
D a |
D D |
ix |
♦ |
• |
••• • |
• |
• |
• |
* |
||
|
600 |
♦ |
♦ |
||||||||||
|
800 |
D 1* |
• |
||||||||||
|
1000 |
||||||||||||
|
1500 |
||||||||||||
|
b |
»0 |
20 |
30 |
AO B •
C O D D E -tt
F -k
F ISH H n
Figure 32. Diurnal vertical distribution of A. hemigymnus, pigment form B (see Fig. 31).
1888: 313: Culloch, 19 foed, 1961 Argyropelecus Argijropelecus 1936: 246; Argyropelecus Argyropelecus
Goode and Bean, 1896: 127; Mc- 23: 118; Whidey, 1940: 404; Koe-
7; Schultz, 1961: 607, 1964: 241.
caninus Garman, 1899: 235. acanthurus (not Cocco) Fowler, Maul, 1949b: 13; Dollfus, 1955: 24.
rnicracanthus Parr, 1937: 49.
antrorsospinus Schultz, 1937: 5.
Species distinction. See A. olfersi (p.
52).
Description. D. 9; A. 12; P. 10-11; total gill rakers 15-17; vertebrae 34-36.
Large species often exceeding 70 mm SL; body very deep, depth at end of
dorsal less than 1.4 into SL; dorsal spine qnite high, its height about equal to its length; post-temporal spines present; dor- sal surface of post-temporal with distinct serrations; postabdominal spines well de- veloped, the posterior much larger than anterior; ventral keel extends ^^'ell below body margin near postabdominal spines; preopercle spines short, both pointing ven- trally; jaws large, teeth long, recurved, with two enlarged canines in lower jaw; spines present below and in front of sub- caudal photophores; gill rakers medium to
50 Bulletin Museum of Co7nparative Zoology, Vol. 142, No. 1
Table 6. Regression statistics for various populations of A. hemigymnus.
Regression
Character
A
NE Atlantic (36°-39°N, 27°W)
Pigment Form B
Body depth
Caudal peduncle width
Jaw lengtli
Jaw width
NW Atlantic (36°N, 55-60°W)
Pigment Form A
Body depth
Candal pednncle width
Jaw length
Jaw width
Indian Ocean (5°-35''S, 55°-65°E)
Pigment Form B
Body depth
Dorsal blade
Jaw length
Jaw width
Mediterannean Pigment Form A Body deptli Dorsal blade Jaw length Jaw width
1.44 0.97 1.69 0.39
-0.33 0.55
-0.04 0.43
-0.96 0.35 0.99 0.83
1.44 -0.24
1.16 -1.38
0.45 0.06 0.17 0.11
N
0.51 0.08 0.24 0.12
N
0.55 0.11
0.21 0.09
N
0.45 0.09 : 0.20 0.19
N:
.069 .016 .036 .035 35
.081 .019 .052 .034 24
.266 .129 .158 .123
7
.104 .030 .056 .045 15
short, with dentate inner surfaces; pigment diffuse on trunk, no marked pigment on midline, pigment concentration above sub- caudals present, pigmentless bar anterior to caudal peduncle in young.
Distribution. Horizontal distribution ( Fig. 36 ) : Taken abundantly in the Carib- bean and Gulf of Mexico; in the western North Atlantic to about 40°N and 35°W; occurs in the northeastern Atlantic south
Table 7. Slope comparisons of regressions of several characters between various populations of a. hemigymnus. a = pigment form a; b = pigment form b.
|
Character |
Population 1 |
Population 2 NE Atlantic B NE Atlantic B NE Atlantic B NE Atlantic B |
T 2.073 2.222 2.211 2.854 |
P |
|
|
Jaw length |
Gulf of Mexico B Gulf of Me.vico A NW Atlantic A Soutliern Ocean A |
.05 .035| .034] .00^ |
|||
|
Jaw width |
Gulf of Mexico B |
Southern Ocean |
A |
2.109 |
.04 1 |
|
Candal peduncle depth |
Gulf of Mexico A Gulf of Mexico A |
NE Atlantic B Southern Ocean |
A |
2.659 2.098 |
.01 1 .051 |
|
Jaw width |
Mediteiranean A |
California A |
2.548 |
.02I |
Marine Hatchetfishes • Baird 51
Figure 33. Geographic variation in gill raker number in A. hemigymnus, pigment forms A and B. A =: NE Atlantic; B = NW Atlantic; C = Gulf of Mexico; D = Mediterranean; E = Indian Ocean; F = Soutfiern Ocean (Pacific); G = N Pacific; H =: Pacific (California). Numbers refer to sample size.
of about 35°N along the North African coast and associated islands; essentially absent from the tropical Atlantic; small to moderate catches in the southwestern At- lantic represent this species; taken in the central Indian Ocean from about 10°S to 40°S and reported abundant off the easteni South African coast; a few records scat- tered along the western Pacific from north of New Guinea to Japan represent it in the western Pacific; a number of moderate catches indicate its presence in the north central Pacific; these are matched by similar catches off Chile and one large haul off Sidney, Australia.
Vertical distribution (Fig. 37): Occurs between 200 m and 550 m by day with the greatest concentrations from 350-450 m; marked diunial movement with major con- centrations from 80-200 m at night; Sar- gasso Sea captures indicate concentrations !at about 520 m by day (Dr. James Crad- :lock, WHOI, personal conversation).
Geog^raphic variation. Because of large samples available this species was used for 1 detailed population study in the Atlantic, [t allowed checks to be made of within- lopulation variation both from different vcars and as subsamples of the same catch; urthermore, an examination of samples
in the northwest Atlantic provided an op- portunity to look at variations over at least 15° of longitude in the same biogeographic region. Table 8 records these results. In the northwest Atlantic, results indicate that population parameters remain constant in
A B
C D
E F G
|
A |
B |
|
rio 10 |
1 i_j 14 |
|
ni] s |
|
|
I iLl 10 |
|
|
r^£} 17 |
Cfc 7 |
|
rin 13 |
|
|
(ii \o |
r^ 12 |
|
liQ 15 |
liL 15 |
|
rin 13 |
jM 12 |
|
cir ^ |
r^ 12 |
Figure 34. Geograpfiic variation in dorsal blade fieight in A. hem\gymnu%, pigment forms A and B, for standard lengths 23-28 mm. A = NW Atlantic; B = Cape Verdes Islands; C = NE Atlantic; D = Gulf of Mexico; E =: Mediterranean; F = Pacific (California); G = Southern Ocean; H ^ Indian Ocean; I = Caribbean and Tropical Atlantic. Numbers refer to sample size.
52 Biillctin Museum of Comparative Zoology, Vol. 142, No. 1
Figure 35. Argyrope/ecus aculeatus; R/V CHAIN, Cruise 60; Station 1266; SL 46 mm.
the same locality from year to year. In addition, populations in this area taken at the same latitude but separated by 15° of longitude show no indication of changes in values of parameters measured — in fact they appear to remain remarkably con- stant. Once again division of a large haul from the Caribbean into two subsamples gave little variability with adequate sample sizes. In a given area, populations seem to remain distinctive both from year to year and over a broad range in the same biogeographical region. There appears to be a clinal variation between populations in the Atlantic, going from the Caribbean, to the Gulf of Mexico, to the northwest Atlantic. Gill raker number and body
depth ( Figs. 38 and 39 ) show a clinal vari- ability and possibly jaw length (Table 8) as well. Six separate populations are identified and statistically defined in Table 8 and Figures 38 and 39. The South Pacific (Chile) population is quite distinct from the others. Differences in slope between the Caribbean and northeast Atlantic popu- lations were significant in several char- acters (Table 9).
Argyropelecus olfersi (Cuvier) Figure 40
Sternoptyx olfersi Cuvier, 1829: 316 (holotype MNHNP 1889; Cape of Good Hope; not seen).
Argyropelecus olfersi, Cuvier and Valenciennes, 1849: 408; Collett, 1903: 108; Brauer, 1906: 69; Regan, 1908: 218; Zugmayer, 1911: 52;
Marine Hatchetfishes • Baird
53
D
c
<D O
75 >
-Q
E
o
0)
o u
"o
-^ 'o
~5
o
X
54 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
|
• • • |
•0- -ttH |
♦ |
|
|
Zoo |
■♦t |
n* * |
|
|
400 |
♦ |
• • • |
|
|
t>00 . |
D |
||
|
8oo |
D |
||
|
lOOO |
|||
|
15 00 |
N
A O
B O
C • D -d
E f
F D
10
20
30
FISH
HR
Figure 37. Diurnal vertical distribution of A. aculeatus determined by rate of capture with depth during day (D) and nighf (N). A = Pacific (Chile); B = N Central Pacific; C = Caribbean and Tropical Atlantic; D = NW Atlantic; E = Gulf of Mexico; F = NE Atlantic.
Holt and Byrne, 1913: 120; Jespersen, 1915: 23; 1934: 15; Roule and Angel, 1933: 46; Buen, 1935: 52; Parr, 1937: 49 (spines); Bertin, 1940: 314 (holotype); Nybelin, 1948: 23; Bertelsen and Grontved, 1949: 163 (light organs); Maul, 1949b: 13; Dollfus, 1955: 1; Holgersen, 1958: 120 (population density); Koefoed, 1961: 10; Schultz, 1961: 610; 1964: 241; Wheeler, 1969: 136.
Species distinction. Differs from A. aculeatus by absence of subcaudal spines,
less deep body (see regressions, body dcptli, Tables 8 and 10), lower dorsal spine, higher \'ertebral count and post- abdominal spine characteristics; differs from A. lychnus by presence of enlarged canines, lighter pigment, no subcaudal spines, preopercle and post-temporal spine characteristics and first anal photophore; differs from A. sladeni by presence of enlarged canines; no definite pigmented
Marine Hatchetfishes • Baird
55
Table 8. Regression statistics for various populations of A. aculeatus.
Regression
Character
SE Pacific (Chile)
Body depth Jaw length
Gulf of Mexico (24°N, 83°W)
Body depth
Caudal peduncle depth
Jaw length
Caribbean (13°N, 71°W) (Sample 1)
Body depth
Caudal peduncle depth
Jaw length
Caribbean (13°N, 71°W) (Sample 2)
Body depth
Caudal peduncle depth
Jaw length
N\V Atlantic (42°N, 47°W) (9/64)
Body depth
Caudal peduncle depth
Jaw length
NW Atlantic (41°N, 62°W) (9/64)
Body depth
Caudal peduncle depth
Jaw length
NW Atlantic (42°N, 62°W) (9/62)
Body depth
Caudal peduncle def)th
Jaw length
NE Atlantic (32°N, 13° W)
Body depth
Caudal peduncle deptli
Jaw length
-1.64 1.27
0.49 0.12 0.70
2.32 0.20 0.70
2.85 0.39 0.66
0.69 -0.06 -0.05
0.60
0.23
-0.17
0.25
0.57
-0.47
1.32
1.34
-0.72
0.77 ± .274 0.22 ± .087
N = 8
0.67 ± .084
0.12 ± .015
0.23 ± .032
N = 23
0.64 ± .064 0.12 ± .013 0.24 ± .025
N=r23
0.63 ± .080 0.11 ± .018 0.25 ± .034
N = 26
0.66 ± .087 0.12 ± .020 0.26 ± .031
N = 28
0.67 ± .061 0.12 ± .012 0.26 ± .025
N = 30
0.67 ± .055
0.11 ± .010
0.27 ± .038
N = 40
0.69 ± .123
0.10 ± .019
0.29 ± .042
Nzz29
Table 9. Slope comparisons of regressions of various characters between two populations of
A. ACULEATUS. ThE CARIBBEAN POPULATION CONSISTS OF TWO SUBSAMPLES ( SEE TabLE 8).
Character
PopulaHon 1
Population 2
|
Caudal peduncle depth |
NE Atlantic |
Caribbean 2 |
2.009 |
.05 |
|
Jaw length |
NE Atlantic |
Caribbean 1 |
2.266 |
.038 |
|
Jaw lengtli |
NE Atlantic |
Caribbean 2 |
2.059 |
.048 |
56 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
15
A
B
Table 10. Regression statistics for two popu- lations OF A. OLFERSI.
□
|
1 |
15 11 15 1 |
Character |
Regression |
|
|
^51^ |
A |
B |
||
|
NE Atlantic Body depth Jaw length Southern Ocean Body depth Jaw length |
(Pacific) |
0.51 0.56 1.74 -0.19 |
0.64 ± .231 0.26 ± .099 K = 8 0.61 ± .158 0.28 ± .075 N = 10 |
|
|
T-» • . .7 |
1 |
1 -I . . |
||
|
\h 18 |
Figure 38. Geographic variation in gill raker count in A. aculeatus. A = NW Atlantic; B = NE Atlantic; C = N Central Pacific; D = Pacific (Chile); E = Caribbean; F =1 Gulf of Mexico. Numbers refer to sample size.
midline; deeper body (see regression, body depth, Table 11), spine characteristics, anal photophores, and lower gill raker count.
Description. D. 9; A. 12; P. 10-11; total gill rakers (15) 16-17; vertebrae 36-37 (38).
Large species often exceeding 70 mm SL; body deep, depth at end of dorsal usually greater than 1.5 times into SL; first preanal photophore with pointed dorsal margin; dorsal spine high, its height nearly one-half its length; post-temporal spines well developed; postabdominal spines nearly equal, anteriormost spine curves smoothly forward; lower preopercle spine long, curving fonvard, upper very short; jaws large; teeth recurved with two large canines in lower jaw and a somewhat smaller pair in the upper jaw; pigment diffuse over whole of trunk; no marked midline f)igment spots; less marked con- centration of pigment in caudal peduncle; dark pigment present on outermost caudal rays (this often lost in handling).
Distribution. Horizontal distribution (Fig. 36): Restricted to the northeast Atlantic between latitudes 35°N and 65°N and east of longitude 35°W; occurs in a broad band across the southern Pacific between 30° S and 50° S from Chile to New Zealand; re- ported southwest of the Cape of Good Hope suggesting a bipolar distribution in the Atlantic; not reported from the North Pacific or southern Indian Ocean.
Vertical distribution (Fig. 41): Data variable by day with relatively low con- centrations from 200 m to 750 m; by night depths are concentrated between 200 m and 450 m with most records from LSO m to 300 m; no indications of marked geo- graphic variation in depth.
Geographic variation. Analysis of small sample sizes from the two major widely separated populations indicate no statisti- cal differences and little evidence of separation (Table 10; Fig. 42).
Argyropelecus sladeni Regan Figure 43
Ar'^ijropelccus sladeni Regan, 1908: 218 (liolotype BMNH; Central Indian Ocean; not seen); Jespersen, 1934: 15; Fowler, 1936: 1208; Parr, 1937: 49 (fig., incorrectly cites Norman, 19.30 as original description); Norman, 1939: 19;
Marine Hatchetfishes • Baird
57
45
15-
20
BO
S L
Figure 39. Geographic variation in the regression of body depth (BD) on standard length (SL) in A. aculeatus. A _ NW Atlantic; B = Gulf of Mexico; C = NE Atlantic; D = Caribbean; E = Pacific (Chile).
Marr, 1948: 140; Misra, 1952: 367; Haig, 1955:
321; Fowler, 1956: 27; Koefoed, 1961: 1. \rgtiropdecus olfersi: Barnard, 1925: 153; Smith,
1957: 37 (?); Bright and Paquegnat, 1969: -29. Argtjropelccus Itjchntis hjchnus Schultz, 1961: 587
(in part); 1964: 241;" Blache, 1964: 71; Backus
et al., 1965: 139; Bright and Paquegnat, 1969:
30. Argijwpelecus lycliniis sladcni Schultz, 1961: 587;
1964: 241 (incorrectly cites Norman, 1930, as
original description); Kotthaus, 1967: 22
(photo., otoliths). Argyropelecus hjchnus hawaicnsis Schultz, 1961:
587; 1964: 241. Argyropclcciis hawaicnsis Berry and Perkins, 1965:
625; Lavenberg and Ebeling, 1967: 185.
Species distinction. See A. olfersi (p. 52) and A. lychnus (ix 63) .
Description. D. 9; A. 12; P. 10-11; total gill rakers 17-21; vertebrae 35-37.
Medium size species seldom exceeding 60 mm SL; body less deep, depth at end of dorsal about two or more times into SL; dorsal blade low, height about three or more times into its length; postabdominal spines of equal size, anterior one occasion- ally straight, usually squared or blunted; upper preopercle spine long, directed posteriorly and usually dorsally, lower directed ventrally and often slightly posteriorly; jaws medium; teeth small, re- curved, no large canines present; gill rakers medium to long, slightly dentate; first pre- anal photophore raised well above second
58 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
E E o
CN
n o
U
Z
<
— 1 I— <
>
a. o
Marine Hatchetfishes • Baird
59
Zoo. AOO.
600 •
800. 1000. 1500
B *
N
¥■¥■
¥ *
10 FISH HR
20
Figure 41. Diurnal vertical distribution of A. olfersi determined by rate of capture witfi depth. A (Pacific); B = NE Atlantic.
30
Soutfiern Ocean
which is even with or above third; anal pterygiophore gap with three haemal spines lacking pterygiophores; in preserva- tive pigment often quite dark; large dis- tinct pigment spots present along midline, especially evident in smaller specimens; there may be a diurnal pigment difference similar to A. hemigymnus in this species.
Distribution. Horizontal distribution (Fig. 44): In the Atlantic this species is found in abundance along the African coast from
about 15 °S northward into the Gulf of Guinea; it occurs in moderate numbers across the equatorial Atlantic in a belt from 5°S to 15 °N latitude; it is abundant in the Caribbean in the vicinity of the Venezuelan coast, absent from the northern Caribbean, appearing again in numbers in the western and northern Gulf of Mexico and the straits of Florida; a few small catches have been taken in the North Atlantic and along the Brazilian coast. In
60
Bulletin Museum of Comparative Zoologtj, Vol. 142, No. 1
9
1
Table 11. Regression statistics for variousI populations of a. sladeni.
14
17
Figure 42. Geographic variation in gill raker count in A. oliersi. A := NE Atlantic; B z:^ Southern Ocean. Numbers refer to sample size.
the Pacific a somewhat biantitropical dis- tribution is indicated, with large populations represented in the North Pacific to about 175°W longitude, and off the California coast; another large population occurs off the coast of Chile; the species occurs north of New Zealand and south of Hawaii. A. sladeni is abundant in the northern Indian Ocean to about 15 °N and along the African coast to about 10° S; while not reported from the Bay of Bengal, it is represented by several small catches south of Java.
Vertical distribution ( Fig. 45 ) : Concen- trated between 350 m and 600 m by day, with the major concentrations between 350 m and 450 m; by night concentrated be- tween 100 m and 375 m, \\'ith the major concentrations between 100 m and 300 m; no marked indication of geographic vari- ation with depth.
Geographic variation. This species, like A. fiigas, has low variability in those b()d\-
|
Regression |
||
|
Character |
A |
B |
|
N Pacific (42°N, 165° W) |
||
|
Body depth |
0.38 |
0.52 ± .156 |
|
Dorsal blade |
2.02 |
0.01 ± .027 |
|
Jaw lengtli |
-0.03 |
0.25 ± .076 |
|
faw width |
0.48 |
0.13 ± .070 N = 10 |
|
£ Pacific (CaHfoiiiia) |
||
|
Body depth |
1.25 |
0.53 ± .074 |
|
Dorsal blade |
2.62 |
0.00 ± .022 |
|
Jaw lengtli |
0.92 |
0.23 ± .027 |
|
Jaw widdi |
0.62 |
0.10 ±.028 N = 27 |
|
Indian Ocean (05°N, 65°E) |
||
|
Body depth |
0.80 |
0.56 ± .142 |
|
Dorsal blade |
1.89 |
0.01 ± .010 |
|
Jaw length |
0.22 |
0.24 ± .069 |
|
Jaw width |
-0.73 |
0.13 ± .043 N = 11 |
|
Caribbean |
||
|
Body depth |
0.87 |
0.52 ± .163 |
|
Dorsal blade |
1.29 |
0.02 ± .013 |
|
Jaw length |
-0.07 |
0.26 ± .081 |
|
Jaw width |
0.75 |
0.12 ± .044 N = 9 |
|
Gulf of Guinea |
||
|
Body depth |
-0.91 |
0.57 ± .110 |
|
Dorsal blade |
1.89 |
0.01 ± .014 |
|
Jaw lengtli |
-0.33 |
0.27 ± .055 |
|
Jaw width |
1.14 |
0.11 ± .027 N = 13 |
|
SE Pacific (Chile) |
||
|
Body depth |
0.51 |
0.54 ± .097 |
|
Dorsal blade |
2.04 |
0.01 ± .012 |
|
Jaw length |
1.04 |
0.23 ± .040 |
|
Jaw width |
-0.63 |
0.14 ± .035 N = 16 |
Table 12. Comparisons between mean slopes of two ch.'^racters for all populations of A.
SLADEM and a. LYCHNUS FOR WHICH REGRESSION STATISTICS WERE CALCULATED. PoP. # = NUMBER OF
populations; x Slope = unweighted mean slope; Total # = total number of fish measured
OVER all populations.
Species
Character
Pop. #
Slope
Range
Total #
|
A. sladeni |
body depth |
6 |
0.54 |
0.52-0.57 |
86 |
|
A. hjchnus |
body depth |
3 |
0.61 |
0.57-0.64 |
38 |
|
A. sladeni |
jaw widtli |
6 |
0.12 |
0.10-0.14 |
86 |
|
A. hjchnus |
jaw width |
3 |
0.15 |
0.14-0.16 |
38 |
Marine Hatchetfishes • Baird
61
CN
o
^0
u
<
X
u
c
01
o
-2
ID
a.
o
>s
O)
62
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
o
J3
E
o
O
c O
o
X
zl
Marine Hatchetfishes • Baird
63
Zoo
Aoo
600
800 1000
D iSOO
E P T H
|
D |
||||||
|
• |
||||||
|
• ••• |
n |
•k • |
it |
•k |
• |
|
|
• |
0 |
|||||
|
0 |
10
20
30
|
2o |
• • |
• |
N |
||
|
2oo 400 |
■k |
D |
• |
* |
|
|
^00 |
|||||
|
600 |
|||||
|
1000 |
|||||
|
1500 |
A •
B D
C O
D •
E O
F if.
F ISH H R
20
30
Figure 45. Diurnal vertical distribution of A. sladeni determined by rate of capture with depth during day (D) and night (N). A — Pacific (California); B = Pacific (Chile); C = Gulf of Guinea; D = Gulf of Mexico and Caribbean; E = N Central Pacific; F = N Atlantic.
proportions measured. Overlap is broad and sample sizes are small. Six populations were statistically defined (Table 11; Fig. 46) but only gill raker counts gave much separation. Certainly the Atlantic popu- lation is distinct from the Indian Ocean and several Pacific populations; within the latter, distinctions are not marked. The Indian Ocean, Chile, and California popu- lations show some separation, although not statistically significant. Other characters
and larger sample sizes are required to better define populations in this species.
Argyropelecus lychnus Gorman
Figure 47
Argyropelecus lycJmus Garman, 1899: 234 (lecto- type USNM 57885, designation Schultz, 1961; tropical east Pacific, not seen; paialectot>pe MCZ 35193, seen); Ledenfeld, 1905: 170 (light organs); Berry and Perkins, 1965: 625; Grand- perrin and Rivaton, 1966: 36; Lavenberg and Ebeling, 1967: 185.
64 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
A
B C D
E
F
16
16
20
2Z
Figure 46. Geographic variation in gill raker count in A. sladeni. A ^ Gulf of Guinea; B = Caribbean; C = Indian Ocean; D ^ N Central Pacific; E =: Pacific (Chile); F r= Pacific (California). Numbers refer to sample size.
Argyropclccus olfersi: Weber and DeBeaiifort,
1913: 1 (?); Clemens and Wilby, 1949: 106;
Koumans, 1953: 186 (?); Morrow, 1957: 56;
Koepcke, 1962: 145; Bussing, 1965: 185. Argyropclccus lycJmus lychnus Sehultz, 1961: 587
(in part); 1964: 241." Argyropeleciis sp., Kotthaus, 1967: 11 (?)
(photo.).
Species distinction. See A. olfersi (p. 52); differs from A. sladeni by its higher dorsal lilade, preopercle spine character- istics, presence of two rather than three haemal spines in anal pterygiophore gap,
lack of distinct dark pigment spots on midline, broader body, and generally lower gill raker count (Figs. 46 and 49). Tables 12 and 13 and Figure 50 illustrate the nature and degree of difference in several of the characters mentioned above.
Description. D. 9; A. 12; P. 10-11; total gill rakers 16-18; vertebrae 35-37.
Medium to large species often exceeding 60 mm SL; body deep, depth at end of dorsal greater than 1.5 into SL; dorsal blade high, height about 2.5 times into its length; postabdominal spines of about equal size, anterior one slightly smaller, not smoothly curving but blunted or squared; upper preopercle spine long, di- rected posteriorly and usually ventrally; lower spine usually curved slightly ante- riorly or straight down; jaws large, teeth recurved especially in lower jaw, no large canines; gill rakers medium to short, den- tate; first preanal photophore usually lower than third; spiny scales present in adults below subcaudal photophores; the gap made by the anal pterygiophores contains two haemal spines lacking pterygiophores; ill preservative, pigment dark dorsally, diffuse on trunk with small, light pigment spots on midline.
Distribution. Horizontal distribution (Fig. 44): Absent from the Atlantic; represented possibly by a single sample from the Indian Ocean (04°S, 66°E, Kottliaus, 1967). Pri-
Table 13. Slope compabisons of the regression of dorsal blade height on standard length for various populations of a. lychnus (l) and a. sladeni (s).
|
Character |
Population 1 — L |
Population 2 — S |
T |
p |
|
Dorsal blade |
C Pacific |
Chile |
3.179 |
.005 |
|
C Pacific |
N Pacific |
2.452 |
.025 |
|
|
C Pacific |
Indian |
2.904 |
.01 |
|
|
Chile |
Chile |
3.903 |
.001 |
|
|
California |
Chile |
3.965 |
.001 |
|
|
Chile |
N Pacific |
2.171 |
.045 |
|
|
Chile |
California |
3.514 |
.001 |
|
|
Chile |
Indian |
3.366 |
.005 |
|
|
California |
N Pacific |
2.272 |
.035 |
|
|
California |
California |
3.355 |
.005 |
|
|
California |
Indian |
3.444 |
.005 |
Marine Hatchetfishes • Baird 65
E E o
O
If)
o
CO
CO (U
u
X
O
X
J) "S
Q. O
O)
N. ■^
3
66
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
H
|
o |
N |
|||||
|
Zoo. ^00. |
o |
o • . * * |
• * |
* |
* |
|
|
6oo. |
||||||
|
8oo. |
||||||
|
1000. |
||||||
|
1500 |
BO
C •
fc 10
F ISH H R
20
30
Figure 48. Diurnal vertical distribution of A. lychnus determined by rate of capture with deptfi during the day (D) and night (N). A = Pacific (California); B = Pacific (Chile); C = Tropical E Pacific.
marily restricted to the tropical Pacific; found in abundance in the eastern Pacific between 35°N and 35°S; distribution nar- rows across the equatorial Pacific as far as 160^W; a moderate catch from tlic lesser Sunda Islands indicates a possible trans- equatorial distribution in the Pacific.
Vertical distribution (Fig. 48): Concen- trated between 300 m and 400 m off California by day, with the highest con- centration near 400 m; by night major
concentrations occur from 200 m to 350 m, with no marked indication of geographical variation in depth.
GcQfiraphic variation. Three samples from widely separated areas in the tropical east Pacific and its northern and southern boundaries gave no indication of any sig- nificant variation (Table 14; Fig. 49). Horizontal distribution data indicates an essentiallv continuous distribution in this
area.
Marine Hatchetfishes • Baird
67
A
B C D
Table 14. Regression statistics for various populations of a. lynchus.
Regression
\h
18
Figure 49. Geographic variation in gill raker count in
A. lychnus. A = Central Pacific; B = Tropical E Pacific;
C ^ Pacific (Chile); D = Pacific (California). Numbers refer to sample size.
Genus Sfernoptyx Hermann, 1781
SteiiiOjityx Hermann, 1781: 8 (type species: Stc'inoptyx dicijilxana Hermann, 1781, by mono-
typy).
Diagnosis. Ten abdominal, three anal, three branchiostegal and five isthmus photophores; a single large dorsal pterygio- phore spine with an anterior, serrated ex- tension; first anal pterygiophore greatly enlarged, forms support for triangulate membrane above anal fin rays; premaxilla without anterior pedicels; anteriormost gill rakers reduced to toothed ridges; post- temporal and supracleithrum separate; hy- pural elements fused to form single caudal plate; haemal and neural spines greatly elongate in trunk region.
Description. Photophores: PO 1; PTO 1; PRO 1; SO 1; SP 3; PAN 3; SAN 1; AN 3; SC 4.
Spines: Preopercle with single ventrally oriented spine; retroarticular bears spine, preabdominal spine present; basipterygia fused to form a set of four postabdominal spines; base of first anal pterygiophore bears ventral spines; no well-developed post-temporal spines.
Character
SE Tacific (Chile) Body depth Dorsal blade Jaw length Jaw width
E Pacific (California) Body depth Dorsal blade Jaw length Jaw width
2.38
1.56
0.84
-0.46
0.94
1.53
-0.15
-1.02
Central Pacific (10°N, 145°W)
Body depth 1.42
Dorsal blade 1.79
Jaw length 0.88
Jaw width -0.77
0.57: 0.05: 0.26: 0.16
N:
0.61 0.04 0.29 0.16
N
0.64 0.05 0.28 0.14
N
.120 .018 .055 .043 12
.107 .013 .058 .031 15
.156 .023 .068 .052 11
Eyes: Large, well developed, nontele- scopic.
Gill rakers: Total seven to nine; well developed, \\'ith rough spiny margins; an- teriormost rakers reduced to spiny tooth- like plates extending into mouth cavity.
Jaws and dentition: Jaws vertically oriented, premaxilla small, heavily toothed; maxilla heavily toothed and major upper jaw bone in gape; lower jaw heavily toothed, teeth small, sharp, triangulate; palatine teeth present; first epibranchial extended anteriorly and ventrally forming toothed arms at dorsal, posterior end of mouth.
Meristics: Vertebrae 28-3 J; C. 9+10; D. 8-11; A. 14-16.
Color: Bright silvery in life, dark pig- ment especially evident on dorsal surface;
68
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
30
B
D
20
3 O
60
S L
Figure 50. Regression of body depth (BD) on standard length (SL) In A. sladeni (A) and A. lychnus (B) off California.
.silver (pickly lost in formalin preservative; sides can be very dark, often black.
Internal anatomy: Air bladder and gas gland well developed and fills much of the body cavity (see Marshall, 1960); digestive system simple, consisting of a heavily pigmented, single sectioned stom- ach, capable of considerable distension,
five to seven pyloric caecae of which only two to three are long and well developed, relatively small liver and short, uncoiled, thin-walled intestine. The gonads, when mature, lie against the posterior wall of the body cavity. This cavity is large, unlined with pigment, and appears capable of some expansion.
Key to the Species of Sternoptyx
la. Dorsal long, its lengtli greater than 1.3 times height of dorsal spine; trunk long and narrow, SL more than 3.0 times l^ody depth at end of dorsal (see regression, body depth. Table 15); body very dark, pigment forms broad band at base of caudal rays S. ohscura ( p. 69 ) .
b. Dorsal short, its length less or equal to height of dorsal spine; trunk broad, SL less than 2.8 times body deptli at end of dorsal (see regression, body depth, Tal:)les 16 and 17); body pigment less uniformly dark, pigment absent or in very narrow band at base of caudal rays 2
2a. Supra-anal photophore high, its height greater than one-half the distance from ventral body margin to midline ( often raised to midhne ) ; gill raker tooth plates with prominent spines; secondary anal pterygiophores long, e.xtending posteriorly on same level as anal photophores _.._ S. pseudobscura (p. 72).
b. Supra-anal photophore low, its height less than one-half distance from ventral body margin to midline; gill raker tooth plates lacking prominent spines; secondary anal
pterygiophores short; not extending posteriorly on same level as anal photophores
S. duiphana (p. 75 ).
Marine Hatchetfishes • Baud 69
a.
Sfernopfyx obscura Garman Figure 51
Sternoptyx obscura Garman, 1899: 63 (lectntype USNM 177888; designation Schultz, 1961; tropical east Pacific; not seen; paralectotype MCZ 28532; seen); Ledenfeld, 1905: 170 (light organs); Follett, 1952: 409.
Sternoptyx diaphana Schultz, 1961: 587 (in part); 1964: 241 (in part); Berry and Perkins, 1965: 625 (in part).
Species distinction. Differs from both S. diaphana and S. pseudohscura in its shorter dorsal spine and longer dorsal fin; longer, narrower trunk, slight extension of body
Figure 51. Sfernopfyx obscuro; R/V ANTON BRUUN, Cruise 3; Station 215; SL 30 mm.
70 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
margin in front of anal photophorcs, broad pigment band on base of caudal rays, and generally dark pigment; differs from S. pseiidohsciira in its lo\\'er supra-anal photo- phore, small teeth, low gill raker tooth plates, and smaller mouth; from S. dia- phana in its extension of the ventral body margin at same level behind anal photo- phores. Tables 18 and 19, and Figures 5.3, 56, and 61 illustrate the degree of differ- ence between the three species in several of the above characters. Note especially the significant differences between sym- patric populations.
Description. D. 10-11; A. 14-15; P. 10- 11; total gill rakers 7-9; vertebrae 29 (30).
Small species, seldom exceeds 40 mm SL; trunk long and narrow, its length usually longer than depth; dorsal fin long, its length more than 1.3 times the length of dorsal spine; abdominal length along midline from supra-anal photophore to caudal peduncle, less than or equal to body depth at end of dorsal; postabdominal and anal pterygiophore spines long; posterior anal pterygiophores extend behind and at same level \\\\\\ anal photophore group; supra-anal photophore raised above anals one half or less the distance to midline; body margin extends slightly in front of anal photophores before curving ventrally; jaws medium; teeth small; gill raker tooth plates consist of multiple low spiny ridges; anterior dorsal surface of tongue between branchial arches smooth; few-to-no raised nodules; in preservative, pigment very dark over whole of body; pigment extends in broad band at base of caudal fin rays.
Distribution. Horizontal distribution (Fig. 52): Tliis species has not been recorded from the Atlantic; it is concentrated north of 10°S latitude in the Indian Ocean, al- though small catches occur as far as 40°S; occurs in the eastern Bay of Bengal, and abundantly south of Java; occurs off the Philippines and scattered but large catches indicate in all probability a continuous distribution across the equatorial Pacific; occurs abundantly in the tropical east
Table 15. Regression statistics for various populations of s. obscura.
|
Regression |
||
|
Character |
A |
B |
|
Indian Ocean (5°N, 60°E) |
||
|
Body depth |
-0.35 |
0.31 ± .051 |
|
Abdominal length |
-1.16 |
0.40 ± .074 N = 23 |
|
Indian Ocean (3°N, 67°E) |
||
|
Body dcpdi |
0.45 |
0.29 ± .133 |
|
Jaw length |
2.41 |
0.09 ± .067 N= 10 |
|
Java (10°S, 114°E) |
||
|
Body depth |
0.98 |
0.27 ± .071 |
|
Abdominal length |
0.68 |
0.31 ± .090 N = 25 |
|
Central Pacific (11°N, 163°E) |
||
|
Body length |
-1.72 |
0.37 ± .067 |
|
Abdominal length |
0.12 |
0.37 ± .074 |
|
Jaw length |
0.85 |
0.14 ± .028 N = 20 |
|
Central Pacific (7°S, 135°W) |
||
|
Body depth |
-0.71 |
0.34 ± .104 |
|
Abdominal lengtli |
0.71 |
0.35 ± .141 |
|
Jaw length |
0.10 |
0.15 ± .057 N = 11 |
|
East Pacific (California) |
||
|
Body depth |
-0.94 |
0.33 ± .075 |
|
Abdominal length |
-1.23 |
0.42 ± .091 |
|
Jaw length |
1.85 |
0.12 ± .034 N = 15 |
Pacific from California to the Chile-Peru border.
Vertical distribution: Depth data is spotty and no depth rate plot \\'as made, however, data (Appendix B) indicates a depth range of 650 m to at least 1000 m; tropical Pacific maximum net depth figures concur in general with this range.
Geographic variation. Analysis of catches from many widely scattered areas in the Pacific and Indian Ocean gave no indi- ^ cation of population variation (Table 15, | Fig. 53). This, coupled with horizontal distribution data, indicates a probable single trans-Indo-Pacific population.
Marine Hatchetfishes • Boird
71
<u a
a -a
J2
E
o
o u
J3 O
■D o o
O
-O
o
3
o
X
72 Bulletin Museum of Compamtwe Zoology, Vol. 142, No. 1
25
A B C D
E F G H
Table 16. Regression statistics for various populations of s. pseudobscura.
W
Z4
J*1
\s
s
Figure 53. Geographic variation in gill raker count in S. obscura. A =: E Indian Ocean; B ^ Central Indian Ocean; C = Tropical E Pacific; D ^ Banda Sea; E ^ Central Tropical Pacific; F =: Marshall Islands; G := Pacific (California); H =; Indian Ocean — S of Bali. Num- bers refer to sample size.
Sternopfyx pseudobscura n. sp. Figure 54
Holotype MCZ 46400, 1° 20'S, 27° 37'W; 2/27/ 63; R/V CHAIN, Cruise 35; Station 977.
Sternoptyx diaphana: Braiier, 1906: 69 (in part); Maul, 1949b (in part); Blache, 1964: 71; Backus et al., 1965: 139 (in part).
Specie.9 distinction. See S. obscura (p. 69), differs from S. diapliana by its larger mouth (see jaw length, Fig. 62), longer teeth and gill raker tooth plate spines; markedly higher supra-anal photophore; extension of long anal pterygiophores be- hind and at same level with anal photo- phores. Tables 18 and 19, and Figures 56, 61, and 62 illustrate the nature and degree of difference between the two species.
Description. D. 10-11; A. 14-16; P. 10- 11; total gill rakers 7-9; vertebrae 29.
Regression
Character
Gulf of Guinea
Jaw length 1.23
Photophore 0.43
Caribbean
Jaw length 1.70
Photophore 2.28
Florich (30°N, 76°W)
Jaw length 1.85
Photophore 0.58
Indian Ocean (6°-35°S, 55°-65°E)
Jaw length 2.04
Photophore 2.21
Central Pacific
B(Kly depth -1.06
Abdominal length 0.21
Jaw length 1.63
0.15 ± .243 0.19 ± .076
N = 8
0.15 ± .045 0.11 ± .052
N = 13
0.16 ± .080
0.15 ± .091
N = 6
0.16 ± .200
0.17 ± .098
N = 6
0.49 0.32 0.15
N
.078 .066 .061
13
Largest species in genus, often exceeds 55 mm SL; trunk broad, its depth greater than length; dorsal spine long, its length about equal to, or less than, length of dor- sal fin; posterior anal pterygiophores long, extend behind and at same level as anal photophores; supra-anal photophore very high, its height more than one-half the' distance from ventral body margin to mid- line (often raised to midline); no body margin extension in front of anal photo- phores; jaws large; teeth well developed and recurved; gill raker tooth plates with long spines; usually one much longer than others; anterior dorsal surface of tongue with small nodules; postabdominal and anal pterygiophore spines long; in preserv- ative, pigment dark over most of body except lighter in trunk region; if present, pigment band very narrow at base of caudal rays.
Marine Hatchetfishes • Baird 73
Figure 54. Sternopfyx pseudobscuro; R/V CHAIN, Cruise 60; Station 1310; 22 mm.
Holotvpe: measurcment.s (mm), SL 4.S.1, BD 15.0, JL 07.4, CP 04.6, Ab. length 14.0; meristics: GR 7, D 9, A 15, anal photo- phores 3; name derivation: pseiidobscura refeis to this species' close resemblance to S. ohscura.
Distribution. Horizontal distribution (Fig. 52): The limited distributions seen here may be artifacts resulting from the vertical distribution of this species; widely scattered but moderate-to-high catch numbers are additional indications of sampling prob- lems. This species occurs in the South Atlantic off Brazil, and in the Gulf of
Guinea off Africa with ^\idely scattered occurrences in the tropical Atlantic; it is abundant in the northern Gulf of Mexico and the straits of Florida; scattered catches indicate its presence in the southern Carib- bean; it has not been reported from the western North Atlantic, but occurs in num- bers off the northwest coast of Africa and near the Azores; small catches indicate its presence in the southwestern Indian Ocean from 5°S to 40°S latitude; isolated small- to-moderate catches south of Java, near the Marshall Islands, in the North Pacific, off California, and in the southeast Pacific
74 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
loo.
Qoo looo
D 1^00
E
p
T H
|
D |
||||||||
|
• |
||||||||
|
• o |
• |
oo |
||||||
|
6 |
10 |
2 0 |
30 |
AO
Figure 55. Diurnal vertical distribution of S. pseudobscura determined by rate of capture with depth during day (D) ondi night (N). A ;::: Gulf of Guinea; B =: NE Atlantic; C =: Gulf of Mexico and Caribbean.
indicate a broad range in the Pacific which future collecting should better define.
Vertical distribution ( Fig. 55 ) : The deepest living species in the genus; data indicate a depth distribution from 500 m to 1500 m; greatest concentrations recorded occur between 800 m and 1500 m; overall small rates of capture plus relatively few deep stations may be indicative of a depth preference below 1000 m; no diurnal movement is indicated.
Geographic variation. Small sample sizes, few characters, and the few populations represented precluded a detailed exami- nation of variability in this species. No differences were noted in the Atlantic populations, or in gill raker counts, or jaw length among all populations (Table 16, Fig. 56). The Indian Ocean population could be distinguished from the Atlantic ones by the supra-anal photophore mea- surement (Fig. 57).
Marine Hatchetfishes • Baird 75
-r-
6
—I— 8
-igure 56. Geographic variation in gill raker count in j. pseudobscura. A ^ Caribbean; B =: N Central Pacific; Z ;= Indian Ocean; D := Marshall Islands. Numbers refer o sample size.
Sternoptyx diaphana Hermann Figure 58
Sternoptyx diaphana Hermann, 1781: 33. In accordance with article 75a(i) Int. Code Zool. Nomen., a neot>'pe is hereby designated: Neo- type MCZ 46402; 11° 06'N, 78° 2rW; 8/7/66; R/V ANTON BRUUN, Cruise 19; Station 813. Cuvier and Valenciennes, 1849: 415; Giinther, 1864: 387; Goode and Bean, 1896: 127; Al- cock, 1896: 331; Gilbert, 1905: 601; Brauer, 1906: 69 (in part); 1908: 175 (eye muscles); Holt and Byrne, 1913: 20; Weber and DeBeau- fort, 1913: 1; Jespersen, 1915: 12; Jespersen
and Tiining, 1919: 220 (eye); Borochn, 1931: 68; Jespersen, 1934: 15; Roxas, 1934: 287; Buen, 1935: 52; Fowler, 1936: 1208; Beebe, 1937: 22; Parr, 1937: 49; Norman, 1937: 82; 1939: 19; Nybelin, 1948: 25; Maul, 1949a: 17; 1949b: 13 (in part); Wilimovsky, 1951; Misra, 1952: 367; Koumans, 1953: 186; Mead and Taylor, 1953: 570; Smith, 1953: 102; Haig, 1955: 321; Rass, 1955: 328; Grey, 1959: 326; Koefoed, 1961: 11; Schultz, 1961: 617 (in part); 1964: 241 (in part); Backus et al., 1965: 139 (in part); Berry and Perkins, 1965: 682 (in part); Bussing, 1965: 185; Haedrich and Nielsen, 1966: 909; Bright and Paquegnat, 1969: 34.
Species distinction. See S. obscnra (p. 69) and S. pseudobscura (p. 72).
Description. D. 9-11; A. 14-16; P. 10- 11; total gill rakers 7-8 (9); vertebrae 29 (30). _
Medium size species, seldom exceeds 55 mm SL; trunk very broad; its depth usually greater than its length; dorsal spine long, its length greater or equal to dorsal fin length; posterior anal pterygiophores short, little extension behind and on same level with anal photophores; supra-anal photo- phore low, not reaching more tlian one-half the distance from ventral body margin to midline, no body margin extension in front of anal photophores; jaws medium to small; teeth short and low; gill raker tooth plates with low spinate ridges; anterior dorsal
ID
P
a
o
o
SI
o
o
-Q_
O
O
o *
o
A • B* CO
3 O
S L
5 Q
■igure 57. Geographic variation in distance from dorsal body margin of supra-anal photophore (PO) with standard length SL) in S. pseudobscura. A =: Indian Ocean; B =Gulf of Guinea; C = Gulf of Mexico and Caribbean.
76 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Figure 58. Sternoptyx diaphana; GALATHEA; Station 494; 33 mm.
surface of tongue with small nodules; post- abdominal and anal pterygiophore spines usually shorter than others in genus; in preservative pigment dark dorsally, often light and dispersed in trunk region, usually little pigment present at base of caudal rays.
Neotype: measurements (mm): SL 28.4, BD 12.1, JL 05.3, CP 03.5, Ab. length 09.0; meristics: GR 7, D 10, A 15, anal photo- phores 3.
Distribution. Horizontal distribution (Fig. 59): Broadly distributed in the Atlantic, caught in moderate numbers in the South Atlantic off Brazil and from 20°W to the African coast at about 35^ S; abundanth'
present in the Gulf of Guinea and the tropical Atlantic; taken abundantly in the southern Caribbean, the Gulf of Mexico,! and straits of Florida; taken in the western North Atlantic; a large population occurs in the northeastern Atlantic from 25°N to 45°N latitude. In the western Indian Ocean small to moderate catches extend from 5°S to 35 °S latitude, a single catch has been observed from the eastern Indian Ocean; numerous catches indicate this species present south of Java, near Borneo, and in the Banda Sea; known also between New Guinea and the Solomon Islands, it occurs in the westem Pacific near the Philippines and along the coast of Japan, with a small
I
Marine Hatchetfishes • Baird 77
75
JQ
E
o
a U
o
c o -c a o
c O
o
X
78 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
AC B ••< C •• D C
F ISH HR
Figure 60. Diurnal vertical distribution of S. diaphana determined by rate of capture with depth during day (D) an night (N). A ^ Southern Ocean; B ^ Gulf of Guinea; C =: Gulf of Mexico and Caribbean; D =: NE Atlantic.
sample taken in the North Pacific; this species occurs in small numbers off lower California and has been reported abun- dantly southeast of Hawaii; a large popu- lation extends across the South Pacific from Chile to about 160° E longitude.
Vertical distribution (Fig. 60): Distrib- uted between 400 m and 1200 m; major concentrations occur between 700 m and 900 m; no diurnal movement or marked
geographical depth xariation could be dei tected.
Geographic variation. With the excep tion of the Pacific southern ocean popm lation, little geographic variation could b* detected; characters were few and sampl( sizes small, however (Table 17, Fig. 61); The Southern Ocean population is quit«| distinct and certainly represents an ini stance of incipient speciation. Tables 1'
Marine Hatchetfishes • Baud
'able 17. Regression statistics for various populations of s. diaphana.
|
Rf |
jgression |
|
|
Character |
A |
B |
|
laribbcan |
||
|
aw length |
-0.24 |
0.18 ± .066 |
|
hotophore |
-2.04 |
0.37 ± .101 N = ll |
|
ava (10°S, 114°E) |
||
|
lody depth |
-2.93 |
0.54 ± .107 |
|
ibdominal length |
0.78 |
0.31 ± .056 N = 22 |
|
ndian Ocean (5°-40°S, 55° |
-65°E) |
|
|
lody depth |
0.10 |
0.51 ± .095 |
|
Lbdoniinal length |
1.18 |
0.31 ± .093 |
|
hotophore |
0.20 |
0.29 ± .085 |
|
aw length |
1.00 |
0.15 ± .088 N = 12 |
|
oiithern Ocean |
||
|
lody depth |
-0.77 |
0.42 ± .044 |
|
abdominal length |
1.18 |
0.31 ± 093 |
|
aw length |
1.47 |
0.16 ± .041 |
|
hotophore |
2.23 |
0.11 ± .033 N = 40 |
|
outhern Ocean (Chile) |
||
|
lody depth |
-1.50 |
0.41 ± .113 |
|
aw length |
1.46 |
0.14 ± .037 N=ll |
nd IS indicate the degree of difference letween this population and others in the pedes. In body depth it falls somewhat »et\veen most populations of S. diaphana nd S. ohscura (Table 19). Phenotypically t has supra-anal photophore characteristics
resembling some populations of S. pseud- ohscura. In most characters it falls closest to other populations of S. diaphana, espe- cially in mouth and gill raker character- istics. Considering the lack of sympatry with other forms and the degree of distinctness between the species, it is presently con- sidered to represent a distinct form of S. diaphana.
Genus Polyipnus Gijnther, 1887
Polyipnus Ciinther, 1887: 170 (type species: Polyipnus spinosus Giintlier, 1887, by mono-
typy).
Diagnosis. Ten abdominal, three supra- abdominal, and a lateral photophore; post- temporal spine(s) well developed; a fused double dorsal pterygiophore forms short spines anterior to dorsal fin rays; cleithrum projects below pectoral fin forming fanlike, spine-bearing extension posteriorly; otoliths very large with characteristic armlike ex- tension (see Kotthaus, 1967); three to four hypural elements in upper caudal lobe; lower jaw noticeably expanded dorsally.
Description. Photophores: PO 1; PTO 1; BR 6; I 6; PRO 1; SO 1; SP 3; SAB 3; AB 10; L 1; PAN 5; AN 6-14; SC 4.
Spines: Post-temporal extends posteriorly to form from one to three prominent spines; preopercle spined, the lateral sur- face often bearing spiny elements; retro- articular spined; ventral surface of lower jaw often serrate; cleithrum bears pre- abdominal spine; bony keel scales often bear spines ventrally; four postabdominal
ABLE
18. Slope comparisons between regressions of sever,\l characters in species of
StERNOPTYX. D = S. DIAPHANA; O = S. OBSCURA; P = S. PSEUDOBSCURA.
|
Character |
Population 1 |
Population 2 |
T |
P |
|
5ody depth |
Java D Central Pacific P Java D |
Java O Central Pacific O Southern Ocean D |
2.780 2.362 2.611 |
.001 .025 .010 |
|
Uadominal Icngdi |
Java D |
Southern Ocean D |
2.381 |
.023 |
|
'hotophore |
Caribbean P Gulf of Guinea P Florida P |
Caribbean D Caribbean D Caribbean D |
4.977 3.240 2.767 |
.001 .005 .018 |
80
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
IS
A
B
C □
38
rin
22
10
6
— r-
8
Figure 61. Geographic variation in gill raker count in S. diaphana. A = Caribbean; B = Indian Ocean; C =^ Southern Ocean; D = Pacific (Chile); E = Indian Ocean — - S of Bali; F z= Java Sea. Numbers refer to sample size.
spines; first anal pterygiophore may extend ventrally to form small spines.
Eyes: Large, well developed, nontele- scopic, essentially laterally oriented.
Gill rakers: Number 10-28; well de- veloped, long, often quite close together; first branchial arch considerably larger
than succeeding arches; the inner surfaces of second and thir'd cerato- and epi- branchials bear tooth plates.
Jaws and dentition: Mouth small and vertically oriented; premaxilla long, toothed, and major upper jaw bone in gape; arm of first supramaxilla elongate; dorsal margin of lower jaw greatly ex- panded which, with broadening in the meso-and metapteryoid, make the mouth cavity a long conelike basket with a sub- stantial distance between the mouth en- trance and the beginning of the branchial arches; the maxilla is toothed, but essen- tially excluded from gape; teeth small to minute, no canines, vomer and palatines bear teeth.
Meristics: Vertebrae 31-36; C. 9+10; D. 10-17; A. 13-19.
Color: Bright silvery in life; dark dorsal pigment band often extends ventrally and may reach lateral midline; dark pigment may form lateral striated bands on pos- terior trunk.
Internal anatomy: Swim bladder and associated gland well developed, gland quite large, with grainy appearance; l)lad- der thick walled, and often heavily in- vested with fatty tissue (see Marshall, 1960); digestive system simple with bi- partate stomach, anterior section thicl< walled, the lining often raised into heavy
Table 19. Comparisons between mean slopes of several characters among the species or Sternoptyx. S.O. = Southern Ocean populations of S. diaphana; Pop. # = number of popula- tions; X Slope = unweighted mean slope; Total # = total number of fish measured over ali
populations.
|
Species |
Character |
Pop. # |
X Slope |
Range |
Total # |
|
S. obscure |
body depth |
6 |
0.32 |
0.27-0.37 |
104 |
|
S. diaphana |
body depth |
2 |
0.525 |
0.51-0.54 |
34 |
|
S. diaphana ( S.O. ) |
body depth |
2 |
0.415 |
0.41-0.42 |
51 |
|
S. pseudob.scura |
body depdi |
1 |
0.49 |
0.49 |
13 |
|
S. diaphana |
photophore |
4 |
0.31 |
0.29-0.37 |
35 |
|
S. diaphana (S.O.) |
photopliore |
1 |
0.11 |
0.11 |
40 |
|
S. pseudohscura |
photophore |
4 |
0.15 |
0.11-0.18 |
33 |
|
S. obscura |
jaw length |
4 |
0.13 |
0.09-0.15 |
56 |
|
S. diaphana |
jaw length |
5 |
0.196 |
0.15-0.23 |
41 |
|
S. diaphana (S.O.) |
jaw length |
2 |
0.15 |
0.14-0.16 |
51 |
|
S. pseudohscura |
jaw length |
5 |
0.15 |
0.15-0.16 |
46 |
Marine Hatchetfishes • Baud
81
30
SL
so
Figure 62. Regression of lower jaw length (JL) on standard length (SL) in S. diaphana (A) and S. pseudobscura (B) from the Tropical Atlantic.
ridged folds; die posterior section diin walled and extensible, six to ten pyloric caecae, short straight intestine, and a rel- atively large well-developed liver. Gonads, when mature, lie horizontally and laterally in the body cavity; cavity lined with pig- mented membrane; nepliritic tissue is not as well developed as in other genera.
Species complexes: As with Arfiyropele- ciis, there has been considerable radiation within the genus. There are three distinct species complexes, two closely related. The third complex, P. spinosiis, is quite distinct. The latter appears more primiti\e in terms of axial and caudal skeleton characteristics. The P. spinosws complex differs from the other two groups as follows: greater de- velopment of the post-temporal spine complex; otoliths (Weitzman, personal con- versation); four hypural elements in the upper caudal lobe; serrate lower jaw mar-
gin; spine-bearing abdominal keel plates; the second and tliird pretenninal neural spines wedge shaped; extension of the cleithrum below the pectoral relati\ely more broadened; and the anal pterygio- phore gap is reduced.
Within the P. spinosus complex there is a further dichotomy. P. nuttingi, P. oJiiohis, and P. indicus with peculiar reductions in post-temporal spine characteristics, fomi one group; P. spinosus, P. sterope, and P. tridentifer with a well-developed, post- temporal complex, fonn the other.
The P. asteroides and P. htenwtus spe- cies complexes are more closely related. They differ primarily in post-temporal spine characteristics, body shape and size, photophore number and pattern, and den- tition. Both complexes have similar otoliths and resemble each other osteologically.
The species P. later natiis — P. omphiis
82 BuUctin Museum of Comparative Zoology, Vol 142, No. 1
and P. iinispinus — P. aquavit us form a dichotomy within the P. latenuitus complex. The differences inckide preopercle spine length, body shape, photophore pattern, and some meristic differences (gill rakers, vertebral nnmber).
P. astcroides — P. poUi — P. triphanos and P. matsu])ami — P. meteori — P. kiiciensis — P.
rugg,eri form a similar dichotomy within the P. asteroides complex. The latter group is characterized by an extremely long, narrow, posterior vomerine shaft, closely allied and fitting into the parasphenoid. This shaft bears teeth anteriorly, in addi- tion to the normal lateral vomerine teeth. Other minor differences are also joresent.
Key to the Species of Polyipnus
la. Post-temporal spine complex, bearing one or two basal snpplementary spines; dorso-lateral edge of fused, post-temporal-supracleithrum serrate; abdominal keel scales with spiny ventral surfaces; supra-anal photophore group usually not distinctly separated from anal
group P- spinosus complex 3
1). Post-temporal spine simple, bearing no basal spines; lateral edge of post-temporal-supra- cleithrum smooth; keel scales with smooth ventral surfaces (except P. imispinus); supra- anal photophores separate, usually raised well above anal group ._ __ — _ 2
a.
b.
2a. Post-temporal spine long, its length greater than one-fourth the diameter of orbit; anal photophore number 10 to 13 (adults only); body long and narrow, SL greater than 1.7 times body depth; first supra-anal photophore even with or raised above the second
(except P. late maim in which the first is slightly lower than die second)
P. laternalus complex 8
b. Post-temporal spine short, its length less than one-fourth the diameter of orliit; anal photophore number 7 to 9 (occasionally 10); body more robust, SL less than 1.9 times
body depth; first supra-anal photophore markedly lower than second -
P. asteroides complex 1 1
1
^
A
a.
Marine Hatchetfishes • Baird
83
3a. Post-temporal spine with two distinct basal spines; anal-subcaiidal photophore distance less than one-third the length of the subcaudal group; anal photophore number 10 to 13 4
b. Post-temporal spine with a single distinct basal spine (this reduced in P. oluohis); anal- sul)caudal distance greater than one-half of the length of the subcaudal group; anal photopbore number 6 to 9 6
4a. Anal photophore number 10; SL less than 3.6 times body depth at end of dorsal fin;
caudal peduncle broad, head length less than 2.8 times narrowest peduncle depth 5
b. Anal photophore number 12 to 13; SL greater than four times body depth at end of dorsal; caudal peduncle narrow, head length greater than three times narrowest
peduncle depth P. tridentifer (p. 86).
5a. Post-temporal basal spines well developed, ventralmost basal spine lengtli greater than
one-half length of post-temporal spine; gill raker numlaer 24-28 P. sterope (p. 88).
b. Post-temporal basal spines short, ventralmost basal spine less than one-fourth post- temporal spine length; total gill raker nmnber IS to 21 P. spinosus (p. 89).
a. b.
6a. Alidominal keel scales triangulate, with one or two large ventral spines; iDost-temporal spine long, heavily spinose dorsally and laterally; first supra-anal photophore markedly
lower than third; anal photophores 6 to 7 P. oluohis (p. 90).
b. Abdominal keel scales rectangular, with many small ventral spines; post-temporal spine long and smootli or short and spinose dorsally only; first supra-anal photophore about even with or raised above tliird; anal photophores 8 to 9 ._ 7
a.
7a. Ventral margin of subcaudal photophores with spines (adults); anal-subcaudal photo- phore distance less than three-fourths length of subcaudal group; first supra-anal photophore higher than the last; post-temporal spine long, greater than one-half
diameter of orbit P. iudiciis (p. 91).
b. Ventral margin of subcaudal photophores smooth; anal-subcaudal distance greater tlian three-fourths length of subcaudal group; first supra-anal photophore lower than last; post-temporal spine short, less than one-half diameter of orbit .. . P. nuttingi (p. 92).
84 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
8a. Gill rakers 18 to 21; supra-anal photophores distinctly raised above tlie anal group; preopercle spine short, somewhat triangulate 9
b. Gill rakers 12 to 14; supra-anal photophores not markedly raised above anal group; preopercle spine long and slender - — ....10
'^^
Aap\
m=^
^^Peep^
9a. Sulicaudal photophores compact, length of subcaudal group less tlian narrowest caudal peduncle depth; distance from top of last supra-anal photophore to top of first anal greater than three-fourths length of preanal group P. latcrnatus (p. 92).
b. Subcaudal photophores somewhat spread; length of subcaudal group equal or greater than narrowest peduncle depth; distance from top of last supra-anal photophore to top of first anal less than one-half preanal length P. omphus (p. 94).
10a. Posterior lateral margin of preopercle smooth; xentral keel scales smooth; dorsal
.spine short, length less than one-fourth diameter of orbit P. aquavitiis (p. 96).
b. Posterior lateral margin of preopercle serrate; ventral edge of keel scales with small
spinelets; dorsal spine long, length greater than one-half diameter of orbit —
. P. iini.spinus (p. 97).
11a. Teeth absent on posterior vomerine shaft; second supra-anal photophore e\en with or raised above third; supra-abdominal photophores nonsymmetrical, first photophore
raised well above other two 12
b. Teeth present on posterior vomerine shaft; second supra-anal photophore lower than third; supra-abdominal photophores symmetrical, first not markedly raised above third 14
Marine Hatchetfishes
Baird
85
f\
^
ft
,^
f\
flA
a.
%(\
12a. Gill rakers 16 to 18; dorsal rays 11 to 12; lower margin of dark dorsal pigment markedly
raised above supra-anal photophores P. triphanos (p. 97).
b. Gill rakers 20 to 24; dorsal rays 14 to 16; lower margin of dark dorsal pigment straight from dorsal spine to caudal peduncle , 13
13a. Anal pterv'giophores form circular margin below anal photophores; anal photophores
number 7 to 8; SL less than 3.5 times body depth at end of dorsal P. polli (p. 98).
b. Anal pterygiophore margin essentially straight; anal photophore nimiber 9 (rarely 10); SL greater than 3.7 times body deptli at end of dorsal P. asteroides (p. 99).
86
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
14a. Gill rakers 22 to 24; dark pigment bar extending to midline very narrow, its width
less tlian greatest widtli of lateral photophore (see illustration p. 101)
P. matsuharai ( p. 101 ) .
b. Gill rakers 13 to 18; dark pigment bar absent or much wider than width of lateral photophore 15
15a. Dark pigment bar greatly reduced or absent; supra-abdominal photophores essentially
in straight line (see illustration p. 102); gill rakers 18 P. ruggeri (p. 102).
b. Dark pigment bar present, extending to midline; supra-abdominal photophores tri- angulate, the second markedly lower than the otlier two; gill rakers 1.3 to 17 16
16a. Gill rakers 13 to 15; light stripe behind dark pigment bar extends to mid-dorsal line; ventral border of dark dorsal pigment markedly raised above supra-anal photophores
(taken from photo and description, Kotthaus, 1967) P. mctcori (p. 104).
b. Gill rakers 16 to 17; light stripe behind dark pigment bar not extending to mid-dorsal line; ventral border of dark dorsal pigment not markedly raised above supra-anal photophores (see illustration p. 103) P. kiwiemis (p. 103).
PoJyipnus tridentifer McCulloch Figure 64
Pohjipnus tridentifer McCulloch, 1914: 78 (lecto-
type AM E.3.543; designation Schultz, 1961;
Australian Bight; not seen); Schult/., 1961: 619;
1964: 247. Pohjipnus spinostis: Weber and DeBeaufort, 1913:
1; Matsubara, 1950: 192; Okada and Suzuki,
1956: 297; Suzuki, 1964: 1. Pohjipnus frazeri Fowler, 1933: 2.57; Schultz,
1961: 620.
Species distinction. Differs from P. spinostis and P. sieropc in its long, narrow trunk and caudal peduncle; long, smooth post-temporal spine; more sharply angled dorsal spine; multispinose subcaudal scales; less spinose abdominal keel scales; differs from P. spinosiis by its much longer third basal post-temporal spine.
Description. D. 13-14; A. 15-17; P. (12) 13-14; total gill rakers (20) 21-24; verte- brae 33-34. '
Medium size species, not often exceed- ing 60 mm SL; trunk tapering into long, narrow, caudal peduncle; its depth less or equal to length of subcaudal photophore group; post-temporal spine long, its length more than one-half the distance from its base to point of dorsal spine; second basal i post-temporal spine long; dorsal surface of post-temporal spine relatively smooth, lack- ing marked serrations; supra-abdominal photophores arranged in steplike fashion, each raised an approximate ec^ual distance, above the next; abdominal keel scales spinose, although spines very short; scales below subcaudal photopliores with several prominent spincvs; preoperclc spine di-
Marine Hatchetfishes • Baird
87
I® a! .-
E Q."
a
t/i o
E
<i
X
• - c
° ■>
o
;5 o
c
n
^
D- O
3
Q.
5^
|
a. |
o c |
|
|
t/> |
a |
|
|
3 |
-cr |
|
|
t- |
0 |
|
|
0) |
||
|
O) |
||
|
0) |
o' |
|
|
■ , |
(•) |
|
|
o |
||
|
c |
^- |
|
|
() |
(11 |
|
|
3 |
c- |
|
|
0) |
||
|
^ |
D |
|
|
^ |
_i- |
|
|
0) |
Q.' |
|
|
1— CO |
• |
O |
|
VI |
||
|
ri |
o |
|
|
(11 |
"^ |
|
|
3 |
o |
Cl, |
|
cn |
a to |
^ |
88
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
-,-%r.::-.
Figure 64. PoJyipnus tridentifer; after Matsubara, 1950.
rected, at its base, posteriorly, curving distally to point ventrally or slightly an- teriorly; jaws medium; teeth minute; gill rakers long; pigment in preservative dark dorsally with narrow, dark bar extending toward midline; pigment diffuse on trunk. Distribution (Fig. 63). Restricted to the western Pacific, taken abimdantly around the Philippines, off the south China coast, off Japan, in waters north of the Strait of Malacca, and in the Great Australian Bight.
Polyipnus sferope Jordan and Storks Figure 65
PoJyipnus stcropc Jordan and Starks, 1904: 581 (holotype USNM 51451; Sagami Bay, Japan; seen); Matsubara, 1941: 2; Haneda, 1952: 12 (light organs); Okada and Suzuki, 1956: 297; Suzuki, 1964: 1 (X-ray).
Polyipnus spinosus: Kamohara, 1952: 17.
Poli/ipnus spinosus sterope Schultz, 1961: 621; 1964: 247.
Species distinction. See P. triclcntifer
Figure 65. Polyipnus sterope; modified from Jordan and Starks, 1904.
Marine Hatchetfishes • Baud
89
igure 66. Po/yipnus sp/nosus; modified from Gijnther, 1887.
(p. (S6); differs from P. spinosus in its onger basal post-temporal spines; shorter oreopercle spine; more raised first supra- ibdominal photophore, somewhat shorter 30st-temporal spine in relation to its base- :o-dorsal spine length, and higher gill •aker count.
Description. D. 13-14; A. 15-17; P. 13- L5; total gill rakers (23) 24-28; vertebrae 33-34.
Medium to large species, seldom ex- ceeding 70 mm SL; trunk broadly tapering; :audal peduncle broad, its depth more :han length of subcaudal photophore group; [lost-temporal spine spinose dorsally, its length substantially less than one-half dis- tance from its base to point of dorsal spine; third basal post-temporal spine long, sec- ond basal spine prominent; dorsal spipe bigh, with flangelike anterior portion not rising sharply from dorsal surface; supra- ibdominal photophores positioned in a itep-wise arrangement, \\'ith first photo- phore raised above other two; abdominal keel scales very spinose, including those ventral to preanal photophores; subcaudal scales either smooth or with single short spine; jaws medium; gill rakers long; pre- opercle spine curves slightly anteriorly; pigment in preservative dark dorsally with very narrow bar extending toward midline; pigment diffuse on trunk.
Distribution (Fig. 63). Known only from the waters around Japan, where it has been taken less abundantly than P. tridentifer.
Polyipnus spinosus GUnther Figure 66
Polyipnus spinosus Giinther, 1887: 170 (holotype BMNH, East Indies; not seisn); Alcock, 1896: 331; 1899: 135; Brauer, 1906: 69 (larvae, fig.) (in part); 1908: 175 (eye muscles); Roxas, 1934: 287; Misra, 1952: 367; Koumans, 1953: 186 (?); Samuel, 1963: 101 (?).
Polyipnus spinosus spinosus Schultz, 1961: 624; 1964: 247.
Species distinction. See P. tridentifer (p. S6) and P. sterope (p. 88).
Description. D. 13-14; A. 15-17; P. 13- 15; total gill rakers 18-21; vertebrae 33-34.
Medium to small species, seldom exceed- ing 70 mm SL; trunk and caudal peduncle broad, its depth greater than or equal to length of subcaudal photophore group; post-temporal spine spinose dorsally, its basal spines reduced; post-temporal spine nearly equal to one-half the distance from its base to dorsal spine; dorsal spine similar to P. sterope; supra-abdominal photophore positioned with first photophore only slightly raised above other two; first two supra-anal photophores slightly raised from third; preopercle spine long, curving an- teriorly; abdominal and preanal keel scales
90
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Figure 67. Polyipnus o/uo/us; R/V HUGH M. SMITH, Cruise 37; Station 43; SL 33 mm.
spinosc; siibcaudal scales with no spines to a single small spine; jaws medium; gill rakers long; pigment in preservative dark dorsally, with narrow dark bar extending toward midline; pigment less dark above anal photophores.
Distribution (Fig. 63). Taken in num- bers off the Philippines and off the south China coast; reported from peninsular India, although these reports may repre- sent P. tridentifer.
Polyipnus o/uo/us n. sp. Figure 67
Holotvpe BCFH 2562; 11° 18'N, 162° 06'E; 12/9/56; R/V HUGH M. SMITH, ciaiise 37; Station 43.
Species distinction. Differs from P. indi- cus and P. nutfingi in its much broader body; post-temporal spine characteristics; triangular abdominal keel scales; supra- anal, supra-abdominal, and subeaudal
photophore characteristics; and posterior extension of dorsal fin rays to end of anal photophores.
Description. D. 14; A. 15; P. 13; total gill rakers 19; vertebrae 33.
Known only from holotype, 33 mm SL; body very broad, narrowing abniptly to short narrow caudal peduncle; body depth 1.3 times into SL; post-temporal spine length more than one-half diameter of orbit, extends to origin of dorsal spine, very spinose dorsally and laterally; frontal ridges almost vertical, spinose; postabdomi- nal spines well developed; abdominal keel scales extend well beyond ventral body margin; these scales sharply triangulate, coming to a single or double point ven- trally; subeaudal scales smooth; first supra- abdominal photophore raised considerably above other two; subeaudal photophores raised well above anals; first supra-anal photophore noticeably lower than second
Marine Hatchetfishes • Baird
91
-igure 68. Polyipnus indicus; after Schultz, 1961.
which is lower than third; anal photophores in two distinct groups; jaws large; teeth small, several recurved ones in upper jaw; vomerine teeth well developed; gill rakers medium, spinose; in preservative pigment somewhat darker dorsally; pigment stri- lations present on tiimk. I Holotype: measurements (mm): SL 32.9, |BD 25.7, JL 08.3, CP 03.9; meristics: GR jl9, D 14, A 15, anal photophores 7; name: from the Hawaiian "oluolu," which means I happy.
Distribution (Fig. 63). Known only from a single capture near the Marshall Islands.
Polyipnus indicus Schultz Figure 68
Pohjipnus indicus Schultz, 1961: 645 (holotype BMNH; off Zanzibar; not seen; paratype USNM 179897; seen); 1964: 241.
Pohjipnus nuttingi: Norman, 1939: 20.
Species distinction. See P. ohioJiis (p.
90); differs from P. nuttingi by its longer,
^sharper post-temporal spines, less extended
abdominal keel scales, less spinose ventral border of lower jaw, photophore character- istics,, presence of spiny subcaudal keel scales, generally lower gill raker number, and shorter post-temporal base-to-dorsal spine length compared to post-temporal spine length.
Description. D. 13-14; A. 15-16 (17); P. (12) 13-14; total gill rakers 20-21 (22); vertebrae 33-34.
Largest specimen less than 55 mm SL; trunk tapering to long caudal peduncle; post-temporal spine long, thin, its length greater than one-half the diameter of orbit; basal post-temporal spine short; preopercle spine long, curving anteriorly; frontal ridges minutely spinose; abdominal keel scales do not extend much below ventral body margin, these scales with multiple spines; subcaudal scales spinose; supra- abdominal photophores in steplike ar- rangement with first photophore raised substantially above second; supra-anal photophores not well separated from anals; jaws large; teeth minute; underside of
92 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
i
Figure 69. Polyipnus nuttingi; modified from Gilbert, 1905.
lower jaw smooth to slightly spinose; gill rakers short to medium with rough spinose internal surfaces; pigment in preservative dark dorsally and dark above anal photo- phores, dark pigment bar extends to mid- line; pigment in myomerelike striations on trunk, with definite pigment spots along posterior midline.
Distribution (Fig. 63). Known from three localities in the Indian Ocean along the east African coast from the equator to 30°S; reports of P. spinosus from this area may represent P. indicus.
Polyipnus nuttingi Gilbert Figure 69
Pohjipmts mittingi Gilbert, 1905: 609 (holotype USNM 51599; Hawaii; seen); Fowler, 1949: 42; Haig, 1955: 321; Schultz, 1961: 640; 1964: 247.
Species distinction. See F. oluolus (p. 90) und P. indicus (p. 91).
Description. D. (12) 13-14; A. 15-16; P. 13-14; total gill rakers (21) 22-24; verte- brae 3.3-34.
Largest specimen less than 65 mm SL; body broad, tapering to long narrow caudal peduncle; post-temporal spine stout, rel- atively short (less than one-half eye diam- eter), slightly spinose dorsally; frontal ridges more vertically oriented than P.
indicus and minutely spinose; preoperck spine short, curving anteriorly; abdominali keel scales with multiple spines; these scales extend well below ventral bod>i( margin; post-temporal spine length less than one-half the distance from its base to point of dorsal blade; subcaudal scale." smooth; supra-abdominal photophores ar- ranged in a straight line, steplike arrange- ment; the three supra-anal photophoret separated slightly but definitely from ana. photophore group; jaws large; teeth mi- nute; undersurface of lower jaw markedl) spinose; gill rakers long, spinose on interna surface; dorsal spine high; pigment ir preservative similar to P. indicus, althougl dorsal pigment bar is longer and broader Distribution (Fig. 63). Known only fron the Hawaiian Islands where it appears tc be an endemic.
Polyipnus laternatus Garman Figure 70
Pohjipnus laternatus Garman, 1899: 238 ( holo type MCZ 27945; off Barbados; seen); Parr 1937: 49; Scbnltz, 1961: 639; 1964: 241.
Polyipnus si)inosus: Brauer, 1906: 121 (in part) Goode and Bean, 1896: 127; Rivero, 1936: 5(
Species distinction. Both P. Jaternatm and P. omphus differ from P. aquavitus anc P. unispinus by their higher meristic
Marine Hatchetfishes • Baird 93
CN O ^0 CO
O P
if o o
LU
o
>
c
Q.
O
0)
94 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Figure 71. Polyipnus omphus; R/V DISCOVERY; Station 5509; SL 43 mm.
counts; shorter preopercle and dorsal spines; l^roader body, photophore and pig- ment characteristics. P. laternatiis differs from P. ompluts in its broader caudal peduncle; shorter, more compact subcaudal photophores, supra-anal and supra-abdom- inal photophore characteristics, slightly shorter preopercle spine, and relatively larger eye.
Description. D. 13-14 (15); A. (15) 16-17; P. 1.3-14; gill rakers (18) 19-22; vertebrae 32-33 (34).
Small to medium size species, rarely ex- ceeding 55 mm SL; body relatively long and narrow, tapering into broad caudal peduncle, its width greater than width of subcaudal photophores; eye large, orbital diameter usually less than six times into SL; post-temporal spine long, thin, its total length variable (usually about one-half the diameter of orbit); dorsal spine short; preopercle spine short, broad, triangulate; abdominal keel scales smooth, not ex- tended far beyond body margin; subcaudal photophores closely allied, little space bc- tvveen each photophore; supra-anal photo- phores raised well above anals, with first supra-anal slightly lower than second; first supra-abdominal photophore raised well above other two; second supra-abdominal even with or lower than third; jaws medium to small; teeth minute; vomerine
and palatine teeth small but prominent; gill rakers long, spinose on inner surface; pigment in preservative dark dorsally, dark i:)igment bar usualK' does not reach mid- line; prominent, dark spots along trunk midline; myomerelike pigment striations dorsally and vertically from midline.
Distribution (Fig. 63). Restricted to the western Atlantic; abundant in the Carib- bean off Venezuela and the central Amer- ican coast, in the lesser Antilles, off Puerto Rico, Cuba, and in the straits of Florida; not reported from the Culf of Mexico oi the east coast of North America.
Polyipnus omphus n. sp. Figure 71
Holotype BMNH: 11° 2rS, 48° 58'E; 8/21/64
R/V DISCOVERY; Station 5509. Pohjipntis Jatcmafiis: Kobayashi, 1963: 179; Kott
haus, 1967: 22 (otoliths, photo. ).
Species distinciion. See P. laternatus (p 92). Roth P. laternatus and P. omphm differ from P. acjuavitus and P. unispinm by their higher meristic counts; shorter preopercle and dorsal spines; broader body photopore and pigment characteristics. P laternatus differs from P. omphus in iti broader caudal peduncle; shorter, mo« compact subcaudal photophores; supra anal and supra-abdominal photophore
Marine Hatchetfishes • Baird
95
CO
o
<
X I— <
<
>
0£
a
D- □
.a
o a.
CN
tN.
96 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Figure 73. Polyipnus unispinus; after Schultz, 1938.
characteristics; slightly shorter preopercle spine; and relatively larger eye. The single specimen from tlie Pacific (SIO 60-236-101) appears slightly different phenotypically from the Indian Ocean forms. These two populations should be further examined when such material is available.
Description. D. 14-15; A. 16; P. 13-14; gill rakers 18-21; vertebrae (33).
Largest specimen less than 50 mm SL. body narrow, tapering into narrow caudal peduncle; its greatest depth less than length of subcaudal photophore group; eye relatively small, orbital diameter greater than six times into SL; post-temporal spine long, about one-half the diameter of orbit (or greater); dorsal spine short; preopercle spine short, narrowly triangulate; abdom- inal keel scales smooth, not extending far beyond body margins; subcaudal ^Dhoto- phores spaced apart (about width of a photophore between them), distance be- tween subcaudal and anal photophores about the same as length of one of the former; supra-anal raised only slightly above anals, with first supra-anal higher than second; supra-abdominal photophores in an oblique straight line; jaws medium; teeth minute; definite vomerine teeth present; gill rakers long, spinose; in pre-
servative, pigment dark dorsally with broad, dark bar reaching to or near mid- line; dark pigment spots along trunk midline with pigment striations radiating from them; dark pigment above ventral photophores.
Holotype: measurements (mm): SL 40.1 BD 20.0, JL 06.1, CP 03.4; mcristics: GE 19, D 14, A 16; anal photophores 11; name:] from the Marathi word "omphus," roughlv translated as "unwanted."
Distribution (Fig. 63). Extremely dis- junct range; known from a few specimeni north of Madagascar in the Indian Ocear and from a single capture in the Centra Pacific north of the Marquesas Islands Additional record: 00^00', 165°42.5'W.
Polyipnus aquavitus n. sp. Figure 72
Holotype ZMUC P20969; 33° 42'S, 151° 51'E 11/13/51; R/V GALATHEA: Station 551.
Species distinction. See P. laternatiis (p 92). P. iniispiniis differs from P. aquavi ills by its longer dorsal and preoperck spines, spinose preopercle and ventral kee plates, shorter subcaudal to anal photo phore distance, longer postabdominal anc
Marine Hatchetfishes • Baiid 97
anal pterygiophore spines, and its narrower trunk and caudal peduncle.
Description. D. (11) 12-13; A. 15-16; P. (12) 1.3-14; gill rakers 12-14; vertebrae (35).
Largest specimen less than 45 mm SL; body narrow, tapering into narrow caudal peduncle; its least depth less than length of subcaudal photophores; post-temporal spine long, length greater than one-half the diameter of orbit; dorsal spine short; preopercle spine short, sharp, length less than one-half the diameter of orbit; second preopercle spine reduced; lateral surface ji preopercle smooth; abdominal keel scales smooth, not extended ventrally; mpra-anal photophores only slightly raised from anals; first supra-anal photophore 'aised above second and third; anal-sub- 3audal photophore distance one-fourth or greater than length of latter; mouth small; :eeth minute; gill rakers short to medium; n preservative, body pigment is dark over abdomen and trunk; pigment often present :n band at base of caudal rays, few dark ,Digment spots along lateral midline.
Holotype: measurements (mm): SL 38.5, BD 17.7, JL 07.0, CP 03.6; meristics: GR 13, D 13, A 15, anal photophores 10; name: TOm the Danish national drink, akvavit.
Disfrihution (Fig. 63). Taken abun- dantly off Sidney, Australia, and known from single captures in the Banda Sea and jetween Tasmania and New Zealand.
°olyipnus unispinus Schultz Figure 73
Foltjipniis unispinus Sclmltz, 1938: 137 (holotvpc USNM 103153; Philippines; seen); 1961: 643; 1964: 247.
Species distinction. See P. latematus (p. )2). Differs from P. aquavitus by its onger dorsal and preopercle spines, spinose ircopercle and ventral keel scales, shorter iubcaudal to anal photophore distance, onger postabdominal and anal pterygio- :)hore spines, and its narrower trunk and audal peduncle.
Description. D. 12-13 (14); A. 1.3-15; P. 12-13; gill rakers (11) 12-14; vertebrae (.3.5-,36).
Small, possibly a "dwarf" species, none yet exceeding 40 mm SL; body narrow, tapering into long narrow trunk and caudal peduncle; post-temporal spine long, almost equal to orbital diameter; dorsal spine long, high; its length about equal to one-half of orbital diameter; preopercle spine long, greater than one-half of orbital diameter; a well-developed second pre- opercle spine usually present; dorsal lateral surface of preopercle spinose; abdominal keel scales spinose ventrally; postabdom- inal and anal pterygiophore spines well developed; supra-anal photophores almost continuous with anals ; first two supra-anals raised markedly above third; distance be- tween subcaudal and anal photophores less or equal to one-fourth the length of the latter; mouth small; teeth minute; vomerine teeth prominent; gill rakers short to medium, well spaced; in preservative pig- ment slightly darker dorsally; abdomen and trunk relatively dark;
Distribution (Fig. 63). Taken in small numbers off the Philippines, and repre- sented by two small samples from the Banda Sea.
Polyipnus triphanos Schultz Figure 74
Polyipnus iriphanos Schultz, 1938: 140 (holotype USNM 103027; Pescador Islands; seen); 1961: 640; 1964: 247.
Species distinction. See P. asteroides (p. 99); differs from P. polli by its higher anal photophore number, lower gill raker and dorsal ray counts, a straight ventral anal photophore margin, and the raised dorsal pigment border above the supra- anal photophores.
Description. D. 11-12; A. 17 (18, 19); P. 1.3-14; gill rakers (15) 16-18 (19); vertebrae (33).
Body broad, tapering into narrow caudal peduncle; its height slightly greater than length of subcaudal photophore group;
98 BuUctin Museum of Cuuiparative Zoology, Vol. 142, No. 1
Figure 74. Polyipnus tripbanos; after Schultz, 1938.
post-temporal spine short, ncedlelike; dor- sal spine short; preopercle spine short, triangulate; abdominal keel scales extend only slightly below ventral body margin; first siipra-anal photophore markedly lower than other two; first supra-abdominal raised above others, second lower than third, jaw medium to small; teeth minute; gill rakers medium, spinose; in preservative, pigment dark dorsally; dark pigment bar reaches almost to midline followed by light stripe reaching towards mid-dorsal line; ventral margin of dark dorsal pigment markedly raised on trunk above supra-anal photophores; pigment spots present on trunk midline, striations not distinct.
Distribution (Fig. 63). Known only from a few captures off the Philippines. Addi- tional Record: 05^^ 01.0'S, 127° ST'E.
Polyipnus polli Schultz Figure 75
Palyijmtis polli Schultz 1961: 635 (holotype MRAC 95092; south east Atlantic; not seen; para type USXM 179878; seen); 1964: 247; Blache, 1964: 71; Backus et al, 1965: 139.
Pohjipmis latcniatus: Norman, 1930: 305; Fowler,
1936: 1208; Poll, 1953: 65. Pohjipniis spinofius: Smith, 1953: 102 (?).
Species distinction. See P. osteroidcs (p. 99) and P. triphanos (p. 97).
Description. D. 14-15 ( 16); A. ( 15) 1&- 17; P. 13-14; gill rakers (20) 21-23; verte- brae 32-33.
Medium to small species, seldom exceeds 50 mm SL; body and trunk broad, narrow- ing abruptly to small, short caudal pedun- cle; post-temporal spines short, needlelike; dorsal spine short; preopercle spine short, triangulate; abdominal keel scales smooth; not extended ventrally; first supra-anal photophore lower than other two; first supra-abdominal raised well above second which is approximately even wnth third; subcaudal photophore group short, about equal to width of dorsal pigment bar at itsi center; body margin below anal photo- phores markedly curved; anal ptery- giophores extend well beyond body margin;' jaws small; teeth minute; gill rakers medium to long, and spinose; in preserva- tive, pigment dark dorsally with pigment
Marine Hatchetfishes • Baird
99
Figure 75. Polyipnus polli; after Norman, 1930.
bar reaching to\\ard midline; ventral border of dorsal pigment in straight line, from lateral photophore to caudal pe- duncle; dark pigment spots on midline and between midline and border of darker dorsal pigment; pigment striations present on trunk.
Distribution (Fig. 63). Restricted to the southeastern Atlantic along the west African coast from the Gulf of Guinea to 10°S latitude.
Polyipnus asteroides Schultz Figure 76
Polyipnus asteroides Schultz, 1938: 138 (holotype USNM; West Indies; not seen): 1961: 640: 1964: 247; Scott, 1963: 1303.
Polyipnus laternatus: Jespersen, 1934: 15.
Species distinction. P. asteroides, P. tri- phanos, and P. poUi differ from P. meteori, P. matsiiharai, P. kiiciensis, and P. rw^geri by their lack of teeth on the posterior vomerine shaft, and by supra-abdominal and supra-anal photophore characteristics; P. asteroides differs from P. polli by its greater number of anal photophores, less
broad trunk, longer subcaudal photophore group, relatively straight anal photophore margin, and attainment of greater size; differs from P. triphanos by its less broad body, higher gill raker and dorsal ray counts, and bodv pigment characteristics.
Description. D. 14-16 (17); A. (15) 16- 17 (18); P. (12) 13-14 (15); gill rakers 20-23 (24); vertebrae 32-33.
Large to giant species, often exceeds 70 mm SL; body relatively broad, tapering evenly to narrow but short caudal pe- duncle, its greatest depth greater than length of subcaudal photophores; post- temporal and dorsal spines short (less than one-fourth eye diameter); preopercle spine very short, triangulate; abdominal keel scales extend slightly below ventral body margin; first supra-anal photophore mark- edly lower than second; first supra-abdom- inal photophore raised well above other two, second and third supra-abdominals usually about same height; jaws medium to large; teeth minute; gill rakers medium to long, spinose on internal surface; in preservative, pigment dark dorsally; dark pigment bar extends toward but never
100 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
in
CN
O
U
<
I
u >
o
c a
o
Marine Hatchetfishes • Baird 101
Figure 77. Polyipnus matsubarai; after Schultz, 1961.
reaches midline; lateral border of dark- dorsal pigment straight from dorsal spine to caudal peduncle; dark pigment spots mark lateral midHne, pigment striations present on trunk.
Distribution (Fig. 63). Restricted to the western North Atlantic; abundant in the Caribbean and Gulf of Mexico from the coast of Venezuela to the straits of Florida; occurs off the outer islands of the West Indies and less abundantly along the east coast of North America; a single capture has been reported as far north as the Gulf 3f Maine.
Polyipnus matsubarai Schultz Figure 77
Polyipnus matsubarai Schultz, 1961: 641 (holo- type USNM 179793; Kumanonada, Japan; seen); 1964: 247.
Polyipnus japonicus Schultz, 1961: 643; 1964: 247.
'^olyipnus asteroidcs: iMatsubara, 1941: 2; 1950: 192.
Species distinction. See P. asteroides (p. 99); differs from P. ruggcri, P. kiwiensis, and P. meteori by its higher gill raker count, long narrow caudal peduncle, and very narrow dorsal pigment bar.
Description. D. 12 (13); A. 16-17; P. (12) (13) 14-16; gill rakers 22-24; verte- brae 33.
Largest specimens have not exceeded 50 mm SL; body broad, tapering into long, relatively narrow caudal peduncle; its length equal to or greater than its greatest depth; post-temporal spine rather long and needlelike, its length about one-fourth the orbital diameter; dorsal spine short; pre- opercle spine short, triangulate; abdominal keel scales smooth, with no \entral ex- tension; first supra-abdominal photophore raised above second which is equal to or slightly raised above third; supra-anal photophores in steplike arrangement, the third being highest; first three anal photo- phores even and parallel to midline; jaws
102 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
"'■ -.IC'.*.'.' • .' ■ ■ N*-*
■ // //'■'/■■/ •'/'••■'■•■'.-''//■.•■■■ .■•■'.■• .-•'^'^Whv^^
Figure 78. Po/yipnus ruggeri; R/V TUl; New Zealand; SL 47 mm.
medium; teeth present on long posterior shaft of vomer lying ventral to parasphe- noid, resulting in three distinet tooth hear- ing areas on the vomer; gill rakers medium; in preservative, dorsal pigment har is ex- tremely narrow and reaches to midline; dorsal pigment horder is hroken hy light stripe hehind pigment har, reaching hroadly to mid-dorsal line; dorsal pigment border raised slightly above supra-anal photophores; small, dark pigment spots mark lateral midline.
Distrihution (Fig. 63). Abundant in the waters off Japan in the North Pacific; a single capture in the mid-North Pacific represents this species.
57'S, 177° 38'E;
Polyipnus ruggeri n. sp. Figure 78
Holotype DMNZ 4670; 31' 7/24/62; R/V TUI.
Species distinction. Differs from the P. osteroides (p. 99) group by dentition and photophore characteristics and from P. matsuharai (p. 101) by dorsal pigment and gill raker characteristics; differs from P. meteori and P. kiwiensis by its dorsal pig-
ment characteristics, higher gill raker count, and photophore patterns; further differs from P. kiwiensis by its smaller, rounder eye, longer, narrower caudal pe- duncle, and lesser distance between frontal crests ( interorbital ) .
Description. D. 12; A. 16-17; P. 15; gill, rakers 18; vertebrae (33).
Largest specimen less than 60 mm SL;] body broad, tapering into somewhat long and narrow caudal peduncle; its length greater than depth; post-temporal spine short, rough surfaced dorsally, less than one-fourth of the diameter of orbit; dorsal spine short, low; preopercle spine triangu- late; eye large, round, its length about equal to width; greatest distance betweer frontal crests (interorbital) less than oi equal to length of subcaudal photophort group; abdominal keel scales not extendec ventrally; supra-abdominal photophores ir essentially straight line, first may be slighth raised above third; first supra-anal photo phore noticeably lo\\'er than second, whicl is lower than third; jaws medium to large' teeth present on posterior vomerine shaft gill rakers medium, slightly spinose; in pre servative, pigment dark dorsally with ;
Marine Hatchetfishes • Baird 103
■igure 79. Polyipnus kiwiensis; R/V TUI; New Zealand; SL 60 mm.
^'ery reduced pigment bar; much reduced light stripe behind bar does not reach mid- ilorsal Hne; ventral border of dorsal pig- nent raised above supra-anal photophores; >mall dark pigment spots present on lateral nidline.
Holotype: measurements (mm): SL 46.8, BD 30.3, JL 09.7, CP 05.5; mcristics: GR L8, D 12, A 17; anal photophores 9; name: lamed in honor of New Zealand's national iport, rugby.
Distribution (Fig. 63). Kno\vn only from I few small captures off Wellington, New Zealand, and west of the Kermadec Islands.
^olyipnus kiwiensis n. sp. Figure 79
Motype DMNZ 4802; 36° 50'S, 176° lO'E- 9/26/62; R/V TUI.
Species distinction. Differs from P. isteroides (p. 99) group by photophore characteristics and teeth on posterior omcrine shaft; from P. matsuharai by gill aker number and dorsal pigment char- icteristics (F. matsuharai, p. 101); from P. ■uggeri (p. 102) by dorsal pigment char- icteristics, eye size, gill raker number,
caudal peduncle, and interorbital crests; differs from P. meteori by its higher gill raker counts, larger eye and mouth, photo- phore and dorsal pigment characteristics.
Description. D. (11) 12; A. 16-17; P. 15-16; gill rakers 16-17; vertebrae (32) 33 (34).
Largest specimens less than 70 mm SL; body broad, tapering rather abruptly into short caudal peduncle; its depth about equal to its length; post-temporal spine short, less than one-fourth the diameter of orbit; dorsal spine short, preopercle spine triangulate; eyes extremely large, their diameter less than seven times into SL; greatest distance between frontal crests (interorbital), greater than length of sub- caudal photophore group; abdominal keel scales not extended ventrally; first and third supra-abdominal photophores about even and raised well above second; first supra-anal photophore noticeably lower than second, which is slightly lower than third; jaws large, broad; teeth well de- veloped on posterior vomerine shaft and lo\\er yAw; gill rakers medium; in pre- servati\e broad, dark, dorsal bar reaches to midline; light stripe posterior to dorsal
104 Bulletin Museum of Coniparatwe Zoology, Vol. 142, No. 1
bar not reaching to mid-dorsal line; ventral border of dark dorsal pigment only slightly raised above supra-anal photophores; small dark pigment spots on lateral midline.
Holotype: measurements (mm): SL 59.5, BD 36.4, JL 14.3, CP 09.7; meristics: GR 17, D. 12, A. 17; anal photophores 10; name: from Kiwi — a New Zealand bird; in the vernacular, a Kiwi is a native of New Zealand.
Distribution (Fig. 63). Taken in moder- ate numbers near Red Mercury Island off the northeastern coast of North Island, New Zealand.
Polyipnus mefeori Kotthaus
Polyipniis meteori Kotthaus, 1967: 27 (liolotype lOES 20; off Seychelles, Indian Ocean; not seen ) .
Species distinction. See P. asteroides (p. 99), P. matsubarai (p. 101), P. ruggeri ( p. 101 ) , and P. kiiciensis ( p. 103 ) .
Description. (From description of holo- type (Kotthaus, 1967) and photograph.) D. 12; A. 16; P. 15; Still raker number 13-15.
Known only from holotype ( SL 37 mm ) ; body broad, tapering to relatively long caudal peduncle (appears shorter than P. matsubarai); post-temporal spine needle- like, about equal to one-fourth the eye diameter; dorsal spine short; abdominal keel scales smooth, not extended ventrally; first and third supra-abdominal photo- phores about even and raised above second; first supra-anal markedly lower than second, which is lower than third; jaws medium; in preservative broad, dark, dorsal bar reaches to midline followed by light stripe which reaches mid-dorsal line: ventral border of dorsal pigment raised considerably above supra-anal photo- phores; dark pigment spots present on midline.
Distribution (Fig. 63). Known only from a specimen taken near the Seychelle Islands in the Indian Ocean; two juvenile Fohjipmis from the east coast of Africa
may represent this species. Note: Key characters checked with holotype through the courtesy of Dr. Verner Larsen, ZMUC.
DISCUSSION AND CONCLUSIONS
Patterns of Distribution
The ecological distinctness of the family and the basic stiuctural modifications in- volved in the peculiar body form were dis- cussed above. Given this basic structural) similarity, the respccti\'e genera have di- verged morphologically and ecologically. This is apparent in the distinctive distribu- tion pattern of each genus and is indicative of the types of distributions to be found in deep-sea fishes.
Fohjipnus. Although Poh/ipntis has the' basic adaptive attributes of a midwater fish, the genus — with the exception of isolated expatriates — is associated with land areas. Land-oriented distributions, have been reported in midwater fishes (Ebeling, 1962; Nafpaktitis, 1968), but these have involved mc>mbers of essentially pelagic genera. Polyipnus is a moderately speciose genus which has adapted solely to land associated environments. While continental slope areas are important, this genus occurs abundantly near oceanic islands well away from continental margins. Depth data are generally sparse, but indi- cate that Polyipnus is found from 50 m to 400 m. The extent of diurnal migration is unknown, although certain species have been reported near the surface at night off Japan (Haneda, 1952). The pseudo- pelagic" environment of this species has not been extensively sampled in most areas. Species ranges are therefore incomplete, and little is known about population struc- ture and vertical distribution. New species can be expected and additional revision will be required as collecting proceeds.
The peculiar distribution of this genus may be related to land-oriented food' chains. There is an extensive amount of literature on the increased productivity associated with land areas and on the
Marine Hatchetfishes • Baird 105
Dccurrence of neritic forms of zooplankton. Foh/ipnus has specialized feeding habits, and two peculiar morphological features may be involved in its adaptation to a specialized niche. These features are the jaw and branchial morphology, and the enlargement of the otic region. An addi- tional indication of biological differences From the other genera is the small number of juN'eniles collected mth the adults. Much remains to be known about the biology and ecology of this genus, as well IS its "pseudopelagic" environment.
Comparison of the essentially tropical and-oriented distribution of Pohjipnus with >ther tropical shore species provides some nteresting parallels. Tropical reef and ihelf fish are diverse in the Indo-west Pacific region, with the Indo-Malayan area 'he most speciose (e.g., Ekman, 1967: 17). riie number of species declines as one proceeds from this area. Wliile present in nany of the islands of outer Polynesia 'Hawaii, Marquesas, Tuamotu archipelago). Few shore species reach the western coast of the Americas. This is attributed to the wide stretch of open water in the eastern Pacific (the zoogeographic east Pacific barrier). Contributions of Indo-west Pacific elements to the tropical Atlantic are re- duced by a similar, although not as re- strictive, central Atlantic barrier, in addi- tion to the African continent ( Briggs, 1960, 1961). The tropical shore fauna is further characterized by its "modernness." It con- sists primarily of the most advanced and latest evolved fishes, with relicts and more primitive groups less well represented. Geographic endemics are common, espe- cially near the more isolated island groups.
The largest number of Pohjipnus species have been collected around the Philippine Islands. Eight of 17 kmown species occur in the tropical west Pacific. Endemics occur in New Zealand, Hawaii, and the Marshall Islands at the limits of the range in the Pacific. Three other species occur in the western Indian Ocean, thus account- ing for 14 of 17 species in the Indo-west
Pacific. No species are reported from the eastern Pacific. Tlie P. spinosiis species complex is not found in the Atlantic; only three species occur there. Two are re- stricted to tropical and temperate America, and one to the west African coast. There are no trans-Panamanian species. Speci- ation tends to be geographic and endemics are numerous. Extensive sympatry be- tween species complexes is rare. Life history features apparently restrict species to land-associated waters. No open-water pelagic populations are known, and bar- riers to gene flo\^' among discontinuous populations appear considerable.
Here, then, is a classic tropical shore distribution in what appears to be the most primitive genus of the family (Ebeling, 1962, indicates some of the same features in MeJamphaes) . Since such a distribution is characteristic of lately evolved groups, it is interesting to speculate on the possible recent origin of Pohjipnus. While primitive maurolicid gonostomatids are identified from the early Tertiary, Pohjipnus as pres- ently defined, is not. (It is not present in Tertiary Tethys or California deposits.) ArgyropeJecus is kno\\'n from the Oligo- cene. Its distribution is worldwide (includ- ing the Tethys fauna), as are a number of gonostomatids (admittedly a different ecology and distribution pattern). Polij- ipnus, ^^'hile primitive in axial skeleton characteristics, is nevertheless highly spe- cialized in the cranial region. These char- acters may be the major adaptive features allowing Pohjipnus access to its specialized niche, resulting in a new adaptive type w^hich possibly arose relatively recently.
Ars.ijropeIecus. Argijwpelecus species are characterized by broad worldwide high seas distributions. Tlie genus is found in all tropical and temperate oceans, and is absent from polar seas. The limits of dis- tribution are bounded approximately by the 5° isothenn at 200 m. Within these broad limits, however, distribution can be quite restricted with the result that world- wide species are broken up into a series of
106
Bulletin Museum of Com par a live Zoology, Vol. 142, No. 1
disjunct populations whicli appear more or less isolated from each other.
In general, species occur vertically over the same depth range wherever they are found. With the exception of A. il,iil.as, Ariiijropclecus species are partial or in- complete diurnal migrators. At night many species ascend to above 300 m, often to about 200 m from their daytime depths of 400-500 m. Catches in the upper 100 m seldom involve large numbers of in- dividuals. A. aculcaiiis is most distinct in its vertical migration, while A. gigas, the deepest living species, migrates very little. Within these broad limits (150-600 m) depth variability is high, indicating considerable microcomplexity (Appendix B). From batluscaphe obser\-ations during the day, Peres (1958) reports A. hemigymnm from 250-600 m, with large concentrations from 400-500 m. Peres' and odier bathyscaphe observations (Drs. R. Rosenblatt, R. Haed- rich, and R. Richards, personal conver- sations) indicate that Argywpelccus species do not school in the classical sense, but are somewhat isolated from one another. Catch data (Table 2.3) show the wide range in size distribution with large catches, an- other indication of nonschooling behavior.
Unlike many midwater fishes, the larvae and ju\'eniles of Argyropeleciis are found in the adult environment (Table 23) (Ahls- trom, 1959). Over the range of a species distribution some gravid females and young ju\eniles were usually found. Large scale expatriation does not appear to be im- portant. Wherever a species is found in an area in numbers it seems to represent a breeding population.
Argiiwpchcus is represented by seven species in three species complexes. Species are moqjhologieally distinct in most cases and, as with Poh/ipnm, broad sympatry within complexes is uncommon. Sympatry is limited to zones of mixing between allo- patric species ranges w hen it occurs within species complexes. Dwarf and giant spe- cies occur. The giant species (A. gigas) is quite restricted in distribution, limited
essentially to zones of water-mass bound- aries. The dwarf species (A. Iwmigymmis), while occurring in the relatively unproduc- tive central water masses, is abundant in highly productive temperate and eastern b()undar\- current waters.
Sfcnwptyx. Sfemoptyx species have broad worldwide pelagic distributions similar to Argyropelectis and with approxi- mately the same geographic limits. The juveniles are found in the adult environ- ment, although larger individuals may be found slightly deeper. There is no indi- cation of expatriation. Sternopty.x is less speeiose than ArgyropeJecus and species distinctions are much less marked. Two of the three species (S. diaphana and S. ohscura) have wide allopatric ranges, with restricted areas of overlap. S. pseudohscura and S. diaphana are broadly sympatric over much of their respective ranges.
Vertically, all species are deep living (500-1500 m) and show little diurnal mi- gration. Variability in catch size ranges^ indicates that Stenwptyx probably does not school.
Geographic Variation
Nhiyr (1963: 333) makes the following points in a discussion of geographic varia- tion: Every population of a species differs from all other populations genetically, and when sufficiently sensitive tests are em- ployed, also biometrically. The degree oB divergence between different populations' of a species ranges from near complete identity to distinctness almost of speciest level. Various characters of a species may and usually do differ independently. The characters of a given population have atl least a partial genetic basis, and in most' cases tend to remain rather constant through the years.
The absence of detectable differences! between horizontally disjunct populations is not necessarily indicative of no popu-. lation differences. In the present study, methods were not particularly sensitive, nor were many characters used. However.
Marine Hatchetfishes • Baud 107
w'liere differences do exist one can delimit populations which, when coupled with dis- tributional data, should add to our under- standing of the environmental and bio- logical factors which are important in restricting species distributions.
Most of the patterns of geographic vari- ation outlined by Mayr are present in Ar<i,yropeJecus and Stcrnoptyx. A. iiigas, ivhich appears to have the most disjunct distribution, displayed no detectable differ- snces between widely separated popu- lations. Characters which arc constant in me species, vary in another. In general, however, population differences were "ound between geographically isolated populations as indicated by horizontal dis- ributions. The statistical characteristics if a population of at least one species (A. iciileotus) remained constant over a period 3f two years. Population limits can be ^uite broad, and usually population bound- iries are correlated with species boundaries in the same area. However, where species' ranges cross major zoogeographic bound- aries, populations on either side of this boundary may be quite distinct (e.g., A. hemiiUjiimus in the North Atlantic).
Population boundaries and morpholog- ical diversification are more obvious in the shallow-dwelling Argyropelcciis than in the deeper-living and nonmigrating Sterno))tyx (Ebeling and Weed, 1963, noted this for Melamphaes and Scopelo- godtis). Distributions of deeper living pe- lagic invertebrates also show this pattern (David, 1963; Grice and liulsemarin, 1967) which is apparently correlated with the decrease in environmental differences with depth between different areas of the oceans. Additional study is necessary to fully appreciate and delimit the population ^structure of both genera.
Distributional Factors
j A considerable amount of literature is 'now available emphasizing the importance of water masses in the distribution of deep- sea organisms (Bieri, 1959; Ebeling, 1962;
Johnson and Brinton, 1963, among others). Discussions of water masses — their forma- tion, location, and identification — are numerous (e.g., Sverdrup et al., 1960; Ebeling, 1962), and each year knowledge of the extent, boundaries, and origin of discontinuities in the oceans increases. Table 24 relates the distribution of the species of Argyropelcciis and Sternoptyx to the various water masses as presently defined. Several interesting conclusions result. One species (A. aculeahis) is limited to central water masses within the great gyre systems of the central oceans. Two species (A. hjchmis, S. oh- scura) are limited to the east Pacific equatorial water mass and the transitional waters at its boundary. As presently de- fined, water masses are too broad to ac- curately describe many distributions; this is particularly so in the Atlantic. While the label "central" or "equatorial" is indic- ative of similarities between water masses in different areas, there can be significant differences in the faunal components (e.g., Indian equatorial versus east Pacific equa- torial; east north Pacific central versus west north Pacific central).
Since the distributions of Argyropelcciis and Sternoptyx are disjunct yet worldwide, a detailed look at the range of each species was made in an attempt to define some of the important distributional parameters. Temperature was considered, as it is often correlated with the distributional limits of fishes. Table 22 represents the tem- perature range of each species within arbitrarily selected depth limits, cor- responding roughly to the depth limits of the species. Tables 20 and 21 illustrate temperature depth profiles from various parts of the ocean \\'here different species occur. As Table 22 illustrates, absolute temperature ranges widely overlap, al- though certain species tend to be high while others are low. Species occur in colder waters in the Pacific in comparison with the Indian and Atlantic oceans. Dis-
108 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Table 20. Temperature-depth profiles for vaiuous parts of the Pacific and Indian oceans.
Depth (m)
4o°N, isonv
Pacific subarctic
34°N, 122°W California
10°N, 120°W Eq Pacific
18°N, 142''W NE Pacific Cent
25'-N, 160°E NW Pacific
|
200 |
9.5 |
9 |
12 |
13 |
17 |
|
|
400 |
6.5 |
6.5 |
9.5 |
8 |
14 |
|
|
600 |
5 |
5 |
6 |
6.5 |
8 |
|
|
800 |
<5 |
<5 |
5.5 |
5 |
5 |
|
|
1000 |
<5 |
<5 |
4.5 |
<5 |
4.5 |
|
|
Depth |
30°S, 85»W |
25°S, 130°W |
30°S, 160°E |
4005, 140''E |
14 |
'S, IIS-E |
|
(m) |
SE Pacific Chile |
SE Pacific |
SW Pacific |
Subantarctic |
Java |
|
|
200 |
11 |
17 |
21 |
8.5 |
14 |
|
|
400 |
7 |
11 |
12 |
7.5 |
9 |
|
|
600 |
6 |
6.5 |
9 |
7 |
||
|
800 |
4.5 |
5.5 |
7 |
5 |
||
|
1000 |
<5 |
5 |
4.5 |
4.5 |
||
|
Depth |
4°N, es^E |
35°S, 65°E |
35°S, 40°E |
12°S, 65°E |
20' |
s inn°F |
|
(m) |
NC Indian 14 |
SC Indian 13 |
SW Indian |
Eq Indian |
SE Indian |
|
|
200 |
17 |
15 |
18 |
|||
|
400 |
11 |
12 |
15 |
10 |
11 |
|
|
600 |
10 |
11 |
13 |
7.5 |
||
|
800 |
8 |
9 |
10 |
6 |
||
|
1000 |
7 |
6 |
7 |
5 |
||
|
1200 |
5.5 |
<5 |
5 |
<5 |
||
|
tiibiitional |
generalities become more defi- |
paring the |
various species. |
A. |
gigas i5 |
nite, however, upon examination of the horizontal di.stribution of the various spe- cies, coupled with the temperature profiles and general hydrographical characteristics. The following pattern emerges by corn-
limited to transitional waters at the bound- aries of tropical central or warm water masses and colder temperate waten where roughly the 5° isotheim is deepei' than 800 m. A. affinis and A. sladeni are
Table 21. Temperature-depth profiles for various parts of the Atlantic Ocean.
|
Depth |
28°N, 87°W |
24°N, 93='W |
IT^N, 60= |
W 16° |
N, 79= |
w |
32° |
N. 15 |
'W |
40°N, 20°W |
40°N 50°W |
|
(m) |
N Gulf Mex |
Gulf Mex |
Venezuela Caribbea |
n |
Trop NE Atl |
N NE Atl |
NW Atl |
||||
|
200 |
14 |
20 |
15 |
18 |
16 |
13 |
16 |
||||
|
400 |
9 |
10 |
9 |
11 |
13 |
12 |
12 |
||||
|
600 |
6 |
8 |
7 |
7.5 |
11 |
10.5 |
8 |
||||
|
800 |
5 |
7 |
5 |
6 |
10 |
9.5 |
6 |
||||
|
1000 |
<5 |
6 |
<5 |
5 |
9 |
8 |
<5 |
||||
|
1200 |
<5 |
5 |
<5 |
<5 |
8 |
7 |
<5 |
||||
|
Gulf of |
8°N, 35°W 16 |
°N, 25°W |
16 |
°S, 60°W |
24°S. 70°W |
24°S, 5°E |
|||||
|
Guinea |
Trap Atl |
Verdes Isl |
SW Atl |
SW Atl |
SE Atl |
||||||
|
200 |
14 |
10 |
15 |
20 |
14 |
16 |
|||||
|
400 |
9 |
8 |
11 |
10 |
10 |
13 |
|||||
|
600 |
6 |
7 |
9 |
6 |
8 |
9 |
|||||
|
800 |
<5 |
6 |
7 |
<5 |
<5 |
5.5 |
|||||
|
1000 |
<5 |
5 |
6 |
<5 |
<5 |
<5 |
|||||
|
1200 |
<5 |
<5 |
5.5 |
<5 |
<5 |
<5 |
Marine Hatchetfishes • Baird
109
Pable 22. Temperature ranges at arbitrarily chosen depths for the various species of
\rgyropelecus and Sternoptyx. Figures were obtained by comparing horizontal species ranges
WITH known temperature-depth profiles over this range.
Species
Depth (ill)
Temperature Range (°C)
\. aculeatus \. olfersi \. Jychnus \. slacJeni \. hemigymnus \. affinis \. gigas i. diaphana ). obscura
>. pseudohscura
200 400 200 400
200 400
200 400
200 400
200 400
400 600 600
SOO
600
800
1000
800 1000
15-21 (all oceans) 10-15 ( all oceans )
12-13 ( all oceans )
9 (Atlantic) 7 (Pacific)
7-12 (Pacific) 6-10 (Pacific)
9-14 (all oceans) 6.5-11 (all oceans)
9-18 (all oceans) 6.5-13 (all oceans)
9-14 ( all oceans ) 6.5-11 (all oceans)
7-12 (all oceans) 6-10 (all oceans)
5-7 (Pacific); 9-11 (Indian & Atlantic) 5-4.5 (Pacific); 7.5-10 (Indian & Atlantic)
7.5-10 ( Indian & Pacific)
6-8 (Indian); 4.5-5.5 (Pacific)
4.5-5 ( Indian & Pacific )
8-5 (all oceans) 4-5 ( all oceans )
estrictcd primarily to the eastern boundary 'urrcnts and areas of upwelling which are •haracterized by cool water between 300 n and 400 m. A. olfersi is restricted to the -vanner areas of subpolar waters character- zed by 12-13° temperatures at 200 m. A. lemiii^ijmmis is excluded only from ec^ua- orial waters, although biometric data ndicates a population structure which :;orresponds to water mass boundaries. A. iciileatus is restricted to wami central kvater masses in areas bounded approxi-
mately by the 15° isotherm at 200 m. A. hjcliniis occurs only in the Pacific equa- torial water mass characterized by cool temperatures between 200 m and 400 m. S. diaphami is excluded from the equatorial water masses only. S. obscura is limited to equatorial water masses and their bound- aries while S. pseudohscura is similar to S. diaphana, although more restricted to tropical and subtropical waters.
The above distribution pattern is a strong argument for the importance of
fABLE 23. Catch statistics for two large samples of Argyropelecus from the North Atlantic.
5IZE class figures REFER TO THE NUxMBER OF INDIVIDUALS IN THE SAMPLE WHOSE STANDARD LENGTH ?"ALLS BETWEEN THE SIZE LIMITS; I.E., THERE WERE TWO INDIVIDUALS OF A. ACULEATUS WHOSE STANDARD
LENGTHS WERE FROM 21 TO 25 MM.
|
Total Catch |
Size Class |
(mm) |
|||||||||
|
Species |
10 |
15 |
20 |
25 |
35 |
45 |
55 |
65 |
90 |
90+ |
|
|
4.. hemigynmus* \. aculeatus** |
240 75 |
9 |
61 |
37 3 |
43 2 |
80 6 |
10 13 |
29 |
15 |
8 |
1 |
* Atlantis II 13, station 1040, 0940-1125, 320-375 m. ** Chain 32, station 859, 0835-1305, 380 m.
110 Bulletin Museum of Coinpaiatice Zoology, Vol. 142, No. 1
Table 24. OccimnExcE of the species of Argy ropelecis and Sternopty.x in various wateh
MASSES. X = TAKEN IN NUMBERS; S = REPORTED IN SMALL NITAIBERS USUALLY NEAR AVATER MASS
BOUNDARIES; O = UNRECORDED; ? = POSSIBLE RECORD.
|
Species |
||||||||||
|
00 |
1 |
on s K |
1 |
00 |
09 3 1 "a u |
i |
-a |
g 3 1 |
1 o •a 3 1 |
|
|
Water Mass |
't |
•< |
< |
< |
■X |
^ |
^ |
W3 |
(yj |
c/i |
|
N Atlantic Central |
X |
X |
X |
o |
X |
X |
X |
X |
o |
X |
|
S Atlantic Central |
X |
X |
X |
o |
X |
X |
s |
X |
o |
X |
|
EN Pacific Central |
o |
X |
X |
o |
X |
o |
o |
X |
o |
X |
|
WN Pacific Central |
o |
o |
s |
o |
o |
s |
o |
X |
o |
0 |
|
ES Pacific Central |
s |
o |
s |
o |
o |
X |
s |
X |
o |
X |
|
^^'S Pacific Central |
o |
o |
s |
o |
X |
X |
o |
0 |
o |
0 |
|
NE Pacific Transitional |
o |
X |
X |
X |
X |
o |
o |
s |
X |
o |
|
SE Pacific Transitional |
X |
X |
X |
X |
X |
s |
s |
X |
X |
o |
|
Pacific Equatorial |
o |
s |
o |
X |
o |
o |
o |
s |
X |
0 |
|
Indian Eciuatorial |
s |
X |
X |
o |
X |
o |
o |
o |
X |
s |
|
Indian Central |
X |
s |
X |
? |
s |
X |
o |
X |
s |
X |
|
Subantartic |
X |
o |
X |
6 |
o |
o |
X |
X |
o |
0 |
|
N Pacific Subarctic |
o |
o |
X |
o |
X |
o |
o |
X |
o |
0 |
temperature in defining distribntions (e.g.. MeGowan. 1960; Xafp^aktitis, 196S). ItYs apparent, however, that absolute tempera- ture values per se are not the sole limiting factor and that eaeh water mass can be defined by a host of other physical and biological factors, all of which ma\ be important in limiting distributions.
A number of recent studies indicate that water masses ha\e a biological identit\- and man\" widely diverse forms are limited to them (Bieri, 1959; Aron, 1962; Brinton. 1962; McGowan, 1963; Fager and Mc- Gowan, 1963). Additional aspects to be considered are the Indrographic features such as boundar\- areas, transitional waters and upwelling areas which pro\ ide further heterogeneity of biological importance. Pelagic hatchetfish distributions are par- ticularly illustrati\e of the biological simi- larities of areas with corresponding h\dro- graphic properties. For instance, eastern boundar\- currents which are quite similar h\drographicall\ (\A'ooster and Reid. 1963) contain the same hatchetfish species wher- ever they are found. The same can be said of central g>re areas or subpolar waters.
While the physical and biological proper ties of the whole water column are im portant in the ecolog\- of a gi%en wate mass, barriers to distribution in hatchet fishes appear to be primarily a function oi the en\ironmental properties over th« depth range of the species. Furthermore the barriers become less marked witl depth so tliat discontinuities at 800-100* m occur less often than those from 200- 400 m. This is reflected in the broac distributions of the deep living Sternoptyxi with three closely related species com pared to the highl\- disjunct and mon speciose Arii.iiropclecii.s w ith its more shal low distribution. This same pattern is evil dent in other deep living forms (Davi 196:3; Ebeling, 1962; Grice, 1963; Grice an Hulsemann, 1967) where life history' features of juveniles or lar\ae do not com plicate the distribution.
i
Zoogeographic Regions
Se\"eral features make the pelagic hatche fishes particularK" well suited to zoogeo graphical studies. Their distributions ar' broad, vet limited to waters of simila
J
Marine Hatchetfishes • Baiid
111
Table 25. Zoogkographic Regions.
Region
Species Assembly
1. E Pacific Equatorial
2. N Pacific Transitional
3. Pacific Subarctic
4. EN Pacific Central
5. WN Pacific Central
6. SE Pacific Transitional
7. Pacific Subantarctic
8. S Pacific Central
9. Indian Equatorial
0. Java-Indonesian
1. Indian Central
2. Tropical Atlantic
3. SE Atlantic Transitional
4. Venezuelan-Caribbean
5. Caribbean-Gulf Central
6. Gulf Peripheral
7. NW Atlantic Pocket
8. WN Atlantic Central
9. EN Atlantic Central
:0. NE Atlantic Subarctic
',1. SW Atlantic Central
;2. W Mediterranean
13. N New Zealand Pocket
A. SE Atlantic Subantarctic
A. lijchnit.s, S. ohscura.
A. affinis, A. hemigyinnus, A. sladeni, S. ohscura, A. lychnus (S. diaphana).
A. sladeni, A. hemigtjmnus (S. diaphana).
A. affinis; A. sladeni, A. hemigtjmnus, S. diaphana.
A. acidealus, A. hemigymnus, S. diaphana, S. pseudohscura.
A. affinis, A. sladeni, A. hemigymnus, A. hjcJinus, A. gigas, S. diaphana (A. olfersi, S. ohscura).
A. olfersi, A. hemigymnus, A. gigas, S. diaphana.
A. aculeatus, A. hemigymnus (S. pseudohscura).
A. affinis, A. sladeni, A. hemigymnus, S. ohscura.
A. affinis, A. sladeni, A. hemigymnus, S. diaphana, S. ohscura.
A. aculeatus, A. hemigymnus, A. gigas, S. diaphana, S. pseudohscura.
A. sladeni, S. diaphana, S. pseudohscura (A. hemigymnus).
A. affinis, A. sladeni, A. gigas, A. diaphana, S. pseudohscura.
A. affinis, A. sladeni, A. hemigymnus, S. diaphana, S. pseudohscura, A. aculeatus.
A. aculeatus, A. hemigymnus, S. diapharm, S. pseudohscura, A. sladeni.
A. affinis, A. sladeni, A. hemigymnus, A. gigas, S. diaphana.
A. affinis, A. sladeni, S. diapliana, S. pseudohscura.
A. aculeatus, A. hemigymnus, S. diaphana (A. gigas).
A. aculeatus, A. hemigymnus, S. diapliana, S. pseudohscura (A. gigas).
A. olfersi, A. hemigymnus, A. gigas, S. diaphana, S. pseudohscura.
A. aculeatus, A. hemigymnus, S. diaphana, S. pseudohscura (A. sladeni).
A. hemigymnus.
A. sladeni.
A. olfersi, A. hemigymnus, S. diaphana.
iydroc;iaphic properties; they are relatively luinerous and easily eaiight; expatriation is iniited; adults and ju\eniles share the same mvironment; they are only partial migrators it best and are thus less affected by seasonal luctuations; and they occur over much of he depth range of the "mesopelagic" en\'i- onment. As we have seen above, the water masses
as presently defined are too broad to ex- plain species distributions as we find them. However, the concept of water masses as bodies of water with similar hydrographic and biological properties is important, and seems to be the most significant one in explaining much of the heterogeneity in the midwater environment. The pelagic hatchetfishes are used in Figure 80 as
112
Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
indicator species of waters of similar properties and their associated discontinui- ties. The results may add to greater appreciation of water masses — both con- ceptuall)' and geographically. Table 25 and Figure 80 list the zoogeographic regions and are also an attempt to indi- cate similarities between regions. The characteristic species assemblages which occur in these regions are listed under the appropriate area. No boundaries were drawn because in most cases they could not be defined. Important isotherms are in- cluded and may serve as rough boundaries. Presently defined water mass boundaries (see Sverdrup et al., 1960) in many cases mark the limits of these areas.
Several attempts at defining oceanic zoogeographic regions have been made (Ebeling, 1962; Clarke, 1966) and Figure 80 represents an additional one. No at- tempt has been made to categorize these regions as primary or secondary, but cer- tainly some regions involve the whole of the mesopelagic environment, while others seem important only at shallower depths. Considerable variation exists in the sharp- ness of the boundaries and, to some extent, in the degree of species overlap. As knowl- edge of the oceans and their fauna in- creases, the nature and extent of these regions and their boundaries will become more apparent.
Areas which are zoogeographic regions and have boundaries which appear throughout the "mesopelagic" environment are the tropical east Pacific, the Indian equatorial region, the northeast Atlantic, and the subantarctic, especially the Pacific portion. There is a wide subtropical belt that is continuous at deeper depths, but is broken into smaller regions above ap- proximately 600 m. The tropical east Pacific has been recognized as a major zoogeographic region, and it seems to have an endemic fauna at all levels (Brin- ton, 1962; Ebeling, 1962; Johnson and Brinton, 1963). The Indian equatorial region, while not as well known, appears
to be somewhat similar to the equatorial Pacific, at least in some species of hatchet- fishes and other fishes as well (Ebeling, 1962; Gibbs and Hmwitz, 1967). The northeast Atlantic is quite different from the western Atlantic in a number of groups (Haffner, 1952; Clarke, 1966; Nafpaktitis, 1968). Additional evidence from this and other studies ( Alvarino, 1965; Gibbs, 1968) indicates that the conver- gence area, especially in the South Pacific, is a major zoogeographical region which may be quite restricted in the South At- lantic and Indian Ocean. The 5° isotherm is much closer to the central water masses at 200 m (Fig. 80), and the distance be- tween the convergences is generally less broad (Sverdrup, 1960).
Regions which are distinctive for the upper 500 m are the warm central water masses of the major gyre systems, and the eastern boundary currents which are cold water areas of transition and upwelling. There are other smaller areas that are im- portant zoogeographically and are faunally similar to the major regions. These include pockets of cold water around the Gulf of Mexico, off South Africa, off the southeast United States (see Haffner on Chauliodm, 1952), in the southern Caribbean and tropical Atlantic, off Java, off New Zealand, and southeast of Hawaii.
The Sternoptychidae are represented by a single species in the western Mediter- ranean, an area which seems distinct from the warmer eastern end. Tliis population is distinct from the North Atlantic one, and this distinction has been documented for other midwater fishes (Marshall, 1963). Hatchetfishes have not been taken in the Red Sea proper (Marshall, 1963) or the Gulf of California (Lavenberg and Fitch. 1966).
Ecological niches and diversity. Speci- ation pattern, distribution, and population structure are three indicators of diversity, niche breadth, and heterogeneity in the mesopelagic environment. While the world- wide midwater environment is heterogene-
Marine Hatchetfishes • Baird
113
CD O CN
E
"5
Q.
|
n |
|
|
(It |
LO |
|
-13 |
CN |
|
"D |
(U |
|
£ |
XI n |
|
a) |
1— |
|
_c |
-n |
|
n |
|
|
o |
|
|
c |
X |
|
o |
|
|
(1> |
|
|
O) |
01 |
|
<u |
|
|
*- |
|
|
u |
t/i |
|
_c |
H |
|
C2 |
|
|
n |
o |
|
^ |
|
|
o |
o |
|
o |
E |
|
o M |
o o |
|
"^ |
|
|
V |
|
|
CO |
n |
|
(U |
E |
114 Bulletin Museum of Comparative ZooJo<i,ij, Vol 142, No. 1
ous, it nevertheless appears to be relatively constant, at least in measurable physical parameters, over broad areas. This is re- flected in the patterns observed in the three indicators mentioned above.
Measurable niche parameters appear quite broad m hatchctfishes in comparison with frcsh\\'ater or shore faunas. Overlap between congeneric species is not ex- tensive, and where it does occur there are usually major morphological or vertical distributional differences. Allopatric ranges are the rule. Congeneric coexistence usually refjuircs either a major shift in depth distri])ution or marked morpho- logical change. Thus Pohjipniis — often geographically isolated and land associated — is the most speciose genus, while Ster- noptyx — the deepest living — is the least. There are indications that this broad niche phenomenon occurs in other groups as well (Marshall, 1963).
ACKNOWLEDGMENTS
I wish to acknowledge with appreciation the following sources of support for my study at Harvard University: a two year Aquatic Sciences Fellowship from the Bureau of Commercial Fisheries; travel funds from the United States — Japan Co- operative Program in Science (GF 147) to Harvard University; and assistance from the Harvard Committee on Evolutionary Biology.
My thanks go to the peopl(> listed below from numerous institutions \\'hose kind cooperation made this study possible.
C. Robins, T. Devany, J. Stieger, Institute of Marine Science, University of Miami
F. Berry, W. Richards, Bureau of Com- mercial Fisheries, Miami
R. Tiavenberg, Los Angeles County Museum
J. Savage, J. Paxton, Hancock Foundation, University of Southern California
K. Suzuki, University of Mie, Japan
R. Rosenblatt, D. Hoese, Scripps Institute of Oceanography
E. Ahlstrom, Bureau of Commercial Fish- eries, La Jolla J. Marr, B. Rothschild, Bureau of Com- mercial Fisheries, Honolulu J. Nielsen, Zoological Museum, Copen- hagen P. Foxton, J. Badcock, National Institute of
Oceanography, England P. Greenwood, G. Palmer, British Museum
(NH), England J. Moreland, Dominion Museum, New Zea- land D. Cohen, N. Gamblin, L. Schultz, U. S.
National Museum, Washington B. Zaranuhee, OccanograjDhic Data C(>nter, Washington
For their understanding discussions and for such gracious hospitality, I am par- ticularly grateful to T. Alie of the Uni- versity of Tokyo, and to E. Bertelsen of the Danish Carlsberg Foundation.
I wish to thank the staff of the Woods Hole Oceanographic Institute, especially Richard Backus and James Craddock, who offered me every encouragement and gave freely of their time and facilities.
My thanks also to Jules Crane of Cer- retos College, who kindly provided fossils from his collection; to Stanley Weitzman of the U. S. National Museum for hiS' thoughts and ideas on the subject of osteol- ogy; and to W. Bossert of Harvard Uni- versity, who introduced me to the worldi of computer science.
I wish to acknowledge the staff of the Museum of Comparative Zoology, par- ticularly Myvanwy Dick; the graduate stu- dents of the fish department whose assist- ance and discussion were invaluable; B. Nafpaktitis and M. Eckardt of the Uni- versity of Southern California; Sharon Horn, and Nancy Smith (illustrations); M. HowbcM-t, who pr()\ided many of the original illustrations through the courtesy of the Woods Hole Oceanographic Insti- tute; K. S. Baird, and Penelope Lasnik fori her many hours before the c^ditorial mast.
Finally, I owe a special debt of gratitude and thanks to Giles W. Mead, my major
Marine Hatchetfishes • Baird
115
professor, for his guidance, encouragement, criticism, and support in making my gradu- :ite years at Harvard a full and rich ex- perience.
SUMMARY
1. The Sternoptychidae are primitive stomiatoid fishes closely related to the Gonostomatidae, but different from them morphologically; most of this difference is related to the peculiar deep body shape of the former.
2. The Sternoptychidae probably arose during the early Tertiary as part of an early stomiatoid radiation. Miocene fossils of Argyropelecus could not be distinguished from their modern coun- terparts, indicating little osteological evolution in this genus since then.
3. The three genera in the family are widely divergent; each has specialized in a separate direction.
4. Polyipniis occurs only in close associ- ation with land. Its pattern of distri- bution and speciation closely parallels that of many tropical shore species.
5. Ariiijropelecus is distributed widely in all tropical and temperate seas. It is a partial migrator not often entering the upper 100 m at night. Adults and juveniles are found in about the same depth range. Argijropelecus inhabits
the upper "mesopelagic" zone (100- 600 m).
6. Sternopfi/x is distributed horizontally within the same limits as Argyropele- cus. It inhabits the lower "mesopelagic" zone (500-1500 m) and does not appear to migrate diurnally.
7. ArgyropeJecus and Sternoptyx species are restricted in distribution, each spe- cies seemingly restricted to waters with similar hydrographic and biological properties.
8. Argyropeleciis is more speciose and shows more morphological variation than Sternoptyx. Species ranges in Sternoptyx are much broader, indicat- ing that barriers to distribution and heterogeneity may be more pronounced in the upper "mesopelagic" than in the lower.
9. Certain species assemblages occur in waters which are hydrographically similar. These assemblages are used to zoogeographically define distinct areas of the world's oceans.
10. Ecological niches in the Sternoptychi- dae are broad over measurable niche parameters. Allopatric species ranges are the rule and, where congeneric sympatric species occur, there is usually a considerable amount of morpho- logical or vertical distinctness.
Appendix A Institutions and Cruises from Which Material Was Examined or Recorded
1. Institutions and their abbreviations.
Collections of T. Abe and O. Suzuki, Tokyo, Japan.
Australian Museum, Sidney, Australia.
Bureau of Commercial Fisheries, Honolulu, Hawaii.
Bureau of Commercial Fisheries, La Jolla, California.
Bureau of Commercial Fisheries, Miami, Florida.
British Museum (Natural History), London, England.
Dominion Museum, Wellington, New Zealand.
International Indian Ocean Expedition.
Biologische Anstalt Helgoland (Meteor Indian Ocean Expedition), Hamburg, Cermany.
Los Angeles County Museum, Los Angeles, California.
Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts.
ABE
AM
BCFH
BCFL
BCFM
BMNH
DMNZ
HOE
lOES
LACM
MCZ
116 Bulletin Museum of Comparatwe Zoology, Vol. 142, No. 1
Appendix A (Continued)
Musee National d'Histoire Naturelle, Paris, France.
Musee Royalc d'Afriqne Central, Tenuren.
National Institute of Oceanography, Sin-rey, England.
Oceanographic Data Center, Washington, D.C.
Scripps Institute of Oceanography, La Jolla, California.
Institute of Marine Science, University of Miami, Florida.
John Hancock Foiuidation, University of Southern CaHfonu'a, Los Angeles, CaHfornia.
U.S. National Museum, Washington, D.C.
Woods Hole Oceanographic Institute, Woods Hole, Massachusetts.
Zoological Museum, Copenhagen, Denmark.
MNHNP
MR AC
NIO
ODC
SIO
UMML
use
USNM WHO I ZMUC
|
2. Institutions and |
Cruises from Wliich Material Was Examined an |
d Recorded. |
|
Institution |
Ship and Cruise |
Location |
|
ABE |
local fishing vessels |
Japan |
|
AM |
holotypes only |
Australia |
|
BCFH |
HUGH M. SMITH 30 |
C Pacific |
|
HUGH M. SMITH 31 |
C Pacific |
|
|
BCFL |
larval and juveniles only |
EN Pacific |
|
BCFM |
GERONIMO |
Tropical Atlantic |
|
SILVER BAY |
Florida, Gulf of Mexico |
|
|
OREGON |
Gulf of Mexico |
|
|
BMNH |
DISCOVERY and others |
Worldwide |
|
DMNZ |
TUI |
New Zealand |
|
IIOE |
ANTON BRUUN 3 and 6 |
Indian Ocean |
|
lOES |
METEOR (holotype description) |
Indian Ocean |
|
LACM |
VELERO |
NE Pacific |
|
MCZ |
ATLANTIS, CAPTAIN BILL III |
N Atlantic |
|
GOSNOLD, BRUUN 13, |
Chile, Tropical Atlantic |
|
|
CHAIN 17-49, miscellaneous collections |
Mediterranean |
|
|
MNHNP |
holotypes only |
Atlantic |
|
MRAC |
holotype |
W African Coast |
|
NIO |
DISCOVERY (1955-1965) |
NE Atlantic, Indian Ocean |
|
ODC |
USNS GILLISS |
N Atlantic, Caribbean |
|
SIO |
COBB 208, 303 |
NE Pacific |
|
BLACK DOUGLASS 203, 303 |
NE Pacific |
|
|
HORIZON |
N Pacific |
|
|
HOLIDAY, TETHYS |
C & S Pacific |
|
|
MONSOON, BAIRD |
C & S Pacific |
|
|
UMML |
GERDA |
Florida, Gulf of Mexico |
|
PILLSBURY |
Caribbean, Gulf of Guinea |
|
|
use |
USNS ELTANNIN |
Subantarctic, Pacific |
|
USNM |
ALBATROSS, OREGON SILVER BAY, COMBAT |
Atlantic, Pacific |
|
WHOI |
CHAIN 60, 72, ATLANTIS II 13 |
Caribbean, Gulf of Mexicq |
|
ATLANTIS II 31 |
N & SW Atlantic |
|
|
ZMUC |
GALATHEA |
World Cruise |
|
THOR and miscellaneous |
N Adantic, |
|
|
collections |
Mediterranean |
Marine Hatchetfishes • Baird
117
Appendix B )epth tables of the species of Argyropelecus axd Sternoptyx. Depth data for these tables
lEPRESEXT MAXIMUM NET DEPTHS ONLY. "CaTCh" REFERS TO THE NUMBER OF HAULS IN WHICH THE
jumber of fish taken is shown horizontally between the figure listed and the previous figure. Depth" refers to the maximum net depths and is recorded vertically between the depth listed
.ND the one listed ABOVE IT. "ZeRO" CATCH REFERS TO THE NUMBER OF NEGATIVE TOWS FOR THAT SPECIES. "+" MEANS GREATER THAN OR EQUAL TO THE VALUE LISTED.
Argyropelecus affinis Night
|
Catch |
Catch |
||||||||||||||||
|
)epth |
0 |
5 |
10 |
20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 |
20 |
50 |
100 |
100+ |
N |
|
100 |
36 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
100 |
20 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
|
200 |
5 |
1 |
0 |
1 |
1 |
0 |
0 |
o |
200 |
10 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
300 |
13 |
7 |
1 |
1 |
0 |
0 |
0 |
9 |
300 |
9 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
|
400 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
400 |
1 |
1 |
0 |
0 |
1 |
0 |
0 |
2 |
|
500 |
4 |
2 |
0 |
0 |
0 |
0 |
1 |
3 |
500 |
2 |
2 |
0 |
1 |
0 |
0 |
0 |
3 |
|
600 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
600 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
700 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
700 |
1 |
2 |
0 |
0 |
0 |
0 |
0 |
2 |
|
000 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
1000 |
n O |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
000 |
0 |
O |
0 |
1 |
0 |
0 |
0 |
4 |
2000 |
0 |
2 |
0 |
0 |
0 |
0 |
0 |
2 |
|
001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
jalifornia, Chile { Pacific )
Gulf of Mexico, Gulf of Guinea, Caribbean
Day
|
Catch |
Cat( |
:h |
|||||||||||||||
|
)epth |
0 |
5 |
10 |
20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 |
20 |
50 |
100 |
100+ |
N |
|
100 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
100 |
9 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
200 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
200 |
10 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
300 |
5 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
300 |
14 |
3 |
0 |
0 |
0 |
0 |
0 |
3 |
|
400 |
19 |
1 |
1 |
0 |
0 |
0 |
0 |
2 |
400 |
7 |
0 |
0 |
0 |
0 |
1 |
0 |
1 |
|
500 |
7 |
2 |
1 |
1 |
0 |
0 |
0 |
4 |
500 |
5 |
0 |
0 |
1 |
2 |
0 |
0 |
3 |
|
600 |
1 |
2 |
0 |
1 |
0 |
0 |
0 |
600 |
4 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
700 |
2 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
700 |
0 |
•"1 o |
0 |
0 |
0 |
0 |
0 |
3 |
|
000 |
5 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
1000 |
10 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
000 |
6 |
4 |
0 |
0 |
0 |
0 |
0 |
4 |
2000 |
19 |
3 |
0 |
0 |
0 |
0 |
0 |
3 |
|
001 + |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
2001 + |
7 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
^ulf of Mexico, Gulf of Guinea, Caribbean
California, Chile (Pacific)
Argyropelecus gigas
|
Night |
Day |
||||||||||||||
|
Catch |
Catch |
||||||||||||||
|
)epth |
0 |
5 |
10 20 |
50 |
100 |
100 + |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
100 |
30 |
0 |
0 0 |
0 |
0 |
0 |
0 |
100 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
|
200 |
26 |
0 |
0 0 |
0 |
0 |
0 |
0 |
200 |
7 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
300 |
19 |
0 |
0 0 |
0 |
0 |
0 |
0 |
300 |
10 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
400 |
9 |
0 |
0 0 |
0 |
0 |
0 |
0 |
400 |
34 |
0 |
1 0 |
0 |
0 |
0 |
1 |
|
500 |
21 |
1 |
0 0 |
0 |
0 |
0 |
1 |
500 |
25 |
0 |
1 0 |
0 |
0 |
0 |
1 |
|
600 |
7 |
0 |
1 0 |
0 |
1 |
0 |
2 |
600 |
12 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
700 |
4 |
1 |
0 0 |
0 |
0 |
0 |
1 |
700 |
8 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
000 |
14 |
0 |
0 0 |
0 |
0 |
0 |
0 |
1000 |
20 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
:000 |
4 |
1 |
0 0 |
0 |
0 |
0 |
1 |
2000 |
5 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
itlantic |
Atlantic |
118 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Appendix B ( Continued )
AriiUroi)clcciis hcmigtjtnmifi (Form A) Night
|
<:at(.li |
Catch |
||||||||||||||
|
Dcplli |
0 |
5 |
1 0 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
100 |
20 |
1 |
0 0 |
0 |
0 |
1 |
100 |
23 |
7 |
1 1 |
0 |
0 |
0 |
9 |
|
|
200 |
9 |
1 |
0 0 |
0 |
0 |
1 |
200 |
29 |
0 |
0 2 |
0 |
0 |
0 |
2 |
|
|
300 |
11 |
0 |
0 0 |
0 |
0 |
0 |
300 |
20 |
0 |
0 0 |
0 |
0 |
0 |
2 |
|
|
400 |
2 |
0 |
0 0 |
0 |
0 |
0 |
400 |
8 |
1 |
1 0 |
2 |
0 |
0 |
4 |
|
|
500 |
4 |
0 |
0 0 |
1 |
0 |
1 |
500 |
23 |
0 |
1 1 |
0 |
0 |
0 |
2 |
|
|
600 |
0 |
0 |
0 0 |
0 |
0 |
0 |
600 |
14 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
|
700 |
3 |
0 |
0 0 |
0 |
0 |
0 |
0 |
700 |
5 |
3 |
1 0 |
0 |
0 |
0 |
4 |
|
1000 |
0 |
0 0 |
0 |
0 |
0 |
0 |
1000 |
22 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
|
2000 |
2 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2000 |
5 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
0 |
0 0 |
0 |
{) |
0 |
0 |
|
Carihht' |
an anc |
1 c;uii |
of Mexico |
North Atl |
antic |
(see |
Day) |
||||||||
|
Catch |
Catch |
||||||||||||||
|
n.ptii |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
J 00 |
5 |
0 |
0 0 |
0 |
0 |
0 |
100 |
35 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
|
200 |
3 |
0 |
0 0 |
0 |
0 |
0 |
200 |
7 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
|
300 |
4 |
1 |
0 0 |
0 |
0 |
1 |
300 |
7 |
7 |
2 0 |
1 |
0 |
10 |
||
|
400 |
16 |
2 |
1 1 |
1 |
0 |
5 |
400 |
0 |
0 |
0 0 |
0 |
0 |
0 |
||
|
500 |
3 |
0 |
0 1 |
0 |
0 |
1 |
500 |
7 |
0 |
0 0 |
0 |
0 |
0 |
||
|
fiOO |
2 |
0 |
0 0 |
0 |
0 |
0 |
600 |
0 |
0 |
0 0 |
0 |
0 |
0 |
||
|
TOO |
5 |
2 |
0 0 |
0 |
0 |
2 |
700 |
0 |
0 |
0 0 |
0 |
0 |
0 |
||
|
1000 |
8 |
2 |
1 0 |
0 |
0 |
3 |
1000 |
0 |
0 |
0 0 |
0 |
{) |
0 |
||
|
2000 |
11 |
2 |
2 0 |
0 |
0 |
4 |
2000 |
3 |
1 |
0 0 |
0 |
0 |
1 |
||
|
2001 + |
6 |
1 |
1 0 |
0 |
0 |
2 |
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
Southern Ocean (Pacific)
Pacific ( California )
ArgtjrojK'lccu.s hcmigyniuus (Form A)
Dav
|
Depth |
0 |
5 |
Catch 10 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
Catc 10 |
■h 20 |
50 |
100 |
100 t |
N |
|
100 |
2 |
0 |
1 0 |
0 |
0 |
0 |
1 |
100 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
200 |
8 |
0 |
0 0 |
0 |
1 |
0 |
1 |
200 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
300 |
10 |
0 |
0 1 |
1 |
0 |
0 |
2 |
300 |
2 |
1 |
0 |
2 |
0 |
0 |
0 |
3 |
|
400 |
12 |
■J |
1 2 |
2 |
0 |
2 |
10 |
400 |
11 |
3 |
3 |
0 |
0 |
0 |
9 |
|
|
500 |
15 |
7 |
3 4 |
3 |
3 |
1 |
21 |
500 |
4 |
3 |
1 |
1 |
1 |
0 |
0 |
6 |
|
600 |
10 |
2 |
1 3 |
2 |
1 |
0 |
9 |
600 |
2 |
1 |
0 |
0 |
1 |
0 |
0 |
1 |
|
700 |
4 |
0 |
0 0 |
2 |
1 |
0 |
3 |
700 |
3 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
|
1000 |
25 |
5 |
0 1 |
0 |
0 |
0 |
3 |
1000 |
6 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
|
2000 |
6 |
0 |
1 0 |
0 |
0 |
0 |
1 |
2000 |
10 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
2001+ 0 North Athrntic |
0 0 0 (20°-40°N, 5 |
0 0 =-70°\V) |
0 |
0 |
2001 + Culf of M |
1 exico |
0 c:a |
0 rihlx' |
0 an |
0 |
0 |
0 |
0 |
|||
|
Dei.th |
0 |
5 |
Catch 10 20 |
50 |
100 |
1 00 }- |
X |
De|-,th |
0 |
5 |
Catt 10 |
h 20 |
50 |
100 |
100+ |
N |
|
100 |
0 |
1 |
0 0 |
0 |
0 |
1 |
100 |
8 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
200 |
1 |
1 |
0 0 |
0 |
0 |
1 |
200 |
8 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
300 |
3 |
1 |
0 0 |
0 |
0 |
1 |
300 |
11 |
1 |
0 |
1 |
0 |
0 |
0 |
2 |
|
|
400 |
0 |
3 |
0 0 |
0 |
0 |
3 |
400 |
4 |
0 |
1 |
0 |
0 |
0 |
0 |
1 |
|
|
500 |
0 |
2 |
0 0 |
0 |
0 |
2 |
500 |
3 |
2 |
1 |
0 |
0 |
0 |
0 |
3 |
|
|
600 |
1 |
1 |
0 0 |
0 |
1 |
2 |
600 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
700 |
0 |
1 |
0 0 |
0 |
0 |
1 |
700 |
1 |
2 |
0 |
0 |
0 |
0 |
0 |
2 |
|
|
1000 |
0 |
4 |
2 1 |
1 |
0 |
8 |
1000 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
2000 |
4 |
6 |
1 0 |
1 |
0 |
8 |
2000 |
3 |
7 |
0 |
0 |
0 |
0 |
0 |
7 |
|
|
2001 + |
3 |
3 |
0 0 |
0 |
0 |
0 |
3 |
2001 + |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
Southern Ocean ( Pacific )
Pacific ( California )
Marine Hatchetfishes • Baird 119
Appendix B ( Continued )
Argtjropclccu.s hcniii^ynimi.s (Form B) Night
|
Depth |
0 |
5 |
Catch 10 20 |
50 ] |
00 |
100+ |
N |
Depth |
0 |
5 |
Catch 10 20 |
50 |
100 |
100+ |
N |
||
|
100 |
21 |
0 |
0 0 |
0 |
0 |
0 |
0 |
100 |
27 |
3 |
1 0 |
1 |
0 |
0 |
5 |
||
|
200 |
10 |
0 |
0 0 |
0 |
0 |
0 |
0 |
200 |
21 |
6 |
2 2 |
0 |
0 |
0 |
4 |
||
|
300 |
9 |
2 |
0 0 |
0 |
0 |
0 |
2 |
300 |
17 |
1 |
0 0 |
2 |
0 |
0 |
2 |
||
|
400 |
2 |
0 |
0 0 |
0 |
0 |
0 |
0 |
400 |
7 |
1 |
2 0 |
0 |
1 |
0 |
4 |
||
|
500 |
2 |
0 0 |
0 |
0 |
0 |
2 |
500 |
2 |
1 |
1 1 |
0 |
0 |
0 |
3 |
|||
|
600 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
600 |
0 |
2 |
4 0 |
1 |
0 |
0 |
5 |
||
|
700 |
3 |
0 |
0 0 |
0 |
0 |
0 |
0 |
700 |
4 |
0 |
1 0 |
0 |
0 |
0 |
1 |
||
|
1000 |
3 |
0 |
0 0 |
0 |
0 |
0 |
0 |
1000 |
14 |
•1 |
0 3 |
1 |
0 |
0 |
6 |
||
|
2000 |
2 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2000 |
1 |
0 |
1 0 |
0 |
0 |
0 |
1 |
||
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
||
|
Gull' of M |
exico, |
Ca |
libbean |
North Atlantic |
(see |
Day) |
|||||||||||
|
Depth |
0 |
5 |
Catch 10 20 |
50 100 |
100+ |
N |
|||||||||||
|
100 |
5 |
0 |
0 |
0 |
0 0 |
0 |
0 |
||||||||||
|
200 |
2 |
1 |
0 |
0 |
0 0 |
1 |
|||||||||||
|
300 |
3 |
2 |
0 |
0 |
0 0 |
2 |
|||||||||||
|
400 |
8 |
8 |
1 |
2 |
1 1 |
13 |
|||||||||||
|
500 |
0 |
1 |
0 |
1 |
0 2 |
4 |
|||||||||||
|
600 |
1 |
0 |
0 |
1 |
0 0 |
1 |
|||||||||||
|
700 |
4 |
1 |
2 |
0 |
0 0 |
0 |
3 |
||||||||||
|
1000 |
3 |
4 |
2 |
1 |
1 0 |
0 |
8 |
||||||||||
|
2000 |
5 |
9 |
1 |
0 |
0 0 |
0 |
10 |
||||||||||
|
2001 + |
5 |
0 |
2 |
1 |
0 0 |
0 |
3 |
Southern Ocean (Pacific)
Argtjropc'Icciis licinigytniius (Form B'
Day
|
Depth |
0 |
5 |
Catch 10 20 |
50 |
00 |
100+ |
N |
Depth |
0 |
5 |
Catch 10 20 |
50 |
100 |
100+ |
N |
|||
|
100 |
4 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
100 |
1 |
0 |
0 0 |
0 |
0 |
0 |
0 |
||
|
200 |
10 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
200 |
2 |
0 |
0 0 |
0 |
0 |
0 |
0 |
||
|
300 |
14 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
300 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|||
|
400 |
22 |
0 |
0 |
0 |
0 |
0 |
1 |
400 |
3 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|||
|
500 |
31 |
1 |
1 |
0 |
0 |
0 |
0 |
500 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|||
|
fiOO |
14 |
3 |
o |
0 |
0 |
0 |
0 |
6 |
600 |
2 |
0 |
0 0 |
0 |
1 |
0 |
1 |
||
|
700 |
11 |
0 |
0 |
1 |
0 |
0 |
0 |
700 |
1 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|||
|
1000 |
29 |
1 |
0 |
0 |
0 |
0 |
0 |
1 ^ |
1000 |
9 |
1 |
0 0 |
0 |
0 |
0 |
1 |
||
|
2000 |
7 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
2000 |
10 |
2 |
0 0 |
0 |
0 |
0 |
2 |
||
|
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
2001 + |
6 |
0 |
0 0 |
0 |
0 |
0 |
0 |
||
|
N Atlan |
tic (20°-40°N, 5° 70° W) |
Southern |
Ocean (Pacific) |
|||||||||||||||
|
Depth |
0 |
5 |
Catch 10 20 |
50 100 |
100 + |
\ |
||||||||||||
|
100 |
1 |
0 |
0 |
0 |
0 0 |
0 |
0 |
|||||||||||
|
200 |
0 |
0 |
0 |
0 |
0 0 |
0 |
0 |
|||||||||||
|
300 |
5 |
0 |
0 |
0 |
0 0 |
0 |
0 |
|||||||||||
|
400 |
20 |
0 |
0 |
0 |
0 0 |
0 |
0 |
|||||||||||
|
500 |
10 |
0 |
0 |
0 |
0 0 |
0 |
0 |
|||||||||||
|
600 |
3 |
1 |
0 |
0 |
0 0 |
0 |
1 |
|||||||||||
|
700 |
4 |
0 |
0 |
0 |
0 0 |
0 |
0 |
|||||||||||
|
1000 |
7 |
0 |
0 |
0 |
0 0 |
0 |
0 |
|||||||||||
|
2000 |
10 |
0 |
0 |
0 |
0 0 |
0 |
0 |
|||||||||||
|
2001 + |
1 |
0 |
0 |
0 |
0 0 |
0 |
0 |
Gulf of Mexico, Caribbean
120 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
|
Cat |
:h |
|||||||
|
Depth |
0 |
5 |
10 |
20 |
50 |
100 |
100+ |
N |
|
100 |
25 |
4 |
2 |
1 |
0 |
0 |
0 |
7 |
|
200 |
18 |
10 |
2 |
0 |
1 |
0 |
0 |
13 |
|
300 |
21 |
0 |
1 |
0 |
0 |
0 |
0 |
1 |
|
400 |
8 |
3 |
1 |
0 |
0 |
0 |
0 |
4 |
|
500 |
17 |
7 |
1 |
0 |
0 |
0 |
0 |
8 |
|
600 |
8 |
6 |
0 |
0 |
0 |
0 |
0 |
6 |
|
700 |
0 |
7 |
2 |
0 |
0 |
0 |
0 |
9 |
|
1000 |
20 |
3 |
0 |
0 |
0 |
0 |
0 |
3 |
|
2000 |
5 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
North Atl |
antic |
Appendix B ( Continued ) Argyropelecus aculcatus
Night
Catch
|
Depth |
0 |
5 |
10 |
20 |
50 |
100 |
100+ |
N |
|
100 |
8 |
6 |
1 |
0 |
2 |
0 |
0 |
9 |
|
200 |
0 |
3 |
3 |
1 |
1 |
2 |
0 |
10 |
|
300 |
5 |
3 |
0 |
2 |
0 |
0 |
0 |
5 |
|
400 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
500 |
2 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
|
600 |
0 |
0 |
0 |
{) |
0 |
0 |
0 |
0 |
|
700 |
2 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
1000 |
2 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
|
2000 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
Catch
Dav
Cailf of Mexico, Caribbean
Catch
|
Depth |
0 |
5 |
10 20 |
50 |
100 100+ |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
|
100 |
4 |
0 |
0 0 |
0 |
0 |
0 |
0 |
100 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
200 |
7 |
2 |
1 0 |
0 |
0 |
0 |
3 |
200 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
300 |
9 |
5 |
0 0 |
0 |
0 |
0 |
5 |
300 |
0 |
3 |
1 0 |
0 |
0 |
0 |
4 |
|
400 |
6 |
8 |
4 0 |
4 |
1 |
0 |
17 |
400 |
4 |
3 |
2 0 |
7 |
0 |
0 |
12 |
|
500 |
10 |
17 |
1 4 |
2 |
0 |
0 |
24 |
500 |
3 |
3 |
0 1 |
2 |
0 |
0 |
6 |
|
600 |
10 |
8 |
0 0 |
2 |
0 |
0 |
10 |
600 |
1 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
700 |
9 |
4 |
0 0 |
0 |
0 |
0 |
4 |
700 |
1 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
1000 |
24 |
6 |
0 1 |
0 |
0 |
0 |
7 |
1000 |
2 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
2000 |
6 |
0 |
0 1 |
0 |
0 |
0 |
1 |
2000 |
4 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
North Atlantic |
(20° |
-40°N; 5 |
°-70°W) |
Gulf of M |
exico |
, Caribbean |
|||||||||
|
Argyropelecus olfersi |
|||||||||||||||
|
Night |
Dav |
||||||||||||||
|
Catch |
Catch |
||||||||||||||
|
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
100 |
34 |
1 |
0 0 |
0 |
0 |
0 |
1 |
100 |
4 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
200 |
23 |
4 |
2 0 |
0 |
0 |
0 |
6 |
200 |
8 |
0 |
1 0 |
0 |
0 |
0 |
1 |
|
300 |
21 |
2 |
1 0 |
0 |
0 |
0 |
3 |
300 |
13 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
400 |
22 |
6 |
0 0 |
0 |
0 |
0 |
6 |
400 |
20 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
500 |
21 |
2 |
1 0 |
0 |
0 |
0 |
1 |
500 |
26 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
600 |
7 |
2 |
0 0 |
0 |
0 |
0 |
2 |
600 |
13 |
0 |
1 0 |
0 |
0 |
0 |
1 |
|
700 |
4 |
0 |
0 0 |
0 |
0 |
0 |
0 |
700 |
9 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
1 000 |
24 |
1 |
0 0 |
0 |
0 |
0 |
1 |
1000 |
26 |
0 0 |
0 |
0 |
0 |
3 |
|
|
2000 |
17 |
1 |
0 0 |
0 |
0 |
0 |
1 |
2000 |
13 |
4 |
0 0 |
0 |
0 |
0 |
4 |
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
Atlantic |
and S( |
3 II the |
rn Ocean |
Atlantic and Southern Ocean |
|||||||||||
|
A |
rgijrapclecus |
sladcni |
1 |
||||||||||||
|
Night |
|||||||||||||||
|
Catch |
Catch |
||||||||||||||
|
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 100+ |
N |
|
|
100 |
19 |
1 |
0 1 |
0 |
0 |
0 |
2 |
100 |
35 |
3 |
0 0 |
0 |
0 |
0 |
3 |
|
200 |
10 |
0 |
0 0 |
0 |
0 |
0 |
0 |
200 |
6 |
1 |
2 0 |
0 |
0 |
0 |
3 |
|
300 |
3 |
5 |
1 1 |
0 |
0 |
0 |
7 |
300 |
7 |
10 |
2 0 |
0 |
0 |
0 |
12 |
|
400 |
0 |
1 |
0 0 |
0 |
0 |
0 |
1 |
400 |
13 |
6 |
1 0 |
1 |
0 |
0 |
8 |
|
500 |
2 |
0 0 |
0 |
0 |
0 |
2 |
500 |
7 |
4 |
0 0 |
0 |
0 |
0 |
4 |
|
|
600 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
600 |
2 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
700 |
2 |
1 |
0 0 |
0 |
0 |
0 |
1 |
700 |
7 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
1000 |
3 |
0 |
0 0 |
0 |
0 |
0 |
1 |
1000 |
11 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
2000 |
0 |
2 |
0 0 |
0 |
0 |
0 |
2 |
2000 |
2 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
Gulf of Mexico, Caribbean, Gulf of Guinea
Pacific ( Cah'fornia ,
Marine Hatchetfishes • Baird
121
Appendix B (Continued)
Depth
Catch
10 20 50 100 100+ N
Day
Depth
Catch
10 20 50 100 100+ N
|
100 |
1 |
0 |
0 0 |
0 |
0 |
0 |
0 |
100 |
8 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
200 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
200 |
7 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
300 |
4 |
2 |
0 0 |
0 |
0 |
0 |
2 |
300 |
6 |
8 |
0 0 |
0 |
0 |
0 |
8 |
|
400 |
8 |
6 |
3 2 |
1 |
0 |
0 |
12 |
400 |
4 |
0 |
0 1 |
0 |
0 |
0 |
1 |
|
500 |
6 |
3 |
1 0 |
0 |
0 |
0 |
4 |
500 |
3 |
3 |
0 1 |
0 |
0 |
0 |
4 |
|
600 |
7 |
1 |
1 0 |
0 |
0 |
0 |
2 |
600 |
1 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
700 |
1 |
3 |
0 0 |
0 |
0 |
0 |
3 |
700 |
1 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
1000 |
3 |
4 |
0 0 |
0 |
0 |
0 |
4 |
1000 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
2000 |
7 |
3 |
0 0 |
0 |
0 |
0 |
3 |
2000 |
8 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
2001 + |
0 |
1 |
0 0 |
0 |
0 |
0 |
1 |
2001 + |
1 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
Gulf of M |
exico |
Caribbean, Gulf of Guinea |
Pacific |
(California) |
|||||||||||
|
Argyropelecus |
lychnus |
||||||||||||||
|
Night |
Day |
||||||||||||||
|
Catch |
Catch |
||||||||||||||
|
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
100 |
42 |
2 |
0 0 |
0 |
0 |
0 |
2 |
100 |
10 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
200 |
10 |
3 |
0 1 |
0 |
0 |
0 |
4 |
200 |
10 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
300 |
17 |
4 |
2 1 |
0 |
0 |
0 |
7 |
300 |
13 |
3 |
1 0 |
1 |
0 |
0 |
5 |
|
400 |
23 |
9 |
0 4 |
0 |
0 |
0 |
13 |
400 |
9 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
500 |
7 |
2 |
2 0 |
0 |
0 |
0 |
4 |
500 |
6 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
600 |
2 |
0 |
0 0 |
0 |
0 |
0 |
0 |
600 |
4 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
700 |
7 |
0 |
0 0 |
0 |
0 |
0 |
0 |
700 |
2 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
1000 |
11 |
0 |
0 0 |
0 |
0 |
0 |
0 |
1000 |
10 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
2000 |
18 |
1 |
0 0 |
0 |
0 |
0 |
1 |
2000 |
21 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
2001 + |
8 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2001 + |
7 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
California |
, Chile, Tropical Pacific |
California, Chile, T |
ropical Pacific |
||||||||||||
|
Sternoptyx o |
hscura |
||||||||||||||
|
Night |
Dav |
||||||||||||||
|
Catch |
Catch |
||||||||||||||
|
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
100 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
100 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
200 |
8 |
2 |
0 0 |
0 |
0 |
0 |
2 |
200 |
2 |
0 |
0 1 |
0 |
0 |
0 |
1 |
|
300 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
300 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
400 |
4 |
1 |
0 0 |
0 |
0 |
0 |
I |
400 |
1 |
3 |
0 0 |
0 |
0 |
0 |
3 |
|
500 |
2 |
1 |
0 0 |
0 |
0 |
0 |
1 |
500 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
600 |
4 |
2 |
0 0 |
0 |
0 |
0 |
2 |
600 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
700 |
4 |
1 |
1 0 |
0 |
0 |
0 |
2 |
700 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
1000 |
5 |
5 |
0 1 |
0 |
0 |
0 |
6 |
1000 |
4 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
2000 |
4 |
1 |
0 0 |
0 |
0 |
0 |
1 |
2000 |
7 |
2 |
1 0 |
0 |
0 |
0 |
|
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 ^ |
2001 + |
4 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
Indian 0( |
;ean |
Indian |
Ocean |
||||||||||||
|
Stciuoptyx pseudobscura |
|||||||||||||||
|
Night |
|||||||||||||||
|
Catch |
Catch |
||||||||||||||
|
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
10 20 |
50 |
100 100+ |
N |
|
|
100 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
100 |
50 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
200 |
10 |
0 |
0 0 |
0 |
0 |
0 |
0 |
20O |
36 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
300 |
16 |
1 |
0 0 |
0 |
0 |
0 |
1 |
300 |
30 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
400 |
5 |
0 |
0 0 |
0 |
0 |
0 |
0 |
400 |
9 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
500 |
3 |
0 |
0 0 |
0 |
0 |
0 |
0 |
500 |
24 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
600 |
6 |
0 |
0 0 |
0 |
0 |
0 |
0 |
600 |
9 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
700 |
6 |
0 |
0 0 |
0 |
0 |
0 |
0 |
700 |
7 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
1000 |
9 |
2 |
0 0 |
0 |
0 |
0 |
2 |
1000 |
12 |
4 |
1 0 |
0 |
0 |
0 |
5 |
|
2000 |
5 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2000 |
2 |
1 |
1 0 |
1 |
0 |
0 |
3 |
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
Indian Ocean |
North Atlantic and Gulf of C |
.uinea |
122 Bulletin Museum of Cumparative Zoology, Vol. 142, No. 1
Appendix B (Continued)
Dav
|
Depth |
0 |
5 |
Catt 10 |
h 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
Gate 10 |
1 20 |
50 |
100 |
100 + |
N |
|
100 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
100 |
4 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
200 |
■1 o |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
200 |
7 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
300 |
10 |
2 |
0 |
0 |
0 |
0 |
0 |
2 |
300 |
15 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
400 |
4 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
400 |
39 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
500 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
500 |
35 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
600 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
600 |
14 |
1 |
0 |
0 |
0 |
0 |
1 |
|
|
700 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
700 |
12 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
1000 |
5 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
1000 |
16 |
7 |
2 |
1 |
0 |
0 |
10 |
|
|
2000 |
8 |
2 |
0 |
0 |
0 |
0 |
0 |
2 |
2000 |
12 |
0 |
0 |
0 |
0 |
3 |
||
|
2001 + |
4 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
2001 + |
0 |
1 |
0 |
0 |
0 |
0 |
1 |
|
|
Indian Ocean |
North Atlantie |
and Gulf |
of G |
ninea |
|||||||||||||
|
Slcrnuplijx (lidpliaiia |
|||||||||||||||||
|
Dcptli |
0 |
5 |
Catt 10 |
h 20 |
50 |
100 |
100+ |
Day N |
Depth |
0 |
5 |
Gate 10 |
h 20 |
50 |
1(10 |
100 t |
N |
|
100 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
100 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
200 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
200 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
300 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
300 |
4 |
0 |
0 |
0 |
0 |
0 |
||
|
400 |
4 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
400 |
14 |
2 |
0 |
0 |
0 |
2 |
||
|
500 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
500 |
5 |
4 |
0 |
0 |
0 |
4 |
||
|
600 |
0 |
1 |
0 |
0 |
0 |
1 |
0 |
2 |
600 |
1 |
0 |
0 |
1 |
0 |
1 |
||
|
700 |
1 |
J |
0 |
0 |
1 |
0 |
0 |
2 |
700 |
0 |
1 |
0 |
0 |
0 |
1 |
||
|
1000 |
0 |
0 |
0 |
0 |
1 |
0 |
4 |
1000 |
2 |
0 |
0 |
1 |
0 |
1 |
|||
|
2000 |
2 |
1 |
1 |
1 |
1 |
0 |
0 |
4 |
2000 |
1 |
3 |
0 |
0 |
0 |
3 |
||
|
2001 + |
0 |
0 |
0 |
0 |
1 |
0 |
0 |
1 |
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
Ciilf of G |
lint'a |
GuLf of Mexico |
and Cari |
)l)eai |
1 |
||||||||||||
|
Depth |
0 |
5 |
Gate 10 |
h 20 |
50 |
100 |
100+ |
N |
Depth |
0 |
5 |
Gate 10 |
h 20 |
50 |
100 |
100 \- |
N |
|
100 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
100 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
||
|
200 |
7 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
200 |
3 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
300 |
9 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
300 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
400 |
19 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
400 |
3 |
1 |
0 |
0 |
0 |
0 |
1 |
|
|
500 |
20 |
4 |
1 |
0 |
0 |
0 |
0 |
5 |
500 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
600 |
7 |
0 |
1 |
2 |
1 |
0 |
0 |
4 |
600 |
0 |
0 |
u |
0 |
0 |
0 |
0 |
|
|
700 |
4 |
3 |
0 |
0 |
1 |
0 |
0 |
4 |
700 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
1000 |
8 |
4 |
2 |
2 |
3 |
0 |
0 |
11 |
1000 |
5 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
2000 |
3 |
1 |
1 |
0 |
0 |
0 |
0 |
2 |
2000 |
8 |
2 |
0 |
0 |
0 |
0 |
2 |
|
|
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
2001 + |
5 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
North Atlantic |
Indian |
Ocean |
|||||||||||||||
|
Sternoptyx diaphana |
|||||||||||||||||
|
Depth |
0 |
5 |
Catt 10 |
h 20 |
50 |
100 |
100+ |
Night N |
Depth |
0 |
5 |
Gate 10 |
1 20 |
50 |
100 |
100 i |
\ |
|
100 |
2 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
100 |
3 |
2 |
0 |
0 |
0 |
0 |
■•> |
|
|
200 |
5 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
200 |
3 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
300 |
3 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
300 |
4 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
400 |
4 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
400 |
12 |
4 |
0 |
0 |
0 |
0 |
4 |
|
|
500 |
4 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
500 |
2 |
1 |
0 |
0 |
0 |
0 |
1 |
|
|
600 |
3 |
1 |
0 |
0 |
1 |
0 |
0 |
2 |
600 |
1 |
0 |
1 |
0 |
0 |
0 |
1 |
|
|
700 |
1 |
3 |
0 |
0 |
1 |
0 |
0 |
4 |
700 |
2 |
2 |
1 |
0 |
0 |
0 |
3 |
|
|
1000 |
0 |
2 |
1 |
3 |
3 |
0 |
0 |
9 |
1000 |
3 |
1 |
0 |
1 |
1 |
0 |
3 |
|
|
2000 |
1 |
0 |
0 |
1 |
0 |
0 |
0 |
1 |
2000 |
2 |
4 |
0 |
0 |
1 |
0 |
5 |
|
|
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
2001 + |
0 |
4 |
0 |
0 |
0 |
0 |
4 |
NE Atlantic (20°-40°N, 0°-30°W)
Southei^n Ocean (Pacific)
Marine Hatchetfishes • Baird
123
|
Catch |
|||||||
|
Depth |
0 |
5 |
10 20 |
50 |
100 |
100 + |
N |
|
100 |
2 |
1 |
1 0 |
0 |
0 |
0 |
2 |
|
200 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
300 |
0 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
400 |
0 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
500 |
0 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
600 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
700 |
0 |
0 |
0 0 |
0 |
1 |
0 |
1 |
|
1000 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
2000 |
0 |
0 |
0 1 |
1 |
0 |
0 |
2 |
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
Culf ot( |
Guinea |
Catch |
s |
ter |
|||
|
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
100 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
200 |
8 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
300 |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
400 |
4 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
500 |
3 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
600 |
6 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
700 |
5 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
1000 |
8 |
2 |
1 0 |
0 |
0 |
0 |
3 |
|
2000 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
Western |
Indian |
Oce |
Catch |
||||
|
Depth |
0 |
5 |
10 20 |
50 |
100 |
100+ |
N |
|
100 |
1 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
200 |
3 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
300 |
3 |
1 |
0 0 |
0 |
0 |
0 |
1 |
|
400 |
4 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
500 |
7 |
2 |
0 0 |
0 |
0 |
0 |
2 |
|
600 |
3 |
2 |
2 2 |
0 |
0 |
0 |
6 |
|
700 |
2 |
1 |
0 1 |
1 |
0 |
0 |
3 |
|
1000 |
2 |
3 |
2 4 |
1 |
0 |
0 |
10 |
|
2000 |
0 |
0 |
1 0 |
1 |
0 |
0 |
2 |
|
2001 + |
0 |
0 |
0 0 |
0 |
0 |
0 |
0 |
|
NE Atlantic |
Appendix B (Continued)
Depth
Catch
10 20 50 100 100+ N
|
100 |
15 |
1 |
0 |
0 |
0 |
1 |
0 |
2 |
|
200 |
10 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
300 |
4 |
3 |
3 |
0 |
0 |
0 |
0 |
6 |
|
400 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
500 |
2 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
|
600 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
700 |
1 |
0 |
0 |
0 |
1 |
0 |
0 |
1 |
|
1000 |
0 |
2 |
0 |
1 |
0 |
0 |
0 |
3 |
|
2000 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
Gulf of Mexico and Caribbean
SternopUjx diaphana
Night
Day
Depth
Catch
10 -2
50 100 100+ N
|
100 |
30 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
200 |
26 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
300 |
19 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
400 |
6 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
|
500 |
18 |
2 |
0 |
0 |
0 |
0 |
0 |
2 |
|
600 |
5 |
■J |
0 |
0 |
1 |
0 |
0 |
4 |
|
700 |
0 |
2 |
0 |
1 |
1 |
0 |
0 |
4 |
|
1000 |
7 |
1 |
1 |
2 |
o o |
0 |
0 |
7 |
|
2000 |
2 |
0 |
0 |
1 |
0 |
0 |
0 |
1 |
|
2001 + |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
North Atl |
antic |
(30° |
-45' |
N, 20°-70°W) |
||||
|
Catch ■ |
||||||||
|
Depth |
0 |
5 |
10 |
20 |
50 |
100 100+ |
N |
|
|
100 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
200 |
2 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
300 |
3 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
400 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
500 |
2 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
600 |
2 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
700 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
1000 |
3 |
1 |
0 |
1 |
1 |
0 |
0 |
3 |
|
2000 |
3 |
3 |
0 |
1 |
3 |
0 |
0 |
7 |
|
2001 + |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
1 |
Southern Ocean (Pacific'
LITERATURE CITED
Ahlstrom, E. H., and R. C. Counts. 1958. De- velopment and distriliution of Vinci^iicnia lucetia and related species in tlie eastern Pacific. U. S. Fish Wildlife Serv. Fish. Bull., 58: 363-416.
. 1959. Vertical distril)iition of pelaj^ic
fish eggs off California and Baja California. U. S. Fish Wildlife Serv. Fish. Bull., 60: 107-146.
Alcock, a. 1896. Supplementary list of the marine fishes of India. J. Asiatic Soc. Bengal, 65(2): 331.
. 1899. A Descriptive Catalogue of the
Indian Deep-sea Fishes in the Indian Museum . . . Collected by the Royal Indian Marine Survey Ship INVESTIGATOR. Calcutta, Baptist Mission Press, p. 211.
Alvarino, a. 1965. Chaetognaths. Oceanogr. Mar. Biol. Ann. Rev., 3: 115-194.
Arambourg, C. 1929. Argyiopclecus logeaiti, un nouveau poisson bathypelagique du Sahelien. Bull. Soc. Geo!. France, 29: 11-15.
Aron, W. 1962. The distribution of animals in the eastern North Pacific and its relationship to physical and chemical conditions. J. Fish. Res. Bd. Canada, 19: 271-314.
124 Bulletin Museum of Comparative Zoology, Vol. 142, No. 1
Backus, R. H., G. W. Mead, R. L. IIaediuch, AND A. W. Ebeling. 1965. The mesopelagic fishes collected during cruise 17 of the R 'V CHAIN, with a method for analyzing faunal transects. Bull. Mus. Conip. Zool. Harvard, 134(5): 139-157.
Badcock, J. 1969. Colour variation in two meso- pelagic fishes and its correlation with ambient light conditions. Natine (London), 221: 383-385.
Bahamoxde, N. 1963. Argtjropdecus en Chile. Rev. Univ. Catol. Santiago, 48: 83-86.
Barnard, K. H. 1925. A monograph of tlie marine fishes of South Africa. Ft. 1. Ann. South African Mus., 21: 1-153.
Bassot, J. M. 1966. On the comparative morphol- ogy of some luminous organs. In Johnson, L., and Y. Haneda (eds.), Bioluminescence in Progress. Princeton, N. J., Princeton Univ. Press. 650 pp.
Beebe, W. 1927. Preliminary list of Bermuda deep-sea fish. Zoologica (New York), 22: 197-208.
Berg, L. S. 1940. Classification of fishes both recent and fossil. Trav. Inst. Zool. Acad. Sci. URSS, 5: 34(>-517.
Berry, F. H., and H. C. Perkins. 1965. Suney of pelagic fishes of the California current area. U. S. Fish Wildlife Serv. Fish. Bull., 65(3): 625-682.
Bertelsen, E., and J. Grontved. 1949. The light organs of a bathypelagic fish A. olfersi (Cuvier) photographed by its own light. Vid. Medd. Dansk Naturh. Foren. Kjoben- havn. 111: 163-167.
Bertin, L. 1940. Catalogue des types de poissons du Museum d'Histoire Naturelle. Bull. Mus. Hist. Nat. Paris, Ser. 2, 12: 244-322.
BiERi, R. 1959. The distribution of the planktonic Chaetognatha in the Pacific and their re- lationship to the water masses. Limnol. Oceanogr., 4: 1-28.
BiGELOw, H. B., and W. C. Schroeder. 1953. Fishes of tlie Gulf of Maine. U. S. Fish Wildlife Serv. Fish. Bull., 74: 1-577.
Blache, J. 1964. Poissons bathypelagiques de la famille de Sternoptychidae (Teleostei, Clupei- formi, Stomatiodei). O.R.S.T.O.M. Ocean- ographie, 5: 71-87.
Borodin, N. A. 1931. Atlantic deep-sea fishes. Bull. Mus. Comp. Zool. Harvard, 72: 55-89.
Brauer, a. 1901. iJber einige von die Valdivia- Expedition gesammelte Tiefsee-Fische und ihre Augen. Sitz. Gesellsch. Befiird. Gesam. Naturw. Marburg, 1901(7): 120.
. 1906. Die Tiefsee-Fische. I. Systemati-
scher Tiel. Wiss. Ergeb. Deut. Tiefsee-Exped. VALDIVIA, 1898-99, 15: 69-122.
. 1908. Die Tiefsee-Fische. II. Anato-
mischer Teil. Wiss. Ergeb. Deut. Tiefsee- Exped. VALDIVIA, 1898-99, 15: 1-175.
Briggs, J. C. 1960. Fishes of worldwide (cir- cumtropical) distribution. Copeia, (3): 171- 180.
. 1961. The East Pacific Barrier and the
distribution of marine shore fishes. Evolution,
15: 545-554. Bright, T. J., and W. E. Paquegnat. 1969. Deep
sea hatchetfishes of the Gulf of Mexico.
Quart. J. Florida Acad. Sci., 32: 26-37. Brinton, E. 1962. The distribution of Pacific
euphausiids. Bull. Scripps Inst. Oceanogr.,
Univ. California, 8(2): 51-270. BuEN, R. DE. 1935. Fauna Ictiologica. Pt. 1.
Notas y Resumenes. Inst. Espaiiol Oceanogr.,
88: 1-182. Bussing, W. A. 1965. The midwater fishes of
the Peru-Chile trench. Antarctic Res. Ser.,
5: 185-227. Cervigon, F. 1964. Nuevas citas de Peces para
Venezuela y datos sombre algunas especies
poco conocidas. Novedades Cient. Contrib.
Ocas. Mus. Hist. Nat. La Salle, Caracas.,
Ser. Zool., 31: 1-18. Clarke, F. E. 1878. On two new fishes. Trans.
Proc. New Zealand Inst., 7: 91-300. Clarke, M. R. 1966. Review of the systematics
and ecology of oceanic squids. Advan. Mar.
Biol., 7: 91-300. Clemens, W. A., and G. V. Wilhy. 1949. Fishes
of the Pacific coast of Canada. Fish. Res.
Bd. Canada, Bull., 67: 1-368. Cocco, A. 1829. Su di alcuni nuovi pesci del
mare di Messina. Giorn. Sci. Lett. Arti Sicilia,
26(77): 46-50. CoLLETT, R. 1903. Meddelelser om Norges Fiske
i Aarene 1884-1901. Vid. Selsk. For. Chris-
tiania, 9: 108. Crane, J. M., Jr. 1966. Late tertiary radiation of
viperfishes (Chauliodontidae) based on a
comparison of Recent and Miocene species.
Los Angeles County Mus. Contri. Sci., No.
115: 29. CuviER, G. 1829. Poissons. Le Regne Animal.
2d ed., vol. 2. Paris, p. 316. , AND A. Valenciennes. 1849. Histoire
Naturelle des Poissons. Paris. Vol. 22: 398-
415. Daniltshenko, p. G. 1960. Bony fishes of the
Maikop beds of the Caucasus. Trudy Paleo.
Inst. Moscow, 78: 1-208 (In Russian). . 1962. Fishes from the Dabakhan seriest
of Georgia. Paleon. Zhurnal, 1: 111-126 (Im
Russian ) . David, L. R. 1943. Miocene fishes of Southerni
California. Geol. Soc. America Spec. Pap.,
43: 1-733.
Marine Hatchetfishes • Baird
125
David, P. M. 1963. Speciation in Chaetognatlia. Syst. Assoc. Publ., 5: 129-143.
Denton, E. J., and J. A. C. Nicol. 1965. Studies on reflection of light from silvery surfaces of fishes, with special reference to the bleak Alhiimus alhiirnus. J. Mar. Biol. Assoc. U. K., 45: 711-738.
DoLLFus, R. p. 1955. Premiere contribution a I'etablissement d'un fichier ichtliyologique dn Maroc Atlantique de Tanger a I'embouchure de I'oued dra. Trav. Inst. Sci. Cherifien (ser. Zool.) Rabat, 6: 1-226.
i]BELiNG, A. W. 1962. Melamphaidae I. Systc- matics and zoogeography of the species in the bathypelagic fish genus Melamphaes Gunther. DANA Rept., 58: 1-164.
, AND W. H. Weed, III. 1963. Melamphai- dae III. Systematics and distribution of the species in the bathypelagic fish genus, Scopelogadits Vaillant. DANA Rept., 60: 1-58.
i]HRENBAUM, E. 1909. Eicr mid Larven von Fischen, 1. Nordisches Plankton, Lief. 10: 327.
Hkman, S. 1967. Zoogeography of the Sea. Lon- don, Sidgwick and Jackson, Ltd. 417 pp.
''ager, W. E., and J. A. McGowAN. 1963. Zoo- plankton species groups in the North Pacific. Science, 140: 453-460.
^OLLETT, H. 1952. Annotated list, central Cali- fornia fish. Proe. California Acad. Sci., 27: 409.
''owLER, H. W. 1933. Descriptions of new fishes obtained 1907-1910, chiefly in the Philip- pine Islands and adjacent seas. Proc. Acad. Nat. Sci. Philadelphia, 85: 233-367.
. 1936. The marine fishes of West Africa.
Bull. American Mus. Nat. Hist., 70: 1-275.
. 1949. Fishes of Oceania. Mem. Bernice
P. Bishop Mus., 12: 1-540.
. 1956. Fishes of the Red Sea and South-
ern Arabia. Vol. 1. Jerusalem, Weizmann Press. 240 pp.
^UGLiSTER, F. C. 1960. Atlantic Ocean atlas of temperature and salinity profiles and data from the International Geophysical Year 1957-1958. Woods Hole Oceanographic In- stitution Atlas Series, 1 : 1-209.
jArman, S. 1899. Reports on an exploration off the west coasts of Mexico, Central and South America, and off the Galapagos Islands. Mem. Mus. Comp. Zool. Harvard, 24: 1-431.
jIrbs, R. H., Jr. 1968. Photoncctcs miinificus, a new species of melanostomiatid fish from the South Pacific subtropical convergence, with remarks on the convergence fauna. Los Angeles County Mus. Contri. Sci., 149: 1-6.
, and B. a. Hurwitz. 1967. Systematics
and zoogeography of the stomiatoid fishes
Chauliochis pammelas and C. sloani, of the Indian Ocean. Copeia, (4): 798-805.
Gilbert, G. H. 1905. Fishes of tlie Hawaiian Islands. Bull. U. S. Fish Gomm., 23: 575- 713.
Gill, T. 1884. Note on the Stemoptychidae. Proc. U. S. Nat. Mus., 7: 349-351.
GooDE, G. B., and T. H. Bean. 1896. Oceanic Ichthyology. Mem. Mus. Comp. Zool. Har- vard, 22: 1-554.
Grandperrin, R., and J. RivATON. 1966. "Cori- olis": Crosiere "Alize." Individualization de plusieurs ichthyofaunes le long de I'equateur. Cab. O.R.S.T.O.M. (ser. Oceanogr.), 4(4): 36-49.
Greenwood, P. L., D. E. Rosen, S. H. Weitz- MAN, AND G. S Myers. 1966. Phyletic studies of teleostean fishes with a provisional classi- fication of living forms. Bull. American Mus. Nat. Hist., 113(4): 341-455.
Gregory, W. K., and G. M. Conrad. 1936. Pic- torial phylogenies of deep sea Isospondyli and Iniomi. Copeia, ( 1 ) : 21-36.
Grey, M. 1959. Deep sea fishes from the Gulf of Mexico with tlie description of a new species. Fieldiana, Zool., 39: 326-339.
. 1964. Family Gonostomatidae. In Fishes
of tlie Western North Atlantic. Sears Found. Mar. Res., 4: 78-238.
Grice, G. D. 1963. Remarks on the diversity and distribution of deep living Copepoda. Proc. Int. Congr. Zool. Washington, 16: 312-327.
-, and K. Hulsemann. 1967. Bathypelagic
copepods of the western Indian Ocean. Proc. U. S. Nat. Mus., 122: 1-22.
GiJXTHER, A. 1864. Catalogue of the Fishes in the British Museum. London, Taylor and Francis. Vol. 5: 384-392.
. 1887. Report on the scientific results of
H.M.S. CHALLENGER during the years 1873-76. Rep. Sci. Res. H.M.S. CHALLEN- GER, 22: 1-335.
Haedrich, R. L., and J. G. Nielsen. 1966. Fishes eaten by Alepisaiiris (Pisces, Iniomi) in the southeastern Pacific Ocean. Deep-Sea Res., 13: 909-921.
Haffner, R. E. 1952. Zoogeography of the bathypelagic fish, Chatiliodus. Syst. Zool., 1: 112-133. '
Haig, J. 1955. Fishes killed by the 1950 eruption of Maima Loa. III. Stemoptychidae. Pacific Sci., 9: 321-329.
Handrick, K. 1901. Zur Kenntnis des Nerven- system und der Leuchtorganc des Argyropele- cus hemigymnus. Zoologica (Stuttgart), 13 (32): 1-68.
Haneda, Y. 1952. Some luminous fishes of the genera Yarella and Pohjipnus. Pacific Sci., 6: 13-21.
126 BiiUcfin Museum of Comparative Zoology, Vol. 142, No. 1
Harrissox, C. M. H. 1967. On nietliods for sampling niesopelagic fishes. Symp. Zoo!. Soc. London, 19: 71-126.
Hermann-, D. J. 1781. Uber cin neues anierikan- isc'hes Fischgcschlecht, Sternoptyx diapliaua, der durchsichtige Brust-Falten-Fisch. Dcr Natuiforscher, 16: 33-34.
HoLGERSEN, H. 1958. Sjeldne fisker ved soinestky- sten. Arbok Stavanger Museum, 59: 120- 124.
Holt, E. W. L., and L. W. Byrne. 1913. Sixth report on tlie fishes of the Irish Atlantic slope. Fisheries, Ireland, Sci. Invest., 1912: 1-28.
HuBBS, C. L. 1953. Synonymy of the liatln- pelagic fish genus Rhynchohyalus, referred to the expanded family Argentinidae. Copeia, (2): 96-97.
, AND K. F. Lagler. 1947. Fishes of the
Great Lakes Region. Bull. Cranbrook Inst. Sci., 26: 1-186.
Jerzmanska, a. 1960. Ichthyofauna from the Jaslo shales at Solmiow (Poland). Acta "Paleont. Folonica, 5: 367-420 (English Sum- mary ) .
, and S. Jucha. 1963. L'affeurment de la
faune de poissons dans les schistes de Jaslo a Tubno pres de Dynow ( Karpates Polonaises ) . Ann. Soc. Geol. Pologne, 33: 159-180.
Jespersen, p. 1915. Sternoptychidae. Kept. Dan- ish Oceanogr. Exped. 1908-10. Vol. II: Biology. A, 2. p. 41.
. 1934. Sternoptychidae. In Joubin (ed.).
Faune Ichthvologique de I'Atlantique Nord. No. 15.
, AND A. V. Taning. 1919. Some Mediter- ranean and Atlantic Sternoptychidae. Vid. Medd. Naturhist. For. Kjobenhavn, 70: 220- 228.
1926. Mediterranean Sternoptychidae.
Kept. Danish Oceanogr. Exped. 1908-10. A., 12: 1-59. Johnson, M. W., and E. Brinton. 1963. Bio- logical species, water masses, and currents. In Hill, M. N. (ed.). The Sea. Interscience Publ., 2: 381-414.
Jordan, D. S., and B. W. Evermann. 1896. The
fishes of North and Middle America. U. S.
Nat. Mus. Bull., 47 (Pt. 1): 1-604. Jordan, D. S., and E. C. Starks. 1904. Japanese
fishes. Bull. U. S. Fish Comm., 22: 581. Kamohara, T. 1952. Revised descriptions of the
off shore bottom fishes of Prov. Tosa,
Shikoku, Japan. Repts. Kochi Univ. Nat. Sci.,
3: 16-24. , and T. Yamaicwva. 1965. Fishes from
Amami-Ishimi and adjacent regions. Rept.
Usa Mar. Biol. Stn., 12: 1-27. Kobayashi, D. 1963. First record of PoIyii>uus
lalcmattts from the Pacific Ocean. Copeia,
(2): 170-180. Koefoed, E. 1961. Isospondyli 2. Heterophoto-
dermi, Sternoptychidae. Rept. Sci. Res.
Michael Sars. N. Atlantic, 4: Pt. 2(10): 1-12. Koepcke, H. W. 1962. Lista de los peces marinos
conocidos del Peru con datos de sa distri-
bucion geografica. Biota, 4: 145-154. KoTTHAUS, A. 1957. Fische des Indischen Ozeans.
Sonderheft aus METEOR Forschungsergeb-
nisse. Reihe D, 1: 1-22. , AND G. Krefft. 1957. Fischefaunenliste
der Fahrten mit F.F.S. "ANTON DOHRN"
nach Island — Gronland. Ber. Deut. Wiss.
Komm. Meeresforsch., N.F., 14: 169-191. KouMANS, F. D. 1953. Biological results of the
Snellius Expedition. XVI. The pisces and
leptocardii. Temminckia, 9: 177-210. Lavenberg, R. J., AND J. E. Fitch. 1966. An- notated list of fishes collected by midwater
trawl in the Gulf of California, March-April
1964. California Fish and Game, 52: 92-110. , AND A. W. EiiELiN'G. 1967. Distribution
of midwater fishes among deep water Ixisins
of the Southern California shelf. Proc. Symp.
Biol. California Islands, Santa Barbara, 1 :
185-201. Ledenfeld, R. V. 1905. The radiating organs of
the deep sea fishes. Mem. Mus. Comp. Zool,)
Harvard, 30: 170. Lowe, R. T. 1843-1860. History of the Fishes
of Madeira. London, Bernard Quaritch.
p. 64. Marr, J. C. 1948. Two additions to the known
fauna of California. Copeia (2): 140—141. Marshall, N. B. 1960. Swimbladder structure
of deep-sea fishes in relation to their system-
atics and biology. DISCOVERY Rept.s., 31:
1-122. . 1963. Diversity, distribution, and speci-
ation of deep-sea fishes. Svst. Assoc. Publ..
5: 181-195. Matsubara, K. 1941. Studies on the deep sea
fishes of Japan. Suisan Kenkiu-shi, Japan,
36: 1-8. . 1950. Studies on the fishes of the genus
Pohjipntis found in Japan. Japan J. Ichthy.,
1: 192-197. Maul, G. E. 1949a. Lista Systematica dos Peixe.'-"
Assinalados nos Mares de Madeira e Indicf
Alphabetico. In de Noronha, A. C, and A.
A. Sarmento (eds.), Vertebrados de Madrini
2. Peixes. Funchal, Tipografia Experanca:
1-53. . 1949b. Monografia dos peLxes de Museii
Municipal de Funchal. Bol. Mus. Mun
Funchal, 4: 9-17. Mavr, E. 1963. Animal Species and Evolution
Cambridge, Harvard Uni\ersity Press. 797 pp
I
Marine Hatchetfishes • Baird
127
, E. G. LiNSLEY, AND R. L. UsiNGER. 1953.
Methods and Principles of Systematic Zool- ogy. New York, McGraw-Hill. 538 pp.
VIcCuLLOCH, A. R. 1914. Report on some Fishes obtained by the F.I.S. ENDEAVOUR on the Coasts of Queensland, New South Wales, Victoria, Tasmania, and South and South- western Australia. Ft. 2. Zoological Results F.I.S. ENDEAVOUR 1909-10, 2: 78-87.
. 1923. Fishes from Australia and Lord
Howe Island. Rec. Australian Mus., 14: 118.
VlcGowAX, J. A. 1960. The relationship of the distril)ution of the planktonic worm Pocohius meseres Heath to water masses of the North Pacific. Deep Sea Res., 6: 125-139.
. 1963. Geographical variation in Linia-
cina helicina in the North Pacific. Syst. Assoc. Publ., 5: 109-128.
\Ieao, G. W., and F. H. C. Taylor. 19.53. A collection of oceanic fishes off Northeastern Japan. J. Fish. Res. Bd. Canada, 10: 570.
\lisRA, K. 1952. An aid to the identification of the fishes of India, Burma, and Ceylon. II. Clupeiformes, Bathyclupeiformes, Galaxi- formes, Scopeliformes, and Ateleopiformes. Rec. Indian Mus. Calcutta, 50: .367-422.
VIoRROw, J. E. 1957. Mid-depth fishes of the Yale South American Expedition. Bull. Bing- ham Oceanogr. Coll., 16: 56-71.
. 1964. Suborder Stomiatoidea; families
Gonostomatidae and Sternoptychidae. In Fishes of the Western North Atlantic. Sears Found. Mar. Res. 4(1): 71.
VluROMTSEV, A. M. 1963. The principal hydro- logical features of the Pacific Ocean. Jerusalem, Israel Program for Scientific Trans- lation. 790 pp.
Mafpaktitis, B. G. 1968. Taxonomy and dis- tribution of the lanternfishes, genera Lohi- anchia and Diaphus, in the North Atlantic. DANA Rept., 73: 1-131.
VicoL, J. A. C. 1967. The luminescence of fishes. Symp. Zool. Soc. London, 19: 27-55.
VoRDEN, C. R. 1961. Comparative osteology of representative salmonid fishes with particular reference to the grayling (Thyiuallus arcti- cus) and its phylogeny. J. Fish. Res. Bd. Canada, 18: 679-789.
VoRMAX, J. R. 1930. Oceanic fishes and flat- fishes collected in 192.5-27. DISCO\'ERY Repts., 2: 261-370.
. 19.37. Fishes. British, Australian, and
New Zealand Antarctic Res. Exped. 1929-31, 1(2): 51-88.
. 1939. Fishes. Sci. Repts. lohn Nburav
Exped. 1933-34, 7(1): 1-116.
. 1944. A draft synopsis of the orders,
families, and genera of recent fishes and fish-
like vertebrates. London, British Museum
( Nat. Hist. ) 649 pp. NusBAUM-HiLARowicz, J. 1923. fitudes d'anato-
mie Comparee sur les Poissons . . . Pt. 2.
Res. Camp. Sci. Alliert ler Monaco, 65:
1-100. Nybelin, O. 1948. Fishes collected by the
Skagerak Expedition, 1946. Goteborgs K.
Vet. Vitt. Samh. Handl, ser. B, 5(16): 1-95. Ogilby, J. D. 1888. Description of a new genus
and species of deep sea fish from Lord
Howe Island. Proc. Linn. Soc. New South
Wales, 3(2): 1313. Okada, Y., AND K. Suzuki. 1956. Taxonomic
considerations of tlie lantern fish Pohjii)nii.s
spinosiis and related species. Pacific Sci., 10:
297-302. Parr, A. E. 1937. Concluding report on fishes
with species index for articles 1-7. Bull.
Bingham Oceanogr. Coll., 3: 1-79. Pauca, M. 1931. Neue Fische aus dem Oligoziin
von Piatra-Neamt., Bull. Acad. Roumaine
Bucarest, 14: 29-34. Pearcy, W. G. 1964. Some distributional features
of mesopelagic fishes off Oregon. J. Mar.
Res., 22: 83-102. Peres, J. M. 1958. Remarques generales sur un
ensemble de quinze plongees effectuees avec
le Bathyscaphe F.N.R.S. III. In Resultats
Scientifiques des Campagnes du Bathyscaphe
F.N.R.S. Ill 1954-1957. V. Ann. Inst.
Oceanogr. Monaco, N. S., 35: 259-285. Poll, M. 1953. Expedition oceanographique
Beige dans les eaux cotieres Africaines de
I'Atlantique Sud, 1948-1949. Resultats Scien-
tifique. Poissons III. Teleosteens, Malacop-
terygiens. 4: 1-390. Rass, T. C. 1955. Deep water fishes of the Kurile-
Kamchatka Trench. Trud. Inst. Okeanol.,
12: 328-339. Rechnitzer, a. G., and J. Bohlke. 1958. Ichthtj-
ococcus irrcp,ularis, a new gonostomatine fish
from the Eastern Pacific. Copeia, ( 1 ) : 10-15. Regan, C. T. 1908. Report on the marine fishes
collected l^y J. S. Gardiner in the Indian
Ocean. Trans. Linn. Soc. Zool. London, 12:
218. . 1923. The classification of the stonii-
atoid fishes. Ann. Mag. Nat. Hist., 9(11):
612-615. Rivero, L. H. 19.34. Paces nuevos para la fauna
Cubana. Mem. Soc. Cubana Hist. Nat., 8:
31. . 1936. Some new, rare, and little known
fishes from Cuba. Proc. Boston Soc. Nat.
Hist., 41: 56. RouLE, L., and F. Angel. 1933. Poissons prove-
nant des campagnes du Prince Albert ler de
128 Bullet in Museum of Comparative Zoology, Vol. 142, No. 1
Monaco. Res. Camp. Sci. Monaco, 86:
1-115. RoxAS, H. A. 1934. A review of Philippine iso-
spondvlous fishes. Philippine J. Sci., 55:
287. Samuel, C. T. 1963. Notes on deep sea fishes
collected by R/V CONCH off the Kerala
Coast. Bull. Dept. Mar. Biol. Oceanogr. Univ.
Kerala, India, 1: 101-107. Sanzo, R. 1928. Uova, larvae, e stadi giovanili
di teleostei. Comit. Talassogr. Italiano,
Monogr. 2: 1-60. Sauvage, L. 1891. Histoire Naturelle des Pois-
sons. In Grandidier, A. (ed.), Hist. Phys.
Nat. Pol. Madagascar, 16: 483. ScHROEDER, E. H. 1963. Serial Atlas of the
Marine Environment. Folio 2. North Atlantic
Temperatures at a Depth of 200 Meters.
New York, American Geographical Society. ScHULTZ, L. P. 1937. A new species of deep sea
fish, Argijropclecus antrorsospiniis, of the
family Sternoptychidae. Smith. Misc. Coll.,
91(27): 5-8. . 1938. Review of the fishes of the genera
Polyi))nu.s and Argyropelecus (Family Ster- noptychidae) with description of three new
species. Proc. U. S. Nat. Mus., 86: 135-
168. . 1961. Revision of the marine siKer
hatchetfishes ( Family Sternoptychidae ) . Proc.
U. S. Nat. Mus., 112: 587-649.
1964. Family Sternoptychidae. 7m Fishes
of the Western North Atlantic. Sears Found. Mar. Res., 1(4): 241-273.
Scott, W. B. 1965. A record of the hatchetfish Polyipnus asteroides Schnltz from the Cana- dian Atlantic region. J- Fish. Res. Bd. Canada, 22: 1303.
Smith, J. L. B. 1953. The Sea Fishes of Southern
Africa. South Africa, Central News Agency, Ltd. 550 pp.
Suzuki, K. 1964. Two rare deep sea fishes from Kumano-Nada. Rept. Fac. Fish. Pref. Univ. Mie, 5: 1-8.
Suzuki, O. 1961. Mechanical analysis on the working behavior of midwater trawl. Bull. Japan Soc. Sci. Fish., 27: 903-907.
SvERDRUP, H. U., M. W. Johnson, and R. H. Fleming. 1960. The Oceans: Their Physics, Chemistry, and General Biology. 9th Ed. Englewood Cliffs, N. J., Prentice-Hall, Inc.
Ta\xor, R. W. 1967. An enzyme method of clearing and staining small vertebrates. Proc. U. S. Nat. Mus., 122: 1-7.
Weber, M., and L. F. de Beaufort. 1913. The Fishes of the Indo-Australiau Archipelago. Vol. 2. Leiden, E. J. Brill. 428 pp.
Weitzman, S. H. 1967a. The osteology and relationships of the Astronesthidae, a family of oceanic fishes. DANA Rept., 71: 49-52.
. 1967b. The origin of the stomiatoid fishes
with comments on the classification of sal- moniform fishes. Copeia, ( 3 ) : 507-540.
Wheeler, A. 1969. The Fishes of the British Isles and North-west Europe. London, Mac- millan. 613 pp.
Whiteley, G. p. 1940. Illustrations of some Australian fishes. Australian Zool., 9: 404.
Wilimovsky, N. J. 1951. The correct generic and family name of the deep sea hatchet- fishes (Stemoptyx and Sternoptychidae). Copeia, (3): 247-248.
Wooster, W. S., and J. L. Reid, Jr. 1963. East- ern boundary currents. In Hill, NL N. (ed.), The Sea. New York and London, Inter- science Publishers, 1: 253-279.
Zugmayer, E. 1911. Poissons provenant des campagnes du yacht Princesse Alice ( 1901- 1910). Res. Camp. Sci. Monaco, 35: 1-174.
ulletin OF THE -
Museum of
omparative Zoology
The Carabid Beetles of New Guinea.
Part IV. General Considerations; Analysis and
History of Fauna; Taxonomic Supplement
p. J. DABLINGTON, JR.
HARVARD UNIVERSITY VOLUME 142, NUMBER 2
CAMBRIDGE, MASSACHUSETTS, U.S.A. 18 OCTOBER 1971
PUBLICATIONS ISSUED
OR DISTRIBUTED BY THE
MUSEUM OF COMPARATIVE ZOOLOGY
HARVARD UNIVERSITY
Bulletin 1863-
Breviora 1952-
Memoirs 1864-1938
JoHNSONiA, Department of Mollusks, 1941-
OccAsiONAL Papers on Mollusks, 1945-
Other Publications.
Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth.
Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth.
Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth.
Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth.
Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. )
Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae ( Mollusca : Bivalvia ) . $8.00 cloth.
Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth.
Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. )
Publications of the Boston Society of Natural History.
Authors preparing manuscripts for the Bulletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A.
© The President and Fellows of Harvard College 1971.
THE CARABID BEETLES OF NEW GUINEA. PART IV. GENERAL CONSIDERATIONS; ANALYSIS AND HISTORY OF FAUNA; TAXONOMIC SUPPLEMENr
p. J. DARLINGTON, JR.
CONTENTS
Introduction to Part IV. 133
1. Purpose; previous parts; acknowledg- ments 133
2. Sources, disposition, and adequacy of material 133
3. Preparation for work on New Guinean Carabidae; my collecting 135
4. Basic literature 136
5. Localities 138
Policies and Methods 143
6. Modern taxonomy 143
7. Types 146
8. Taxon concepts: subfamilies and tribes 147
9. Genera 147
10. Species 148
11. Subspecies 149
12. Nature of taxonomic characters 150
13. Secondary sexual and genitalic char- acters 151
14. Preservation of material 152
15. Taxonomic methods and procedures __ 153
16. Data sheets 154
Analysis and Discussion 156
17. New Guinea 156
^The author suggests that, when the four parts of TJie Caiabid Beetles of New Guinea are bound together, the present part be divided, and the whole bound in the following order: general con- siderations, analysis, and history of the fauna from Part IV; Part I; Part II; Part III; and the taxonomic supplement from Part IV.
■ Work and publication supported by National Science Foundation Grant GB-12346.
Bull. Mus. Comp,
18. The New Guinean carabid fauna: tax- onomic composition 157
19. Numbers of species 158
20. Size of individuals 161
21. Wings and wing atrophy 165
22. Explanations of wing atrophy 170
23. Summary of wing state and wing atrophy 172
24. Ecology: habitats 172
25. Ecologic composition 173
26. Altitude 175
27. Ants : 176
28. Ecologic interactions 176
29. Ecologic ranges 177
Zoogeography: Existing Geographic Patterns .. 178
30. Geographic patterns, relationships,
and origins 178
31. Existing geographic relationships: problems and procedures _. 178
32. Distribution of carabid tribes from
Asia to Australia 180
(Cicindelinae) _ 182
33. Ozaenini 183
34. Paussini 183
35. Scaritini 183
36. Bembidiini 184
37. Trechini 185
38. Panagaeini 186
39. Pterostichini 186
40. Agonini 188
41. Perigonini 189
42. Licinini 189
43. Chlaeniini 190
44. Oodini 190
45. Harpalini 190
46. Anaulacini 192
47. Gyclosomini 192
48. Lebiini _____ __. 193
49. Pentagonicini 196
50. Hexagoniini 197
Zool., 142(2): 129-337, October, 1971 129
130 Bulletin Museum of Comparative Zoology, Vol 142, No. 2
51. Odacanthini -- -- 197
52. Dnptiiii 198
53. Zuphiini - 198
54. Ilelluodini 198
55. Hclluonini _ 198
56. Brachinini _ .- 198
57. Pseudomorphini 200
58. Smiiinaiy of goM),c;raphic units 200
59. Cco^iaphic units by major habitats — . 201
60. Transitions of carabid faunas from Asia to Australia - 202
61. Wallace's Line and Celebes 202
62. Moluccas 203
63. Transition from New Guinea to Aus- tralia 203
64. Summary of transitions — - — . 207
65. Faunal regions — - 210
66. Geographic patterns within New Guinea 211
Zoogeography: Dispersals and Geographic
Origins 216
67. The apparent main pattern of dis- persal 216
68. Dominance, competition, and extinc- tion in dispersal 216
69. Complexity of dispersal 217
70. Place of New Guinea in the main dis- persal pattern 217
71. Directions of dispersal ____ 218
72. Direction and vagility 218
73. Carabid versus mammalian dispersals 219
74. Directions of dispersal of tribes, gen- era, and species - 220
75. Summary of direction to this point _-_. 221
76. Direction and dominance 224
77. Direction and size 224
78. Direction and wings — _ 224
79. Direction and ecology 225
80. Direction and altitude: mountain- hopping across the Malay Archipelago 225
81. Direction and age 228
82. Australian-American discontinuities . .. 229
83. Summary of directions of dispersal .^ - 230
84. Barriers; filtering at Wallace's Line, Moluccas, Cape York 232
85. Amount of dispersal now and in the past 233
86. Summary of geographic origins of New Guinean Carabidae 236
E\olution 237
87. Evolutionary perspective 237
88. Evolution of New Guinean carabid fauna as whole _ 237
89. Relative age of the New Guinean fauna 239
90. Evolution of the mountain fauna 240
91. Evolution and adaptation of separate carabid stocks on New Guinea 243
92. Agonine radiation on New Guinea — 244
93. Evolutionary trends: not toward in- crease of size 245
94. Atrophy of wings, and associated trends ___ 246
95. Loss of setae ..._ 246
96. Modification of legs and tarsi 247
97. Modification of ecology and behavior 248
98. Parallelism and convergence; develop- ment of elytral spines; color patterns .. 248
99. Mimicry _ 250
100. Mutation and dimorphism 251
Taxonomic Section 254
101. Tribal classifications 254
Taxonomic Supplement .— 255
Cicindelinae 255
Carabinae - 255
Ozaenini 255
Scaritini 255
Harpalinae 258
Bembidiini 258
Trechini _ 262
Panagaeini 262
Pterostichini 265
Agonini 272
Perigonini 322
Licinini — 322
Chlaeniini 322
Oodini 322
Harpalini 324
Lebiini 325
Pentagonicini 330
Hexagoniini 331
Odacanthini 331
Zuphiini ( Leleupidiini ) ___ 332
Helkumini 333
Brachinini 334
Bibliography 335
In lieu of index 338
ABSTRACT
This is the fourth and final part of an extended work on the Carabidae (predaceous ground beetles) of New Guinea. About 24,000 specimens from New Guinea have been examined, enough to show the general characteristics of the New Guinean carabid fauna, although many details remain unknown. Collecting done in New Guinea and in adjacent eastern Australia, the collections accumulated at the Museum of Comparative Zoology, and preparations for their study are described; working collections of New Guinean Carabidae for future use are deposited in the MCZ, the British Museum (Natural History), the Bishop Museum in Honolulu, and with CSIRO ini Canberra, Australia. Localities are mapped, and those where most Carabidae have been found
The carabid beetles of New Guinea • Darlington 131
(Dobodiira, Wau, Mt. Wilhelm and vicinity, and the Snow Mountains ) are briefly described. Then discussed are modern taxonomic methods in rela- tion to those of the past; the continuing usefulness of types; concepts of tribes, genera, species, and subspecies; the nature of taxonomic characters; and the relative importance of secondary sexual and genitalic characters in carabid classification. Methods and procedures of second-stage faunal taxonomy (of which this work is an example) are discussed, with emphasis on the limitations imposed by time and available material.
Under "Analysis and discussion," the immense, tropical, mountainous island of New Guinea is described briefly, and its carabid fauna is de- scribed and analyzed. The taxonomic compo- sition of the fauna is summarized. The number of species of Carabidae now known from New Guinea is 667, of which 434 occur in the low- lands, below 500 m altitude, and 376 in the mountains, above 1000 m; 161 of these species occur in both the lowlands and the mountains; and 18 additional species are iniknown as to altitude. At one lowland locality ( Doliodura, Papua), 217 species have been found; at one mid-altitude locality (Wau and vicinity, North- east New Guinea, 1000-2000 m), 170 species; but numbers of species decrease sharply at higher altitudes. In size, New Guinean Carabidae are small, rarely as much as 25 m in length. The lowland fauna is bimodal in size distribution, species 2.00-2.95 and 6.00-6.95 mm long being most numerous, with a deficiency of species especially at 4.00-4.95 mm. This bimodality is a result of an apparently recent arrival in New Guinea of many small Tachtjs, which have im- posed a second mode on an otherwise unimodal fauna, but the bimodality may reflect also a vulnerability of 4 to 5 mm carabids to compe- tition with ants. The mountain carabid fauna of New Guinea is unimodal in size distribution, with tlie mode at 9.00-9.95 mm. Almost ,all lowland Carabidae in New Guinea are winged, but incidence of wing atrophy increases with altitude, reaching 95 per cent on the highest mountain tops. The causes of flightlessness at high altitudes are complex; blowing or straggling away of winged individuals is apparently unim- portant, for most of the flightless species live in dense, wind-free montane forest; limitation of irea on mountains, by requiring carabid popu- lations to be dense and stable, may be the most mportant single factor in reducing the value of light and inducing wing atrophy. Ecologically, he New Guinean carabid fauna at low altitudes •onsists of roughly one-Uiird mesophiles (ordinary I ground-living forms), one-third hydrophiles (as- ociated witli standing or running water), and >ne-third arboreal forms. Altitude affects the
carabid fauna in several ways: in taxonomic com- position (Agonini become predominant), in num- ber of species (fewer at higher altitudes), in size distribution (see above), in wing state (see above), and in ecologic composition (relatively more mesophiles at higher altitudes), but size of individuals in specific stocks apparently does not decrease with altitude; altitude probably exerts its effects partly directly and partly indirectly, ])y modifying vegetation and limiting areas and habitats. Ants probably compete with and re- place especially flightless mesophile Carabidae at low altitudes in New Guinea.
Under "Zoogeography," existing geographic patterns are first discussed, witli emphasis on problems and procedures. The distribution of Carabidae in the Asia-New Guinea-Australian area is described by tribes, and the relationships of the New Guinean carabid fauna are expressed in "geographic units" (separate geographic re- lationships). The totals are 173 Oriental to 120 Australian units, giving a ratio of approximately 3/2 Oriental/Australian relationships; this ratio holds (roughly) for all ecologic fractions of the New Guinean carabid fauna except in opener country in southern New Guinea, where Austra- lian elements are more numerous. Change of carabid faunas from Asia to Australia is sum- marized; change in ratio of Pterostichini/ Agonini is most striking: tliese tribes are approximately equally represented in the Oriental area (Java), but Agonini are overwhelmingly dominant in New Guinea, and Pterostichini overwhelmingly dominant in Australia; this reversal of dominance has complex historical and ecologic causes. The three principal barriers to carabid dispersal be- tween Asia and Australia seem to have been of different sorts: an old ivater gap between Borneo and Celebes ( Wallace's Line ) , a bottleneck in the Moluccas caused by the relatively small size of the islands, and ecologic barriers between New Guinea and Australia even when there was a land connection. Although New Guinean Carabidae ( and some other insects ) are more Oriental than Australian in relationship, entomologists should not put New Guinea in the Oriental Region but should accept the conventional regions, perhaps adding New Guinea to the transition area ( an extended Wallacea) between tlie Orient and Australia. Within New Guinea, some Carabidae are restricted to the western end or to the southern edge of the island (suggesting that they are recent arrivals from the Orient or from Australia), and patterns of complex differenti- ation of subspecies, species, and genera exist; but no special centers of speciation and no specially important barriers to dispersal at low altitudes within the island are indicated. Endemic genera are concentrated toward the eastern end
i
132 BiiUetin Museum of Comparative Zoology, Vol. 142, No. 2
of the island; tliis may l)e a result of invasion of western New Ckiinea by Oriental stocks arri\ing from the west. Complex patterns of differenti- ation and local radiation occur on separate moun- tain ranges.
The dispersals that have made the existing patterns are discussed under "Dispersals and geographic origins," with emphasis on the prob- able importance of dominance, competition, and extinction, and on complexity. Apparent direc- tions of dispersal of tribes, genera, and species of Carabidae represented in New Guinea are sum- marized. Carabid dispersal in this region is com- pared with that of mammals, and dispersal is related to dominance, size of insects, wings and flight, ecology, altitude, and the relative time of dispersal of different carabid groups. Con- clusions in any single case are tentative, but all the cases together form a pattern of multiple dispersals mainly from Asia toward Australia. Over short distances the preponderance of east- ward and southward over northward and west- ward movements has not been great, but over the route as a whole movements from Asia and the Oriental area to New Guinea and Australia ha\'e apparently been several or many times more numerous than movements from Australia and New Guinea to the Oriental area and Asia. Cara- bids have apparently been coming into New Guinea in numbers for a considerable time. There is no good evidence that arrivals have been more numerous at some times than at others; arrivals have been very numerous recently, but faunal overturns with extinctions may have obscured the evidences of earlier arrival rates. Both Oriental and Australian stocks have probably reached New Guinea at all times, but (except in the relatively dry, open areas of southern New Guinea) in- coming Oriental stocks have apparently always been more numerous than Australian ones re- gardless of size of insects (almost all were small), regardless of wing state (almost all were winged), in all main habitats, and at all altitudes ( most were lowland forms ) . Continual extinc- tions have probably been correlated with the arrivals, and the extinctions ( "withdrawals" ) too have probably tended to begin at the Oriental end of the area and proceed toward Australia. This general history accounts for the nature of the New Guinean carabid fauna as a whole and perhaps for the distrilmtion patterns of "mountain- hopping" groups and for Australian-American dis- continuities in some cases. The trend of dispersals and extinctions from Asia toward Australia fits into an apparent world-wide pattern of evolution of successive dominant groups of Carabidae in the great, climatically favorable area of the main Old-World tropics and of successive dispersals into smaller and/or less favorable areas. And the ap-
parent pattern of o\erturn of the New Guinean carabid fauna itself and the pattern of dispersal of Carabidae from New Guinea to the smaller islands to the east fit the MacArthur-Wilson theory of overturn and faunal etiuilibrium on islands. The New Guinean carabid fauna thus fits into and connects both world-wide and local dispersal and equilibrium patterns in a very satisfying way.
As to evolution, the New Guinean carabid fauna as a whole has evolved from an ancestral accumulation of relatively unspecialized, small, winged, vagile ancestors selected by dispersal across barriers and including relatively numerous species living in water-side habitats; carabids in water-side habitats tend to be vagile, and these habitats are less widely interrupted by climatic factors than rain forest is. The ancestral accumu- lation probably did not have a single starting time but has been an evolving continuum, added to by arrivals and subtracted from by extinctions from time to time. The existing segment of this continiumi is relatively recent; the New Guinean carabid fauna is more recent in its origins than the faunas of tropical Asia, Australia, New Cale- donia, or New Zealand; this may be because faunal overturn has been more rapid on New Guinea. From this changing continuum different Carabidae have evolved on New Guinea to different extents and in different ways; some multiplications of species and ecologic radiations are described. The principal general results have been to increase greatly the number and diversity of both ground-living and arboreal Carabidae in rain-forest, partly by ecologic shifts including shifts from water-side habitats to the forest floor and partly by nuiltiplications of species, and especially to form on the higher mountains a complex alticoline fauna which, in ecology and in superficial adaptations (including wing atrophy), is like the carabid faunas of mountains elsewhere but which consists largely of genera and species apparently derived independently from lowland ancestors on New Guinea. The evolution of the mountain fauna, evolution and adaptation of separate carabid stocks, and radiation of Agonini on New Guinea are discussed in more detail. Evolutionary trends do not include increase of size of individuals on the island but do include trends toward atrophy of wings and associated structures especially on mountains, loss of setae especially on mountains, and modification of legs and tarsi especially on mountains. Parallelism or convergence have occiured among New Guinean' Carabidae not onl>' in atrophy of wings, loss of setae, and modifications of tarsi, but also in modifications of eyes, modification of body form, development of ventral pubescence ( especially in diverse Agonini), development of elytral spines
The carabid beetles of New Guinea • Darlington 133
(which have evolved in 40 apparently separate stocks of carabids in New Guinea), and in some elytral color patterns. These parallelisms and convergences seem to inxolve in different cases parallel adaptations to montane environments (wing atrophy, etc.), adaptations to specific environmental details, mechanical protection (ely- tral spines), mimicry (some color patterns), and occurrence of homologous or parallel mutations (some other color patterns). Occurrence ot dimorphism apparently resulting from mutation among New Guinean and other Carabidae is de- scribed, and the relation of mutational changes to taxonomic characters is discussed.
In the "Taxonomic section," tribal classifications of Carabidae are briefly referred to (but no new classification is offered), and a Taxonomic sup- plement lists important new records and new species. Seventy-three new species are described, most in the tribe Agonini, and most from high altitudes. No new genera are described, but the Oriental Phijsodcra and Omohms (both in tribe Lebiini) are recorded from New Guinea for the first time.
INTRODUCTION TO PART IV
[1]' Purpose; previous parts; ac- knowledgments. This is the fourth and final part of my work on l:)eetles of the family Carabidae of the island of New Guinea. The first three parts (see Bihli- ographij at end of present part) were pri- marily taxonomic. Part I (1962) covered the Cicindelinae ( tiger beetles, which were treated relatively superficially) and the tribes of Carabidae proper from the be- ginning through the Pterostichini in the order of the Junk-Schenkling Coleopt&ro- rum Catalogus (Horn, 1926; Csiki 1927- 1933). Part II (1952), which was published before Part I, covered the tribe Agonini, which is dominant in New Guinea and in which I have a special interest; my intro- duction to Part II went beyond taxonomy to discuss the general nature and evolution of the New Guinean agonine fauna. Part III (1968) covered the remaining tribes of Carabidae, from the Perigonini through
To avoid excessive insertions in page proof, I nave assigned numbers to successive items, and shall use these rather than page numbers in cross references.
the Pseudomorphini. The present part (Part IV) is a general summary, analysis, and discussion of the New Guinean carabid fauna as a whole, with a taxonomic supple- ment. This part can be divided: working copies of "The Carabid Beetles of New Guinea" can be bound with the general portion of Part IV first, then Parts I, II, and III in order, and finally the taxonomic supplement, bibHography, and statement in Heu of index of Part IV. See this statement (p. 33(S) for suggestions to users.
The present part actually begins with a review of material used in my work, of the history of work on New Guinean Carabi- dae, of localities (with new maps), and of my methods of work and taxonomic con- cepts. Then follow analysis and discussion of results of the work from several points of view: numbers of species, size of insects, state of wings, taxonomic composition, ecologic composition, existing geographic patterns, and origin and evolution of the fauna. Because I am a biologist (as every taxonomist should be and many are), the analysis of the fauna as a whole has been, for me, the most exciting part of my work with New Guinean Carabidae. And be- cause I am a biogeographer, the geographic patterns and geographic histories and their significance have been most exciting of all, and I have treated them in greatest detail. I am indebted for careful typing and other work done on the manuscript to Miss Wilmoth Peairs; for drawing done patiently under my direction to Mrs. Mary Catron and Mrs. Sarah Landry; and for support both of the work while in progress and of publication of it, to the National Science Foundation (Grant GB-12.346).
[2] Sources, disposition, and adequacy of material. The principal sources of material used in my work on New Guinean Carabidae are listed in Part II, pp. 90-91; Part I, p. 323; and Part III, pp. 2-3. I have prepared (on cards) a consohdated list of all the museums and other institu- tions and all the persons from whom material has been received, and of the
134 BuUctin Museum of Comparative Zoology, Vol. 142, No. 2
names of collectors, but the list includes more than one himdred items and seems to me not worth the space and cost of publication. The names of pertinent mu- seums, etc., and of collectors are given under the separate species throughout my work. I need say only that I am deeply indebted to the persons there named and to the authorities of the museums and other organizations concerned. And I should add that useful material has been received, but mostly too late to be included in my work, especially from Mr. and Mrs. G. W. Cot- trell and Mr. Fred Parker.
Tlie kind of work that I have been doing with New Guinean Carabidae re- quires and receives international cooper- ation on a scale which persons who are not taxonomists do not always appreciate. I haxe received specimens or information about New Guinean species not only from many sources in the United States in- cluding Hawaii, but also from Canada; England, France, Belgium, Italy, Czecho- slovakia, and Hungary; Japan and Java; Australia (several persons in different cities) and New Zealand; and New Guinea itself and the Solomons. And basic ma- terial or information that has formed part of the background of my New Guinean work has come also from persons in Sweden, Germany, Switzerland, and India.
In general, borrowed material has been returned to the sources from which it was received, with duplicates kept for the Museum of Comparative Zoology (MCZ). In a few cases additional important specimens, including holotypes, have been deposited in the MCZ for safekeeping. I am especially indebted to Dr. J. J. H. Szent-Ivany for the Department of Agriculture, Port Moresby, and to Dr. R. W. Hornabrook for permitting holotypes from their ma- terial or from material in their charge to be deposited in the MCZ. This arrange- ment safeguards irreplaceable specimens and makes them more easily available to specialists, and is a real contribution to
future work on the carabid beetles of New Guinea.
On the other hand, duplicates of my owTi material have been widely distributed. I have tried especially to build up the best possible working collections of New Guinean Carabidae in the British Museum; the Bishop Museum in Hawaii; and with the Commonwealth Scientific and Indus- trial Research Organization (CSIRO), in Canberra, Australia. For further discussion of the place of types and of working col- lections of specimens in modern taxonomy see [6, 7].
The number of specimens of Carabidae proper (excluding Cicindelinae) actually recorded from New Guinea in Parts I-IV is about 22,500. In addition I have seen perhaps 1000 or 2000 additional specimens from New Guinea without counting them ( an exact accounting has not seemed worth the trouble), and of course I have seen many thousands more from the Oriental Region and from Australia (see Part I, pp. 325-328). Most of the specimens from New Guinea were collected during or after the war, and most of them have exact localities, often altitudes, dates, and names, of collectors. About 8000 of the New Guinean specimens were collected by my- self (see [3] ). I have at least a rough idea of the habitats of most of the species that I obtained, and some specimens collected by other persons have indications of habi- tat. Many of the recently collected speci- mens were taken at light; these probably flew, at night. However, there is often no way of kno\\'ing whether individuals taken at light came from forest or grassland or swamps, or whether tht>y lived on the ground or were arboreal.
All this material is adecjuate to show the general nature of the New Guinean carabid fauna as a whole. The material probably includes most existing primarily lowland species and good samples from a few mountains, although hundreds of mountain- living species localized at different levels, on different ranges, and on different peaks
I
The carabid beetles of New Guinea • Darlington 135
of the same ranges surely still remain to be discovered. In all, enough species are known from enough material to allow sig- nificant statistical analysis of the fauna as a whole. But the material of most single species is not sufficient for statistical study of either individual or geographic variation. Much more collecting and much more study will have to be done to make known the variation and geographic distribution of most species even at low altitudes, and the mountain-living species are much less well known. And, although so much still remains to be done on the structure, vari- ation, and distribution of the species, much more remains to be done in the insects' biology. About all that is known of the ecology e\'en of the best kno\\'n lowland species is their gross habitat and whether or not they fly to light, and not even this much is known of most high-altitude spe- cies. And the life histories of most New Guinean Carabidae are wholly unknowm, excepting only Pseudozaena and Morion (Gressitt, 1953) and a few genera and species of "Truncatipennes" which occur also in Japan and for which Habu (1967) gives biological notes.
[3] Preparation for uork on New Guin- ean Carabidae: my collecting. Previous work on Carabidae of New Guinea is briefly described in Part I, pp. 324-328. Points emphasized and worth repeating are the small amount of work done on actual New Guinean specimens in the past, and the importance of work done on Carabidae of adjacent areas, especially by T. G. Sloane on the Australian fauna and by H. E. An- drewes on the fauna of the Oriental Region including the western part of the Indo- Australian Archipelago. Sloanc's and An- drewes' collections (which I ha\'e seen and studied) and the descriptions and revisions published by these two persons are the hasis for study of the relationships and history of the New Guinean carabid fauna.
My own interest in and work on Carabi- dae of the Oriental Region and Australia IS well as of New Guinea, and the col-
lections accumulated at the Museum of Comparative Zoology, are briefly described too in Part I, pp. 325-328. As a result mainly of my own activities beginning in 1931-1932 (when I was a member of the Harvard Australian Expedition), the MCZ now possesses a good working collection of Carabidae not only from New Guinea but also from both major source areas ( Oriental and Australian) from which the New Guinean fauna has been mainly derived.
Although my collecting is summarized in Part I (pages cited above), a few addi- tional details are worth giving here. In collecting around Dobodura I was first struck by the very slow rate at which species accumulated. The first day I found, I think, only three or four species, and few individuals. The next day I found perhaps two or three additional species. And so forth. But when I came to sort out the species and study them at the MCZ, I found that I had obtained 217 species at this one diverse but strictly lowland lo- cality! The slowness with which species accumulated was, I think, due only partly to a temporary physical handicap of mine. It was probably due partly to the fact that, although species are diverse in the tropics, many have sparse populations and (even in relatively small, ecologically homo- geneous areas) patchy distributions (Wil- son, 1958).
My impression is that populations of Carabidae tend to be most sparse and most scattered on the rain-forest floor. Collect- ing in the leaf litter and loose soil on the floor of rain forest does in fact call for patience and ingenuity if the diverse fauna which lives there is to be adequately sampled. Old-fashioned sifting handles too little material to yield an adequate sample of the thinly dispersed fauna in a short time. Berlese funnels are better and are indispensible for collecting some insects, but are probably still too slow to obtain Carabidae in adequate numbers — and I had no Berlese equipment.
The method that I did use, and that I
136 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
think is most effective in rapid sampling of rain-forest-floor Carabidae, is what I call washing or drowning. It involves rak- ing or scraping up large quantities of leaf litter and loose surface soil and throwing them into still water where the light is good. The debris is spread out on the water surface, thoroughly wetted and gently stirred, and perhaps stirred again at intervals. The Carabidae, of course, came to the surface of their own accord, and usually run across the floating debris toward the shore, where they are picked up by the waiting collector. Some take flight and must be caught the instant they appear. Others run down the floating debris again if they are alarmed on the surface. And some, especially some fos- sorial forms, come to the surface only after considerable delay. Sometimes the col- lecting is complicated by small birds and lizards, which wait close by and make dashes for moving insects.
The collector who uses this method for the first time is likely to be disappointed in it. It requires judgment and experience in selecting the right places and in raking up debris in the right way. However, properly used, it yields good collections even from material scraped from ordinary, thin, rain-forest leaf litter. And the yield is sometimes multiplied in washing out piles of leaves that have been concentrated by some natural means, for example under the head of a tree that has fallen and from \\hich the dead leaves have dropped thicklv on the ground. Flash floods in the rain forest, which wash masses of leaves and other debris together, are best of all. provided the collector can \\ork them im- mediatelv, while the water is rising or at least before it starts to fall, before the Carabidae have had time to redisperse and before ants and other predators have had time to decimate them. A flash flood at Dobodura, which brought down the bank of a gully and dammed water back over a piece of flat ground in heavy rain forest, gave me the best collecting I had in New
Guinea. In it, I found (I think) all my Oclontomasoreus humcraU.s (Part III, p. 76) at Do])odura, all my Nototarii,s papua (Part III, p. 186), my only specimen of Cola.sidio papua (present part, Tax. suppL), probably two of my three specimens of Phcropsophus catuhis (Part III, p. 328), and series of other ground-living carabids that I rarely found at other times.
One additional note: at Dobodura, as elsewhere in the tropics, virtually no Carab- idae were found under stones on the ground. Under stones is, of course, where northern collectors first look for Carabidae, and failure to find them there in tropical rain forests has perhaps contributed to the idea that Carabidae are scarce insects in the tropics. But the Carabidae are there in numbers and diversity — just not under stones (see following paragraph).
On the Bismarck Range and Mt. Wil- helm [5] my time was so limited that I got only a skimpy sample of the carabid fauna. I was impressed by the fact that at these altitudes (c. 2000 m and higher), Carab- idae did commonly occur under stones (cf. preceding paragraph). In fact some of the same species that in my experience; were never found under stones at low alti- tudes were found there in the Waghi Valley on th(> Bismarck Range. It therefore seems not that the requirements of tropical Carabidae are different from those of temperate ones, but that the microclimate or microhabitat under stones in the full tropics is somehow inhospitable to most Carabidae.
Of my collections made outside New Guinea (Part I, pp. 327-328), the most im- portant is a comprehensive collection of wet-forest Carabidae made along the wholcj eastern edge of Australia, from northern Cape York to southern Tasmania, during 19 months in 1956-1958 (Darlington. 1960).
[4] Basic literature. Work done on any fauna is carried over from generation to generation in the form of collections, publi- cations, and sometimes unpublished manu-
The carabid beetles of New Guinea • Darlington 137
scripts. Collections of New Guinean Carab- idae available for future work are noted elsewhere [2]. Publications and manu- scripts essential to or resulting from work- on New Guinean Carabidae are listed in more detail in the BihJiopaphy (pp. 334- 337 ) and under many of the species treated in Parts I-IV, but the more important items may usefully be summarized here.
Basic to work on Carabidae in any part of the world are the Junk-Schenkling Coleoptewrum Catalogtis (Horn, 1926; Csiki, 1927-1933) and the Zoological Record. The former (often called the "Junk Catalogue") lists all Carabidae of the world up to within a year or two of the dates given, with very few omissions and not many errors of citation, and with bibliographies and indications of distri- bution, both sometimes incomplete. This work is still available from W. Junk, Pub- lisher, 13 van Stolkweg, The Hague, Netherlands. The annual volumes of the Zoological Record list additional papers and new genera and species published from year to year, and enable students to com- pile preliminary Inbliographies and pre- liminary faunal lists.
Basic works on the Carabidae of the Oriental Region and Indo-Australian Archi- pelago include H. E. Andrewes' papers (see Part I, p. 325) and especially his (1930) "Catalogue of Indian Insects, Part 18, Carabidae," which in many genera lists all species known from the Indo-Australian Archipelago including New Guinea; Lou- werens' papers, especially his (1953) re- vision of Oriental Colpodes; studies by the late Amost Jedlicka and by Akinobu Habu, especially their long papers on Oriental "Truncatipenncs" (see Bibli- ography); and work now in progress by Shun-Ichi Ueno especiallv on Trechini. Also useful to future workers should be my manuscript list of Carabidae of the Indo-Australian Archipelago including New Guinea, based on the Coleopterorum Catalogtis and the Zoological Record, but amplified and brought up to date. This
manuscript is not prepared for publication, but I expect to deposit Xerox copies of it in the Department of Entomology at the British Museum, in the Bishop Museum in Honolulu, and with CSIRO at Canberra in Australia. The original manuscript will, of course, be kept at the MCZ.
Basic work on Australian Carabidae has been done principally by T. G. Sloane ( see Part I, pp. 324-325), who published also two short papers ( 1907 ) on New Guinean species. Work done and being done on Australian Carabidae by B. P. Moore should also be followed by those interested in the New Guinean as well as the Austra- lian faunas. My own papers on certain groups of Australian carabids may be useful in some cases, especially since they are related to my New Guinean work; papers on my Australian collecting locali- ties ( 1960 ) and on transition of wet-forest carabid faunas from New Guinea to Tas- mania (1961) may be especially useful. My manuscript list of Australian Carabi- dae, based again on the "Junh Catalogue" and Zoological Record but amplified by search of all Sloane's papers and much other literature, may save time for later workers; a copy of it will be deposited with CSIRO at Canberra.
Works on Pacific Carabidae that should be considered in studies of New Guinean species include H. E. Andrewes' (1927) paper on Carabidae of Samoa; E. C. Zim- merman's "Insects of Hawaii," especially his introductory volume (1948) and his projected volume on Hawaiian Carabidae; and other work done and being done at the Bishop Museum or on Bishon Museum material, including my paper (1970) on the Carabidae of Micronesia.
As to New Guinea itself, my "The Cara- bid Beetles of New Guinea," of which the present paper is Part IV, coordinates earlier work and should be the basis for future work on New Guinean carabids, especially for third-stage taxonomic study of selected groups (see Part I, pp. 328- 330 ) . Three volumes that will be important
138 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
lOS-E
3rr
ISO-E
ASIA
TRORC
of CANCER
EQUATOR
'1 ..-•-'■ .t.\^^^^
^New Britain t\
15' S
195' E
Solomon t^'' Island's
TROPIC
of CAPRrcORN
j^ New
?>ji Coledonio
FijiV
ISO"^^
Figure 1. Map to show relation of New Guinea to Asia and Australia. Broken lines (at c. 100 fathoms depth) show ap- proximate limits of continental shelves and some other areas of shallow water which were probably land at times in the Pleistocene.
to entomologists working on New Guinean Carabidac and other insects are in press or are planned by J. L. Gressitt and J. J. H. Szent-Ivany (joint authors). The first (now published, 1968) is a bibliography of New Guinean entomology. The second, now being planned, is a history of entomological exploration in New Guinea, with a list of localities. And the third, also being planned, is on the environment of New Guinea from an entomological standpoint. (This in- formation is from a letter from Dr. J. L. Gressitt dated December 26, 1967.)
[5] Localities: Ideally, I should like to map all New Guinean localities at which Carabidae have been collected, but this has proved beyond my power. I can, how- ever, present the following new maps, pre- pared by Miss Sally Babb (now Mrs. Joseph Landry) under my direction. Figure
1 is a small-scale orthographic map de- signed to show the relation of New Guinea to other land areas from southern Asia to northern Australia. Figure 2 is a map of New Guinea as a whole showing general features of the island, some localities, and outlines of limited areas which are mapped in more detail. And Figures .'3-5 are more detailed maps of parts of New Guinea in which important carabid localities are too numerous to show on Figure 2. Many New Guinean localities are spelled in different ways by different authorities, but I cannot list alternative spellings here. Some lo- calities are put in slightly different places on different maps and by different gazet- ' teers; some margin of error should there- fore be allowed for in using my maps. And where detailed localities are too crowded to distinguish, I have shown only general
The carabid beetles of New Guinea • Darlington 139
|
J 1 I o |
■ |
a •* 1 ' |
o |
' .,0 |
|||||
|
Q |
o |
i ^. |
|||||||
|
Z |
_j |
- tr ;1j |
|||||||
|
< |
A |
N_^ |
Q O |
S 8 |
|||||
|
UJ c- |
y |
D |
^^ |
il^ |
|||||
|
/^ |
f |
"i |
^oX |
i i i £ 19? |
5 |
||||
|
o |
( zl |
||||||||
|
-\^ y |
!1 |
r \ |
|||||||
|
i \ UJ / |
Sq |
k ^ |
|||||||
|
s Jz V |
o <^ |
||||||||
|
5 ) |
-^ |
||||||||
|
■ |
(/I o |
° 3 22s II |
/^ ; |
||||||
|
>--! |
A • l^n° J. |
^ |
y-' |
||||||
|
^ |
IT ^1 |
/to t- ' |
■' r |
o s |
■ |
||||
|
/7 |
I, o 0- |
||||||||
|
^ |
// |
Or O Or |
J |
»- |
|||||
|
■* |
X^^: |
a: o |
M |
||||||
|
/ |
^ 5 ,' |
1' |
|||||||
|
^ |
/ |
~ I |
« X |
s?t? |
|||||
|
r |
(Q ' « |
i^ |
|||||||
|
^ / 3 |
ej |
^ |
|||||||
|
1/ |
i>i |
z |
/ S |
Mp |
A = |
||||
|
\r* |
|||||||||
|
1 ^ |
■ 4^ 1 / i? |
p\? |
^_^ |
||||||
|
/ *^ |
2 !/ / |
\ |
f |
||||||
|
Q / |
c E |
^ - V, <? 0 \ o . |
|||||||
|
\ \\r- |
5 c |
-- |
III ^ §^ 1 / |
= 1 ) 1 1 |
■ |
||||
|
„ i-* |
h^ , |
•jj |
/ |
^ |
|||||
|
■? |
< = % |
o •J- |
|||||||
|
1 o |
o \ |
||||||||
|
7* «' - |
/ |
||||||||
|
\1? |
I |
||||||||
|
tfi |
|||||||||
|
> |
2 |
- " jJT^ |
|||||||
|
F o |
• -J' |
UJ _^ ^ ( ^, |
|||||||
|
yo oD |
o > |
o |
o * /^ V |
||||||
|
1 / ^ |
|||||||||
|
ft |
) |
3 |
i > |
||||||
|
r [( |
z |
"1 ' |
' . i. 3£ |
/ o |
|||||
|
^. |
J lO |
/ "^ |
;i'„ |
||||||
|
1.^ |
"' |
^ |
/ r^^n^So |
||||||
|
oE |
^ |
^ |
'Or^ Mr\ |
||||||
|
°ci , |
/-^ |
2-, |
?(r>^^ |
||||||
|
/-^"'^A, |
-^ |
O |
?" |
.'>'* C^IO |
|||||
|
/e- |
<. ?^ |
'■■* |
|||||||
|
o |
"'V |
'tl.ss^ |
% |
||||||
|
T |
(^ _ |
»^-\_^ |
|||||||
|
/ o /o ^ / o UJ |
7 |
15^ |
) |
||||||
|
C f |
t/^ |
m |
|||||||
|
V ^ O |
) |
f |
< <V 5 |
||||||
|
Y^ > |
y |
1 |
:<r • - |
||||||
|
V |
J |
||||||||
|
1 ^/ |
■V. |
||||||||
|
•^' ,i 1 |
el |
o |
I CO
o
E
a
Q.
o
E
D
O ■D
J3 D
E o
o
(U
0)
en
O
<u
Z
C2. O
140 Bulletin Museum of Cotnparntive Zoology, Vol. 142, No. 2
-^
^JCepilam WopenamonO ^'■■^S Mt^Hagen
WESTERN •*"'™' ^°" ^HIGHLANDS
^^Tombul
Jimmi .^ Valley .Tsenqo
0 - 'i'^^Si "-^
Erima-V^STROLASe S4y
<*,„ Mt .Wahelm
<?, ''ANSE
:^e7 '^Zs *wo"°
Ml Gilu««A
■x^^
SOUTHERN
HIGHLANDS
^Mt Murray
Koiapit
^£0 RAN©£
Mt Sdrawdtel
Mt Elandoro Mt Piora
^-^^^.^
Bubio MS«i<^^
HEReoG
Mrs 1 ^"ON
GULF
;;;:- Mumeng Sum Sum
Bulolo*
Mt Missim
, 'S""*"
, (y °" ,MI Chapmon (
Edie Creek AMI ftminqwing
Figure 3. Map of part of eastern Nortfi-east New Guinea and adjacent corner of Papua, sfiowing especially corabid lo calities. Higfi mountain areas (above c. 2500 m) ore shaded, but tfieir limits are only approximate.
Figure 4. Mop of part of eastern Papua and adjacent corner of North-east New Guinea, showing especially carabid lo calities. High mountain areas (above c. 2500 m) are shaded, but their limits are only approximate.
The carabid beetles of New Guinea • Darlington 141
areas: e. g., Wissel Lakes, Cyclops Moun- tains, Sepik River, and Waigeo Island.
The best map of New Guinea for general puiposes is (I think) the current set of USAF Operational Navigational Charts (see Bibliography). For spellings of well- known islands, mountain ranges, and localities I have usually followed Webster's Geographical Dictionary. To find less- known localities in New Guinea I have used primarily the blue, paper-bound gazetteers of the United States Board on Geographic Names (see Bibliography). Also very useful is the "Bishop Museum List of New Guinea localities, 1966," which is a provisional list of localities at which collections have been made by Bishop Museum entomologists and by some other persons. This will be replaced by Gressitt and Szent-Ivany's book (referred to above) on entomological exploration and localities in New Guinea. A shorter list of localities, which is useful because it covers collect- ing done by soldiers during the war, is in Arthur Loveridge's paper on New Guinean reptiles and amphibians (1948: 310-314).
Sources of information about the lo- calities of important carabid collectors include a summary of Miss L. Evelyn Cheesman's work in New Guinea, in Charles Barrett's The Pacific, pp. 63-65. Miss Cheesman spent two and a half years (chiefly in 1933-19:34) in Papua, mostly on the Owen Stanley Range and in the moun- tainous area southwest of the Albert Ed- ward Complex. Then she was about a ^^ear (chiefly in 1936) in the Cyclops Mountains and near Lake Sentani (vicinity :>f Hollandia). And later (in 1938-1939) ?he collected extensively in Waigeo and fapen Islands, and briefly in the Torricelli Mountains. Most of her localities will be found in the Bishop Museum list of locah- ies referred to above.
A source of useful information on limited ireas in New Guinea is the series of re- ports of the Archbold Expeditions in the bulletin of the American Museum of
SNOW MOUNTAIN RANGE
^^t J Scrte Volley
WILHELMINA \ Comp
Figure 5. Part of the Snow Mountain range. West New Guinea, showing especially the localities at which Carabidae were obtained by the Netherlands Indian-American (3rd Archbold) expedition of 1938-1939. Redrawn from Toxopeus, 1940.
Natural History, including items by L. J. Brass (1941 and 1964) on the expedition to the Snow Mountains and on an expedi- tion to Mt. Wilhelm and vicinity and to the Lae-Edie Creek area (vicinity of Wau). Localities of the Archbold Expedition to the Snow Mountains have been published separately by Toxopeus (1940).
See Gressitt and Szent-Ivany (1968) for references to accounts of New Guinea col- lecting by Biro, Cheesman, D'Albertis, Gressitt, MacLeay, Maindron, Szekessy
142 BiiUciin Mitscin)i of Comparotivc Z(X)logij, Vol. 142, No. 2
(concerning Biro), Szcnt-Ivany, Toxopeus, Wallace, Wilson, and others.
I should at this point say something about altitudes. Exact or approximate altitudes of many localities are given on maps, or in the Bishop Museum list, or on locality labels on specimens. However, even modern altitude determinations are sometimes inaccurate, and the altitudes given on maps are not necessarily the alti- tudes at which collecting has actually been done. The risk of error is greatest in the case of material collected long ago. We are not justified in assuming that old speci- mens labeled (for example) Sattelberg came from the immediate vicinity of the town or from the same altitude. They may have been collected many miles away and many hundreds of meters above or below the altitude of the town itself. I have usually omitted old locality records in dis- cussing the distribution of New Guinea Carabidae in relation to altitude [26].
Several localities where especially im- portant collections of Carabidae have been made are \\'orth separate notice here.
Dobodura and vicinity (which includes Oro Bay), Papua, are briefly described in Part I, pages 325-326, and a few more details are given in the present part [3]. It is a strictly lowland area; my collecting there probably did not extend above 200 meters altitude. But the area is otherwise diverse: habitats within easy walking dis- tance of Dobodura included heavy lowland rain forest, \'arious kinds of second growth, grassland, swamp margins, and the banks of streams including a small river, sluggish brooks, and rapid brooks in foothill topog- raphy. Important lowland habitats missing in this area were, I think, only special coastal ones, those associated with really large rivers, and those confined to the open eucalyptus country of southern New Guinea. My collection of Carabidae from Dobodura comprises 217 species [19] and more than 4,000 specimens.
A second very important locality or group of localities is Wau, with Edie Creek,
Mt. Mis(s)im, Mt. Kaindi, etc., in the Morobe District of North-east New Guinea. This is a diverse, mid-altitude area, indicated in Figure 3. Of this area, Gressitt (letter of December 26, 1967) says, "A description of the environment just above Wau is found on pages 182-185 of Volume 127 ( 1964) of the Bulletin of the American Museum of Natural History, by Brass. This is under the heading 'Kaindi, Morobe District.' Part of the area de- scribed in that section is what we call Edie Creek. And most of our material labeled Kaindi is from just above the area de- scribed in that section. Wau proper is just below these areas, at the foot of Kaindi, around the lower part of the Edie Creek Road. Several of our localities like Kunai Creek, Nami Creek, Delias Creek, are along the Edie Creek Road above our field station.
"There is a tremendous range of vegeta- tion types from the bottom of the Bulolo Gorge just below Wau to the top of Mt, Kaindi and to the top of Mt. Missim, the higher mountain on the north side of the Valley (from which MCZ has some old material). Among the main differences ol the lower part of the valley with the ares described in Brass' paper is the fact thai Araucaria is dominant in many parts ol the valley to just about the level of oui field station. Also, palms and many othei tropical trees are in the lower forests, ex tending above the station. In Wau Vallej itself, there is a great deal of coffee grown Also, a lot of the Araucaria has been cu and pure stands of both species have beer planted after destruction of the remaining native forest. Still, there are extensivi areas under natural vegetation. And mon particularly so going higher on the tw( mountains."
Brass's (1964) paper should be referrec to for further infonnation about the Wai area, especially its vegetation, anel fo photographs. Carabidae collected betweei 1000 and 2{X)0 meters altitude in this are now number 170 species, and 35 additions
The carabid beetles of New Guinea • Darlington 143
pecies have been found in the vicinity ither at higher or at slightly lower alti- Lides. This is an outstandingly fine col- jction — a notable accomplishment by the iishop Museum entomologists using their ield station at Wau.
My brief visit to and the collection made n the Bismarck Range and Mt. Wilhelm in )otober 1944 are briefly described in Part , pages 326-327, and Part IV ( the present >art) [3]. A sketch map of my route to it. Wilhelm is in Part I, p. 326, and ►resent Figure 3 shows the position of the lountain in relation to localities at which ther collectors have obtained Carab- iae more recently. Tlie altitude of Mt. Vilhelm, previously considered to be about 5,400 ft., is now considered to be about 4,600 ft. ( c. 4450 m ) . Specimens collected ly me in this area bear three different abels. Those from the lower, more open ountry are labeled "Chimbu Valley, Bis- narck Range, 5,000-7,500 ft.," and were aken in the densely inhabited valley, most I which has been highly modified by man, nd which is now largely grass or gardens, Ithough very limited habitats including he edges of small streams and of the Chim liver are still natural. Specimens taken in he montane forest, which changes from ain forest to moss forest with increasing Ititude. are labeled "Mt. Wilhelm, Bis- aarck Range, 7,000-10,000 ft.," and were aken on the ground in the forest, mostly mder stones and logs or beside running vater. Finally, specimens taken above the orest line, in "subalpine" habitats, are abeled "Mt. Wilhelm, Bismarck Range, hove 10,000 ft.," and were taken either in the ground vmder various kinds of cover acluding tussocks of grass, or beside small treams in tussock-grass country, or ( Macii- igomim oltipox only) in a grass tussock.
or further information and photographs f the Bismarck Range and Mt. Wilhelm, ee Brass (1964) and Brookfield (1966: 79-183). Brookfield (pages cited) sum- larizes the geology as well as the egetation of the area. This range (like
many of the other mountains of New Guinea, 1 think) is described as "a recent fold-structure which was uplifted in a series of stages culminating in the late Tertiary," with complexly faulted strata which include limestone.
The localities in the Snow Moimtains, West New Guinea, at which L. J. Toxopeus obtained his fine collection of Carabidae, are briefly described by him (1940), with altitudes and very brief descriptions of the vegetation. I here reproduce Toxopeus' map, somewhat simplified (Fig. 5). His localities cover virtually all the important montane habitats from cultivated valleys below 2,000 m, through various types of forest at increasing altitudes, to "alpine" areas above timberline. The highest alti- tude at which collecting was done was 4,250 m, but "results were few" this high up.
Evidence that "Dor(e)y" labels have been wrongly placed on many Carabidae, col- lected by Alfred Russell Wallace, that probably really came from Celebes or the Moluccas is given in Part I, pages 330-331, and Part III, page 5. Wallace did collect at Dor(e)y in West New Guinea, but he or someone else evidently labeled as from there many specimens which really came from other islands.
POLICIES AND METHODS
[6] Modern taxonomy. "Modem taxon- omy" means different things to different persons. To me, it means taxonomy as practiced now, and it is worth considering what the policies and methods of taxonomy now are. In general, I think modem taxon- omy, as compared with the taxonomy of one or two generations ago, is more care- fully calculated to reflect real situations in nature and more carefully designed for intelligibility and utility. Modem taxon- omy also employs new techniques and new procedures appropriate to the material and purpose of each particular piece of taxo- nomic work, but the new techniques and procedures are for the most part added to,
144
Biilhtin Museum of Comparative Zoology, Vol. 142, No. 2
not used in place of, the procedures of the past. Taxonomic work has therefore be- come increasingly complex and difficult, but also more precise and more useful. See Mayr (1969) for detailed discussion of the principles and procedures of this kind of work. And sec^ the volume on Systematic Biologij published by the Na- tional Academy of Sciences-National Re- search Council (1969) and current numbers of the journal Systematic ZooIo<i.y for dis- cussion of various aspects of modern taxonomy.
Within my general definition several levels and many different methods of taxonomy are possible. My methods are specifically those of second-stage faunal taxonomy as described in Part I, pages 328-330, limited by the amount and kind of material available from New Guinea and by the amount of time I have been able to devote to the project.
The methods I use are essentially sub- jective: comparison of specimens, detection of similarities and differences, and reach- ing of conclusions based primarily on per- sonal judgment and experience rather than on statistical analyses or other objective tests. I have tried to make my taxonomy conform to phylogeny and reflect the existence and variability as well as the interrelationships of populations in nature. For practical purposes, Carabidae have no fossil record. Their phylogenies must therefore be deduced, and taxonomists are sure to make mistakes about them. Never- theless the results of the kind of work I am describing have a gross phylogenetic reality perhaps best demonstrated by its utility: zoogeography based on this kind of taxonomy makes sense, and it would not be expected to make sense if the phylo- genetic basis of the taxonomic work were seriously in error.
The methods I use are, of course, essenti- ally the same as those used by Darwin and by a multitude of taxonomists before and after him, although his predecessors were not aware that their classifications reflected
phylogenies. The method is fundamentally simple. Different kinds of animals (or of plants) are compared; characters they share are assumed to be primitive (unless there is evidence to the contrary); and characters by which they differ are as- sumed to be derivative and to indicate di- vergent lines of evolution. Of course use of these basically simple criteria can be- come very complex in practice, as a result of parallelism and convergence, loss of structures, and other complexities and ir- regularities in the phylogenies of complex groups. In fact, classifications almost al- ways extremely oversimplify phylogenies. This is a necessary result of the complexity of evolution, and it does not spoil the usefulness of the classifications, if the latter are consistent with phylogenies. In modern terms (Mayr, 1969) classifications like mine are phyletic but not cladistic, and ol course not phenetic. In a few cases, especially among some Agonini that have radiated on New Guinea, I have been unable to recognize phyletic lines and have therefore temporarily classified the insects by grades — my "genera of convenience' [9] are of course grades.
Hennig (various publications, culminat ing 1966) and others (notably Brundin 1965; 1966) have proposed a formal systen of "phylogenetic" (cladistic) systematics Their methods are not fundamentally dif ferent from those that have always beei and still are used by most taxonomists The cladists, like the rest of us, look fo similarities and differences among tin organisms they study, and make subjectivi judgments about relationships and diver gences and about primitive and derivativ characters on the basis of what their com parative studies show. Unless their sub jecti\e judgments are of a different orde from other persons' judgments (of cours they are not!), their fundamental method, are no different from and no better tha other persons' methods. It seems to in extraordinary that they do not seem t realize this fact! However, the cladisi
The carabid beetles of New Guinea • Darlington 145
then force their classifications to fit an in- flexible hierarchy of taxonomic categories determined by the (supposed) sequence of branching points of the phylog- enies they have worked out by conven- tional means. The new methods (so far as they are new) and the new terms are (I think) not useful but merely make taxon- omy rigid and unadaptable. Among the assumptions that I object to in "cladism" are that evolution proceeds at a fairly con- stant rate, that phylogenies are fairly regu- lar successions of dichotomies which the cladists can reconstruct and date even without a fossil record, and that primitive ("plesiomorphous") characters are rather easy to recognize. Criteria of primitive- ness have, in fact, been known since Dar- win, but are often more difficult to use than Hennig and Brundin seem to realize. Even more dubious is the assumption by at least some cladists that there is a simple correlation between primitiveness and geo- graphic distribution and that the most primitive existing forms of any given group persist at the group's place of origin. Do persons who make this assumption suppose that primitive stocks cannot disperse, or that evolution is suspended at places of origin? Among Carabidae this whole con- cept is made ridiculous by the fact that possession of wings is primitive and atrophy of wings derivative, but that it is the primitive (winged) stocks that do in fact disperse most readily. Of course the matter is really much more complex than this, but I cannot take space to discuss it further here. In the present work I reject the idea that the places of origin of widely distributed groups of Carabidae are re- vealed by the present occurrence of primi- tive forms. Other methods of tracing geographic histories are briefly discussed in Section [74]. In general, I think taxon- omists using less arbitrary methods can do as well or better than cladists in fitting their classifications to phylogeny and in interpreting geographic patterns. For a more detailed practical criticism of Hen-
nig's and Brundin's ideas and procedures, see Darlington, 1970a.
Numerical taxonomy is not adapted to the kind of work I do. It requires more and better material and much more time than I have. And I would not use it in any case unless I thought (I do not) that the results would be more useful than the methods I do use. Of course some nu- merical taxonomists do not even pretend that their classifications are phylogenetic.
To return to modern taxonomy, it is worth some further consideration of what present-day taxonomy is, without the arti- ficial rigidity of Hennig's system and without the aberrations of numerical taxon- omy.
One important characteristic of modern taxonomy is that very much more material is used than was usually available in the past. For example, when Sloane (1907) wrote on Carabidae from New Guinea, he had only 30-odd specimens from that island and New Britain, while I have had about 24,000 specimens from New Guinea alone [2]. In fact I have had to cut off the flood of incoming material in order to finish my work at all!
Another way in which modern taxonomy often differs from that of the past is that taxonomists themselves are more likely to have collected substantial amounts of the material they study. For example, Sloane was never able to collect in New Guinea, and Andrewes, who worked so extensively on the Carabidae of the Oriental Region and Indo- Australian Archipelago [4], col- lected only in India and only a handful of specimens there, while I have been fortu- nate enough to collect about one-third of the carabid specimens I have studied from New Guinea, including individuals of more than half the species. Tlie basing of work on large amounts of material, substan- tial parts of it collected by the taxonomists themselves, gives the latter first-hand knowledge of populations in nature and of the variation, ecology, and distribution
146 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
of species, and surely should contribute future when only one specimen of a species
to realistic, useful taxonomy. is known, even though unic^ue types are
Modern taxonomy, even the nonnumeri- now loaned more often than they used to
cal kind, does include methods that ( added be. I should add that, in faunal works on
to the old methods ) improve the procedures remote areas like New Guinea, descriptions
and results of taxonomic work. For ex- of species based on single specimens are
ample, statements of proportions are now still often necessary. To put off the work
usually based on careful measurements until more material becomes available
made under the microscope. Proportions, might well mean that the work would
such as the relative widths of head and never be done. And of course the chance
prothorax and the ratio of width to length of more material being obtained is greatly
of prothorax, were often simply estimated increased when descriptions are published,
or guessed at by taxonomists working on so that collectors can know what species
Carabidae two or three generations ago. to look for in what localities. To return to
New and useful kinds of characters have more ordinary cases, I think that descrip-
been found and utilized in carabid taxon- tions should be short, partly because com-
omy, for example presence or absence of plete descriptions are impossible, partly
specific setae on many parts of the body, because very long ones waste time both
the patterns of surface microsculpture as in the writing and in the using, and partly
seen under the microscope at 50 X or because the cost of publishing very long
100 X magnifications, and the state of the descriptions is excessive. But, to compen-
inner wings. Use of genitalic characters sate for the shortness of descriptions, I
too has advanced our understanding of think that it is an important function of
the natural classification and phylogeny modern taxonomists to distribute working
of Carabidae, although I have not been sets of specimens and to tell readers where
able to make much use of genitalic char- the sets are. Working sets of my New
acters in my work on New Guinean cara- Guinean Carabidae are being distributed
bids (see [13]). as described [2]. ,
One other characteristic of modem [7] Types. In this context, of increasing taxonomy seems to me to be that emphasis emphasis on animals rather than descrip- has shifted from descriptions to actual tions, type specimens remain decisively im- specimens, or from words to animals. De- portant. They are necessary to stabilize scriptions cannot be made full enough and nomenclature. (If names of genera and accurate enough to satisfy later workers, species have not been stabihzed, this is Each generation of taxonomists must see more often a result of bad judgment by the actual specimens used by earlier taxonomists than of failure of the existing generations, and I think the tendency now code of nomenclature — but this is a deli- is, or should be, to make descriptions short, cate matter which cannot be adequately but of course explicit and carefully calcu- expounded here. ) And, more important, lated, and to make specimens widely avail- types (which are animals warranted to be able. This is facilitated by the very large properly identified ) are the best means of amount of material now often available assuring that in the future biologists will (see third paragraph above). 1 have know (or can know, if they want to take tended to make descriptions shorter and the trouble) what animals biologists are shorter during the course of my work, and talking about now. Types or equivalent . I have also varied their length according "voucher specimens" are therefore as im- to the number of specimens that I have portant to ecologists, behaviorists, geneti- for distribution. Relatively long descrip- cists, and physiologists as to taxonomists. tions are most likely to be needed in the So far as New Guinean Carabidae are |
The carabid beetles of New Guinea • Darlington 147
concerned, the great collections of older types are in the London and Paris Museums, while the MCZ possesses the largest collection of more recent types of these particular beetles.
Taxonomists who designate and use "types," and who sometimes have to work with limited material or single specimens, are sometimes accused of being "typol- ogists," but I think this is a confusion of terms. The original typologists were Greek, and they dealt in abstractions rather than reality. Their types were idealized per- fections never attained in reality. But the types of taxonomists are real specimens that can be put under the microscope and described, and put under it again and the details checked. I do not know any taxon- omists now who really treat types as abstractions or who do not understand that types are individuals which represent populations but uhich do not show all the populations' characters. If there are typol- ogists in taxonomy now, I think they are the mathematical biologists who take limited samples of populations and from them derive mathematical formulae in- tended to represent the characteristics and limits of the whole populations. The formulae are abstractions which cannot be wholly correct. Samples are not likely to show all the variation of whole populations. And formulae based on samples are often extrapolated unjustifiably, as if variation were continuous, although in fact much variation is discontinuous. Mathematical typologists produce idealized models which are useful for reference but which should not be confused with reality.
[8] Taxon concepts: subfamilies and tribes. My concepts of subfamilies, tribes, genera, species, and subspecies are worth brief discussion, not so much to defend them as because readers should know my usage in order to understand my taxonomy.
In the case of subfamilies, I have fol- lowed the Coleopteroritm Catalogus (Horn, 1926; Csiki, 1927-1933) as a matter of practical convenience, because this is the
latest arrangement that covers the Carab- idae of the whole world. However, the great subfamilies Carabinae and Harpa- linae are surely unnatural ( polyphyletic ) and should be ( and by some recent authors have been) broken up.
The tribes now recognized [18] are, I think, mostly natural, although many de- tails are undecided. The number of tribes to recognize is a matter of utility and intelligibility rather than of fact. It seems to me that the number should be small enough, and the names and limits of the tribes stable enough, so that most of them will be recognized by most students of Carabidae everywhere. My use of tribes is therefore conventional, approximately the same as that of Andrewes, Sloane, and Ball [101]. Jeannel (1940-1941) has, I think, split tribes too much. He has divided what most of us call the single family Carabidae into about 50 separate families, many of which are further divided into subfamilies and tribes. I do not mean that the splitting is wrong in itself. The old tribes can and should be subdivided when natural characters can be found to do it. But the subdivisions can be treated as subtribes or groups of genera. This allows specialists to refine classifications without imparing their intelligibility,
[9] Genera. The genus is not a natu- rally limited entity. It should conform to phylogeny when possible (see "genera of convenience," below), but natural (phylo- genetic) genera can be broad or narrow, and the decision how broad to make them is often a matter not of fact but of intelligibility and utility. In general, I think genera should be broad, with names and limits as stable as increase of knowl- edge will allow, but large genera can often usefully be divided into subgenera.
For example, among tiger beetles (Cic- indelinae) I think it is useful to recognize the immense genus Cicindela in the old- fashioned sense (see Part I, pp. 340ff). This genus then includes about half the existing tiger beetles, and it is world-wdde
148 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
in distribution and known by name to entomologists everywhere. What is writ- ten about it by different persons on differ- ent continents is immediately intelligible to persons on other continents, and papers published on the biology of members of the genus are easily catalogued and can easily be found by interested persons everywhere. Also, some aspects of the broader zoogeography of these beetles are best understood if generic lines are broad. On the other hand, specialists working on Cicindelo know that the genus can be divided into natural groups, many of them confined to single continents (see Rivalier, 1950-1963), and recognition of the groups as subgenera helps specialists catalogue the species, indicate phylogenies, and de- scribe local distribution patterns.
For another example, CJilaetiitis (Part III, pp. 20ff) is a huge, world-wide genus of 700 or 800 or more known species. The genus as a whole seems natural (mono- phyletic), and although it can and should be subdivided, the taxonomic level of the subdivisions should be determined by utility and intelligibility. Chlaeniiis too, like Cicindela, is known to many entomol- ogists who are not specialists in Carabidae, and there seems much to lose and little to gain by splitting it into many small genera with new and unfamiliar generic names. Some of the small genera segregated from Chlaeniiis by recent authors have names that even I do not recognize, although I am a specialist in Carabidae! Here again recognition of one genus for general intelli- gibility, and division into many small sub- genera for the utility of specialists, seems the most useful taxonomic treatment.
Those who dislike subgenera may of course use species groups instead.
As to characters which distinguish genera, I know none which, of itself, is always of generic value. I have discussed this fact in detail for the Agonini (Part II, p. 105), and it is true of Carabidae in general. The primary criterion which I have tried to use is actual relationship as
shown by a sum of characters. In practice, in cases in which I have been forced to make new generic classifications, I have usually treated as genera groups of species which seem to be closely related among themselves but much less closely related to other species, and which share at least two distinctive characters, one of which may be a distinctive form or appearance. In some cases I have given weight to continuity of variation; that is, I have included in one genus species which are very unlike if the differences between them are bridged by a series of intermediate species. An example of this is discussed in notes under the genus Nebriagomim (Part II, pp. 236-237). Fortunately, except in the Agonini (for which see Part II), most New Cuinean carabids can be as- signed to previously known, often well- known genera. In fact I have been forced to describe only nine new non-agonine genera from New Guinea (cf. [66] and Table 13).
Tlie criteria briefly outlined above fail in some cases, especially among some Agonini that are apparently actively evolv- ing and diversifying in New Guinea now. In these complexly evolving assemblages, in which relationships are obscured by absence of well-defined group characters and by probable occurrence of much parallelism (tor example in loss of setae), I have found the most useful treatment is to make "genera of convenience," each containing a number of species which have certain common characters but which may not be directly related among themselves, Three such genera of convenience are used for New Guinean agonines: Notagonum Colpodes (in a restricted sense), and Alia- gonum (see Part II, pp. 127ff, 158ff, and 185ff). Of course genera of convenience are temporary, and should be abandonee when the real relationships of the specie;, concerned have been elucidated by third stage taxonomic work.
[10] Species. In contrast to the genus the species does have natural limits, a
The carabid beetles of New Guinea • Darlington
149
least in theory and probably also often in fact. So far as it can be simply defined, a species is a population or group of inter- breeding populations which is reproduc- tively isolated from other populations. I cannot here discuss all the practical dif- ficulties and complexities that taxonomists find in applying this definition, but I must mention a few that concern New Guinean Carabidae.
New Guinea is an island. On relatively small islands, for example on some of the West Indies where areas are smaller and where species of Carabidae are fewer than on New Guinea, specific populations on a given island seem to be better defined and easier to distinguish than species of the same genera on continents. The chief dif- ficulty on such islands may be to decide whether slightly different populations on different islands should be considered sub- species or species. Some taxonomists, applying the criterion of reproductive iso- lation strictly, and assuming that genetic exchange does not occur among the island populations, call every island population a species, no matter how slightly defined it may be. But I think it is more intelligible and more useful to treat slightly differ- entiated populations on different islands as subspecies, for this seems to me to clarify distribution patterns and relation- ships.
However, New Guinea is so large, and the species of Garabidae on it are so numerous, that the situation is more liTke that on a continent than on a small island. Some carabid species apparently are fairly uniform over the whole of New Guinea, but geographic differentiation of many other species occurs in different parts of the island even at low altitudes, and very extensive radiation of species has occurred on New Guinea in some genera [92]. especially Demeirida (Part III, pp. 140ff). Even in fairly simple cases it is not safe to assume that all New Guinean individuals of a species form one population, which can be compared with and differentiated
from (say) all Australian individuals of the same or a related species. An illustra- tion of this point will be found in Notes under Pheropsophits verticali.s (Part III pp. 236-237).
The difficulty of distinguishing species of Carabidae in New Guinea has been increased by the nature of available ma- terial. In spite of the large total number of specimens [2], the representation of many species is still inadequate, and the specimens are sometimes in poor condition. In general, in treating the more complex situations on New Guinea (as on conti- nents), I have followed what I have re- ferred to elsewhere (Part III, p. 146) as a rule of the trade: a taxonomist doing an extensive piece of work must do the best he can with the available material in the available time, and leave details for third- stage taxonomy later.
Although, because of the large size of the island and the large number of species of Carabidae on it, situations on New Guinea as a whole are often complex and difficult to analyze, local situations are more obvious. (This is true on continents too.) At Dobodura, for example, although a number of species of Notag,onum oc- curred there (Part II, pp. 127ff), the spe- cies were almost all clearly different from each other at that locality, and different related species often occupied different habitats. Tlie difficulty has been to decide how slightly-different populations in other parts of New Guinea are related to the Dobodura populations.
In summary, my species are primarily subjective, but my subjective decisions are based not only on comparison of characters visible in museum specimens but also to some extent on occurrence of the insects in the field, and I think they conform reasonably well to reality.
[11] Subspecies. The subspecies is now usually defined, in zoology, as a recogniz- able geographic population. I have dis- cussed subspecies in connection with New Guinean Carabidae in Part I, pages 331-
150 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
332, and Part II, pages 111-112. They seem to me to be populations which have begun to diverge and which are potentially capable of becoming species, although many or most subspecies probably never do become species. I think it adds to the intelligibility and usefulness of taxonomic work to recognize subspecies, in moder- ation.
Three kinds of situations occur in New Guinea that can be expressed by trinomials — by use of subspecies. First, a population that is spread over the whole of New Guinea may be slightly differentiated from related populations on other islands. An example is Chlaenius himaculatus pon- graczi (Part III, p. 27). Second, a species that is widely distributed on New Guinea may occur in a habitat that is discontinuous on the island. Such a species may be broken into slightly different, geographi- cally isolated subpopulations or subspecies on different mountain tops, or on different river systems, etc. The subspecies of Litha<ionum annuUconie (Part II, pp. 176- 180) illustrate such a pattern of differenti- ation on different river systems. ( However, acquisition of more material from more localities may show this kind of pattern to be too complex for subspecific treatment within the limits of New Guinea.) Finally, a species may be widely and more or less continuously distributed on New Guinea but may vary from locality to locality. Probable examples are found in AJtagonum vallicoki (Part II, pp. 190-191) and A. grossuhim (Part II, pp. 191-193). (In these cases, too, acquisition of more material is likely to show that, even within the limits of New Guinea, situations are too complex to be reduced to simple systems of sub- species.) I have used subspecies in all the three kinds of situations described and exemplified above. But I have used them sparingly, only when I think they really do clarify situations, or sometimes as a stratagem to emphasize that geographic differentiation occurs even though I am not sure subspecific treatment will be best
in the end. I have in fact recognized sub- specific differentiation within New Guinea in only 16 of the 667 full species of Carabidae on the island (cf. Table 12 in section [66]).
[12] Nature of taxonomie characters. It is a good working taxonomic principle that no characters are inherently of generic, or specific, or subspecific value. Characters used by taxonomists are significant not in themselves but because, in particular cases, they do in fact characterize what seem on the sum of all evidence to be natural and useful genera, or natural species, or useful subspecies. Key characters, including many of the characters used in my keys to New Guinean Carabidae, are therefore pri- marily tags for the identification of taxa which have other, fundamental but less obvious bases.
For example, the inner wings of Carab- idae may be either fully developed on vestigial. Carabid taxonomists (including; Sharp, when he wrote the carabid volume of Fauna Haicaiiensis about 1900) formerly assumed that atrophy of wings must be at profound evolutionary process and that' winged and "wingless" carabids must goi in separate genera no matter how similar the insects might be otherwise. Now we know that wing atrophy begins by muta- tion from a long-winged to a short-winged condition, that fully developed and vestig- ial wings may be inherited in simple Mcndelian fashion, and that long- and short-winged individuals of a species often occur together in nature [21]. Under these circumstances the taxonomic significance of state of wings depends on the situation in each case.
In some cases, all known species of what appear (on the sum of other characters) to be natural genera have either fully de- veloped or atrophied wings, and state of wings is then useful in defining and. recognizing genera. Among apparently natural genera of New Guinean Agonini (Part II), for example, Plica<!,ontim, Irida- gonum, and Maculagonum (and others)
The CARABiD BEETLES OF New Guinea • Darlington 151
are always fully winged; Idiagonum, Nebriagonum, and Laevagonum (and others) are always vestigially winged; and only Gastragontim is, as a genus, dimor- phically winged. (I have omitted genera of convenience and doubtfully natural genera from these lists.)
In other cases, within what appear to be natural genera, some species are always (so far as known) fully winged while other species always have atrophied wings, and in these cases state of wings is useful in defining and recognizing species. For example, some species of Gastragonum (Part II, pp. 222ff) seem always to have fully developed wings, others always atrophied ones (although one species of the genus is known to be dimorphic), and in Clivina, although most are fully winged, two New Guinean species (toxopei. Part I, p. 36, and kuhor, see Taxonomic supple- ment) probably always have atrophied wings (but tvvo others are dimorphic). Both long- and short-winged species occur in New Guinea also in the genera Lesticus, Tachijs, Oodes (of the terrestris group), Scopodes, and Fheropsophus (and di- morphically winged species too occur in all these genera, except perhaps in Pherop- sophus.)
In still other cases, what appear to be different geographic populations of single species differ in wing state, and in these cases state of wings becomes useful in defining subspecies. Examples among New Guinean Carabidae are Clivina dedlata (Part I, pp. 372-374) and Tachys serrula (Part I, pp. 408-409).
Finally, in a few species in New Guinea (and in many more species in some other parts of the world ) long-winged and short- winged individuals occur together, in single populations, and in these cases wing- state has no taxonomic significance. Ex- amples among New Guinean Carabidae include Clivina crugatella, Tachys avius, Gastragonum terrestre, and others [21].
These examples show that state of wings of Carabidae has no inherent taxonomic
value, but can be used to characterize genera, species, or subspecies in appro- priate cases. The same generalization can be made about presence or absence of setae. Certain setae, especially those over the eyes, on the prothoracic margins, and on the third intervals of the elytra, are very useful in carabid taxonomy. However, the setae, like the wings, are often lost ap- parently by mutation; this is indicated by the fact that some species of Carabidae are dimorphic wdth respect to setae [100]. So, although presence or absence of par- ticular setae often characterizes genera or species, each case has to be carefully ex- amined to determine what taxonomic value (if any) the setae really have. And when it is found that setae do in fact characterize genera or species, it must be remembered that they are not inherently significant but just happen to be useful taxonomic tags in the cases in question. The effects of mutation and dimoqDhism on certain taxo- nomic characters are further discussed in [100].
[13] Secondary sexual and genitalic characters. Although as a general principle no taxonomic characters are inherently sig- nificant at particular levels, experience shows that certain classes of characters are more likely than others to define broad groups. This is true of characters drawn from secondary sexual structures and from the male genitalia.
Secondary sexual characters, among Carabidae especially the form and cloth- ing of the male front (and often middle) tarsi, are likely to characterize groups of genera. For example, in the tribe Pterosti- chini a group of genera centering on Loxandrus (Part I, pp. 549ff) is character- ized partly by having the male front tarsi obliquely dilated. In the tribe Harpalini (Part Ili, pp. 38ff, esp. Key on pp. 40-41) the clothing of the male tarsi (whether dense pads or double rows of squamae) helps divide the tribe into what seem to be natural groups. And in many other Carabidae the male tarsal clothing sug-
152 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
gests the relationships of difficult genera. However, use of this character is limited by the fact that the tarsal clothing is often lost, the male tarsi being then secondarily simple.
Characters drawn from the male geni- talia are more diverse and even more likely to be broadly significant. Their significance varies. The form of the organs as a whole and also the form of the parameres often yield characters of tribal or even sub- familial value. On the other hand, the form of the apex of the middle lobe is likely to yield characters of only specific value, and in some cases the apex varies so much that it can hardly be used in taxonomy at all. Characters drawn from the armament of the internal sac are often intermediate in value; they often define not only single species but also groups of species within genera.
Although genitalic characters are useful and often decisive in carabid taxonomy, and although no group of Carabidae can be considered thoroughly worked out taxonomically until genitalic characters have been utilized or at least considered (they fail in exceptional cases), I have not attempted to use them in my work on New Guinean Carabidae. Cenitalic characters (like other characters) do vary, and the extent of variation cannot be predicted. Many specimens throughout the range of each species must be dissected before the amount and distribution of variation can be determined. For example, the apex of the middle lobe is obviously different in selected individuals of Trichotichmis alius and T. dux (Part III, p. 250, Figs. 172, 173), but proper use of this character to distinguish these and related species would require dissection of several hundred in- dividuals. This kind of study belongs to third-stage taxonomy. I have not time to do it properly. To use genitalic characters improperly, dissecting only a few individu- als, too often leads to overemphasis of slight differences and the making of far
too many "species" based on characters of no real value.
Although I have for the most part ignored genitalia in my work with New Cuinean Carabidae, I have figured them in special cases, especially for the type species of new genera and for a few im- portant new species of which my material is too limited for wide distribution. In these cases genitalic characters are given to aid in placing the genera and species, not to distinguish the species from related ones.
[14] Freserxjution of material. Discussion of genitalic characters leads to consider- ation of methods of killing and preserving Carabidae. In specimens killed dry, for example in ethyl acetate, the genitalia are not everted and the internal sac is usually fully retracted. The complex patterns of setae and bristles foimed by the armament of the sac are then comparable from speci- men to specimen. But my New Guinean material was killed and preserved in alcohol, and in alcohol-killed specimens the internal sac is often partly everted but not wholly so, and the spine-and-bristle pat- terns are confused and cannot be compared from specimen to specimen. This is a very serious disadvantage of alcohol-killed ma- terial, and is a secondary reason for my not attempting to use genitalic characters ^ in my work on New Guinean Carabidae.
I should add that I was forced to use alcohol in New Guinea, and that it has positive advantages as well as disadvan- tages, especially in the tropics. Properly pre- served in alcohol, the specimens are safei from mice, insect pests, mold, and rotting. The mouthparts and surfaces are clean and ready for examination with a minimum of trouble after the material is mounted. And killing and prescr\'ing in alcohol re- quires a minimum of time in the field, when time is critical. These advantages of alcohol exist only when alcohol is properly used. It should be grain alcohol of about 70 per cent. Not too many speci- mens should be put into one vial; the
The carabid beetles of New Guinea • Darlington
153
specimens will be distorted and may rot if packed too closely. And if the alcohol in which the specimens are killed becomes diluted or greasy, it should be poured off and fresh 70 per cent alcohol poured in.
Every good set of instructions for col- lectors emphasizes that each vial or other unit of specimens should be clearly labeled immediatehj with locality, date, and col- lector's name. Field numbers should NEVER be used in lieu of this minimum information. Numbers can legitimately be added to the basic data to refer to note- books with additional information too ex- tensive to put on the field labels.
[15] Toxonomic methods and pro- cedures. My work is second-stage faunal taxonomy as described in Part I, pages 328-330. My methods are deliberately de- signed to allow me to cover the whole carabid fauna of New Guinea at this taxonomic level within a reasonable time. Actually, my study of New Guinean Carab- idae has been spread over more than twenty years, although this work has been interrupted for long periods from time to time while I did other things.
My methods of drawing descriptions are described in Part II, pages 92-94; Part I, page 330; and Part III, pages 3-4. The descriptions follow a generally consistent model but are flexible in detail, and I have varied the detailed treatment to fit the importance or interest of each group. I have treated the Cicindelinae (Part I, pp. 330ff) comparatively briefly, because they are outside my usual range of interest, and I have treated the Agonini (Part II) It greatest length because of my special interest in them and because they have radiated most extensively on the mountains 3f New Guinea. Other tribes have been [?iven something like average treatment. For reasons given elsewhere [6] I have nade my descriptions shorter and shorter hiring the course of my work, because I think that future work should be based 3n re-examination of specimens more than 3n descriptions.
For reasons given in Part I, page 4, I have not attempted to see the types of all previously described species of Carabidae that occur in New Guinea. However (under a fellowship of the John Simon Guggenheim Memorial Foundation — see Part II, p. 91), I have been able to study the types and other material in the British Museum, including the H. E. Andrewes Collection, which contains specimens com- pared with most of the older types of Oriental Carabidae in European museums, and recently (in March, 1968, as part of work done under National Science Foun- dation Grant GB-93) I have been able to examine also pertinent material in the Museum National d'Histoire Naturelle in Paris, including types of Dejean and Chaudoir in the Oberthiir Collection.
My statements of proportions are based on actual measurements made under a stereoscopic microscope. As I have said before but cannot repeat too often, pro- portions cannot be satisfactorily estimated by eye. The proportions given in my descriptions are usually those of an average-looking male and female, and the specimens actually measured are usually specified in a separate paragraph, Mea- sured specimens, although this paragraph is omitted in descriptions based on only one or two individuals. In practice, in making identifications, I keep a slide rule on my desk and calculate proportions on it as I use keys or make comparisons with descriptions.
My statements of total length cover the entire size range of each species. Total lengths have usually been read di- rectly from a millimeter ruler set beside reasonably straight specimens under a stereoscopic microscope. Since length of individuals depends partly on position at death, I think nothing is gained by trying to state it more precisely.
The outline drawings have been care- fully prepared as described in Part I, page 4, and are intended primarily to show form, w hich is verv difficult to describe in words.
154 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
The eyes, the shape of tlie prothorax and especially of its posterior angles, the shape of the elytra and especially of their apices, and the positions of supraocular, lateral- pronotal, and dorsal elytral setae or punc- tures are shown accurately and have been checked by me. Other details are semi- diagrammatic: the mandibles, antennae, legs, for example, are indicated only to show the general appearance and "build" of the insects and are not accurate in detail. A drawing with these characteristics can be made in an hoiu' or two. A drawing of a whole carabid accurate in all details would require at least a day or two of the time of a highly skilled artist, and would require very time-consuming checking by me. So, my outline drawings, like my descriptions, are calculated for present purposes, and have been held within the limits of the time and funds available.
My actual procedure in attacking taxo- nomic problems is exemplified by the very difficult problem of New Guinean Detne- trida, as described in Part III, pages 145- 146. The method is to alternate between the general and the particular: first to sort individuals into apparent species in a general way and to make a preliminary key, then to draw detailed descriptions of each particular species to determine its characters and variation, then to make an improved key and a tentative classification, then to check characters again species by species, etc. This is what Hennig (1966: 21) calls the "method of reciprocal illumi- nation," and it is the method that taxon- omists always have used and always should use. This method leads to continual im- provement, never to perfection. Perfection is not attainable in taxonomy.
In selecting new specific and generic names, I have used especially Roland W. Brown's (1956) Composition of Scientific Words. I have tried to keep new names short, and (I am not a purist!) I have occasionally shortened them by informal elision, by dropping out one or more syllables of too-long words. I have not
thought it necessary to explain the deri- vation of new names, except in a few cases. Every working taxonomist should have Brown's volume or something like it, and should be able to find the meanings of new names in it. I have also used as specific names the names of appropriate localities, as nouns in apposition.
[16] Data sheets. As a basis for sum- marizing and analyzing the New Guinean carabid fauna, I have prepared a set of 30 data sheets (Fig. 6) on which are listed all the Carabidae proper (excluding Cicindelinae) now known from New Guinea. Sets of these sheets will be de- posited at the British Museum, the Bishopi Museum in Honolulu, and CSIRO in Can- berra, Australia.
On these sheets, the first regular column lists the names of all New Guinean Carab-^ idae treated in Parts I-IV, arranged in* taxonomic order, with species recorded or' described in the Taxonomic supplement oh Part IV inserted in their proper positions. The "n's" in the left-hand margins of the sheets indicate which species are described as new.
The second column ("No.") gives the total number of specimens of each species or subspecies actually recorded from New^ Guinea in Parts I-IV.
The third column gives the mean size of each species to the nearest 0.05 mm.
The next column ("Wings") indicates,! by signs described in the text [21], whether' the wings are fully developed, reduced, or dimorphic.
The column "Ecology" indicates the gross habitat of the lowlwul species, so far as habitats are known or can reasonably be deduced. "Meso" indicates found on the ground not associated with surface water; "Hydro," associated \\'ith surface water (streams, or swamps, or other wet places); "Arb," arboreal. Additional words- in some cases indicate habitat more pre- cisely: whether arb(oreal) species occur on trunks or in foliage, which meso(philes) occur chiefly in wood debris or (rotting)
The carabid beetles of New Guinea • Darlington 155
|
!!«m« |
So. |
SI IS |
Wings |
Ecology |
End- solo? |
Coogr. relationships |
Low/ mt-s |
By ae |
Bis'k Rga. |
Wau ato. w. »U. |
||
|
/-■»,kj|.tT*',-J ^.^J",r |
J |
y-r; |
/ |
A/..'.. |
//^ |
(5.-..:'./, ... |
/,.-, |
,->/.„ |
||||
|
1 |
iO-, |
-♦ |
t: |
o>--,. •/ |
/' ■ ■ |
^'/.. ">j'.-. |
;r |
X.' 1 |
||||
|
f^t, :'^ir:-^.-- f'J.Jor- |
^■0 |
J.vr |
+• |
(^..^ |
^^' - ,■ |
1. |
n 3r |
|||||
|
n |
2-. |
3-0 |
+ |
•hi |
^l..-A./../^, - ) |
it'.' |
3H |
|||||
|
1 . ifr: |
||||||||||||
|
n |
/ |
/:■,; |
-J. |
|||||||||
|
f\ |
ii.^!;*.;.:^" n<^. ', <,j |
:„ |
^ |
A' |
1 |
•'"v ■ . ■'■ / |
1 |
|||||
|
J |
M,tr:. -,.-_: *.,v ../.^:., |
i" |
/!. -% |
/ ■• . |
j_^i^, , ... |
|||||||
|
1 |
1,1 n |
A f. : - |
7" |
/' |
.' ■'■■■ -h |
» |
||||||
|
"1 |
('J "J |
|||||||||||
|
Hi- 1 |
||||||||||||
|
'^ s^ |
1 |
'/ : |
+ |
Ah le-'/- .'. |
f' |
/; ■ ' . .> |
111. If 1 ^ |
|||||
|
i |
/J-L |
|||||||||||
|
i |
//- |
|||||||||||
|
1 |
/ ( * Uft! ' / ■ |
/-■ |
i-/ |
T^ |
/■/.' , |
\o. iv- /v.; |
;'j(^/-;. |
^.•^■.' + |
li r ,■ ■ |
1 X |
||
|
h |
/t r*-^r,>„^ ... ; |
r/ |
•i" O |
+ |
r ' |
j |
'•v.; |
J.: ■ + |
r./;,,. |
-\ ^7 |
||
|
n £ |
j-^rj--'^; |
5. |
H.<f |
-1. |
//, „ |
■}:■: |
J^ |
/.vA-, |
||||
|
w |
' '/ |
M; -t- |
||||||||||
|
g 3 |
^^-"-- "":-.-t-.'' - ' ■ ' ■ ,, ■■ ■'■•/ |
/r |
i- H |
T- |
^ /,,,,, |
; /i..' |
/) : ..■ .., |
/»^,i' |
l ri ■ |
/2,1'f M |
^^.. , 2, 1 |
|
|
,. ^r. .,.,.„, |
/J |
7. J |
-^ |
/' '■ • |
/' ' |
/»../.*. |
■0,L.- |
/2^,- k-i |
V |
|||
|
4...--.A.,,:.-..._ |
3r,^ |
i- / |
+ |
/, ,,. |
/t.-. |
i'.^j- |
3V . /i*f- |
|||||
|
J.'/ rj/r^. |
'/ .■- |
-'■ / |
. ^'M |
igure 6. One of 30 data sheets on which details of New Guinean Carabidae have been tabulated, xplanation.
See text for further
ogs, etc. These indications of habitat are )ased primarily on my own observations, lowever, my field notes are scanty; some pecies that I collected were not recog- lized in the field; and the habitats of some ither species are simply assnmed to be he same as the habitats of their closest elatives (but see [25]). There are there- ore probably some errors in the "Ecology" ■olumn, although the ecologic composition if the New Guinean carabid fauna as a vhole, based on this information, is surely ipproximately correct.
The next column ("Endemic?") indi- ates, by "Yes" or "No," whether or not the pecies are endemic to (=confined to) ^ew Guinea. In the case of species which )ccur mainly in New Guinea but \\'hich 'xtcnd to certain not- too-distant islands, he "Yes" in the "Endemic" column is :)laced in parentheses, and occurrence out- :ide New Guinea is indicated in the next •olumn.
The column headed "Geogr(aphic) re-
lationships" indicates the principal occur- rence outside New Guinea, or the closest geographic relationships of each species of which the relationships can be determined. "Oriental," for this purpose, is taken to include the Philippines and Celebes, and "Australian" indicates occurrence in Aus- tralia beyond Cape York. These words without parentheses indicate that the New Guinean species themselves occur in the Oriental area or Australia. The same words in parentheses indicate that the apparent closest relatives of the New Guinean spe- cies occur in the areas indicated. Also in this column are indicated slight range ex- tensions of New Guinean species to New Britain, or to the Solomons, or to the Cape York peninsula but not farther into Aus- tralia.
The next column ("Low/mts.") indi- cates altitudinal occurrence. "Low" means recorded below 500 m (1640 ft.); "Mts.," recorded above 1000 m (3280 ft.); and
156 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
"Lovv/mts.," recorded both below 500 m map in Figure 1. This map shows also,
and above 1000 m. "Int." is used for spe- by broken lines, the extent of the conti-
cies thus far reported only between 500 nental shelves of Asia and Australia,
and 1000 m, and not assignable to either The whole backbone of New Guinea is
the lowland or the mountain faunas. A mountainous, with many ranges rising from
few species, known only from old speci- 12,000 to over 14,000 feet (3660 to 4270
mens, are from unknown altitudes. m), the highest point being over 16,000
The next column ("By me") indicates feet (over 4880 m) in the Snow Mountains
species that I collected, those found at of West New Guinea. The island includes
Dobodnra being indicated by "Dobo," and extensive lowlands too, with several enor-
those that I took only at other localities, mous river deltas,
by "Other." New Guinea is tropical, and the most
The next column indicates species taken, obvious and most widely spread vegeta-
chiefly by myself, on the Bismarck Range, tion is tropical rain forest. However,
\\'ith altitudes indicated; and the next, extensive opener woodlands, grasslands,
species found at or in the vicinity of Wau, and swamp forests occur also at low alti-
again with altitudes indicated. tudes, and altitudinal vegetations are
The last regular column is used for stratified on the mountain slopes, with bare
special notes: if the New Guinean popu- rock at still higher altitudes and permanent
lation is an endemic subspecies of a more snow on a few of the highest summits im
widely distributed species; if a species West New Guinea. More details are given,
widely distributed outside New Guinea under Ecolop,!/: habitats [24].
reaches only the western, southern, or As to history, the geologic evolution of
eastern part of the island; if a species that New Guinea is surprisingly Httle known,
is winged on New Guinea is dimorphic Umbgrove ( 1949 ) indicates the complexity
elsewhere; etc. And the wide right-hand of the history of the whole Indo-Australian
margin of the sheets is used for various Archipelago. David (1950, Vol. 1, p. 681)
calculations, especially for reckoning the says that the physiography of New Guinea
number of specimens taken by myself at and adjacent islands "reflects the dominat-
Dobodura and elsewhere — the extent of ing influence of young and intense orogenic
my own collecting is of some interest as processes." And Browne ( 1958, in The
indicating the extent of my knowledge of Aii.stralian Encyclopaedia, Vol. 6, pp. 452-
habitats. 453) summarizes chronologically what is
For some columns, but not all, totals known of the geologic history of the eastern
have been calculated and are entered at half of the island. From these sources we
the bottom of the sheets, and the sheet may conclude, probably correctly, that the
totals are added on a separate page labeled history of New Guinea has been complex,
"Data summary of New Guinean Carabi- that parts of the island may be old, but that
dae." many features including the principal
mountain ranges are geologically recent,
ANALYSIS AND DISCUSSION New infoi-mation and a new synthesis ol
[17] New Guinea. New Guinea is a huge what is actually known about the geology
island, the largest fully habitable island in ^^nd geologic history of the whole island
the world, stretching about 1500 miles ^re very much needed. i
(2400 km) from the tip of the Vogelkop As to geographic history, when sea level'
to Milne Bay, with an area of roughly fell at times in the Pleistocene, the conti-
300,000 square miles (nearly 800,000 sq. nental shelves outlined in Figure 1 were
km). The position of New Guinea in re- mostly or wholly land. New Guinea was
lation to other land areas is shown in the then surely connected to the continent ol
The carabid beetles of New Guinea • Darlington 157
Australia; the last such connection ended only about 10,000 years ago. At the same times water gaps toward Asia were nar- rower than now. The situation before the Pleistocene is not clear: connections with Australia are likely, but a complete land connection with Asia can hardly have existed in the Tertiary (that is, during the age of mammals), and even a pre- Tertiary connection with Asia is doubtful. This is said of New Guinea as a whole. Of the mountains it can be said more specifi- cally that there seems to be no geologic evidence that mountain ranges on New Guinea have ever been connected with ranges on any other land, and this evidence is consistent with that of the mountain :arabid fauna of New Guinea, which has little relationship with montane faunas else- ivhere, and most of which has apparently ?volved in situ from lowland ancestors [90].
Finally, the biotic history of New Guinea bas evidently been complex. The island has been on the dispersal routes of many plants and animals moving, one way or the 3ther, between Asia and more-southern lands (New Caledonia and New Zealand is well as Australia). Present relationships Ure complex. For example, northern-related 3aks and southern-related Nothofogiis ("southern beeches") occur together in the mountains of New Guinea. And it is likely that many groups of plants and inimals that dispersed across the island in the past have disappeared there. The fos- sil record on New Guinea is too scanty to reveal many details, but indirect evidence illows us to make hypotheses about the )rigin and evolution of parts of the New Guinean fauna, including the Carabidae l(see items under Zoogeograplnj and Evo- 'iition in Contents).
The history of man in New Guinea is briefly traced by Biskup et al. (1968). Al- hough prehistoric men may have reached Vew Guinea as much as 50,000 years ago, And although Europeans began to explore he coast of the island more than 400 years
ago, man has (I think) had only a moder- ate impact on the plant cover and native fauna. Some areas of grassland have been increased at the expense of forest by clear- ing and burning for primitive agriculture and, later, for the establishment of coconut plantations, but enormous stretches of rain forest still remain. Hunting for food by primitive man has presumably reduced populations of some mammals, birds, and reptiles, and may have caused extinction of a few species in New Guinea. For example, the Tasmanian Wolf, which is now con- fined to the island of Tasmania ( if it is not extinct even there), has been found sub- fossil at a prehistoric camp site in New Guinea. And introduction of pigs, which run wild and have modified large areas of the rain-forest floor by their rooting, has probably affected the distribution of some ground-living invertebrates. But I think the sum of all these factors on Carab- idae in New Guinea has probably not been great. The most important effect may have been to facilitate dispersal of some common species that live in grass- land, including some primarily lowland species that have apparently invaded mid- altitudes in places after the clearing of forest [26].
Further information about the eastern half of New Guinea can be obtained from The Australian Encyclopaedia ( 1958; un- der "Papua and New Guinea") and from a compact, well-illustrated small book by Lea and Irwin (1967).
[18] The Neiv Guinean carabid fauna: taxonomic composition. The Coleoptero- runi Catalogus (Csiki, 1927-1933), cover- ing the world as a whole, recognizes 70 tribes of Carabidae (exclusive of Cicindel- inae), or 71 if the Agonini (considered a subtribe of Pterostichini by Csiki) are given tribal rank, and the number is in- creased to 72 if the Paussini (placed in a separate family by Csiki) are added. Of these 72 tribes, 25, or about one third, are represented in New Guinea. These tribes are listed, and the number of New Guinean
158 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Table 1. Tribes of Carabidae proper repre- sented IN New Guinea, with numbers of New
GUINEAN species
Tribe
No. of species in New Guinea
Ozaenini
Paussini
Scaritini
Bembidiini
Trechini
Panagaeini
Pterostichiiii
Agonini
Perigoniiii
Licinini
Chlaeniini
Oodini
Harpalini
Anaulacini
Cyclosoniini
Lebiini
Pentagonicini
Hexagoniini
Odacanthini
Dryptini
Zuphiini
Helluodini
Helliionini
Brachinini
Pseudomoqihini
1 1
30
72
3
6
46
160
14
8
12
13
53
4
1
160
15
2
18 5 8 10 10 7 8
667
species in each indicated, in Table 1. Some changes in tribal classification have been made since publication of the Catalogus, and some authors have greatly increased the number of tribes by splitting ("taxo- nomic inflation"), but the splitting has not increased the proportion of tribes repre- sented in New Guinea, and the Cata]Qg,us is still the most convenient basis for ar- ranging the data that I want to present. These tribes are individually discussed in the following pages under item numbers [3.3-57].
Of these 25 tribes represented in New Guinea, the Scaritini include most of the fossorial species. The Bembidiini include most of the very small species, especially on the ground ( some small arboreal species are lebiines). The Pterostichini are notable for representing perhaps a relatively old
fraction of the fauna (but apparently no New Guinean Carabidae are very old), and for including several endemic genera at high altitudes. The Agonini are dominant on the ground (but a few are arboreal) and especially at higher altitudes. The Harpalini are dominant on the ground in grassland and in other relatively open places (but some occur in rain forest too). And the Lebiini are dominant in arboreal habitats (although some are ground-liv- ing), especially in rain forest. This is, I think, a sufficient introduction to the taxo- nomic composition of the New Guinean carabid fauna as a whole.
[19] Numbers of species. The number of full species of Carabidae proper (exclusive of Cicindelinae) now known from New Guinea is 667. These species are tabulated on the data sheets described above [16].
The lowland and mountain-living cara- bid faunas of New Guinea can usefully be distinguished and compared. I have arbi- trarily considered as lowland species all those actually recorded below 500 m (c. 1640 ft.), and as mountain-living all those recorded above 1000 m (c. 3280 ft.). Many species that occur below 500 occur also above 1000 m: 161 such species are now known, and the number will probably be much increased in the future. Some of these species are common over a wide range of altitudes. Others are primarily lowland forms which occasionally occur above 1000 m in suitable habitats. And others are primarily mountain-living forms which descend below 500 m in suitable habitats. (See discussion of altitudinal range in relation to habitat under Altitude [26].) A few (18) species that have thus far been found only between 500 and 1000 m, or at localities of which the altitudes are doubt- ful, cannot now be assigned to either low- land or mountain faunas.
Of the total of 667 full species of Carab-. idae now known from New Guinea, 434 have been found in the lowlands, below 500 m altitude. My impression is, judging from the small number of additional spe-
I
The carabid beetles of New Guinea • Darlington 159
cies that turn up in new collections, that relatively few primarily lowland species remain to be discovered, although many primarily mountain-living species will probably be found to occur below 500 m n places.
The number of species at a single low- and locality in New Guinea may be more lignificant than the total number of low- and species on the island. At Dobodura, -^apua, I found 217 species in about four nonths of collecting. Dobodura is a strictly owland locality; my collecting was done mder difficulty; and my collection from here is surely incomplete, being especially leficient in arboreal species. (See Part I. )p. 325-326, and the present part [3, 5] or further description of this locality and )f my collecting there.) I think that a ear's unhampered collecting by a compe- ent carabid specialist would surely raise he number of species at Dobodura to nore than 250, and perhaps to more than )00. Although a few species are still known mly from Dobodura, the majority found here are widely distributed over much or ill of New Guinea.
Because the Carabidae of most other slands in the Indo-Australian Archipelago, 'xcepting Java (see [64] and Fig. 13), are lot well known, I have to go to the other ide of the world for comparative data. Tlie Carabidae of the Greater Antilles, the four arge islands of the West Indies, are well :nown (better known than those of New Guinea), and I have tabulated numbers of pecies in relation to areas of the islands in I paper on Carabidae of mountains and slands (1943: 42, Table 1). Inspection of his table yields a rule of thumb: among hese islands, if one island is ten times as arge as another, the larger island will lave about twice as many species of >arabidae.
Of the Greater Antilles, Hispaniola
called also Haiti or Santo Domingo) is
,fiost nearly comparable to New Guinea,
'eing ecologically diverse and having ex-
ensive mountains. New Guinea (roughly
800,000 sq. km or 300,000 sq. mi. in area) is about ten times as large as Hispaniola (roughly 80,000 sq. km or 30,000 sq. mi.). Hispaniola has a known carabid fauna of 148 species, of which 97 are lowland forms. By my rule of thumb. New Guinea should have a total of about 300 species, of which about 200 should be lowland forms. But figures given in preceding paragraphs show that New Guinea has in fact more than twice as many species as expected by this rule.
The greater richness of the New Guinean fauna may be due partly to the greater ecologic richness and diversity of the island. New Guinea has very large areas of fine rain forest, several big rivers, ex- tensive swamps, etc., while the Greater Antilles have relatively little, relatively poor rain forest, no really large rivers, and few large swamps. The greater richness of New Guinea may be due in part also to the greater accessibility of the island, which was connected by land to Australia at times in the Pleistocene, and which many Carabidae have reached from the west too.
The New Guinean carabid fauna is in fact continental in size and diversity. Satis- factory figures for numbers of species in continental areas of Asia and Australia are not available, so again I have to go to the other side of the world for comparisons. In 1943 (p. 41) I counted or estimated num- bers of species of Carabidae in several areas in eastern North America including New Jersey, Indiana, and North Carolina. Each of these states has more than 300 but less than 400 species of lowland Carabidae. Most of the species are widely distributed and are shared by all three states, so that the total number of lowland species in all three states together does not much exceed 400. This is considered a rich continental carabid fauna. However, the lowland fauna of New Guinea is evidently still richer — it must be remembered that even the lowland Carabidae of New
160
Bulletin Mttscttm of Comparative Zoology, Vol. 142, No. 2
Guinea are still much less well known than those of eastern North America.
In 1943 (p. 41) I thought I detected a diminution of numbers of Carabidae from temperate North America into the Ameri- can tropics, although exact figures were not available. New Guinea can now be seen to have a carabid fauna rich out of all expectation for a lowland tropical area. Evidently Carabidae are very numerous in species in the lowland tropics at least in some regions. (But in proportion to the total insect fauna, Carabidae may still be less numerous in the tropics than in temperate areas.) However, they are evi- dently much more difficult to find in the tropics than in the north temperate zone, probably because they are more diverse ecologically in the tropics, and perhaps because the tropical populations tend to be sparser. (For further discussion of numbers of Carabidae in the tropics includ- ing New Guinea, and of the possible effect of competition with ants, see the present paper [27] and Fig. 11.)
Mountain-living Carabidae, known above 1000 m (c. 3280 ft.) in New Guinea, now total .376 full species, of which 161 are and 215 are not known below 500 m. However, this is probably a small fraction of the total number of Carabidae existing on mountains in New Guinea. Until good collections have been made not only on different mountain ranges but also on successive peaks along single ranges, we have no basis for estimating the amount of geographic replacement of localized species that occurs from point to point in the mountains of New Guinea, and no basis for making a real estimate of total number. I can there- fore only guess, from the fragmentary collections available, that the total num- ber of mountain-living species of Carabidae in New Guinea will run to many hundreds, perhaps thousands, of species.
The number of species of Carabidae oc- curring at a single locality at middle alti- tudes in New Guinea is surprisingly large. For example, 170 species have been taken
between 1000 and 2000 m at and near Wau [5], in the Morobc District of North-east New Guinea. (Thirty-five additional spe- cies have been found in the same general area either below 1000 or above 2000 m.) The collections at Wau were made by persons (the Sedlaceks and others) who, although fine collectors, are not carabid specialists. Their collections are surely deficient in small ground-living species, al- though strong in arboreal ones and in light- trap material. The total number of Carab- idae existing at and near Wau betw'een 1000 and 2000 m is probably really con- siderably more than 200 species. Many of the species that occur at Wau occur also in the lowlands, but many others are either confined to the mountains or at least have not yet been found at lower altitudes; some are wide-ranging in New Guinea, others apparently confined to the general vicinity of Wau.
At still higher altitudes in the mountains of New Guinea so little carabid collecting has been done that counts of species are hardly significant. However, the number of species at single localities obviously de- creases very sharply with increase of alti- tude. Incomplete samples (Table 2) show the Snow Mountains in West New Guinea with nine species above 3000 m, including only one above 4000 m, and Mt. Wilhelm in North-east New Guinea with seven species above 3000, including two above 4000 m.
In summary of numbers of species: New Guinea has an unexpectedly large and diverse carabid fauna, large even by con- tinental standards. Of the total number of 667 species of Carabidae proper now known from the island, 434 have been found be- low 500 m altitude, and many of these are widely distributed on the island. At mid- altitudes (1000-2000 m) species are still numerous; some of them are widely dis-J tributed, others localized. At still higher altitudes, numbers of species at single lo- calities decrease sharply, but most of the high-altitude species are localized, and
The carabid beetles of New Guinea • Darlington 161
Table 2. New Guinean Carabidae found above 3000 M (above C. 10,000 ft.)
Table gives names, mean lengths, and altitudes of all Carabidae thus far recorded above 3000 m (above c. 10,000 ft.) in New Guinea. All species are (probably) ground-living mesophiles except Nebriagoniim percephahim, which is found beside mountain torrents, and Maculagonum altipox, which apparently lives in tussock grass. All species listed are flightless, with atrophied wings, except Maculagonum altipox, which has full wings.
Snow Mts., West N. G. (9 species) Mectjclothorax ioxopei (4.7 mm), 4200 m Rliytiferonia nigra (19.75 mm), to 3300 m Analoma fortis (12.85 mm), to 3850 m Analoma gracilis (14 mm), 3800 m Gastragonum laevisculptum (8.3 mm), 3600 m Montagonum toxopeanum (9.0 mm), 3600 m Nebriagonum subcephalum (9.9 mm), 3300 m Chydaeus papua (9.9 mm), 3600 m Scopocles altus (3.4 mm), to 3800 m Mt. Wilhelm, N-E N.G. (7 species) Mecijclothorax sedlaceki (4.3 mm), 4250 m Maculagonum altipox (7.0 mm), above 3000 m
( winged, in tussock grass ) Nebriagonum cephalum (8.55 mm), to 4250 m
(2 collections) Nebriagonum percephahim (9.9 mm), to above
3000 m ( beside running water ) Nebriagonum transitum (9.35 mm), to 3400-3500
m (2 collectors) Laevagonum subcistelum (5.95 mm), above 3000
m Chydaeus papua (9.9 mm), to 3400-3500 m (sev- eral collectors ) Mt. Albert-Edward, Papua (3 species) Analoma rosenburgi (12.6 mm), 4026 m Montagonum filiolum (10.5 mm), 3660 m Fortagonum antecessor (9.0 mm), 3660 m Mt. Giluwe, Papua ( 1 species ) Laevagonum giluwe (7.0 mm), 3750 m Mt. Amangwiwa, N-E. N. G. (1 species) Montagonum fugitum (11 mm), 3355 m
much geographic replacement occurs, so that the total number of high-altitude Carabidae on the mountains of New Guinea is surely very great, although the final number can not yet even be guessed at closely. This situation is diagrammed in Figure 7.
[20] Size of individuals. New Guinean Carabidae are small. Of 434 lowland spe- cies (exclusive of Cicindelinae), 388, or
89 per cent, have a mean length of 12 mm (c. V2 inch) or less, and only five species exceed a mean length of 20 mm. Of these five, three are characteristic, endemic New Guinean species : Lesticiis politiis ( 24 mm ) , Colpodes rex (21 mm), and Chlaenius pan (25 mm). The other two, still larger spe- cies are marginal or introduced: Gigadema maxilhre (32 mm) is an Australian species of which a single specimen has been found on the southern edge of New Guinea, and Catadromus tenehroides (mean 51 mm) is an Australian species of which two speci- mens were taken at military ports in New Guinea during the war (and which has been found in Java too). So, characteristic lowland Carabidae of New Guinea are all small, none more than an inch long, and most much less than that. Mountain-living Carabidae in New Guinea are small, too, none having a mean length of more than 22 mm (less than 1 inch). However, minute species (mean length less than 3 mm), although numerous in the lowlands, become relatively fewer with increasing altitude and disappear at highest altitudes. The few Carabidae known above 3000 m (Table 2) range from 3.4 to 19.75 mm mean length.
Small size is a characteristic of the Carab- idae of some other East Indian islands including Celebes and the Philippines. (Sumatra, Java, and Borneo have a few larger Carabidae, notably species of Mor- molyce.) West Indian Carabidae are all small, too.i On Cuba, no carabid has a mean length of more than 25 mm (1 inch), and only four (2 species of Calosoma, a Scarites, and a Chlaenius,) exceed a mean of 20 mm; and on Hispaniola only the two Calosoma and possibly a Scarites exceed a mean of 20 mm. However, the situation is
''■ Since this was written, a very large, large- headed Scarites has been discovered in the moun- tains of eastern Puerto Rico (Hlavac, 1969). The first specimen found measured about 35 mm. It far exceeds in size any of die approximately 350 species of Carabidae previously known from the West Indies.
162 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
C 3000 M
C 2000 M
C 1000 M
LOW
LANDS
Figure 7. Diagram of distribution of species in relation to altitude among New Gulnean Corabidae. Tfie diagram is not exactly quantitative, but each fiorizontal line, whether broken or not, represents c. 50 species; each unbroken vertical line, c. 10 species; and each broken vertical line, less than 10 species. Unbroken horizontal lines represent species that extend over c. the whole length of New Guinea; broken horizontal lines, species that are more localized. The diagram emphasizes that increase in altitude is accompanied both by decrease in number of species at single localities and by increasing locali- zation of species.
strikingly different on Madagascar. The Carabidae of Madagascar include many very large forms especially in the tribes Scaritini and Pterostichini, some of the Madagascan scaritines being among the largest Carabidae in the world.
The small size of New Guinean Carab- idae is not easy to explain. Mere existence on a large island does not necessarily favor smallness: witness the gigantic carabids on Madagascar. Nor does existence in rain forest necessarily favor smallness: witness Mecynognathus and the large Tricho- stermis in tropical rain forest in Australia ( following paragraph ) ,
Comparison of New Guinean species with the same or related species in Austra- lia and southern Asia indicates no general decrease of size on New Guinea [93] . New Guinean Carabidae are small because small rather than large Carabidae have reached and established themselves on the island. This suggests that difficulty of access has barred large forms, but this explanation is too simple or at least incomplete. New Guinea was connected to Australia at
times in the Pleistocene. Many Australian Carabidae including many Scaritini and Pterostichini are large: one of the largest and finest Carabidae in the world, Mectj- nognathus dameli Macleay (mean length c. 50, maximum length 63 mm), is confined to a small area of poor rain forest on the tip of Cape York, just opposite New Guinea; other large Pterostichini. espe- cially species of Trichosterntis (Darlington, 1961), occur in the tropical rain forests at the base of the Cape York Peninsula; and large Carabidae occur in the drier parts of Australia too. Why did not large forms invade New Guinea when there was a land connection in the Pleistocene? Ecologic barriers evidently existed between Austra- lia and New Guinea even when land was continuous [84], but why were the bar- riers so effective against large carabids living in both rain forest and dry country?. Predation, by mammals and birds, does.i not explain the absence of large Carabidae in New Guinea. Insectivorous predators do occur there, but they occur also in Australia and Madagascar, where large
The carabid beetles of New Guinea • Darlington 163
Carabidae are common. And on the West be due to increase of size, the increase
Indies, where mammahan predators at having required a relatively long time
least are relatively few, large Carabidae and relatively effective isolation,
do not occur. The size distribution of New Guinean
In spite of what has been said in the Carabidae is, unexpectedly, bimodal. A
several preceding paragraphs, and al- histogram (Fig. 8) of the distribution by
though the situation is evidently complex, size of all 434 lowland species shows not
I think the small size of New Guinean only that most are small but also that
Carabidae may be a result of a combi- separate modes occur at 2.0-2.95 and at
nation of three factors: (1) difficulty of 6.0-6.95 mm, with a deficiency especially
access, (2) the greater dispersibility of at 4.0-4.95 mm. A highly speculative and
>mall as compared with large Carabidae, oversimplified explanation can be derived
md (3) time. New Guinea has been from the possible relation of Carabidae
somewhat isolated, on the west by water with ants. Very small Carabidae may be
aarriers which have probably been nar- able to hide from ants, and relatively large
-ower than now at times in the past but ones may be able to protect themselves,
A^hich have existed for a very long time, while Carabidae in the 4.0-4.95 mm size
md from Australia by fairly effective class may be especially vulnerable to com-
jcologic barriers even when what is now petition with or predation by ants. If so,
he island was connected with the main- this is only one aspect of a probably com-
and. Small insects do disperse through plex impact of ants on Carabidae and
he air more readily than large ones; the carabid faunas which is further discussed
mcestors of New Guinean Carabidae were under Ants [27],
lot only all small (or at least not very A simpler explanation of the bimodal
arge ) but almost all of them were or may size distribution of lowland Carabidae in
lave been winged when they reached the New Guinea can be derived from the
sland [23, 88]. The small, winged an- predominance of the single genus Tachijs
•estors of New Guinean Carabidae may ( taking the genus in a broad, old-fashioned
lave crossed barriers which large Carab- sense). This is a huge genus of small
dae and flightless ones did not cross. Carabidae. Although it is well represented
rhis explanation does perhaps account well in some temperate areas, it is primarily
•nough for the initial small size of New tropical, and includes great numbers of
Guinean Carabidae. To account for the species in the tropics of all continents,
iresent smaller size of Carabidae on New Most of the really small Carabidae in
Guinea as compared with Madagascar, New Guinea belong to this genus: the
ime must be invoked, I think. The New portion of the bimodal histogram (Fig.
Guinean fauna is evidently relatively re- 8) represented by Tachys is shown by
■ent in its origins; the Madagascan fauna, hatching in the first three columns of the
)robably much older. Also, Madagascar histogram. Without Tachys, the lowland
s much more effectively isolated than Carabidae of New Guinea have a regular
^ew Guinea; invasions and replacements size distribution, with a single mode at
)robably occur less often; and old stocks or near 6-6.95 mm. The bimodality of the
)robably survive longer and have more fauna is therefore apparently due to the
ime to increase in size. (See [89] for addition of a large number of species of
liscussion of relative age and rate of the predominant genus Tachys to what
uniover of the New Guinean fauna.) So, would otherwise be a simpler, unimodal
vhile the small size of New Guinean Carab- size distribution. But this explanation does
dae is not due to decrease of size, the not necessarily preclude the preceding
irge size of many Madagascan forms may one. The small size of individual Tachys
164 BuUetin Museum of Comparative Zoology, Vol 142, No. 2
CO
UJ
o
UJ
a
UJ OQ
ID
0- I- 2- 3- 4- 5- 6- 7- 8- 9-10- 1 1- 12-13-14-15-16- 17- 18-19-2021- SIZE CLASSES (EACH N-N.95 MM)
cm. 24-25- 32- 51
Figure 8. Histogram of size distribution of lowland Carabidae in New Guinea. Each species is counted once, according tci its median lengtfi. Eacfi column represents number of species of which the mean length falls between N-N.95 mm; the actual numbers of species are given above the columns. The hatched portions of the first three columns represent Tachys. See text for further explanation.
may be one reason for their great success, and may enable them to Hve in the pres- ence of dominant ants more successfully than slightly larger Carabidae can do. Although several groups of Tachijs have certainly radiated to some extent in New Guinea, the number of species there is not due primarily to radiation but to the large number of separate stocks that have reached the island: my data sheets indi- cate at least 23 separate relationships be- tween different New Guincan Tachys and those of other areas, and this suggests at least 23 separate invasions of New Guinea by members of this one genus.
In an attempt to get a more detailed explanation of the two size modes of low- land New Guinean Carabidae, I have made a size histogram (Fig. 9), of the 217 species that I found at Dobodura, indicat- ing the ecologic composition of the size classes. I have limited this diagram to species that I myself collected at one
locality, because I know something ol their ecology. (See Ecology: habitats [24 _ for further discussion of habitats of Ne^^' Guinean Carabidae.) This histogram sug gests that lowland mesophile Carabidae do have a bimodal size distribution in New Guinea. Mesophiles compete more di- rectly with ants than hydrophiles anc arboreal Carabidae do, and would be more likely to show the effects of competition.
My tentative conclusion is that th(" double-moded size distribution of lowlanc Carabidae in New Guinea is due to the presence of excessive numbers of smal species of Tachys, but that this explanatioi still allows the possibility of an underlying effect of competition with or predation b) ants.
Mountain-living Carabidae in Nev, Guinea have a different size distributioi from the lowland forms. Figure 10 is ; histogram of the sizes of the 215 known strictly mountain-living species. This figure
The carabid beetles of New Guinea • Darlington
165
CO
lij o
UJ
a.
CO
UJ CD
z
0-1- 2-3-4-5-6-7- 8-9-1011-12-13-14-15-16-17-18- SIZE CLASSES (EACH N N 95 MM)
igure 9. Histogram of size distribution of Carabidae found at Dobodura. Explanotion as for Fig. 8, except hatched portions f columns represent ecologic groups: right-oblique hatching at bases of columns, hydrophiles; no hatching, mesophiles; nd left-oblique hatching at top of columns, arboreal forms.
hows a single mode at a larger size — at ir near 9-9.95 mm — than the second mode •f the lowland species. Very few Tachys iccur in the mountains; their place in the listogram is represented by the hatched >art of the first column; and removal if Tachys from this histogram does not •hange it significantly. However, ants too ire relatively few at higher altitudes in 'Jew Guinea, and their fewness may have omething to do with the size distribution )f mountain-living Carabidae.
In summary of the size of New Guinean Carabidae: all the characteristic ones are mall. The absence of large forms may )e due to a combination of ( 1 ) difficulty )f access, (2) the greater dispersibility of mall as compared with large Carabidae, md (3) the fact that the existing New Guinean carabid fauna is relatively recent n its origins and has not had time to ^volve large forms. Tlie lowland Carabidae
of New Guinea have a bimodal size dis- tribution, with modes at 2-2.95 and 6-6.95 mm; this bimodality is due primarily to the presence of many small species of Tachys, but may also reflect an underlying relation with ants, which may compete with or prey on Carabidae especially in the 3 to 5 mm size classes. Mountain-living New Guinean Carabidae have a unimodal size distribution, with the mode at or near 9-9.05 mm; l)oth Tachys and ants are relatively few in the mountains, and their absence may partly account for the dif- ferent size distribution of mountain-living as compared with lowland Carabidae.
[21] ^^^//ii,'.s and wing atrophy. I have a long-standing interest in the wings, wing atrophy, and flight of Carabidae (Darling- ton, 1936; 1943), and I have been con- stantly on the lookout for cases of wing reduction among the New Guinean species. An advantage of specimens killed in alco-
166 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
if) liJ
O liJ Q. 0)
U.
o
Ul OQ
D Z
|
CM ro |
||||||||||||||||||||
|
lO in CM (M |
||||||||||||||||||||
|
ro — |
^ |
|||||||||||||||||||
|
o o |
IV |
CD |
cvl |
1 r |
||||||||||||||||
|
\ |
1 |
^ ^ |
CJ |
|||||||||||||||||
|
OJ CVJ |
||||||||||||||||||||
|
1 1 |
0- 1- 2- 3- 4- 5- 6- 7- 8- 9- 10- II- 12-13- 14-15- 16-17-18- 19-2021-22-
SIZE CLASSES
(EACH N-N.95 MM)
Figure 10. Histogram of size distribution of mountain-tiving Carobidoe in New Guinea. Explanation as for Fig. 8. The hatched portion of the first column represents lachys.
hol, as most of mine were, is that the elytra are likely to be slightly separated and the inner wings easily visible. This is not usu- ally the case in dry-killed specimens, but if the latter are in good condition and properly mounted, it is usually possible to separate the elytra with the point of a pin and see whether or not the wings are fully developed. I have thus been able to see the wings of thousands of specimens of New Guinean Carabidae with com- paratively little difficulty, and in species in which wing reduction has been detected or suspected, I have supplemented this routine examination with more careful examination, by relaxing specimens and raising their elytra. However, I have not usually extended my examination of wings to specimens collected at light, because as a rule only winged, flying individuals can come to hght, so that light trap material is very deficient in flightless forms. In discussing state of wings, I use three
symbols: +w means with wings fully de- veloped or at least long and folded at apex; ±w, wings dimorphic, full in some and reduced in other individuals; and -w, wings reduced in all individuals.
At low altitudes in New Guinea most Carabidae arc +w. Many are known to fly (to Hght), and most probably do so. Of the 434 known lowland species (exclusive of Cicindelinae) only 17 (Table 3), or about 4 per cent, have been found to have any sort of wing reduction at low altitudes in New Guinea.^ (Two additional species
^ Although I have excluded tiger l)eetles (Cicindelinae) from my analysis and discussion of the New Guinean carabid fauna, one species should be mentioned in the present connection. It is Tricondyla apteia (Part I, p. 334). This is tlie only flightless tiger beetle in New Guinea, so' far as I know. Not only is it flightless itself, but it belongs to a wholly flightless Oriental genus, and has almost surely reached New Guinea with- out flying. It occurs in rain forest, but not on the forest floor. It li\'es on the trunks of trees,
The carabid beetles of New Guinea • Darlington 167
Table 3. Lowland Carabidae of New Guinea in which reduction of wings has been found
( Scaritini )
Clivina dedlata (Part I, p. 372), 4-w and -w popu- lations at different localities
Clivina enigatella (Part I, p. 380), +w and ±w populations at different localities
( Bembidiini )
Tachijs serrula (Part I, p. 408), +w and -w popu- lations at different localities
Tachijs amhulatus (Part I, p. 425), -w at single known locality
Tacht/s aviits (Part I, p. 426), ±w at Dobodura
{Tachijs truncatus (Part I, p. 431; Darlington 1970: 15), apparently always +w in New Guinea but ±w in Micronesia)
{Tachijs brachtjs (Part I, p. 433; present part, Tax. suppl.), -fw in New Guinea (only 4 specimens) but -w or ±w in the Moluccas and Formosa ( Taiwan )
( Pterostichini )
Lesticus politus (Part I, p. 526), 2 specimens only seen, 1 +w and 1 -w, from different localities
Platijcoehis depressiis (Part I, p. 534), ±w, geo- graphic distribution of wing forms not de- termined
Loxandriis latiis (Part I, p. 551), ±w at Doborura
( Agonini )
( No lowland species with reduced wings, although many -w in mountains )
( Oodini )
Oodes terrestris {laevlssimiis of Part III, p. 34), apparently ±w at Dobodura, but wings only slightly reduced in "— w" individuals
Oodes rossi (Part III, p. 34), single known speci- men -w
( Harpalini )
Hijphaereon timidus (Part III, p. 67), ±w (wings
actually polymorphic) at Dobodura; state of
wings elsewhere not determined
( Lebiini )
Nototariis papua (Part III, p. 186), -w in series from Dobodura
( Pentagonicini)
Parascopodes cijaneus (Part III, p. 196), ±w at Dobodura
( Zuphiini ( Leleupidiini ) )
Colasidia papua (present part. Tax. suppl.), single known individual -w
( Brachinini )
Pheropsophus apfinomorphus (Part III, p. 237), -w Pheropsophus catulus (Part III, p. 238), -w Pheropsophus canis (Part III, p. 238), -w
Total: 17 species, of which 10 ±w, 7 -w (so far as known); product of at least 14 separate wing- reductions ; little or no radiation in -w stocks
that are +w at low altitudes are ±w on the Bismarck Range — see Table 5.) The patterns of occurrence of +vv and -w individuals of these 17 species are sum- marized in Table 4. These 17 species rep- resent at least 14 separate stocks, in each of which wing reduction has occurred independently. None of these stocks has radiated much at low altitudes, although speciation may have begun in the -w Tachys {amhulatus and avius), Oodes, and Pheropsophus {catulus and canis) (see again Table 3). It is noteworthy that no
reduction of wings seems to have occur- red among Agonini at low altitudes in New Guinea, although the wings have atrophied in a number of stocks of this tribe on the mountains, and although the -w agonines dominate the mountain carabid fauna of the island. In being composed almost wholly of (small) winged species, the low- land carabid fauna of New Guinea is like the faunas of some lowland areas in the continental tropics and of some other tropical islands (see fifth following para- graph ) .
and it is active and conspicuous on them by day. It is about 25 mm ( 1 inch ) long, slender, cylin- drical, with large eyes and long appendages. It is strikingly antlike in appearance and movements. The exceptional success of this insect may be due to its ability to compete with large preda- ceous ants. It combines antlike agility and strik-
ing power ( it can bite severely ) with protective annor and perhaps better vision than ants have. It may be a biological example of the well-known principle, "if you can't beat them, join them." But it is exceptional, and its existence in the lowland rain forest of New Guinea emphasizes how few other oarabids are flightless tliere.
168 BuUctin Museum of Comparative Zoology, Vol. 142, No. 2
Table 4. Summary of distribution of +W and
-W indi\tduals in 17 lowland Carabidae listed
IN Table 3
±\\' populations at single localities ( in
single populations) 6 species
+w and -\v populations at different
localities — 3 species
±\v, hut distribution of wing forms not
determined 1 species
-w in all specimens seen 7 species
Total 17 species
Wing reduction among Carabidae is often followed by other structural changes, especially by narrowing of humeri and shortening of metepisterna. However, most of the lowland species (listed in Table 3) in which wing reduction has oc- curred either are still dimorphic (±w) or do not yet show the structural changes that tend to follow wing reduction. Of the 17 species listed in Table 3, only five begin to show the secondary structural changes (indicated above) that suggest that the insects have been flightless for a consider- able time. Tliese five are Oode.^ rossi, Nototarus papua, Cola.sidia paptia, Pherop- ■sophu.s catiihi.s, and P. cants. I have col- lected three of these species and close relatives of the other two. All five probably live in leaf litter on the floor of rain forest. This fact suggests that selective factors are relatively favorable to flightlessness on the ground in rain forest, although the small number of species that have become flight- les.s even there at low altitudes suggests that the selective advantage is limited. Special factors in favor of flightlessness on the floor of rain forest perhaps include the stability and continuity (both in space and time) of that habitat. Factors limiting flightlessness there may include patchiness of distributions [22] and liability to flood- ing. Floods often do occur on flat ground in rain forest, and Carabidae do fly to escape from them. Competition wdth ants, which are numerous on the ground in rain forest as well as in many other habitats in
the lowland tropics (but which are fewer beside water and at high altitudes), may be an additional factor limiting flightless- ness.
Known mountain-living species of Carab- idae in New Guinea (found above 1000 m ) total 376 species. Of these, 161 species occur also in the lowlands (below 500 m), and all of these species with wide alti- tudinal distributions are +w, except that two normally +w species of T richotichnus are locally ±w on the Bismarck Range (see Table 5). There is here a correlation, perhaps to be expected, between wide altitudinal distribution and possession of fimctional wings.
Of the 215 species of New Guinean Carabidae found above 1000 m but not below 500 m, 69, or 32 per cent, exhibit wing-reduction (Table 5). Only one of these species is known to be ±w (Ga.stra- goniim teircstre. Part II, p. 226), the other 68 species being unifomily -w so far as known. In fact Gastragonum is the only strictly mountain-living genus of Carabidae in New Guinea in which the wangs are dimorphic, most of the other strictly alti- coline genera being apparently products of^ radiation of -w ancestors. (Exceptions are' Plicagoniim, riparian Potamagonum, and' probably-grass-living Maculagomim, which are uniformly +w.) The phylogenies of the mountain-living forms have not and probably can not be fully worked out, but at least 20 different stocks (probably more) ha\'e undergone wing reduction indepen- dently to produce the -w mountain forms, and secondary structural modifications and' radiation in some -w stocks indicate that wing atrophy is long-standing in many of them.
At highest altitudes, above 3000 m, only 21 species of Carabidae have yet been found in New Guinea (Table 2). Of' these, 20, or 95 per cent, are -w, the only +w species on the list being Maculagomim altipox.
The altitudinal distribution of +w and -w Carabidae on New Guinea is consistent
The carabid beetles of New Guinea • Darlington
169
Table 5. Mountain-living Car.a.bidae of New Guinea in which reduction of wings has been
FOUND
Scaritini )
'Hivina toxopci (Part I, p. 363), -w (Australian
relatives both +w and — w ) Uivina kiibor ( present part, Tax. suppl. ) , -\v,
(relatives +w)
Bembidiini )
Jwnastis inops (Part I, p. 486), -w (genus in- cludes +w species )
Panagaeini )
Iraspedophonis gressittorum (present part, Tax. suppl. ) , -w ( Australian species -w, some Oriental -fw)
Pterostichini )
iecijclothorax toxopei (Part I, p. 506), -w (Aus- tralian relatives both +w and -w)
Iecijclothorax sedlaceki (present part, Tax. suppl.), -w (see preceding species)
^estkus (Part I, pp. 521ff; present part, Tax. suppl.), 5 species -w (genus includes +\v and ±\v species )
Ihytifcronia (Part I, pp. 533ff), entire genus (2 species) -w, (related Australian genus +w)
'rosopoginus (present part. Tax. suppl.), 2 species -w ( genus includes +w species )
inaloma (Part I, pp. 538ff; present part, Tax. suppl. ) , entire genus ( 4 species ) -w ( rela- tives undetermined)
Agon in i )
^otagonuru ambulator (present part. Tax. sujipl.)
-w (genus otherwise entirely -|-\v) l,astragonum (Part II, pp. 222ff), 1 species ±\v,
3 -w ( genus includes -fw species ) diagonum (Part II, pp. 229ff; present part. Tax.
suppl.), entire genus (6 species) -\v dontagonum (Part II, pp. 233ff; present part, Tax.
suppl.), entire genus (8 species) -w
Nebriagonum (Part II, pp. 235ff; present part, Tax. suppl.), entire genus (7 species) -w
Laevagonum (Part II, pp. 243ff; present part, Tax. suppl. ) , entire genus ( 7 species ) -w
Fortagonum (Part II, pp. 247ff; present part. Tax. suppl.), entire genus (11 species) -w
( Licinini)
Microferonia baro (Part III, p. 19), -w (genus includes +\v and ±w species in Australia)
( Oodini )
Oodes wilsoni (Part III, p. 35), -w (most species
of genus -\-w [A -w Coptocarpus just received from New Guinea
has not \et been described]
( HariDalini )
Chydaeus (Part III, pp. 47ff; present part. Tax. suppl.), 2 species -w (genus includes -fw and ±w Oriental species)
Tricliotichnus nigricatis (Part III, p. 52), ±w on Bismarck Range ( 4-w elsewhere )
Trichotichnus alius (Part III, p. 54), ±w on Bis- marck Range ( +vv elsewhere )
( Pengatonicini )
Scopodes alius (Part III, p. 198), -w (Australian relatives include +\v and ±w species )
( Zuphiini ( Leleupidiini ) )
Col-asida madang (present part. Tax. suppl.), -w ( known relativ es all -w, but group descended from +w stock )
Total: 71 species, of which 3 ±\\', 68 -w; product of at least 22 separate wing reductions; radiation in -w stocks in some Pterostichini and especially Agonini
\ith distributions in some other, compa- able parts of the world. Some other (but lot all) tropical lowland carabid faunas ire wholly or almost wholly +w, for ex- imple in the Santa Marta region of Co- ombia, northern South America, and on 3arro Colorado Island in the Panama "anal Zone (DarHngton, 1943: 41). And n the West Indies the situation is extra- )rdinarily like that on New Guinea, ilthough the Carabidae concerned are en- irely separate in their immediate deri-
vations and local evolutions. On Cuba, for example (DarHngton, 1943: 49), of 144 known lowland species, almost all are +w, and of the five (3/r ) lowland species that do show wing reduction, four are still ±w; but -w stocks appear with in- creasing altitude on the mountains, and on the summit of Pico Turquino (the highest Cuban mountain, c. 6500 ft. or 2000 m), all se\'en known species are exclusi\ely -w. I collected on Turquino myself for two weeks in 1936 and secured 459 specimens
170 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
of these seven species, every specimen -w. On Jamaica, all known lowland Carabidae (63 species) are +w, but again transition to -w fornis occurs with increasing alti- tude, and above 5000 ft. (c. 1500 m) the ten known species are 80 per cent -w, onK' 20 per cent +w (Darlington, 1943: 49)'.
Wing atrophy of Carabidae in New Guinea is correlated not only with altitude but also with habitat. Carabids may be roughly classified as mesophiles (ordinary ground living species, called also geo- philes), hydrophiles (species specifically associated with open water, especially swamps and the banks of streams), and arboreal species. (See Section [24] for further discussion of this classification of habitats.) Wing atrophy has occurred fre- quently among mesophiles but rarely in the other ecologic groups in New Guinea. A few -w species are or may be hydro- philes, including 'Nehria'^omim cepholum and N. percephalum (Part II, pp. 239, 240), which live beside moimtain torrents on the Bismarck Range. But, so far as I know, not one of the many arboreal Carabidae of New Guinea is -w at any altitude. :
The situation among New Guinean Carabidae suggests that mutations from a long- to a short-winged condition are con- tinually occurring at all altitudes. In the lowlands, selection apparently usually elim- inates -w individuals, and most species remain +w, although wing atrophy does sometimes occur among mesophiles living on the floor of the rain forest. On moun- tains, however, -w mutants more often survive, species often become -w, and the -w stocks persist and sometimes radiate. Tliis process probably occurs increasingly often with increasing altitude; it occurs most often among mesophiles, sometimes among hydrophiles, rarely if at all among arboreal Carabidae in New Guinea. In extreme cases this process has produced groups of interrelated -w species, which I have considered genera, confined to small
areas in the mountains of New Guinea. Examples are Nehriagomtm and Laevag,o- niim at high altitudes on the Bismarck Range, etc. (Part II, pp. 235, 246; present part. Tax. siippL).
[22] Ex))lantions of wing atrophy. Wings and wing atrophy of Carabidae in various environments and on mountains and islands have been discussed by me in 1936 and 1943. I want now to summarize and apply my conclusions to the situation in New Guinea.
Their relationships indicate that most -w Carabidae on mountains in New Guinea are derived from ancestors that were +w when they reached the island [23, 88]. Tlie -w mountain faunas are mainly products of conversion rather than of concentration (see my 1943 paper, pp. 52-53); that is, they have been produced by atrophy processes that have occurred locally, not by accumulation of -w stocks from other regions.
Certain factors should be noted that do not induce wing atrophy among mountain Carabidae or that are unimportant. Alti- tude itself (thinness of air) apparently does not induce wing atrophy or flightless- ness (see my 1943 paper, pp. 50-51). Cold apparently does not directly inhibit de- velopment of wings (ibid., p. 51). And exposure to wind — selection of -w form? by blowing away of -f w flying individuals (Darwin's factor) — seems at best relatively unimportant (ibid., p. 51), for many -w mountain-living Carabidae, including mosi of the ones on mountains in New Guinea live on the ground in dense montane foresi and are not exposed to winds.
Some other factors that probably do in- crease incidence of wing atrophy anc flightlessness on mountains seem to be oi| only minor importance. One is freedoirj from flooding. In the lowlands of New Guinea, flat ground is often flooded by^ standing water, and winged Carabidat often do fly to escape the floods, whik flightless individuals are obviously mor<i likelv to be drowned or taken by predator;'
The carabid beetles of New Guinea • Darlington 171
(which gather over floods) before they can escape. Flood water does not stand on mountain slopes, and absence of flood- ing may reduce the selective advantage of flight on mountains, but it can hardly be primarily important. Another, indirect fac- tor that increases incidence of wdng atrophy and flightlessness on mountains is the increased proportion of mesophiles there. Some important classes of hydro- philes (those associated with extensive swamps and with the margins of big rivers) are absent at high altitudes, and arboreal species too are relatively few. High-mountain carabid faunas do consist mainly of mesophiles [25], and carabid wings do atrophy much more often among mesophiles than among other ecologic groups. However, wing atrophy and flight- lessness are rare at low altitudes even among mesophiles, so that increase in pro- portion of mesophiles in mountain habitats i-annot be the principal explanation of the very great increase of flightlessness with increasing altitude on the mountains of New Guinea.
There remain, I think, three factors that may be more important in favoring wing atrophy and flightlessness of Carabidae in mountain habitats in New Guinea, as elsewhere. First is an indirect effect of temperature. Gold, although it does not directly inhibit development of wings, does reduce ease and usefulness of flight. Cold at high altitudes, especially cold at night, probably shifts the selective advantage toward flightlessness.
A second probably important factor (inversely) correlated with altitude is in- tensity of competition with ants ([27], Fig. 11). Ants decrease very strikingly with increasing altitude on mountains in New Guinea (as elsewhere); they may be entirely absent at highest altitudes; and their fewness or absence may significantly reduce the selective advantage of wings and flight among mountain-living Garab- idae. Perhaps this decrease should be considered a supplementary effect of cold.
if cold is an important factor in reducing the incidence of ants at higher altitudes.
Third and last, but perhaps most im- portant of all in reducing usefulness of flight and thus favoring atrophy of wings on mountains, is limitation of area. I have discussed this factor at some length in 1943 (pp. 42-44, 53). Its effect is due to the fact that Garabidae at low altitudes seem often to live on an unstable, con- tinually changing mosaic or checkerboard of habitats, and that the principal function of wings and flighi of ordinary mesophile Garabidae may be, by continual redisper- sals, to maintain scattered populations in large but unstable areas. This function may become even more important in the lowland tropics, where distributions ap- parently often are patchy even in relatively continuous habitats (Wilson, 1958). With increasing altitude in the mountains, areas become smaller. My impression (based on what I have seen in the West Indies as well as in New Guinea ) is that many high- mountain species of Garabidae may be confined to areas of a few or a few hundred square miles, while many lowland species are spread over tens of thousands or hundreds of thousands of square miles, a difference of the order of 1:1,000. And the mountain areas may also be ecologically more favorable and more stable, with cara- bid populations denser and more stable. Under these conditions the primary func- tion of wings and flight in maintaining discontinuous populations is probably largely lost, allowing the wings to atrophy among those Garabidae (especially meso- philes) which have no other special need to fly.
[Dr. Eugene Munroe adds (in a recent letter) that "there is very possibly also a positi\'e selection for winglessness in con- serving local concentrations of individuals sufficient to ensure a reasonable probabil- ity of mating under unfavorable conditions. Reductions of wings in arctic, mountain, insular, subantarctic and winter insects seem to be related in this respect and also
172 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
related to simplification of courtship J5at- terns, to parthenogenesis and (in Diptera) to autogcny."]
[23] Summon/ of wing .state and wing atrophy. In summary of wing state among New Guinean Carabidae: of 434 lowland species, about 96 per cent are fully winged and only about 4 per cent have dimorphic or reduced wings, and in most of these cases wing atrophy has not been long- standing and has not been followed by much secondary structural change. The 161 species that occur in both the lowland and the mountain faunas are all fully \\ ingcd except that two species are locally dimorphic on one mountain range. But of the 215 species found only above 1000 m, 32 per cent have reduced wings, and of the 21 species known to occur above 3000 m, 95 per cent have reduced wings, and secondary structural changes and radiations indicate that many mountain- living groups have had atrophied wings for a relatively long time. However, all or almost all the wing-atrophied carabid stocks at all altitudes on New Guinea seem to have been derived by reduction of wings of ancestors which were +w when they reached the island. In explanation, I think (for reasons given in more detail in 1943) that wings of carabids atrophy in moun- tain habitats not primarily because flying is disadvantageous (although it may be- come so) but because the advantages of flight are largely lost in small but favorable and stable areas on mountains. Wings then complicate development and waste energy, so that selection favors -w in- dividuals and evolution and radiation of -w stocks. Of course all this is still prob- ably an extreme oversimplification. Ecol- ogists (e. g., Greenslade, 1968) are just beginning more detailed studies of the complex effects of altitude on carabid faunas.
[24] Ecology: habitats. Although I agree with Brass (1964: 208) that the time has perhaps arrived for leaving New Guinean ecology to the ecologists, I want to say
something about major habitats in New Guinea, because knowledge of habitats is important to understanding of the carabid fauna. The infomiation I can give is not very detailed, and some of the general- izations are only approximations, but they are substantially correct and useful. I say this in explanation, not in apology.
Several useful papers describing New Guinean vegetations are listed in my bibliography, but I have chosen to follow Brass (1941; 1964) because his system is based on practical field experience and is relatively simple and well adapted to my purposes. Actually, my classification of vegetations, although based on Brass, is even simpler than his, because I know so little of the ecology of New Guinean Carabidae that I can fit them into only aij very simple system. For my purposes, P recognize the following principal vege- tations, with which the distribution of Carabidae is correlated.
"Rain forest" includes coastal forest, monsoon forest, true interior lowland rain forest, and various types of wet mountain forest. It covers very large areas of New Guinea from sea level to high altitudes on the mountains, with tongues or patches of "subalpine forest" even up to 13,000 ft. (nearly 4000 m) in places. The best lowland rain forest has a high, closed can- opy; vines and epiphytes are numerous but underbrush is usually scanty; and leaf debris and leaf mold are thin on the ground, except under the heads of recently fallen trees. At higher altitudes the forest trees are lower, and at highest altitudes they are extremely stunted; the trees them- selves may be so low that there is not much clearance between them and the ground; and their trunks and branches are often clotheel or buried in moss, especially where clouds supplement rainfall to main- tain almost constant heavy moisture.
Savanna forest includes what I call "open eucalyptus forest." It is drier than rain forest and has a relatively open canopy (with the trees spaced well apart), few
I
The carabid beetles of New Guinea • Darlington 173
or no vines or epiphytes, and usually a ground cover of grass. This kind of forest is primarily Australian, but a virtual ex- tension of the Australian forest (presum- ably dating from the recent land con- nection) covers parts of southern New Guinea too, including the area around Port Moresby.
"Grassland" consists (to my untrained ?ye) of two principal types. In the low- lands, tall grass ("kunai") covers more or less extensive areas around or interspersed in the still more extensive areas of rain forest. Some of this grassland is natural; >ome, the result of cutting and burning of -ain forest by native farmers. What seemed :o me to be similar grassland, at least offer- ng a similar habitat to Carabidae, occurs jp to at least 2000 m in inhabited valleys Dn the Bismarck Range and elsewhere, .vhere man has cleared the mid-altitude nountain rain forest. The second, surely latural type of grassland, with the grass much lower than "kunai" and forming tus- socks, occurs above the forest line on high nountains, including the Bismarck Range.
"Wetlands" can be used to include a variety of habitats where running or stand- ing water or simply wet ground support special vegetations and special Carabidae. Swamps are widespread at low altitudes in New Guinea, but are relatively few and small in the mountains where drainage is more rapid. The lowland swamps include large areas of sago palms, especially on the deltas of the big rivers. The enomious, spine-bearing leaves of these palms fall into the water and make working there difficult and dangerous. This habitat has therefore not been well collected for Carabidae, although it is evidently rich in subaquatic species. Other types of swamp with more diverse vegetations occur here and there, especially inland. And salt swamps, man- grove zones, and other special habitats occur along the sea coast and probably support special Carabidae, although very few of them have been collected. The edges of running water have also special
Table 6. Ecologic composition of lowland New Guinean Carabid fauna
No. in
Ecologic No. of ecologic
groujis species subgroups
Fossorial 27
( Scaritini ) Mesophile 133
Rain forest majority
Open places minority
Hydrophile 136
Wet lands 105
By streams 31
Arboreal 129
In foliage 84 ( incl. 29
Demetrida )
Trunks/logs 45
Unclassified 9
habitats ranging from fringes of dense forest to grass, reeds, and virtually sterile banks and bars, which may be stony, sandy, or muddy. Different habitats on the banks of large rivers, the edges and debris-blocks of smaller streams nmning slowly in heavy shade in rain forest, and the stony and sandy edges of rapidly run- ning brooks all have different carabid faunas. Many of these habitats are found only or chiefly in the lowlands, but rapid brooks occur at all altitudes up almost to the peaks of the highest mountains. Be- sides these larger and more obvious sub- divisions of wetlands. New Guinea presents an almost endless variety of wet spots sometimes only a few square meters in area: rain pools, overflow of streams, ac- cumulations of water in holes left by the roots of fallen trees, etc., and some com- mon carabids inhabit these places.
[25] Ecologic composition. The ecologic composition of the louland carabid fauna of New Guinea (Table 6) is shown by figures compiled from my data sheets [16], from the column headed "Ecology." Be- cause my information about habitats is incomplete, I have assigned some species according to the habitats of their nearest relatives; these assignments are probably correct in most cases, but perhaps wrong
174 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
in a few.^ And in nine cases I have not wanted even to guess about the habitats of isolated species. The totals are there- fore only approximations, but they are close approximations, based on my field observations.
Even when habitats are known, some species are still difficult to assign to eco- logic groups. The fossorial species might be counted as hydrophiles or (a few) as mesophiles. The line between mesophiles and hydrophiles is not sharply drawn in any case; it is crossed by various species with wide ecologic ranges. The distinction between mesophiles, rotten-log forms, and tree-trunk-arboreal forms is not sharp either. Nor is the line between mesophiles and fohage-arboreal forms; it is crossed by species like Vioh^onum viohceum, which occurs in piles of dead leaves on the ground as well as in foliage. Never- theless, these doubts and difficulties do not seriously affect the approximations given in Table 6.
In Table 6, the fossorial Scaritini are separated first, although most of them are included with the hydrophiles in statistical analyses elsewhere in the present paper. Without the fossorial fonns, the lowland Carabidae of New Guinea divide almost equally into three main ecologic groups: mesophiles, hydrophiles, and arboreal forms. This is probably a fair sample of the ecologic composition of lowland cara- bid faunas in the wet tropics elsewhere,
^ Cases in which more or less closely related carabid species occur in different haliitats include the following in New Guinea. Although most Tachijs occur on the ground in wet places, T. aeneus (Part I, p. 46.3) occurs on dnj ground, and r. wallacei (Part I, p. 479) is arl)()real. Most Oodes are more or less aquatic, but O. tene.s- tris lives in leaf litter on the ground in rain forest. And, although most E^adroma li\'e in wet places, E. robusta (Part III, p. 71) li\'es on dry ground. For some examples of carabids which have made ecologic shifts within tlie limits of New Guinea, see [91]. All these cases of ecologic divergence have been allowed for in compiling Table 6, but some other, similar cases may still be undetected.
although hydrophiles may be relatively more numerous in some places ( Darlington, 1943: 41).
In the case of the New Guinean carabid fauna, all three main categories can be usefully subdivided.
The mesophiles can be divided into rain- forest forms (the majority) and those that inhabit open places (a minority). I have not tried to give exact figures for these subgroups, because my information about the occurrence of some of the species is insufficient.
The hydrophiles divide into those that occur on wetlands in general (105 species) and those that occur only by running water ( 31 species ) . The line between these two subgroups is not sharp. However, the division is important because the stream- side species are the only ones that are likely to range far up the mountain slopes. The distribution of hydrophiles depends on distribution of surface water and of special water-side habitats, and some of these habitats are independent of the type of forest cover. Some hydrophiles therefore occur in both rain-forested and opener country, and their dispersal is compar- atively little affected by discontinuities in the distribution of rain forest. Also, most of them are winged and many fly actively. They are therefore likely to be good dis- persers in areas (like the Malay Archi- pelago) where forests and opener country alternate to some extent (see [88, 91]).
Finally, the arboreal forms divide into those that live in foliage (84 species, in- cluding 29 lowland Dernetrida) and those that live on tree trunks and recently fallen logs ( 45 species ) . Almost all these arboreal carabids occur in rain forest, of course.
With increasing altitude, the ecologic composition of the New Guinean carabid fauna changes strikingly. At higher alti- tudes, most Carabidae are mesophiles living in (montane types of) rain forest,{ Most hydrophiles have disappeared, ex- cept a few that are strictly associated with running water. And arboreal Carabidae
The carabid beetles of New Guinea • Darlington 175
are reduced in numbers too, but include relatively more species of Demetrida. I know too little about the habitats of most ?pecies at higher altitudes to give exact Figures.
[26] Altitude. Altitude affects the dis- tribution of Carabidae in New Guinea in several ways. The change of genera and species from the lowlands to the highest altitudes is so great that, although ex- tensive overlapping occurs at intermediate altitudes, no species are common to the lowland and highest-mountain faunas and virtually all the genera are different. (See Table 2 for composition of the highest faunas. ) The only genus that is represented at all at highest and lowest altitudes is Scopodes, but this genus is not (in New Guinea) a regular component of the low- land fauna, and the one or two mid-alti- tude species of the genus that do descend to the lowlands at least locally are not closely related to the single very-high- altitude species. The effect of altitude is shown also by the striking reduction of numbers of species from the lowlands to the mountain tops (Fig. 7); by the dif- ferent size distributions of lowland and mountain species ( Figs. 8, 10 ) ; by the different ecologic compositions of lowland and mountain faunas [25]; and by the increased proportion of species with atro- phied wings, from about 4 per cent in the lowland fauna to about 95 per cent at highest altitudes [21]. See Szent-Ivany (1965) for further notes on the vertical distribution of some beetles in New Guinea. See Greenslade ( 1967 ) for an indication of the correlation of habitats and of insect distributions with altitude in the Solomon Islands. And see Mani ( 1968 ) for a more general discussion of high-altitude insects. Note, however, that even at highest altitudes New Guinean Carabidae do not show all the modifi- cations of alticoline forms that are found in some other parts of the world; for example, I have found no general reduction
of size of individuals even at highest alti- tudes (cf. Mani, 1968: 58ff).
The effect of altitude on New Guinean Carabidae is apparently exerted partly in- directly, through the effect of altitude on habitats. Some species that are common in the lowlands occur also up to at least about 2000 m where forest has been cleared and where for this or other reasons habitats, including grassland, approximate lowland habitats. For example, I found the following common lowland species of Carabidae in cleared country in the Waghi and Chimbu valleys at between about 1500 and 2100 m altitude: Tachijs fasciatus and T. aeneiis (Part I, pp. 414, 463), Notop,o- niim angusteUum, vile, and margaritum (Part II, 133, 135, 145), Chlacnius flavi- giittatus (Part III, p. 26), Egadroma quinqucpustidata and rohusta (Part III, pp. 70, 71). It seems likely that these and other winged lowland species have invaded the Waghi and Chimbu valleys from below only after the mountain valleys had been cleared. The climatic differences have not prevented the invasions, although the ecologic ranges of some of the species did apparently change with increasing altitude: some species that apparently never hide under stones at low altitudes do so in the Waghi-Chimbu area [3]. Many of the lowland species that have been found at Wau [5] at 1200 m altitude or higher have probably invaded cleared land there.
In undisturbed forest and other natural habitats the changes of carabid faunas with altitude are apparently more clear-cut. My infonnation about occurrence of Carab- idae in forest at mid-altitudes is scanty. However, I do know that all species of Carabidae found in montane forest on the Bismarck Range between about 2100 and 3000 m and all species found in open tussock-grass country above about 3000 m are different from lowland species, and many of the genera are different.
It is not yet possible to assess the relative importance of the direct effects of high-
176 Bulletin Museum of Comparative Zoologij, Vol. 142, No. 2
ANTS
ANTS
C A R A B
D A
GEO PHILES
iHYDROPHILES 'ARBOREAL
|
-w :=; 1 1 |
+ W |
4-W |
+W |
|
|
COOL |
TEMPERATE |
E. |
N. A. |
Ti- ll
II ii
\ ( \
\
V
I t
\ \
\
s.
+w
i; TROPICAL
+ W
-I- W
LOWLANDS OF NEW GUINEA
I
C A R A B
DAE
Figure 11. Diagram of (hypothetical) relation of Carabidae to ants in cool temperate eastern North America and in the tropical lowlands of New Guinea. Redrawn and modified from Darlington, 1943, p. 42, fig. 4. Diagram is intended to sug- gest that ants compete most severely with flightless (-w) ground-living (geophile) Carabidae; that the effects of competition are greater in the tropics than in the temperate zone; and that dominant ants replace most flightless ground-living Carab- idae in tropical lowland New Guinea.
altitude climate (cold, continual dampness in some habitats, perhaps thinness of air) on carabid faunas. I can only say that, judging from the restriction of high- mountain species to special montane habi- tats, the indirect effects are probably more important than the direct ones. The in- direct effects are probably exerted not only by control of vegetation but also by limi- tation of habitats on mountains (absence of large swamps, large rivers, etc.), by hmitation of area on mountains (which probably affects nature of populations and, indirectly, state of wings [22]), by reduc- tion of competition with ants [27], and perhaps in other ways.
[27] Ants. Ants are dominant insects, especially in the tropics, and their impact on other insects must often be tremendous. I have suggested that they may affect the size distribution of Carabidae in New Guinea [20] and that their fewness or absence at high altitudes may modify carabid faunas there [26]. Ants may also
modify the ecologic composition of the lowland carabid fauna of New Guinea. This fauna, as compared with the faunas of temperate regions, is deficient in meso- philes and especially deficient in -w mesophiles, which probably compete most directly with ants. The ecologic compo- sitions of the carabid faunas of cool temper- ate eastern North America and of the tropical lowlands of New Guinea are diagrammed and compared in Figure 11. The diagram is intended to suggest that the tropica] fauna, which is larger than the temperate one in any case, would be still larger if it were not for the presence of ants, and that the latter take the place mainly of mesophiles and especially ol flightless mesophiles.
[28] Ecologic interactions. The differeni ecologic factors discussed above do noi. act independently but must interact ir very complex ways on carabid faunas Major habitats and vegetations profoundh affect occvnrence of Carabidae but an
The carabid beetles of New Guinea • Darlington 111
themselves dependent on climate. Altitude may act partly in relatively direct ways (by reducing temperature, increasing pre- cipitation, modifying insolation, and per- haps in other ways) and partly indirectly (by modifying habitats and vegetations). And the effect of ants varies with habitat and altitude. So, climate, habitats, altitude, and ants all interact complexly to modify the environments in which Carabidae live and to determine, directly or indirectly, how many Carabidae and what kinds of Carabidae live in different places. This statement does not do justice to the actual complexity of the environment in New Guinea. The ecologic ranges of different Carabidae on the island are surely deter- mined by the interactions of very many factors derived from the inorganic environ- ment, vegetations, and animal prey, preda- tors, parasites, and competitors. Actual details are almost w^iolly unknown and their investigation must (as I have indi- cated at the beginning of [24] ) now be left to the ccologists.
[29] Ecologic ranges. I have used the phrase "ecologic range" deliberately, rather than "niche." "Niche" is used by many ecologists, and I have used the word my- self, but I think it is inaccurate and mis- leading. It suggests that the environment is full of pigeonholes with fixed bound- aries, and that different kinds of animals and plants are in fact neatly pigeonholed in nature. Experienced ecologists know that this is not so (at least I hope tHey know it!), but the concept of niche some- times confuses them just the same.
For example, ecologists sometimes speak of an island as having a certain number of niches for animals, with some of the niches unfilled, as if the niches existed before the animals were there. But if niches exist for animals at all, it is the animals themselves that make them; dif- ferent kinds of animals surely in part de- termine the ecologic limits of other kinds. And the limits apparently change accord- ing to the number of animals present.
When only a few^ species are present on an island, each is likely to occupy a wider segment of the environment than each will occupy when more species are present. To say that, in the first case, each species occupies several niches does not avoid the difficulty; there is no ultimate niche, no division of the environment so fine that it cannot be further divided. One might almost say that, instead of an island possess- ing a certain number of niches some of which may be unfilled, an island has no niches at all until animals come and define them. What the island does have is eco- logic dimensions or ecologic amplitude or ecologic totality, which is not divided be- forehand but which animals do divide into a few or many fractions in the course of time.
It seems to me that ecologic ranges are comparable to geographic ranges. A given piece of land — say a continent — has geo- graphic dimensions, and is potentially able to support a varying number of species of plants and animals, each, of which has a geographic range. The ranges are areas occupied by populations. They do not exist until the populations occupy them, and they often change and often overlap. At- tempts have been made to divide continents into which might be called idealized (and therefore typological! ) geographic ranges corresponding to climatic zones or major biomes, but (as a zoogeographer) I know that actual species' ranges often do not fit the idealized patterns very well. Similarly, to suppose that the enviionment can be divided into niches which exist before animals occupy them is a (typological) idealization which is likely to obscure the real facts. It is better to think of each species as having an ecologic range which is a fraction or fractions of the total en- vironment. It is then easy to think of ecologic ranges as changing from time to time, or as being discontinuous, or as over- lapping each other, as they often do. The concepts of "changing niches" and "over- lapping niches" are confusing!
178 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
ZOOGEOGRAPHY: EXISTING GEOGRAPHIC PATTERNS
[30] Geographic patterns, relatiomhips, and origins. Because I am a zoogeographer, especially interested in the patterns and significance of animal distribution, I shall make a geographic analysis of the New Guinean carabid fauna in much more detail than would ordinarily be attempted in a faunistic paper. In doing this, I shall begin by making a clear distinction be- tween description of existing geographic patterns and discussion of past origins. This is just the beginning of the sub- dividing of the subject that is necessary to describe and understand the zoogeog- raphy of any complex fauna. Few persons, in fact few zoogeographers, fully under- stand how complex the subject it, or how necessary it is to recognize the complexities if zoogeographic data and analyses are to be informative and significant. I doubt if faunal relationships and other zoogeo- graphic problems can yet be handled satis- factorily by simple numerical methods. These methods are too likely to give a superficial picture that hides more than it reveals.
The gross taxonomic compositions of major carabid faunas from the Orient (Java) to New Guinea and tropical Aus- tralia are described and diagrammed in [64] and Figure 13. Now to be considered are the finer details of distributions and relationships that underlie the gross pat- tern.
[31] Existing geographic relationships: problems and procedures. I shall begin by describing existing geographic patterns and relationships without intending to imply anything about past geographic origins (except sometimes parenthetically). The first question in this connection is, at what level are geographic patterns most signifi- cant: at the level of tribes, genera, or species, or of geographically separate stocks?
The choice of levels can profoundly
affect the results of analysis, and the choices are often complex. For example, Coptodera (Part III, pp. llOff), as I have treated it, is one genus; splitters would divide the New Guinean species among about five smaller genera; eight species of the genus occur in New Guinea; and these eight species represent seven geographi- cally separate stocks, each with its own relationships outside New Guinea. In mak- ing faunal comparisons, should this genus be counted as one or five or eight or seven units? This is the kind of situation that numerical zoogeography too easily hides.
For another example, the genus Deme- trida (Part III, p. 140) is represented on New Guinea by 59 known species, all apparently interrelated among themselves (apparently produced on New Guinea by radiation of perhaps only one or certainly not more than a very few ancestors). Out- side New Guinea, this genus is principally Australian, but most of the Australian species are interrelated among themselves (and probably represent separate evolu- tionary radiations). There are certainly very few, perhaps only one, separate pri- mary bond(s) of relationship between thci New Guinean and Australian members of the genus. In making analyses and geo- graphic comparisons, the zoogeographer must decide whether the New Guinean Demetrida are to be treated as 59 units with Australian relationships or as one unit with Australian relationships. This de- cision makes a substantial difference in the conclusions. This genus compriseSj nearly 9 per cent of all known New Guinean carabid species. If it is decided to count all the species separately in making faunal comparisons, this decision alone in- creases the "Australian" fraction by about 9 per cent of the whole fauna. But if it i^ decided to count Demetrida in New Guinea as one geographic unit, its weight in the] sum of geographic relationships of the New Guinean carabid fauna is a fraction of 1 per cent. The decision how to count Demetrida is even more important ir
I
The carabid beetles of New Guinea • Darlington 179
analysis of the geographic relationships of the arboreal fraction of the carabid fauna, for Demetrida makes up nearly one- third of the arboreal Carabidae of New Guinea!
My solution of the complex problem of choice of units for zoogcographic analyses and comparisons is two-fold. First, I think it is essential to describe situations in words, giving counts of genera, species, and other units, but going beyond mere presentation of figures. I shall do this for a succession of tribes of Carabidae that occur in the Asiatic-Australian area [32ff]. And second, after the situations have been described, I think arbitrary decisions have to be made and the reasons for them given. In the present case I think the decision should be to base statistical analyses pri- marily on geographic units — stocks with independent geographic relationships — be- cause they best show the degree of actual interrelationship (and also the geographic histories ) of faunas. In Coptodera and Demetrida (above) the geographic units — the number of separate bonds between the New Guinean carabid fauna and other famias — are seven and one (or very few) respectively.
Decisions about geographic relationships are complicated by differences in degree of relationship. For example, Cilleniis (Part I, p. 399) is represented in both the Orient and Australia, but the New Guinean spe- cies is evidently more closely related, to Oriental than to Australian forms. I have scored it as one (Oriental) geographic unit; and in other similar cases in which single stocks have relatives in more than one region, and in which closeness of re- lationship varies, I have scored only the closest relationship in each case. When, however, relationships seem equally close with (say) Oriental and Australian forms, I have scored both (see following para- graph). Obviously these decisions are arbitrary in principle and often also in practice.
Another complication occurs in cases like
Mecyclothorax (Part I, p. 505; present part, Taxonomic supplement) and Scopodes altus (Part III, p. 198), in which relation- ships with Oriental and Australian forms seem about equally close, but in which the related forms are numerous in one region (in Australia, in the examples given) and relatively few in the other region. In such cases I have disregarded numbers and have scored geographic units solely ac- cording to apparent closeness of relation- ships. Mecyclothorax and Scopodes altus therefore each score one (Oriental) and one (Australian) geographic unit.
Statements of existing relationships are still further complicated by great differ- ences in extent of the areas occupied by different genera, species, or other units. For example, Demetrida as a genus occurs throughout New Guinea and throughout Australia. If the New Guinean species are treated as a single geographic unit (see above), it is surely an Australian-related one. But besides the underlying Austra- lian relationship of all the New Guinean Demetrida, a second link with Australia is formed by one New Guinean species which occurs also on the extreme tip of the Cape York Peninsula (see Demetrida angidata. Part III, pp. 143, 159). Is this to be counted as a second Australian- related geographic miit in spite of the fact that the area occupied in Australia is very small? Or take the case of Syleter (Part I, p. 356), a genus which ranges from the southeastern comer of Asia to the Philippines and New Guinea, with the single New Guinean species occurring again on the extreme tip of Cape York. In terms of present distribution is this to be counted as one Oriental-related unit and also as one Australian-related unit, and if so, are the two units to be given equal weight in spite of the different extents of distribution in the Oriental Region and Australia? I think that the solution of this problem too (and of others like it) is, first, to make the situation clear, and then to make an arbitraiy de-
180 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
cision, and give the reason for it. In the present case my decision is to count as Australian-related only those geographic units that occur below the Cape York Peninsula in Australia, and to treat sepa- rately those New Guinean species that ex- tend only to Cape York ( see Table 9 [63] ) . This distinction seems to me to reveal im- portant differences in distribution patterns (and to facilitate deductions about geo- graphic histories). In the other direction, it seems to me most useful to treat as Oriental-related those geographic units that occur in Celebes or the Philippines as well as units that occur in the Oriental Region proper, but to treat separately those that occur west onl\ to the Moluccas [62; 64, footnote].
The best way to establish a basis for zoogeographic analysis of the New Guinean carabid fauna — the way that will give the most information — is, I think, to describe the relationships of New Guinean carabids tribe by tribe, emphasizing the separate geographic units as far as they can be recognized, including but not overempha- sizing counts of genera and species, and adding other significant details where pos- sible. I can do this only because I myself did much of the taxonomic work from which the details are derived.
Most tribes of Carabidae are widely dis- tributed. A few small tribes are localized in various parts of the world, but no tribe is confined to or represented mainly in New Guinea.
As background to understanding the relation of the New Guinean carabid famia to other famias, I shall begin by simimariz- ing the distributions of certain tribes which do not occur in New Guinea but which form part of the broader pattern of geographic relationships and of change of faunas from Asia to Australia. First will be noted tribes which (in this part of the world) occur only in Asia above or chiefly above the tropics, then tribes which occur in tropical Asia and extend toward but not to New Guinea, then a few tribes that
occur in both Asia and Australia but skip New Guinea, and finally tribes that (in the part of the world under consideration) occur only in Australia. And then I shall consider the tribes that are represented in New Guinea, taking them in the order of the Coleoptewrum CataJoiius, and giving in each case all the geographically signifi- cant details that I can. The nature (size, wings, etc.) of the members of each tribe will usually be noted too, and also major habitats. (Although this survey is con- cerned only with existing distribution pat- terns, some details will be referred to again in discussion of the origins and di- rections of dispersal of the New Guinean carabid fauna. )
[32] Distribution of earahid tribes from A.sia to Australia. I shall now attempt to summarize, mainly by tribes, the distri- bution of Carabidae along a strip of the earth's surface extending almost from pole to pole, from northern Asia south and east across the Indo-Australian Archipelago (in- cluding New Guinea) to southern Aus- tralia. ( See Fig. 1 for a map of the tropical portion of this area.) Details of distri- bution in the islands and Australia are mostly from my manuscript lists (see [4]).
Asia north of the tropics possesses several tribes of Carabidae wliich do not reach tropical Asia south of the Himalayas or are very poorly represented in the tropics. These northern tribes include Cychrini, Nebriini, Notiophilini, Opisthiini, Elaphrini, Loricerini, Patrobini, Amarini, and Zabrini. Two additional tribes well represented in Asia north of the tropics but not or poorly represented in tropical Asia, but present in Australia, are the Carabini and Broscini; these tribes are referred to again below. Several genera of Carabidae are outstand- ingly dominant in temperate Asia but absent or veiy poorly represented in the tropics; they include Carabus, BemI)idion (which occurs also in Australia, see be- low), and llarpalu.s. Asia north of the tropics is notable also for possessing relatively numerous Pterostichini but rel-
The carabid beetles of New Guinea • Darlington
181
atively fewer Agonini and Lebiini. The mainly northern tribes and genera of Carabidae include many mesophiles and many hydrophiles but few arboreal forms. Tropical Asia and associated continental islands (Sumatra, Java, Borneo) possess a rich and diverse carabid fauna. Six small tribes are represented by single genera on the mainland of tropical Asia (most are represented in Africa too) but do not reach the islands; they are Enceladini, Melacnini, Anthiini, Disphaericini, Grani- gerini, and Idiomorphini. One additional tribe of a single genus, the Mormolycini, is confined to the Malay Peninsula, Sumatra, Java, and Borneo. Other tribes represented in tropical Asia (and more or less widely distributed elsewhere) extend for varying distances eastward across the Indo-Aus- tralian (Malay) Archipelago. The Hiletini (Camaragnathini) (one genus) have been found ( in the area in question ) only on the southeastern corner of Asia, Sumatra, and Borneo; the Omophronini (one genus), on the mainland of Asia and on Luzon in the Philippines (Darlington, 1967). The Sia- gonini (one genus) are numerous in tropical Asia and diminish eastward, the easternmost knowm species occurring on Java and on Mindoro in the Philippines (Darlington, 1967). And the Orthogoniini (one principal genus), with many tropical Asiatic species, diminish eastward, reaching the Philippines and Moluccas but probably not New Guinea. Some important genera of other tribes have similar patterns of diminution from tropical Asia eastward. For example, the dominant, widely dis- tributed genus Scarites (large fossorial Carabidae) is well represented in the tropics of Asia, has tvvo species in the Philippines, at least one on Celebes, and one on Timor, but does not reach New Guinea or Australia. The primarily Oriental Trigonotoma (rather large, ground-living, mesophile Pterostichini ) reaches the Philip- pines and Moluccas but not New Guinea or Australia. And CoIIida (medium-sized, arboreal Lebiini) occurs in tropical Asia
(and other tropical regions) and eastward to the Philippines and Celebes but not New Guinea or Australia (where its place is taken by Demetrida).
Asia and Australia share several tribes and important genera that are absent on most or all of the intervening islands in- cluding New Guinea. In the tribe Carabini, the genus Calosoma (large, ground-living, mesophile and xerophile Carabidae) is well represented in Asia above the tropics (and one or two species reach the tropical part of India), and two species occur in Australia, but the genus does not reach the southeastern corner of Asia and does not occur on the islands between Asia and Australia (except for the extension of one Australian species to New Caledonia and to the eastern Lesser Sunda Islands); this enormous gap in distribution is difficult to explain, for the genus does occur con- tinuously across the tropics in Africa and America. Tlie tribe Broscini (medium- sized and large, ground-living or jDartly fossorial, mesophile and xerophile Carab- idae) includes several genera in temper- ate Asia and several others in southern Australia, but few enter even the edges of the tropics, and none occurs on the islands between Asia and Australia (see Ball, 1956; and Darlington, 1965: 39-42, for discussion of the genera and zoogeog- raphy of this tribe). Pogonini (rather small, ground-living Carabidae often as- sociated with saline or alkaline habitats), too, include several genera on the main- land of Asia and several in Australia, including some in the tropics in both places, but are nearly or quite absent on the intervening islands. T\\'o tribes, Apo- tomini and Amblystomini (each consisting of one principal genus of small, ground- living, mesophile Carabidae), are well represented in Asia and extend eastward across the islands to the Philippines and Celebes or the Moluccas and occur again in Australia but probably do not occur in New Guinea; both have been recorded from New Guinea, but probably in error
182 Bulletin Museum of Comparative Zoology, Vol 142, No. 2
(Part I, p. 398; Part III, p. 19). And the following genera of small Carabidae (all inclnding both hydrophile and mesophilc species) are dominant north of the tropics, extend into (different parts of) the west- em Indo-Australian Archipelago, and occur again in Australia, but are absent in New Guinea: Dyschirius (see under tribe Scari- tini, below): Bembidion (excluding Cil- lenus, see under tribe Bembidiini); and Trechus (see under tribe Trechini).
Finally, five tribes are (in the region under consideration) confined to Austra- lia: Pamborini (one genus), to eastern Australia from the base of the Cape York Peninsula to south-central New South Wales; Migadopini, Merizodini, and Agoni- cini (with two to four genera each in Australia), to the southeastern comer of Australia and Tasmania; and Cuneipectini (one genus), to southwestern Australia. These tribes all consist primarily of ground- living mesophiles. They include both large (Pamborini, Cuneipectini) and small (Meri- zodini) forms. All existing Pamborini, Agonicini, and Cuneipectini have atrophied wings; the other tribes are winged or in- clude winged fomis.
I shall now proceed from the tribes that occur in adjacent areas but that do not reach New Guinea to tribes represented on the island itself, taking them in the order of the Coleopterorum CataJo'^ius (Csiki, 1927-1933). AU the tribes rep- resented on New Guinea are primarily winged, although some include secondarily flightless forms. See again Figure 1 for a map of the area chiefly concemed in the following descriptions of distributions.
[Although tiger beetles (Cicindelinae) are not included in my summary and analysis of New Guinean Caribidae, their distribution may usefully be summarized parenthetically.
The tribe Collyrini (Part I, p. 334) is confined to the Oriental Region except for extensions eastward described below. The tribe consists of only two genera, both arboreal. CoUyris, of which the species are
winged and (probably) live on foliage, is rich in species throughout the Oriental Region and the Philippines, with one spe- cies on Timor and one across the central part of Wallace's Line on Celebes (and supposedly reaching the Aru Islands) but none on New Guinea. Tricondyla, which is flightless and lives on tree trunks, is also well represented throughout the Oriental Region; one species is endemic on Celebes and two occur on the Philippines; and one of the Philippine species extends south and east to New Guinea (Part I, p. 334), the Solomons, and Timor, and has been found in the mid-peninsular rain forest of Cape York, Australia, although it apparently does not occur in the forest on the tip of Cape York. For further com- ments on the distribution (and history) of this species see also footnote, section [21].
The tribe Megacephalini (Part I, p. 335) consists of terrestrial, chiefly nocturnal tiger beetles. The principal genus is Mega- cephala, which is discontinuously dis- tributed in the wanner parts of the world In the Asiatic-Australian area, one species extends into northern India (from the Mediterranean region); otherwise the genus is absent in Asia and absent on the islands between Asia and Australia, except that two (winged) Australian species are doubtfullv recorded from southern New Guinea (Part I, p. 336); but about 20 species occur in Australia.
The tribe Cicindclini (Part I, p. a36) includes chiefly winged, terrestrial and subarboreal, chiefly diurnal tiger beetles The tribe is worldwide in distribution. 01 smaller genera, Prothyma is represented in New Guinea by one species with probably Oriental relationships (Darlington, 1947) Caledonomorpha, with two closely relatec species, is endemic to eastern New Guinea, I do not know its geographic relationships' Distipsidera is an eastern Australian genu.' of which two species have been describee from southern New Guinea; these specie; are not directly related to each other bu
The carabid beetles of New Guinea • Darlington 183
Form two separate Australian geographic the AustraHan units being of doubtful
Linits; all members of the genus live on tree occurrence in New Guinea. However, the
trunks, so far as I know. And Therates is arboreal Cicindelinae that live in rain
in Oriental genus of which five or more forest (Tricondyla and Therates, with to-
ipecies, representing at least three Oriental gether four geographic units) are all
geographic units, occur in New Guinea, Oriental; while the terrestrial tiger beetles
although none reaches Australia; these in- of New Guinea are at least as much Aus-
;ects live on under-story foliage in rain tralian as Oriental in present relationships,
-orest. The single remaining genus of the and the Australian units include apparendy
ribe ( so far as New Guinea is concerned ) older stocks as well as ( presumably more
s the dominant, cosmopolitan Cicindela recent) shared species.] Tart I, p. 340), which includes many terres- [33] The pantropical OZAENINI (Part
:rial as well as ( in the tropics ) subarboreal I, p. 351) includes medium-sized and
)r arboreal species. New Guinea possesses small Carabidae which live on the ground
37 known species and three additional sub- usually in forest and are often associated
species of this genus, representing at least with logs and rotting wood. Four winged
3ight separate geographic units (Part I, p. genera occur in the Oriental Region and
341). Two of the (older?) units are one very distinct (primitive?), flightless,
\ustralian in present relationships. One endemic genus is confined to eastern
)r more stocks of very small, at least partly Australia. Two of the Oriental genera,
irboreal species have radiated on New Danija and Eiistia, reach Java and Borneo,
Guinea; their geographic relationships are and the Philippines, but do not cross
lot clear. Two Oriental stocks are rep- Wallace's Line to Celebes, but a third
•esented on New Guinea by endemic spe- Oriental genus, Pseiidozaena, extends east-
:ies {maino and denticoUis) but do not ward to New Guinea etc. but not Australia,
■each Australia. Cicindela discreta occurs The single species of this genus on New
From Sumatra etc. to New Guinea and Guinea is the only member of the tribe
lorthern Australia; C. decern fi,uttata, from there.
::elebes to New Guinea etc. but not Aus- [34] The tribe PAUSSINI (Part I, p.
:ralia; and C. semicincta, in New Guinea 353) contains medium-sized and small,
;tc. and northeastern Australia and on mostly winged, chiefly ground-living, pri-
Sfew Caledonia. marily myrmecophilous carabids derived
In summary of the geographic relation- from Ozaenini. The tribe is pantropical,
>hips of New Guinean Cicindelinae: of but is best represented in the main part
Gollyrini, the island possesses one Oriental of the Old World. Ten genera occur in
>pecies which occurs also on mid-Cape the Oriental Region; all of them extend
York; of Megacephalini, two Australian east to Java and/or Bomeo and in some
species (doubtfully) recorded from south- cases the Philippines, but none is known
?m New Guinea; and of Cicindelini, from Celebes; one species of the Oriental
Prothyma with one and Therates with at genus Etiplatyrhopoliis (Part I, p. 354)
least three Oriental geographic units and has been recorded from New Guinea, but
Cicindela with apparently four Oriental its occurrence there is doubtful. One prin-
ind five Australian units. The presence cipal genus, Arthropterus, occurs in Aus-
of an endemic genus confined to eastern tralia; it is primitive and is not directly
New Guinea and the occurrence of an related to any existing Oriental genus;
extensive radiation of very small species of one species of it has been described from
Cicindela on the island are also note- (southern?) New Guinea and may be the
worthy. Tlie totals are nine Oriental and only paussid really present on the island, seven Australian geographic units, two of [35] The tribe SCARITINI (Part I, p.
184 Bulletin Museum of Comparative Zoology, Vol 142, No. 2
355; Part IV, Tax. suppl ) is the principal, dominant tribe of terrestrial-fossorial Carab- idac. The trilie is worldwide in distri- bution although much more diverse in the tropics than in temperate regions. About 20 genera are represented in the Oriental Region and about 17 in Austraha, but only three genera have been found in New Guinea. For zoogeographic purposes, the large and small members of the tribe should be grouped separately.
Oriental scaritines include about 11 genera of large and nine of small forms. Among the large fornis is Mouhotia, a fhghtless genus of three species confined to the Indochinese Peninsula; this genus is apparently related to the American Pasimachus and perhaps more distantly the Australian carenums (Biinninger, 1950). Most of the other Oriental genera of large scaritines are confined to the mainland of Asia or to Ceylon, but Oxy<^nathus reaches Java; Dlsiichus, Java and Borneo; Thli- bops, Java and the Philippines; and Scarites, limits given in [32]. The small Oriental scaritines are less well known. Most genera of them are apparently local- ized within the Oriental Region (some are represented also in Africa), but the pri- marily northern Dyschirius reaches Java and Luzon and occurs again in Australia (but not in New Guinea), and Syleter and Clivina are discussed below.
Australian Scaritini include about 12 genera of large and five genera of small forms. The large forms are mostly "care- nums," which include about 11 genera and several hundred species, all flighdess, all confined to the continent of Australia and closely associated continental islands (and all probably products of radiation in Australia); a few reach Cape York, but none has yet been found on New Guinea. The carenums may be related to Mouhotia (above) in a general way, but probably not directly and not closely. The only other large scaritines in Australia are a few winged species of Geoscaptus, one of which extends to New Guinea. Small Australian
scaritines include three monotypic, en- demic genera as well as a few Dyschirius (see above) and many Clivina (see be- low); these genera are all primarily winged, although the wings have atrophied in some Clivina. It will be seen that there is no direct or at least no close relationship bet\\'cen the large Scaritini of the Oriental Region and of Australia, but that a few genera (and even some species, see be- low) of smaller, winged forms are com- mon to the Orient and Australia.
New Guinea itself possesses only one Australian species of Geoscaptus (the only "large" scaritine on the isUmd ) ; one species of the African-Oriental genus Syleter, be- longing to an Oriental species group, and extending to the tip of Cape York; and about 30 known species of the worldwide genus Clivina. Of the latter, zehi, wallacei, and basalis occur also in the Oriental Region and subfusa apparently has Ori- ental relationships; basalis, selhta, and jerru'^inea occur also in Australia, and zchi and inopaca reach Cape York; toxopei and vii!,il apparently have Australian rela- tionships; and a small radiation in the "australasiae group" has occurred or is oc- curring in New Guinea, the group ap- parently having Australian relationships. The relationships of some of these species are doubtful, but zebi clearly represents an Oriental stock that extends to New Guinea and Cape York.
In summary: of large Scaritini, New Guinea possesses no Oriental and only one Australian species; but of small Scaritini four species that occur also in the Orienta area (as here defined) and two additional Oriental relationships, and three specie' that occur also in Australia, two more thai reach only Cape York, and three additiona Australian relationships; and some radi- ation has occurred in one of the Australian related stocks.
[.36] The tribe BEMBIDIINI (Pari I, p 398) contains small, chiefly ground-living often water-loving (hydrophilc) carabids There are two principal genera widel)
The carabid beetles of New Guinea • Darlington
185
distributed over the \\'orld including New Guinea, and a few smaller genera of which only one is represented in New Guinea.
Bcmbidion has very many species in the north-temperate zone including temperate Asia, fewer in south-temperate areas in- cluding five in Australia (Darlington, 1962a), and still fewer in the tropics. Several temperate Asiatic stocks of the genus extend into the western part of the Malay Archipelago, reaching Celebes and/ or the Philippines (Darlington, 1959a); one species (sobriniim Boheman) may range more or less continuously from Asia to Australia along the Lesser Sunda Islands (Darlington, 1959a: 339-340); and one southern Australian species extends into the tropical northern part of Australia; but none of these reaches New Guinea. However, the island does have one Bem- bidion of a different stock, a member of the coastal (including intertidal) subgenus (or separate genus) Cillenus. This stock extends from Asia to Australia and New Zealand; the New Guinean species is re- lated to Oriental rather than Australian fomis. The absence of more-ordinary Bcmbidion in New Guinea is especially noteworthy.
The other principal genus of Bembidiini, Tachys (Part I, p. 400; Part IV, Tax. suppL), is virtually cosmopolitan, but is best represented in the warmer regions of the world. Many species occur in tropical Asia and many in Australia, and some species range from Asia to Australia, but the Oriental and Australian Tachys faunas are remarkably different as wholes. New Guinea possesses 68 species, of which the geographic relationships (so far as they can be determined) are summarized in Part I, in notes under the genus and in discussions under the ten species groups. The New Guinean species include 17 that occur also in some part of the Oriental area (which for present purposes includes Celebes and the Philippines) plus three additional stocks related to Oriental forms, and eight species that occur also in Aus-
tralia. (Several species that range from Asia to Australia are included in both these totals. )
The third bembidiine genus in New Guinea, Limnasiis, is widely distributed in the Old World and reaches the West Indies and Hawaii. It is represented in New Guinea by two winged, eyed species, \\'hich range respectively from southern Asia to Australia and from southern Asia to New Guinea, and by one endemic, flightless, blind species known only from the Bismarck Range. (No aniline Bem- bidiini have yet been found in New Guinea, but some probably occur there. They are minute (often c. 1 mm long), blind carabs which usually occur in soil and which are usually obtained only by special collecting methods.)
In summarv. New Guinean Bembidiini include 19 species ^^'hich occur also in the Oriental area and four additional stocks with Oriental relationships, and nine spe- cies which occur also in Australia.
[37] The TRECHINI (Part I, p. 487) are small, hydrophile and geophile cara- bids. In this tribe. New Guinea is notable principally for what it lacks. Tlie tribe as a whole is worldwide. Terrestrial, meso- phile members of the tribe are numerous in the north-temperate zone and numerous also south of the tropics in southeastern Australia ( and in southern South America ) ; many of those in both Asia and Australia are now flightless, but winged species occur too in both places; the ancestral fomi(s) presumably dispersed by flight. A few species of Trechus have been found at high altitudes in northern Luzon in the Philippines (Darlington, 1959a), but other- wise mesophile trechines are entirely unknown in the Malay Archipelago, includ- ing New Guinea, at any altitude. Their absence in New Guinea is remarkable. However, hydrophile, stream-side trechines are sparingly represented in New Guinea. Of the widely distributed Old-World ( and West Indian) genus PcriJeptus, one Ori- ental species reaches western New Guinea;
186 Bulletin Museum of Comparative Zoology, Vol 142, No. 2
and Perileptodea is confined to New Guinea, except that one of the New Guinean species occurs also in the Solo- mons. It is noteworthy that PeriJeptiis (above), although it occurs from Asia to western New Guinea and again in Aus- tralia, is apparently absent in the main part of New Guinea, where its place may be taken bv Perileptodcs.
[38] The tribe PANAGAEINI (Part I, p. 492; Part IV, Tax. suppl.) includes chiefly medium-sized, terrestrial, mesophile forms, widely scattered over the world. Geographically significant Asiatic genera not represented on New Guinea include winged Fana<^acus, which occurs around the north-temperate zone (and in Central America and the West Indies) including temperate Asia and Japan but not in tropical Asia; flightless Bracht/onychus, confined to the tropical southeastern corner of Asia; and Avinged Euschizomerus, an African-Oriental genus that reaches Java, Borneo, and the Philippines. Five genera have been found in New Guinea, all the New Guinean species being winged except the Craspedophorus. Of Trichisio, the one New Guinean species is supposedly en- demic but closely related to both Oriental and Australian species. Of the tropical Oriental (and African) genus Dischi^ssiis, one widely distributed Oriental species reaches New Guinea. Of the African- Oriental Microcosmodes (Microschemtis), one species (doubtfully assigned to the genus) occurs in New Guinea and Austra- lia. The Oriental genus Peronomcrus is represented in New Guinea by one Oriental and one endemic species, the latter ap- parently representing a second Oriental stock. And Craspedophorus, although well represented in both the Oriental Region (and Africa) and Australia, is known in (eastern) New Guinea by only one en- demic species (a single individual) prob- ably related to Australian rather tlian Oriental forms. (Although most species of Craspedophorus including probably all the Australian ones are now flightless, at least
one of the Oriental (Philippine) species i; still winged.) In summary. New Guineai panagaeines include two Oriental specie; and two additional species with Orienta relationships, and one Australian specie: and two additional species with Australiai relationships.
[39] The tribe PTEROSTICHINI (Par I, p. 497; Part IV, Tax. siippl.) includes i large number of carabids, diverse in size some winged and some not, many of then mesophiles, some hydrophiles, but ven few arboreal. Tlie tribe is worldwide bu is both taxonomically and geographically diverse, with different genera distributee in many different patterns. The Australia! genera of the tribe have been reclassifiec by Moore (1965), who previously (1963 transferred some supposed pterostichine: to the tribe (or subfamily) Psydrini. O the latter (primitive?) group, 18 genen are now known in Australia (Moore, 1963) Nomitis occurs in North America, Europe and locally in Africa; and Psydrtis is ii North America. The only psydrine genu; that reaches New Guinea is Mecyclothorax which (for statistical purposes) I hav( continued to count as a pterostichine.
In the north temperate zone, mesophih pterostichines are dominant carabids; tem perate Asiatic genera that do not ente: the tropics are too numerous to list. Oi the other hand, only a few, small generj are confined to the tropical Asiatic main land. A few additional Oriental genen reach Wallace's Line but not Celebes Moriouidiiis (two species) is known onh from Tonkin ( North Vietnam ) and Borneo Diceromenis (one species), from India t( to Java and Timor; Dicaelindus (few spe cies), from Ceylon and Burma to Java anc Borneo; and Metabacetiis (few species) from Burma to Java, Borneo, and th( Philippines. (The pterostichine genera bes represented in tropical Asia and the west ern Malay Archipelago are actually Lesti ciis, Trigonotoma, and CaeJostomus (meso philes), Ahacetus (mostly hydrophiles) and Morion (subarboreal), all referred t(
The carabid beetles of New Guinea • Darlington 187
gain below.) The mesophile genus Trigo- lotoma, which includes some winged spe- ies and which is primarily Oriental, occurs ast to Celebes and the Moluccas but not 0 New Guinea.
In Australia, mesophile pterostichines are gain dominant, but the genera are almost 11 different from those of Asia. Australian ;enera that do not reach New Guinea are
00 numerous to list.
Genera of Pterostichini that do occur in ■Jew Guinea may, for zoogeographic pur- loses, be grouped according to habitats, ^ost are winged; some -w species occur
1 primarily winged genera, and some small ndemic genera are now wholly -w, but 11 are or may be derived from ancestors hat were winged when they reached New Guinea.
Ten genera of New Guinean Pterosti- hini are (or probably are) mesophiles. Tiey include two endemic, mountain-living ;enera, of which Rhytiferonia (with two pecies) may have Australian relationships, nd Analoma (Paraloma) (with four spe- ies). Oriental relationships. A third en- lemic genus, HapJofcroniu, with one pecies, may be related to (derived from) Australian ) Loxandrus ( for which see inder hydrophiles, below). Monotypic h-achijdius ranges from southeastern Asia 0 New Guinea etc. but not Australia- In he African and Oriental genus Caelosto- mis, picipes ranges from Asia to Australia perhaps carried by man); the other four 'Jew Guinean species are endemic but epresent at least one additional Oriental geographic unit; and one of the endemic pecies reaches Cape York. Meci/clotliomx Part I, pp. 498, 505; present part [SO], ^ax. siippl. ) , is mainly Australian ( and iawaiian) but two species occur at very ligh altitudes in New Guinea, and related pecies, on mountains in Java; I score this IS one Australian and one Oriental geo- graphic relationship. Catadromus is a nainly Australian genus of very large :;arabids of which two individuals of one Australian species have been taken in New
Guinea; this species occurs also on Java; I think it may be carried by man; however, in terms of present distribution, it repre- sents both one Australian and one Oriental geographic unit. Prosopogmus is mainly Australian; three endemic species on New Guinea comprise one Australian relation- ship; and one species is on the Moluccas. Lesticus is a primarily Oriental genus with eight New Guinean species; one of them occurs in Australia (and is the only Lesti- cus there); the other seven are endemic but their relationships are Oriental, and I count them as one Oriental unit of relation- ship. And the Oriental genus Cosmodiscus includes two New Guinean species which have independent relationships with Ori- ental species, and one of the New Guinean species occurs also in Australia. In sum- mary: the mesophile Pterostichini of New Guinea apparently include three species shared with the Orient plus six additional Oriental relationships; four species shared with Australia plus four additional Austra- lian relationships; and one New Guinean species that reaches Cape York.
Six genera of Pterostichini on New Guinea are primarily hydrophiles, or prob- ably so. Two of them, endemic Homa- loncsiota and Nchriojeronia (with together three species), probably live beside run- ning water; they are related to each other and to Loxandrus (below); I count them as one Australian unit of relationship. Of genera associated with standing water, the African-Oriental-Australian genus Aha- cetus is represented on New Guinea by two principal Oriental species, one of which reaches Australia; the primarily Australian Platycoelus, by four endemic species, probably representing at least two Australian relationships (one species reaches the Moluccas); Australian-Amer- ican Loxandrus (Part I, pp. 498, 549; present part [82] ), by four endemic species probabK' representing at least two Austra- lian and one Oriental ( Celebes ) geographic relationships; and Tiferonia, by one en- demic species, the only other known species
188 Bulletin Museum of Comparafive Zoology, Vol. 142, No. 2
of the genus being Oriental (Philippine), of which one Oriental species reaches New
In summary: h\dr()philc Pterostichini in Guinea. Other agonine genera that are
New Guinea include two Oriental species well represented in temperate Asia but are
and two additional Oriental relationships; absent or few in the tropics include Cala-
one species shared with Australia and thus, Sphodnis, etc., and Syniichus (Lind-
probabK' at least five additional Australian roth, 1956).
relationships; and the Australian-related Most of the agonines of tropical Asia
I.oxandrus and its apparent derivatives, and the western part of the Malay Archi-
llomalouesiola and Nehrioferunia and per- pelago are now placed in the catch-all
haps T if crania, have apparently radiated genus CoJpodes (Louwerens, 1953); their
on New Guinea, while the Oriental-related classification is difficult and their geo-
hydrophile pterostichines have scarcely graphic relationships are largely unknown;
done so. the few distinct, small. Oriental agonine
The single subarboreal (bark-and-log- genera include Onycholahis and Dirotiis, inhabiting) genus of Pterostichini in New represented from the Malay Peninsula to Guinea is pantropical Morion. The tvvo Java and Borneo, with the foiTner genus New Guinean species belong to a primarily reaching also the Philippines. Oriental rather than Australian group of The genus Homothes (including Aeolo- the genus; one of the species is endemic dcrmus) is discontinuous, with one species and the other extends to Australia. In distributed from the Malay Peninsula to terms of present distribution, the New Celebes and the Philippines, none in New Guinean members of the genus score one Guinea, but several in Australia. Other- Oriental relationship and one species wise, the extremely limited Australian shared with Australia. agonine fauna includes only the endemic.
The sum of existing geographic relation- monotypic, flightless Odonta'j^onum (Dar-
ships of all New Guinean Pterostichini lington, 1956) in North Queensland, and
(three preceding paragraphs) is five spe- a small number of species all of which I
cies shared with the Orient and nine ad- tentatively assign to Notagonutn or Col- >
ditional Oriental relationships, six species podcs. i
shared with Australia and nine additional Known New Guinean Agonini (Part II;
Australian relationships, and one additional present part. Tax. stippl.) now total 21
species extending to Cape York. genera and 160 species. They are medium-
[40] The distribution of the tribe sized Carabidae, diverse ecologically. All
AGONINI is discussed in the present part those found at low altitudes in New Guinea
[92]. The tribe is worldwide, better repre- are winged; many of the mountain-living
sented in the tropics than in the temperate groups now have atrophied wings (but all
zones, and especially dominant in New may be derived from ancestors that were
Guinea (Fig. 13 [64]), but deficient in \\'inged when they reached New Guinea).
Australia. They may be grouped according to habitat,
In the north-temperate zone, agonines i^s follows,
arc moderately numerous and include the Mesophiles among the New Guinean i
dominant, primarily winged, hydrophile Agonini include the Oriental genus Arhy-
and mesophile genus Ai:,onuJ7}, with nu- tinu.s, with three endemic New Guinean
merous subgenera; the geographic limits of species, representing at least two Oriental
this genus are doubtful (because of doubt stocks, and Tarsagonum, with one endemic
about the assignment of some tropical species, the genus otherwise known only
species), but the only subgenus that ranges from Borneo; these two genera occur at
far into the Oriental tropics is probably low altitudes. Also mesophiles are certain
Sericoda ( pres(>nt part [80], Tax. suppL), species of the "genera of convenience"
The carabid beetles of New Guinea • Darlington
189
Notagonum and Colpodes (of which the geographic relationships, of the mesophile species, have not been determined in detail) and some or all the species of nine endemic genera, most but not all of which are momitain-living, and most but not all of which have atrophied wings; this whole assemblage is Oriental in general relationships, but I cannot decipher the separate geographic units.
Hydrophiles among the New Guinean Agonini include at lower altitudes the single New Guinean Lorostemma, which is closely related to an Oriental species; some species of Notaganum; and the one species (several subspecies) of the endemic genus Lithagomim; and at higher altitudes some additional Notagonuni; the endemic genus Potamagonum; and some species of the endemic genus Nebrkigomim. Excepting the Lorostemma (which has been derived independently from the Orient), these agonines may all be products of the com- plex radiation of Agonini that has occurred on New Guinea. I cannot state their geo- graphic ties in detail, except to repeat that they are all basically Oriental in relation- ships, and to add that two of the species of Notagonum occur also in Australia. These two are N. submetalliciwi (present part. Tax. siippl.) (which is distributed as if it has reached New Guinea from Australia) and N. dentellum (present part. Tax. siipp].) (which has apparently reached North Queensland from New Guinea).
Arboreal agonines in New Guinea in- clude the African-Oriental genus Euplenes, with two endemic species probably repre- senting two Oriental geographic units; the Oriental genus Dicranoncm, of which the one species that occurs in New Guinea ranges also from southern Asia to Australia; the subgenus Sericoda of the primarily northern genus Agomwi, of \\hich one Oriental species reaches New Guinea (present part, Ta.x. siippL)-, Violagomnn, a mainly insular genus (Darlington, MS), with the one New Guinean species oc- curring also in Australia; and some species
of the "genus of convenience" Colpodes, of which three New Guinean species occur also in the Oriental area and one reaches Australia.
In summary of the geographic relation- ships of New Guinean Agonini: the meso- philes include four specific Oriental units of relationship; the hydrophiles, two spe- cies shared with Australia; and the arboreal forms, five Oriental species, two additional specific Oriental relationships, and three species shared with Australia. But it should be repeated again, and emphasized, that the whole diverse assemblage of Agonini on New Guinea is Oriental in general relationships, although only a few of the separate geographic units can be dis- tinguished.
[41] The tribe PERIGONINI (Part III, p. 5) contains small, winged, mesophile carabids which live mainly in rotting logs and leaf htter. Only the worldwide genus Perigona {.semu Jato) is represented in New Guinea. It includes numerous species in the Oriental Region, 14 in New Guinea, and only five in Australia (Darhngton', 1964). Of the New Guinean species, five occur also in the Oriental area (which includes the Philippines) and three more have separate Oriental relationships, and one species occurs also in Australia.
[42] The tribe LICININI (Part III, p. 14; present part, Tax. suppJ.) includes both medium-sized and small, mostly winged Carabidae, of which some are hydrophiles and some mesophiles. The tribe is nearly worldwide in distribution. A few genera, including Licinus, occur in north-temperate areas but do not reach tropical Asia. DipJochila {Remhiis) is widel\- distributed in the north and in Africa and the Orient; several species reach Java, Borneo, and the Philippines, but none farther east. Several genera are confined to or occur mainly in Australia, which has more licinines than any other part of the world. Five genera are represented on New Guinea: the widely distributed genus Badister, by one species which occms also
190
BuUctin Museum of Comparative Zoology, Vol. 142, No. 2
in the Orient and Australia; primarily Aus- tralian PJnjsolacstJius, by one species which occurs also in the Orient; monotypic Omestes, by O. torta, which is also in the Oriental area (Celebes, Philippines); primarily Australian Dicrochile, by one Australian and three endemic species rep- resenting at least one additional Australian relationship; and primarily Australian Microf crania (Part III, p. 18; present part, [80] ) by one species on the Bismarck Range in New Guinea, and one on moun- tains in Java. The sum of geographic units seems to me to be three species shared with the Orient (chiefly the Philippines) plus one additional Oriental relationship, and two species shared with Australia plus two additional Australian relationships. How- ever, regardless of these unit scores, all the New Guinean licinines except the Badister belong to groups that are characteristically Australian rather than Oriental. (See Ball, 1959, for further discussion of the classifi- cation and distribution of some members of this tribe.)
[43] The tribe CHLAENIINI (Part III, p. 20; present part [9] ) includes medium- sized (less often small or large), primarily winged, hydrophile and mesophile Carab- idae. Tlie tribe is most diverse in Africa and Eurasia. A few small genera (each with one or very few species) are confined to temperate Eurasia or the Asiatic tropics. Very small chlaeniines of the African- Oriental genus CaUifitomimii.'i reach Java, Borneo, and the Philippines (and Timor), and of the Oriental genus Chlaeminus, Java and the Philippines. However, only the worldwide genus Chkienius (.sensii lato) crosses the central part of Wallace's Line to Celebes, New Guinea, and Austra- lia, and species decrease in number in this direction, Java having about 30, New Guinea 12, and Australia 10. Tlie New Guinean species include the following geographic units (see under the species in Part III, pp. 23ff for details): six species shared with the Oriental area plus three additional Oriental relationships, and four
species shared with Austi^alia plus two additional Australian relationships. Several of the species range from the Orient to Australia or have close relatives in both, and score as both Oriental and Australian units.
[44] The tribe OODINI (Part III, p. 30; present part, Tax. suppl.) contains mostly medium-sized carabids most of which are subaquatic, although a few have left the water and become terrestrial. Most are winged, but some terrestrial fonns have atrophied wings. Oodines occur in all the warmer parts of the world, but are rel- atively few in temperate areas. In the Oriental Region, Systolocrauius (also in Africa) and Holcocoleus are confined to the mainland, and Simous- (see under Oodes hevissimus. Part III, p. 34) ranges from India to Java and Borneo. Australia possesses numerous species of the endemic genus Coptocarpus, which reaches the Cape York peninsula but not New Guinea^ Only two genera are represented on New Guinea: Amitrichis, which is Oriental- Australian (and American, see [82]), and Codes, which is nearly worldwide. New Guinean species include the following geographic units (see under the species in Part III, pp. 32ff): four species that occur also in the Oriental area (as here defined ) and two species that occur also in Australia. The geographic relationships of the other species are undetermined. i
[45] The tribe HARPALINI (Part III, p. 38; present part. Tax. suppl.) contains a large proportion of the common, medium- sized Carabidae that live on the ground especially in open country in all parts of the world, and the tribe includes also some common smaller forms that live especially in wet places. Most are winged. The dis- tribution of the tribe in the Asiatic- Austra-! lian area including New Guinea is best^ summarized by subtribes (see Part III, p
^ A Coptocarpus has now been found in Nev' Guinea (see Taxonotnic supplement, footnot< under Oodini ) .
The carabid beetles of New Guinea • Darlington 191
39), although this classification is an over-simplification. One additional harpa- line not belonging to any of the following subtribes should be mentioned: one species of the Old- World-tropical genus Pachytra- cltehis (subtribc Pachytrachelina; Csiki, 1932: 1082) ranges from India etc. to Australia perhaps along the Lesser Sunda Islands, although the genus does not occur in New Guinea.
Harpalini of subtribe Anisodactylina are chiefly medium-sized mesophiles. Tlie prin- cipal genus in the north-temperate zone is Anisodactylus (sensii lato), which is Hol- arctic; it enters the edge of the tropics in southeastern Asia but does not reach the tropical islands; however, two related, monotypic, primarily insular genera, Rluj- sopiis and Harpalomimetes, do reach Java and/or Borneo (Wallace's Line). Also in this subtribe is Chijdaeus (Part III, p. 47; present part. Tax. suppL), which occurs at high altitudes on the Himalayas and on mountains in Formosa, Sumatra, Java, the Philippines, and New Guinea, but does not reach Australia. It has apparently "mountain hopped" some 4,000 miles ( more than 5,000 km), by steps, from the moun- tains of Asia across the Malay Archipelago (see [80]). Three other genera of Ani- sodactylina are best represented in Austra- lia but reach New Guinea and extend westward into and across the Malay Archi- pelago, chiefly in relatively open country including open eucalyptus woodland, iil- though some species occur in rain forest too. Of these primarily Australian genera, Gnathaphanus has five species in New Guinea all of which occur also in Australia and two of which reach only the southern edge of New Guinea, the others reaching (respectively) West New Guinea, the Malay Peninsula, and India; DiapJwromerus, two endemic species in New Guinea closely related to (different) Australian species, one of the New Guinean species reaching also the Moluccas; and IlypJiorpox, one New Guinean species which apparently occurs also in northeastern Australia and
extends west to Java and Sumatra. The geographic units of New Guinean Ani- sodactylina are therefore three species that occur also in the Orient plus one additional Oriental relationship, and six species that occur also in Australia plus two additional Australian relationships. (However, ex- cepting Chydaeus, the Anisodactylina that reach New Guinea are all primarily Aus- tralian, and they show an exceptionally clear pattern of spread westward across the islands.)
Harpalini of subtribe Pelmatellina are represented in New Guinea by only three small, endemic, water-loving species of the diverse Australian genus Lecanomerus (Part III, p. 45); they may all be derived from one ancestor and I count them as one Australian geographic relationship.
Harpalini of subtribe Harpalina are (in the area under discussion) primarily Asi- atic-Oriental and do not occur in Australia at all, except that one or two genera reach just the northern edge of the continent. Harpahis itself is a dominant Holarctic genus which occurs south to Java ( one spe- cies) but does not cross Wallace's Line. Oxycentrus reaches at least Celebes; Dio- ryche, at least Celebes and Timor; but these few details do not do justice to the numbers and complexity of limits of the Oriental Harpalina. Members of this sub- tribe are dominant on the ground in New Guinea, especially in rain forest, but the New Guinean Harpalina represent merely the fringe of the much richer Asiatic- Oriental fauna. The principal genus of the subtribe that does reach New Guinea is Trichotichmis (Part III, p. 48), with 14 species on the island showing at least three separate Oriental relationships {^^Carhanus," "Lampetes" and the others). Harpaloxenus, with five species on New Guinea, is ap- parently closely related to Trichotichntis and counts as at least one additional Ori- ental relationship. Lyter is a new, endemic, monotypic genus confined to New Guinea (and perhaps derived from Trichotichnus on the island). Coleolissus has two en-
192 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
demic species on New Guinea representing perhaps only one Oriental stock; one of the species of Colcolissii.'i is represented on Cape York. Finally, Platymctopus is represented on New Guinea by one Ori- ental (Philippine) species that apparently reaches only western New Guinea. In summary: the geographic units of New Guinean Harpalina are one species shared with the Orient (Philippines) and at least eight additional Oriental relationships, and one New Guinean species reaching Cape York. (This is a remarkably clear pattern of multiple dispersal of a dominant Asiatic- Oriental group south and east across the Malay Archipelago.)
Harpalini of the subtribe Acupalpina are relatively small forms most of which live in wet places on the ground. Their generic classification is not very satisfac- tory; all the genera represented in New Guinea are widely distributed at least in the Old World, but their zoogeographic limits (except as they concern New Guinea) need not be discussed here. Of Egadwma, New Guinea possesses four species, of which three apparently occur also in both the Orient and Australia and the fourth is closely related to an Oriental species (present part. Tax. suppL). Anoplo- genius is represented in New Guinea by one species that probably ranges from the Orient (Sumatra) to Australia; Stenolo- phu.s', by t\vo species both probably present in the Oriental area and in Australia. Finally, the six New Guinean Acupalpus include apparently two species that occur also in the Orient plus two additional Oriental relationships, and one species that occurs also in Australia plus two additional Australian relationships (see details given in Notes- under the species, Part III, pp. 73ff). The sum of geographic units of New Guinean Acupalpina is therefore eight species plus three additional relationships Oriental, and seven species plus two re- lationships Australian. (But the pattern as a whole is clearly one of multiple dis-
persal of Acupalpina from the Orient to New Guinea and Australia.)
The sum of geographic units of New Guinean Harpalini of all subtribes is 12 species shared with the Orient plus 12 additional Oriental relationships, 13 species shared with Australia plus five additional Australian relationships, and one additional species reaching Cape York.
[46] The small, pantropical tribe ANAU- LACINI (Part III, p. 76) contains small, winged, mesophile carabs which usually live in leaf litter on the ground in rain forest. New Guinea possesses four genera: Anauhicus is represented by one species, which is widely distributed also in both the Orient and Australia; Caphora, by one Oriental species that has been foimd also on the tip of Cape York; Acphnidius, by one Oriental species that does not reach Australia; and Odontomasorcus, which is endemic and of unknown relationships, by one species and one additional subspecies. The geographic units are therefore three species that occur also in the Orient, one also in Australia, and one additional occur- rence on Cape York. The few additional members of the tribe that occur in the Orient are geographically insignificant; no additional ones occur in Australia.
[47] The CYCLOSOMINI (Part III, p. 78) is another small tribe of rather small, winged Carabidae widely distributed in! the warmer parts of the world. Of the fewi genera represented in tropical Asia, Tetra- gonodenis (which is widely distributed in' the Old World and the Americas) and Cijclosomus (which is African and Ori- ental) reach Java and the Philippines but not Celebes. These genera are terrestrial In Australia the tribe is represented only by numerous species of Sawthrocrepis Most of them live on shaggy tree trunks but one group of small species is foliage' arboreal, and this group is represented b> endemic species on New Guinea, Java, the Philippines, and some Lesser Sundi Islands. The New Guinean species is ap
I
The carabid beetles of New Guinea • Darlington
193
parently closely related to both Australian and Oriental (Javan, etc.) forms.
[48] The tribe LEBIINI (Part III, p. 80; present part, Tax. suppl. ) is, among Carab- idae, equalled in number of species on New Guinea only by the Agonini. Ho\\'- ever, the Lebiini are far more numerous and diverse at low altitudes on the island and far more complex in their discernible geographic relationships. Most tropical lebiines are arboreal, and all the 160 New Guinean species except Nototarus papua are winged. In discussing the New Guinean forms in detail (below) I shall group them by major habitats. First, how- ever, I shall state briefly the distribution of genera that occur in neighboring regions but do not reach New Guinea. The Ori- ental forms are especially numerous and significant, but I cannot group them by habitats, because I do not know the habi- tats of many of them.
Tlie north-temperate zone possesses com- paratively few Lebiini. Of the genera that do occur in temperate Asia, the most im- portant is Holarctic, terrestrial Cijmindls, a genus well represented in temperate Asia but which scarcely enters the Asiatic tropics.
Tropical Asia possesses relatively more Lebiini, but only half a dozen or so small genera of the tribe are actually confined to the tropical Asiatic mainland.
A much larger number of genera occur on the Malay Archipelago. Some of them are represented on the mainland of tropical Asia too or even in Japan, and some reach New Guinea (as detailed below), but about 15 do not reach New Guinea. Of these 15 genera that are represented in the western part of the Malay Archipelago but that do not reach New^ Guinea, 12 reach Java and/or Borneo (and sometimes also the Philippines and/or some of the Lesser Sunda Islands) but (so far as known) do not cross the central part of Wallace's Line to Celebes. These are mostly small genera of one or very few species; they include AUocota (three spe-
cies), distributed from Burma to Java, Borneo, and the Philippines; Lehidia (four species), from eastern Asia and Japan to Java; and Dromius, with many species in the northern hemisphere, fewer in tropical Asia, and one species (probably unde- scribed) on Sumatra and two on Java.
Two other, important genera of Lebiini do cross Wallace's Line but do not reach New Guinea: Lioptera, with about eight species, is distributed from southeastern Asia and Japan to Borneo, Celebes, and the Philippines; and Callida, well repre- sented in all the warmer parts of the world except the Australian region but including tropical Asia, reaches Java, Borneo, Cele- bes, and the Philippines, and also some of the Lesser Sundas.
Australia has comparatively few Lebiini, and many of those that do occur in Aus- tralia are the same as or related to New Guinean species and are noted in my enumeration (below) of the geographic relationships of New Guinean genera. Additional significant lebiines in Austra- lia include FhiJophJoeus, with many species occmiing through most or all of the continent, chiefly on the shaggy trunks of eucalypts and other trees. Agonochila and Demetrido are numerous in Australia (and have apparently radiated separately there) chiefly on tree trrmks, but are represented on New Guinea too, of course. And FhJoeocarahus and Trigonothops (these t\\'o genera I think mainly on tree trunks) and Nototoriis and Anomotoms (I think chiefly on the ground ) are well represented in Australia too, although present also in New Guinea. Other genera of lebiines in Australia seem geographically unimportant.
New^ Guinea has representatives of 32 genera of Lebiini, and the geographic re- lationships even of some single genera are complex. In discussing their relationships, I shall divide them into ecologic groups comparable to those into which I have divided some preceding tribes, although the basis of division is not quite the same. Among New Guinean Lebiini few genera
194 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
are terrestrial, none is hydrophilous, but many are arboreal; and the arboreal forms are divided, those that live on tree trunks and those that live in foliage being grouped separately. And several genera which do not fit into this ecologic classification or of which the habitats are unknown are noted last of all.
Terrestrial (mesophile) or probably ter- restrial Lebiini in New Guinea include one species of the Asiatic-Australian genus Lachnoderma, the New Guinean species occurring also in Australia; one Oriental species of the Oriental genus Sintirus; one endemic species of Peliocypas, this being the easternmost representative of a diverse Oriental group; one Oriental-Australian species of Syntomiis; two species of Micro- lestes, one with Oriental and one with apparent Australian relationships; two en- demic species of Apri.stus, the easternmost representatives of this African-Oriental (and North American) genus, representing perhaps a single Oriental geographic unit; one endemic species of the Australian genus Nototaru.s; eight endemic species of the Asiatic-Australian genus Anomotarus, the geographic relationships of one of the species being Oriental and the others undetermined (except that one extends to the Moluccas); and one Oriental species of the Oriental genus Omohrus. In sum- mary, the terrestrial or probably terrestrial Lebiini of New Guinea include three spe- cies that occur also in the Orient plus four additional Oriental relationships, and two species that occur also in Australia plus two additional Australian relationships.
New Guinean Lebiini that live mainly on tree trunks and on fallen logs include one endemic species of the Oriental genus Stenotelus; one Oriental species of Miscelus that has been found also in mid-Cape York, and two endemic species of the same genus representing at least one additional Oriental relationship; nine species of the mainly New Guinean genus Miniithodes, one of the new Guinean species having a relative in the Orient (Celebes) and one
a relative in Australia; 14 species of the widely distributed tropical genus Catas- copus related as detailed below; one en- demic species of Pericahts, the easternmost (except for another endemic species on New Britain) of a species-rich Oriental genus; eight species of the widely dis- tributed genus Coptodera, detailed below; one endemic species of the Oriental genus Mochtherus; and two species of the pri- marily New Guinean genus Stricklandia, one of the New Guinean species reaching the Moluccas, with a closely related species in Australia. In more detail, the New Guinean Catascopus (Part III, pp. lOlff, see especially Notes under the genus) in- clude elegans, which ranges from tropical Asia to northern Australia; smaragdulus, from tropical Asia to New Guinea and mid-peninsular Cape York; facialis, from tropical Asia to Western New Guinea;, laevig^atus, which occurs also in the Moluc- cas and has a close relative in Australia; and aniensis, which occurs also in mid- peninsular Cape York. The geographic units of the genus listed in the preceding sentence are three species shared with the Orient; and one species shared with Aus- tralia, one additional Australian relation- ship, and two additional extensions to Cape York. (Nevertheless, Catascopus is most diverse in the Orient; several additional stocks have probably reached New Guinea in the past, although I cannot now dis- tinguish them, and all the few Australian species seem to represent more or less recent invasions from New Guinea.) Coptodera (Part III, p. llOff) includes endemic i!,rossa, with no recognized close relatives; cijaneUu, lincolata, eluta, and oxyptera, all of which occur also in some part of the Oriental area; ornatipennis, which occurs also in the Moluccas and has a relative in Celebes (Oriental area); and pajmella and tcau, \\'hich together form' one Australian-related unit: total units in this genus: four species that occur also in the Oriental area plus one additional Ori- ental relationship, and one Australian re-
The carabid beetles of New Guinea • Darlington
195
lationship. Geographic units for all the tree-trunk-living New Guinean Lebiini are eight also-Oriental species plus six Oriental relationships, and one also-Australian spe- cies plus four Australian relationships, and three additional species reaching Cape York.
Arboreal Lebiini that probably live mainly in foliage on New Guinea include the following: three endemic species of the Oriental genus Aristolehia perhaps representing one Oriental stock, one of the species occurring also on the tip of Cape York; seven species of the worldwide genus Lebia, their geographic relationships detailed below; one endemic species (the easternmost of the genus) of the Oriental genus Physodera; one endemic species of the mainly Oriental genus Holcodenis, the New Guinean species having close relatives in both the Orient and Australia; mono- typic Oxydontus, the one species occurring also in the Orient; 13 species of the mainly Oriental DoUchoctis, related as described below; monotypic Celaenephes, the one species occurring also in the Orient and Australia; three species of the mainly African-Oriental Parena, one of the species occurring also in the Orient, one also in Australia, and one in the Orient and Aus- tralia; and 59 known endemic species of the otherwise mainly Australian Demetrida (all perhaps derived from one Australian- related stock) with one of the New Guinean species found also on the tip, of Cape York. In more detail, the New Guinean Lehia include karenia, which is also Oriental; melanonota (present part, Tax. siippl. ) which is Oriental and Aus- tralian; papucUa, with a close relative in Australia; and additional species of which the relationships are undeteiTnined or are with other New Guinean species. And of DoUchoctis, striata ranges from southern Asia to New Guinea and Australia; rnicro- dera, from Sumatra to New Guinea; and the remaining 11 New Guinean species of the genus, all members of or perhaps de- rived from the aculeata group, include one
species (actdeata) that occurs also in the Oriental area (Celebes) and reaches mid- peninsular Cape York, and at least one additional Oriental relationship. The sum of geographic units of the foliage-living New Guinean Lebiini is nine species that occur also in the Orient plus four Oriental relationships, and five species that occur also in Australia plus three Australian re- lationships and three additional occur- rences on Cape York.
Besides the Lebiini listed in preceding paragraphs, four lebiine genera of which I do not know the habitats occur in New Guinea. These genera are probably arboreal, but I do not know whether (in New Guinea) they live on tree trunks or in foliage. Agonochila is a primarily Aus- tralian genus with seven small endemic species in New Guinea representing per- haps only one Australian-related stock; most members of this genus in Australia live on tree trunks, but a few small tropical Australian species inhabit foliage, and the New Guinean ones may do so too. Minii- phloeus is a monotypic genus confined to New Guinea; its relationships and habitat are unknown. Phloeocarahus is primarily Australian but is represented on New Guinea by one species that occurs also in Australia and one that is endemic and represents a separate Australian-related unit. And Trigonothops is primarily Aus- tralian but includes one New Guinean species closely related to an Australian species. The geographic units of these four genera total one species shared \\'ith Aus- tralia plus tliree Australian relationships.
Finally, as far as Lebiini are concerned, two genera are represented in New Guinea by single species that are carried by man and that cannot be placed in the preceding habitat classification. They are Anchista and Endynomena, each represented in New Guinea by one species that has been found also in the Orient but not in Australia. (Plochioniis pallens, if it turns up in New Guinea, will be a third man-distributed lebiine in New Guinea.)
196 Bulletin Museum of Comparative Zoology, Vol 142, No. 2
Table 7. Summary of geographic relationships of New Guineax Lebiini
Habitat
Species also Or.
Additional Or. units
Species also Au.
Additional An. units
Additional to C. York
|
Ground |
3 |
4 |
2 |
2 |
— |
|
Trunks |
8 |
6 |
1 |
4 |
3 |
|
Foliage |
9 |
4 |
5 |
3 |
3 |
|
Prob. arb. |
— |
— |
1 |
3 |
- |
|
Man-carried |
2 |
- |
— |
- |
- |
|
Total |
22 |
14 |
9 |
12 |
6 |
Table 7 sums up the geographic relation- ships of New Guinean Lebiini.
The distribution and relationships of Lebiini from Asia to Austraha may be summarized as follows. The tribe is rela- tively poorly represented (and chiefly terrestrial) in temperate Asia. It is better represented (and more arboreal) in trop- ical Asia, with some small genera confined to the tropical Asiatic mainland and others, including larger genera, extending onto the Malay Archipelago for various distances. The lebiines of the Western (Oriental) part of the archipelago are numerous and include about a dozen (mostly small) genera that reach Java and/or Borneo but do not cross Wallace's Line to Celebes, and also two important genera that do extend to Celebes but not New Guinea. The Lebiini of New Guinea are numerous too, and include 22 species that occur also in the Oriental area plus at least 14 Ori- ental relationships, nine species that occur also in Australia plus at least 12 Australian relationships and six additional species that reach Cape York. Some New Guinean genera are represented also in Australia, but additional Australian genera are few, the important ones being mostly tree-trunk- living or terrestrial.
This whole main pattern of distribution from Asia to Australia reflects the fact that Lebiini are primarily arboreal Caiabidae, far more numerous in the complex arboreal habitats of tropical rain forest than they are in thinner and less complex temperate woodlands.
The localization of a number of small
genera toward the Oriental end of the Malay Archipelago and the extensive radi- ation in several genera in New Guinea ( see below) suggests that, although some species have been able to disperse from southern Asia to Australia, some other lebiines have dispersed less effectively than might have been expected of such active, winged Carabidae. Their dispersal may have been limited either by limitation of their power of dispersal or by discontinuities in the distribution of the rain forests in which most of them live. It is noteworthy that the lebiines that have radiated in New Guinea have diverse geographic relation- ships: the foliage-living Demetrido (59 species on New Guinea derived from one or a few ancestors! ) is Australian-related; the foliage-living DoJichoctis and tree- trunk Catascopiis ( in both of which moder- ate radiations have occurred in New Guinea) are primarily Oriental in relation- ships; and the foliage-living Dicraspeda *" (six New Guinean species) and tree-trunk Mimithodes (nine New Guinean species) are primarily New Guinean. The only ground-living lebiine genus in which radi- ation seems to have occurred on New Guinea is Anomotarus (eight New Guinean species, seven of them perhaps products of local radiation); the primary geographic relationship of the New Guinean members of this genus is undetermined.
[49] The small tribe PENTAGONICINI" (Part III, p. 191) includes only four genera, of which one is confined to New Zealand, the other three being represented in New Guinea. Pentagonica occurs in all
The carabid beetles of New Guinea • Darlington 197
the warmer regions of the world and in- ckides six species on New Guinea: two of them apparently range from southern Asia to Australia; two more, from southern Asia to New Guinea and Cape York; and two are endemic. The one known species of Paroscopodes occurs in eastern New Guinea and northern Australia. And Scopodes is divisible into two stocks: one stock is mainly Australian but includes one species at very high altitudes on the Snow Mountains of New Guinea and one on high mountains in Java; and the other stock is confined to New Guinea and includes seven species which are perhaps all prod- ucts of one (Australian-related) radiation. Pentagonicines are small Carabidae. The New Guinean species of Penfa^onica are arboreal, in foliage; of Parascopodes, ground- or grass-living; and of Scopodes, ground- or log-living,
[50] The small, African-Oriental tribe HEXAGONIINI (Part III, p. 202; present part, Tax. siippl.) includes two genera in the Orient: Dinopelma (about 11 species) is apparently confined to the western part of the Malay Archipelago, with several species on Java, Borneo, and the Philip- pines but none on Celebes; Hexa^onia (which occurs also in Africa) is rather diverse in the Orient; two Oriental stocks are represented in New Guinea by single endemic species; and one of these stocks has one endemic species in Australia, too. The members of this tribe that I have collected are rather small, winged Carabi- dae which live under the leaf sheaths of tall grasses.
[51] Tlie tribe ODACANTHINI (Part III, p. 203), which is worldwide in dis- tribution, includes small and medium- sized, winged carabids of which some are terrestrial and some arboreal, some of the subarboreal forms being also hydrophilous. Few occur in temperate Eurasia. The Orient possesses many, diverse species of the worldwide genus CoUiurls ( see below ) and representatives of a few smaller genera (some named below). New Guinea has
eight genera, 19 species, rather diverse in ecology and in geographic relationships (see below). And Australia has several small endemic genera as well as representa- ti\es of several of the genera that occur on New Guinea.
New Guinean Odacanthini are as fol- lows. The genus CoJUuris (see above) in- cludes four species in New Guinea, of which one occurs also in the Orient, one is endemic but related to an Australian spe- cies (it and the preceding probably live in grass especially in wet places), and two are endemic but probably represent one Oriental stock, and one of these species occurs also on the tip of Cape York (these species live on or near the ground in wet places). Of the Oriental-Australian genus Casnoidea, only two species surely occur in New Guinea: one is endemic, and one occurs also in Australia (the members of this genus live in grass and reeds over water). The one knowoi species of Basisti- cus occurs in northeastern Australia and southern New Guinea (it lives on the ground in more or less open country). Clarencia is an Australian genus with two species on New Guinea: one occurs also in Australia, the other is endemic but Aus- tralian-related (they live on or near the gromid in wet places). Dicraspeda is pri- marily New Guinean, with six rather di- verse species; three geographic stocks can be distinguished; hrunnea ranges from Java and the Philippines to New Guinea and northeastern Australia; longiloba, dnhia, and hispinosa are endemic and perhaps represent one stock which is confined to New Guinea except that longiloho reaches New Britain and dnhia Cape York; and quadrispinosa and violacea are related to each other (and probably derived from the same stock as the preceding) and are confined to New Guinea except that one or both reach the Moluccas, New Britain, and/or the Solomons. (All species of this genus li\e in under-story foliage of rain forest.) One Oriental species of the Afri- can-Oriental genus Lachnothorax reaches
198
Bulletin Musetim of Comparative Zoology, Vol. 142, No. 2
New Guinea but not Australia (I think it lives beside running water). One New Guinean speeies is tentatively assigned to the Australian genus Eiidalia but may also have Oriental relationships (this is prob- ably water-loving too). And Dohodiim is a monotypic genus confined to New Guinea but perhaps derived from the Australian Eudalia (it lives among stones beside turbulent brooks). The sum of geographic units of the New Guinean Odacanthini is three species that occur also in the Oriental area plus one Oriental relationship, and four species that occur also in Australia plus foiur Australian relationships, and two additional species reaching Cape York.
[52] The DRYPTINI (Part III, p. 216) is a small tribe of medium-sized, usually winged Carabidae which usually live on the ground or in grass, I think. Of this tribe, the widely distributed, terrestrial genus Galeritiila (Reichardt, 1967) crosses Wallace's Line to Celebes; the Old-World (and Brazilian) genus Dnjpta is repre- sented in New Guinea by two endemic species probably related to different Ori- ental forms, by one Australian species which reaches only the southern edge of New Guinea, and by one additional en- demic species; and the Old-World genus Desera includes one New Guinean species which is endemic (except that it reaches also New Britain and New Ireland) but closely related to species in both the Orient and Australia. Besides these, the tribe contains only a few small genera none of which occurs in the area under consider- ation except that Pseiidaptinus (Thalpius), a genus of small terrestrial dryptines widely distributed in the Americas, has also one species in Australia.
[53] Of the small tribe ZUPHIINI ( Part III, p. 218; present part. Tax. siippl.) which includes small, mesophile and hvdro- phile carabids, New Guinea possesses one endemic species of the worldwide genus Zuphium and one Australian species of the same genus that reaches only southern New Guinea; four species of the African-
Oriental genus of Planetes, one being also Oriental and one also Australian; and two endemic species of Colmidia representing one Oriental stock. The few other (small) genera of the tribe include Oriental A<:,astiis, which reaches Java, Borneo, and the Philippines, and Acrogemi.s, confined to Australia.
[54] Of the small, Oriental-Australian tribe HELLUODINI (Part III, p. 222), which includes only one principal genus of medium-sized, winged, mesophile carabids. New Guinea possesses nine species of Fo'^ono<jJossus. The species of this genus are moderately di\'erse in the Oriental Region as well as New Guinea, less diverse in Australia. All the New Guinean spe- cies are endemic, and I cannot determine their individual relationships outside New Guinea. The few species that I have col- lected were found in leaf litter on the ground in rain forest.
[55] Tlie tribe HELLUONINI (Part' III, p. 228; present part. Tax. siippl.) in- cludes medium-sized (rarely large), usually winged, ground- or tree-tmnk-living Carab- idae widely distributed over the world. The genus Omphra is confined to India and Ceylon; monotypic Colfax, to the tropi- cal mainland of Asia; the African-Oriental genus Macrocheilus reaches Celebes; and one species of the Oriental genus Creogris extends to New Guinea and Australia. A dozen additional Australian genera fonn a special Australian group of the tribe; of these genera, three are represented in New Guinea by single Australian species (two; of these reach only southern New Guinea ) , and Ilelhionidius includes four New Guin- ean species which are diverse but may represent one Australian-related stock. Helluopapua (two species) is endemic but may represent the same stock. I therefore score New Guinean Helluonini as includ- ing one Oriental species, four species that occur also in Australia, plus one Australian relationship.
[56] The BRACHININI (Part III, p 2^34; Erwin, 1970) are medium-sized and
The carabid beetles of New Guinea • Darlington
199
Table 8. Summary of geographic relationships of New Guinean Carabidae proper by tribes
Tribe
Ecology
Or. spp. (+ Or. rel.)
Au. spp. (+ Au. rel.)
Cape York
|
Ozaenini |
Meso |
|
Paussini |
Meso |
|
Scaritini |
|
|
large |
Hydro |
|
small |
Hydro+ |
|
Benibidiini |
Hydro+ |
|
Trechini |
Hydro+ |
|
Panagaeini |
Meso |
|
Pterostichini |
Meso |
|
Hydro |
|
|
Subarb |
|
|
Agonini |
Meso |
|
Hydro |
|
|
Arb |
|
|
Perigonini |
Meso (incl. logs) |
|
Licinini |
Hydro, Meso |
|
Chlaeniini |
Hydro, Meso |
|
Oodini |
Hydro |
|
Harpalini |
|
|
Anisodactylina |
Meso ( open ) |
|
Pelmatellina |
Hydro |
|
Harpalina |
Meso+ |
|
Acupalpina |
Hydro + |
|
Anaulacini |
Meso |
|
Cyclosomini |
Arb |
|
Lebiini |
|
|
Ground |
|
|
Trunks |
|
|
Foliage |
|
|
Prob. arb |
|
|
Man-carried |
|
|
Pentagonicini |
Arb, meso |
|
Hexagoniini |
Arb |
|
Odacanthini |
Arb, meso, hydro |
|
Dryptini |
Arb |
|
Zuphiini |
Meso |
|
Helliiodini |
Meso |
|
Helluonini |
Meso (+ ?) |
|
Brachinini |
Meso, hydro |
|
Pseudomorphini |
Arb |
(1)
|
4 |
(2) |
|
L9 |
(4) |
|
1 |
|
|
2 |
(2) |
|
3 |
(6) |
|
2 |
(2) |
|
(1) |
|
|
(4) |
|
|
2 |
|
|
6 |
(2) |
|
5 |
(3) |
|
3 |
(1) |
|
6 |
(3) |
|
4 |
2
4
1)
|
1 |
(8) |
|
8 |
(3) |
|
3 |
|
|
(1) |
|
|
3 |
(4) |
|
8 |
(6) |
|
9 |
(4) |
1)
2) 1) 3) 1)
2) 1)
104(69) 173
(1)
1
3 (3) 9
1 (2)
4 (4) (5)
1
3 1
2 (3)
4 (2) 2
6 (1)
(1)
7 (2) 1
2 (2) 1 (4)
5 (3) 1 (3)
3 (2) (1)
4 (4)
1 (1)
2
4 (1) 1 (5)
69(51) 120
1[1]
1
[1]
3 3
2 2
13(2)
small, mostly winged, ground-living, hydro- phile and mesophile "bombardiers." Most of those in the north-temperate zone are speeies of Brachinus (see below). Small fonns of the African-Oriental genera Styphlomerus and Mastax reach Java and the Philippines (Styphlomerus, Timor too) but not Celebes. New Guinea possesses
only six species of the pantropical genus PheropsopJuis, one occurring also in Aus- tralia and the others endemic and compris- ing at least one Oriental relationship, and one species of Brachinus. Tlie latter genus occurs over most of the world except Australia; species are diverse in the Ori- ental tropics but become fewer eastward.
200
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
the New Guincan species (perhaps con- fined to western New Guinea) being the easternmost.
[57] The final trilie, PSEUDOMOR- PHINI (Part III, p. 239, with footnote 4), consists of aberrant, small and medium- sized, winged, chiefly tree-tiiink-living carabids.
New Guinea possesses representatives of two Australian genera with together seven endemic New Guinean species prob- ably representing at least five independent Australian relationships, and one endemic species of a genus that is otherwise known only from the Oriental end of the Malay Archipelago and the Malay Peninsula.
[58] Summary of geoiiraphic units. The geographic units (see [31]) itemized in the preceding survey of distribution of tribes of Carabidae that are represented in New Guinea (sections [33-57]) are sum- marized and totaled in Table 8. The table shows that New Guinean Carabidae in- clude 104 species that occur also in the Oriental area (including Celebes and the Philippines) plus 69 additional Oriental relationships, and 69 species that occur also in Australia (beyond Cape York) plus 51 additional Australian relationships, or a total of 173 Oriental and 120 Australian geographic units. (Occurrences of addi- tional New Guinean forms on the Moluccas in one direction and Cape York in the other are noted in sections [62] and [63].) This is, of course, a summary of the present relationships of New Guinean Carabidae without regard to their geographic his- tories.
It should be emphasized again that the geographic patterns are complex, that the limits of the areas compared are arbitrary, and that the selection of units (geographic units, rather than species or genera) is arbitrary too. It will be interesting to see how changes in these arbitrary decisions change the findings summarized in Table 8.
If Celebes and the Philippines were removed from the "Oriental area," and comparison made between the Oriental
Region (in a strict sense) and Australia, the totals of Table 8 would be changed in favor of Australia. On the other hand, if the rain-forested areas of North Queens- land, which are botanically Malaysian and to which many of the New Guinean Carab- idae that do reach Australia are confined, were removed from the "Australian" area, the change would be in favor of the Orient. However, I see no reason to go further with this juggling of boundaries. The bound- aries as selected, which allow comparison of the New Guinean carabid fauna with the faunas of all the major islands to the west (excluding only the Moluccas) and of virtually the whole of Australia (ex- cluding only the Cape York peninsula) seem to me to permit a reasonably bal- anced assessment of the situation.
The probable effect of increase of knowledge of the fauna of Celebes should be noted. The Carabidae of Borneo and especially of Java and the Philippines and also those of New Guinea and Australia are now fairly well known, but those of Celebes and the Moluccas are much less kno\\'n. If the Carabidae of Celebes were better known, they would almost surely be found to include representatives of ad- ditional New Guinean groups, which would increase the relative strength of the "Oriental" relationships of the New Guin- ean fauna as summarized in Table 8.
The effect of using different taxonomic units to quantify faunal relationships is more difficult to assess. If all species of Carabidae on New Guinea were counted as separate units, and if (where details arQ unknown ) species were scored according to their general relationships, most of thei 160 species of Agonini on the island would count as Oriental units, and so would most species of Harpalini of subtribes Ilarpalina and Acupalpina and also most Lebiini except Deinetrida, while most of the spe- cies of Harpalini of subtribes Anisodacty- lina and Pelmatellina and also the 5S species of Demetiida, eight species ol Scopodcs, and most Licinini and Ilelluonin:
The carabid beetles of New Guinea
Darlington
201
would count as Australian. I think it is better not to attempt to make more exact counts of species which would (I think) conceal ignorance of details rather than add to knowledge of faunal relationships. It is enough to say that counting all species as separate units would not decrease but would probably increase the relative strength of the Oriental relationships of the New Guinean carabid fauna.
[59] Geofiraphic units by major habi- tats. Table 8 can be broken down by major habitats, to see whether Carabidae in dif- ferent habitats in New Guinea have differ- ent geographic relationships. I shall do this only in a general way (because too many details are unknown to allow c^uanti- fication ) and only for groups that occur in the lowlands (because Carabidae may change habitats as they move to higher altitudes). Among the teiTestrial (meso- phile) lowland Carabidae of New Guinea, those in rain forest, including various Agonini, Pcri'^ona, Harpalini of subtribe Harpalina, and others, seem to be mostly Oriental in relationships, while those (fewer) in opener, drier areas especially in southern New Guinea, including most Harpahni of subtribe Anisodactylina, most Helluonini (if they are terrestrial) and a few others, are Australian. Among hydro- philes, Abacetus, the water-loving Agonini, and most Chlacnius, are basicallv Oriental, while the pterostichine genera PJatycoehis and Loxandrus and its derivatives and most of the (few) Licinini are Austrahan. The water-loving species of Clivina and Tachys are divided, but more of the Tachys probably have Oriental than have Austra- lian relationships. Among arboreal Carab- idae, the arboreal Agonini and manv Lebiini {Catascopus, Misceliis, Doliclwc- tis, etc.) are Oriental in relationships, while the lebiine genus Demetrida and the Pseudomorphini are mostly Australian (see Table 7 for summary of geographic re- lationships of the Lebiini). This brief statement of the geographic relationships of New Guinean Carabidae by major
habitats omits a number of small groups and groups of which the habitats or rela- tionships are undetermined. Also omitted are important groups that have radiated primarily on New Guinea (see [91]). However, what has been said does justify the following tentative conclusions. Among lowland, ground-living (mesophile) Carab- idae on New Guinea, those in rain forest are predominantly Oriental; those in opener country, more Australian. Among the hydrophiles, relationships are partly Oriental and partly Australian, but the Oriental relationships probably are more numerous. And among the arboreal forms, which occur principally in rain forest, rela- tionships are much more Oriental than Australian, if the comparison is made in terms of geographic units. (But if the 59 species of Demetrida were counted as- 59 Australian units, then Australian would outweigh Oriental relationships among the arboreal forms.)
To return to Table 8, the totals there given are approximations and would be changed to some extent if arbitrary deci- sions about the limits of areas and about the units to be counted were changed. How- ever, on any reasonable basis of compari- son, the carabid fauna of New Guinea is more Oriental than Australian, the pro- portion of Oriental to Australian relation- ships being, very roughly, as three to two. This is probably true of all the principal ccologic segments of the fauna except ot terrestrial mesophiles in the opener, drier part of southern New Guinea, where Aus- tralian elements predominate. However, it should be remembered that this is an oversimplified summary of an excessively complex situation, and that although the Australian relationships of the New Guin- ean carabid fauna as a whole are less numerous than the Oriental relationships, nevertheless some of the Australian rela- tionships involve important fractions of the fauna. It should be remembered too that this is a summary of existinii relationships, without regard to origins and directions of
202 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
movement, which will be considered sepa- rately.
[60] Transition of carahid faunas from Asia to Australia. Ideally, the transition of carahid faunas across the Malay Archi- pelago should be described in terms of actual counts of genera and species as they drop out, appear, or change from island to island, but this is not yet possible. The details of distribution of many species are still too little known; the faunas of Celebes and the Moluccas are too little known; and too many genera need taxonomic revision. To obtain for Carabidae the kind of in- formation that is now available for birds (Mayr, 1944) \\'ould require, I think, at least a lifetime of work by a competent specialist. The work would have to in- clude years of collecting, then years of taxonomic study, then probably further years of collecting to fill gaps discovered during the taxonomic work, and finally years of assembling, mapping, and pre- senting the geographic data. The best I can do now is to make a general descrip- tion of the apparent principal changes of carabid faunas from Asia to Australia, emphasizing major boundaries (if any) and major transition zones, and emphasiz- ing that the description is provisional. It is based, of course, mainly on infonnation given under the tribes in preceding pages (sections [32-.57]).
Temperate Asia has a carabid fauna of which many of the dominant elements are confined to the temperate zone or at least are relatively unimportant in the tropics. Some tribes are primarily north temperate, and so are such dominant genera as Cara- J)us, Bemhidion, Amara, many subgenera of Pterostichus, Harpahis, and Anisodactij- lus.
The change from temperate to tropical Asia is very great, as far as Carabidae arc concerned. Many dominant northern groups disappear or become insignificant; many tropical tribes and genera appear; and arboreal Carabidae, especially Lebiini, become relatively numerous.
Within the tropics, many genera and species of the Asiatic mainland reach adjacent islands of the Malay Archipelago, where a number of additional genera ap- pear or become important; there are of course many minor changes from the mainland to the islands, but the transition as a whole is not striking.
[61] Wallace's Line and Celebes. Many carabid genera that reach Java and Borneo (and often the Philippines, too) do not occur on Celebes across Makassar Strait, which is the central part of Wallace's Line. Striking examples are the huge, flattened carabids of the genus Mormolyce ( Mormo- lycini), and also the myrmecophilous Paussini, of which 27 species in eight Ori- ental genera have been recorded from Java but none from Celebes. The small tribes Hiletini and Siagonini also reach Wallace's Line from the Orient but do not cross it. Additional Oriental genera that reach this part of Wallace's Line but are not kTiown to cross it are named in the survey of tribes [3.'3-57] in the Ozaenini, Scaritini, Pana- gaeini, Pterostichini, Agonini, Licinini, Chlaeniini, Oodini, Harpalini, Lebiini, Hexagoniini, Zuphiini, and Brachinini. In the other direction, a few (but very few) Australian or New Guinean groups reach Celebes without crossing Makassar Strait to the Orient; they include Loxandrtis (subaquatic Pterostichini), Minuthodes (tree-trunk-living Lebiini), and Dolichoctis of the aculcata group (foliage-living ar- boreal Lebiini). Wallace's Line, at least the central part of it (the distributions of Carabidae at the northern and southern ends of the Line are too Httle known to be discussed now), is evidently an im- portant boundary for many, but not all, Carabidae. Actually, the Line is primarily the eastern boundary of the full-scale Ori- ental carabid fauna. East of it begins the main transition from Oriental to Australian i faunas. Among Carabidae and many other insects the transition area extends from Celebes to New Guinea and includes the latter.
The carabid beetles of New Guinea • Darlington
203
Tlie fact that the Carabidae of Celebes are less well known than those of Ja\a and Borneo presumably increases the apparent importance of Wallace's Line, but enough Carabidae are known from Celebes to show that important fractions of the fauna do change with passage from the Oriental Region proper to this island. Zoogeo- graphically, Celebes is in fact an anomalous island for Carabidae, as it is for many other animals. Oriental genera of Carabidae that do cross Wallace's Line to Celebes but are not known to reach the Moluccas etc. include tlie fossorial hydrophile Scarites, the mesophile haqDaline Oxycentrus, the foliage-arboreal lebiines Lioptera and Callida, the mesophile dryptine Golcritula (Galerito), and the helluonine MacrocJiei- lus. A few Australian or New Guinean carabids that reach Celebes but are not found farther west are named in the pre- ceding paragraph. The carabid fauna of Celebes seems as a w^hole to be more Oriental than Australian or New^ Guinean in relationships, but it is still much too poorly known for detailed analysis. (For a more general discussion of Wallace's Line see Mayr, 1944, and Darlington, 1957, and for the relation of the Line to some insect distributions, see Gressitt, 1959.)
[62] Moluccas. The Carabidae of the Moluccas too are poorly knowm. I have made a rough list of them, based on pub- lished records and on a collection made by myself on Morotai Island, but the total scarcely exceeds 100 species. This is surely a minor fraction of the total Moluccan carabid fauna. However, a few important Oriental groups of Carabidae do reach these islands but not New Guinea. Tliev include t\\'o conspicuous genera, OiiJw- gonius (tribe Orthogoniini ) and Trigono- toma (Pterostichini), and one very distinct Tachijs (interpunctotus Putzeys), the latter collected by myself on Morotai. On the other hand, Australian and New Guinean Carabidae that reach the Moluc- cas but are not known from Celebes in- clude a TacJujs of the serra group, the
pterostichine genera Prosopogmus and Platijcoehis, probably the agonine Viola- gonum, the haqialine Diaphoromerus, and the lebiine Demetrida (for the last see Darlington, 1968a). These details suggest that the Moluccas share more Carabidae with New Guinea than with Celebes. However, the Carabidae of the Moluccas are inadequately known, and different Moluccan islands probably have different carabid faunas with somewhat different geographic relationships. (I think the Moluccas may have been a bottleneck in dispersal, because of their small area. This possibilitv is further discussed in section [84].)
The change of Carabidae from Celebes through the Moluccas to New Guinea is considerable. It involves (for example) change among large Scaritini from an Oriental Scarites in Celebes to an Austra- lian Geoscaptus in New Guinea, and among arboreal Lebiini from Oriental Callida in Celebes to Australian Deme- trida in New Guinea. But these examples should not be overemphasized. Our rela- tively poor knowledge of the Carabidae of Celebes prevents an exact assessment of the change of Carabidae that surely does occur from this island to New Guinea.
[63] Transition from New Gidnea to Australia. Within New Guinea itself there is transition from the still predominantly Oriental carabid fauna of the rain forest to the more Australian faima of the opener country especially of southern New Guinea. This transition is most obvious among Harpalini [59]. It involves also disappear- ances of many arboreal rain-forest species that do not enter opener habitats. How- ever, the Carabidae that do occur in opener habitats in southern New Guinea are so little known (much less well known than those in rain forest) that this transition cannot yet be described in much detail.
Transition of Carabidae from New Guinea to Australia is in part a continu- ation of the transition that begins within New Guinea, from the rain forest to the
204 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
more open country, for the open eucalyp- tus woodland of southern New Guinea is little more than a (depauperate) extension of the much larger areas of eucalyptus \\'Oodland of Australia. However, tran- sition occurs also within the rain forest and is in fact strikin'j^ as one proceeds from the great forests of New Guinea, to the isolated rain-forest tracts on Cape York, to the larger tracts farther south in eastern Aus- tralia. This transition has been described more detail elsewhere (Darlington,
m
1961 ) . I \\'ant now merely to summarize it, and to say something more about the pattern of distribution of New Guinean Carabidae on Cape York.
The transition of carabid faunas from the rain forests of New Guinea to those of Australia involves more than a change of particular genera and species. Two more- profound changes occur. One is a virtually complete change of flightless stocks, which are very few at low altitudes in New Guinea but which are more numerous and of wholly different origins in Australia, even on Cape York ( see below ) . The other is a change from overwhelming dominance of Agonini in New Guinea to overwhelm- ing dominance of Pterostichini in Australia (cf. [64], Fig. 13). And this change too occurs even in rain forest and even on Cape York. To the collector, these changes tend to be concealed by the occurrence of some New Guinean species in the Austra- lian rain forests; entomologists familiar with the Australian insect fauna find the New Guinean species in North Queensland new and specially exciting. But actually, among Carabidae and I think among many other insects too, the New Guinean species form a minor fraction of the Australian fauna even in the northern rain forests.
The number of important lowland New Guinean Carabidae, most of them oc- curring in rain forest, that do not reach Australia is impressive. Among Cicindel- inac, although the Tricondyla does reach Australia (mid-Cape York), Oriental The- rates (with at least five species in New
Guinea) and endemic Caledonomorpha (two species) do not, and the very small subarboreal species of Cicindela that have radiated in the rain forests of New Guinea are, I think, entirely unrepresented in those of Australia. The only New Guinean ozaenine (Oriental Fseudozaena) docs not reach Australia. A few Tochys are common to New Guinea and Australia, but most New Guinean members of the genus, in- cluding the endemic serra group, do not reach even Cape York. The endemic trechine genus Perilcptodes does not reach Australia. Among Pterostichini, the Oriental Bmchidius and the endemic Homalonesiota, Nehrioferonia, and Tiferonia are absent in Australia. Among Agonini, Arhytimis, Tarsa- gonum, and Eiiplencs (all represented also in the Orient) do not reach Australia, nor do the endemic Lithagonuni and Iridago- niun, which are common at low altitudes in New Guinea, and only one of the many New Guinean species of Notagontim and no Altogonum reach Australia; the fact that Notagonum dcntellum, Violagonmn vioJa- ceum, and Colpodes hahilis do occur in the Australian rain forest should not be allowed to obscure the fact that most New Guinean Agonini do not. Of 14 New Guinean Perigomi, only the cosmopolitan nigriceps reaches Australia. Codes of the terrestris group, which inhabit rain-forest leaf litter in New Guinea, do not reach Australia. Of Harpalini, Trichotichnus, llarpaloxemis, and Lytcr, as well as IIy))]iaL'reon, which together form an im- portant fraction of the carabid fauna on the ground in rain forest in New Guinea, are absent in Australia. The endemic New Guinean anaulacine Odontomasorctis does not reach Australia. Of Lcbiini, Syniirus, Stenotehis, Pericahis, Oxyodonttis, Moch- therus, and some other Oriental genera represented in New Guinea fail to reach Australia, and foliage-living Demetrida, dominant in New Guinean rain forest, are very poorly represented in the rain forests of Australia. Among Odacanthini, Dicras- pcda is primarily New Guinean, with onlyi
The carabid beetles of New Guinea • Darlington
205
hrunnca well distributed in the Australian rain forest and duhia only on the tip of Cape York, and endemic Dohodura is unknown in Australia. And among Bra- chinini, Brachinus fails to reach Australia, and only one species of Fheropsophus does so. This is far from a complete list of carabids that occur in lowland rain forest in New Guinea but not in Australia. Those that do reach Australia represent a small fraction of the Ne\\' Guinean carabid fauna, and they are in the minority also in the Australian rain forest, where a majority of the Carabidae are derived from Austra- lian groups.
This is not the place to attempt to list all the Carabidae occurring in the different rain-forest areas in Australia. In any case such lists would be difficult to prepare. My collections from these rain forests are extensive but far from complete, and many of the species found are still unidentified and probably undescribed. However, some significant details are worth giving, and the occurrence of actual New Guinean spe- cies of Carabidae in the Australian rain forests is worth tabulating.
Even on the tip of Cape York, in the tip-of-peninsular rain forest (see map. Fig. 12), the outstanding carabid is an enor- mous, flightless pterostichine (Mecijno- gnothm) of an Australian group, and the only other flightless carabid in this rain forest is a large Clivina closely related to species elsewhere in Australia. (Other flightless Australian Carabidae occur in the adjacent opener forest.) In the mid- peninsular rain forest, the flightless Carab- idae are a large Australian pterostichine {Paranunis) and a probably undescribed (Australian) Coptocarpus. And in the much larger base-of-peninsular rain forests, the dominant Carabidae are almost all Australian; half a dozen flightless Austra- lian genera are conspicuous; and some of them, including several genera of Pterosti- chini, have radiated in the rain forest. See again my 1961 paper for further details. In brief, while the relatively small rain forests
NEW GUINEA
Tip-of-pen. rain forest
Mid-pen rain forest
Cocktown
Base-of-pen. Cairns — °'' n^°in tropical rain forest
Townsville
Figure 12. Distribution of rain forest in tropical Nortfi Queensland, Australia. From Darlington, 1961, p. 8, pi. 3. Heavy lines enclose principal areas of tropical rain forest, but the rain forest is usually not continuous witfiin the boundaries shown. The rain-forested areas are separated by wide barriers of relatively dry, open eucalyptus woodland.
at the tip and middle of the Cape York Peninsula have independent, mixed (but I think more Australian than New Guin- ean) carabid faunas, the main tropical rain forests of Australia at the base of the peninsula are overwhelmingly Australian in their Carabidae. The situation is surely complex, and some other insects may show closer relationships between the New Guinean and Australian rain forests, but nevertheless I think it is a mistake for entomologists to include the rain forests of Australia in a Papuan region or subregion.
The distribution of rain forests in east- ern Australia is mapped in Figure 12, and the known distributions of 24 New Guinean species of Carabidae in the tip-of- peninsular, mid-peninsular, and base-of- peninsular (main) rain-forest areas of tropical North Queensland are shown in Table 9. For further description of these rain forests, see my paper already cited (1961). Note that the rain-forested areas are small; actually, the rain forest is often not continuous even within the boundaries shown, but occurs in discontinuous or scat- tered strips or patches. The extent and continuity of rain forest in North Queens- land has often been greatly exaggerated.
Table 9 is based on my o\\ai extensive
206 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Table 9 Occuri\ence of New Guinean species of Carabidae in rain forest on Cape York, etc.
(See Fig. 13)
Name
Tip-of-pen. forest
Mid-pen. forest
Base-of-pen. forest
-X .X
x x
-X ..X
TricomUjla aptera - — ^
Syletcr papua - x
Clivina zebi x -X — -
Morion longipenne —- x— - — x
Caelostomus alhertisi — - — ^
Notafionum dcntellum x - x - -
V iohigonuDi liolaccuin x — —
Colpodes kahilis ^
Anatrichis pusilla — ^
Coleolixsus nr. papua _ x
Aristolehia wait x
Miscelus unicolor — '^
Catascoptis clegans x — x-__- — — x
II smarafidulus x
I > antensis ^
Dolidwctis striata -- ^
Demetrida angulata _.__ — x
Pentagonica pallipes ^
II blanda x -- x
II erichsoni x
CoUitiris par x
Clarcncia quadriguttata ^
Dicraspeda brunnea x x. x
II dubia — - X
collecting on Cape York, and on previously published records. Further collecting would probably fill some gaps, but never- theless carabid distributions evidently are irregular in these rain forests. Of the 24 New Guinean species tabulated, six have been found only on the tip of the peninsula, but only six of the other species have been found there, leaving 12 species that occur in more-southeni rain-forest tracts but are apparently absent on the tip of Cape York. The tip-of-peninsular rain forest is small and of comparatively poor quality, and the carabid fauna probably really is much more limited than are the faunas of the larger and better rani forests farther south. Sixteen of the 18 species that do occur farther south have been found in the mid- peninsular rain forest. And only ten have been found in the base-of-peninsular forests (but see below).
The detailed pattern shown by Table 9 should be supplemented by two general
statements. First, a few of the species tabulated, including Viohfionum viola- ceum, extend still farther south in Austra- lia. And second, a considerable number of endemic species derived from New Guinean groups exist in the base-of- peninsular rain forest, and in some cases still farther south in Australia. (This whole situation, of somewhat irregular occurrence of New Guinean species in the isolated rain forests of Cape York, and existence of additional derived species farther south, suggests continual dispersal of rain-forest j species from New Guinea to Australia. The insects probably disperse with diffi- culty from forest tract to forest tract. Some species probably survive only temporarily in some tracts. Survival is probably cor- 1 related with area: extinctions probably occur most often in the smallest tract, which is the tip-of-peninsular one; while survival time in the largest tract, at the base of the peninsula, has often been long
The carabid beetles of New Guinea • Darlington
207
enough for differentiation of species. All this is consistent with a history of con- tinuing dispersal from New Guinea into Australia, across ecologic filter-barriers, even when there was a broad land con- nection. Such dispersal across ecologic barriers to a series of islands of rain forest is comparable to dispersal across water gaps to the islands of an archipelago and should be susceptible to analysis by meth- ods developed by MacArthur and Wilson (1967).)
To complete this general account of transition of carabid faunas in the Asiatic- Australian transect, I should add that within Australia, between the tropical rain forests of North Queensland and the south- temperate rain forests of southern Austra- lia and Tasmania, there is not only an almost complete change of species and genera but also a second partial change of dominant tribes, from Pterostichini as principal dominant mesophiles to (in the far south) dominance shared by Broscini, Trechini, Licinini, and even "antarctic" Migadopini and Merizodini, as well as some Pterostichini. This change too is described in more detail in my 1961 paper.
[64] Summary of tramition,s. In sum- mary of transitions of carabid faunas from north-temperate Asia to south-temperate Australia, there is first a profound change of dominant tribes and genera from the north-temperate zone to the tropics in Asia; then a major transition of tropical faunas from the Orient to Australia, with the most obvious changes at Wallace's Line, between Celebes and New Guinea, and (even in rain forest) between New Guinea and tropical AustraUa; and finally another profound change of dominant tribes and genera from tropical to south-temperate Australia. The carabid faunas in the north and south temperate zones, at opposite ends of this series of transitions, are re- markably similar in certain ways, for ex- ample, in presence of Broscini, of flightless "Trechiis," and of Bemhidion. These groups must somehow have crossed the tropics in
the past. However, they do not occur in New Guinea now, and further consider- ation of them would be out of place here. I have discussed them in more detail else- where (1965).
The gross changes in taxonomic com- position of carabid faunas within the trop- ics, from Java to New Guinea to tropical Austraha (North Queensland), are shown by histograms in Figure 13.^ The histo- gram of the Javan fauna is based on a list extracted from my MS list of Indo- Australian Carabidae [4]. That for the New Guinean fauna is, of course, based on counts of species listed on my data sheets [16]. And that for the tropical Australian (North Queensland) fauna is based on a list extracted from my manuscript list of Australian Carabidae [4]. Many species described from "Queensland" are not known from more exact localities and may not be tropical, but on the other hand I have a number of tropical Queensland spe-
^ The carabid faunas of Celebes and the Moluc- cas are too little known to be included in this comparison. Celebes is about half again larger than Java, but only a]:)out one-third as many Carabidae (only about 150 species) have been recorded from it. (Professor E. O. Wilson calcu- lates, using data provided by me, that if Celebes were as well collected as Java and New Guinea, about 509 species of Carabidae should be known from the island. ) The Moluccas are much smaller, but tlie number of species that occur there is presumably increased by differentiation of species on different islands of the group. About 100 species of Carabidae have actually been found there ( including both those recorded in print and those found by myself on Morotai Island), but this is probably a minor fraction of the whole Moluccan carabid fauna. The following table shows the numbers of species in four principal tribes of Carabidae actually known from Celebes and the Moluccas. The figures suggest that Agonini and Lebiini are the dominant tribes on these islands as they are on New Guinea, but the figures should be considered preliminary indi- cations only.
Celebes Moluccas Pterostichini 8 13
Agonini 23 18
Harpalini 20 8
Lebiini 36 35
208
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
|
< 2 |
||
|
o <0 |
||
|
CD |
||
|
.8 |
V/// |
yiHio iN,iNoan3Sd iNiNiHOvya
INI,NVOVaO
iNiia3"i
INHVddVH
INIIN3V1H0
ININOOtf
iN.isoaaid
iNiiaiaiN39
iNiiiyvos
iNissnvd
iN,woan3Sd iNiNiHovya
INl.NVOVaO
iNiiaan mnvdbivH
lNIIN3VnH0
IN1N09V
IN.lSOySld
iNiiaia^3a
iNiiiavos
iNissnvd
sq;
|
< |
_J |
|
> |
< 1- |
|
< |
2 UJ |
|
-^ |
o |
o c: c
./> ° .2
<U t/i o
E .9 £
E ■- =
o ,„'
D
!= "O
c
O <l> 0) -3; ^ N
o
Q- O O
o E
D
Z ^
_0
o ■- -o
o OJ " E
<u
|
a3H10 |
0 |
■*~ |
C |
|
,»_ |
01 |
— |
|
|
■ ~ |
c |
||
|
N.iNOanssd |
0 |
||
|
ININlHOVda |
Q- E 0 |
c |
0^ F |
|
u |
c 0 0) |
0 |
|
|
INI.NVOVaO |
'e |
3 1/1 |
|
|
iNiia3n |
0 |
3 |
O- |
|
0 |
-n |
||
|
INnVddVH |
p |
Wl |
c 0 |
|
IN1IN3V1H0 |
(a |
0 |
, |
|
c |
|||
|
1NIN09V |
^ |
0 |
"cT |
|
<u |
|||
|
IN,lS0a31d |
f |
0 |
|
|
0 |
F |
0 |
|
|
iNiiaiaiM39 |
en |
0 |
-0 |
|
0 |
-t |
U |
|
|
INiiiavos |
if |
c |
'E |
|
iNissnvd |
r_" |
||
|
CO |
-i£ |
||
|
<u |
c/) (II |
■5; |
|
|
3 |
|||
|
O) |
Q. |
(U |
t I f A
The CARABiD BEETLES OF New Guinea • Dcuiington
209
. Ml - > t
'•m^
IWI831 n
0 U Cl
V«"V4
11:
:3k.C i '
g c -
11(11931
ies which are undetermined and probably lundescribed. I have tried to take these
actors into account in totahng the North Queensland list, which is therefore based partly on estimates rather than on counts, although the estimates are carefully made,
iby scanning lists of species. However, I have made actual counts of species of ^JL.« jE-. vJqji^]^ Queensland Agonini and in most *'"''5» ^?5;JJTenera of Pterostichini, so that the figures "' Jjiven for these two specially significant ribes are exact or nearly so.
Java is a recent continental island and its auna is a fair if probably depauperate ;ample of the fauna of eastern tropical \sia, and the fauna of North Queensland s a good sample of that of tropical Aus- ralia. The three histograms therefore ihow major changes in composition of Topical carabid faunas from Asia to Aus- ralia. The changes that occur in several lominant tribes are worth further dis- cussion.
The tribe Scaritini, which is the princi- pal tribe of fossorial Carabidae, is well •epresented in all three faunas. However, \.-xim ',<>\^'arg,e scaritines are relatively numerous in [ava and (different genera) in North ; Queensland but veiy deficient in New : juinea, where almost all members of the • 'irribe are small. The predominance of imall foHTis in New Guinea probably re- lects their greater dispersability. : Bembidiini, mainly Taclujs, are well ■epresented in all three faunas and include nost of the very small Carabidae in all hree. Many are hydrophiles, some meso- ohiles, and a few arboreal. The large lumber of these small carabids in New Guinea surely reflects their greater dis- persability. They have reached the island n such numbers as to impose a second node on the size distribution of Carabidae here (cf. [20], Fig. 8).
Harpalini and Lebiini are well repre- :ented in all three faunas. The Harpalini nclude many medium-sized mesophiles md a number of smaller hydrophiles. The ^ebiini are mostly arboreal. Members oi
«!(03t
oM
-, ^Tii
N»:::«d i.
c ■
0^^ r:
both tribes are evidently good dispersers. The relatively smaller number of Lebiini in North Queensland probably reflects the smaller extent of rain forest there.
Finally (as far as dominant tribes are concerned) the Pterostichini and Agonini show remarkable geographic changes of dominance. On Java, the two tribes are about equally represented. On New Guinea, Agonini are overwhelmingly domi- nant, being three or four times as numerous in species as Pterostichini are. But in Australia, even in the tropics, dominance is reversed, Pterostichini being many times more numerous in species than Agonini.
I have discussed this reversal of domi- nance before (1956; 1961), saying (1961: 22-23) "Pterostichini and Agonini tend, as dominant tribes, to be complementary over the world as a whole. . . . Both tribes are cosmopolitan, but unevenly so. In some parts of the world they occur in nearly equal numbers; in others, one tribe or the other is over\\'helmingly dominant. The tribes tend to be complementary within the Australian Region, .... In [the whole of] Australia . . . (with Tasmania) Pterosti- chini are dominant, with more than 350 known species against probably less than 20 species of Agonini, a ratio of nearly 20/1. But in New Guinea Agonini are dominant, with considerably more than 100 known full species . . . against about 40 species of Pterostichini ... a reversed ratio of about 3/1.
"One reason for the number of Agonini in New Guinea is that species of this tribe have multiplied on the mountains there. In Australia, however, Pterostichini, not Agonini, ha\'e multiplied in what seem to be comparable habitats on the mountains. This difference can hardly be accounted for in simple ecological terms but is probably due to a complex combination of ecological, historical, and geographical factors. Over the world as a whole, there is a tendency for Agonini to be better represented in the tropics; Pterostichini, in the temperate zones. Also it is probable
I
208 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
|
LlI |
<r |
|
li 1 |
|
|
-1 |
_i |
|
a |
< |
|
(r |
|
|
_j |
h- |
|
< |
U) |
|
o |
ID |
|
Q- |
< |
|
O |
*■ — ' |
|
cr |
|
|
h- |
iN,woan3Sd iNiNiHovya
INI,NVOVaO
iNiiaai
INIlVdaVH
INllNSVnHO
ININ09V
IN,iS0a3id
iNiiaiaiN39 iNiiiavos iNissnvd
a3HiO
iN,woan3Sd
ININIHOVaS
INI.NVOVaO
INIiaS"!
iNintfdavH
INIINSVTHO
IN1N09V
IN,iS0a31d
iNiiai9iN3a
iNiiiavos
iNissnvd
asHio
iN.tNoanasd
NiNiHOtfaa
INl.NVOVaO INIiaS"!
iNinvdavH
IN1IN3VTH0 ININ09V
lN,lS0d31d lNliai9W39 INIliaVOS
iNissnvd
E
o o
a
ID
Z i
o >
■ - - t ;! ,
•5' S. |c
■ I 7T iP.r
A. ^
V f ■ ? 9 ^ i ^ i^ 1^ >
»' f
k ». ..
The carabid beetles of New Guinea
Darlington
209
cies which are undetermined and probably undescribed. I have tried to take these factors into account in totahng the North Queensland list, which is therefore based partly on estimates rather than on counts, although the estimates are carefully made, by scanning lists of species. However, I have made actual counts of species of North Queensland Agonini and in most genera of Pterostichini, so that the figures given for these two specially significant tribes are exact or nearly so.
Java is a recent continental island and its fauna is a fair if probably depauperate sample of the fauna of eastern tropical Asia, and the fauna of North Queensland is a good sample of that of tropical Aus- tralia. The three histograms therefore show major changes in composition of tropical carabid faunas from Asia to Aus- tralia. Tlie changes that occur in several dominant tribes are worth further dis- cussion.
The tribe Scaritini, which is the princi- pal tribe of fossorial Carabidae, is well represented in all three faunas. However, Jorge scaritines are relatively numerous in Java and (different genera) in North Queensland but veiy deficient in New Guinea, where almost all members of the tribe are small. The predominance of small fonns in New Guinea probably re- flects their greater dispersability.
Bembidiini, mainly Tachijs, are well represented in all three faunas and include most of the very small Carabidae in all three. Many are hydropliiles, some meso- philes, and a few arboreal. The large number of these small carabids in New Guinea surely reflects their greater dis- persability. They have reached the island in such numbers as to impose a second mode on the size distribution of Carabidae there (cf. [20], Fig. 8).
Harpalini and Lebiini are well repre- sented in all three faunas. The Harpalini include many medium-sized mesophiles and a number of smaller hydrophiles. The Lebiini are mostly arboreal. Members of
both tribes are evidently good dispersers. The relatively smaller number of Lebiini in North Queensland probably reflects the smaller extent of rain forest there.
Finally (as far as dominant tribes are concerned) the Pterostichini and Agonini show remarkable geographic changes of dominance. On Java, the two tribes are about equally represented. On New Guinea, Agonini are overwhelmingly domi- nant, being three or four times as numerous in species as Pterostichini are. But in Australia, even in the tropics, dominance is reversed, Pterostichini being many times more numerous in species than Agonini.
I have discussed this reversal of domi- nance before (1956; 1961), saying ( 196L 22-23) "Pterostichini and Agonini tend, as dominant tribes, to be complementary over the world as a whole. . . . Both tribes are cosmopolitan, but unevenly so. In some parts of the world they occur in nearly equal numbers; in others, one tribe or the other is overwhelmingly dominant. The tribes tend to be complementary within the Australian Region, .... In [the whole of] Australia . . . (\\'ith Tasmania) Pterosti- chini are dominant, with more than 350 known species against probably less than 20 species of Agonini, a ratio of nearly 20/1. But in New Guinea Agonini are dominant, with considerably more than 100 known full species . . . against about 40 species of Pterostichini ... a reversed ratio of about 3/1.
"One reason for the number of Agonini in New Guinea is that species of this tribe have multiplied on the mountains there. In Australia, however, Pterostichini, not Agonini, have multiplied in \\'hat seem to be comparable habitats on the mountains. This difference can hardly be accoimted for in simple ecological terms but is probably due to a complex combination of ecological, historical, and geographical factors. 0\'er the world as a whole, there is a tendency for Agonini to be better represented in the tropics; Pterostichini, in the temperate zones. Also it is probable
210 Bulletin Museum of Comparative Zoology, Vol 142, No. 2
that Agonini, which are phylogenetically less diverse, are more recent in origin than Pterostichini and that they have dispersed more recently. It is therefore likely that Pterostichini are dominant in Australia partly because Australia is more temperate than tropical in climate and partly because Pterostichini reached Australia before Agonini did, and it is likely that Agonini are dominant in New Guinea partly be- cause the climate there is fully tropical and partly because the carabid fauna of New Guinea is more recent in its origins than that of Australia, as I think it is. Add to this that the mountain carabid faunas of Australia and New Guinea have been de- rived independently, each from the low- land fauna adjacent to it, and not by dispersal along a connecting mountain chain, and we have an adequate and prob- ably correct explanation of the great dif- ference in composition of the carabid faunas on the mountains of Australia and New Guinea."
To extend the comparison to Java, where numbers of species of Pterostichini and Agonini are nearly equal, I suppose the Javan carabid fauna is continental in com- position and includes more relatively old forms than the New Guinean fauna does, the ratio of Pterostichini to Agonini on Java being perhaps near the average for the world as a whole. The increase in relative numbers of Agonini from Java to New Guinea probably reflects the greater dis- persability of Agonini as well as the lesser age of the New Guinean fauna.
[65] Fauna! regions. The preceding summary leads to the question, in what faunal region should New Guinea be placed (if it must be placed in a faunal region), according to its Carabidae? The answer is that // it must be placed in a faunal region, and if the faunal regions are based (as they should be) on present geo- graphic relationships without regard to past movements, New Guinea must be considered part of the Oriental Region, so far as its Carabidae are concerned.
However, the preponderance of Oriental over Australian relationships is not over- whelming, and I prefer to take Wallace's Line as the eastern boundary of the Ori- ental Region and to consider the Carabidae' as forming a broad transition from there to and including New Guinea. The latter can then be considered part of a transition zone, an extended "Wallacea." Or, better. New Guinea can be kept where Wallace put it (and where the distributions ol vertebrates put it) as part of the Australian Region, with the understanding that the transition of Carabidae (and of many other insects) does include New Guinea nevertheless. This solution of the problen' has the advantage that it does not make confusing changes in the boundaries of the conventional regions. The pattern o) faimal regions is a standard known to al. zoogeographers and continually referrec to in describing and comparing the dis- tributions of different groups of animals The pattern would lose much of its value if it were continually changed to make il fit the distributions of special groups. ( Foi more eletailed discussion of the nature ane; usefulness of faunal regions see Darlington 1957: 419ff.) Actually, I do not thin! regional boundaries are worth arguing about. What is important in any giver case is to make the situation clear. The transition of carabid faunas from Asia tc New Guinea and Australia is complex beyond my power of describing it in full but I hope I have said enough to make thej general outlines of it clear.
Some other families of insects in New Guinea probably show a higher proportior of Oriental relationships. This is likely tc be the case among insects that are arborea. and live in rain forest. They include thej Cerambycidae and Chrysomelidae studiec by Gressitt, who finds that the New Guinean faunas of these beetles are more Oriental than Australian in relationships and who therefore puts New Guinea ir' the Oriental Region (Gressitt, 1961, with map on p. 18). I have already given rea-
I
The carabid beetles of New Guinea • Darlington 211
Table 10. Widely distributed oriental species or species groups of Carabidae which reach
ONLY THE WESTERN PART OF NeW GuINEA OR ( BELOW THE BROKEN LINe) WESTERN AND CENTRAL BUT
NOT EASTERN NeW GuINEA, SO FAR AS KNOWN
Perileptiis japonicus (Part I, p. 489), reaches the Vogelkop
Abacetus convexiusculiis (Part I, p. 521), reaches Salawati I. and perhaps Dor(e)y on the Vogelkop
Anaulacus siamensis (Part III, p. 77) reaches Geelvink Bay, West N. G. (but is an inconspicuous carabid possiljly overlooked farther east)
Flatijmetopus hticeps (Part III, p. 48), reaches the Vogelkop and Biak I.
Catascopus facialis (Part III, p. 103), is recorded from Dor(e)y (a locality always somewhat doubt- ful) and from Maffin Bay but has not been found farther east in New Guinea although common on many islands west of New Guinea, including Morotai I. in the Moluccas
Microlestes curtatiis (Part III, p. 136), reaches Dor(e)y on the Vogelkop, if Wallace's labels are correct in this case
Tachijs coracinus (Part I, p. 481; present part. Tax. siippl.) reaches Astrolabe Bay, N-E N. G., but perhaps not extreme eastern New Gmnea.
Chlaeniiis pan (Part III, p. 23), member of an Oriental species group, reaches Sepik District, N-E N. G., but perhaps not farther east ( a conspicuous carab, not easily overlooked )
Brachimis papua (Part III, p. 239), member of an Oriental species group, reaches vicinity of Hollandia, West N. G. (also conspicuous, not easily overlooked)
sons (preceding paragraph) why the con- ventional faunal regions should be ac- cepted by all zoogeographers. If ento- mologists must change regional boundaries, I think the changes should await aquisition of more information. The ground-living and soil-inhabiting insects of New Guinea may not be Oriental in their relationships, or at least not decisively so; most of them are in fact too poorly known for analysis. Surely if a system of faunal regions is to be based on the distribution of insects, .it should be based on a synthesis of the dis- tributions of many different groups and not on a few selected families. In any case I think that the insects of the main (base-of- peninsular) rain forests of Australia will prove to be more Australian than New Guinean in present relationships and probably also in origins.
[66] Geo<^rophic patterns within Neiv Guinea. The distributions of Carabidae within the limits of New Guinea form a number of different geographic patterns at low altitudes, as well as a pattern of diminution and increasing geographic dif- ferentiation with increasing altitude (cf.
[26, 90]). The patterns are real and sig- nificant, although probably still incom- pletely known in most cases.
One set of patterns is formed by species or species groups which are widely dis- tributed outside the island but which are restricted in New Guinea itself. Several Oriental species or species groups which extend to New Guinea have been found only in the western part of the island (Table 10). Additional Oriental species and species groups will probably be found to have this pattern, of occurrence in the western but not in the eastern part of New Guinea, when the Carabidae of western New Guinea are better kno\vn; the western end of the island has been much less well collected than the eastern end. (This pat- tern, of course, suggests that the species that have reached only western New Guinea have arrived more or less recently from the Orient, or at least from the west. That so few recent arrivals are still re- stricted to the western part of New Guinea may be because most Oriental species, if they have sufficient dominance and dis- persal power to reach the island at all.
212 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Table 11. Australian species and species groups of Carabidae (including Cicindelinae in parentheses) that reach only southern New Guinea or (below the line) eastern but not
western New Guinea, so far as known
{Megacephaki (Part I, p. 335), 2 Australian species recorded from Merauke, but the record may be doubtful )
{Dlstipsidera (Part I, p. 337), an Australian genus with 2 endemic species described from southern New Guinea)
[Arthropterus novelliis (Part I, p. 354), locality within tlie island not specified but presumably southern New Guinea]
Clivina ferruginea (Part I, p. 387), Port Moresby
Tachijs bembidiiformis (Part I, p. 464), Port Moresby (occurs also on Java, Sumba, Tanimbar, etc.; may have dispersed via the Lesser Sunda Is. and reaches New Guinea from Australia)
Tachtjs convexus (present part. Tax. suppL), Oriomo R.
Dicrochile gigas (present part, Tax. suppl.), Rouku
Gnathaphamis picipes (Part III, p. 42), Port Moresby & vie. Brown R.
Gnathuphanus pulcher (Part III, p. 42), Port Moresby & vie, Bisianumu
Aciipalpus brunnicolor (Part III, p. 74), Port Moresby, Oriomo R.
Lehia melanonota ( present part. Tax stippL), Rouku (occurs also on Java, Lesser Sunda Is., etc., but presumably reached southern New Guinea from Australia)
Basisticiis micans (Part III, p. 208), Rouku
Drijpta mastersi (Part III, p. 217), Rouku
Ztiphiutn thouzeti (Part III, p. 219), Port Moresby
Helluosoma atrum (Part III, p. 233), Port Moresby & vie, Bisianumu, Rouku
Helluodcma unicolor (Part III, p. 233), Rouku, Merauke
Gigadema maxillare (Part III, p. 234) Rouku
Geoscaptus caciis ( Part I, p. 356; present part. Tax. suppl.), west to vie. Hollandia
Clivina basalts (Part I, p. 383; present part. Tax. suppl.), west to N-E N. G. (occurs also on Celebes and Java, but N. G. specimens from Fly R. probably derived from Australia )
Clivina sellata (Part I, p. 387), at Dobodura
Craspedophoriis (present part. Tax. suppl.), genus with numerous species in Australia and 1 endemic at Tapini, Papua (a separate group of species in tlie Orient)
Gnathaphanus philippensis (Part III, p. 42), Rouku and Kokoda (occurs also in Orient, but distribu- tion in New Guinea suggests arrival from Australia)
Gnathaphanus licinoides (Part III, p. 41), west to vie. Hollandia
Hypharpax dentipes (Part III, p. 44), west to Lae and Wau (occurs also in Java, but distribution in New Guinea suggests derivation from Australia )
Lachnoderma foveolatum (Part III, p. 89), not found west of Papua
Nototarus (Part III, p. 185), an Australian genus with 1 endemic species in eastern New Guinea, at Dobodura
Parascopodes ctjaneus (Fart III, p. 196), not found west of Papua
Casnoidea puncticollis (Part III, p. 207), Fly R.
Pseudomorphini (Part III, p. 239), New Guinean species of Adelotopus and Sphallomorpha are en- demic but apparently related to Australian species and have been found only in Papua and
N-E N. G.
The carabid beetles of New Guinea • Darlington 213
Table 12. Lowland Carabidae in which different ation of subspecies or species pairs has
OCCURRED in DIFFERENT PARTS OF NeW GuINEA
Clivuui dedlata (Part I, 372): a winged population on the Fly R. and different short-winged subspecies in N Papua and N central New Guinea
Clivina erugatella (Part I, 380): fully winged populations at Hollandia and Maffin Bay in West N. G., and a dimorphic population (mostly short-winged) at Aitape in N-E N. G. (although these populations are incipiently differentiated, I have not recognized them as subspecies)
Tachijs serra (Part I, 405): subspecies in Papua, N central N. G., and the Vogelkop
Tachijs serrula (Part I, 408): a fully winged subspecies in N-E N. G. (and New Britain), a short- winged population in N Papua
Tachys suhlohatus (Part I, 418): a subspecies at Lae and Dobodura, another at Maffin Bay
Tachijs mastersi (Part I, 420): a subspecies on the \^ogelkop and another in the main part of New Guinea (and other subspecies or closely related species in Australia and the Philippines)
Tachijs mascidus (Part I, 422): subspecies in N central N. G., and the Vogelkop
Tachys pictus (Part I, 447): subspecies in Papua, N N-E N. G., and Hollandia area, (and on the Bismarck Rge. )
Tachys latissimus (Parti, 474): a subspecies widely distributed in the Oriental Region and occurring also in E New Guinea, and a melanic subspecies in N West N. G.
Ahacetus haplosternus and straneoi ( Part I, 518 ) : a common intemiediate fonii is widely distributed in the main part of N. G. but splits into 2 closely related species in Papua (and New Britain)
Notagonum aitape (Part II, 141): subspecies in N central N. G., and the Vogelkop
Notagomim suhpiinctum (Part II, 146): a subspecies in Papua, and another in the main part of New Guinea and the Vogelkop
Notagonum pahidiim (Part II, 150): a subspecies in Papua, and another in N N-E N. G.
Lithagoniim annulicorne (Part II, 176): a distinct subspecies in Papua, less distinct ones in N N-E N. G., West N. G., (and in the mountains)
Altagonum vaUicola (Part II, 190): subspecies in Papua, Huon Peninsula, and West N. G.
Altagonum grossidiim (Part II, 191): a subspecies in Papua, and 2 more subspecies in different parts of West N. G. (the western subspecies are in low mountains but probably range to the lowlands too, as the Papuan one does )
Odontomasoreus humeralis (Part III, 76): a subspecies at Dobodura (Papua) and another in N central N. G.
Minuthodes sexiialis (Part III, 98) : a subspecies^in Papua, anotlier widely distributed in the west of N. G.
Dolichoctis divisa and huon (Part III, 131): a pair of apparently closely related species occuiTing in Papua and N N-E N. G. respectively
Anomotarus ornatus and fuscipes (Part III, 190): ornatus occurs in the Moluccas and the western part of New Guinea east of Hollandia, fuscipes, in eastern New Guinea west to Hollandia. A species pair, which I should call subspecies expect that they overlap in Hollandia.
have quickly spread through the \\'hole length of it, and often to the comer of Austraha, too. That the Mokiccas tend to be a bottleneck in dispersal, and that the Carabidae that reach New Guinea across them tend to be relatively dominant, is suggested elsewhere [84].)
Another set of geographic patterns is
foiTned by x\ustralian species which reach only the southern edge of New Guinea, some of them perhaps only open eucahp- tus woodland and other "Australian" habi- tats, or \\'hich, although somewhat more widely distributed, reach only the eastern part of the island. These species are listed in Table 11. Besides the actual Australian
214 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
species that occur in southern or eastern New Guinea, several Australian genera have endemic species apparently confined to these parts of New Guinea (see also Table 11). (Tliis pattern, of course, sug- gests that the species in question, or their immediate ancestors, have reached New Guinea more or less recently from Aus- tralia, some of them over the land con- nection that existed at times in the Pleistocene [17].)
Still other sets of geographic patterns are formed by localization and differenti- ation of Carabidae at low altitudes in different parts of New Guinea. Because collecting of Carabidae has been both in- adequate and uneven, geographic patterns cannot yet be fully defined for most spe- cies. However, in 20 cases (listed in Table 12) lowland Carabidae have fonued sub- species or species pairs in different parts of the island. In four cases subspecies occur in the Vogelkop, with different sub- species at one or more localities elsewhere in New Guinea. In 13 cases subspecies occur in Papua, with different subspecies elsewhere in New Guinea. And in seven cases divisions between subspecies or species pairs are somewhere in the middle part of the island. (The sum of these cases is greater than the number of species and species groups listed in Table 12 because several species have divided into more than two subspecies.) It should be em- phasized that even these cases are still incompletely known. That relatively few subspecies have been distinguished on the Vogelkop may be because relatively little collecting has been done there. Allowing for this inequality of collecting, I cannot see that subspecies have tended to form in significantly greater numbers in any one part of New Guinea than in other parts. The geographic patterns of subspeci- ation are more complex in some cases than Table 12 shows. Also, it should be em- phasized that many species of Carabidae are apparently distributed over most or all of New Guinea (in appropriate habi-
tats, of course) without obvious local differentiation.
Because different lowland areas have been unequally collected, I see no use list- ing endemic species kno\vn from different parts of New Guinea. Every part of New Guinea seems to have at least a few local- ized species of its own, in spite of the fact that many other species, probably a great majority of the lowland ones, are widely distributed on the island. However, a table of the distributions of endemic genera may be more significant (Table 13). (The table includes a few genera that are not strictly endemic but have been found to occur also in limited areas outside New Guinea.) This table shows a concentration of endemic genera mainly in the mountains and (at lower altitudes) toward the east- ern end of the island. Of ten genera en- demic to New Guinea and represented in the lowlands, only two are known to reach the Vogelkop and none is peculiar to the Vogelkop. It remains to be seen whether this apparent situation is real, or whether it is a reflection of the relatively small amount of collecting that has been done at the western end of the island. (It mipjit be due to continual invasions of western New Guinea by new Oriental stocks and a resulting tendency of older endemic stocks to survive toward the eastern end of the island.)
In summary of geographic patterns of Carabidae at low altitudes within New Guinea, it can be said that some species and species groups that are widely dis- tributed outside the island reach only the western or only the southern or eastern parts of New Guinea (suggesting more or less recent arrivals from the Oriental area or from Australia). Differentiation and localization of species and even of some genera in different parts of New Guinea form other patterns. Endemic genera and species may be concentrated toward the eastern end of the island, but othenvise no special centers of speciation and no specially important barriers to dispersal at
The carabid beetles of New Guinea • Darlington 215
Table 13. Distributions, relationships, and ecology of "endemic" (see text) genera
( Cicindelinae )
Caledonomorpha (Part I, p. 336): 2 species confined to eastern New Guinea ("the bird's tail"), relationships probably Australian; terrestrial or subarboreal
(Carabidae proper, otlier than Agonini)
Perileptodes (Part I, p. 489): 2 species, 1 or both distributed throughout New Guinea including the Vogelkop (and 1 reaching the Solomons); relationships witli (primarily Oriental) Perileptus; hy- drophiles, by streams
Rhtjtiferonia (Part I, p. 533): 2 species, both at high altitudes on the Snow Mts.; relationships probably Australian; probably mesophiles
Analoma (Paraloma) (Part I, p. 538; present part. Tax. sitppl): 4 species, at high altitudes only; relationships probably Oriental; probably mesophiles
Haploferonia (Part I, p. 547): 1 species, known from 1 specimen from 750 m. West N. G.; relation- ships probably with (Australian) Loxandrus; probably mesophile
HomaJonesioia (present part. Tax. suppl.) : 2 species, of which 1 presumably from lowlands c. 80 km west of Hollandia, the otlier widely distributed at moderate altitudes; relationships probably with (Australian) Loxandnis; probably hydrophiles, by streams
Nebrioferonia (present part. Tax. suppl. ) : 1 species, widely distributed, but not yet known on the Vogel- kop; relationships with preceding; hydrophile, by streams
Tiferonia (Part I, p. 563): 1 species, eastern New Guinea west to Hollandia; relationships doubtful (an apparent relative in the Philippines); hydrophile, swamps
Lyter (Part III, p. 63): 1 species, widely distributed in New Guinea; relationships probably with (Oriental) Trichotichnus; mesophile
Odontomasoreus (Part III, p. 76): 1 species (2 subspecies), in eastern half of New Guinea; relation- ships probably Oriental; mesophile
Minuphloeus (Part III, p. 117): 1 species, widely distributed at moderate altitudes; relationships un- determined; ecology unrecorded (perhaps arboreal, on tree trunks )
Dobodura (Part III, p. 215): 1 species, widely distributed at low altitudes; relationships probably with (Austrahan) Eudalia; hydrophile, by streams
HeUuopapiia (Part III, p. 232; present part. Tax. suppl): 2 species, at moderate altitudes in West N. G.; relationships with (Australian) He//uom'rfms; probably mesophiles ( possibly on tree trunks )
(Agonini, probably not derived from Notagonum/Colpodes ancestors)
Tarsagonum (Part II, p. 127; present part. Tax. suppl): 1 species, Papua and N-E N. G. (and a species from Borneo has now been assigned to tliis genus); relationships Oriental; mesophile
Idiagonum (Part II, p. 229; present part. Tax. suppl): 6 species, at high altitudes on several mountain ranges; relationships Oriental ( probably derived independently of other New Guinean agonines); mesophiles
(Agonini, probably derived on New Guinea from Notagonum/Colpodes ancestors)
Lithagonum (Part II, p. 176): 1 species (several subspecies), probably throughout New Guinea but not yet actually recorded from the Vogelkop; probably related to otlier New Guinean agonines ( which originally derived from Orient ) ; hydrophile, by streams
/nV/agonum (Part II, p. 181; present part. Tax. supp/.) : 7 species, at low and moderate altitudes throughout New Guinea including the Vogelkop; probably related to otlier New Guinean agonines (which originally from Oriental stock); mesophiles
Eight other apparently endemic agonine genera (Part II; present part, Tax. suppl): total of 59 species all confined to the mountains; all probably derived from the (nonendemic) "genera of con- venience" Notagonum, Colpodes, and Altagomim (which have diversified on New Guinea, but which are all probably derived from Oriental stocks); mostly mesophiles, but Potamagonum and some species of Nebriagonum hydrophiles, by streams, and 1 Nebriagonum and perhaps some species of other genera (especially Maculagonum?) subarboreal
216 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
low altitudes within the island are indi- cated. (Apparently many lowland Carab- idae have dispersed freely within the limits of the island. Their dispersals and later differentiations have formed a num- ber of different patterns, l^ut no one pat- tern is uniquely important. Rather, all the dispersals and differentiations together ha\'e formed a complex network of geo- graphic patterns over the lowlands of New Guinea as a whole, and the network is not divided into major parts by well-defined barriers. )
At higher altitudes, every principal mountain range has been a center of dif- ferentiation of species and sometimes of genera, and deep valleys between high ranges have been important barriers. This is surely time. However, so little collect- ing has been done at high altitudes, and collecting on different mountain ranges has been so unequal, that patterns of dis- tribution of Carabidae at high altitudes cannot yet be usefully discussed.
ZOOGEOGRAPHY: DISPERSALS AND GEOGRAPHIC ORIGINS
I shall now turn from description of existing geographic patterns to consider- ation of the movements — dispersals — that have made the patterns. Existing patterns are complex enough and difficult enough to describe fairly (see [31]), and the movements that have made the patterns are even more complex and difficult to reconstruct without prejudice. I suppose that all writers on this subject do have prejudices. One of mine is, in the absence of conclusive evidence, to prefer dispersals from large to small areas, in this case from Asia to Australia, rather than the reverse. Having acknowledged this prejudice, I shall try to guard against it.
[67] The apparent main pattern of dis- persal. All faunas are derived, geographi- cally. Nowhere in the world is there an ex- isting fauna that cannot be accounted for in terms of derivations from other parts, if one goes far enougli liack in evolutionary
and geographic history. Animals have in fact formed a continually changing pattern — a zoogeographic kaleidoscope — of evo- lutions, dispersals, retreats, and extinc- tions, proceeding in many different directions over the world as a whole. Every part of the world is continually invaded or lialjle to invasion by animals from other parts, and every part of the world has some chance of originating groups that disperse to other parts. Tlie details are almost inconceivably complex. Nevertheless, I think the kaleidoscope has a main pattern. Dominant groups apparently evolve most often in certain places, in the largest and most favorable areas, and most often dis- perse in certain directions, into smaller and less fa\'orable areas. I have discussed this fundamental pattern several times, most thoroughly in 1959, and most recently in a careful summary in 1965, Chapter 5, on "Area, climate, number of species, evo- lution, and dispersal." The main area of the Old World tropics seems to me to have been the greatest of all evolution- dispersal centers, from which successive dominant stocks have spread over the world. Australia has been another center, although (from a worldwide point of view) much less important than the main Old- World-tropical one. Australian animals have evolved diversely, and many have dispersed to surroimding islands, but very few of them have gone far.
[68] Dominance, competition, and ex- tinction in dispersal; faanal overturn. The geographic history of animals seems to me to have been primarily the history of dis- persals of successive dominant groups.
Dominant animals are conspicuously suc- cessful ones. Dominant groups are rel- ativelv numerous in individuals, often numerous in species, often diverse in adaptations, and often widely and con- tinuously distributed in more or less diverse habitats. Examples among Carabidae are (within their ecologic limits) CUvina among fossorial forms, Tachijs (in the tropics ) and small Harpalini among hydro-
The carabid beetles of New Guinea • Darlington 217
philes, some Agonini in a variety of habitats (except very dry ones), some medium-sized Harpalini on the ground, and some Lebiini in arboreal liabitats (especially in the tropics). Dominance presumably reflects underlying qualities — "general adaptations" (Darlington, 1948: 109; 1957: 565; Brown, 1958)— that make for initial success and that lead to evo- lution of numerous, varied, successful types.
Dominance implies success in competi- tion, competition being defined as any interaction among organisms that is or may be disadvantageous to any of them. Al- though competition is difficult to demon- strate in particular cases, the general evidences of it in the animal world are overwhelming. The strongest evidence comes from the general level and balance of faunas in all parts of the world (Dar- lington, 1957: 552ff). For example, every habitable part of the world has a carabid fauna roughly proportional to area and climate, and the Carabidae in each part show a reasonable range of size and in- clude representatives of all the principal ecologic groups for which habitats are available. No substantial part of the world is overfull of Carabidae, and no part has a notable deficiency of them. This balance cannot be due to chance. Something must hold the size and composition of carabid faunas everywhere within certain lintits in spite of continual multiplications and dis- persals of successive phylogenetic groups. Only competition can do this, and to do it competition must be a fundamental, omni- present force, resulting in continual extinc- tion of undominant stocks as dominant ones spread and diversify.
I have reviewed this subject here be- cause the role of dominance and compe- tition in spreading, recession, and extinc- tion must be understood if the nature and history of the existing carabid fauna of New Guinea are to be understood. ' [69] Complexity of dispersal. Dispersals jmust often be almost inconceivably com-
plex. The dispersals of most tribes and of many genera are the sums of the dispersals of many species, and the dispersals of spe- cies are the sums of movements of multi- tudes of individuals. Moreover, extinctions (withdrawals) as well as spreadings must be important in the dispersal histories of many groups of animals. We know from their unique fossil record that mammals have had a very complex dispersal history, with successive dominant groups spreading over much or all of the world, the spread of new major groups being accompanied by localizations and extinctions of pre- viously dominant groups. The Carabidae have left virtually no fossil record, but they are probably more diverse taxonomi- cally, older, and (because most of them are winged) more rapid in their dispersals than the mammals are, and their geo- graphic history may have been even more complex than the history of mammals, with more successive dispersals and more ex- tinctions. Along any one line of dispersal, many groups of Carabidae are likely to have spread and many others to have "retreated" (become extinct in parts of their ranges), and both processes are likely to have occurred in both directions in different groups along the line of dispersal. Nevertheless net changes of distributions amounting to directional movements of whole faunas may have oc- curred in the course of time. We want now to find whether there has been a net direction in the movements of Carabidae in the Asiatic-Australian area, whether the direction can be detected by analysis of carabid distributions, and how New Guin- ean Carabidae fit into the dispersal pattern. [70] Place of Neiv Guinea in the main dispersal pattern. The vertebrates of New Guinea are mostly Australian in their closest relationships, while the Carabidae are more Oriental than Australian, but the vertebrates and carabids do share one sig- nificant characteristic: New Guinea has not been a major dispersal center for either of them. Groups that seem to have evolved
218 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
and diversified primarily on the island have usually not spread far. For example, the birds of paradise, which have diversi- fied extraordinarily on New Guinea, are represented (in very small numbers) west- ward only to the Moluccas and southward only to northeastern Australia. Among mammals, murid rodents of the subfamily Hydromyinae seem to have diversified pri- marily on New Guinea; one genus has spread over Australia (and to New Britain) and another genus is localized in northeastern Australia, but New Guinean hydromyines do not extend westward, although a few possibly related fonns are (relict?) in the Philippines.
Among Carabidae, too, very few groups have patterns of distribution that suggest evolution in and spread from New Guinea itself. No tribe of Carabidae is peculiar to or centered on the island. Of genera, the foliage-living section of Demetrida (Part III, pp. 140ff) does seem to have evolved or at least diversified in New Guinea, but it has dispersed (in very small numbers) westward only to the Moluccas and south- ward only to northern Australia; the total distribution of these insects is very much like that of the birds of paradise (above). Minuthodes (Part III, pp. 95ff) may be primarily New Guinean, but if so it has dispersed only as far as Celebes, New Britain, and northern Australia. Dolichoc- tis of the aculeata and (derived?) polita gi-oups (Part III, pp. 128ff), with ten species on New Guinea, reach only to Celebes, New Britain and New Ireland, the Solomons, and northern Australia. Dicra.speda (Part III, pp. 210ff) has six diverse New Guinean species, one of which extends to Java and the Philippines and to northern Australia (another reaches the tip of Cape York); this is, I think, the widest dispersal of any primarily Nev/ Guinean group of Carabidae. These are the principal groups of Carabidae that are distributed as if they have originated in and spread from New Guinea. The .sena group of Tachijs (Part I, pp. 404ff), with
three distinct species on New Guinea, and with outlying fonns reaching the Moluccas and New Britain (not Australia), is an example of a smaller group distributed as if it has diversified in New Guinea and dispersed for short distances beyond the limits of the island.
Of course some Carabidae have differ- entiated and radiated on New Guinea (see especially [91, 92]), but the island has been much less important as an evolution- ary center tlian the continents have been and has not been a major dispersal center for these insects. Most carabids that have reached the island have apparently come in from Asia or Australia, or have dispersed across New Guinea from one continent to the other. The question now is, what has been the net direction of dispersal of these insects between Asia, New Guinea, and Australia?
[71] Directions of dispersal. The sums of nondirectional geographic units at the bottom of Table 8 [58] give a ratio of 173:120 Oriental to Australian relation- ships for New Guinean Carabidae, and this at least suggests more movement from the Orient to New Guinea than from Aus- tralia to New Guinea. However, this situation mi'^ht be explained in other ways, for example by movements of New Guin- ean stocks to the Orient or by extinctions of New Guinean stocks in Australia, and the Australian relationships of considerable numbers of New Guinean Carabidae mipjit be the result of movements either from Australia to New Guinea or from New Guinea to Australia. It is therefore im- portant to look for other evidences of direction of movement to confirm (or con- tradict) the generalization that Table 8 suggests.
[72] Direction and vaii,iUttj. Correlation . of relative dispersal ability with geographic relationships might give evidence of di- rectional movement. The Bembidiini and Perigonini of New Guinea total 31 Oriental to ten Australian geographic units, and the arboreal Agonini and primarily arboreal
The carabid beetles of New Guinea • Darlington 219
Lebiini total 35 Oriental to 16 Australian geographic units; the sums for all these groups together are 66 Oriental to 26 Aus- tralian units. If these figures are subtracted from the grand totals of 173 and 120 (Table 8), the sums of geographic units for all other New Guinean Carabidae are found to be 107 Oriental to 94 Australian. The Bembidiini (principally Tochys, see [20] and Fig. 8) and Perigonini (Ferigona) include most of the very small Carabidae in New Guinea and, because of their small size, they probably disperse through the air more rapidly than larger Carabidae. And the arboreal agonines and lebiines are active and fly more readily than most other Carabidae, and they too are likely to dis- perse rapidly. The geographic relationships of these groups show that they have in fact made multiple dispersals in the Asiatic- Australian area. If there is a net direction in dispersal, the more rapidly dispersing groups might be expected to run ahead of the more slowly moving groups and thus to show net direction of movement, and the groups just specified do in fact show sig- nificantly more Oriental and fewer Austra- lian relationships than other New Guinean Carabidae do, suggesting that the main direction of movement has been from the Orient toward New Guinea and Australia. This is, however, only a tentative con- clusion, which depends on (among other things) the assumptions that there is a net direction of movement and that the whole situation is relatively simple, not distorted by massive extinctions.
[73] Cani])id versus mommalian dis- persals. This same method, of comparing more with less vagile groups, can be ex- tended by comparing the patterns of dis- tribution of Carabidae and of mammals. While the Carabidae of New^ Guinea are (roughly) three parts Oriental to two parts Australian in present relationships, the mammals of New Guinea are overwhelm- ingly Australian, excepting the bats (Dar- lington, 1957: 3.35, Table 8; Keast, 1968). Of flightless land mammals, New Guinea,
like Australia, possesses only a few mono- tremes, many marsupials, and rodents of the family Muridae, and even the murids are (in terms of present relationships) much more Australian than Asiatic. Among these New Guinean mammals, degree of relationship wdth Asiatic mammals is cor- related with power of dispersal: the mono- tremes and marsupials have no existing Asiatic relatives; the rodents, which cross water barriers more often than other ter- restrial mammals (as we know from situ- ations in many parts of the world), do show Asiatic ties, some older and some more recent; and the bats, which obviously have still greater power of crossing water barrriers, are still more Asiatic and less differentiated in New Guinea and Austra- lia. In this case, we know beyond reason- able doubt that six or more stocks of murid rodents (Simpson, 1961) and many stocks of bats have in fact dispersed from Asia toward Australia, and the closeness of relationships of the New Guinean (and AustraHan) to Asiatic forms are roughly in proportion to the powers of dispersal of the different groups. This is the same pattern that has been found [72] in comparing the more with the less readily dispersed groups of New Guinean Carabidae, and the ex- planation is probably the same, that the carabids too have dispersed mainly from Asia toward New Guinea and Australia, the most actively dispersing groups show- ing their Asiatic relationships and origins most strongly.
The pattern of relationships of New Guinean Carabidae can be compared di- rectly with the mammalian pattern. The priinarily winged carabids surely disperse across barriers more easily than terrestiial mammals do. New Guinean Carabidae are much more Oriental and less Australian in their relationships than New Guinean mammals are, and I think this is evidence both that the Carabidae have dispersed more recently and that the direction of their dispersal has been mainly (but of
220 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
course not exclusively) from Asia toward New Guinea and Australia.
Another conceivable explanation of their different geographic patterns in the Asiatic- Australian area is that Carabidae are older than mammals; that existing carabids dis- persed before the mammals did, while New Guinea and Australia were connected by land with Asia; and that the land con- nections were broken before the mammals radiated in the Australian Region. I think, however, that the facts are decisively against this possibility. I recognize 104 actual species of Carabidae common to New Guinea and the Oriental area. Further taxonomic work may show that in some of these cases the New Guinean and Oriental populations are distinguishable, but there can be no serious doubt that many species of Carabidae do now range from the Orient to New Guinea, some species even from the mainland of Asia to Australia. Granted that rates of evolution vary and that speciation may have occurred rela- tively slowly in some Carabidae, it is never- theless unlikely that so many species have not only remained identical in the Orient and New Guinea but have also maintained their geographic ranges since the beginning of the Tertiary, which (as the fossil record of mammals shows) is the latest there can have been a complete land connection be- tween Asia and the Australian Region, if there ever was a connection. I therefore think that the general pattern of relation- ships of different groups of Carabidae, and comparison with the pattern of distribution of mammals, does give strong indications that the carabids have dispersed relatively recently and mainly from Asia toward New Guinea and Australia.
[74] Directimvs of disper.sal of tribes, fl^enera, and species. Besides the general indications of direction derived from the nature and relationships of the fauna as a whole, from comparison of some more- vagile groups of Carabidae with less-vagile ones, and from comparison of the Carabi- dae with the mammals (above), more
specific evidence of direction can be found in the patterns of distribution of some tribes, genera, and species of Carabidae. This evidence is of a different sort from that analyzed above and sometimes contra- dicts it, or rather clarifies it, for present closest relationships (considered above) show only the most recent in what may have been a complex series of dispersals. For example, the closest relationships of the rodents of New Guinea are now with Australian rodents, but the pattern of dis- tribution of the whole family Muridae and also the fossil record show that the an- cestors of New Guinean and Australian murids all came from Asia in the first place. For another example, Chlaenius in New Guinea now includes nine Oriental to six Australian "geographic units" of present relationships, but the pattern of distri- bution of the genus as a whole and also the distributions of several of the species ( Part III, pp. 20ff) suggest multiple movements from Asia toward Australia. In this case the failure of some "geographic units" to show direction decisively seems to be due to the effectiveness of dispersal, for several stocks seem to have dispersed from Asia across New Guinea and into Australia, so that they count as both Oriental and Aus- tralian in tenns of present distribution even though dispersal may have occurred in only one direction. On the other hand. New Guinean Licinini include four Oriental and five Australian geographic units, but the tribe is now much more Australian tham Oriental in its distribution, and it seemS' likely that movements from Australia across' New Guinea into the edge of the "Oriental area" have been relatively more important than the geographic units show.
Criteria for detennining probable di- rections of dispersal of specific groups have been discussed by me in 1957, pages 31- 35. In the absence of a significant fossil record, the best clues to directions of movement are based on relative numbers and relative areas: if a genus is represented by (say) 50 species on one continent and
The carabid beetles of New Guinea • Darlington 221
one species on another, dispersal is likely to have been from the first to the second continent, and if a species occurs over the whole of one continent and only on the comer of another, dispersal is again likely to have been from the first to the second continent. But clues like these must be used judiciously. They are likely to be most trustworthy in dominant (conspicu- ously successful) groups, in which dis- tributions are still continuous, and in which there are definite clines of numbers in given directions, or great inequalities in areas occupied on different continents, for these groups are most likely to have dis- persed recently or to be dispersing now, >o that their distribution patterns really do reflect dispersals rather than local radi- ations, extinctions, and other complications. Carabidae are primarily winged insects vvhich do fly and can disperse rapidly, in ^pite of the fact that some have lost the power of flight. Their movements over the world and between adjacent continents have probably been very numerous and very complex. And, to judge from what has happened among other animals of which we have better fossil records (espe- cially the mammals), the multiple and successive dispersals of dominant carabid groups have probably been accompanied by frequent and widespread extinctions of ather groups. Therefore, I do not trust numbers clues" or "area clues" to show directions of movement of nondominant, discontinuously distributed groups, of vvhich the present distributions may be the result of withdrawals ( partial extinctions ) rather than of initial dispersals (cf. Lox- indrus, [82]). (Zoogeographic tracking is like tracking in snow in that the tracker 3an follow with the most confidence the clearest, most recent trails rather than )lder, partly obliterated ones.)
With these criteria and their limitations n mind, I want now to survey the tribes )f Carabidae that are represented in New Guinea and to try to pick out the groups "hat clearlv show evidences of direction of
dispersal. Evidence might come from the distributions of tribes as wholes, or of genera, or of species. In any single case, no matter how clear the evidence seems to be, the conclusions should be considered tentative. But if, of many separate cases, most seem to show dispersal in the same direction, the probability that dispersal has had a net direction will become strong. However, no matter how great the pre- ponderance of movements in one direction, some counter-movements are to be ex- pected too; this seems always to be the case in complex faunal movements.
The groups of Carabidae that seem to show directions of dispersal are listed in Table 14. The table is derived mostly from the preceding survey of tribes (items [33- 57] ) . Most of the details that seem to me to indicate direction have been given in this survey (or under the groups con- cerned in Parts I-III and the present Taxonomic supplement) and will usually not be repeated here. Evidence of di- rection of movement may, of course, be derived not only from the distribution of a given genus or species but also from the occurrence of related fomis — whether they are Oriental or Australian. Tlae following abbreviations are used: Or, Oriental area; NG, New Guinea; Au, Australia beyond Cape York; CY Cape York; Mol, Moluccas. "Or to NG to Au" means that dispersal has apparently been from the Oriental area to New Guinea to AustraHa. In this table I have indicated movements from New Guinea to Cape York or from New- Guinea to the Moluccas only when they are con- tinuations of longer moxements from the Orient or Australia respectively. To in- clude other short-range movements would weight the results in favor of movements to\\'ard Australia, because the Carabidae of Cape York are better known than those of the Moluccas.
[75] Summary of direetion to this point. The 128 groups (tribes, genera, and spe- cies) of Carabidae listed in Table 14, that seem to show relatively clear evidences of
222 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Table 14. Groups of New Guinean Carabidae that seem to show directions of dispersal
In tribe Ozaenini: Pseudozaena orientalis with siibsp. opaca. Or to NG.
In tribe Paussini: Arthropteriis, An to NG.
In tribe Scaritini: Geoscaptus cacus, An to NG; Syleter, Or to NG to CY; Clivina zebi. Or to NG to CY. ( Other Clivina in New Guinea are in part Austrahan- and in part Oriental-related, but I think they are not w ell enough known to justify deductions about their directions of dispersal. )
In tribe Bembidiini: Tachijs as whole, mostly Or to NG to Au; T. fasciattis. Or to NG to An; T. cetjlanicus. Or to NG; T. lilugi, Or to NG; T. convexus, Au to NG; T. fumicaius. Or to NG (closely related curticollis may not have moved Au to NG but may represent a first invasion from Or to NG to Au of an Or stock that later reinvaded NG as fiimicatus); T. umhrosus. Or to NG; T. coracinus. Or to NG; (some other Tachijs have distributions that suggest dispersal either from Or to NG or from Au to NG, but the evidence seems less clear than in the cases cited); Limnastus atricapillus. Or to NG.
In tribe Trechini: Perileptus, Or to NG to Au.
In tribe Panagaeini: Peronomerns xanthopus. Or to NG; Dischissus notulatiis. Or to NG; Craspedo- phorus gressittorum, Au to NG.
In tribe Pterostichini: Morion of Or group, Or to NG to Au; (Mecyclothorax, see [80]); Brachi- dius crassicornis. Or to NG; Caelostomus (excluding picipes). Or to NG to GY; Ahacetus haplostermis. Or to NG to An; A. convexiusculiis. Or to NG; Lesticus, Or to NG to An; Prosopogmus, Au to NG to Mol; Platycoelus, Au to NG to Mol; Loxandrus, Au to NG to Or (Celebes) (see [82]).
In tribe Agonini: tribe as a whole, mainly Or to NG to An; Euplenes, Or to NG; Dicranoncus ■ queenslandicus. Or to NG to Au; Lorostemma, Or to NG; Agomim (Sericoda) cetjlaniciim. Or to NG [ (see [80]); Notagonum dentellum, NG to An; N. suhmetaUiciim, Au to NG; Violagomim violaceum, NG to Au; Colpodes sapphyrinus with subsp. sloanei. Or to NG; C. habilis, NG to Au; (directions of. dispersal of some other Agonini are indicated, but the evidence seems less clear tlian in the cases cited ) .
In tribe Perigonini: Perigona as whole. Or to NG to An; P. plagiata. Or to NG.
In tribe Licinini: Bodister sundaiciis. Or to NG to Au; PhysolaestJins, Au to NG to Or (Java, Philip- pines); Dichrochile, Au to NG; (Microferonia, see [80]).
In tribe Chlaeniini: Chlaenius ceylanicus. Or to NG to Au; C. tnacuUger, NG to Au; C. gtittula. Or to NG; C amplipennis. Or to NG; C. himacidatus group. Or to NG; (directions of dispersal of. some other Chlaenius are indicated, but the evidence is less clear than in the cases cited ) . j
In the tribe Harpalini: Gnathaphaniis as whole, Au to NG to Or (SE Asia); G. licinoides, Au to; NG; G. upolensis, Au to NG to Or (to Malay Pen.); G. picipes, Au to NG; G. pulcher, Au to NG; Dia- phorurnerits, 2 stocks, An to NG to Mol, and Au to NG; Hypharpax, Au to NG to Or (Java, Sumatra); Lecanomerus, Au to NG; Chydaeus, Or to NG (see [80]); Platymetopus, Or to NG; Trichotichnus, 3 stocks. Or to NG; Coleolistis, Or to NG to CY; Egadroma, 3 stocks Or to NG to Au, and 1 stock Or to NG; Stenolophtis, 2 stocks Or to NG to An; Acupalpus, 3 stocks Or to NG to Au.
In tribe Cyclosomini: Anaulacus siamensis. Or to NG; Caphora litimilis. Or to NG to CY; Saro-\ throcrepis, Au to NG to Or (Java, Philippines).
In tribe Lebiini: Aristolebia, Or to NG to CY; Physodera, Or to NG; Lebia, Or to NG to Au; Holcoderus, Or to NG to Au; Minuthodes, NG to Au, NG to Or (Celebes); Catascopiis as whole. Or to NG to Au; C. facialis. Or to NG; C. elegans. Or to NG to Au; C. smaragdulus. Or to NG to CY; Pericalus, Or to NG; Coptodera as whole. Or to NG to Au; C. eluia. Or to NG; Agonochila, Au to NG; Mochtherus, Or to NG; Dolichoctis as whole. Or to NG to An; D. striata. Or to NG to An; D. aculcata, NG to Or (Celebes); Stricklandia, NG to An; PeUocypas, Or to NG; Syntorntis, Or to NG to An; Apris-i tus. Or to NG; (I have not counted Anchista binotata dnd Eiulynomena pradieri, hoihT^rohahXy curried by man from Or to NG); Demetrida, Au to NG to Mol (but see [80]); Phloeocarabus, Au to NG; Trigonothops, Au to NG; Nototarus, Au to NG.
The carabid beetles of New Guinea • Darlington
223
Table 14. Continued
In tribe Pentagonicini: Pentagonica as whole, Or to NG to Au; P. pallipes, Or to NG to CY; P. hlanda. Or to NG to Au; P. crichsoni, Or to NG to CY; P. rtificoUis, Or to NG to Au; (Scopodes see
p. 228).
In tribe Hexagoniini: Hexagonia, Or to NG to Au.
In tribe Odacanthini: CoIIiuris, Or to NG to Au; C. fuscipennis. Or to NG; Clarencia, Au to NG; Dicraspeda, NG to Au, NG to Or ( Java, Philippines ) ; Lachnothorax, Or to NG.
In tribe Dryptini: Dnjpta mastersi, Au to NG.
In tribe Zuphiini: Planetes, Or to NG to Au.
In tribe Helluodini: Pogonoglossits, NG to Au.
In tribe Helluonini: Creagnis, Or to NG to Au; HeJhiosoma, Au to NG; Helliiodema, Au to NG; Gigadema, Au to NG.
In tribe Brachinini: Pheropsophus, Or to NG to Au; Brachinus, Or to NG.
In tribe Pseudomorphini: Adelotopus, 1 stock Au to NG to Or (Java) and 2 stocks Au to NG; Sphcdlomorpha, 3 stocks Au to NG; (Cryptocephalomorpha occurs from NG to the Malay Pen. and Thailand, but its direction of dispersal is not clear).
directions of dispersal, are classified and totaled in Table 15. The grand totals are 89 groups that seem to have dispersed southeastward, including 44 groups that seem to have spread all the way from some part of the Oriental area to some part of Australia (including Cape York), and 39 groups that seem to have dispersed north- westward, including seven groups that seem to have spread all the way from Australia to some part of the Oriental area. It should be repeated and stressed that the groups considered to show clear evi- dences of directions of dispersal have, necessarily, been selected somewhat arbi- trarily. The totals would vary to some extent with judgments about which groups really show direction clearly. But on any reason- able basis of selection the tribes, genera, and species that seem to have dispersed from the Oriental area to New Guinea and Australia far outnumber those that seem to have dispersed from Australia to New Guinea and the Orient. This disparity is greatest among the groups that have moved the longest distances. Many pri- marily Asiatic or African-Asiatic genera include species that extend eastward across the Malay Archipelago to the mainland of Australia. But very few primarily Austra- lian groups include species that extend
across the islands to the mainland of Asia; in fact Gnathaphanus, with one common Australian species reaching the Malay Peninsula and another reaching India etc., is the only carabid genus that seems to show this reverse patteni clearly. (Tlie
Table 15. Summary of apparent directional
DISPERSALS OF NeW GuINEAN CaRABIDAE
Or to NG 37
Or to NG to CY 9
Or to NG to Au 35
NG to Au 8
SE ■ movements 89
Au to NG 25
Au to NG to Moluccas 4
Au to NG to Or 7
only to Celebes, 1 ( Loxandrus ) to Java, 1 (Adelofopus)
to Java and Philippines, 2 (Physolaesthus,
Sawthrocrepis) to Sumatra, 1 (Hypharpax) to Malay Pen., 1 (G. upolensis) to India, etc., 1 (Gnaihaphanus)
NG to Or 3
only to Celebes, 2 {Minuthodes, Dolichoctis
aculeata group) to Java, 1 ( Dicraspeda )
NW movements 39
224 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
mainly Australian tribe Pseudomorphini reaches the southeastern corner of Asia, but the only genus that does so (Crypto- cephalomorpha) is not Australian and its geographic history is doubtful. ) This whole situation suggests not only that move- ments from Asia toward Australia are much more numerous than the reverse, but also that the reverse movements are usually shorter.
[76] Direction and doniinance. Thus far, I have been considering directions of dis- persal in the whole carabid fauna without distinguishing different fractions of it, ex- cept that I have compared certain relatively vagile fonns (especially very small ones) with the rest of the fauna. Now, I want to compare the apparent directions of dis- persal of different faunal fractions to see if the comparisons will give further in- formation about the origins and history of New Guinean Carabidae.
Among the Carabidae of the Asiatic- Australian area, the genera Chlaenius (Part III, pp. 20ff) and Egadroma (Part III, pp. 69ff) are notably dominant. Each genus includes several species that are very widely distributed, ranging from Asia to Australia, and individuals of some of the species are numerous. The distribution patterns of these two genera have note- worthy characteristics in common. In both genera, the relative numbers and diversity of species on different continents and the relative areas occupied seem clearly to indicate origin in the main part of the Old World and multiple dispersals toward and into AustraHa (Chlaenim, into the Americas too ) . Some species in each genus are now widely distributed in Australia as well as in the Orient. Considered singly, these species scarcely show the directions of their dispersals, although the geographic patterns of the genera as wholes do clearly indicate direction. These and most other really dominant Carabidae, for example Tachys fa.sciatus (Part I, p. 414), seem to have dispersed from Asia to Australia. This fact is significant because the most domi-
nant members of a fauna are most likely to have dispersed recently and to be dis- persing now, and to show the main direction of dispersal of the whole fauna, if there is a main direction.
Several of the most widely distributed and commonest species of these genera, e.g,., Chlaenius jlamguttatus and E<;adro7na smaraiidida, occur in a rather wide variety of wet places, although some other species of both genera have moved into drier haliitats. These are examples of what I think is a fact, that ground-living carabids associated with water ( hydrophilcs ) dis- perse and also cross climatic barriers more easilv than most other Carabidae (see [84] j.
[77] Direction and size. Although small Carabidae disperse more easily than large ones, direction of dispersal is apparently not correlated with size of insects. The very small Carabidae that have reached New Guinea have probably dispersed more rapidly than most larger ones and may show stronger Oriental relationships for this reason [72], but the larger Carabidae too (including for example Lesticiis, Chlaenius, large Colpodes, and Cata- scopu.s) seem to be more Oriental than Australian in derivation. The two largest Carabidae in New Guinea are both Aus- tralian species, but one {Catadromus tene- hroides. Part I, p. 563) is probably man- carried and is probalily not established on the island, and the other ( Gigadenia tnaxil- lare. Part III, p. 234) is known from just the southern edge of New Guinea, and is a unique case and not an integral member of the New Guinean fauna.
[78] Direction and it//i.i^.S'. Although winged, flying Carabidae disperse more easily than flightless ones, direction of dis- persal is not strongly correlated with presence or absence of wings in the present case. Almost all New Guinean Carabidae, whatever the present state of their wings, are or may be derived from ancestors that were winged when they reached the island [84, 88], regardless of the direction from
The carabid beetles of New Guinea • Darlington
225
which they came. The only sure exception is the fhghtless tiger beetle Triconclyla, which came from the Orient, probably by rafting (cf. [21], footnote). The three other carabids most likely to have reached New Guinea without flying are all Aus- tralian, as would be expected from the fact that New Guinea was connected with Australia not long ago [17]. Tliey are species of Craspedophoriis ( Tax. Stippl. under Panagaeini), Coptocarpus (Tax. siippl., footnote under Oodini), and Noto- tarus (Part III, p. 185). These genera are all represented at low altitudes in tropical Australia (although the single individuals of Craspcdophonis and Coptocarpus thus far found in New Guinea were in the mountains) and all are now wholly flight- less in Australia as well as in New Guinea. They may have reached New Guinea with- out flying. However, their ancestors were prt)bably winged not long ago and may have dispersed partly by flight.
[79] Direction and ecoJogij. Direction of dispersal is correlated with ecology to only a limited extent. The principal eco- logic groups of New Guinean Carabidae (mesophiles, hydrophiles, and arboreal forms) are all more Oriental than Austra- lian in relationships and in probable origins. However, the mesophiles divide into two subgroups: those (more numerous) that live on the floor of rain forest are mainly Oriental; those (fewer) that live in open country including open eucalyptus woodland are mainly Australian. The latter, the mesophiles that live in relatively open, relatively dry areas in New Guinea, are apparently the only ecologic group of New Guinean Carabidae in which Australian relationships and probable origins do pre- dominate. Because my ecologic division of the New Guinean fauna is rough at best, with many details in doubt, I see no point in attempting to find finer correla- tions between ecology and geographic origins.
[SO] Direction and oltitude: mountain- hopping, across the Malay Archipelago.
Direction of dispersal is not clearly cor- related with altitude. Most Garabidae on the high mountains of New Guinea seem to have been derived from lowland forms on the island and not to have had inde- pendent geographic origins [90]. Of the few endemic high-mountain genera that do seem to have independent geographic relationships, AnaJoma and Idiagomim probably have Oriental and Rhytiferonia Australian ties, but their origins are far from clear. Less differentiated mountain- hopping carabids that have reached New Guinea are considered in more detail be- low. Doubtful cases which should be dis- posed of first include Notagonum suh- metaUicum (see Tax. siippL), a common, winged, southern Australian carabid which has been found at moderate altitudes both in tropical Queensland and in New Guinea; it may have begun to mountain-hop north- ward and westward, but if so, it has not gone far. Craspedophoriis, Physolaesthus, and Coptocarpus, although thus far found only at middle altitudes in New Guinea, occur at low altitudes in tropical Australia and are probably not mountain-hoppers.
The three genera MecyclotJiorax (Part I, pp. 498, 505; present part. Tax. suppl.), Mia'oferonia (Part III, p. 18), and Sco- podes (Part III, p. 197) have mutually similar distributions in the area mider discussion. All these genera are now chiefly Australian, but all have also en- demic species localized on mountains in New Guinea and in Java. (The distri- bution of MecycIotJwrax is complicated by occurrence also on the Hawaiian Islands, etc., and of Scopodes by occurrence on New Guinea not only of one Australian- related species but of a second endemic stock which has radiated on the island, but these are added complications which do not affect the Australia-New Guinea- Java pattern. ) The question is, have these genera mountain-hopped from Australia across New Guinea (and presumably Celebes ) to Java, or are the isolated species on New Guinea and Java relicts left by
226 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
withdrawals ( by partial extinctions ) of the Guinean species also well defined and
genera into Australia? I cannot answer this flightless, but perhaps less closely related
question, but at least in Mecijclothonix and to any particular Australian species (but
Scopode.'i the New Guinean and Javan the species of this genus need further
species seem fairly closely related to Aus- study), may be oldest of all. Tliis classi-
tralian species, and I think westward fication by age may be wrong in detail,
mountain-hopping is a possible explanation. However, it does seem probable that
All the pertinent New Guinean and Javan Mecijclothorax, Scopodes, and Microfewnia
species of these genera are now flightless, dispersed in the Malay Archipelago before
but all three genera include winged species Sericoda and Clujdaeus did, and it is at
in Australia, and all may have dispersed least possible that they represent stages
by flight. in dispersal by a "sweepstakes" route
A reverse pattern of distribution is across the mountains of the Archipelago
shown by Sericoda (present part. Tax. from Asia to Australia. No single group
suppl.) and Clujdaeus (Part III, p. 47), of mountain-hopping Carabidae occurs
both of which are distributed as if they along this whole route now. However,
have mountain-hopped from Asia across Sericoda and Clujdaeus have covered most
the Malay Archipelago to New Guinea, of the route (and an Asiatic stock of
In these cases direction of dispersal is Bemhidion, not related to the Australian
clearer. An actual Oriental species of Bemhidion, has reached the mountains of
Sericoda reaches New Guinea and is still Celebes — Darlington, 1959a), and Mecijclo-
winged. And Clujdaeus includes slightly thorax, Microferonia, and Scopodes may
differentiated species widely scattered on have follo\\'ed the whole route earlier,
mountains on the Malay islands, some mountain-hopping from Asia to New
flightless, but others still retaining wings Guinea and Australia, becoming extinct in
(for example, C. hakeri Andrewes on Asia, and radiating in Australia especially
mountains in Luzon is still dimoqohically south of the tropics. If so, the Javan and
winged). New Guinean species of these genera may
The five cases of probable or possible be geographic relicts, and the special New
mountain-hopping summarized in the two Guinean group of Scopodes (Part III, p.
preceding paragraphs can be arranged ac- 197, Notes under Scopodes) may represent'
cording to apparent ages of dispersal, a separate radiation of the ancestral stock
relative age being judged by state of wings which has paralleled the Australian radi-
and by amount of differentiation of the ation rather than being derived from it.
isolated species. Sericoda, with a winged This hypothetical history is diagrammed in
species now extending from Asia to New Figure 14. Guinea, has presumably dispersed most I do not know how these five genera
recently. Clujdaeus, with slightly differ- really have dispersed. But I think the
entiated species scattered from the Hima- best way of attempting to decipher their
layas to the Philippines and New Guinea, histories is to put the cases together, see
and with wings still present in some insular whether they fit a common pattern ( as
forms, is presumably a little older. Mecijclo- they seem to do), and see how that pattern
thorax and Scopodes, each with well- compares with the distributions and ap-
diffcrentiated, flightless species on Java parent histories of other Asiatic- Australian'
and New Guinea, but with the Javan and Carabidae and of other animals and even
New Guinean species apparently related of plants.
to each other and related to specific Aus- Among New Guinean Carabidae, Deme-
tralian species, are presumably still older, trida (Part III, pp. 140ff) may fit the pat-
And Microferonia, with Javan and New tern suggested for Scopodes. The ancestor
The carabid beetles of New Guinea • Darlington
227
6 < -
5
4
3
2
I -(N.AMER.)-
ASIA
JAVA
N. TEMP- ERATE
T
■>
? .
? ■ ■ ? - 7
CELEBES
R 0
^
^
->^
NEW GUINEA
c
AUSTRALIA
I
|S. TEMP-
I ERATE
Figure 14. (Hypothetical) diagram of dispersal of successive mountain-hopping Carabidae from Asia to Australia. Broken lines indicate presumed past occurrences; solid lines, present occurrences. Case 1 represents Psydrini (included in Ptero- stichini in text) other than Mecyc/ofhorax, now relict in North America and moderately diverse in Australia; 2, Mecyclotho- rax, now represented on mountains in Java and New Guinea and moderately diverse in south temperate Australia (and well represented in Hawaii, etc.); 3, Scopodes, now on mountains in Java and New Guinea with separate radiations in New Guinea and Australia; 4, Chydaeus, with slightly differentiated species on mountains from Asia to New Guinea but not Aus- tralia; 5, Agonum (Ser/codo) cey/onicum (Motschulsky); 6, 6emb/d/on of the bryanti-pendelburyi group (Darlington, 1959a), which has reached Java and Celebes but not New Guinea. See text for further details.
of Deiiietiida must have reached the Aiis- trahan Region from some other part of the world, probably from Asia, since the genus does not have a subantarctic distribution. The New Guinean radiation of the genus (Part III, pp. 143-144), like that of Scopodes, may be a separate radiation of an original Asia-derived ancestor independent of the main Australian radiation. (However, I have scored Demctrida as probably halv- ing dispersed from Australia to New Guinea, see [74], Table 14.)
A pattern of distribution that seems basically similar is shown by the native murid rodents of the Australian Region, among some of which radiation has pro- ceeded partly independently in New Guinea and in Australia. The subfamily Hydromyinae, for example, has diversified principally in New Guinea. In this case we know that the ancestral stock did come from Asia, and we can be reasonably sure that the ancestor of the New Guinean Hvdromvinae was Asia-derived and did not
come from Australia even by counter- movement.
Another basically similar pattern is shown by a geographically notable genus of plants: trees of the genus Nothofagus (southern beeches) are well represented on the mountains of New Guinea as well as in south-temperate Australia, New Zea- land, and South America. We know (from their pollen record) that the history of these trees in the southern hemisphere has been complex, but we can be reasonably sure that their ancestor came from the north. The numerous species on the mountains of New Guinea may therefore be products of radiation of an ancestor derived from Asia, not from Australia. (For further details and discussion see Darlington, 1965: 29-31 and other pages listed in index.)
In summary of Carabidae that seem to have mountain-hopped for considerable distances in the Malay Archipelago, Seri- coda and Clnjdaeus (and Bemhidion, to
228 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Celebes) have elearly dispersed southeast- ward, from Asia, relatively recently; and Mecijdothorax, Microferoniu, and Scopodes may have dispersed northwestward, from Australia, relatively long ago. However, all these genera may fit a common pattern of successive dispersals southeastward, from Asia toward Australia, with ancestral stocks later becoming extinct in Asia, and with separate evolutionary radiations some- times (in Scopodes) occurring in New Guinea as well as in Australia. Some other Carabidae (notably Demetrida) may fit (a terminal stage of) this pattern. And some rodents (Hydromyinae) and some plants (notably Nothofagtis) may have followed the same pattern of southeastward dispersal, extinction in Asia, and radiation in separate centers in New Guinea and Australia.
[81] Direction and age. Whether di- rection of dispersal is correlated with age (time of arrival) of different groups of Carabidae in New Guinea is an important question. It is conceivable that early arrivers might show different patterns of relationships and origins than later ar- rivers do, and the differences might indi- cate changes in the geographic or ecologic relations of New Guinea to the Orient and Australia. In the absence of a fossil record, age cannot be determined exactly, but it is at least a good working assumption that genera endemic to New Guinea (Table 13) are relatively old (but see [89]). Endemic genera of which the ancestors seem to have come from the Orient are Perileptodes, Analoma, Lyter, Odonto- masoreu.s\ Tarsagonum, Idiagonum, and additional chiefly mountain-living agonine genera, and those of which the ancestors seem to have come from Australia are Rhytiferonia, IleUuonidius plus Ilelluo- papua (one stock), and three endemic genera (perhaps only one stock) related to and perhaps derived from Loxandriis. (The relationships of Tiferonia and Minii- phloeus are doubtful.) Among these pre- sumably relatively older arrivers, therefore.
Oriental stocks seem to be at least twice as numerous as Australian stocks, and the evidences of direction are clearer in the case of some of the Oriental stocks than of the Australian stocks, I think.
The carabid stocks that have radiated on New Guinea may also be relatively old. The radiation of agonines derived from Notagonum- and Colpodes-\ike ancestors is unique on the island. Their ancestors were Oriental. The uni(|ueness of this case lies not only in the amount of differ- entiation of species and genera at all alti- tudes (Part II) but also in the ecologic radiation (of mesophiles, hydrophiles, and even a few arboreal forms) that has occur- red within the limits of New Guinea. A less striking radiation has occurred in the Pterostichini related to (derived from?) LoxandriLS; this group is Australian in pres- ent relationships and may have been derived from Australia (but see [82]). The radiation of this group has apparently produced two or three endemic genera in New Guinea {Haploferonia, Homalonesi- ota, and Nehrioferonia, if the latter is recognizable) but few species, and its ecologic radiation has been relatively slight: its members are confined to low and middle (not high) altitudes, and the habitats occupied are only those associated with standing water (most Loxandriis), the banks of running water (Nehrioferonia and probably Uomaloncsiota), and the rain-forest floor (probably Haploferonia and Loxandriis latus).
Tlie most striking multiplication of spe- cies on New Guinea has occurred in the genus Demetrida, which may be Australian in origin. Most members of this genus in New Guinea apparently live in the foliage of rain forest. Other species-radiations of Carabidae in rain-forest foliage have oc- curred in Dolichoctis of the aculeata and' related groups, which may be derived from one Oriental ancestor, and in Dicraspeda, which is a mainly New Guinean genus ( six rather diverse species on the island, all in lower-story rain-forest foliage) of unde-
The carabid beetles of New Guinea • Darlington 229
termined ancestry. Among the Carabidae differentiated species on Celebes. This
that Hve on tree tiamks in rain forest, pattern suggests recent dispersal from
moderate species-radiations have occurred Australia to New Guinea to Celebes. But
in Catascopus of the icallocei and perhaps Loxandnis is well represented also in the
other groups (Oriental in derivation) and warmer parts of North, Central, and South
in Minuthodes, which now occurs chiefly America, and additional genera related to
on New Guinea (nine species on the or derived from Loxandnis occur in the
island) and of which the ancestor is vm- Americas as well as in Australia and New
determined but was probably Oriental Guinea, but nowhere else in the world, as
rather than Australian. Other carabid far as I know. The most likely explanation
stocks in which moderate radiations of is that Loxandnis once occurred also in
species have occurred on New Guinea and the Old- World tropics or at least in tropi-
of which the ancestors probably came from cal Asia, that it reached Australia from
the Orient include several subgroups of Asia long ago, that it later became extinct
Tachijs (especially the serra group), some in Asia, and that still later it made minor
Perigona, Trichotichnus, and Pogonoglossus, return movements from Australia to New
all primarily ground-living mesophiles. And Guinea to Celebes.
others of which the ancestors probably Most species of Loxandnis are hydro- came from Australia include some Clivina philes; some of them are among the most ( hydrophiles ) , Agonochila (arboreal), a aquatic of Carabidae, although some de- special group of Scopodes (mesophiles, or rived or related fonns are more terrestrial, on rotting logs), and Helhionidius (prob- Another subaquatic genus of Carabidae, ably mesophiles). On the whole, the groups the oodine Anatrichis (Part III, p. 31), has in which species-radiations have occurred a distribution that may correspond to that on New Guinea probably include more of an ancestral Loxandnis. Anatrichis oc- Oriental derivatives than Australian de- curs from southeastern Asia (including rivatives, and the Oriental origins are Japan, according to Ueno, personal com- clearer than the Australian ones, I think. munication) to Australia, and is widely
Facts and probabilities considered in the distributed also in the warmer parts of the
three preceding paragraphs suggest no Americas. Both Loxandnis and Anatrichis
obvious correlation between direction of are primarily tropical. They extend into
dispersal and age of Carabidae on New moderate temperate areas but do not have
Guinea. Some Carabidae have probably the "subantarctic" distribution patterns of
reached the island from Australia as well some other Carabidae that may have dis-
as from the Orient in both older and more persed across Antarctica (for example, the
recent times, but Oriental stocks have Migadopini, Darlington, 1965: 35-37). Tlie
probabK' always been more numerous, close relationship of the American and
This summary concerns only the direction Australian Loxandnis has been established
of dispersal. Possible changes in rate of by modem methods of comparison (Moore,
dispersal (numbers of stocks reaching New 1965), but Anatrichis needs study not only
Guinea regardless of direction) from time of species now assigned to the genus but
to time are considered in [85]. of some other Australian species that may
[82] Australian-American discontinuities, prove to be related (Darlington, Part III,
A special pattern of distribution is exempli- P- 31, paragraph 3).
fied by Loxandnis (Part I, pp. 498, .549- Two other, nonaquatic genera of Carab- 557), which occurs in two widely separated idae should be mentioned in this con- regions. The genus includes numerous nection. The harpaline genus Ne7nagIos.sa species in Austraha, four Austrahan-related or Lecanomerus (Darlington, Part III, p. species on New Guinea, and one slightly 45) is supposed to occur in South America
230 Bulletin Mitseinn of Comparative Zoology, Vol. 142, No. 2
and Australia, but the relationships of the South American and Australian fomis need further study; whatever the earlier history of the group, the small species of Lecano- meriis in New Guinea have presumably been derived from an Australian stock rather recently. And Psetidaptimis {ThaJ- piiis) occurs in the warmer parts of the Americas (numerous species) and in Australia (one species); this genus too needs study.
This pattern of Australian-American dis- continuity occurs also in leptodactylid and hylid frogs, chelyid turtles, and especially marsupials. These animals do not have "subantarctic" distributions. All of them, like the Carabidae just discussed (Loxan- clrus, etc.), inhabit principally tropical and warm-temperate areas, and none of them occurs on New Zealand (except as recently introduced by man). The pattern may have been fonned in somewhat dif- ferent ways in different cases, but it is probably usually a relict pattern, the result of widespread extinctions in the main part of the world.
[83] Sirmmaiy of direction.'^ of dispersal. New Guinean Carabidae have been found to include 173 Oriental and 120 Australian "geographic units" [58]. This finding (to- gether with the fact that the carabid fauna of New Guinea seems to be mainly deriva- tive, and not a source-fauna from which many groups have radiated geographically [70] ) suggests that Carabidae have moved from the Orient to New Guinea more than from Australia to New Guinea. A com- parison of the distributions of rapidly dis- persing groups of Carabidae with those of more slowly dispersing groups [72] shows relatively strong Oriental relationships among the rapid dispersers, which again suggests movement mainly from the Orient toward Australia. Comparison of the dis- tributions of Carabidae with those of mammals [73] suggests that many cara- bids have moved from the Orient to New Guinea and Australia while the terrestrial mammals were isolated in the Australian
Region; a number of carabid species seem to have dispersed from Asia to Australia so recently that populations are not or not much differentiated on the two contin- ents. Selected tribes, genera, and species of Carabidae that seem most clearly to show directions of movement [74, 75] include 89 groups that seem to have dis- persed southeastward and only 39 groups that seem to have dispersed northwest- ward, and the disparity is greatest over the longest distance: 44 carabid stocks seem to have dispersed all the way from the Orient to some part of Australia, while only seven stocks seem to have dispersed from Australia to the Oriental area, and only one primarily Australian genus seems to have reached the mainland of Asia.
Before making a final summary, I want to re-emphasize the complexity of the situation, the difficulty of finding and assessing real evidence, and the tentative nature of the conclusions. Faunal move- ments are statistical, not co-ordinated one- way movements. They are the sums of very complex movements and countermove- ments of many families, tribes, genera, and species, which in turn are the sums of almost inconceivably complex movements of individuals. However, in spite of the difficulties and complexities, the evidences of net direction of movement of carabids in the Asiatic-Australian area are surpris- ingly good. The movements have been complex. Large numbers of Carabidae have apparently dispersed from Asia and the Oriental islands to New Guinea and Australia over a long period of time, while the numbers that have apparently dis- persed from Australia to New Guinea and Asia are significantly smaller. Conclusions about direction are more or less tentative in single cases, but when all the cases are put together they form a pattern which (I think) is as a whole overwhelmingly prob- able: dispersal has been predominantly toward the southeast, from Asia toward Australia. At any one point along the route the preponderance of southeastern
The carabid beetles of New Guinea • Darlington 231
I ORIENTAL I AREA
<
I
I
MOLUCCAS |NEW GUINEA 'CAPE YORK I ,
I
37
T
AUSTRALIA BELOW C. Y.
35
z.
25
V
4
4
7
I
Figure 15. Diagram of apparent directional movements o'f Carabidae in the Oriental-Australian area. Lengths of arrows indicate distances moved; widths, numbers of stocks that seem to have made the movements; and numbers of stocks are given in figures on the arrows. See text for further details.
against northwestern movements may not have been very great, but over the route as a whole movements from Asia to New Guinea and AustraHa seem to have been several times more numerous than move- ments from Australia and New Guinea toward Asia. The resulting, coniple.x but di- rectional faunal movement is diagrammed in an oversimplified way in Figure 15. I My conclusion is that there has in fact 'oeen direction in the sum of movements ,)f Carabidae in the Asia tic- Australian area:
a continual flooding of Asiatic stocks into the Australian region, with much less movement of Australian stocks toward Asia, except for short distances. This process has probably been going on for a very long time, and is still going on. I think that it is only one feature — but a major one — of a complex pattern of directional movements which Carabidae are continu- ally making over the world as a whole, but which (in the absence of a fossil record) are very difficult to demonstrate. The
232 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
apparent world-wide pattern of movements has been briefly described in [67].
One other deduction: if, as I suppose and as apparently must be the case if Carabidae are not to multiply without limit everywhere, additions to carabid faunas tend to be balanced by extinctions, then the movements toward Australia of many new Asiatic stocks must have been accompanied by withdrawals of many other Carabidae, the withdrawals tending to begin in Asia and to progress toward New Guinea and Australia. Some sort of balance between spreading of new groups and withdrawal of old ones would explain several characteristics of the New Guinean carabid fauna. It would explain, for ex- ample, the various cases in which New Guinean Carabidae seem to find their closest relatives in the Philippines; I have not stressed these cases, because most of the groups concerned are not well enough known taxonomically, but possible ex- amples will be found imder Tachijs exiil (Part I, Notes on p. 421), Tifcronia (Part I, Notes on p. 561), Peri<^ona erimae (Part III, Notes on p. 12), and Lehia popiiella ( Part III, Notes on p. 88 ) . Tlie explanation, of course, is or may be that these groups once occurred on the Greater vSunda Islands or even on the continent of Asia, but have become extinct there as competing groups have moved from Asia into the archipelago. Another, more important characteristic of the New Guinean carabid fauna is that it seems to include few or no phylogenetic rehcts [88, 89], and this may be not be- cause New Guinea is a young island but because there has been a relatively rapid overturn of the famia caused by the con- tinual, massive influx of new stocks, chiefly from Asia, with extinction of older stocks. And some kind of balance between arrixal of new stocks and progressive extinction of older ones may also explain the dis- tribution of the "mountain-hoppers" [80], of which new stocks seem to be coming from the Asiatic end of the archipelago
while older stocks may be "retreating" to- ward Australia (Fig. 14).
[84] Barriers; filtering, at Wallace's Line, Moluccas, and Cape York. During disper- sal across the Malay Archipelago, Carab- idae seem to have encountered principal barriers at Wallace's Line, in the Moluccas, and between New Guinea and Cape York. These have probably been main filter points in the "sweepstakes route" (Simpson, 1940) which many Carabidae have followed, in one direction or the other, between Asia and Australia. These three bankers are of different sorts, and their filtering effects have probably been somewhat different.
At Wallace's Line, the barrier has been a gap of salt water — Makassar Strait — which has probably sometimes been narrower than now (perhaps only 25 miles wide in the Pleistocene) but which has separated Celebes from the continental shelf of Asia for a very long time. This is the first major barrier to eastward dispersal of Oriental stocks, and it is a relatively simple but effective barrier. Its effect on Carab- idae has probably been primarily to block more sedentary forms while more vagile forms, especially small, winged, active ones, have often crossed it.
Among the Moluccas, too, water gaps have probably existed for a very long time. However, their filtering effect on dispers- ing Carabidae may have been less than the effect of Makassar Strait. Tlie most sedentary, least vagile elements of the carabid faunas of both the Orient and. Australia were probably filtered out by other barriers, so that the Carabidae that reached the Moluccas have been pre- selected for active dispersal. However, the Moluccas are relatively small islands. Their limitation of area has probably limited the numbers of species that could occur, and has probably given an extra advantage to dominant, successful stocks. The islands have therefore probably been a bottleneck through which only Carabidae with some degree of general dominance have been able to pass. (For further consideration of
The carabid beetles of New Guinea • Darlington
233
the Caiabidac of the Moluccas see [62]. And for discussion of dominance among Carabidae see [68].)
Finally, although New Guinea and Aus- tralia are separated by water now, they were broadly connected by land at times in the Pleistocene. Nevertheless only small fractions of the New Guinean and Austra- lian carabid faunas crossed the land con- nection. The New Guinean rain-forest carabids that reached Cape York or farther into tropical Queensland were not very numerous and some of them are patchily distributed now (see [63] and Table 9). And the Australian open-country carabids that reached even southern New Guinea were not very numerous and were ap- parently severely filtered, for very few of the many flightless carabid stocks of Aus- tralia reached New Guinea. Tlie barrier in this case seems to have been primarily ecologic, perhaps an alternation of areas of rain forest and open country like that which actually occurs on the Cape York Peninsula now, and which may have im- posed (different) barriers to both rain- forest and open-country stocks, regardless of the direction of their dispersal.
Tliat the three principal barriers to dis- persal of Carabidae betAvcen Asia and Australia have probably acted in somewhat different ways is, I think, important. To- gether, however, they have had a common, net result: the accumulation on N^w Guinea of a carabid fauna composed almost entirely of small, winged, active carabids, many of them belonging to inherently successful, widely distributed, actively dis- persing groups.
Among these actively dispersing groups, the hydrophiles are outstanding. Hydro- phile Carabidae, that live beside water, are usually winged and do fly relatively often either to escape rising water or to main- tain populations in shifting water-side habitats. Because they live in unstable, shifting habitats, they presumably have to maintain relatively large populations, and , this may be an additional advantage to
them in dispersal. Water-side habitats are to some extent independent of forest cover, so that Carabidae associated with them can disperse relatively easily across areas like present Cape York, where the forest cover is discontinuous. And presence of water probably gives some protection against extremes of climate, so that hydro- philes can cross climatic barriers relatively easily too. Hydrophile Carabidae do in fact seem to be exceptionally good dis- persers. They have reached New Guinea in sufficient numbers not only to occupy their own habitats but also in several cases to invade the floor of rain forest, where they have apparently compensated for a deficiency of less vagile forms primarily adapted to the forest floor [97]. I have elsewhere (1959a; 1962a; 1965) suggested that not only the winged Australian Bembi- dion but also the now chiefly flightless and chiefly mesophile "Trechiis" of south- temperate Australia have been derived from winged ancestors which crossed the tropics from Asia to Australia at low alti- tudes in water-side habitats.
[85] Amount of dispersal, noiv and in the past. Amount of dispersal — number of stocks dispersing between New Guinea and other areas regardless of direction — can be correlated with time, although the data are necessarily imprecise and the correlation is rather rough. Tlie method is to count and compare the numbers of stocks of New Guinean Carabidae at three taxonomic levels: at the level of non- endemic species, which are shared with the Oriental area and/or Australia, and each of which has had its own separate dispersal; at the level of endemic species or groups of species not in endemic genera, counting only species or species groups that seems to have separate relationships outside New Guinea; and at the level of endemic genera or groups of genera, count- ing only those with separate relationships outside New Guinea. A general correlation of taxonomic level with time is assumed. That is, it is assumed that, in spite of
234 Bulletin Museum of Comparative Zoology, Vol 142, No. 2
probable differences in rate of evolution of different stocks, nonendemic species have in general dispersed more recently than the ancestors of endemic species, which in turn have in general dispersed more recently than the ancestors of en- demic genera.
The kind of unit to use for this puqoose must be considered carefully. To count only stocks that show direction of dispersal [74, 75] would weight the results in favor of the more recent stocks, because clues that show direction are likely to be clearest in the stocks that have dispersed most recently. And the stocks to be counted now are not the same as my "geographic units" [31, 58]. The latter were designed to show existing relationships, and some wide-ranging stocks were therefore counted twice: e. g., a New Guinean species that occurred also in both the Orient and Aus- tralia was counted as both one Or and one Au unit. Now, each stock will be counted only once. Also, in scoring "geo- graphic units," I have counted only stocks with discernible geographic relationships, while now I shall count all stocks that seem independent within the New Guin- ean fauna, whether or not their geographic relationships are discernible. The resulting units might be called amount-of-dispersal units. They are designed simply to show the numbers of separate stocks that make up the existing New Gmnean carabid fauna, correlated with taxonomic level and therefore presumably with age. Since the New Guinean carabid fauna is as a whole a relatively recent one [89], most amount- of-dispersal units probably represent move- ments into New Guinea, but this is not assumed in the calculation.
Table 16 summarizes the amount-of-dis- persal units in the existing New Guinean carabid fauna. Before interpreting it, I should ask whether changes in my con- cepts of species and genera would sig- nificantly change the results — whether, if I "split" species and/or genera, the table would be significantly changed. I think
Table 16. Summary of aaiount-of-dispersal
units: NUMBERS OF SEPARATE STOCKS OF NeW
Guinean Carabidae at three taxonomic levels
|
Nonendemic species |
Endemic species and species groups in nonendemic genera |
Endemic genera and groups of genera |
|
|
Other than Agonini Agonini |
128 9 |
±129 ±9 |
11 ±5 |
the answer is that the counts would be changed, and that it might become neces- sary to tabulate units at additional taxo- nomic levels, including perhaps subgenera and natural groups of genera, but that if the classification were approximately phylogenetic, the table would still show many more dispersals at more recent than at the oldest level.
Table 16 indicates that the existing New Guinean carabid fauna consists of many stocks that have not differentiated specifi- cally, many that have become distinguish- able species but not genera, but very few that have become distinct genera. This suggests that stocks that have dispersed more or less recently far outnumber those that dispersed longer ago. The difference in number of dispersals in proportion to time may have been greater than the figures show: the nonendemic species may have made their movements within a relatively short time; the endemic species may represent dispersals over a much longer time; and the endemic genera, a still longer one, so that the number of dispersals now represented in the New Guinean fauna by endemic genera may have been widely spaced over a very long period. This is diagrammed in Figure 16.
For this and other reasons actual amounts of dispersal cannot safely be calculated from Table 16. Nevertheless, the table does strongly suggest that one of two things has happened. Either amount of dispersal — number of groups moving into or out of New Guinea — has increased
The cababid beetles of New Guinea • Darlington 235
(0
o o
a:
UJ CD
2
iNON-E., ANCESTORS OF . |SPEC. , ENDEMIC SPEClESi
IN NON-ENDEMIC ' GENERA
ANCESTORS OF
ENDEMIC GENERA
I
T HIS
MORE RECENT
SITUATION
HYPOTHETICAL
OLDER
MORE RECENT
OLDER
i
TIME
Figure 16. Histogram of numbers of stocks in relation to time in tfie existing New Guinea carabid fauna. Division of the an- cestors of the endemic species and endemic genera into "more recent" and "older" categories is arbitrary. See text for further explanation.
enormously among Carabidae in relatively recent times. Or faunal overturn [68] has eliminated a large proportion of older stocks as new ones have come in. I prefer the latter explanation. It is consistent with the relatively recent nature of the New Guinean carabid fauna as a whole [89], and it is consistent also with my general bypothesis of dispersals and replacements if successive carabid groups over the world [67].
My conclusion is that, although the ex- isting New Guinean carabid fauna is made up of many recently dispersed stocks and comparatively few old ones, this situation need not be the result of a recent increase in amount of dispersal. Carabidae may have been reaching New Guinea in num- bers for a very long time, and the present fewness of old stocks may be due to ex- tinction and replacement. I see no indi- cation that New Guinean Carabidae form
236 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
t\vo groups, one old and one relatively re- cent, separated by a time when few or none of the beetles reached the island. On the contrary, the concept of continual ar- rivals with continual overtinns seems to me to fit the observed situation. Tliere is therefore, as I shall say again [88], no initial starting point to be looked for in the accumulation of Carabidae on New Guinea.
Of course this simple conclusion about the dispersal history of New Guinean Carabidae covers an immense ignorance of details and also ignores many complicat- ing factors and processes. For example, although I have noted that rates of evo- lution probably vary among different groups of Carabidae, I have not sufficiently emphasized how differences in rates of evolution may have affected some parts of the New Guinean carabid fauna. This is perhaps best shown by comparing the Agonini with the other carabids on the island. The Agonini are tabulated sepa- rately in Table 16. In general, the agonine figures confonn to the table as a whole in that number of dispersals seems to have been greater among more recent than among older stocks, but in the case of the Agonini the correlation has been blurred by radiations on New Guinea, the radi- ations having proceeded to the point where the number of initial ancestors can no longer be determined with any accuracy. But I think the Agonini do fit the main pattern reasonably well, and that so far as they differ, the difference is due not to their being older on New Guinea but to their having evolved there either more rapidly or more diversely.
[86] Suuimanj of ii,eop,raphic origins of Netc Guinean Carabidae. The history of Carabidae in the Asiatic-Australian seg- ment of the world seems to have conformed in general (but of course not in detail) to the better documented histoiy of verte- brates. The largest favorable area (the Old-World tropics, including tropical Asia) has apparently been the principal
center of evolution and dispersal of domi- nant Asiatic-Australian Carabidae. The smaller and less favorable area of Australiai has been a less important center. And the still smaller area of New Guinea has been least important in the evolution and dis- persal of dominant carabid stocks.
Carabids have apparently been coming: into New Guinea continually during considerable period. There is no good evi- dence that arrivals were more numerous at some times than at others; arrivals have apparently been very numerous recently, but faunal overturns, with extinctions per- haps of many species, may have obscured the evidences of earlier arrival rates. Both Oriental and Australian stocks have prob- ably reached New Guinea at all times. The incoming Oriental have probably always exceeded the incoming Australian stocks in number and importance, except that among terrestrial mesophiles living in relatively dry, relatively open country Australian arrivals have been more nu- merous at least recently. Otherwise, incom- ing Oriental stocks have probably been more numerous than Australian ones re- gardless of size of insects, regardless of wing-state (almost all were winged), in all main habitats, and at all altitudes, as well as at all times. There must also, 1 think, have been continual extinctions correlated with the arrivals, and if the ar- rivals have been more from the Orient than from Australia, extinctions ("withdrawals") also have probably tended to begin at the Oriental end of the area and proceed toward Australia. The result has probably been a gradual shift of major distribution patterns from Asia toward Australia, caused by the procession of both dispersals and extinctions in this direction. But the details of this pattern have been excessively complex and have been further compli- cated and partly obscured by local evo- lutions and radiations of some groups within New Guinea and elsewhere.
The broad movement of Carabidae from Asia toward Australia is, I think, part of
The carabid beetles of New Guinea
Darlington
237
the world-wide pattern of evolution of successive dominant groups in the great, climatically favorable area of the Old- World tropics (Africa and tropical Asia) and of dispersal into smaller and/or cli- matically less favorable areas, with re- placement of older by more recently dominant groups [67, 68]. And indications af extensive overturn in the New Guinean Fauna itself [85] and evidences (not given here) that Carabidae have dispersed from New Guinea eastward to smaller islands, 3n which numbers of species of Carabidae are at least roughly proportional to area \nd isolation, are consistent with Mac- Arthur and Wilson's (1967) theory of directional dispersal, faunal overturn, and "aunal equilibrium on small islands. (In ^act my diagram (Fig. 16) of the age itructure of the New Guinean carabid auna, although differently constructed, ?an be considered a model of faunal equi- librium comparable to the MacArthur- Wilson equilibrium model.) The New Guinean carabid fauna thus fits into and connects both the apparent world^^ide pattern and the local insular pattern of svolution, dispersal, and faunal balance. [ find this a very satisfying, unifying con- :ept.
EVOLUTION
[87] Evohitionanj perspective. For a rhorough discussion of animal species and ?volution, with the necessary background 3f modern biology and genetics, see Mayr, 1963. And for a careful statement of the ipparent role of evolution in detennining lattems of dispersal and resultant distri- jution patterns, see Darlington, 1965, Chap- ters 5 and 6. New Guinea can legitimately 36 treated as an evolutionary center of its u)wn, but it should be remembered that it ilso has its place in worldwide evolution ind dispersal patterns [67]. The patterns ire apparently determined by the relative ireas of different pieces of land, and by 'limate. The fact that New Guinea, al- hough a very large island, is still much
smaller than Asia or Australia has presum- ably limited its importance as a center of origin and dispersal of major groups [70], but has probably not limited either rate or diversity of evolution of species upon the island.
[88] Evolution of Neic Guinean carabid fauna as a whole. Geographic differenti- ation and overturn of Carabidae on New Guinea have been considered briefly in [85]. Now to be discussed in more detail are the evolution of the New Guinean cara- bid fauna as a whole, the differentiation and radiation of particular carabid stocks, and some general trends in the exolution of Carabidae on the island.
The starting point for discussion of the evolution of the New Guinean carabid fauna is the assemblage of ancestral stocks from which the fauna has evolved. These stocks were evidently pre-adapted in several ways. They were pre-adapted for dispersal across barriers, the pre-adapta- tions including usually possession of wings (the whole New Guinean carabid fauna is or may be derived from ancestors that were winged when they reached the island), small (or at least not very large) size, probably other characteristics of structure and behavior including the habit of flying actively, and often adaptation to water-side habitats. Some, but not all, of the initial stocks (especially some Agonini) had also characteristics — general adapta- tions [68] — that made for general domi- nance, i.e., for success in a variety of situations.
The pre-adaptations that favor dispersal and the adaptations that favor continued existence in island habitats are partly dif- ferent and opposed. For example, wings and flight pre-adapt a stock to dispersal, while existence in some habitats on some islands favors atrophy of wings. Or, for another example, small size favors disper- sal, but continued existence on some islands may in some cases ( perhaps on Madagascar [20] but apparently not on New Guinea) favor increase of size. There mav therefore
238 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
be a reversal of direction of selection and stocks in these habitats, the deficiency adaptive evolution after carabids reach an being compensated for in part by multi- island, and tliis in part explains the extraor- plication of the comparatively few rain- dinary distinctness of many insular cara- forest stocks that have reached the island bid faunas, including that of New Guinea and in part by ecologic shifts of hydro- especially at higher altitudes. To restate philcs onto the rain-forest floor. This is this important generalization in different consistent with the general rule stated by words: dispersal from a source fauna to Wilson (1961) for the ants of New Guinea an island selects as ancestors of the island and other islands: that dispersals tend to fauna fractions of the source fauna that occur in marginal habitats, and that after happen to be pre-adaptcd for dispersal; dispersal some stocks penetrate the rair but when these fractions evolve and radiate forest and evolve and diversify there, to form a whole fauna on the island, di- Water-side habitats apparently are mar- rection of selection is partly reversed, and ginal so far as the rain forest is concerned the new fauna that evolves is likely to be They are certainly less stable and also more very different from the source fauna in widely distributed than habitats on the gross taxonomic composition as well as in ground in rain forest. Carabidae that live many details. beside water do seem to disperse relativel) Most of the ancestral stocks from which easily and do apparently undergo e-cologic the New Guinean carabid fauna has shifts onto the rain-forest floor [84, 97]. evolved apparently reached the island at So, I take as the starting point of evo- low altitudes anel were adapted to lowland lution of the New Guinean carabid faunn tropical habitats. Carabidae in such habi- an initial assemblage of immigrants: small tats, including tropical rain forest, often do winged, dispersible forms concentrated al seem to be small and winged [21]. This low altitudes and adapted to existence ir may be partly because carabid populations the lowland tropics, but "disharmonic' in these places tend to have "patchy" distri- taxonomically, with a surplus of Agonini butions [22] maintained by continual redis- and disharmonic also ecologically, with a persals, so that the characteristics that surplus of hydrophiles but a moderate pre-adapt for dispersal to an island also shortage of stocks living in rain forest ane pre-adapt Carabidae to survive in lowland an extreme shortage of stocks adapted tc tropical habitats on the island. However, mountain habitats. I call this an initial dispersal across a tropical archipelago not assemblage, but I elo not think of it as ha\' only strongly selects small, winged, low- ing a single starting point. It was, rather land Carabidae but also probably has other a changing or evolving continuum, a sorl filtering effects. Carabidae tend to be of faunal germ plasm, of relatively un more active and more liable to dispersal in specialized forms. We do not know when some habitats than in others, so that some it began. We do know or at least suppose e^cologic groups disperse more readily anel that it has been continually changing b\ reach islands more often than the others "faunal overturn," by arrival of a succession do. Dispersal may be more rapid in rel- of ne^w stocks and extinction of old ones atively open country than in heavy forest, although some of the older stocks have And carabids that are associated with been evolving and radiating even while water apparently disperse more readily new ones have been continuing to arrive- than those that are not. The multiplication I think that overturn has probably beer of species in some groups (e. fi,., Deme- so extensive as to destroy most or all of the trida, see Part III, p. 140) in foliage and earlier stage\s of accumulation of Caraliieku on tree trunks in rain forest in New Guinea on New Guinea. There are two reasons foi suggests an initial deficiency of carabid thinking that overturn has been extensive
i
i
The carabid beetles of New Guinea • Darlington 239
First, many species of Carabidae have From this changing continuum different
reached New Guinea so recently that they carabid stocks on New Guinea have
are not yet differentiated there, and if evolved to different extents and in differ-
one accepts the idea of faunal balance as ent ways. The most striking processes have
applying to all faunas everywhere ( except been multiplications of species and ecologic
extremely young or very isolated ones), radiations. These processes are further
New Guinea must have lost many older discussed below [91, 92]. As far as the
stocks to compensate for the arrival of new fauna as a whole is concerned, the princi-
ones. And second, the New Guinean cara- pal results of the multiplications and radi-
bid fauna seems to possess no striking ations have been greatly to increase the
evolutionary or geographic relicts, nothing number and diversity of both ground-living
that seems very old, or taxonomically iso- and arboreal Caralyidae in rain-forest habi-
lated, or geographically very distant from tats, and especially to fonn on the higher
its relatives, as if all really old members mountains of New Guinea a complex alti-
of the changing and evolving fauna have coline fauna which is ecologically like the
been eliminated. New Guinea possesses carabid faunas of mountains elsewhere
no endemic tribes; endemic genera are (Darlington, 1943) but which consists
few; and all of them have or may have largely of genera and species which have
relationships with other Carabidae still apparently evolved on and are confined to
existing in the Oriental or Australian the island. Most of this alticoline fauna
Regions. If such genera as Mecyclothorox seems to have been derived primarily from
and Loxandras are geographic relicts, they the surrounding lowlands of New Guinea,
have survived primarily in Australia rather by differentiations of alticoline stocks from
than in New Guinea. Australia and New lowland ancestors and by ecologic radi-
Zealand possess a number of geographi- ations especially of Agonini at high alti-
cally isolated carabid stocks (e. g., the tudes. The "momitain hoppers" [80] that
"carenums" [35] and various Pterostichini have reached New Guinea make up a com-
[39] in Australia, and certain stocks with paratively small part of the mountain
apparent northern relationships on New carabid fauna [90].
Zealand [89] ) which may have come long So, the New Guinean carabid fauna can
ago via New Guinea but are not repre- be thought of as derived from a changing
sented there now, and this is at least con- continuum of relatively unspecialized,
sistent with disappearance of older stocks small, winged, lowland ancestors, including
on New Guinea. I shall note again this many hydrophiles and some relatively un-
point — the apparent absence of relicts specialized Agonini, which have been
among New Guinean Carabidae — in con- coming in over a long period of time, and
sidering the relative age of the fauna [89]. from which have evolved on New Guinea
The actual rate of overturn (arrivals/ex- a great, partly endemic rain-forest fauna
tinctions ) can perhaps be calculated and a diverse, unique mountain fauna. The
eventually by formulae something like complexity of this process and of the whole
those devised by Mac-Arthur and Wilson carabid fauna it has produced is, of course,
( 1967 ) . But neither existing data nor my far beyond what I can describe or even
mathematics are adequate now. satisfactorily indicate.
To summarize: the "initial assemblage" [89] Relative age of the Neic Guinean
From which the New Guinean carabid fauna. If the New Guinean carabid fauna
fauna has evolved should be thought of as has evolved not from an initial set of an-
a changing continuum of which we cannot cestors beginning at one point of time but
>ee the beginning but only a relatively from a constantly changing accumulation
recent segment. of ancestors, including relatively general-
240 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
ized Agonini, to which additional incoming stocks have continually been added while other stocks have continually been elimi- nated, an exact determination of the age of the fauna may be impossible even in theory, and absence of a fossil record makes dating it in terms of geologic time impossible in practice in any case. We do not know the geologic age of New Guinea; we do know that some of the mountain ranges are geologically recent, but we do not know how long a significant piece of land has existed where New Guinea now is; we do not know when Carabidae were first able to reach and exist on proto-New Guinea, if there was one; and we do not know whether the ancestors of the existing fauna were the first carabids on the island or whether they were preceded by others that have disappeared during faunal over- turns. However, although we cannot de- termine absolute age, we can say some- thing about the relative age of the New Guinean carabid fauna, its age in relation to the faunas of other land areas.
Both tropical Asia and Australia have carabid faunas that include noteworthy evolutionary and geographic relicts. Tropi- cal Asia, for example, has an endemic tribe (Idiomoqohini), as well as at least a few genera which are isolated taxonomically and/or widely separated geographically from their closest relatives (for example, Mouhotia, a genus of enormous, flightless scaritines confined to the Indo-Chinese Peninsula), and Australia has two endemic tribes (Agonicini in the southeast and Cuneipectini in the west), as well as Famhorus, a striking endemic assemblage of large scaritines [35], diverse and isolated pterostichines (and psydrincs) [39], and others. New Guinea has no comparable relict or isolated Carabidae. But perhaps no island, not even a large one, should be compared with continents.
A more significant comparison can be made with the carabid faunas of New Caledonia and New Zealand. New Cale- donia has several genera of Carabidae so
distinct that their relationships are doubt- ful, or so isolated geographically that their dispersal routes are lost. For example, the New Caledonian genus C\iphocoleus (eight species) is so distinct tliat it is not clear whether it belongs in the Agonini or in the Lebiini, and the New Caledonian scaritine genus Anomophaenus (eight species) is not related to anything now existing in Australia or New Guinea, its nearest relatives being (perhaps) in tropi- cal Asia. And New Zealand has an endemic tribe ( Zolini ) , an extraordinary relict genus {Maoripamhorus, related to the Australian Pamhorus), and endemic groups of Tre- chini, Agonini (Sphodrini), and perhaps Bemlndion of which the closest existing relatives seem to be in norf/j-temperate areas (for discussion of these cases see Darlington, 1965: 64). In contrast. New Guinea, although it has a much larger carabid fauna than either of the other islands, has no endemic tribe of Carabidae and relatively few (few in proportion to the size of the fauna) endemic genera, and all of the latter have or may have relatives in adjacent areas, in either the Oriental Region or Australia. I conclude that the carabid fauna of New Guinea is more recent in its origins than the faunas of tropical Asia or of Australia or of New Caledonia or New Zealand.
It should be re-emphasized that the relative ages of the faunas do not neces- sarily indicate the relative ages of the islands. What the relatively recent age oJ the New Guinean carabid fauna probably does indicate, I think, is that, because oi the greater accessibility of the island. Carabidae have flooded into New Guinea in much greater numbers than into New Caledonia or New Zealand, and that fauna! overturnis have therefore been more rapic and more thorough in New Guinea.
[90] Evolution of the mountain fauna The general characteristics of the moun- tain carabid fauna of New Guinea are the same as those of mountain faunas else- where (Darlington, 1943). The mountair
The carabid beetles of New Guinea • Darlington 241
species are indicated on my data sheets [16]. Species are relatively few in any single mountain faunule, but geographic replacements are frequent, and the total number of species on all the mountains of New Guinea is very great [19]. The size range of the mountain species is within the range of the lowland species but with a single mode at 9-9.5 mm, which is larger than the larger of the two lowland modes; but this is apparently due not to increase af size of Carabidae on the mountains but to failure of many small forms to reach bigh altitudes [20]. Incidence of species ^vith atrophied wings increases with alti- ude, reaching about 95 per cent on the highest mountains; this is a result of itrophy of wings (and multiplication of >pecies following wing atrophy) in New Guinea, not of accumulation of wing- itrophied stocks from outside the island [21-23]. Ecologically, most of the moun- tain species are mesophiles, of which the majority live in wet montane forest, fewer in grassland above the forest zone; a few are hydrophiles living on the banks of nountain streams; and a few arc arboreal [24, 25]. The effects of altitude on Carab- idae [26] may be partly direct, but in- direct effects seem more important and probably include climatic control of vege- tation, limitation of areas, reduction of competition with ants, and presumably 3ther factors.
The mountain carabid fauna of New
i^uinea, of which some characteristics are
reviewed above, is composed of very many
separate stocks, derived mostly from the
iurrounding lowlands, over a considerable
period of time as indicated by their dif-
erent degrees of differentiation. Tlie most
["ecent movements, probably still going on,
nay be of lowland species into mid-altitude
?rass areas deforested by man [26]. Many
)ther of the 161 species of Carabidae
cnown to occur both lielow 500 and above
1000 m [19] may also have moved (spread)
rom the lowlands to mid-altitudes re-
t'litlv. In a few cases altitudinal differ-
entiation has apparently just begun, for example, in Brachidius crassicornis (Part I, p. 508; Tax. siippL), in which mountain individuals are relatively large; in Proso- pogmiis garivagliae (Part I, p. 537), in which lowland and mountain individuals differ slightly in elytral striation and width of intervals; and in T richotichmis nigri- cans (Part III, p. 52), which is fully winged at low altitudes but dimorphically winged on the Bismarck Range. In many other cases differentiation has proceeded to the point where mountain-living (usu- ally mid-altitude) species now exist in genera which are well represented also in the lowland fauna. Such genera in which the mountain-living species are still winged include (this is far from a full list) Tachys, Notagonum (see also below), Colpodes, Iridagomim, T richotichmis, Horpcdoxenus, Hyphaereon, Catascoptis, Agonochila, and especially Demetrida. A few primarily winged lowland genera include mountain species in which the wings have atrophied; for examples see Clivino ]<ubor (Tax. siippL), Lesticus (Part I, p. 521), and Notagonum ombidator (Tax. siippL). And differentiation has gone still further in some other genera (chiefly of Agonini, Part II) which are now confined to the mountains of New Guinea and of which all species now have atrophied wings, although they are probably descended from winged lowland ancestors.
These (incomplete) details are, I hope, enough to give some idea of how the mountain carabid fauna of New Guinea has evolved, primarily by accumulation of many stocks derived from the surrounding lowlands, and by their differentiation at and adaptation to high-altitude habitats and in some cases their further geographic differentiation and ecologic radiation on different mountain ranges.
The additional Carabidae that seem to have reached the mountains of New Guinea from outside the island, by "moun- tain-hopping" [80], have apparently done so at different levels. Agonum (Sericoda)
242 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
ceylanicum (Tax. .siippl.) and Nota^onum suhmctalUcum {Tax. supph) occur at mid- altitudes; Microferonia (Part III, p. 18), in the highest montane forest; and Mecy- clothorax (Part I, pp. 498, 505; present part, Tax. suppL), Chydaeus (Part III, p. 47), and Scopodes altu.s (Part III, p. 198), above the forest zone. The mountain- hoppers thus make up an insignificant fraction of the carabid fauna at intermedi- ate altitudes but a relatively larger fraction of the small faunules on the exposed summits of the highest mountains (see [19] and Table 2). Perhaps climatic bar- riers keep most lowland tropical groups of Carabidae from populating these summits, and perhaps "mountain-hopping" occurs most often at this altitude, where carabids adapted to temperate climates are exposed to wind.
I am not sure that I have sufficiently stressed the isolation of the mountains of New Guinea from the rest of the world so far as Carabidae are concerned. It is emphasized by the fact that no Bembidion [36] and no Trechus [37] have reached these mountains, although these genera (in a broad sense) are dominant on mountains in the northern hemisphere and have invaded or are invading mountains in southern Australia and Tasmania. It is of course the isolation of the mountains, the difficulty which Carabidae adapted to montane habitats in other parts of the world have had in reaching them, that has allowed an independent mountain fauna derived mainly from the surrounding lowlands to accumulate and evolve, with only minor additions received directly from outside the island (preceding para- graph).
As to age, the mountain carabid famia of New Guinea is not necessarily very old. The mountains themselves are apparently geologically young [17], and the distri- bution of Carabidae on them is essentially orderly, as if the insects have evolved (as mountain-living forms) on the moun- tains that now exist. If the beetles had
evolved on older mountains and occupied the present ones later, I would expect patterns of distribution to be much less orderly. A few endemic, mountain-living genera such as RJujtiferonia (Part I, p. 533), Analoma (=Paraloma, Part I, p. 538), and Idia^onum (Part II, p. 229) are relicts of a sort, since they do not seemi to have ancestral stocks still existing im New Guinea, but they are not very isolated taxonomically, and I think they are likely to have been derived from lowland Newi Guinean ancestors that have disappeared during overturns of the lowland fauna [89]. Some overturn may have occurred' on the mountains too, but the orderliness of distribution patterns suggests that re- placements have not been extensive.
The enormous ecologic shift that Carab- idae (and other insects) make in moving from the tropical lowlands to higher alti- tudes in New Guinea should be re- emphasized. It involves formidable changes of climate, of vegetation, and probably of micro-habitats, and often great reduction of the areas inhabited by populations. Limitation of area may be the most im- portant of these factors [22]. It presumably affects the structure of populations, which in turn affects number of species, state of wings, and so forth. The mountain Carab- idae of New Guinea show other diverse specializations. Tlie body fonn is often strikingly modified (see figures on pages 263, 295, 312). Loss of setae is common [95]. The eyes are often reduced, be- coming either small and flat or small and abruptly prominent (see Part II, pp. 95- 96, and [98]). The mandibles are some- times modified, becoming either long and curved or exceptionally slender and almost straight, presumably in adaptation to special foods. In a few cases, for example Brachidius cras.sicornis (Part I, p. 508) and Nota^onum dentellum with subspecies chimbu (Part II, p. 149), an increase of size of mountain as compared with lowland individuals has been noted, but in Lesticus (Part I, p. 521) the mountain species are
The carabid beetles of New Guinea • Darlington 243
Table 17. Notable species-radiatioxs among New Guinean Carabidae, excludlng Agonlni ( for
wtiich see [92] )
Demetrida: 59 known species on New Guinea from 1 or few ancestors; at low and moderate altitudes, chiefly in foliage in rain forest; this case put first and separated for emphasis
Clivina: probably minor radiation of australasiae and some other groups; chiefly at low altitudes; fossorial, mostly in wet places
Tachijs: minor multiplications of species; chiefly at low altitudes; chiefly on the ground in rain forest (serra group) and beside running water (po/jY(/.v group, in part)
Perigona: probably minor radiation of species in subgenus Trechicus- at low altitudes; in leaf litter in rain forest
Trichotichnus: moderate species radiations; at low and middle altitudes; especially on the ground rain forest
m
Minuthodes: primarily New Guinean genus (9 species on the island); chiefly at low altitudes; on tree tnmks etc., in rain forest
Catascopus: moderate radiation in tcaUacei group (a primarily New Guinean group of 5 species) and perhaps some other groups; at low and middle altitudes; on tree trunks in rain forest
Dolichoctis: moderate radiation especially in aculeata + polita groups (about 10 interrelated species on New Guinea); chiefly at low altitudes; in lower-story foliage in rain forest
Agonochila: 7 apparently interrelated, small species on New Guinea; chiefly at mid-altitudes; habitat unreported but probably arboreal
Scopodes: moderate species-radiation (7 interrelated species in special New Guinean group); at mid- altitudes; probably on logs etc., in montane forest
Dicraspeda: genus primarily New Guinean (6 rather diverse species on the island); chiefly at low altitudes ( 1 at mid-altitudes ) ; all probably in imder-story foliage of rain forest
Pogonoglossus: 9 species on New Guinea, interrelationships not detennined but probably in part product of local radiation; at low and middle altitudes; in leaf litter etc., in rain forest
Helluonidius-Helluopapua: probably related genera with together 6 species probably product of radia- tion on New Guinea; at low and middle altitudes; habitat not reported but probably on the ground or possibly on tree trunks
smaller than the lowland ones, and in most cases I find no general increase in size in comparing the same or related species at different altitudes (see first paragraph of present section, and [20]).
The mountains of New Guinea are, I think, a magnificent natural laboratory for the study of evolution. Selective forces are probably intense there; the adaptive changes that occur are profound; and populations are probably often confined to limited areas within which they can be measured and perhaps even manipu- lated. As a taxonomist doing second-stage
faunal work (Part I, pp. 328-330), I can emphasize this situation, but I cannot go further in investigation of it.
[91] Evolution and adaptation of sepa- rate carabid stocks. Within the main pro- cess of continual accumulation, survival, e\olution, extinction, and overturn of the fauna as a \\ hole, some particular carabid stocks have had notable evolutionary his- tories on New Guinea. In some cases the histories have involved mainly multipli- cations of species, in others more or less extensive adaptive radiations into new habitats.
244 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Speciation with relatively little adaptive radiation seems to have oceiirred in Demetrida (Part III, p. 140), in which 59 known species have evolved on New Guinea from one or a very few ancestors. Some species of this genus have differenti- ated at different altitudes and probably in slightly different niches, but the ecologic range of the genus as a whole is (in New Guinea) probably confined to foliage chiefly in rain forest, with perhaps a minor invasion of grass at high altitudes. Less extensive but still notable species-radiations that have occurred partly or wholly within the limits of New Guinea are listed in Table 17. Most of them have occurred in rain forest. Note that Agonini are not included in this table; their radiations will be considered separately [92].
Adaptive radiations into new habitats are (except in the Agonini) fewer and less easy to demonstrate than multiplications of species. The extent of adaptive radiation among New Guinean Demetrida, for ex- ample, although probably not great, is in fact unknown, because the actual habitats of many of the species are unknown. In two other groups limited adaptive radiation is suggested but not yet actually known. In one of them, Dicraspeda (Part III, p. 210), the six species all occur in under- story foliage of rain forest, but striking differences in size, color, and tarsal struc- ture suggest that different species occupy different niches within the rain-forest- foliage habitat. And in Helhionidiiis/IIel- luopapua the six known species exhibit such striking differences in tarsal and antennal structure as to suggest ecologic differentiation, although the habitats of the species are unknown.
A limited but noteworthy ecologic shift from water-side habitats to the floor of rain forest seems to have occurred in several stocks of New Guinean Garabidae (cf. [84]). In Platycoehis, for example (Part I, pp. 541ff), most species are found in wet places, but P. depressu.^ apparently lives in leaf litter on the ground in rain
forest, and it is the only member of the genus in which the wings have begun to atrophy (the species is now ±winged), loss of wings probably being correlated with shift from a less to a more stable habitat. In Loxandrus (Part III, pp. 549ff), most species arc closely associated with water, but L. latu.s probably occurs in rain- forest leaf litter; this species too is ±winged, apparently becoming flightless in the more stable habitat. In 'Nota<ionum, most species are found in wet places (or in montane habitats), but N. spinulum (Part II, p. 157) lives in rain-forest leaf litter; it is still fully winged. And in the mainly subaquatic genus Oodes, O. ter- restris (see Tax. stippL; — hevissimus of Part I, p. 34) has left the water and now occurs commonly and (in my experience) exclusively in rain-forest leaf litter. Wing reduction may have begun in this species (the wings are slightly reduced in some individuals), and two related species local- ized in different places in New Guinea are -winged, suggesting that geographic dif- ferentiation of the flightless terrestris stock has begun.
In two other cases species of Garabidae belonging to primarily hydrophile groups have become adapted to drier habitats in opener places rather than in rain forest. Tliey are Tachijs aeneus (Part I, p. 463) and E^adroma robiista (Part III, p. 71). However, their ecologic shifts may not have occurred in New Guinea, for the spe- cies concerned have extensive ranges out- side as well as inside the island.
[92] Afs^onine radiation on New Guinea. The radiation of Agonini on New Guinea (Part II; present part. Tax. siippl.) goes far beyond that of any other group of Garabidae. Although a number of agonines have probably reached the island inde- pendently from time to time ( a few species of Arhytinus, Euplenes, Dicranoncus, Loro- stemma, A^onuin, Notag^omim, and Col- podes are not or not much differentiated and have probably come in recently), the great majority of the 160 species of the
The carabid beetles of New Guinea • Darlington 245
tribe known on New Guinea have appar- subarboreal; this ecologic radiation has ently evolved from very few ancestors, apparently occurred within the limits of The principal ancestors were generalized, the Bismarck Range. Of other mountain- tropical, lowland agonines of moderate living genera that may have been derived size, fully winged, and with full comple- from Notafiommi-like ancestors, Gastrago- ments of standard setae. (In fact it is still num apparently includes mesophiles living true that all Agonini that occur at low alti- in relatively open places; Altagonum, tudes on New Guinea are winged, and Montagonum, Laevagonum, and Fortago- almost all the lowland species still have num, mesophiles living mostly in montane full counts of setae. ) Such agonines, placed rain forest; Potamagonum, hydrophiles liv- in the "genus of convenience" Nofagomim, ing beside turbulent mountain brooks; and are still numerous especially at low alti- Maculagonum, perhaps subarboreal species tudes in New Guinea. They are presumably living at least partly in grass. Tliis is a derived from Oriental stocks, but the very inadequate description of the extent details of their relationships have not yet and diversity of radiations of Agonini in been worked out. Such relatively un- New Guinea. In fact, the habitats and specialized stocks of agonines have entered habits of most of the mountain species are every ordinary New Guinean habitat, and unknown (but can often be deduced from have evolved in many directions. At low the insects' structure or from what is altitudes, different hydrophiles have be- known of the localities where they occur), come adapted to deep swamps, to the and very many — perhaps hundreds — of banks of rapidly flowing streams, and to high-mountain species probably still re- various more-ordinary water-side situations, main to be discovered, and some (Indogonum as well as Notago- Although the Agonini have diversified num spinuhim) have invaded leaf litter so much more than other Carabidae on on the floor of rain forests. (Some are New Guinea, I do not think they need be arboreal, but most arboreal agonines at low older than the others. Their jjresent altitudes in New Guinea, including several patterns of complex and often close inter- Colpodes, have probably reached the relationships suggest an explosively radiat- island independently and are not derived ing group rather than an old one. And from Notagomim-Mke ancestors. ) From the their patterns of distribution on the moun- lowlands, different Agonini have evidently tains of New Guinea suggest rapid evo- invaded the mountains at different times, lution in situ, with clusters of distinct but Some mountain-living forms are still only related species and even small genera often subspecifically differentiated from their confined to single supposedly geologically lowland relatives (e. g., ]Voto,f,'onf/m f/enfe/- recent ranges. I therefore think that the lum chimbu, Part II, p. 149); others are extraordinary radiation of Agonini on New distinct species of primarily lowland genera Guinea indicates not age but inherent (<?. g., Notagonum altum. Part II, f). 144); dominance derived from characteristics — others have differentiated generically, often general adaptations [68] — which pre-adapt changing their form and often losing their the insects for success in diverse situations, wings and some of their standard setae. What these general characteristics of Some of the alticoline genera have speci- Agonini are, whether structures or physio- ated or even radiated ecologically in small logical processes or behavior patterns, I areas at high altitudes. The best known cannot even guess. Study of the living example is Nehriagonum (Part II, p. 235), insects both in their habitats and in the 3f which six species now occur on the laboratory may suggest them. Bismarck Range, some of them being [93] Evolutionary trends: not toward in- nesophiles, some hydrophiles, and one crease of size. Several different, presum-
246 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
ably adaptive evolutionary trends shared by several or many different carabid stocks on New Guinea can be detected, and one trend that has not occurred is noteworthy too.
A trend toward increase of size has not occurred in evolution of Carabidae on New Guinea. I base this statement on comparisons (made during the course of my taxonomic work) of many New Guin- ean Carabidae with closely related forms in southern Asia and in Australia. The aver- age size of mountain-living Carabidae on New Guinea is greater than the average size of lowland forms, but this is appar- ently a result of a deficiency of small species at high altitudes, not a result of increase in size during evolution of moun- tain-living stocks [90]. Certainly no trend toward gigantism has occurred among Carabidae at any altitude on New Guinea. There is no indication here of evolution toward the situation on Madagascar, where very large carabids dominate the fauna.
[94] Atrophy of wings and associated trends. A trend toward atrophy of wings is conspicuous among New Guinean Carab- idae, but only on the mountains. Wing atrophy is insignificant among the lowland forms, but increases with increasing alti- tude until something like 95 per cent of the carabids on the highest momitain-tops have atrophied wings. The state of wings, the incidence of wing atrophy, and the process of atrophy have already been sufficiently discussed [21-23]. Now, I want merely to re-emphasize two points. The first is that the atrophy process is ap- parently inhibited in most cases at low altitudes on New Guinea, so that the low- land carabid fauna continues to consist almost entirely of small, winged species. Such species are pre-adapted to dispersal [88]. Their predominance at low altitudes in New Guinea might mean that small, winged carabid stocks come in in such numbers, and that the lowland fauna is overturned so rapidly, that species do not have time to evolve either large size or
flightlessness at low altitudes on the island. However, I think it is more likely that small, winged Carabidae dominate the lowland fauna in New Guinea simply be- cause they are well adapted to conditions there, as they are to conditions at low altitudes in the tropics on continents [21].
The second point is that the very strong trend toward atrophy of wings at high altitudes has been accompanied or fol- lowed by secondary trends. Some are direct results of wing atrophy, which di- rectly causes or favors shortening of the metathorax and co-aptive shortening of the elytra and (sometimes) narrowing of the elytral humeri. The elytra also tend to become locked together along the suture. Less direct effects of wing atrophy include, often but not always, a reduction in size of eyes. The changes of form and reduction of eyes together tend to produce a more compact, less active carabid \\'hich often can be recognized at a glance as flightless.
[95] Loss of setae. A strong trend toward loss of certain setae (specified below) parallels atrophy of wings. Loss of setae, like loss of wings, is correlated with alti- tude, and in fact the two processes are to some extent correlated with each other. Setae are not often lost by active, winged Carabidae even at considerable altitudes; no loss of setae has occurred among the (winged) alticoline species of Notagonum, and most of the numerous (winged) alti- coline species of Altagonum have lost only one pair of pronotal and sometimes one pair of elytral setae (although additional setae have disappeared in a few species of this genus). But in alticoline stocks that have lost their wings and have been other- wise modified [94], setae have disappeared more often and in a variety of patterns.
The setae concerned are the two pairs of supraoculars, the two pairs of lateral pnniotal setae, and the three standard setae on the third interval of each elytron. The anterior supraoculars are lost often; the posterior, rarely, and only when the an-l' terior ones are lost too (in New Guinea,
The carabid beetles of New Guinea • Darlington 247
only in Fortagonum bttfo and Perigona pronotal setae). In some cases loss of rex). Of the lateral pronotal setae, the particular setae characterizes species or median (anterior) pair are often lost; the genera; in other cases setae vary individu- posterior pair, rarely, unless the median ally in single populations (see especially pair are lost too (exceptions are Lithago- Prothorox and Elytra of Nebriagonum num, Part II, p. 176, and Laevaganum sub- cepholum, and Notes under Fortagonum citum, Part II, p. 245), but both pairs have fortellum). Loss of setae is usually bi- disappeared in some cases. Of the setae laterally symmetric, although unsymmetric (seta-bearing punctures) of the third in occasional individuals. And in some elytral intervals, sometimes only the an- cases, but apparently not usually, loss of terior one on each elytron is lost, leaving different pairs of setae may be genetically the elytra together 4-punctate (e. g., in correlated: e. g., cases have been found (in Iridagonum quadriptmctum, Altagomim the West Indies and Australia) in which vaUicola and grosstthim, and some Macu- single mutations eliminate both the anterior lagonum); sometimes the anterior and supraocular and the median-lateral pronotal middle ones are lost, leaving each third setae, or both the median-lateral pronotal interval with only the posterior puncture and anterior elytral setae (see [100]). and seta {e. g., in some Maculagonum); Why setae should be lost so often, sometimes these elytral setae disappear in especially by mountain-living Carabidae some other order or are inconstant in single that have lost their wings, is difficult to species (e. g., in Nebriagonum cephahim say. The setae are probably tactile, and and transitum and FoHagonum fortellum); they may tend to be lost by beetles which and in some cases all these elytral punc- have become relatively sluggish and which tures have been lost (e. g., in Altagomim may have less need of tactile warnings exutum and fotuum, Nebriagonum perce- than small, winged, active carabids do. It phalum, and most Laevagonum). In the is also possible that presence of setae has extreme case (only in Fortagonum bufo, some outright disadvantage — possibly as on New Guinea), all these standard setae foci of infection, especially in humid situ- of head, pronotum, and elytra have been ations in the cloud zone on mountains — lost, leaving the insect without dorsal setae, which accelerates their loss when the need Tliis is an incomplete survey of the for them is lost. But loss of setae is some- extent of loss of setae by New Guinean thing to be investigated by ecologists, be- Carabidae and of the diversity of setal haviorists, and physiologists. I can only patterns that has resulted. Most of the loss emphasize the fact of the loss. I cannot has occurred in mountain-living forms, but satisfactorily explain it. some loss of setae has occurred in lowland [96] Modification of legs and tarsi. Evo- species too. Most of the loss has occurred lutionary trends in modification of the legs in flightless (-winged) species, but some and tarsi of New Guinean Carabidae setae have sometimes been lost in winged especially on mountains surely exist. One species too. And most of the loss has oc- trend is toward loss of lobes of the fourth curred in Agonini, but this is probably hind-tarsal segments. These segments are because most mountain-living carabids in often (not always) long-lobed in active New Guinea are agonines; setae have been lowland Carabidae, especially arboreal lost by mountain-living species of some ones, and the lobes often are shorter or other tribes in New Guinea too {e. g., of absent in less active, ground-living, mon- Pterostichini, Rhytiferonia, Analoma, and tane forms. The legs may be shortened, some Lesticus have lost the setae of the too; perhaps this shortening is a secondary third intei-vals, and of Perigonini, Perigona trend following wing atrophy in some rex has lost the supraocular and lateral cases. And the clothing of the lower sur-
248 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
face of the tarsi, and the accessory setae of the fifth tarsal segments, are often modified or partly lost. These details are so diverse among New Guinean Carabidae, and the evolutionary trends are so complex and so obscure, that I shall not attempt to say more about them now, but can only hope that persons who have the oppor- tunity will consider them in more detail in the future.
[97] Modification of ecoloiiy and be- havior. Besides trends in evolution (espe- cially loss) of structures, trends in ecology and behavior probably occur among New Guinean Carabidae, especially on moun- tains. One trend is probably from water- side to mesophile habitats, especially to the floor of rain forest. Several lowland stocks have made this ecologic shift [91], and montane stocks have probably made it too. However, the montane Carabidae are so complexly interrelated and their habitats are so little known that I cannot give details.
Connected with the trend into meso- phile habitats may be a trend toward reduced activity and possibly a trend to- ward increasingly nocturnal habits, at least in montane forests. I base this very tenta- tive suggestion on what I saw of carabid behavior during the less than two weeks that I collected on the Bismarck Range [3].
Finally, there are presumably trends in density and distribution patterns of popu- lations correlated with habitats and with altitude. At low altitudes in New Guinea, carabid populations probably tend to be sparse and patchy. This seems to be the case especially in the stable environment of the lowland rain forest, which suggests a trend toward patchy distributions there. Maintenance of wings by most lowland Carabidae in New Guinea is perhaps an adaptation to sparse, patchy distribution of populations [22]. However, with in- creasing altitude this trend is probably reversed. Species (populations) at higher altitudes often have very small total ranges
Table 18. New Guinean Carabidae with small
BUT abruptly prominent EYES
( Agonini)
Notagonum reversutn group (eyes abrupt in 3 of the 4 species; present part, Fig. 28)
Iridagontim (2 of the 7 species, subfusum and vigil)
Maculagomim (3 of the 13 species)
Gastragonum ( 1 of the 6 species, laevisculptum )
Idiagomnn (all 6 species)
Montagomim (1 of the 8 species, toxopeantim)
Nebriagotitnii ( some )
Fortagonum (some, especially Umum and cJistor-
tiim; present part, Fig. 76) ( Non-agonines )
Dolichoctis distorta (Part III, Fig. 81)
Demetrida vigil (eyes abrupt but scarcely "popped"; Part III, Fig. 107)
(a single mountain-top, against the whole of New Guinea for many lowland species), but within their ranges these species pre- sumably maintain relatively dense, rel- atively continuous populations, which can survive without flight. I think a trend like this, toward increasing concentration of individuals in populations at increasing altitudes, must occur, although (as far as I know) ecologists have not yet demon- strated it, at least not among Carabidae.
[98] Parallelism and convergence; de- velopment of ehjtral spines; color patterns. Multiple parallelisms and/or convergences that have occurred among New Guinean Carabidae in atrophy of wings, in changes that follow wing atrophy, in loss of setae, and in modification of tarsi are described or implied in preceding pages. Some other structures or patterns \\'hich have been duplicated in the evolutions of different New Guinean Carabidae are worth noting. For example, the eyes have been reduced, in many different stocks, especially in flightless ones, and, although the reduced eyes are usually flattened, they have become abnormally abruptly prominent ("popped") in about eight separate cases
The carabid beetles of New Guinea • Darlington 249
Table 19. Apparently separate stocks of New Guinean Carabidae with elytral spines
( Agonini )
Tarsagomim latipes (Part II, Fig. 1)
Notagonum suhrufum
It spinulum
II sectum (present part, Fig. 33)
Violagonum violaceum Colpodes s. sloanei
It hellito (present part, Fig. 35)
II rex
II aniedens
Plicagonum nigifrons (spines short, present only in some individuals )
Iridagomim quadripunctum (some individuals)
Altagonum tiittnn
II cracens (present part, Fig. 44)
avium (present part. Fig. 45) cheesmani, sororium (present part. Figs. 39, 40) scapha, regiscapha erugatum (present part. Fig. 46) steUaris (present part. Fig. 43) higenum, suhconicoUe (spines short; present part. Figs. 47, 48)
Potamagonum diaphanum (spines short, present only in some individuals; Part II, Fig. 8), etc.
Nebriagonum arhoreum (spines short)
(Licinini)
Omestes torta (spines short; Part III, Fis. 5)
II It
II II It
It
Dicrochile acuta ( spine short, present only in some individuals; Part III, Fig. 6)
( Lebiini)
Stenotelits spinosus (Part III, Fig. 44) Catascopus lattis (Part III, Fig. 60) It luevigatus
It sidits (Part III, Fig. 61)
II snmragduhis (some individuals)
dohodura (Part III, Fig. 62), hiroi icallacci, etc. (Part III, Figs. 63, 64) Pericaliis jiguratus (Part III, Fig. 65)
Dolichoctis distorta, aculeata group, polita group (Part III, Figs. 81-85)
Stricklandia pericalloides, lata (Part III, Fig. 86) Demetrida tessehta (Part III, Fig. 98) It genicula (Part III, Fig. 101)
It many other species (Part III, Figs.
103-109, VI, IX-XII)
( Odacanthini )
Clarencia qiiadridem (spines short; Part III, Fig. 128)
Dicraspeda bispinosa (some individuals; Part III, Fig. 129) 11 quadrispinosa, violacca
Dobodtira armata (Part III, Fig. 131)
among New Guinean Agonini (Table 18; see also Part II, p. 96) and in the lebiine Dolichoctis distorta; most of the carabids concerned are mountain-living, but Doli- choctis distorta is a lowland species. Con- vergence of body form, especially evolution of a strikingly fusiform body outline, has occurred in isolated species of at least four genera of Agonini ( listed in Part II, p. 95 ) . And ventral pubescence has been de- veloped in probably ten independent cases in New Guinean Agonini (listed in Part II, pp. 100-101). Abruptly prominent eyes, fusiform body, and ventral pubescence are presumably adaptive, but I do not know their function among New Guinean Carab- idae. However, something more can be said about development of elytral spines. Most Carabidae have the apices of the
elytra rounded or oblique or simply sinuate, but some have the sutural angles dentic- ulate or the apices angulate, and some have evolved spines either as elongated denticles at the sutural angles or as acute prolongations of the apical angulations approximately opposite the ends of the third intervals, or at the outer-apical angles, or elsewhere. Elytral spines have apparently been evolved indepcndenthj in at least 21 separate stocks of Agonini, 13 of Lebiini, two of Licinini, and four of Odacanthini listed in Table 19. Tlie total number of cases in which elytral spines have evolved among New Guinean Carab- idae is at least 40, and in most of these cases the spined forms have apparently originated on the island from spineless ancestors.
250 Bulleiin Museum of Comparative Zoology, Vol. 142, No. 2
Elytral spines have evolved relatively Carabidae. Most striking is development
often in Agonini and Lebiini but rarely of patterns formed by short longitudinal
in some other large tribes. For example, pale lines arranged in three transverse
New Guinean Pterostichini are never series on a dark ground. These patterns
spined, the nearest approach being in occur in some (not all) species of small
Lesticus p,racilLs (Part I, p. 524), in which lebiines of the not-directly-related genera
the elytral apices are pointed or denticulate MiniitJiodes, Coptodem, and Agonochiki
but not much produced. And New Guin- (cf. Part III, Figs. 48, 68, 72, etc.). The
can Harpalini, although numerous and pertinent species of the first two of these
dominant, are never spined, the nearest genera occur on the trunks and branches
approach being in CoIeoUs.nis papua (Part of trees, and the A'^onochila may do so too,
III, p. 65, Fig. 26), in which the sutural although I do not know the habits of this
angles are acutely denticulate. It is a genus in New Guinea. This convergence
ciuestion whether the tendency of some of patterns may be mimetic (see below),
tribes to evolve spines more often than Also possibly mimetic may be the striking
others reflects genetic differences or dif- coloration (red head and prothorax, blue
ferences in ecology or behavior. Most or elytra) of the rare lebiine Phloeocarahiis
all of the spined Lebiini and some of the euplcnes (Part III, p. 184), which differs
spined Agonini are arboreal, while no New strikingly from other members of its genus
Guinean Pterostichini or Harpalini are fully but resembles in color the relatively com-
arboreal. However, some other spined mon agonine Euplenes apicalis. A pattern
Agonini {Tarsagonum at low altitudes and of fine light and dark speckling has evolved
probably some of the mountain-living convergently in Moculafionitm (Agonini)
fornis) are ground-living; the spined and in certain Demetrida (Lebiini), e. g.,
licinine Omestes lives on the ground in D. paUens (Part III, PI. 1, Fig. II). This
swamps; and Dobodtim (one of the most pattern may be not mimetic but cryptic in
strongly spined of all New Guinean Carab- places where light and shade are finely
idae) lives among spray-drenched stones divided, for example in grass. A different
beside turbulent brooks. The correlation pattern of convergence occurs in Chluenhis,
of spined elytra with arboreal habits is of which six species that normally have
therefore not complete. pale markings on the elytra elsewhere in
Elytral spines presumably tend to pro- their extensive ranges are losing or have'
tect the insects against predators, including lost the markings in New Guinea (see Part'
lizards and birds. It may be significant III, p. 22, last paragraph of first column),
that, although spines have been developed These species lix'c in somewhat diverse
in many different stocks of Agonini on New habitats (although all on the ground, of
Guinea, there are no spined species among course), and I think this loss of elytral
the very numerous Agonini on the Ha- markings is probably not mimetic but may
waiian Islands (E. C. Zimmerman, personal be the result of ecologic factors acting
communication, 1968). Hawaii, of course, similarly on the several species. Conver-
lacks native terrestrial predators (frogs gence of an elytral color pattern consisting
and lizards), although some insectivorous of a conspicuous pale square mark before
birds are present. On the other hand spines the apex of the elytra in some Trichotich-
are very rarely developed among Agonini mis (Harpalini), e. <i,., T. g,iittula (Part III,,
or any other Carabidae on the West Indies, p. 57), and in some individuah of Altago-^
where frogs and insectivorous lizards are ntiin grossidum or grossuloide.s (Agonini)
numerous. (see present Taxonomic supplement) may
Some convergence of elytral color pat- be due to parallel mutations and may
terns has occurred among New Guinean have no direct adaptive significance.
The carabid beetles of New Guinea • Darlington 251
[99] Mimicry. The parallelisms or con- vergences discussed in preceding para- graphs are of several sorts. Some, including wing atrophy, the secondary structural changes that often follow it, and loss of setae are probably at least in part adapta- tions to complex new environmental factors encountered especially at higher altitudes. The development of spines may be directly protective. Speckled coloration may be cryptic in special situations. Loss of color patterns by Chlaeniiis may be a selective response to a relatively simple (but un- known ) climatic factor. Convergence in some details of markings may be due to parallel mutation. And parallelism or con- vergence in color pattern of certain small Lebiini and of Phloeocarahus euplenes may be mimetic.
Mimicry has recently been usefully re- discussed by Wickler ( 1968 ) . It is surely very common among insects in the tropics. Nevertheless, it is not always easy to decide whether particular resemblances are mimetic or not. After seeing many New Guinean Carabidae alive (mimicry is partly a matter of behavior of living in- dividuals), I doubt if mimicry is common among them, but I can suggest the follo\\'- ing possible cases for consideration in the future. Mimicry may be involved not only in the convergence of color patterns of the small, tree-trunk-living Lebiini referred to above, but also in the evolution of form and color of some Detnetrida, perhaps most likely in the similarity of D. imitatrix to the very common, spined (and perhaps For other reasons protected) Violap^onum viohcciim (see Part III, pp. 176-177). The iintlike form of Colasidia ( present part. Tax. suppJ.) may also be mimetic. (And the antlike tiger beetle Tricondyki may derive some advantage from its similarity to a large ant, although I am not sure that my ants comparable to Tricondyla in size uid behavior occur in New Guinea, and ilthough Tricondyla itself is a fomiidable insect. )
[100] Mutation and dimorphism. Mu-
tations resulting in Mendelian dimorphism have apparently occurred among New Guinean Carabidae. Known cases are worth listing and describing, for mutation and dimorphism sometimes mark the be- ginning of important evolutionary processes and sometimes affect characters supposedly important in taxonomy.
Mutation from a long- to a short-winged condition, and resultant dimorphism of wings, have been discussed in preceding sections [21-23]. Change from a long- to a short-winged condition probably usually or always begins (among Carabidae) with mutation, and dimorphism of wings is com- mon. Lowland New Guinean Carabidae known to have dimorphic wings are indi- cated in Table 3, section [21]. Altliough wings have atrophied much more often at high than at low altitudes, actual dimoq^h- ism is rare at high altitudes in New Guinea. The only Carabidae known to be wing- dimoqDhic at high altitudes on New Guinea are Trichotichnus niiiricans (fully winged at low altitudes but dimorphic on the Bismarck Range), T. alius (a mountain- living species which may be fully winged at some localities but is dimorphic at others), and Gastrap,omnn ferrestre (which is dimorphic on the Bismarck Range). Even most genera of Carabidae on the high mountains of New Guinea are uni- formly long-winged or uniformly short- winged. The only known exceptions are Ga.stra'^onum, a natural genus which in- cludes +w, ±w, and -w species, all lixing at considerable altitudes, and Notagonum, a "genus of convenience" of many winged species but including the "reversum group" (present part. Tax. sti))pL), a natural group of four mountain-lixing species of which three are long- and one short- winged. This general situation suggests that, \\'hile some Carabidae at low alti- tudes in New Guinea are still unstable as to wings, with mutations continually oc- curring and dimorphism becoming estab- lished at least for short periods, the situation is more stable at higher altitudes.
252 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Mutations presumably continue to occur at high altitudes too, but the wings in each group have apparently become adapted to the way of life of that group and are usually no longer liable to establishment of dimorphism and further evolutionary change. This is to suggest what may well be true, that some carabid populations at low altitudes in New Guinea are relatively unstable and liable to change or overturn, while many of the alticoline groups are more stable, more exactly adapted to spe- cial habitats, and less liable to significant changes or to overturns.
Occurrence of setae is known to be dimorphic in one species of carabid on New Guinea. It is Feria^ona a.^rolahica (Part III, p. 9), in which the posterior- lateral pronotal setae are present or absent. The absence is presumably due to a single mutation inherited in Mendelian fashion. ( Similar mutations may have produced the same pattern of reduction of pronotal setae in Lithaiionum (Part II, p. 176) and Laevo{i,onum suhcitum (Part II, p. 245).) Strict dimorphism of setae is not known in any other Carabidae on New Guinea. (In some species especially of Nebriago- num occurrence of setae varies erratically but is not simply dimorphic.) However, patterns of setae do vary strikingly from species to species in some genera, and it seems likely that the variations begin by mutation and that the mutations are fol- lowed by seta-dimorphism at least for short periods.
Outside New Guinea, seta-dimorphism has been found in a few Carabidae. In the lebiine Phloeoxena deahita Darlington (1937: 136) of Guba, the anterior supra- ocular setae and median-lateral pronotal setae are apparently inherited as a group, these four setae being either all present or all absent in all of 33 individuals from a single mountain top. Tliis dimoqihism presumably originated by a single mutation. A similar mutation may have produced the same pattern of seta-loss in ^^Alta^onum" bigenum (present part, Tax. suppL). In
Notaii.onu7n madcaiji (Sloanc) of North Queensland, the anterior-lateral (=median- lateral) pronotal setae and the anterior seta-bearing punctures of the third elytral intervals are apparently inherited as a group: all four of these setae are present or all absent in all of 48 individuals, with one partial exception (see Darlington, 1963: 4 for further details ) . In this case, too, ab- sence of the setae is probably due to one Mendelian mutation. Homologous muta- tions may have produced the same pattern of loss of setae found in New Guinea in Altogoniim vallicola and grossulmn and Iridogontim quadripunctum.
Dimorphism of elytral color pattern ap- parently occurs in Mimithodes sexiialis and perhaps in M. papuana (Part III, pp. 99, 97), in Demctrida diversa (Part III, p. 172) and perhaps other species of Deme- irida, and possibly but less obviously in some other New Guinean carabids includ- ing perhaps Tachtjs acuticollis (see present' part. Tax. suppl. ) .
In summary, mutations from a long- to a short-winged condition probably occur commonly among New Guinean Carabidae, and wing dimorphism, presmnably follow- ing mutation, has been found in some species, mostly lowland forms. Mutations eliminating setae arc probably common too and are probably diverse, producing many different patterns of setal reduction, al- though knowai cases of strict dimorpliism of setae are few. And mutations are probably continually affecting a variety of other characters, including color patterns, although they arc hard to detect in most
I
cases.
State of wings and patterns of setae are sometimes important taxonomic characters. If mutations affect them, as they do, they probably affect many other characters used by taxonomists. It is therefore of practical taxonomic importance as well as of theo- retical evolutionary interest to detect oc- currences of mutation and dimorphism in nature. Their occasional occurence does not spoil the taxonomic usefulness of the
The carabid beetles of New Guinea • Darlington 253
I'haracters concerned, provided proper al- owance is made for them (cf. Nature of 'axonomic characters [12]).
Variation which is presumably primarily genetic, beginning with mutation and pass- ng through stages of dimorphism (al- hough the dimorphism may be obscure or jrief ) occurs even in characters supposedly )f generic and tribal importance. For ex- imple, presence of a subapical fold of the jlytral margin in Pterostichini and its ibsence in Agonini distinguish most mem- )ers of these tribes. However, the fold is variable or lost in a few pterostichines. It s present or absent in different individuals )f AnaJoma fortis (Part I, p. 539); present n Lestictis ambulator but absent in the )resumably related L. toxopei: and variable n the Loxandrus-Homalonesiota-Nebrio- eronia-HapIoferonia group of genera (Part , pp. 547ff; present part. Tax. suppL).
Number of supraocular setae — whether me or two on each side — is a useful and, vith proper allowance for exceptions, latural character for distinguishing some ribes of Carabidae. The anterior supra- )culars are often lost, sometimes evidently )y mutation (see above), and the posterior etae then usually remain, usually near or liagonally behind the posterior corners of he eyes. To the experienced carabid axonomist this position of the posterior upraocular setae indicates that the species n question are derived from and belong n tribes which normally have two pairs )f supraoculars. (Of course this guide fails n the cases, very few in New Guinea, in vhich both pairs of supraocular setae have )een lost.) However, in the special case of iliscelus (Part HI, p. 91, Figs. 168, 169) he supraocular setae behave differently. "^o one species of this genus is dimorphic, o far as I know, but different species that ire otherwise extremely similar have either )ne or two setae over each eye, and when )nly one setae is present, it is in median )osition, not in the normal position of the posterior setae. Miscelus is a taxonomically solated carabid, placed sometimes in the
tribe Lebiini and sometimes in a tribe of its own. Regardless of its evolutionary history ( which I do not know ) , it is at least a striking example of an exception, which has probably originated by mutation, in a character ordinarily of tribal value.
Modification of the clothing of the male tarsi in Lijtcr (Part IH, p. 63) is another example of an exception to a character usually of broad taxonomic significance. In most Harpalini the males have the lower surface of the front and often also of the middle tarsi clothed either with two rows of broad scales or with a dense pad of very narrow scales; this difference is the basis for distinguishing major subtribes of Harpalini (Part III, p. 40, in Key). In Lifter, however, the male tarsi are loosely clothed with several irregular rows of moderately slender scales. This condition is intermediate between the 2-seriate and densely padded types of tarsal clothing. Lyter is probably derived from an ancestor, perhaps a Trichotichnus, with two rows of male tarsal scales. No transitional con- ditions have been found in the numerovis species of Trichotichnus that I have seen from New Guinea and the Orient; the change from 2-seriatc to multiseriate tarsal clothing may therefore have been by mu- tation. And abrupt, perhaps mutational loss of the scales of the male middle (not front) tarsi is suggested in Trichotichnus semimas and Harpaloxenus fortis ( Part III, pp. 52, 60), in both of which middle-tarsal scales are absent although present in ap- parently closely related species on New Guinea. (Mutation should be suspected, too, in loss of male tarsal clothing by various other Harpalini outside New Guinea. )
Mutation is to be suspected and di- morphism looked for in other taxonomic characters that vary abruptly. Striking dif- ferences in presence or absence of lobes on the fourth tarsal segments of different species of Notuii^onum (Part II, pp. 127ff, especially p. 128) and different species of Dicraspeda (Part III, pp. 210ff) may be
254
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
in part mutational. The difference between the pectinate claws of De-^era and the simple claws of Drypta ( Part III, pp. 216ff, especially p. 218) may be mutational. And abrupt differences in some details of the male genitalia may be mutational although, because I have made little use of genitalic characters in the present work, I cannot give examples from New Guinean Carab- idae. All these cases need further study based on or following more thorough, third- stage taxonomic treatment. For the moment I can only do what I have had to do else- where in preceding pages: point out that important problems exist, although I can- not solve them.
TAXONOMIC SECTION
[101] Tribal classifications. I should like to have included here a classification and key to the tribes of Carabidae that occur in New Guinea, but I have decided against it. To make such a classification based on New Guinean Carabidae would be labo- rious, and would not be a very important contribution to carabid classification, since the New Guinean fauna is very limited. In fact it includes representatives of only about one-third of existing carabid tribes
[18]. For practical identification of tribes of New Guinean carabids, the best key is probably Andrewes' (1929: 43-46), which is based on tropical Asiatic fonns and which includes, I think, all tribes that are represented in New Guinea. This key is only four pages long, and interested per- sons can easily secure copies of it. Sloane's (1923) "Classification of the family Carab- idae" is concerned with the Australian fonns, and can be tried for New Guinean ones when Andrewes' table fails. Ball's ( 1960 ) classification of Carabidae includes many improvements but is concerned pri- marily with the North American fauna and therefore omits some tribes that are found in New Guinea. Jeannel's (1941-1942) classification (in Faunc de France, Part I. pp. 9-10, 77-88, supplemented by foot- notes scattered throughout the work) is important and should be studied by all persons interested in tribal classifications of Carabidae, but it seems to me that it splits families, tribes, and lower groups fai beyond practical usefulness. And Basi- lewsky ( 1953 ) , in keys to higher categories^ of Carabidae represented in Africa anc Madagascar, follows Jeannel's classification in a general way but makes some changes,
I.
The carabid beetles of New Guinea • Darlington 255
TAXONOMIC SUPPLEMENT
This supplement consists of important new records and new species of Carabidae received from New Guinea too late to be in- cluded in preceding parts of my work. Many tribes, genera, and species about which I have nothing important to add are omitted. For some other tribes and genera only an indication of additional material available for future study is given, usually with a list of the species that have been found at Wau, which has become a locality of special importance. But in still other groups, notably some mountain-living Ptero- stichini and Agonini, new material necessi- tates complete revisions of genera, with new keys and complete lists of species. Refer- 3nces are usually limited to Parts I, II, and [II of the present work, where additional references will usually be found.
Subfamily CICINDELINAE
Darlington 1962, Part I, p. 333.
See Rivalier (1950-1963), "Demembre- ment du genre Cicindela L. ," especially parts IV (1961), "Faune indo-malaise," and V (1963), "Faune australienne." This is an important contribution to understanding 3f the Cicindela (sensu loto) of New Guinea and adjacent areas. I have not myself anything further to say about New Guinean tiger beetles.
Subfamily CARABINAE
Tribe OZAENINI
Genus PSEUDOZAENA Castelnau
Pseudozaena orientalis opaca (Chaudoir)
Darlington 1962, Part I, p. 352.
Additional material. Seventy-two, from numerous localities, including 22 from Wau, 1200 m (none higher); and 4, Waigeo Is., Gamp Nok, 2500^ft. (c. 760 m), Apr. 1938 ( Cheesman ) .
Tribe SCARITINI
3enus GEOSCAPTUS Chaudoir
3eoscapfus cacus (Macleay)
Darlington 1962, Part I, p. 356.
Additional material. N-E. N. G.: 3, Main River, Sepik, Feb. 1965 ( Hornabrook ) . West N. G.: 1, Ifar, Cyclops Mts., 300- 500 m, June 2.3-25,1962 (Sedlacek).
Genus CLIVINA Latreille
Darlington 1962, Part I, p. 358.
Notes. Of this genus, 218 additional specimens have been received from New Guinea. They include the following species from Wau and vicinity:
fcssa Darlington : 1200 m ( many ) , 1500 m
(1) brandti Darlington: 1200 m (1)
szekessiji Kult: 1200 m ( 2 )
Besides these, tripuncta Darlington and sid)- fusa Darlington might be expected to oc- cur at Wau but have not yet been found there.
Additional important records from other parts of New Guinea, and 2 new species, follow. Besides the additional material identified (but not all listed), I have seen 5 specimens which I cannot identify satis- factorily and which I have labeled '^Clivina spp." They are all returned to the Bishop Museum.
Clivina wallacei Putzeys Darlington 1962, Part I, p. 365.
Additional material. Papua: 3, Palmer R. at Black R., June 7-14, Julv 2.3-31, 1936 (Archbold Exp., AMNH).
Notes. Putzeys' original material may have come partly from (western) New Guinea (see discussion in my Part I, cited above), but the present specimens are the first surely recorded from the island. Two of the specimens have the labrum 6-setose, one 5-setose, but I am confident they are all wallacei. Variation in number of labral setae occurs in Philippine individuals too.
Clivina kulfi Darlington Darlington 1962, Part I, p. 366.
Additional material. Papua: 1, Popon- detta, 60 m, Sept. 1-4, 1963 (Sedlacek).
256 BiiUetin Museum of Comparative Zoology, Vol. 142, No. 2
Notes. Previously known onlv from the types from Aitape, N-E. N. G.
Clivina delefa Darlington Darlington 1962, Part I, p. 375.
Additional material. Papua: 1, Mt. Lam- ington, 500 m, June 1966 ( F. Shanahan, C.^E. Lippert, Bishop Mus.). N-E. N. G.: 1, Siaute, Torricelli Mts., sea level, Nov. 9-17, 1958 (W. W. Brandt, Bishop Mus.). West N. G.: 1, Waris S of Hollandia, 450- 500 m, Aug. 1-7, 1959 (C. T. Maa, Bishop Mus.), m. V. (mercury vapor) light trap; 1, Archbold L., Central Mts., 760 m, Nov. 26-Dec. 3, 1961 (S. & L. Quate, Bishop Mus. ) .
Notes. Previously known only from the types from Dobodura, Papua.
Clivina basalis Chaudoir Darlington 1962, Part I, p. 383.
Additional material. Papua: 2, L. Davi- umbu, Fly R., Aug. 19-30, 1936 (Archbold Exp., AMNH). N-E. N. G.: 10 Aiyura, 1600 m, June 6, 1966 (Gressitt), light trap; 1, Kainantu, 1650 m, Oct. 20-26, 1959 ( Maa, Bishop Mus. ), m. v. light trap.
Notes. Although hasalis occurs in Java, Celebes, and Australia as well as New Guinea, on the latter it has been found only in the eastern part of the island. Most New Guinean specimens have the elytra red anteriorly and black posteriorly. However, the 2 from L. Daviumbu have the elytra red with the black area reduced to 2 large post- median maculae as in specimens from tropi- cal Australia, suggesting that the Fly R. population may be independently derived from Australia.
Clivina fessa Darlington
Darlington 1962, Part 1, p. 388.
Additional material. Seventy-four, from all 3 political divisions of New Guinea; a few at low altitudes but chiefly in the moun- tains, including 58 specimens from Wau and
vicinity (most at 1200, 1 at 1500 m), some taken at light.
Notes. I now think fessa is closely re- lated to subfusa Darlington (Part I, p. 384), which is another New Guinean mountain- li\'ing species. These 2 species are some- what similar in form, and both have the middle tibia with spur very near apex. However, fessa differs from subfusa in having eyes larger, elytral margins usually subcrenulate behind humeri, and mid-tibial spurs a little longer.
Clivina australasiae group
Darlington 1962, Part I, pp. 361, 380-397.
Notes. Specific characters in this group are poorly understood and perhaps variable. Additional material ( included in total noted under genus ) has been identified according to key characters, but I have not attempted further study of it except to record the second known specimen assignable to C. csikii Kult and to describe 2 obviously dis- tinct new species.
Clivina csikii Kult
Darlington 1962, Part I, pp. 362, 390.
Additional material. West N. G.: 1, Hollandia area, W Sentani, Gyclops Mts., 150-250 m, June 18, 1959 (Gressitt & Maa), in Malaise trap oxer stream.
Notes. This specimen agrees very well with my redescription of the type, which is from Madang, N-E. N. G. The species may well be distinct. It is distinguished, among other similar ones in New Guinea, by rela- tively slender form and small size (c. 5 mm ) .
C/
ivina
kub
or n.
sp.
Description. Form as figured (Fig. 17), large, with oval-elongate elytra; black, ap- pendages dark reddish; moderately shining, reticulate microsculpture irregular on front' and pronotum and faint on elytra, but sur- face sparsely irregularly punctulate. Head 0.66 width prothorax; eyes rather small, genae c. long as eyes, arcuate; antennae
The carabid beetles of New Guinea • Darlington 257
rather long (in genus), middle segments c. long as wide; mandibles rather short; la- brum 7-setose; clypeus weakly evenly emar- ginate, with wings not separated by notches; front irregularly con\cx, frontal sulci sharply defined, nearly straight, converging anteri- orly, reaching neck posteriorly; neck con- striction scarcely impressed; supraocular con\'e.\ities separated from preocular plates by oblique impressions. ProtJwrox nar- rowed anteriorly; width length 1.03; disc convex, middle line fine, anterior transverse impression deep. Elytra elongate-oval, nar- rowed anteriorly as well as posteriorly; width elytra prothorax 1.25; humeri not dentate; striae deep anteriorly, shallower posteriorly, faintly punctulate; 3 striae free at base; intervals strongly convex anteriorly, 7th briefly finely carinate at base, 3rd with 4 or more punctures on outer edge (4 on left, 6 on right elytron). Inner wings evi- dently atrophied but not examined in the single specimen. Lower surface-, proepis- terna weakly punctate inwardly; last ventral weakly trans\'ersely wrinkled but not punc- tate, with 2 setae on each side widely sepa- rated. Legs: front femur moderate; front tibia 2-dentate above a long curved spur; middle tibia with rather short spur on outer edge near apex. Measurements: length c. 9 mm; width 2.S mm.
Ti/pe. Holotype (sex not determined) (MCZ, Type No. 31806) from "Sarua Kup," Kubor Rge., N-E. N. G., Oct. 31, 1965 (altitude and collector not given, but alti- tude probably high, and specimen received from Dept. Agr. Port Moresby); the type is unique.
Notes. This new species is comparable in form with C. toxopei Darlington ( Part I, p. 363) but is much smaller, with smootli rather than wrinkled head, and different in many other ways. The similarit)' in form is presumably convergent, a result of loss of wings and resulting changes of form. The new species is in fact not related to toxopei but is apparently a deri\'ati\e of the australasiae group (see Darlington, Part I, p. 361, in key). It is perhaps most closely related to and possibly deri\'ed from C.
subfusa Darlington (Part I, p. 384), which it resembles in form of spur of middle tibia, but the new species is larger, with more oval and more deeply striate elytra, and it differs in other details. Also, suhfusa is winged, while kubor obxiously has the wings atro- phied. It is likely that the present new species will prove to live at higher altitudes than subfusa, although the latter is a mountain-living species.
Clivina alternans n. sp.
Description. Form as figured (Fig. 18); slightly depressed; irregularly rufous; rather dull, much of upper surface irregularly punctulate, elytra microreticulate. Head 0.67 and 0.66 width prothorax; eyes rather small, genae arcuate-subtiimcate, meeting neck abrupt!}'; antennae moderate, middle segments almost long as wide; mandibles average; labrum 7-setose; clypeus truncate (\'ery weakly arcuate-emarginate ) with wings not separated by notches; supraocular convexities separated from preocular plates by deep impressions; front flat, ± punctate; frontal sulci shghtly sinuous, reaching pos- teriorly nearly to neck; neck not impressed at middle. Prothorax: width/length 1.08 and 1.05; disc moderately con\'ex with fine middle line and shghtly deeper transverse impression. Elytra: width elytra prothorax 1.21 and 1.19; humeri not much narrowed, not dentate; margin scalloped (much in- terrupted) almost to apex; 3 striae free at base; striae well impressed, faintly punctu- late; intervals alternating in width, odd in- tervals (especially 3 and 5) wider than even intervals; 6th and 7th finely carinate at base, 3rd 4-punctate. Inner wings fully developed. Lower surface, especially proepisterna and abdomen, roughened. Legs: front femora moderate; front tibiae 2-dentate above the apical process (a 3rd tooth indicated by angulation and seta); middle tibiae with moderate spur c. V* from apex. Measure- ments: length 5.7-5.9 mm; width 1.8-1.9 mm.
Types. Holotype (sex not determined) (Bishop Mus.) from Eliptamin \'alley.
258 BiiUetin Museum of Comparative Zoology, Vol. 142, No. 2
N-E. N. G., 1200-1350 m, July 1-15, 1959 (W. W. Brandt); 1 paratype (Bishop Mus.), Mobitei, Torricelli Mts., N-E. N. G., 750 m. Mar. 5-15, 1959 (W. W. Brandt); 1 para- type (MCZ, Type No. 31807), Bokodini, 40 km N of Baliem Vv., West N. G., c. 1300 m, Nov. 16-23, 1961 '(S. & L. Quate), light trap; 1 paratype (Bishop Mus.), Kebar Vy., W of Manokwari, West N. G., Vogelkop, 550 m, Jan. 4-31, 1962 (S. & L. Quate), taken in Malaise trap.
Notes. In my key to Clivina of New Guinea (Part I, pp. 359-362) this runs to the second part of couplet 14 ( the australasiae group), but differs from all species named thereunder in having elytral intervals alter- nating in width. Otherwise possibly most similar to (but not necessarily related to) fessa Darlington (Part 1, p. 388) but differ- ing in alternation of intervals. The flatten- ing of the front, scalloping of the elytral margins (indicated only behind humeri in fessa), and rugose lower surface of alternons (partly rugose also in fessa) are also dis- tinctive characters. It seems likely that this species is widely distributed at moderate altitudes in the lower mountains of New Guinea.
Subfamily HARPAUNAE
Tribe BEMBIDIINI
Genus TACHYS Stephens
Darlington 1962, Part I, p. 400.
Notes. Additional specimens of Tachys received from New Guinea total 456. Most are from light-trap material, and many are therefore damaged or matted with scales of Lepidoptera. These specimens are difficult to study; some I have simply labeled, "Not in condition to determine." Nevertheless this material was well worth collecting and sub- mitting. I have been able to pick out from it important new records of poorly known species as well as several new species.
The following 11 species of Tachys have been found at or near Wau:
T. fasciatus (Motschulsky), at 1200 m ( 1 specimen )
T. ochrioides Darlington, at 1200 m (2) T. reticiiloides Darlington, at up to 1200
m ( only 2 at this altitude ) T. reticidatus Andrewes, at 900 m (6) T. pictus Andrewes, subspecies, altitudes
from 900 to 1800 m (15 in all) T. })emhidiiformis Jordan, at 1200 m (1) T. erotyloides Andrewes, at 1100, 1500 m
(3) T. khigi Nietner, at 1200 m ( 1 ) T. fumicatiis Motschulsky, at 1200 m (1) T. umbrosits Motschulsky, at 1050, 1100,
1150,1200m (6) T. acuticoUis Putzeys, at 1000, 1050, 1100 m(4) Of these 11 species from the vicinity of Wau, retictdoides, reticidatus, pictus and erotyloides usually occur (at other locaUties) by running water; umbrosus and acuticoUis, on or under bark; the others, on the ground in various wet places. The small number of specimens of most species suggests that additional TacJiys are still to be found at Wau.
The following additional records and new species are all that seem to be worth re- cording in detail from the new material.
Tachys apex Darlington
Darlington 1962, Part I, p. 414.
Additioncd material. West N. G.: 1, Waris, S of Hollandia, 450-500 m, Aug 16-
23, 1959 (T. C. Maa, Bishop Mus.); 1, Japen Is., SSE Sumberbaba, Dawar R., Oct.
24, 1962 (H. Holtmann, Bishop Mus.), in jungle.
Notes. Previously known only from the types from Nadzab, N-E. N. G.
Tachys brachys Andrewes Darlington 1962, Part I, p. 433.
Additional muterial. Papua: 1, Bisiatabu, Port Moresby (W. N. Lock, South Austra- lian Mus.).
Notes. This individual, like the 3 previ- ously recorded from New Guinea, is fully winged.
The carabid beetles of New Guinea • Darlington
259
Tachys ceylanicus (Nietner) Darlington 1962, Part I, p. 446.
Additional material. West N. G. : 1, Nabire, S of Geelvink Bay, 10-40 m, Oct. 7, 1962 (H. Hoffmann, Bishop Mus.), jungle.
Notes. Of this species, which is very common from South Asia to the Moluccas, only one specimen has been found hereto- fore in New Guinea (at Hollandia).
Tachys par Darlington
Darlington 1962, Part I, p. 452.
Additional material. West N. G.: 6,
Bodem ( 11 km SE of Oerberfaren, 100 m), July 7-17, 1959 (T. C. Maa, Bishop Mus.); 1, River Tor (mouth), 4 km E of Hoi Maffen, July 2-5, 1959 (T. C. Maa, Bishop Mus. ), m. V. light trap.
Notes. The unique type is from Maffin Bay, West N. G.
Tachys klugi (Nietner)
Klti^ii Nietner 1858, Ann. Mag. Nat. Hist., (3) 2:
423 (Bembidium). klugi Andrewes 1925, Ann. Mus. Civ. Genoa, 51:
401,414, PI. 3, fig. 6.
Description (for recognition only). Very stout and convex member of politus group; black or brassy, elytra usually each with small pale spot near apex; shining, reticulate microsculpture virtually absent; all elytral striae present, deeply impressed, punctulate, 1st entire, others abbreviated anteriorly and posteriorly; length (outside N. G.) c. 2.7- 3.0 mm.
Types. From Ceylon; in Berlin Zool. Mus. ( not seen ) .
Occurrence in New Guinea. N-E. N. G. : 1, Wau, Morobe Dist., 1200 m, Jan. 5-10, 1962 (Sedlacek), in Malaise trap.
Notes. This common Oriental species has been known from India, etc., north to Japan and east and south across the islands to the Philippines and to the Moluccas (specimens from Morotai Is. in MCZ), but the present record is the first from New Guinea.
Tachys convexus (Macleay)
convexum Macleay 1871, Trans. Ent. See. New
South Wales, 2 : 1 15 ( Bembidium ) . convexus Darlington 1963, Psyche, 70: 29.
Description (for recognition only). Mem- ber of politus group; very convex; 4- maculate; prothorax subglobose with pos- terior angles reduced to minute tubercles; basal transverse sulcus of pronotum 3- foveate at middle; elytra 1-striate; length c. 2.4 mm.
Type. From Gayndah, Queensland, Aus- tralia; presumably in Macleay Mus., Syd- ney ( not seen ) .
Occurrence in New Guinea. Papua: 1, Oriomo River, 6 m, Feb. 12, 1964 ("H. C.," Bishop Mus.).
Notes. T. convexus is a common East and North Australian species which ranges north to the tip of Cape York ( Darlington, 1963: 31) and (according to the present record, the first from the island) occurs also on the southern edge of New Guinea.
The form of this species, especially the almost globose prothorax with posterior angles reduced to small tubercles, is very different from that of any other Tachys known from New Guinea.
Tach
ys parasenarius n. sp.
Description. With characters of genus and of politus group ( Darlington 1962, Part I, p. 435); form slightly more slender than average, normally convex; slightly reddish testaceous with suture and elytral margins darker, the dark areas not sharply defined; shining, reticulate microsculpture faint or absent. Head 0.78 and 0.79 width prothorax, without unusual characters. Prothorax: width length 1.40 and 1.45; base apex 1.14 and 1.13; sides broadly arcuate, broadly moderately sinuate to c. right well defined basal angles; disc normal; transverse basal sulcus well defined, subpunctate, inter- rupted at middle and with median fovea. Elytra: width elytra prothorax 1.45 and 1.40; sides subparallel to behind middle, not noticeably fusiform, with humeri nor-
260
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
mally rounded; each elytron with 6 im- pressed striae, 1st entire, others abbreviated at both ends, shorter externally, 7th stria faintly indicated, 8th (submarginal) stria deep, entire, not bowed away from margin. Mea.siirements: length c. 2.3 mm; width 0.9-1.0 mm.
Types. Holotype S ( Bishop Mus. ) and 2 paratypes ( S in MCZ, Type No. 31808, ? in Bishop Mus. ) all from Waris, S of Hollandia, West N. G., 450-500 m, Aug. 16-2.3, 1959 (T. C. Maa), at light.
Measured specimens. The S holotype and 2 paratype.
Notes. In my key to Taclujs of the politus group of New Guinea (1962, Part I, pp. 437-439), this would run to couplet 26 but would fit neither species there named, being much more slender than beinJ)idiiformis and differently colored; and smaller, more shining, and relatively larger-headed than senarius. The new species may be related to senarius and, like the latter, probably lives beside running water, to judge from the long series of other stream-side species collected with it.
Tachys tafei n. sp.
Description. With characters of genus and of politus group ( Darhngton 1962, Part I, p. 435); form c. average except less con- vex than usual; piceous, head and prothorax slightly reddish, elytra more distinctly red at apex, legs pale, antennae dark except pale at base; rather shining, reticulate microsculpture distinct and isodiametric on front (less distinct in 6 ), less distinct and somewhat irregular but apparently c. iso- diameti'ic on pronotum, very fine and strongly transverse on elytra, which are subiridcscent. Head 0.82 and 0.81 width prothorax; without unusual characters. Pro- thorax: width/length 1.51 and 1.52; base/ apex 1.09 and 1.14; sides broadly arcuate, broadly sinuate before c. right well formed basal angles; disc normal with basal trans- verse sulcus moderately impressed, not dis- tinctly punctate, more or less interrupted at middle and with median fovea. Elijtra:
width elytra prothorax 1.47 and 1.50; hu- meri rounded; sides subparallel to behind middle; each elytron with 4 or more well- impressed striae (number that seem well im- pressed depends partly on lighting and angle of view, but is somewhat variable ) , 1st stria entire, others abbre\'iated, progressively shorter and less impressed externally, but trace even of 7th stria present in some in- dividuals, 8th stria entire but not bowed away from margin. Measurements: length 2.8-3.3 mm; width 1.2-1.4 mm.
Types. Holotype c5 (AMNH) and 6 9? paratypes; (2 in MCZ, Type No. 31809) all from Peria Creek, Kwagira River, Papua, 50 m, "[camp] No. 7," Aug 14-Sept. 6, 1953 (Geoffrey M. Tate); all specimens in poor condition, but most good enough to show characters.
Measured specimens. The $ holotype and one 2 paratype.
Notes. How this new species would run in my key to Tachys of the politus group of New Guinea (1962, Part I, pp. 437-439) would depend on how many striae were considered present and impressed. If it were considered 4- or 5-striate, it would probably run to J)orneensis, but the outer elytral striae are much more impressed than in horneensis, and tlie coloration is much darker; nevertheless, the new species mayi be a relative of horneensis. If the elytron were considered 6-striate, the new species would probably run to senarius, but it differs from senarius and also from para- seruirius (above) in darker color and in details of form.
Tachys tor n. sp.
Description. With characters of Tachysi and politus group ( Darlington 1962, Part I, p. 435); form about average in group andl of average convexity; head and prothoraxi reddi.sh testaceous, elytra brownish piceous, each with large macula covering most of apical third of elytral length testaceous, appendages testaceous with antennae browner except at base; rather dull, reticu- late microsculpture distinct, isodiametric on front and pronotum, somewhat transverse
The carabid beetles of New Guinea • Darlington
261
on elytra. Head 0.72 and 0.74 width pro- thorax; without unusual characters. Pro- thorax: width/length 1.55 and 1.41 (width more variable than usual); base/apex 1.27 and 1.24; sides broadly arcuate, moderately sinuate before right or acute well-defined posterior angles; disc normal, basal trans- verse sulcus rather shallow, subpunctate, interrupted and with fovea at middle. Elytra: width elytra/pro thorax 1.43 and 1.48; widest behind humeri, then slightly narrowed but not strongly fusiform; humeri rounded; elytra each 7-striate, 1st stria entire, others abbreviated at both ends and progressively shorter externally and also less impressed externally, 7th being an irregu- lar lightly impressed line, 8th (sub- marginal) stria entire, not much bowed away from margin. Measurements: length 2.3-2.6 mm; width 1.0-1.1 mm.
Types. Holotype S ( Bishop Mus. ) and 3 paratypes ( 6 9 'in MCZ, Type No. 31810, 9 in Bishop Mus. ) all from River Tor (mouth), 4 km E of Hoi Maffen, West N. G., July 2-5, 1959 (T. C. Maa), in light trap (2) and m. v. Hght trap (2).
Measured specimens. The 6 holotype and 1 9 paratype.
Notes. This new species does not fit into my key to species of Tachys of the politus group of New Guinea (1962, Part I, pp. 437-439), because 7-striate elytra are not found in any Tachys previously known from the island (except Mugi, not in the key but recorded in the present paper — the present new species is much more slender aftd not related to kJup^i). Actually, the new species is probably related to senarius and also to parasenarius ( described above ) but differs in color ( the 2 species just named are testaceous with at most rather poorly defined elytral markings) and in ha\'ing the 7th striae better developed, although still weak.
Tachys fumicafus Motschulsky
Darlington 1962, Part I, p. 469.
Additional muterial. New material in- cludes West N. G.: 2, Kebar Vallev, W of
Manokwari, Vogelkop, 550 m, Jan. 4-31, 1962 (S. & L. Quate, Bishop Mus.), light trap; 1, Nabire, S of Geelvink Bay, 20-50 m, July 9, 1962 (H. Holtmann, Bishop Mus.).
Notes. These 3 specimens have elytra entirely unspotted; previously known New Guinean individuals are all 2- or 4-spotted, as described in detail by me in 1962. Super- ficially, the unspotted individuals are re- markably similar to Tachys par Darlington, but of course the frontal sulci are very dif- ferent.
Tachys acuficollis Putzeys Darlington 1962, Part I, p. 480.
Additional nmterial. New material in- cludes N-E. N. G.: 6, Okapa, Aug. 28, 1964 (Hornabrook). WestN. G.: 7, Sibil Valley, Star Mts., 1245 m, Oct. 18-Nov. 8, 1961 ( L. W. Quate, Bishop Mus. ).
Notes. The series listed above consist of unifonnly 2-spotted individuals, with post- humeral elytral spots but without subapical spots. The interrelation of 4-spotted acuti- collis^ the present 2-spotted fonn, and un- spotted coi acinus Putzeys requires further study based on series from additional local- ities. The color differences may prove to be Mendelian.
Tachys wallacei Andrewes Darlington 1962, Part I, p. 479.
Additional material. N-E. N. G.: 1, Adelbert Mts.: Wanuma, 800-1000 m, Oct. 25, 1958 (Gressitt); 1, Finisterre Rge., Saidor: Gabumi Village, July 1-21, 1958 ( W. W. Brandt, Bishop Mus.). West N. G.: 1, Nabire, S of Geelvink Bay, 1-20 m, July 2-9,1962 (Gressitt).
Notes. This rare, arboreal Tachys is evidently widely distributed in New Guinea, from sea level into the lower mountains.
Tachys yunax Darlington
Darlington 1939, Mem. Soc. Cubana Hist. Nat.,
13: 87.
1962, Part 1, p. 482.
1970, Insects of Micronesia. Coleop.:
Carabidae, p. 19.
262 BiiUeiin Museum of Comparative Zoology, Vol. 142, No. 2
Louwerens 1967, Ent. Meddelelser, 35: 197 (re- ferred to under T. singularis Andrewes).
Notes. T. ijunax is now known from New Guinea and the Aru Is. ( Darlington, 1962 ) and Micronesia (Darlington, MS) as well as from the Dominican Republic in the West Indies (Darlington, 1939). Louwerens ( 1967 ) suggests that T. sing^idmis Andrewes (1925, Ann. Mus. Genoa, 51: 388, 393) of Celebes may be the same, in which case yunax will become a junior synonym of singular is. In any case this very widely distributed insect, which may be native in the Malay Archipelago, is probably carried by man, perhaps in wood or wood products of some sort.
Tribe TRECHINI
It is worth re-emphasizing that no Trechus and in fact no flightless mesophile Trechini of any sort have been found in New Guinea at any altitude.
Genus PERILEPTODES Jeannel
Dadington 1962, Part I, p. 489.
Additional material of this genus will be reported on by Dr. Shun-Ichi Ueno. The genus has not been found at Wau.
Tribe PANAGAEINI
Darlington 1962, Part I, p. 492.
Jedlicka 1965, Annotationes Zool. et Bot. (Brati- slava), No. 12: 1-15 (monograph of East Asiatic fnmis ) .
In 1962 I had seen only 7 specimens of this tribe from New Guinea, representing 3 genera and species. Eight additional speci- mens from the island are now before me, including 2 genera and 3 species not re- corded before. Two of the species are striking novelties, based on unique speci- mens. The habitats of the members of this tribe in New Guinea are still wholly un- known.
Genus TRICHISIA Motschulsky Trichisia papuana Csiki
Darlington 1962, Part 1, p. 494.
Additional material. Papua: 1, Popon- detta, 25 m. May 1966 ( Shanahan-Lippert, Bishop Mus.), light trap; 1, Mt. Lamington, 1300-1500 ft. (c. 400-450 m) (C. T. Mc- Namara, South Australian Mus.).
Genus PERONOMERUS Schaum
Schauni 1853, Ann. Soc. Ent. France, (3) 1: 440.
Csiki 1928, Coleop. Cat., Carabidae, Harpalinae 1, p. 364 (see for additional references).
Jedlicka 1965, Annotationes Zool. et Bot. (Brati- slava), No. 12: 2, 10.
De.scription. Small, unicolorous pana- gaeines; fonn characteristic (Fig. 19); sur- face punctate, pubescent; labrum with mid- dle pair of setae not much farther forward than lateral setae; paraglossae not pro- longed beyond apex of ligula; 4th hind- tarsal segments emarginate, with lobes not more than V-; total length of segment.
Type .species. P. fumatus Schaum, of SE Asia, etc.
Generic distribution. SE Asia including Japan to the Philippines and New Guinea ( not Australia ) .
Notes. Among genera of Panagaeini previously recorded from New Guinea (keyed out by me, 1962, Part I, p. 493) this is most like Trichisia, with which it shares short paraglossae (and imiform color- ation), but Peronomerus differs from Trich- isia in having the labrum with middle pair of setae not advanced and the prothorax more wedge-shaped.
Although not recorded before, this genus does occur in New Guinea, where 1 Oriental and 1 endemic species are now known.
Key to Species of Peronomervs of New Guinea
1. Black; elytra coarsely punctate with striae well impressed; length c. 8-9 mm xanthopus
- Bluish green; elytra finely punctate with striae fine, superficial; length 10.8 mm hornahrooki
Peronomerus xanthopus Andrewes
Andrewes 1936, Treubia, 15: 217.
The CARABiD BEETLES OF New Guinea • Darlington 263
23
36
37
38
Figures 17—40 (see text, section [15]): 17, Clivina kubor n. sp., holotype; 18, C. o/fernons n. sp., holo.; 19, Perono- merus /lornobroo/ci n. sp., $ holo.; 20, Cratpedopborus gressiftorum n. sp., c5 holo.; 21, Mecyctothorax sedlaceki n. sp.,
9 holo.; 22, Lesticus medium n. sp., S holo.; 23, Prosopogmus hornobroo/c; n. sp., £ holo.; 24, P. sec//ace(torum n. sp., S parotype; 25, Analoma rosenburgi n. sp., 9 holo.; 26, A. hornabrooki n. sp., 9 holo.; 27, Homo/onesiofo karawari Maindron, 9 holo.; 28, Notagonum curiosum n. sp., S holo.; 29, N. ambulator n. sp., 9 hole.; 30, N. astrum n. sp.,
6 holo.; 31, N. exactum n. sp., <£ holo.; 32, N. quadruum n. sp., S para.; 33, N. sectum n. sp., 9 holo.; €34, Col- podes guego n. sp., 9 holo.; 35, C. helluo Darl., 9, Koibugo; 36, Iridagonum fessum n. sp., 9 para.; 37, /. v/g/7 n. sp., (3 holo.; 38, /. septimum n. sp., c^ holo.; 39, Altagonum cheesmani Dark, S , Wau; 40, A. sororium n. sp., 9 para., Tsenga.
264 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Description. Fomi as usual in genus (cf. Fig. 19); black, appendages brownish tes- taceous, antennae browner except at base; extensively coarsely punctate above and below, moderately shining between punc- tiu-es; elytral striae well impressed; length c. 8-9 mm.
Type. From Java, in British Mus. ( seen ) .
Occurrence in New Guinea. Papua: 1, Mt. Lamington, 1300-1500 ft. (c. 400-450 m) (C. T. McNamara, South Australian Mus.); 1, Popondetta, Northern District, Jan. 29, 1965 (Homabrook).
Notes. Outside New Guinea, xanthopus is known from Java and the Philippines, and closely related fonns occur north to South Asia and Japan.
Peronomerus hornabrooki n. sp.
Description. Form as in Figure 19; dark bluish green, legs red, antennae dark brown with 1st segments red; most of upper sur- face punctate and pubescent, shining be- tween punctures. Head 0.58 width pro- thorax; impressed across base, irregularly channeled each side, punctate only at base especially laterally, smooth at middle; an- tennae with 3rd segments more than 2x long as 2nd; mentmn with broad truncate tooth; ligula blunt, 1-setose each side; para- glossae not clearly visible. Prothorax: width/length 1.34; base/apex c. 2.29; sides abruptly sinuate before minute right pos- terior angles; setae at base and behind mid- dle on each side; lateral marginal lines fine anteriorly, obsolete posteriorly; disc slightly depressed at sides posteriorly, with irregular middle line and deeper basal im- pressions; entire surface of disc closely coarsely punctate. Elytra: width elytra pro- thorax 1.48; surface entirely closely finely punctate witli rows of slightly coarser punc- tures indicating superficial striae. Inner icings fully developed. Lower surface punc- tate. Legs: 4th hind-tarsal segments shal- lowly emarginate. Secondary sexual charac- ters: S front tarsi not obviously modified; 9 unknown. Measurements: length 10.8; width 4.9 mm.
Type. Holotvpe S (MCZ, Tvpe No. 318il) from Main Rixer, Sepik, N-E. N. G., Feb. 24, 1965 (Hornabrook).
Notes. The relatively fine punctation of the elytra ( contrasting with the coarse punctation of the pronotum) distinguishes this from all previously known species of the genus.
Genus MICROCOSMODES Strand
Strand 1936, Folia Zool. et Hydrobiologica (Riga), 1936: 169.
Tedlicka 1965, Annotationes Zool. et Bot. ( Brati- slava), No. 12: 2,6.
Microcosmus Chaudoir 1878, Ann. Soc. Ent. Bel- gique, 21: 85, 139 (not Microcosmus Fee 1830 nor Heller 1877).
Microschemiis Andrewes 1940, Ann. Mag. Nat. Hist. (11), 5: 536.
Darlington 1962, Part I, p. 495.
Notes. I am indebted to Dr. Shun-Ichi Ueno for calling my attention to the preced- ing generic synonymy.
Microcosmodes quadrimaculafus (Csiki) Darlington 1962, Part I, p. 496 (Microschemiis).
Additional material. Papua: 2, Mt. Lamington, 1300-1500 ft. (c. 400-450 m) (C. T. McNamara, South Australian Mus.).
Notes. One of these specimens is not spotted (except that the prothoracic mar- gins are pale posteriorly, as usual); the other, 4-spotted, each elytron with a very small anterior-lateral (chiefly on interval 8) and a subapical testaceous mark.
Genus CRASPEDOPHORUS auct.
Darlington 1962, Part I, pp. 492-493 (in text and key).
Notes. This genus, previously known from two separate areas ( SE Asia, etc., and Australia) has not previously been recorded from New Guinea, but the following new species establishes its presence there.
Craspedophorus gressifforum n. sp.
Description. Form as figured (Fig. 20), moderately convex; black, appendages pic-
The carabid beetles of New Guinea • Darlington
265
eous in part slightly reddish; elytra each with 2 yellow spots as figured; moderately shining, disc of pronotum and elytra finely lightly c. isodiametrically microreticulate, head and marginal and lateral channels and base of prothorax irregularly punctate. Head 0.60 width prothorax; antennae long ( but outer 5 or 6 segments missing ) ; mandi- bles short, with points overlapping; labrum emarginate, 4-setose, with 2 inner setae behind level of outer setae; clypeus truncate, 1-setose each side well behind anterior mar- gin; front irregularly impressed each side; neck transversely impressed; mentum \\'ith truncate tooth; ligula and paraglossae fused, fonning a single structiue truncate ( slightly ircuate) anteriorly with 1 seta each side. PwfJiomx: width length 1..34; base apex 1.15; margins wide; posterior angles obtuse but well-defined, each with seta at angle (no setae anteriorly); disc convex, middle line deep, transverse impressions obsolete. Elytra: width elytra/prothorax 1.32; humeri roundly prominent with margins elevated; striae moderately impressed on disc but very light ( indicated by rows of punctures ) externally; intervals c. equal. Inner it/ni^s presumably atrophied (not examined). Lower surface: metepisterna short, scarcely longer than wide; much of lower surface irregularly rather sparsely punctate, and abdomen also sparsely pubescent at mid- dle (sterna also sparsely pubescent at mid- dle); apical ventral segment with 2 seta- bearing punctures each side before apex. Legs slender; 4th hind-tarsal segments emarginate but not long-lobed. Secondary sexual characters: S front tarsi slightly dilated, densely clothed below with slender squamules; 6 apical ventral segment with 2 punctures close together on each side before apex; 5 unknown. Measurements: length c. 12 mm; width 4.9 mm.
Type. Holotype c^ ( Bishop Mus. ) Tapini (Goilala), Papua, 1100 m, May 18, 1961 (J. L. &M. Gressitt).
Notes. This new species has the elytra much more lightly striate than C. australis Dejean or any other member of the genus previously known to me.
Tribe PTEROSTICHINI Genus MORION Latreille Morion longipenne Putzeys
Darlington 1962, Part I, p. 503.
Additional material. Ninety-four, from many widely scattered localities; included are many from W'au, etc., altitudes up to 1700-1800 m, dates in every month except April ( Sedlaceks and others ) .
Notes. The numerous specimens assigned to this species vary considerably, as noted by me in 1962. This and related species in the Malay Archipelago require third-stage taxonomic study.
Genus MECYCLOTHORAX Sharp
Darlington 1962, Part I, p. 505.
Mecyclofhorax sedlaceki n. sp.
Description. Fonn as figured (Fig. 21); rather strongly convex; black, appendages slightly rufescent; reticulate microsculpture light or indistinct on front and pronotum, more distinct and irregular but c. iso- diametric on elytra. Head 0.77 wddth pro- thorax; eyes moderate, genae shorter than eyes, oblique, very slightly arcuate; anten- nae rather short, middle segments not more than 1^/2 X as long as wide; front convex, clypeal sutiue well impressed, frontal im- pressions slightly irregular but linear and subparallel; 2 setae over each eye; mentum wdth strong rounded tooth. Prothorax slightly transverse-rounded; width/length 1.36; base apex 0.98; base /head 0.94; sides c. rounded to base except very briefly strongly sinuate just before right subdenti- fonn posterior angles; margins narrow, each with seta at basal angle and Vs from apex; disc with fine middle line and trans\'erse impression and with basal area slightly de- pressed and conspicuously punctate. Eh/tra 1.46 width prothorax, quadrate-oval, slightly narrowed anteriorly to broadl\- rounded humeri; subapical sinuations slight; anterior margin entire, connected to scutellar striae; striae light, slightly impressed near suture, reduced externally to rows of punctures; 7th
266 Bulletin Museum of Comparative Zoologij, Vol. 142, No. 2
striae scarcely indicated; 3rd intervals each with 2 or 3 seta-bearing pnnctures, 5th in- tervals apparently without such punctures. Inner wing.s apparently atrophied. Lower sm-face virtually impunctate except a few punctures at sides of mesosterna. Lefis without noteworthy special characters. Sec- ondary sexual characters: 9 last ventral segment with 2 pairs of setae each side near apex and an additional pair of smaller setae close together at middle farther from apex; S unknown. Measurements: length 4.3 mm; width 1.8 mm.
Type. Holotype 9 (Bishop Mus. ) from Mt. Wilhelm, Bismarck Rge., N-E. N. G., 4250 m, June 3, 1963 (J. Sedlacek); the type is unique.
Notes. The only species of the genus previously known from New Guinea is M. toxopei Darlington ( 1962, Part I, p. 506), known from a single specimen from 4200 m on Wilhelminatop, Snow Mts., West N. G. The present new species evidently repre- sents the same stock but differs from toxo- pei slightly in details of form ( more distinct posterior prothoracic angles, etc.) and es- pecially in having the base of the pronotum extensively and conspicuously punctate and the elytral striae punctate, and in lacking seta-bearing punctures on the 5th elytral intervals. See section [SO] of the present paper for further discussion of the distri- bution and probable history of this genus.
Genus BRACHIDIUS Chaudoir Brachidius cross/corn/s Chaudoir
Darlington 1962, Part I, p. 508.
Additional material. Twenty-one, from various localities, including 1, Wan, 1200 m, Jan. 4-15, 1965 (M. Sedlacek).
Notes. Individuals of this species vary considerably in size. Those from montane localities tend to be larger.
Genus CAELOSTOMUS Macleay Darlington 1962, Part I, p. 508.
Notes. Additional material (65 speci-
mens) includes the following from vicinit) of Wan.
C. novae-guineae Straneo, 1000-1300 rr
( 1 individual ) C. albertisi Straneo, 1200 m ( 1 ) C. subsinuatus (Chaudoir), altitude;
from 1000 to 1700 m (7) C. picipes Macleay, 1200, 1300 m (9)
Genus COSMODISCUS Sloane Cosmodiscus rubripictus Sloane
Darlington 1962, Part I, p. 514.
Additional material. Papua: 1, Tapin (Goilala, Owen Stanley Rge.), May 17-19 1961 (Gressitt), in light trap; 1, same lo cality, 975 m, Nov. 16-25, 1957 (W. W Brandt, Bishop Mus. ) . N-E. N. G. : 8, Wau 1200 m, dates in Jan., Feb., Apr., July, Sept. 1963, 1964 (Sedlaceks), some in m. v. lighi trap; 1, Okapa, June 13, 1965 (Hornabrook)' 1, Torricelli Mts., Mobitei, 750 m, Apr. 16- 22, 1959 (W. W. Brandt, Bishop Mus.); 1 Wapenamanda, West Highlands, Mar. 21 1960 (J. Barrett, Dept. Agr. Port Moresby)] at m. V. light.
Notes. All the specimens listed above are rather small (like the pair previously re-i corded from Dobodura) and all have ai least vestiges of a red elytral pattern. Be- sides these specimens I have seen 2 othen as follows. N-E. N. G.: 1, Damanti Finisterre Rge., 3550 ft. (c. 1080 m), Oct. 2-11, 1964 (Bacchus, British Mus.), "Station No. 46"; 1, Budcmu, Finisterre Rge., c. 4000 ft. (1220 m), Oct. 15-24, 1964 (Bacchus. British Mus.), "Station No. 51." Tliese 2 specimens are identified doubtfully. The one from Damanti is more parallel-sided than most rubripictus and lacks red marks, and the surface seems to be modified per- haps by a chemical preservative. The one from Budemu is much larger (c. 8.3 mm), unmarked except for slightly reddish suture and slight translucent lateral prothoracic margins, and very deeply striate. However, these specimens show no decisive differ- ences from rubripictus, and they may repre- sent extremes of a population that varies in size, markings, and some other details.
The carabid beetles of New Guinea • Darlington 267
Cosmodiscus brunneus Darlington Darlington 1962, Part I, p. 515.
Additional material. N-E. N. G. : 4, Wau, 1200 m, Sept. 17, Nov. 21, 1961 ( Sedlaceks ) , some at m. v. light trap; 1, Lae, Singiiawa R., 30 m, Apr. 15, 1966 (Lippert, Bishop Mus.); 1, TorricelH Mts., Mobitei, 750 m, Feb. 28-Mar. 4, 1959 (W. W. Brandt, Bishop Mus. ) .
Genus LESTICUS Dejean
Darlington 1962, Part I, p. 521.
Notes. Thirty additional specimens from New Guinea have been seen. They are referred (in some cases doubtfully) to chloronotus Chaudoir (subspecies?), gracilis Darlington, poUtus Chaudoir, and a new species from the W'issel Lakes described below. Third-stage taxonomic study, based on much more material, is needed to deter- mine specific variation and specific limits in this genus. No species of the genus has yet been found at or near Wau.
Lesficus medius n. sp.
Description. With characters of genus; form as figured ( Fig. 22 ) , with elytra more narrowed in front than behind; moderately depressed; black, elytra slightly aeneous, appendages dark; moderately shining, retic- ulate microsculpture indistinct on head and pronotum, light and very fine on elytra. Head 0.77 and 0.79 width prothorax; eyes moderate; antennae with middle segments 3x long as wide; mandibles strongly curved and acutely produced as usual; front weakly irregularly convex, deeply impressed each side anteriorly; neck constriction slight. Prothorax cordate; width length 1.37 and 1.45; base apex 0.95 and 0.87; lateral mar- gins rather narrow, each with usual 2 setae; disc weakly convex, with fine middle line and weak ( subobsolete ) transverse im- pressions; basal foveae shallow and poorly defined, impunctate. Elytra: width elytra/ prothorax 1.30 and 1.31; humeri narrowed, obtuse; subapical marginal interruption
present; striae virtually absent or indicated as faint fine lines; each elytron with 2 mi- nute punctures on (position of) 3rd interval, on outer edge less than Vi from base and inner edge near middle (posterior dorsal punctures absent). Inner icings vestigial, reduced to thin strips c. % length elytra. Lower sut^ace impunctate except for group of punctures each side mesosternum; met- episterna shortened, slightly longer than wide. Secondary sexual characters: 6 front tarsi slightly dilated, 3 segments squamulose below; c5 with 1, 5 with 2 setae each side last ventral segment. Measurements: length 19.0-20.5 mm; width 6.7-7.5 mm.
Types. Holotype S (MCZ, Type No. 31813) from Kamu Valley, Wissel Lakes, West N. G., June-Aug. 1959 (L. J. Pospisil); I am indebted to Professor C. T. Remington for this specimen. Also 10 paratypes (some in MCZ) all from Lake Paniai, Wissel Lakes, 1750 m, Sept. 23, 28, 30, Oct. 1, 1939 ( H. Boschma, Leiden Mus. ) .
Measured specimens. The i holotype and one 2 paratype.
Notes. In my key to New Guinean Lesti- cus ( 1962, Part I, pp. 523-524 ) , this species would run to couplet 5, but would fit neither half of the couplet, being less broad and less depressed than depressus but broader and less convex than toxopei and ambulator. Actually, the present new species is inter- mediate between the nomially fomied, fully winged members of the genus and the more strongly modified flightless forms just named. In other words, the present new species represents an intermediate stage in wing atrophy leading toward the more highly divergent montane species.
Genus PROSOPOGtMJS Chaudoir
Darlington 1962, Part I, p. 536.
Key to Species of Prosopogmus of New Guinea
1. Fully winged; pronotum punctate baso- laterally garivagliae
- Wings atrophied; pronotum not or not much punctate baso-laterally 2
2. Eyes larger (cf. Figs. 23 and 24); prothorax relatively wider and with wider base ( cf . pro- portions in Descriptions) honiabrooki
268
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
- Eyes smaller; prothorax relatively narrower and with narrower base sedlacekoiuiu
Prosopogmus garivaglioe Straneo Darlington 1962, Part I, p. 537.
Additional material. Thirty-one, from widely scattered localities; altitudes given from 50 to 1950 and 1665-2530 m; included are 6 from Wau, 1050, 1200 m, dates in Jan., Sept., Oct., Nov., Dec, 1961-1963 (Sedlaceks and others), some in light traps.
Notes. This species varies in depth of striae and in other ways and requires third- stage study.
Prosopogmus hornabrooki n, sp.
Description. With characters of genus; form as figured (Fig. 23); black, append- ages reddish black; shining, reticulate micro- sculpture indistinct on head, very light and slightly more distinct and transverse on elytra. Head 0.68 width prothorax; eyes slightly larger and meeting neck more abruptly than in other New Guinean mem- bers of genus; front convex except for transverse clypeal suture and curved an- terior frontal impressions. Prothorax moder- ately transverse; width/length 1.37; base/ apex 1.33; lateral margins rather narrow; disc weakly convex, distinctly flattened each side near basal angle, with middle line fine and abbreviated both ends, transverse im- pressions obsolete, and basal impressions sublinear; surface of disc impunctate except for a few punctures basally especially near lineal basal impressions (not near angles). Elytra subquadrate, not noticeably nar- rowed basally; width elytra/ prothorax 1.18; striae entire, deeply impressed, not punc- tate; intervals convex, 3rd 3-punctate with anterior puncture on outer and others on inner edge of interval. Inner win<!_s reduced to thin strips c. % length elytra. Lower siniace: last 3 ventral segments transversely impressed near base. Secondary sexual characters: i front tarsi moderately dilated, 3 segments squamulose; 6 with 1 seta each side last ventral segment; 9 unknown. Mea- surements: length 9.5; width 3.5 mm.
Type. Holotype S (MCZ, Type No. 31814), from Okapa (SW of Kinantu, East Highlands), N-E. N. G., (altitude probably between 1650 and 1800 m), June 12, 1964 ( Horaabrook ) ; the type is unique.
Notes. This new species is close toj fi,arivagliae Straneo (preceding) but has' a more transverse prothorax with pronotum less punctate basally, relatively larger eyes, and reduced inner wings. P. garivagliae occurred at Okapa too, which suggests that hornabrooki is a real species, not a geo- graphic form. The fact that the eyes are relatively larger in hornabrooki in spite of the fact that the wings are reduced also suggests specific distinctness.
Prosopogmus sedlacekorum n. sp.
Description. With characters of genus; form as in Figure 24, rather slender; black, appendages reddish black; shining, reticu- late microsculpture faint or light and iso- diametric on head, more distinct and more transverse on pronotum and elytra. Head 0.71 and 0.73 width prothorax; front evenly convex except for deep clypeal suture and curved anterior impressions each side, with surface finely rather sparsely punctulate. Prothorax subquadrate; width/length 1.29 and 1.33; base/apex 1.23 and 1.18; lateral margins relatively narrow; disc normally convex, less impressed than usual baso- laterally, with fine middle line, transverse impressions obsolete; surface impunctate, except vaguely punctate baso-laterally. Ely- tra slightlv narrowed anteriorlv; width elytra prothorax 1.24 and 1.24; striae entire, moderately convex, 3rd 3-punctate with 1st puncture on outer and others on inner edge of interval. Inner wings vestigial, reduced to narrow strips c. % length of elytra. Loicer surface: last 3 ventral segments conspicu- ously transversely impressed near base. Sec- ondary sexual characters: $ front tarsi moderately dilated, 3 segments squamulose ' below; $ with 1, 9 2 setae each side last ventral segment. Measurements: length 9.3-10.4 mm; width 3.5-3.8 mm.
Types: Holotype 6 (Bishop Mus.) and
The carabid beetles of New Guinea • Darlington
269
4 paratypes (2 in MCZ, Type No. 31815) all from Mt. Giluwe, Papua, with additional details as follows: holotype, 2500-2750 m, May 30, 1963; paratypes, 2500, 2550, 2800- 3280 m, dates in May 1963 (all specimens collected by J. and/or M. Sedlacek).
Measured specimens: the $ holotype and 9 paratype from 2550 m.
Notes. This distinct, localized, flightless species is distinguished from other New Guinean members of the genus in the pre- ceding key.
Genus ANALOMA new name
Paraloma Darlington 1962, Part I, pp. 500, 538 (not Paraloma Cope 1863, Proc. Acad. Nat. Sci. Philadelphia for 1962: 181 ).
Diag,nosis and description. As for Para- loma Darlington 1962.
Ty))e species. Paraloma fortis Darlington 1962, p. 539.
Generic distribution (revised). High mountains of New Guinea.
Notes. The name Analoma, now pro- posed, is a new name for Paraloma Darling- ton, which is preoccupied. I am indebted to Mr. W. J. Brown for calling my attention to Cope's earlier use of Paraloma.
The 2 new species described below ex- tend the range of the genus, previously known only from the Snow Mts., West N. G., to mountains in the eastern half of the island.
Key (Revised) to the Species of Analoma
1. Elytra with basal margin 2
- Elytra without basal margin gracilis
2. Sides of prothorax not sinuate hornabrooki
- Sides of prothorax sinuate posteriorly 3
3. Sides of prothorax sinuate farther before base (Darlington 1962, fig. 56) fortis
- Sides of prothorax sinuate very near base (present Fig. 25) rosenburgi
Analoma rosenburgi n. sp.
Description. With characters of genus; form as in Figure 25 (but possibly di- morphic); black (slightly reddish), append- ages reddish brown; shining, reticulate microsculpture of front and pronotum in-
distinct, of elytra very fine, lightly im- pressed, slightly transverse. Head 0.79 width prothorax; eyes small (but longer than genae), scarcely prominent; mentum tooth broad, short, emarginate; other mouth- parts as described for genus ( 1962 ) . Pro- thorax very broad anteriorly (but perhaps dimorphic); width/length 1.28; base/apex 0.84; base/head 0.96; lateral margins nar- row, slightly broader at anterior angles, still broader (but not wide) at base, each with seta-bearing puncture c. % from apex just inside margin and on face (not edge) of margin at base; basal sinuations of mar- gin strong, very near c. right (slightly acute) basal angles; disc with usual impressions weak, baso-lateral impressions slight and poorly defined; base and apex not mar- gined; apical angles prominent, bluntly acute. Elytra narrowed toward base; width elytra/prothorax 1.36; base margined, mar- gin obtusely but distinctly angulate at humeri; humeri not dentate; subapieal mar- ginal interruptions present; subapieal sinu- ations slight; apices conjointly rounded; striae slightly impressed, virtually entire, irregularly punctulate; no ocellate puncture at base; 3rd intervals impunctate (as usual in genus). Lower surface and legs as de- scribed for genus ( 1962). Secondary sexual characters of 9 normal, of c^ unknown. Measurements: length 12.6; width 4.8 mm.
Type. Holotype ? (MCZ, Type No. 31817) from Mt. Albert Edward, Papua, 13,200 ft. (c. 4025 m), July 1, 1963 (Guy Rosenburg), in alpine grassland; the type is unique.
Notes. This species probably represents the same stock as Analoma fortis ( Darling- ton ) ( 1962, Part I, p. 539) of the Snow Mts., West N. G., but differs as indicated in the preceding Key. A. rosenburgi may prove to be dimorphic in form, as fortis is.
I take pleasure in naming this interesting new, high-altitude species for the collector.
Analoma hornabrooki n. sp.
Description. With characters of genus; form as in Figure 26 (but perhaps di-
270 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
morphic); black, appendages dark brown- ish; shining (althongh surface partly ob- scured), reticulate microsculpture light, fine and irregular on head, indistinct on prono- tum, rather strongly transverse on elytra. Head 0.76 width prothorax; eyes rather large (in genus), slightly prominent and longer than genae; mentum tooth very wide, short, emarginate; other mouthparts as de- scribed for genus ( 1962 ) . Frothorax sub- cordate; width length 1.30; base ape.x 1.00; base /head 1.00; lateral margins very narrow anteriorly, wider posteriorly, running into shallow poorly defined baso-lateral impres- sions, with setae as described for preceding species {rosenhur^); base and apex not margined; disc with fine middle line, very weak transverse impressions, scarcely sculp- tured at base. Elytra narrowed toward base; width elytra/prothorax 1.34; base mar- gined, margin c. right (slightly obtuse) at humeri; subapical marginal interruptions present; subapical sinuations moderate; apices conjointly rounded; striae slightly impressed, entire, irregularly punctulate at least in part; intervals scarcely convex, ir- regularly sparsely punctulate; ocellate punc- ture at base 1st stria, but 3rd interval impunctate. Lower surface and legs as de- scribed for genus (1962) except proepi- stema in part finely punctate. Secondary sexual characters of 9 normal; of c^ un- known. Measurements: length 12.0; width 4.7 mm.
Type. Holotype 9 (sex determined by dissection) (MCZ, Type No. 31816) from Okapa (Purosa), N-E. N. G., (altitude prob- ably between 1650 and 1800 m), Nov. 30, 1965 ( Hornabrook ) ; the type is unique.
Notes. Although this species differs con- siderably in fonn from A. rosenhurgi, both species may prove to be dimorphic, as fortis is known to be on the Snow Mts., so that form may prove not to be specifically diag- nostic. However, the present new species differs from the preceding one (rosenhurgi) in other ways, including presence of an ocellate puncture at the base of the first striae (absent in rosenhurgi) and much more transverse elytral microsculpture. I have
no doubt that the single individuals on which these species are based represent distinct, local, high-altitude forms of a group of which other representatives will probably be found on other mountains widely scattered over New Guinea.
This species, like several others in the present paper, is dedicated to Dr. R. Horna- brook, in recognition of his discoveries ot^ exciting Carabidae at high altitudes in New Guinea.
Genus HOMALONESiOTA Maindron '
Darlington 1962, Part I, p. 516.
Description, Notes, etc. See under fol- lowing species.
Homalonesiofa karawari Maindron \
Maindron 1908, Nova Guinea, 5: 296. i
Andrewes 1946, Proc. R. Ent. Soc. London, ser. B,
p. 85. Darlington 1962, Part 1, p. 517.
Description (revised). Form and selected details as in Figure 27; head and pronotum without distinct microsculpture, elytra with meshes distinct, irregular, transverse but not close-packed (not iridescent). Head ( as measured in type ) 0.79 width prothorax; clypeus 1-setose each side (and apparent!) impressed or punctate at outer corners of base); labrum 6-setose; mentum with sim- ple triangular tooth; ligula not clearly visi- ble; paraglossae longer, rather narrow, membranous; mandibles without seta in scrobe. Prothorax subquadrate; width length 1.26; base/apex 0.98; lateral margins narrow^ each with seta just inside margin c. ^-i from apex and seta on raised margin just before base; base and apex not mar- gined at middle; disc with median im- pressed line fine, abbreviated both ends, transverse impressions weak, baso-lateral impressions irregularly linear, they and sur- face near them punctate. Elytra parallel; width elytra/prothorax 1.27; base margined, margin obtusely angulate at humeri; basal margin running into 1st striae; no sciitellar striae; ocellate puncture at base 2nd stria; suhapical marginal interruptions ahsent;
The carabid beetles of New Guinea • Darlington 271
striae impressed, impunctate; intervals slightly convex, 7th and 8th slightly nar- rowed and more convex toward apex but not otherwise modified, punctures (ocel- late) of 9th intervals as indicated (Fig. 27); 10th intervals absent; subapical punc- ture between striae 7 and 8; single dorsal puncture on 3rd interval of each elytron as indicated (Fig. 27). Inner tiings full (visible). Lower surface: sterna and 1st ventral segment partly, irregularly punc- tate; last ventral entire, 2-setose each side ( 9 ). Legs moderately slender; hind tro- chanters nearly Mj long as femora, rather narrowly oval, narrowly rounded at apex; bind tibiae grooved each side externally but not grooved at middle; hind tarsi with basal segments subcarinate at middle and 3ach side above, rather broadly slightly depressed (rather than sulcate) between ntiedian and lateral carinae; 4th hind-tarsal segments weakly emarginate; 5th segments with 1 or 2 weak accessory setae each side (additional setae possibly broken off). Length c. 10.2 mm.
Type. Holotype 9 ( Paris Mus. ) labeled 'Tawarw./Juni 1903," "nieuw. Guinea./ Expeditie 1903," "Homalonesiota karawari m. [on small red square]" "Collection Ba- bault." An additional label repeats the name, gives the original reference, and notes that the specimen was submitted to An- drewes 27/7/1922. This specimen (the type) is now in the Paris "Generale" collection in a box with New Zealand Holcaspis, etc., the box being apparently numbered 19.
Occurrence in New Guinea. Known only from the type.
Notes. Andrewes was seriously in error in comparing this insect to Morion. The resemblance is superficial. In fact, many characters indicate an actual relationship with Loxandrus. The general form is rather like that of a parallel-sided Loxandrus, and the linear baso-lateral prothoracic impres- sions and single dorsal elytral punctures are consistent too. Also consistent is the ab- sence of an interruption of the elytral mar- gin toward the apex. This interruption is rarely absent in Pterostichini but does vary
and is sometimes obliterated in Nehriofe- ronia, which is (I think) a Loxandrus- derivative. The form of the $ front tarsi of Homalonesiota karawari is of course un- known, since the type is a 2 . However, this beetle is so similar in other ways to the species that I have described as Nebriofe- ronia straneoi (below) that I feel sure the $ tarsi will be the same. I am therefore transferring straneoi to Homalonesiota. Tlie 2 species, karawari and straneoi, differ mainly in form, the prothorax of straneoi being much more narrowed posteriorly and the elytra coaptively narrowed toward base. The two species agree in lacking or virtually lacking the marginal interruptions of the elytra as well as in many other characters. The type species of Nebrioferonia, strigitar- sis Straneo, differs in some details that may permit retention of Straneo's genus. I shall let future revisers of the Loxandrus complex decide about this.
It seems likely that H. karawari and H. straneoi, like Nebrioferonia strigitarsis, live among cobblestones and under other cover by running water, although there is no actual record of this. It is surprising that no additional specimens of karawari have been found, since it is a winged species occurring at low altitudes. It may prove to be a primarily lowland fonn which is re- placed at higher altitudes by straneoi.
Homalonesiota straneoi (Darlington)
Darlington 1962, Part I, p. 559 (Nebrioferonia).
Additional material. N-E. N. G.: 15, Wau, Morobe Dist., 1200 (most), 1250 (1) m, dates in Jan., Mar., Apr., May, Sept., Oct., Nov., Dec, 1961-1966 (Sedlaceks, and 1 Gressitt & Wilkes), some labeled as taken in light trap or m. v. light trap; 1, 6 km W of Wau, Nami Ck., 1700 m, June 10, 1962 (Sedlacek), in Malaise trap; 1, Mt. Kaindi, 2350 m, Dec. 12, 1964 (Sedlaceks); 1, W Highlands, nr. Minj, 1340 m. May 2, 1966 (Gressitt), light trap; 1, Eliptamin Vy., 1200-1350 m, July 16-31, 1959 (W. W. Brandt, Bishop Mus.); 1, Finisterre Rge., Moro, c. 5550 ft. (c. 1700 m), Oct. 30-Nov.
272 Bulletin Miiscinn of Cuinparative Zoology, Vol. 142, No. 2
15, 1964 (Bacchus, British Mus.). West N. (;.: 1, Japen Is., Camp 2, Mt. Eioii, 2000 ft. (610 m), Sept. 1938 (Cheesman).
Notes. The N-E. N. G. records fill the previous geographic gap in the species' distribution. It was known formerly only From the Snow Mts., West N. G., and Kokoda, Papua.
Genus NEBRIOFERONIA Straneo
Darlington 1962, Part I, p. 557.
Notes. See Notes under Homalonesiota karawari (above).
Nebrioferonia sfrigif arsis Straneo
Additional 7nate rial. N-E. N. G.: 1, Wau, Mt. Missim, 1250 m, Oct. 5, 1961 (Sedlaceks); 2, Main R., Sepik, Feb. 1965 (Hornabrook). West N. G.: 2, Star Rge., Sibil, 1260 m, June 27, 29, 1959 ( Neth. N. G. Exp., Leiden Mus.); 1, Sentani, 90+ m, June 22, 1959 (Gressitt & Maa), m. v. light trap.
Notes. This species may vary geograph- ically, but more material is needed to treat it properly. It lives among cobblestones, etc., beside fairly large streams.
Tribe AGONINI
Darlington 1952, Part II, p. 113.
1956, Australian carabid beetles III. Notes
on the Agonini. Psyche, 63: 1-10.
1963, Australian carabid beetles XIII. Fur- ther notes on Agonini, . . . Breviora ( Museum of Comparative Zoology), No. 183: 1-10.
In 1952, in Part II of my work on carabid beetles of New Guinea, 18 genera, 107 spe- cies, and 14 additional subspecies of Ago- nini were recognized from the island. Since then, "Colpodes" violaceus Chaudoir has been made the type of a new genus (Violagonum) by me in 1956, and 2 addi- tional genera are recorded now. One, Di- cranoncus, is expected. The other, Agonum, is unexpected. This genus is dominant in the northern hemisphere; the single stock (subgenus Sericoda) that has reached New Guinea has evidently done so by "mountain- hopping." One common south-temperate
Australian species, Notagoniim siibmetal- licum (White), has been found to occur in the Wissel Lakes region in the mountains of central New Guinea. And 50 new, chiefly mountain-living species, some of them strikingly characterized, have been dis- covered and are described below; their dis- covery has necessitated complete revisions of several of the mountain-living genera. These additions bring the number of Ago- nini known from New Guinea to 21 genera, 160 species, and 14 additional subspecies. This total will surely be very much in- creased by further collecting on the higher mountains of the island.
The origins and radiation of Agonini on New Guinea are discussed in Part II and summarized in the present part [92].
Agonini are dominant throughout the tropics of the world, but are relatively less, numerous in most temperate areas and notably few in Australia. The change from an overwhelmingly agonine fauna in New Guinea to an overwhelmingly pterostichine fauna in Australia has been discussed by me in 1956 and is rediscussed in the present paper [64].
The following characters are considered normal for Agonini and will usually not be repeated in the individual descriptions, un- less the normal characters are modified. (A more detailed discussion of structures and characters of New Guinean Agonini will be found in Part 11, pages 94ff. ) On the upper surface, the reticulate micro- sculpture is considered normal when the meshes are visible in good light at a magni- fication of 54 X, and when they are iso- diametric on head, moderately transverse on pronotum, and equally or more trans- verse on elytra. On the head, the eyes are normally prominent, with 2 setae over eachi eye; mandibles, of moderate length andl moderately arcuate; the palpi, not or noti much dilated; and the mentum, with a tri- angular tooth. The prothorax normally has' lateral margins moderate, wider posteriorly, each with 2 setae, at or near basal angle and slightly before middle; disc with rather fine impressed middle line usually abbreviated
/
The carabid beetles of New Guinea • Darlington 273
at both ends, less well-defined anterior and posterior transverse impressions, and mod- erate, usually impunctate baso-lateral im- pressions. The elytra normally have mar- gins with moderate subapical sinuations; 8 entire, moderately impressed striae (in addition to the scutellar striae); and inter- vals usually not much modified toward apex, the third with 3 seta-bearing punc- tures more or less evenly spaced along the length of the interval, the anterior puncture being on the outer and the other punctures on the inner edge of the interval. The lower surface is normally impunctate or nearly so, and not pubescent except for fixed setae. And normal secondary sexual characters are i with anterior tarsi slightly dilated, the first 3 segments 2-seriately squamulose; and $ with 1, 9 2 setae each side last ventral segment.
The Key to Genera of Agonini of New Guinea (Part II, pp. 114-116) may be brought up to date by insertion of the following couplets:
9a. Fifth hind-tarsal segments with accessory setae; i parameres tipped with setae Agoimm (Sericoda)
- Fifth hind-tarsal segments without acces- sory setae; S parameres without setae Notagonum
10a. Head relatively short, with short mandibles; ( broad, purple, each elytron with an apical spine c. opposite 2nd interval, length c. 10 mm) _.__ Yiolagonum
- Head and mandibles relatively longer; (form and color various, elytra differently spined or not spined, size usually larger X Colpodes
Note also the following exceptions to char- acters used in the original Key. ISlotagonum now includes 1 species with reduced wings {ambulator, described below), and Altag,o- num now includes 2 species which have lost the anterior supraocular setae {higenum and conicoUe, described below ) .
Genus ARHYTINUS Bates
Arhytinus medius Darlington Darlington 1952, Part II, p. 117.
Additional material. Six, from localities scattered in all three political divisions of New Guinea; most at low altitudes, but 1, Tapini (Papua), at probably c. 1200 m.
Arhytinus granum Darlington Darlington 1952, Part II, p. 119.
Additional material. N-E. N. G.: 1, Okapa, probably between 1650 and 1800 m, June 13, 1965 ( Hornabrook ) .
Notes. This very small agonine has been known before only from the 2 types from Dobodura.
Genus TARSAGONUM Darlington
Darlington 1952, Part II, p. 120. Louwerens 1966, Ent. Tidskrift, 87: 36.
Generic distribution (revised). Known only from New Guinea and Borneo.
Notes. This very distinct genus was orig- inally based on the single New Guinean species; the Bornean species has been de- scribed by Louwerens ( 1966 ) .
Tarsagonum lafipes Darlington Darlington 1952, Part II, p. 120.
Additional material. Papua: 1, Mt. Day- man, Maneau Rge., 700 m, "N. Slope No. 6," July 13-20, 1953 (Geoffrey M. Tate, AMNH). N-E. N. G.: 1, Torricelli Mts., Mobitei, 750 m, Feb. 28-Mar. 4, 1959 (W. W. Brandt, Bishop Mus. ).
Notes. Previously known only from the type series from Dobodura.
Genus EUPLENES Schmidt-Goebel
Euplenes laetus Darlington
Darlington 1952, Part II, p. 122.
Additional material. N-E. N. G.: 1, Wau, Morobe Dist., 1200 m, Dec. IS, 1961 (Sed- laceks); 1, Sepalakambang, Salawaket Rge., 1920 m, Sept. 15, 1956 (E. J. Ford, Jr., Bishop Mus.); 1, Wareo, Finschhafen ( Finch Haven ) ( L. Wagner, South Austra- lian Mus.). West N. G.: 1, Star Rge., 1300
274 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
m, Bivak 39, July 26, 1959, and 1, Star Rge., 1500 ni, "mond Ok tenma," May 18, 1959 (both Leiden Miis., Neth. N. G. Exp.).
Notes. This is evidently primarily a mountain-living species.
Euplenes apicalis Darlington
Darlington 1952, Part II, p. 123.
Additional material: 104, from many lo- calities in Papua (including 44 additional specimens from Dobodura, the type locality) at low altitudes and in mountains up to Wau ( many specimens, but none certainly above 12.50 m); Kassem Pass, N-E. N. G., 1460 m; and Mt. Giluvve, Papua, 2550 m.
Genus DICRANONCUS Chaudoir Dicranoncus queenslandicus (Sloane)
Darlington 1952, Part II, p. 124.
Occurrence in New Guinea. N-E. N. G.: 22, Wau, Morobe Dist., 1050, 1200 m, dates in Feb., Mar., May, July, Oct., Nov., Dec, 1961-1964 (Sedlaceks), some in m. v. light trap; 3, Western Highlands, Baiyer R., 1150 m, Oct. 18, 19, 1958 (Gressitt); 1, Finschha- fen, Huon Pen., 10 m, Apr. 16, 1963 (Sed- laceks), m. V. light trap.
Note.^. As I stated in 1952, this species, although previously unrecorded from New Guinea, is known from many localities from India and Ceylon to NE Australia. It prob- ably inhabits grassland. The length of New Guinean specimens is 6.6-7.2 mm.
Genus AGONUM Bonelli
Subgenus SERICODA Kirby
Kirby 18.37, Richardson's Fauna Boreali- Americana,
4: 14. Tanaka 1960, Kontyu, 28: 90. Lindroth 1966, Opuscula Ent., Supplementum 29:
565.
Diagnosis. Small, often flattened, parallel- sided, black Agonini with rather short an- tennae; with full complement of 2 pairs supraocular and 2 pairs lateral prothoracic setae and 3 punctures on 3rd elytral inter- val; 5th hind-tarsal segments with accessory
setae; i copulatory organs with both para- meres tipped with setae.
Description. None required here.
Tij]ye sj)eciesf. A(:,onum bemhidioides \ Kirby, of North America.
Subgeneric distribution. Previously known to be circunipolar in the northern hemi- sphere south to the Himalayas, Japan, and Formosa (only at high altitudes on latttT, according to Ueno, personal communica- tion ) ; now found to be represented also ( by A. ceylariicum, below) in tropical Asia and on islands east to the Philippines and New Guinea. Besides A. ceylanicum, a second species of Sericoda has been found in the Philippines: Agonum (Sericoda) quadri- punctatum De Geer, a northern circumpolar species previously known to reach the Himalayas and Japan, occurs also on Luzon (specimens in British Museum with Bott- cher's field Nos. 167 and 169, indicating "Haights-pl(ace)," at high altitudes in the mountains near Baguio, March 1912).
Notes. The occurrence of Sericoda in New Guinea is the first proven occurrence of any primarily north-temperate group of Agonum so far east and south in the Indo- Australian Archipelago.
According to Lindroth, the species of Ser- icoda "are excellent flyers and share a bio- logical property, not yet understood: they are attracted by burning wood and some- times appear in great number during and after forest fires, . . . when the ashes are still hot."
Agonum {Sericoda) ceylanicum (Motschul- sky)
Motschulsky 1859, Etude Ent., 8: 36 (Agotw-
tharax ) . Andrewes 1930, Cat. Indian Carabidae, p. 24
(Anchoincnu.s). 1931, J. Federated Malay Mus., 16: 436
(Anchomcnus) .
1933, Cat. Carabidae Sumatra, p. 328.
1947, Arkiv f. Zoologi, 38, No. 20: 9.
Landin 1955, Arkiv f. Zoologi, ser. 2, 8: 403, 441.
philippiiwnsc Jedlicka 1935, Acta Soc. Ent. Czech- oslovakia, .32: 79 (NEW SYNONYMY).
?karasatcai Tanaka 1960, Kontyu, 28: 91, 94, figs. 2, 5 (NEW SYNONYMY).
The CARABiD BEETLES OF New Guinea • Darlington 275
Types. Of ceijlanicum, from Ceylon, should be in Moscow University Museum; of philippinense, from the Philippines, in British Museum (the type has Bottcher's field No. 46, which, according to his not- very-legible notebook, indicates the locality Mt. Polis, apparently in Mindanao, Feb. 1918); of karasawai, from Japan, in Karasawa's collection. (Of these, only the type of pliilippinense has been seen. )
Description (for recognition in New Guinea). A small, black agonine which runs to Notagonum in my key to genera of New Guinean Agonini ( 1952, Part 2, pp. 114-116) but is distinguished by presence of distinct accessory setae on the fifth hind- tarsal segments and by setae on the tips of the S parameres. Length in New Guinea 5.2-6.0 mm.
Occurrence in Neiv Guinea. West N. G.: 1, Wissel Lakes, Arabu Camp, 1800 m, 1939 (H. Boschma, Leiden Mus.). N-E. N. G.: 1, Wau, 1200 m, Sept. 26-27, 1964 (Sed- lacek), m. v. light trap; 4, Okapa, June 12, 16, Sept., 1964 (Homabrook); 1, Purosa (near Okapa), 2000 m, Aug. 21, 1964 (Sedlaceks); 2, 1:3-20 km SE Okapa, 1800- 1900 m, Aug. 27, 1964 (Sedlaceks); 1, L. Sirunki, 2600-2900 m, June 15, 1963 (Sed- lacek). Papua: 1, Mt. Dayman, Maneau Rge., 2230 m, "N. Slope N. 4," May 19- June 19, 1953 (Geoffrey M. Tate, AMNH).
Notes. At the British Museum in March, 1968, I compared a 6 ceijlanicum from Ceylon with the 6 type of philippinense Jedlicka and found no characters to dis- tinguish the latter, and I have compared these with New Guinean specimens and also with Tanaka's excellent description and figures of the unique type of karasawai. My conclusion is that all the specimens con- cerned are probably referable to ceijlani- cum. However, this conclusion is tentative. My material is inadequate for a final study either of the synonymy or of the geographic variation of the species. If my tentative con- clusion is correct, Agonum (Sericoda) ceij- lanicum is now known from Ceylon, Burma, Japan, Sumatra, Borneo, the Philippines (Mindanao, Luzon), and New Guinea.
The fact that ceijlanicum is so widely distributed on New Guinea at moderate and high altitudes suggests that it has reached the island naturally, by island-hopping (or more properly mountain-hopping) across the Malay Archipelago.
Genus NOTAGONUM Darlington Darlington 1952, Part II, pp. 115, 127.
Diagnosis (revised). Small or medium- sized (4.8-10.0 mm); form Agoniim-, Platijnu.s-, or Bemhidion-\ike; never brightly colored, usually not or not strongly irides- cent; wings full (except in ambulator, q. V. ) ; with 2 pairs supraocular and 2 pairs lat- eral prothoracic setae and 3 dorsal punctures on 3rd elytral intervals; elytral apices simple, denticulate, or spined; tarsi slender, 5th segments without obvious accessory setae; otherwise in general without unusual or striking characters.
Notes. In 1952, I assigned 25 species and 5 additional subspecies of relatively un- specialized New Guinean Agonini to this "genus of convenience." One previously known Australian and 6 new species are added to the number now. Three species described by Maindron and listed by me (Part II, pp. 129-130) as "not identifiable from description" have now been seen ( the types, at Paris); they are all slightly differ- ent from anything described by me, but the differences are hardly more than sub- specific; these species are listed and dis- cussed below, but are not included in the totals given above or in the statistical analy- ses of the New Guinean carabid fauna in preceding pages. Five other specimens of Notagonum have been seen that I cannot place satisfactorily but that are not distinct enough to describe now. This whole genus obviously needs third-stage taxonomic study. A new key to the species seems un- necessary now, but I have keyed out the 4 members of the "reversum group" (q. v.).
Species of Notagonum are now known to be diverse at middle altitudes in New Guinea as well as in the lowlands. Eight have been found at or near Wau ( see under
276 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
individual species below), and others may occur there. (I found 12 species at Dobo- dura. )
Notagonum alberfisi (Maindron) Dadington 1952, Part II, p. 130 (Colpodes).
Description (revised). Form and charac- ters c. as in vile Darlington. Head 0.84 width prothorax. Prothorax: lateral mar- gins slightly translucent; sides sinuate, with posterior angles relatively well defined (much better defined than in novae- <iuineae) and baso-lateral impressions punctate/roughened (more than in novae- ouineae). Lower smiace: abdomen with a little pubescence at extreme base (just a few, sometimes only one or two hair-bearing punctures near middle of base). Legs: 4th hind-tarsal segments with good lobes, outer lobe longer than inner. Length 7.6-8.4 mm.
Types. From "Fly River, L. M. D'Albertis 1876-77"; 3 examples in Maindron Coll., Paris Mus. (seen); and 14 additional speci- mens (not types) with same data in Paris Mus. and MCZ.
Notes. This is close to or the same as the species that I have described as N. vile (1952, Part II, p. 135) but has a slightly narrower head and a less pubescent abdo- men than typical examples of vile. I think populations of this species (if it is all one species) in different parts of New Guinea are distinguishable, but they need much more study than I can give them now. In the meantime, I do not care to state synonymy or to designate a lectotype for alhertisi.
Notagonum novae-guineae (Maindron)
Darlington 1952, Part II, p. 130 (Colpodes).
Description (revised). Form and charac- ters c. of N. vile Darlington. Head 0.84 width prothorax (minimum width for vile). Prothorax with lateral margins moderately translucent; side sinuate near base, posterior angles blunted, baso-lateral impressions scarcely punctate. Elytra scarcely im- pressed; subapical sinuations well defined; apices simple; striae slightly punctulate.
Lower surface: abdomen inconspicuously or perhaps not pubescent even at base. Legs: 4th hind-tarsal segment with very short lobes, outer lobes slightly longer than inner. Lengtli 6.5 mm.
Ti/pe. A single 9 labeled "B. Jamour [?] . . . ■12.VIII. 1903"; in Maindron Coll., Paris Mus. (seen).
Notes. This too (like alhertisi, above) is apparently an earlier name for Notagonum vile Darlington, but (for reasons given in Notes under alhertisi) I am not ready to state the synonymy formally.
Notagonum laglaizei Maindron Darlington 1952, Part 11, p. 130.
Description (revised). An ordinary, Agonutn-Mke Notagonum. Head 0.73 width prothorax. Prothorax with lateral margins somewhat (not strongly) translucent; sides scarcely sinuate posteriorly; basal angles obtuse; baso-lateral foveae not distinctly punctate. Elytra not impressed; subapical sinuations distinct; apices obtusely (slightly) angulate, with sutural angles dentate; striae not distinctly punctate. Lower .snaiace: abdomen not pubescent. Legs: 4th hind- tarsal segments with moderate lobes, outer lobe longer than inner. Length 7 mm.
Type. A single i , labeled "Timmena, 6. IV [IX ?] 1904."; in Maindron Coll., Paris Mus. (seen).
Notes. This is probably the species de scribed by me as Notagonum dentellumi (1952, Part II, p. 147), but the latter is a variable species, and I do not like to state the synonymy without thorough restudy of all available material, which I can not undertake now.
Notagonum submetallicum (White)
White 1846, Voyage Erebus and Terror, Ent., p. 2 ( Colpodes ) .
Csiki 1931, Coleop. Cat., Carabidae, Harpalinae 5, p. 873 (see for synonymy and additional refer- ences ) .
Darlington 1963, Breviora (Museum Comparative Zoology) No. 183, pp. 2, 3, figs. 5 (4th hind- tarsal segment), 7 (elytral apex).
The carabid beetles of New Guinea • Darlington 277
Description (for recognition only). With characters of genus; large, rather slender; brown, faintly metallic; not strongly shining, entire upper surface microreticulate, the meshes c. isodiametric or only weakly trans- verse even on elytra; head less than 0.80 width prothorax, with eyes normal; elytra with outer-apical angles not defined, apices c. rounded, not distinctly denticulate; 4th hind-tarsal segments emarginate, not lobed; length c. 9.5-10.0 mm.
Type. From Australia; present location unknown.
Occurrence in New Guinea. West N. G.: 4, Wissel Lakes: Itouda, Kamo Vy., 1500 m, Aug. 14, 1955 ( Gressitt ) ; Waghete, Tigi L., 1700 m, Aug. 16, 1955 (Gressitt), light trap; L. Paniai, 1750 m, 1939 ( H. Boschma, Leiden Mus.); Arabu Gamp, 1800 m, 1939 (H. Boschma, Leiden Mus.). Also 1, Wa- mena, 1700 m, Feb. 10-25, 1960 (T. G. Maa, Bishop Mus.).
Notes. In my key to species of Notago- num of New Guinea (Part II, pp. 130ff ) this runs to couplet 13 but fits neither species there named, differing from oltum Darling- ton in having the elytra not more than usually narrowed toward base, and from sigi Darlington in being heavily micro- reticulate. It is in fact probably not directly related to either of these species.
N. suhmetaUicum is very common in southern temperate Australia, where it is found in a variety of wet places. It occurs much more rarely northward in eastern Australia into the subtropics and tropics r I found it near Brisbane, and on the Atherton Tableland at near 1000 m altitude in grass growing over cool running water (Darling- ton, 1963). Its occurrence in West rather than eastern New Guinea is, however, sur- prising.
Although there may be slight differences between New Guinean and Australian in- dividuals, I do not wish to describe a New Guinean subspecies. Gomparison of the copulatory organs of a <^ from Brisbane, Australia, and a c^ from Wissel Lakes, New Guinea, shows only slight differences which may be individual rather than geographic.
The species needs study as a whole, includ- ing statistical analysis of population samples from different parts of Australia, before its status in New Guinea can be settled. Tliis is third-stage taxonomy, which I cannot undertake now.
Notagonum angusfellum Darlington Darlington 1952, Part II, pp. 130, 133.
Additional material. Ninety-six, from widely scattered localities including the fol- lowing. Papua: 2, Owen Stanley Rge., Goilala: Tapini, 975 m, Nov. 16-25, 1957 (W. W. Brandt, Bishop Mus.). N-E. N. G.: 50, Wau, 1100, 1200, 1300 (1 only), 1500 (1 only), dates in \hir., Apr., May, July, Aug., Sept., Nov., Dec, 1961-1966 (Sed- laceks and others); 14, Finisterre Rge., Moro, Stn. No. 85, c. 5550 ft. (c. 1700 m), Oct. 30- Nov. 15, 1964 ( Bacchus, British Mus. ) . West N. G.: 11, Star Rge., various localities, 1260, 1300, 1500 m, dates in May, June, 1959 (Leiden Mus., Neth. N. G. Exp.).
Notes. Although this species occurs at low altitudes ( for example at Dobodura ) as well as in the mountains, it lives only ( in my experience) on the banks of small, rapidly flowing streams, and this habitat restriction may retard its dispersal. Detailed taxonomic study of available material would probably show that it varies geographically within the limits of New Guinea.
Nofagonum subnigrum Darlington Darlington 1952, Part II, pp. 130, 134.
Additional material. Papua: 21, Mt. Lamington, 500 m, June 1966 ( P. Shanahan, Bishop Mus.); 1, Biniguni, Gwariu R., 150 m, "No. 3," Julv 27-Aug. 14, 1953 (Geoffrev M. Tate, AMNH); 67, Peria Gk., Kwagira R., 50 m, "No. 7," Aug. 14-Sept. 6, 1953 ( Geoffrey M. Tate, AMNH ) .
Notes. This species is very distinct from populations of other species in eastern New Guinea, but its relationships with western New Guinean forms are undetermined (see Notes under original description ) .
278 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Notagonum vile Darlington Darlington 1952, Part II, pp. 130, 135.
Additional material. One hundred and sixty-eight, from all three political divisions of New Guinea, including 33 from Wau & vie, 1050, 1200, 1250, 1200-1300, 1500, 1500^1900, 1600 m, Feb., Mar., Apr., May, June, Aug., Sept., Oct., Nov., 1961-1966 ( Sedlaceks and others ) .
Notes. This material, and that recorded by me in 1952, should be sufficient for third- stage taxonomic study of the geographic variation of this species within New Guinea. Such variation obviously does occur, but I cannot undertake to analyze it now. (See also N. albertisi and novae-guineae, previous pages. )
Notagonum Reversum Group
The four species keyed out below form what may be called the reversum group of Notagonum. They share the following char- acters: head usually more than 0.85 width prothorax, with eyes usually (not always) abruptly prominent; prothorax with sides usually sinuate well before base; elytra with humeri obtusely angulate or sub- angulate, with subapical sinuations weak or absent, and apices simply rounded except often denticulate at suture; abdomen not pubescent. The group is notable for occur- ring at relatively high altitudes, compared with most other members of the genus. One of the species, N. amhuJator (below), is the first known Notagonum with atrophied wings, but is so obviously closely related to the winged forms that it cannot be ruled out of the genus.
Key to Species of Notagonxjm Reversum Group
1. Eyes normal; base of prothorax not punctate _.. reversum
- Eyes more abruptly prominent; base of pro- thorax conspicuously punctate 2
2. Prothorax with sides sinuate c. % of length before base reversior
- Sides of prothorax sinuate nearer base 3
3. Winged; length 9.7 mm curiosum
- Wings atrophied; length c. 7.8 mm — . ambulator
Notagonum reversum Darlington
Darlington 1952, Part II, p. 136.
Notes. Still known only from the type series from the Chimbu Valley, Bismarck Range, N-E. N. G., 5000-7000 ft. (c. 1500- 2300 m).
Notagonum reversior Darlington Darlington 1952, Part II, p. 137.
Additional material. N-E. N. G.: 1, Mt. Wilhelm, 2800-2900 m, July 6, 1963 (Sed- lacek ) .
Notes. The 2 types were from Moss Forest Camp, Snow Mountains, West N. G., 2800 m. The individual recorded above is apparently referable to this species, al- though further material may show that it is a geographic form distinguishable by slightly different proportions, etc.
Notagonum curiosum n. sp.
Description. With characters of genus and of reversum group; form as in Figure 28, slender, appendages slender; brown, head and pronotal disc darker, appendages reddish testaceous; moderately shining, re- ticulate microsculpture faint on front and pronotum, distinct and transverse on elytra. Head 0.87 width prothorax; eyes small, more abruptly prominent than usual (but less so than in reversum). Prothorax: width/length 1.11; base/apex 1.27; base and apex deeply margined; lateral margins narrow, subangulate at median-lateral setae, strongly sinuate c. Vr> from base, with pos- terior angles sharply formed, c. right; disc convex, middle line distinct, anterior im- pression not sharply defined, posterior transverse impression deep, baso-lateral im- pressions small; entire base of pronotum punctate. Elytra: width elytra /prothorax 1.65; base margined, margin subangulate at. humeri; subapical sinuations obsolete; apices apparently broadly rounded ( slightly broken ) except angulate or vaguely sub- dentate at suture; striae impressed, not punctulate; intervals c. flat. Inner wings
The cababid beetles of New Guinea • Darlington
279
fully developed. Lower surface: sides of sterna in part punctate; abdomen not pubes- cent. Legs: 4th hind-tarsal segment emarginate, scarcely lobed. Secondary sex- ual characters of i normal; 2 unknown. Measurements: length 9.7 mm; width 3.3 mm.
Type. Holotype $ (Bishop Mus.) from Owen Stanley Rge., Papua, Goilala, Bome, 1950 m, Mar. 16-31, 1958 (W. W. Brandt); the type is unique.
Notes. For comparisons, see preceding Key to Species of Reversum Group.
Nofagonum ambulator n, sp.
Description. With characters of genus and of reversum group, except wings atro- phied; form as in Figure 29, elytra more narrowed basally than usual; brown, head and pronotal disc darker, appendages yel- low; moderately shining, reticulate micro- sculpture faint on head, very light and transverse on pronotum; irregularly trans- verse on elytra. Head 0.87 and 0.84 width prothorax; eyes small but more abruptly prominent than usual, with posterior supra- ocular setae behind posterior eye level. Prothorax: width/length 1.15 and 1.15; base/apex 1.16 and 1.23; base and apex margined; sides broadly sinuate well before base, with basal angles sharply formed, c. right or slightly acute; lateral margins very narrow; disc convex, with middle line and anterior transverse impression nonnal, pos- terior transverse impression deeper (in the type), and baso-lateral impressions small but rather deep; whole basal area strongly punctate. Elytra: width elytra prothorax 1.60 and 1.55; base margined, margin faintly subangulate at humeri; subapical sinuations slight or obsolete; apices independently rounded then sinuate to denticulate sutural angles; striae deep, not punctulate; intervals convex. Inner wings atrophied, reduced to vestiges c. Mi as long as elytra. Lower stir- face: sides of pro- and mesosterna punc- tate; abdomen not pubescent. Legs: 4th hind-tarsal segment emarginate, with very short lobe on outer side. Secondary sexual
characters of i normal; ? unknown. Mea- surements: length c. 7.8; width 2.7-2.8 mm.
Types. Holotype $ (CSIRO) from Mur- mur Pass (W of Mt. Hagen), N-E. N. G., 8600 ft. (c. 2620 m), Nov. 1961 (W. W. Brandt); and 1 ? paratype (Bishop Mus.), 32 km E Wapenamanda (Western High- lands), N-E. N. G., 2500-2700 m, June 9, 1963 (M. Sedlacek).
Notes. Although evidently related to N. reuersior (Darlington 1952, Part H, p. 137), this species seems distinct by prothorax much less sinuate with the sinuations nearer base, as well as by atrophy of wings. These differences may conceivably all be due to a single mutation, but I cannot assume this without further evidence, and the occur- rence of a short winged species of Notago- num is worth recording now.
Nofagonum externum Darlington
Darlington 1952, Part II, pp. 131, 138.
Additional material. Papua: 2, Popon- detta, 25 m, June 1966 ( Shanahan-Lippert, Bishop Mus.), hght trap; 1, Mt. Lamington, 1300-^1500 ft. (c. 400-450 m) (C. T. Mc- Namara, South Australian Mus.).
Notagonum sigi Darlington Darlington 1952, Part II, pp. 131, 143.
Additional material. N-E. N. G. : 3, Wau, 1100 m, Sept. 9, 1961 (Sedlacek).
Notes. N. sigi is superficially very much like vile but lacks the sparse but distinct ventral pubescence of vile and has a slightly narrower prothoracic base, although the proportions of both species vary and may overlap. The fact that the three \A^au in- dividuals were all taken at one time and place suggests that they came from one particular habitat, while vile is apparently much more widely distributed altitudinally and perhaps ecologically.
Notagonum sinuum Darlington Darlington 1952, Part II, pp. 131, 139. Additional material. Papua: 1, Biniguni,
280 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Gwariu R., 150 m, "No. 3," July 27-Aug. 14, 1953 (Geoffrey M. Tate, AMNH). N-E. N. G.: 1, Adelbert Mts., Wanuma, 800-1000 m, Oct. 24, 1958 (Gressitt); 1, Finisterre Rge., Budemu, Stn. No. 51, c. 4000 ft. (c. 1220m), Oct. 15-24, 1964 (Bacchus, British Mus.); 2, Herzog Rge., Vagau, Stn. No. 137, 147A, c. 4000 ft. (c. 1220 m), Jan. 4-17, 1965 ( Bacchus, British Mus. ) .
Notes. This species, now known from all three political divisions of New Guinea, occurs chiefly at moderate altitudes in the mountains, but evidently descends to or nearly to sea level.
Notagonum altum Darlington
Darlington 1952, Part II, pp. 131, 144.
Additional material N-E. N. G.: 4, Mt. W'ilhelm, 2800-2900 m, July 6, 1963 (Sed- lacck); 2, Sarua Kup, Kubor Rge., Oct. 31, 1965 (Dept. Agr. Port Moresby); 6, Finis- terre Rge., S side Mt. Abilala, Stn. No. 100, 8000 ft. (c. 2440 m), Nov. 17, 1964 ( Bacchus, British Mus. ).
Nofagonum margaritum Darlington Darlington 1952, Part II, pp. 131, 145.
Additional material. One hundred one, from all three political di\'isions of New Guinea, including die following. N-E. N. G. : 16, Wau & vie. (incl. Mt. Missim, Mt. Kaindi), 1100, 1150, 1200, 12.50, 1600-1650, 1800, 2.300 m, dates in Jan., Feb., May, June, Sept., Oct., Dec, 1961-1966 (Sedlaceks and others). WestN. G.: 60, Star Rge., various localities, 1260, 1300, 1500 m, dates in May, June, July, Aug., 1959 ( Leiden Mus., Neth. N. G. Exp.), many at light including ultra- violet light; 1, Waigeu Is., Camp Nok, 2500 ft. (c. 760 m), Apr. 1938 (Cheesman).
Notes. This species is apparently com- mon over an exceptionally wide altitudinal range, from near sea level in some cases (see localities given in 1952) to middle altitudes in the mountains.
Notagonum subpuncfum Darlington Darlington 1952, Part II, pp. 131, 146.
Additional material. Papua: 10, Kiunga, Fly R., dates in July, Aug., Sept. 1957 (W. W. Brandt, Bishop' Mus.'); 1, Kokoda, 360 m, Mar. 20, 1956 (Gressitt), light trap; 1, Mamoo Plantation, Northern Dist., Mar. 22, 1956 (Dept. Agr. Port Moresby), at Hght. N-E. N. G.: 2, Minj, Western Highlands, 5200 ft. (c. 1600 m) May 20, 1960 (J. H. Barrett, Dept. Agr. Port Moresby), at m. v. lamp. West N. G.: 1, Star Rge., Sibil, 1260 m, June 16, 1959 (Leiden Mus., Neth. N. G. Exp.), at light.
Notes. This species, including subspecies capitis Darlington (1952: 147), is now known from nearly the whole length of New Guinea, from sea level to moderate altitudes in the mountains. However, it has not yet been found in the Morobe area.
The specimens from the Fly River repre- sent a population which, in size and elytral microsculptiue, is more like subspecies capi- tis from the Vogelkop than like typical suhpunctum from Dobodura (northern Papua) etc., but which differs slightly from capitis in form of elytral apices, which are usually a little more distinctly angulate than in capitis. However, extremes of the two forms are hardly distinguishable. I do not wish to describe a Fly River form now, but mention it as an example of geographic variation at low altitudes within New Guinea. Tliis kind of local geographic varia- tion will probably be found to occur ini many other lowland carabids when ade- quate series from different parts of the island are available.
Notagonum denfellum Darlington
Darlington 1952, Part II, pp. 132, 147.
1963, Breviora ( Museum of Comparative
Zoolog>0, No. 183: 3,5.
Additional material. Twelve, from all three political divisions of New Guinea and' Goodenough Island, including the following localities. N-E. N. G.: 1, Bulolo (Morobe area), 730 m, Aug. 31, 1956 (E. J. Ford, Jr., Bishop Mus.); 1, Western Highlands: Baiyer, 1150 m, Oct. 19, 1958 (Gressitt), light trap; 1, Torricelli Mts., Mobitei, 750 m,
The carabid beetles of New Guinea • Darlington 281
Apr. 1-15, 1959 (W. W. Brandt, Bishop Mus. ) .
Notes. This variable species is widely distributed in New Guinea at low and mod- erate altitudes (subspecies chimhu Darling- ton (1952: 149) is on the Bismarck Range at at least 5000 ft. ( c. 1525 m ) ) and occurs also in tropical northern Australia.
Nofagonum subimpressum Darlington
Darlington 1952, Part II, pp. 132, 149.
Additional material. Eleven, from all three political divisions of New Guinea and Fergusson Is.; none above 500 m (pre- vious highest record, 800 m, in Snow Mts. ); 1, Sangeman Village, nr. Busu R., NE of Lae, 25 m, Aug. 30, 1957 ( D. Elmo Hardy, Bishop Mus. ), at light; not found near Wau.
Nofagonum paludum Darlington
Darlington 1952, Part II, pp. 132, 150.
Additional material. Papua: 1, Popon- detta, 25 m. May 1966 ( Shanahan-Lippert, Bishop Mus.); 1, Normanby Is., W'akaiuna, Sewa Bay, Nov. 11-20, 1956 {W. W. Brandt, Bishop Mus. ) .
Nofagonum addendum Darlington Darlington 1952, Part II, pp. 132, 153.
Additional material. West N. G.: 5, Waigeo Is., Camp Nok, 2500 ft. (c. 760 m), April 1938 (Cheesman). These specimens are in addition to 2 paratypes from the same locality recorded in 1952.
Notes. For comments on the relation of this species to Altagonum vallicola, see the latter, below.
Nofagonum angulum Darlington Darlington 1952, Part II, pp. 132, 154.
Additional material. West N. G. : 5,
Star Rge., various localities, 1260, 1300, 1500, 1800 m, dates in May, June, July 1959 (Leiden Mus., Neth. N. G. Exp.); 1, Wissel Lakes, Arabu Camp, 1800 m, 1939 (H. Boschma, Leiden Mus.); 1, Bokondini,
40 km N of Baliem Valley, c. 1300 m, Nov. 16-23, 1961 (no collector given), Malaise trap. N-E. N. G.: 20, Wau and vie. (Mt. Missim, Mt. Kaindi, Nami Ck., Edie Ck., Bulldog Rd.), 1100, 1200, 1250, 1700, 2000, 2200, 2300, 2350 m, Jan., Feb., Mav, June, Aug., Sept., Oct., Nov., Dec. 1961-1965 (Sedlaceks), some at light including m. v. light, also in Malaise trap; 8, same area (Edie Ck. and Mt. Kaindi), Stn. No. 10, 20, 7000, 8000 ft. (c. 2135, 2440 m), Sept. 17, 22, 1964 (Bacchus, British Mus.); 2, Fera- min, 1200-1500 m. May 11-22, 2.3-31, 1959 ( W. W. Brandt, Bishop Mus.); 1, Swart Vy., Karubaka, 1450 m, Nov. 12, 1958 (Gressitt), light trap; 3, Okapa, May, June, 1965 ( Hornabrook ) ; 2, same localitv, Stn. No. 170, c. 5000 ft. (c. 1525 m),>eb. 4-15, 1965 ( Bacchus, British Mus. ) ; 1, Wonenara, nr. Kratke Mts., 1450 m, June 14, 1966 (Gressitt), hght trap; 1, 11 km S of Mt. Hagen (town), 2000-2300 m. May 20, 1963 (Sedlacek); 1, Sarua Kup, Kubor Rge., Oct. 31, 1965 (Dept. Agr. Port Moresby); 8, Finisterre Rge., Budemu and Moro, Stn. No. 51, 78, 4000, 5500 ft. (c. 1220, 1675 m), Oct. 15-24, Oct. 30-Nov. 15. 1964 (Bacchus, British Mus.). Papua: 1, Owen Stanley Rge., Goilala: Tapini, 975 m, Nov. 16-25, 1957 (W. W. Brandt, Bishop Mus.).
Notes. The occurrence of this ver\' dis- tinct species seems worth gixing in full, as an example of the distribution of a widely spread but strictly mountain living New Guinean carabid.
Nofagonum subrufum Darlington Darlington 1952, Part II, pp. 133, 156.
Additional material. N-E. N. G.: 1, Finisterre Rge., Moro, c. 5550 ft. (c. 1690 m), Oct. 30^Nov. 15, 1964 (Bacchus, British Mus.).
Notes. This very distinct species has been previously known only from two specimens from Rattan Camp, Snow Mts., West N. G., at 1200 m.
Nofagonum subspinulum Darlington Darlington 1952, Part II, pp. 133, 158.
282 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Additional material. Papua: 1, Good- enough Is., "E. Slope No. 10," 900 m, Oct. 24-30, 1953 (Geoffrey M. Tate, AMNH); 1, Mt. Rill, Sudest Is., "No. 10," 250-350 m, Aug. 22, 1956 (L. J. Brass, AMNH). N-E. N. G.: 1, "Krisa, N. New Guinea, Vanimo," Apr. 1939 (Cheesman, South Austrahan Mus.).
Notagonum astrum n. sp.
Description. With characters of genus; form as in Figure 30; black, legs bicolored (femora dark at base pale at apex, tibiae dark, tarsi pale with small black spots at articulations), antennae dark with basal segment pale; moderately shining, reticulate microsculpture light, isodiametric on front, transverse on pronotum and elytra. Head 0.88 width prothorax; eyes large, normal. Prothorax: width/length 1.39; base/apex 1.18; sides rounded almost to base, then slightly sinuate before very obtuse, narrowly rounded posterior angles; lateral margins narrow; baso-lateral impressions small, vaguely punctate; base indistinctly, apex not margined at middle; disc convex, with distinct middle line, indistinct or irregular transverse impressions. Elytra: width elytra/pro thorax 1.72; margins broadly rounded at humeri; subapical sinuations weak; apices narrowly independently rounded; striae impressed, outer (not inner) ones slightly punctulate; intervals slightly convex. Lower siu-face not punctulate; ab- domen not pubescent. Legs: 4th liind- tarsal segment strongly lobed, outer lobe longer than inner. Secondary sexual char- acters of $ normal; ? unknown. Mea.sure- ments: length 8.0 mm; width 3.1 mm.
Type. Holotype S (Leiden Mus.) from Star Rge., West N. G., 1300 m, "Bivak 39," June 28, 1959 (Neth. N. G. Exp.); the type is unique.
Notes. In my key to species of Notago- num (Part II, pp. 130ff) this new species runs to couplet 2 but fits neither part of the couplet, the sides of the prothorax being sinuate relatively near the base but the abdomen not being pubescent. The new
species may actually be most closely related to N. gibbum ( couplet 12 ) but has a wider head and bicolored legs, the latter distin- guishing the present species from any other
Notagonum known to me
Nofagonum exactum n. sp.
Description. With characters of genus; form as in Figure 31, with elytra more oval than usual (but inner wings long and folded ) ; brownish black, lateral margins of elytra and (less distinctly) of prothorax narrowly testaceous, appendages testaceous except antennae browner from 4th segments; shining, elytra subiridcscent, reticulate mi- crosculpture light and so strongly transverse on elytra as to be scarcely distinguishable. Head 0.82 width prothorax; eyes normal. Prothorax cordate; width/length 1.39; base/ apex 1.07; lateral margins moderate; poste- rior angles right and exactly defined; base margined, apex not margined at middle; disc normal, baso-lateral impressions mod- erately deep, vaguely subpunctate, and base slightly longitudinally wrinkled at middle. Elytra quadrate-suboval, each slightly (in- dependently) impressed before middle; width elytra /prothorax 1.64; subapical sinu- ations moderate; apices subtruncate with sutural angles very briefly dehiscent and weakly subdentate; striae impressed, slightly irregular but not distinctly punctulate; in- tervals slightly convex. Lower surface c. im- punctate; abdomen not pubescent. Legs: 4th hind-tarsal segments rather strongly lobed, outer lobe longer than inner. Secondary sexual characters of c^ normal; 9 unknown. Meastirements: length 6.7 mm; width 2.6 mm.
Type. Holotype 6 (Bishop Mus.) from Wau, Morobe Dist., N-E. N. G., 1200 m, July 16-22, 1962 (M. Sedlacek); the type is unique.
Notes. In my key to the species of Nota- gonum (Part II, pp. 130ff) this runs to dentcllum, but the present species has the posterior angles of prothorax much better defined and the elytra more oval. The dif-
The CARABiD BEETLES OF New Guinea • DarUngtotx
283
ference in shape of prothorax is striking on comparison of specimens.
Nofagonum quadruum n. sp.
Description. \\\\\\ characters of genus; form as in Figure 32, rather depressed, elytra impressed near or slightly before middle; black, lateral margins of prothorax and elytra narrowly testaceous-translucent, appendages testaceous except antennae slightly darker except at base; shining, retic- ulate microsculpture absent or faint on front and disc of pronotum, light but distinct and strongly transverse on elytra. Head 0.76 and 0.75 width prothorax; eyes normal. Fro- thorox: width/length 1.45 and 1.49; base/ apex 1.11 and 1.09; lateral margins moderate and moderately reflexed; apex finely mar- gined, base not or indistinctly so; disc with usual impressions, impunctate except vaguely subpunctate in baso-lateral impres- sions. Elytra: width elytra/prothorax 1.51 and 1.53; subapical sinuations strong; apices acutely angulate then emarginate to acutely denticulate (almost spined) sutural angles; striae impressed, not distinctly punctulate; intervals slightly convex. Loxcer swiace not or not much punctate; abdomen not pubes- cent, he OS: 4th hind-tarsal segment rather strongly lobed, outer lobe longer than inner. Measurements: length 8.2-9.8 mm; width 3.2-3.7 mm.
Types. Holotype <^ (Bishop Mus.) and 9 paratypes (some in MCZ, Type No. 31822) from Wau, Morobe Dist., N-E. N. G., 1200 (1 para type 1200-1300) m, dates in Mar., Apr., Oct., 1961-1964 (holotype, Oct. 11, 1962) (Sedlaceks); and 1 additional para- type from Wau, 3400 ft. ( slightly over 1000 m), "3.8.62" (J. J. H. Szent-Ivany, Dept. Agr. Port Moresby ) .
Additional material. Papua: 1, Tapini, (Owen Stanley Rge., c. 1200 m). May 17-19, 1961 (Gressitt). N-E. N. G.: 1, Torricelli Mts., Mobitei, 750 m, Mar. 16-31, 1959 ( W. W. Brandt, Bishop Mus. ) . West N. G. : 1, Star Rge., Sibil, 1260 m, June 16, 1959 (Neth. N. G. Exp., Leiden Mus. ).
Measured specimens. The i holotype
and 1 2 paratype from Wau, 1200 m.
Notes. In my key to species of Notago- num (Part II, pp. 130ff), the present new species runs to margaritum Darlington (couplet 20) but has elytral denticles much more prominent than margaritum, elytral striae not punctulate, and differs in other ways. The individual from the Torricelli Mts. has dark rather than pale legs and may represent an independent population.
Some of the specimens recorded above were taken in light traps including mercury vapor light traps.
Nofagonum sectum n. sp.
Description. With characters of genus; fomi as in Figure 33, with elytra indepen- dently impressed c. % from apex; black, lateral margins prothorax and elytra nar- rowly slightly translucent, appendages dark; moderately shining, reticulate micro- sculpture in part light or indistinct on head and pronotum, more distinct and transverse on elytra. Head 0.72 width prothorax; eyes large, normal. Prothorax wide-subcordate; width length 1.45; base/apex 1.24; lateral margins rather wide; apex margined, base faintly margined; disc weakly convex, baso- lateral impressions moderate, poorly de- fined, irregular but scarcely punctate. Elytra long-subquadrate, width elytra/ prothorax 1.46; apices strikingly modified, with outer-apical angles c. right and sharply defined, each apex then strongly emarginate to second angulation c. opposite end 3rd stria, then again emarginate to moderate spine at sutural angle; striae moderately impressed, sHghtly irregular but not punctu- late; intervals flat or slightly convex. Lower surface virtually impunctate; abdomen not pubescent. Legs: 4th hind-tarsal segments moderately lobed, outer lobe longer than inner. Measurements: length 9.8 mm (in- cluding spines); width 3.6 mm.
Type. Holotype ? (Bishop Mus.) from Mokai Village, Torricelli Mts., N-E. N. G., 750 m, Dea 8-15, 1958 (W. W. Brandt); the type is unique.
Notes. Although based on a single fe-
284 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
male, this species is so strikingly character- ized by form of elytral apices that it is worth describing. In my key to the species of Notagonum (Part II, pp. 130ff ) it runs to N. externum (couplet 7) because the outer- apical elytral angles are sharply defined, but the elytral apices are otherwise much more modified than in externum. These two species are probably not directly related.
Genus VlOLAGONUhA Darlington Darlington 1956, Psyche 63, p. 8.
Diagnosis and description. See reference given.
Type species. Colpodes violaceus Chaud- oir.
Generic distribution. Sec following Notes.
Notes. This genus of medium-sized Agonini is distinguished from Colpodes by shorter head, and from Notagonum by posi- tion of elytral spines. The principal species of the genus is V. violaceum (Chaudoir), which occurs in New Guinea, New Britain, the Solomons, and NE Australia, with sub- species goa Louwerens (1956, Treubia, 23: 221, 231 ) in the Moluccas. V. CColpodes") piceiis ( Andrewes ) represents the group on Samoa. And an undescribed species occurs on the Palau Islands (Darlington, 1970: 23).
Violagonum violaceum (Chaudoir)
Colpodes violaceus auct. including Darlington 1952, Part II, pp. 159, 160.
Additional material. Two hundred sixty- eight, from all three political divisions of New Guinea and Goodenough, Normanby, Woodlark, Rossel, Fergusson, Sudest, and Biak Is., and including the following. N-E. N. G.: 126, Wau and vie, altitudes from 1050 to 1700-1800 m (most from 1050 to 1300 m), dates in every month, 1961-1963 (Sedlaceks), some at light, some in Malaise trap. Papua: 2, Mt. Giluwe, 2500, 2,550 m. May 1, May 27-June 6, 1963 (Sedlacek). West N. G.: 4, Star Rge., various localities, 1220, 1260, 1500 m, dates in May, July, Aug. 1959 (Leiden Mus., Neth. N. G. Exp.).
Notes. This is one of the commonest Carabidae in New Guinea. It evidently occurs throughout the island from sea level to moderate altitudes in the mountains, and rarely at higher altitudes. It is easily recog- nized by its rather broad form, size ( length c. 10 mm), purple or blue color, and elytra each with a single, moderate apical spine c. opposite the end of the second interval. It is partly arboreal, occurring in vegetation including clumps of leaves on low branches in rain forest and also in piles of dead leaves on the ground in forest. It often flies to light.
Genus COLPODES Macleay
Darlington 1952, Part II, pp. 115, 158.
Notes. As I (and many other authors) have used it, this is a "genus of convenience" to which can be referred relatively large, usually tropical Agonini not now referable to more exactly defined generic groups. The single new species described below is probably not directly related to any previ- ously known New Guinean ^'Colpodes" but may be independently derived from Nota- iionum.
Colpodes guega n. sp.
Description. With characters of genus as restricted (Part II, pp. 158-159); form as in Figure 34; dark brown with lateral mar- gins of prothorax and (less distinctly) of elytra translucent, appendages paler; mod- erately shining, reticulate microsculpture faint and c. isodiametric on front, light and strongly transverse on pronotum, still more strongly transverse on (faintly iridescent) elytra. Head 0.74 width prothorax; eyes normal; front not wrinkled. Prothorax: width length 1.39; base/apex 1.41; lateral margins moderate; apex strongly margined, base very narrowly indistinctly so; disc con- vex, middle line distinct, transverse impres- ' sions slight; baso-lateral impressions poorly defined, not distinctly punctate. Elytra: width elytra prothorax 1.47; base margined, margin rounded-obtuse at humeri; apices
\
The carabid beetles of New Guinea • Darlington
285
with outer angles not defined, subapical sin- uations broad, actual apices acutely angulate c. opposite ends 3rd intervals and obtusely angulate (almost minutely denticulate) at suture; striae moderately impressed, not distinctly punctulate; intervals slightly con- vex, 3rd 3-punctate as usual, Sth and 9th not specially modified at apex. Lower surface c. impunctate, without special pubescence. Legs: 4th hind-tarsal segments lobed, outer lobe longer than inner. Secondary sexual characters: ? with only 1 seta-bearing puncture each side apex last ventral seg- ment; i unknown. Measurements: length 11.5; width 4.3 mm.
Type. Holotype 9 (sex detennined by dissection) (Bishop Mus. ) from Guega, W 3f Swart Valley, West N. G., 1200 m, Nov. 14, 1958 (Gressitt); the type is unique.
Notes. In my key to Colpocles of New Guinea (Part II, pp. 159-160), this runs to couplet 9 but fits neither half of the couplet, differing from acuticauda in having elytra with entire basal margins and elytral 3rd intervals 3-punctate, and differing from sinuicauda and simplicicauda in having ely- tral apices conspicuously angulate. See also Notes under genus, above.
Co/podes saphyrinus sloanei Maindron Dadington 1952, Part II, pp. 160, 161.
Additional material West N. G.: 1, Star Rge., Bivak 39A, 1500 m, July 3, 1959, and 1, Star Rge., Sibil, 1260 m, June 21, 1959 (both Netherlands-New Guinea Exp., Leiden Mus.); 1, Waigeu Is., Camp Nok, 2500 ft. (660 m), Apr. 1938 (Cheesman). N-E. N. G.: 52, Wau, 980-1100, 1200, 1250, 1300, 1200-1500 m, Jan. (most), Feb., Mar., May, Aug., Sept., Oct., Dec, 1961-1961 (Sed- laceks), some in light traps; 1, Mt. Kaindi, 1000 m, July 9, 1963 (Sedlaceks); 1, Karimui, 1080 m, July 11-12, 1963 (Sedlacek); 5, Eliptamin Vy., 1200-1350, 1350-1665, 1665- 2530 m, June, Aug., Sept., 1959 (W. W. Brandt, Bishop Mus.). Papua: 1, Mt. Lamington, 1300-1500 ft. (c. 400-450 m) (C. T. McNamara, South Australian Mus.).
Notes. This species is evidently charac-
teristic of the lower mountain slopes in New Guinea. It rarely occurs near sea level (3 at Dobodura, previously recorded) and has rarely been found above 1500 m.
Co/podes helluo Darlington
Darlington 1952, Part II, pp. 160, 162.
Additiomd material. West N. G.: 1, Star Rge., Tenma Sigin, 1800 m. May 20, 1959 (Leiden Mus., Neth. N. G. Exp.), at light (the types came from Rattan Camp in the Snow Mts. at 1150 m). N-E. N. G.: 6, Wau, altitudes from 1200-1400-1500 m, Mar., Apr., June, Sept., Dec, 1961-1964 (Sed- lacek, Gressitt), 1 taken at light; 9, Elipta- min Vy., 1200-1350 m, dates in June, July, Aug., Sept., 1959 (W. W. Brandt, Bishop Mus. ) ; 5, Okapa & vie, some at 1800-1900 m, Aug. 27, 1964 (Hornabrook and Sedla- ceks), some under stones; 3, Wanatabe Vy., nr. Okapa, Stn. No. 174, c. 5000 ft. (1525 m), Feb. 5, 1965 (Bacchus, British Mus); 1, Koibuga, E Highlands, 1500 m, July 5, 1963 (Sedlaceks); 1, Finisterre Rge., Budemu, Stn. No. 51, c. 4000 ft. ( 1220 m), Oct. 15-24, 1964 (Bacchus, British Mus.). Papua: 2, Owen Stanley Rge., Goilala: Tororo, 1560 m, Feb. 15-20, and Rome, 1950 m. Mar. 8-15, 1958 (W. W. Brandt, Bishop Mus.).
Notes. This very distinct species is easily recognized by form ( Fig. 35 ) and form of elytral apices. It is evidently widely distrib- uted at moderate altitudes in New Guinea.
Co/podes laefus (Erichson) Darlington 1952, Part II, pp. 160, 163.
Additional material. Seventeen, from all three political divisions of New Guinea; most at low altitudes (usuallv near sea level) but 1, Wau, 1200 m, Jan. 29, 1963 (Sedlaceks), light trap, and 3, Eliptamin Vy., 1350-1665 and 1665-2530 m, dates in June 1959 (W. W. Brandt, Bishop Mus.).
Notes. Colpodes laetus laetus is now known on the Bonin Is. as well as the Philippines, Celebes, New Guinea (not Australia), Solomons, and New Hebrides, and C. I. pacificus Andrewes is on Samoa
286 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
and the E Caroline Is. ( Kusaie ) ( Darling- ton, 1970: 24).
Co/podes habilis Sloane
Darlinj^non 1952, Part II, pp. 160, 164.
1963, Breviora (Museum of Comparative
Zoology), No. 183: 5.
Additional material. One hundred thirty- four, from all three politieal divisions of New Guinea, few near sea level, most be- tween 1000 and 2000 m, none specifically higher. Included in this total are 52 from Wau, 1050 to 2000 m, dates in every month, many specimens at light.
Note.s. Occurs also on Biiru, New Brit- ain, Solomons, and Santa Cruz Is., and in tropical northern Australia.
Colpodes bennigseni Sloane Darlington 1952, Part II, pp. 160, 165.
Additional material. N-E. N. G.: 34, Wau & vie, 1200 m (most so labeled, but some without altitude), Feb., Mar., Apr., May, June, July, Aug., Nov., Dec, 1961-1966 (Sedlaceks and others), some in light traps; 1, Okapa, Mar. 20, 1964 ( Hornabrook ) ; 4, Eliptamin Vy., 1200-1350, 1665-2530 m, June, July, Aug., 1959 (W. W. Brandt, Bishop Mus. ); 1, TorricelH Mts., Mokai Village, 750 m, Dec. 16-31, 1958 (W. W. Brandt, Bishop Mus.). West N. G.: 4, Star Rge., Sibil, 1260 m, dates in Apr., May, June, 1959 (Leiden Mus., Neth. N. G. Exp.), some at light.
Notes. This species, like habilis (above), occasionally occurs near sea level ( see local- ities cited in 1952) but is mainly character- istic of middle altitudes in the mountains.
Colpodes rex Darlington
Darlington 1952, Part II, pp. 160, 167.
Additional material. West N. G.: 11, Star Rge., Bivak 34A at 850 m, Bivak 39A at 1500 m (most), Sibil at 1260 m, dates in May, June, July, Aug., 1959 (Leiden Mus., Neth. N. G. Exp.), at least 1 at light; 6, Japen Is.,
Camp 2, Mt. Eiori, 2000 ft. (610 m), Sept. 1938 (Cheesman); 7, Waigeu Is., Camp Nok, 2500 ft. (660 m), Apr. 1938 (Chees- man). N-E. N. G.: 6, Eliptamin Vy., 1200- 1350, 1350-1665, 1665-2530 m, dates in June, Julv, 1959 (W. W. Brandt, Bishop Mus.); 2,'Feramin, 1200-1500 m, June 7-14, 15-18, 1959 (W. W. Brandt, Bishop Mus.); 4, TorricelH Mts., Mobitei & Mokai Village, 750 m, dates in Dec. 1958, Feb., Mar., 1959 (\V. W. Brandt, Bishop Mus.); 1, Finisterre Rge., Saidor: Matoko, Aug. 28-Sept. 5, 1958 (W. W. Brandt, Bishop Mus.).
Notes. It is surprising that this striking endemic species, which has been found in all three political divisions of the island ( Papuan examples are recorded in 1952 ) , has not been found at Wau or anywhere in the Morobe area. Otherwise the species seems to be widely distributed in New Guinea at moderate altitudes, rarely below 1000 m.
Colpodes acuticauda Darlington
Darlington 1952, Part II, pp. 160, 170.
Additional material. N-E. N. G.: 1, Wau, 2500 m, Dec. 28, 1961 (Sedlacek); 1, Finisterre Rge., Saidor, Matoko Village, (c. 1500 m), Sept. 6-24, 1958 (W. W. Brandt, Bishop Mus.). West N. G.: 1, Star Rge., Bivak 39, June 28, 1959 ( Leiden Mus., Neth.
N. G. Exp.).
,y
Colpodes sinuicauda Darlington
Darlington 1952, Part II, pp. 160, 171.
Additional material. N-E. N. G.: 1, Wau,, 1700-1800 m, Oct. 7, 1962 (Sedlaceks); 1, Caves near Telefomin, Aug. 1964 ( B. Craig, South Australian Museum). Papua: 1, Mt. Dayman, Maneau Rge., 1550 m, N Slope No. 5, June 30-July 13, 1953 ( Geoffrey M. Tate,AMNH).
Notes. The 2 types were from Sigi Camp, Snow Mts., West N. G., at 1500 m. This distinct species is therefore now known; from all three political divisions of New Guinea, but from a total of only 5 specimens.
The carabid beetles of New Guinea • Darlington
287
Colpodes simplicicauda Darlington Darlington 1952, Part II, pp. 160, 172.
Additional material. West N. G.: 2, Wissel Lakes, Enarotadi, 1550 m, July 25- 29, 1962 (Sedlacek) and Arabii Camp, 1800 m, 1939 (H. Boschma, Leiden Mus. ). N-E. N. G.: 11, Wau & vie. (incl. Edie Ck. and Mt. Kaindi), altitudes 1200 to 2400 (most over 2000) m, Jan., May, June, Oet., 1961- 1963 (Sedlaceks), some in light traps; 1, 16 km NW of Banz, 1700-2100 m, June 28-29, 1963 (Sedlacek); 1, 32 km E of Wapenamanda, 2500-2700 m, June 9, 1963 (Sedlacek); 1, Finisterre Rge., Saidor, Matoko Village, Sept. 6-24, 1958 (W. W. Brandt, Bishop Mus.). Papua: 2, Owen Stanley Rge., Goilala: Borne, 1950 m, Mar. 8-15, 1958 (W. W. Brandt, Bishop Mus.).
Notes. This species too, like several of the preceding ones, is widely distributed at middle altitudes in the mountains of New Guinea, but it ranges somewhat higher than most of the others.
Genus PUCAGONUM Darlington
Plicagonum rugifrons Darlington
Darlington 1952, Part II, p. 174.
Additional material. West N. G.: 3, Star Rge., Bivak 36, 1220 m, July 29, 1959, and Bivak 39A, 1500 m, July 1, 3, 10, 1959 (Leiden Mus., Neth. N. G. Exp.). N-E. N. G.: 2, Edie Ck. (nr. Wau), 2000, 2100 m, Oct. 4-10, 1961, May 31, 1962 (Sedlaceks). Papua: 1, Popondetta, 60 m, Sept. 3-4, 1963 (Sedlacek); 1, Mt. Dayman, Maneau Rge., 1550 m, N Slope No. 5, June 30-July 13, 1953 (Geoffrey M. Tate, AMNH).
Plicagonum fulvum Darlington
Darlington 1952, Part II, pp. 174, 175.
Additional material. West N. G. : 4, Star Rge., Bivak 40, 2330, 2360 m, July 19, 22, 29, 1959 (Leiden Mus., Neth. N. G. Exp.).
Notes. This species is, so far as known, confined to West N. G.; the types were from the Snow Mts. However, the following spe-
cies from the Morobe area, N-E. N. G., is apparently a geographic representative.
Plicagonum kaindi n. sp.
Description. With characters of genus; form c. as in P. fulvum Darlington ( Part II, Fig. 3); dark brown, head almost black, margins of prothorax and appendages more rufous; reticulate microsculpture absent on front, slightly transverse on pronotum, c. isodiametric ( at most slightly transverse ) on elytra. Head 0.78 and 0.80 width prothorax; front longitudinally wrinkled at sides, scarcely so at middle, irregularly slightly impressed, sparsely punctulate. Prothorax c. as in fulvum; width length 1.33 and 1.31; base apex 1.25 and 1.19; lateral margins wide, moderately reflexed, each with usual 2 setae; disc with usual impressions; base and apex strongly margined; baso-lateral impressions formed mainly by angles be- tween base and lateral margins, not dis- tinctly punctate. Elytra: width elytra/ prothorax 1.60 and 1.73; base margined, margin rounded at humeri; subapical sinua- tions moderate, apices more or less (obtusely but usually distinctly) angulate opposite ends 3rd intervals; striae moderately im- pressed, not distinctly punctulate, intervals flat or slightly convex, 3rd with only 1 ( the posterior) dorsal puncture; outer intervals not specially modified apically; no 10th intervals. Lower sutiace virtually impunc- tate; abdomen not pubescent. Legs: 4th hind-tarsal segment emarginate but not lobed; 5th segments without obvious acces- sory setae. Secondary sexual characters normal. Measurements: length 15.5-17.5 mm; width 5.6-6.3 mm.
Types. Holotvpe 6 (Bishop Mus.) from Mt. Kaindi, 16 km SW of Wau, N-E. N. G., 2300 m, Oct. 5-7, 1962 (Sedlacek), m. v. light trap; 15 paratvpes (some in MCZ, Type No. 31825), same locality, 2200, 2300, 2400 m, dates in Jan., June, Oct., 1962-1963 ( Sedlacek ) , most in m. v. light trap; 1 para- type (AMNH), same localit\-, 2050 m. May 25, 1959 (L. J. Brass); and 2 paratypes labeled W'au (but perhaps actually from
288
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Mt. Kaindi), 2400 m, Jan. 9-12, 1962 (Sed- laceks and others ) .
Additional material N-E. N. G.: 1, Bull- dog Rd., 19-29 km S of ^^'an, 2200-2500 m, May 28, 1962 (Sedlacek).
Measured specimens. The S holotype and 1 9 paratype from Mt. Kaindi.
Notes. This may be a geographic form of Plica<i,onum fulvum Darlington (above) but kaindi is larger, darker, with more strongly angulate (but still obtuse) elytral apices, and (if it proves constant) only 1 puncture on the 3rd elytral interval. The single speci- men from Bulldog Rd. is still slightly larger and has virtually simple elytral apices. More material is necessary to show whether it is a distinguishable form.
Genus UTHAGONUM Darlington
Lithagonum annulicorne (Maindron) Darlington 1952, Part II, p. 176.
Additional material: 546, from all three political divisions of New Guinea, and in- cluding the following. Papua: 485, Peria Ck., Kwagira R., 50 m, "No. 7," Aug. 14- Sept. 6, 1953 (Geoffrey M. Tate, AMNH), evidently taken in light traps; these speci- mens referable to subspecies dihitior Dar- lington. N-E. N. G.: 3, Wau, 1090, 1100 m, Jan. 25, 1963, Sept. 2, 1961 (Sedlaceks); these specimens referable to subspecies poUtior Darlington.
Genus IRIDAGONUM Darlington Darlington 1952, Part II, p. 181.
Generic distriljution (revised). Primarily New Guinea; represented also on New Brit- ain by 1 or more undescribed forms.
Notes. Additional material increases the number of known species of this New Guinean genus from 4 to 7. /. fessum is less strongly iridescent above and less strongly punctate below than the other members of the genus, and thus forms a link with Alta- gomim, from which Iridagonum may be derived.
The 7 species now known may be distin- guished by the following revised key.
Key ( Revised ) to the Species of Iridagonum of New Guinea
1. Third interval of elytron 2-punctate (anterior puncture missing) ..__ 2
- Third interval 3-punctate 4
2. Larger (usually over 10 mm; for specimens between 9.3 and 10 mm cf. descriptions); elytral intervals 7 and 8 deeply longitudinally impressed toward apex qiiadripunctum
- Smaller (usually under 10 mm); intervals 7 and 8 less impressed 3
3. More slender (width /length prothorax c. 1.32); black, strongly iridescent above; in
part strongly punctate below
quadripunctellum
- Broader (width/length prothorax r. 1.40); often browner and we;ikly iridescent above ( but variable ) ; more weakly punctate below fessum
4. Two setae over each eye 5
- One seta over each eye ( anterior setae miss- ing ) ___,„ 6
5. Eyes normal sexpunctum
- Eyes smaller, abruptly prominent vigil
6. Form normal; eyes nomial septiwtim
- Fonii fusifonn; eyes small, abruptly promi- nent siibfti.su m
Iridagonum quadripunctum Darlington
Dadington 1952, Part II, p. 182.
Additional material. Twenty-seven, from all 3 political divisions of New Guinea, andi including the following. N-E. N. G.: 11, Wau, 1200 m, dates in Jan., Apr., Oct., Nov., Dec. 1961-1963 (Sedlaceks), some in light traps. West N. G.: 2, Star Rge., Sibil, 1260 m, June 4, 16, 1959 (Leiden Mus., Neth. N. G. Exp.).
Iridagonum fessum n. sp.
Description. Characters of genus; irides- cence weaker and punctation below weaker than usual; form as in Figure 36, very Agoniim-Wke; brownish black, appendages not much paler; reticulate microsculpture of head and pronotum more distinct than usual, transverse meshes being visible even on pronotum at 80x magnification; head 0.71 and 0.72 width prothorax; eyes normal, 2 setae over each eye. Prothorax wide; width length 1.38 and 1.40; base apex 1.16 and 1.21; margins rather broadly weakly reflexed, running into shallow subpunctate
The carabid beetles of New Guinea • Darlington
289
baso-lateral impressions. Elytra sub- quadrate; width elytra/prothorax 1.61 and 1.59; sutural angles with small but distinct denticles; 3rd intervals 2-punctate (anterior puncture lacking); 7th and 8th intervals usually slightly impressed toward apex (variable). Lower sui'jace: sides of sterna in part subpunctate, abdomen not punctate. Secondary sexual characters normal. Mea- surements: length 7.5-8.8 mm; width 3.1- 3.9 mm.
Types. Holotype S (Bishop Mus.) and 18 paratypes (some in MCZ, Type No. 31826) all from Wau, N-E. N. G., 1200, 1700-1800 (1 onlv) m, dates in Jan., Feb., Julv, Sept., Nov., bee. 1961-1964 (holotype, 1200 m, Nov. 6-7, 1961) (Sedlaceks), some in m. V. light trap.
Additional material. N-E. N. G.: 1, Bulolo, 885 m, Aug. 17, 1956 (E. J. Ford, fr.. Bishop Mus.); 1, Sum Sum, 64 km N 3f Wau, 580 m, Feb. 15, 1963 (W. W. Clissold, Bishop Mus.). West N. G.: 1, Sentani, SW Cyclops (Mts.), 100 m, June 15, 1959 (Gressitt); 1, Hoi Maffen, 22 km E of Sarmi, July 18, 1959 (T. C. Maa, Bishop Mus.).
Measured specimens. The 6 holotype and 1 ? paratype.
Notes. This is the most Agonum-liVe species of Iridagonum. Its recognition char- acters are given in the preceding revised Key.
Iridagonum quadripuncfeilum Darlington Darlington 1952, Part II, p. 183.
Additional material. West N. G.: 1, Waris, S of Hollandia, 450-500 m, Aug. 8- 15, 1959 (T. C. Maa, Bishop Mus.); 1, Hoi Maffen, 22 km E of Sarmi, July 18-19, 1959 (T. C. Maa, Bishop Mus.), m. v. light trap.
Notes. Previously kno\\'n only from the I types from Aitape, N-E. N. G.
Iridagonum sexpunctum Darlington
Darlington 1952, Part II, p. 184.
Additional materml. Seventv', from local- ities widely scattered in all 3 political divi-
sions of New Guinea, at moderate altitudes, including 19, Wau, 1200, 1250, 2400 (1 specimen only) m, dates in every month except Apr., June, 1961-1965 (Sedlaceks and others), some in light trap, m. \. light trap. Malaise trap.
Iridagonum vigil n. sp.
Description. With characters of genus; form as in Figure 37, slightly more convex than usual; dark, strongly iridescent above (as usual). Head 0.70 width prothorax; eyes small but abruptly prominent; 2 strong setae over each eye. Prothorax: width/ length 1.18; base /apex 1.30; lateral margins rather narrow, wider basally, running into poorly defined, finely punctate baso-lateral impressions; disc more convex than usual in genus. Elytra subquadrate (not subfusi- form); width elytra prothorax 1.55; sutural angles scarcely (vaguely) denticulate; 3rd interval 3-punctate; outer intervals (7, 8) slightly impressed toward apex. Lower sur- face extensively punctate. Secondary sexual characters oi i normal; $ unknown. Mea- surements: length c. 12 mm; width 4.8 mm.
Type. Holotype S (Leiden Mus.) from Arabu Camp, Wissel Lakes, West N. G., 1800 m, 1939 (H. Boschma); the type is unique.
Notes. See preceding revised Key for differential characters of this distinct spe- cies.
Iridagonum sepfimum n. sp.
Description. With characters of genus; fonri as in Figiu-e 38, slightly more convex than usual; dark, strongl\- iridescent above on elytra. Head 0.64 and 0.65 width pro- thorax; eyes of moderate size, not unusually prominent, 1 seta over each eye (anterior seta missing). Frothorax: width length 1.12 and 1.12; base/apex 1.25 and 1.30; mar- gins moderate, running into rather long, slightly or vaguely punctate baso-lateral impressions. Elytra subquadrate (not sub- fusifonn); width elytra/prothorax 1.53 and 1.56; apices minutely or scarcely subdentic-
290
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
ulate; striae more deeply impressed than usual in genus; 3rd interval 3-punctate; in- tervals 7, 8 weakly ( scarcely ) impressed at apex. Secondary .sexual character.s of S normal; 9 unknown. Measurements: length c. 12 mm; width 4.8 mm.
Types. Holotype S ( Bishop Mus. ) from Edie Ck., 14 km SW of Wau, N-E. N. G., 2000 m, Feb. 13, 1962 (Sedlacek); 1 S para- type (MCZ, Type No. 31828) from Goilala, Bome, Owen Stanley Rge., Papua, 1950 m, iMar. 8-15, 1958 (W. W. Brandt).
Notes. This is another distinct species adequately characterized in the preceding Key. Although the 2 specimens are from different localities, they agree well in form and significant details.
Iridagonum subfusum Darlington
Darlington 1952, Part II, p. 185.
Notes. The holotvpe, from Ibele Camp, Snow Mts., West N.' G., 2550 m, is still the only specimen of this distinct species known.
Genus ALTAGONUM Darlington Darlington 1952, Part II, pp. 116, 185.
Diag^nosis {revised). Size small or me- dium (not over 14.5 mm); wing-and-seta formula + w, ( + )-f, -( + ), ( + )( + )( + ) (see Part II, p. 107, for explanation); other characters in general as in Notagonum, but details more variable.
Notes. This "genus of convenience" in- cludes a number of New Guinean species, not necessarily all related, but sharing at least the diagnostic characters stated above. Most species of this genus are mountain- living, although a few occur at or descend to sea level. This is in contrast to the species of Notaoonum, most of which occur at low altitudes, although some of these too are mountain-living. Twenty-four species and 4 additional subspecies from New Guinea were assigned to Altagonum in 1952 . Ten additional species are described now, and several more that may be new are repre- sented in the material before me by single specimens that I prefer not to describe now.
In the following pages all species found in the Morobe area, but not previously re- corded from there, are noted, as are other important records of previously described species.
Iridagonum fessum (described above) might be mistaken for an Altagonum, since the dorsal iridescence and the punctation of the lower surface are less obvious than usual in Iridagonum. In my key to the New Guinean species of Altagonum (Part II, pp. 188ff), fessum would run to couplet 8 but differs from the species there named {nox Darlington, etc.) in being usually browner ( less black ) , with sutural angles of elytra conspicuously denticulate, and baso-lateral area of pronotum as well as sides of sterna plainly punctate in clean specimens.
The 10 new species of Altagonum that I am now describing all have (in addition to the generic characters) one or more of the following special characters: either posterior-lateral (as well as the median- lateral) prothoracic setae lacking, or less than 3 (either 2 or 0) punctures on each 3rd elytral interval, or elytra spined. I therefore offer the following key to all known New Guinean species of Altagonum that have one or more of these characters. This new key supplements but of course does not entirely supersede the Key to the Species of Altagonum of New Guinea in Part II, pp. 188-189.
Supplementary Key to Certain New Guinean Altagonum
1. Posterior-lateral pronotal setae present, and 3rcl intervals of elytra 3-punctate, and ely- tra spined 2
- Posterior-lateral pronotal setae absent or 3rd intervals with only 2 setae or none ( elytra spined or not spined) 6
2. Elytral spines c. opposite ends of sutural intervals 3
- Elytral spines c. opposite ends of 3rd inter- vals 4
3. Length c. 8 mm tutum
- Length c. 11 mm cracens
4. Black, iridescent; length c. 14.5 mm __.
avium
- Elytra greenish or bluish; length c. 11-13 nun 5
The carabid beetles of New Guinea • Darlington
291
5. Prothoiacic margins narrow anteriorly, wide posteriorly cheesman i
- Prothoracic margins wide anteriorly as well as posteriorly __. sororium
6. Posterior-lateral prothoracic setae present 7
- Posterior-lateral prothoracic setae absent 13
7. Third intervals of elytra 2-punctate 8
- Third intervals of elytra impunctate (form of grossidum, etc. ) exidum
8. Elytra not spined 9
- Elytra spined 11
9. Subfusiform; prothoracic base/apex c. 1.70 or more parascapha
- Agon i/m -like; prothoracic base/apex c. 1.50 or less 10
'0. Black; length 6.8-9.7; elytral apices angu- late vaUicola
- Brown; length 5.3-7.5 mm; elytral apices not
angulate grossiihi m
( and ) grossuloidcs
LI. Length c. 8 mm or less (brownish black) scapha
- Larger 12
L2. Length c. 9 mm (elytra submetallic)
regiscapha
- Length c. 11 mm (brown) erugatum
Two pairs supraocular setae present 14
Anterior pair supraocular setae absent 17
Elytra spined stellaris
Elytra not spined 15
Fonn broad Ca/af/ius-like; prothorax width/ length c. 1.30 or more nudicolle
- More slender 16
16. Length c. 10 mm fatmim
- Length c. 8 mm , temieUtim
17. Form ]:)road Ccdathus-\ike; prothorax width/ length c. 1.45 bigcnuDi
- Similar but prothorax more wedge-shaped;
prothoracic width/length c. 1.60
stibconicolle
Altagonum vallicola Darlington Darlington 1952, Part II, pp. 188, 190.
Additional material N-E. N. G.: 1, Bulolo, Morobe Dist. 1000 m, Aug. 17, 1956 (E. J. Ford, Jr., Bishop Mus.); 1, Wareo, Finschhafen ( Rex. L. Wagner, South Aus- tralian Mus.); 1, Adelbert Mts.: Wanuma, 800-1000 ni, Oct. 27, 1958 (Gressitt), Hght trap. West N. C: 3, Japen Is., Camp 2, Mt. Eiori, 2000 ft. (610 m), Sept. 1938 (Cheesman); 1, Waigeo Is., Mt. Nok, Camp 2 (Buffelhorn), June 1938 (Cheesman).
Notes. Except for the specimens from
Japen Island (which are virtual topotypes of subspecies subvividiim Darlington), I have not attempted to assign the Additional specimens to subspecies.
The specimen from Mt. Nok, Waigeo Is., is noteworthy. Externally, except for the absence of anterior-lateral pronotal and an- terior dorsal elytral punctures, it is virtually indistinguishable from 5 specimens of Nota- gonum addendum (above) also from Mt. Nok, but from a different locality on the mountain. The similarity extends to pro- portions, exact form of elytral apices, and other usual characters, except for a very slight difference in the 4th hind-tarsal seg- ment, of which the outer angle is slightly longer than the inner in Notagonum adden- dum, but equal to the inner in Altagonum vallicola. The whole similarity is so great that I was prepared to consider these two insects dimorphic forms of one species, until I dissected the male copulatory organs. These are strikingly different, more slender and more arcuate in N. addendum, much stouter and less arcuate in A. vallicola. Fur- ther study, with more material, will be necessary to decide the real relationship ( if any ) between these two species.
Altagonum grossulum Darlington (and) Altagonum grossuloides Darlington
Darlington 1952, Part II, pp. 188, 191, 194.
Additional material Ninetv-one, from all three political divisions of New Guinea, at altitudes from near sea level to 2770 m (Mt. \\'ilhelm), but not yet found in the vicinity of ^^'au.
Notes. This material forms a confusing taxonomic complex in which variation is notable not only in form (e. g., in relative width length of prothorax), elytral micro- sculpture, and size, but also in punctures of third elytral intervals, in depth of elytral striae, and in color. The third intervals are each 2-punctate in most individuals (an- terior pimcture lacking) as originally de- scribed, but 7 specimens from the \\^issel Lakes area, West N. G., have the third in- tervals 3-punctate on both elytra, although
292 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
these specimens seem referable to iircssiilum or grossuloides otherwise. In color, individ- uals vary from uniformly dark (except lateral margins of prothorax and elytra rather narrowly yellowish-translucent) to elytra mainly pale brown with subapical dark areas on intervals 2-6 ( longest on lat- ter), the dark areas together setting off a squarish apical pale spot that recalls the apical spot of Trichotichnus (Lampetes) iiuttula Darlington ( 196S, Part III, p. 57 ) . The entire color variation occurs in a series from Saidor, Finisterre Range (Bishop Mus.) and in another series from other localities on the Finisterre Rge. ( Bacchus, British Mus.). These series give the impression that the color may be partly genetically dimorphic, Mendelian. Analysis of the var- iation in this new material requires third- stage taxonomic treatment, and is more than I can undertake now. The material is mainly in the Bishop Museum, British Mu- seum, Leiden Museum, AMNH, CSIRO, and South Australian Museum. The following new species, although without punctures on the third elytral intervals, seems other- wise closely related to gios.suliim and gros- suloides and should be considered in future studies of this group of Altagomim.
Alfagonum exufum n. sp.
Description. With characters of genus. Form of A. grossuloides (c. of small Ago- num s. s. ) ; brownish black, appendages yel- lowish brown, lateral margins of prothorax and elytra yellowish-translucent; reticulate microsculpture isodiametric on front, trans- verse on pronotum, strongly transverse on elytra; latter faintly iridescent in proper light. Head 0.66 and 0.65 width prothorax. Prothorax rather large; width length 1.38 and 1.39; base/apex 1.55 and 1.50; sides moderately arcuate, moderately converging and sometimes slightly sinuate before obtuse but c. well-defined posterior angles; disc and impressions as in other members of group, virtually impunctate. Elytra: width elytra/prothorax 1.46 and 1.50; not or not much impressed before middle; humeral
margins angulate, blunted; apices normal, c. simple; striae well impressed, impunctate; intervals usually convex, 3rd without dorsal pimctures. Lower swface: virtually im- punctate; abdomen not pubescent. Legs: 4th hind-tarsal segment simply emarginate. Measurements: length 6.0-7.2 mm; width 2.5-2.9 mm.
Types. Holotype S (Bishop Mus.) and 1 9 paratype (MCZ, Type No. 31S29) from VVau, Morobe Dist., N-E. N. G., 1250 m. Feb. 14 (holotype) and Jan. 11, 1962 (Sedlacek), and additional paratypes as follows. N-E. N. G.: 1, Finschhafen (Wag- ner, South Australian Mus.), "trapped b)- sticky seeds of Pisonia hrunoniana"; 1 Herzog Rge., Vagau, c. 4000 ft. (1220 m). Jan. 4-17, 1965 (Bacchus, British Mus.); 2. "No. 10," Purosa Camp, Okapa area, 195C m, Sept. 23, 24, 1959 (L. J. Brass, Sixtl Archbold Exp., AMNPI). Papua: 3, Mt. Dayman, Maneau Rge., N Slope "No. 5,' June 30-July 13, 1953 (Geoffrey M. Tate, AMNH); 1, Doveta, and 1, Komania, Ama- zon Bay Dist., Aug. and Nov., 1962 (W. W. Brandt, CSIRO).
Measured specimens. The S holotvpc and 9 paratype from Wan.
Notes. In form and most characters this' species resembles grossulum Darlington and grossuloides Darlington, but it differs from these species in lacking dorsal punctures on the 3rd elytral intervals. See also pre- ceding Supplementary Key.
Alfogonum nox Darlington
Darlington 1952, Part II, pp. 188, 194.
Additional material. West N. G.: 1, Star Rge., Bivak 39A, 1500 m, Julv 9, 1959 (Leiden Mus., Neth. N. G. Exp.)'.
Alfogonum magnox Darlington Darlington 1952, Part II, pp. 188, 195.
Additiomd material. West N. G.: 2, Star' Rge., Sibil and Bivak 39A, 1260, 1500 m, June and July 10, 1959 ( Leiden Mus., Neth. N.G.Exp.).
Notes. Specimens seen of possibly related
The carabid beetles of New Guinea • Darlington
293
forms are: 1, Bivak 40, 2330 m. Star Rge., West N. G. (Leiden Miis.); 1, Saidor, Finisterre Rge., N-E. N. G. (Bishop Mus.); 2, Mt. Dayman, Maneau Rge., Papua (AMNH).
Aliagonum japenox Darlington
Darlington 1952, Part II, pp. 189, 196.
Additional materia}-. 1 erushcd ?, Japen Is., Camp 2, Mt. Eiori, 2000 ft. (610 m), Sept. 1938 (Cheesman).
Aliagonum pubinox Darlington
Darlinston 1952, Part II, pp. 189, 196.
Additional material N-E. N. G.: 2, Mt. Kaindi, 16 km SW of Waii, 2300 m, Oct. 6, 5-7, 1962 (Sedlacek), light trap.
Aliagonum nociellum Darlington
Darlington 1952, Part II, pp. 189, 197.
Additional material. N-E. N. G.: 2, Waigeu Is., Camp Nok, 2.500 ft. (c. 660 m), Apr."l93S (Cheesman); 1, Wau, 1450 m, Feb. 6, 1963 (Sedlacek); 1, Mt. Kaindi (near Wau), 2350 m, Jan. 10, 1962 (Sedk- ceks); 1, Lae, July 1944 (F. E. Skinner, Purdue U. Coll.); 1, EHptamin Vy., 1200- 1350 m, June 19-30, 1959 (W. W. Brandt, Bishop Mus.); 1, 22 km SE Okapa, 2100 m, Aug. 28, 1964 (Sedlaceks). Papua: 2, Popondetta, 25 m, May, June 1966 (Shanahan-Lippert, Bishop Mus.), light trap; 2, Goodenough Is., E Slope No. 10, 900 m, Oct. 24-30, 1953 (Geoffrey M. Tate, AMNH).
Aliagonum planinox Darlington
Darlington 1952, Part II, pp. 189, 198.
Additional material. N-E. N. G.: 1, Wau. 1250 m, Jan. 5, 1963 (Sedlacek), m. v. light trap. Papua: 1, Goodenough Is., E Slope No. 10, 900 m, Oct. 24-30, 1953 ( Geoffrey M. Tate, AMNH); 2, Fergusson Is., Mts. between Agamoia and Ailuluai, 900 m, "No. 4," June 5-17, 1956 (Fifth Archbold Exp., L.J. Brass, AMNH).
Aliagonum diluiipes Darlington
Darlington 1952, Part II, pp. 189, 198.
Additional material. N-E. N. G. : 4, Wau, 1200 m, dates in Sept., Nov., Dec, 1961-
1965 (Sedlaceks & others), some in m. v. light trap; 1, Karimui, S of Goroka, 1000 m, June 8, 1961 (Gressitts). West N. G.: 4, Swart Vy., Karubaka, 1450 m, Nov. 10, 12, 1958 (Gressitt), light trap; 8, Star Rge., Sibil and Bivak 39A, 1260, 1500 m, dates in May, June, July, 1959 (Leiden Mus., Neth. N. G. Exp.), some at light.
Notes. Some of the specimens listed above, especially those from the Star Range, are intermediate between diliitipes and europhilum Darlington. Only third-stage taxonomic study can decide the status of these two species.
Aliagonum europhilum Darlington
Darlington 1952, Part II, pp. 189, 199.
Additional material. N-E. N. G.: 1, Mt. Piora, (Kratke Rge.), 2100 m, June 12,
1966 (Gressitts), Malaise trap.
Notes. The types were from the Snow Mts., West N. G. The single specimen recorded above is definitely europhilum rather than dilutipes, with prothoracic length width only c. 1.20.
Aliagonum pallinox Darlington
Darlington 1952, Part II, pp. 189, 200.
Additional material. N-E. N. G.: 3, Wau, 1200, 1700 m, dates in Feb., June, Sept., 1961, 1963 (Sedlaceks), 2 of these in light trap; 3, Eliptamin Vy., 1200-1350, 1665- 2530 m, June 19-30, Aug. 16-30, 1959 (W. W. Brandt, Bishop Mus.).
Aliagonum caducum Darlington
Darlington 1952, Part II, pp. 189, 202.
Additional materia]. One hundred thirty- three, from all 3 political divisions of New Guinea, altitudes from 1200-2770 m, mate- rial mainly in Bishop Mus., British Mus.,
294 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Leiden Mus., AMNH. Included are: 23, Wau and Edie Ck., 1200 to 2400 m, Jan., Feb., May, Oct., 1961-1963 (Sedlaceks), some in light trap.
Notes. This material would probably re- pay third-stage taxonomic study.
Altagonum sphodrum Darlington Darlington 1952, Part II, pp. 189, 204.
Additional moterial. N-E. N. G.: 1, Mt. Wilhelm, 2800-2900 m, July 6, 1963 ( Sed- lacek); 2, Laiagam, 2520 m, June 24, 1963 (Sedlacek). Papua: 1, Mt. Giluwe, 2500 m, May 1, 1963 (Sedlacek).
Notes. This distinct species appears to be restricted to relatively high altitudes on and near (west and south of) the Bismarck Rge.
Alfagonum posfsulcatum Darlington Darlington 1952, Part II, pp. 189, 205.
Additional material. Thirty-four, from localities in N-E. N. G. and West N. G., 1200 to 2550 m. Included are: 13, Wau and Edie Ck., 1200 to 2350 m, Oct., Nov., 196.3 (Sedlaceks), some in light trap.
Notes. This is another species that would probably repay third-stage taxonomic study, especially comparison of series now avail- able from the Morobe area (Wau etc.) and the Snow Mts.
Altagonum misim Darlington
Darlington 1952, Part II, pp. 189, 206.
Additional material. One hundred, from all three political divisions of New Guinea, at altitudes from 900 ft. (275 m) to 2400 m. Included are the following. N-E. N. G. : 1, Wau, 2400 m, Jan. 9-12, 1962 ( Sedlaceks and others); 1, Edie Ck. nr. Wau, 2000 m, Oct. 4-10, 1961 (Sedlaceks), light trap; (the types were from Mt. Mis(s)im, near Wau, 6400 ft. (1950 m)); 65, Great Cave, Telefomin, 4700 ft. (1435 m), (Victor Emmanuel Rge.), various dates (B. Craig, T. Hayllar). West N. G.: 7, Star Rge.,
Sibil Vy., 1245 m, Oct. 18-Nov. 8, 1961 (L. W. Quate, Bishop Mus.); 2, Star Rge., Bivak 36, Bivak 39A, 1220, 1500 m, July 10, 29, 1959 (Leiden Mus., Neth. N. G. Exp.). Notes. This species too would repay third- stage taxonomic study. The population on the Star Range is probably distinguishable, characterized by reduced elytral denticles. The ecology of the individuals from Great Cave, Telefomin, will probably be reported on by Mr. Elery Hamilton-Smith, of the South Australian Museum.
Altogonum cheesmani Darlington Darlington 1952, Part II, pp. 189, 207.
Additional material. N-E. N. G.: 5, Wau. 1200-1450 m, Feb., Apr., May, Oct., Nov., 1962, 1963 (Sedlaceks, Shanahan), m. v. light trap. Malaise trap; 1, Mt. Hagen, W Highlands, Oct. 1961 (W. W. Brandt. CSIRO).
Notes. The unique tvpe of this speciej is from Mafulu, Papua,'4000 ft. (1220 m). The color of this species is variable, the elytra of the type being green but of the additional specimens now recorded purple or purplish. The form is as in Figure 39.
Altagonum sororium n. sp.
Description. With characters of genus; form as in Figure 40, rather slender, convex, with large eyes and short-spined elytra; head and disc of pronotum black or red, the pronotum with wide testaceous-translucent margins, elytra bluish or purplish, append- ages reddish or brownish testaceous; shin- ing, reticulate microsculpture absent on front and pronotal disc, strongly transverse on elytra. Head 0.81 and 0.82 width pro- thorax. Prothorax: width length 1.30 and 1.35; base apex 1.73 and 1.63; sides xariably sinuate just before blunted or narrowly rounded basal angles; lateral margins wide, and widely reflexed anteriorly as well as posteriorly, each with seta-bearing puncture on margin at basal angle; apex slightly emarginate; base and apex margined; disc with transverse impressions deep, middle
The carabid beetles of New Guinea • Darlington 295
|
I |
7\\ J3( |
|
60 1 |
|
|
'/ |
|
|
'V |
y- |
Figures 41 — 63A (see text, section [15]): 41, A. parascapha n. sp., (^ holotype; 42, A. tenuellum n. sp., S holo.; 43, A. sfe//or/s n. sp., 6 holo.; 44, A. crocens n. sp., c^ holo.; 45, A. avium n. sp., $ holo.; 46, A. erugafum n. sp.,
? holo.; 47, A. b/genum n. sp., c5 holo.; 48, A. subconico/le n. sp., c5 holo.; 49, Mocu/ogonum seripox n. sp., 9 parotype, Wau; 50, M. daymanpox n. sp., $ holo.; 51, M. canipox n. sp.; i holo.; 52, A/I. afropox n. sp., S holo.; 53, M. depilapox n. sp., $ pore, Mt. Giluwe; 54, M. waupox n. sp., $ pare, Mt. Kaindi; 55, M. kaindipox n. sp.,
S holo.; 56, Pofamogonum posfse/osum n. sp., S holo.; 57, Idiagonum limatulum n. sp., 6 holo.; 58, Monfogonum onox n. sp., c5 holo.; 59, A4. onosso n. sp., 9 holo.; 60, M. nepos n. sp., 9 pore.; 60A, some, copulatory organs, holo.; 61, M. sororcu/o n. sp., 9 holo.; 62, M. pandum n. sp., 6 holo.; 63, M. fugitum n. sp., S holo.; 63A, same, copulatory organs, same individual.
296 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
line rather fine, surface not or not much punctate even in baso-latcral impressions. Elytra long; width elytra/prothorax 1.42 and 1.47; apices each with short acute spine in line of 3rd interval; striae impressed, punctate; 3rd intcr^'al 3-punctate, outer in- tervals not impressed apically. Lower sur- face punctate only at sides of mesosternum; abdomen not pubescent. Legs: 4th hind- tarsal segment rather deeply emarginate but not strongly lobed, outer angle slightly more prominent than inner. Measurements: length 10.8-12.0 mm; width 3.6-4.0 mm.
Types. Holotype S ( Bishop Mus. ) and 2 {S $) paratypes ( 1 in MCZ, Type No. 31830) from Waris, S of Hollandia, West N. G., 450-500 m, Aug. 1-2, 1959 (T. C. Maa), and additional paratypes as follows. N-E. N. G.: 1, Pindiu, Huon Pen., Apr. 20, 1963 (Sedlacek); 1, Korop, Upper Jimmi Vy., 1300 m, July 12, 1955 (Gressitt); 1, Tsenga, Upper Jimmi Vy., 1200 m, July 15, 1955 (Gressitt).
Additional material. Papua: 1, Kokoda- Pitoki, 400 m. Mar. 23, 1956 (Gressitt).
Measured specimens. The S holotype and the 2 paratype from Korop.
Notes. Although probably related to A. cheesmani, the present new species differs in having the prothoracic margins wide anteriorly as well as posteriorly (the mar- gins are relatively very narrow anteriorly in cheesmani) . See also preceding Supple- mentary Key. The present, limited material suggests that clieesmani usually occurs at somewhat higher altitudes ( above 1000 m ) while sororium descends to lower altitudes, although the ranges of the two overlap.
Alfagonum scapha Darlington Darlington 1952, Part II, pp. 189, 208.
Additional material: 12, from all 3 polit- ical divisions of New Guinea (but none from Morobe area), altitudes given from "100-700" to 1700 m.
Notes. This material shows virtually con- tinuous variation from elytra with short spines to elytra only obtusely angulate at
apex. Third-stage taxonomic study is needed to show whether more than one species is involved.
Alfagonum parascapha n. sp.
Description. With characters of genus; form as in Figure 41, more parallel (less fusiform) than scapha, moderately convex; dark rufous to piceous, appendages reddish; shining, reticulate microsculpture indistinct on front and pronotal disc, strongly trans- verse on elytra. Head 0.66 and 0.67 width prothorax. ProtJiorax: width/ltmgth 1.32 and 1.36; base apex 1.70 and 1.75 (but ante- rior angles not distinct, so measurement an approximation ) ; sides c. evenly rounded, at most straighter or faintly sinuate posteriorly; lateral margins moderate, narrow anteriorly, not strongly reflexed, each with seta-bearing puncture a little in from edge of margin near posterior angle; base and apex finely margined; disc with middle line and trans- verse impressions moderate, baso-lateral im- pressions shallow, not or not distinctly punc- tate. Elytra: width elytra/prothorax 1.48 and 1.58; humeral angles c. rectangular; subapical sinuations xirtually absent; apices usually obtusely angulate opposite 3rd in- tervals (but angulation variable), with su- tural angles usually subdenticulate; striae moderately impressed, not distinctly punc- tulate; interxals nearly flat, 3rd 2-punctate (anterior puncture absent), outer intervals not impressed apically. Lower surface vir- tually impunctate; abdomen not pubescent. Legs: 5th hind-tarsal segment moderately emarginate, not distinctlv lobed. Measure- ments: length 8.3-8.5 mm; width 3.0-3.3 mm.
Types. Holotype S (Bishop Mus.) andi 2 paratypes (1 in MCZ, Type No. 31831) from Swart Vy., Karubaka, N-E. N. G., 1550 m, Nov. 8, 1958 (Gressitt); 2 paratypes, Pindiu, Huon Pen., N-E. N. G., 1 labelled! 1200-1450 m, Apr. 18, 20, 1963 (Sedlacek).
Measured specimens. The t^ holotype and 1 5 paratype from Karubaka.
Notes. As compared with A. scapha Dar-
The carabid beetles of New Guinea • Darlington
297
lington, the present new species is less fusi- form, with prothorax relatively slightly wider, slightly more narrowed behind, and with somewhat wider margins, and para- scapha has obtusely angulate or subangulate elytral apices (apices spined or more dis- tinctly angulate in scapha ) . Direct compari- son leaves no doubt that the two species are distinct, and I am not even sure that they are directly related.
Altagonum lofilimbus Darlington Darlington 1952, Part II, pp. 189, 210.
Additional material. N-E. N. G. : 7, Wau, 2400 m ( none lower ) , Jan. 1962 ( Scdlaceks etc.); 8, Mt. Kaindi, 16 km SW of Wau, 2200, 2300, 2350 m, June, Oct., 1962 (Scd- laceks), some in m. v. light trap; 1, Main Finisterre Rge. nr. Freyberg Pass ( N ) , 2550 m, Oct. 1-21, 1958 (W. W. Brandt, Bishop Mus. ). Papua: 1, Owen Stanley Rge., Goilala, Tororo, 1560 m, Feb. 15-20, 1958 (W. W. Brandt, Bishop Mus.).
Notes. Although latilimhus was described from the Snow Mts., West N. G., and al- though another species {paralimbus Dar- lington) is known from Mt. Mis(s)im in the Morobc area, the specimens recorded abo\'e are all latilim])us\ all being relatively wide and having the dark discal color of the ely- tra not extended along the suture toward apex.
Alfagonum nudicolle Darlington Darlington 1952, Part II, pp. 189, 211.
Additional material N-E. N. G.: 2, Daulo Pass ( Asaro-Chimbu Div.), 2800 m, June 14, 1955 (Grcssitt); 1, Mt. Elandora, Oct. 15, 1965 (Dept. Agr. Port Moresby). Papua: 1, Mt. Giluwe, N side, Malgi, 2500 m, May 25-30, 1961 (Gressitt).
Notes: This species (which, like all Alta- gontim, is winged) is evidently widely dis- tributed in the higher mountains ( aboxe 2000 m) of N-E. N.'^G. (south of the Ramu- Markham Valley) and the NW corner of Papua. The long series (the types) that 1
collected on Mt. Wilhelm were under cover on the ground in wet mountain forest.
Alfagonum tenuellum n. sp.
Description. With characters of genus; form as in Figure 42, small, slender; brown, appendages slightly ]:)aler; reticulate micro- sculpture c. isodiametric on front, transverse on pronotum and elytra, in part indistinct or imperfect on latter. Head 0.74 and 0.71 width prothorax. Protlwrax: width/length 1.11 and 1.08; base/apex 1.51 and 1.53; sides weakly arcuate, sinuate near c. right slightly blunted basal angles; lateral margins rather narrow, weakly reflexed, somewhat wider and more reflexed toward base, without setae; disc normally impressed, baso-lateral impressions weak, surface slightly irregular but not punctate. Elytra: width elytra/ prothorax 1.72 and 1.81; margins c. right (slightly obtuse) at humeri; subapical sin- uations slight, apices slightly irregularly rounded, sutural angles not distinctly den- ticulate; striae impressed, not punctulate; intervals slightlv convex, 3rd \\'ithout dor- sal punctures, outer intervals not impressed apically. Lower surface virtually impunc- tate; abdomen not pubescent. Legs: 4th hind-tarsal segment weakly emarginate, not lobed. Measurements: length c. 8 mm; width 2.9-3.1 mm.
Types. Holotype c5 (MCZ, Type No. 31832) from Mt. Maneo, vie. Milne Bay, Papua, "8-10-1961" (John Latter), at light; and 1 2 paratype (AMNH), Mt. Dayman, Maneau Rge., Papua, 1550 m, N Slope No. 5, June 30-July 13, 1953 (Geoffrey M. Tate).
Notes. See Supplementary Key following Genus Altagonum for differential characters of this species.
Alfagonum sfellaris n. sp.
Description. With characters of genus; fonn as in Figure 43, large, slender, with long appendages; black, margins of pro- thorax slightly reddish, femora dark, tibiae, tarsi, and antennae reddish brown; reticulate microsculpturc light and c. isodiametric on
298
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
front, light and trans\'eise on pronotum, more distinct and strongly transverse on ely- tra. Head 0.81 width prothorax. Prothorax: width length 1.01; base/apex 1.22; lateral margins rather wide, scarcely narrower an- teriorly than posteriorly; rather strongly re- flexed especially posteriorly, without seta- bearing punctures; base and apex finely margined; disc rather strongly convex, im- pressed as usual, with poorly defined baso- lateral impressions slightly irregular but not distinctly punctate. Elytra: width elytra/ prothorax 1.55; margins obtusely angulate at humeri; apices each with moderate spine c. in line of 3rd intei"val; striae impressed, not punctulate; intervals slightly convex, 3rd 3-punctate (but middle puncture lacking on left elytron, and 2nd and 3rd striae pulled together and connected by impressions at anterior and posterior punctures ) ; outer in- tervals not impressed apically. Lower sur- face virtually impimctate; abdomen not pubescent. Legs: hind tarsi missing. Mea- surements: length 14.8 mm; width 4.3 mm.
Type. Holotype S (Leiden Mus.) from Star Rge., West N. G., Bivak 39A, 1500 m, July 9, 1959 (Neth. N. G. Exp.); the type is unique.
Notes. See Supplementary Key following Genus Altagonum for differential characters.
Alfagonum cracens n. sp.
Description. With characters of genus; fomi as in Figure 44, moderately large, slender, with long appendages; browaiish black, femora and outer sides of tibiae scarcely paler, tarsi and antennae more brownish; reticulate microsculpture light and c. isodiametric on front, more distinct and transverse on pronotum and elytra. Head 0.90 width prothorax. Trothorax: width length 0.99; base/apex 1.34; lateral margins moderate, slightly broader poste- riorly, moderately reflexed, each with seta- bearing puncture on edge of margin at base; base and apex finely margined; disc nor- mally convex, with usual impressions, baso- lateral impressions moderate, poorly de- fined, not punctate. Elytra: width elytra/
prothorax 1.96; margins at humeri veiy ob- tuse; apices each with short spine near su- ture, with sutural interval running onto spine, but spines slightly dehiscent; striae impressed, not punctulate; intervals slightly con\'ex, 3rd 3-punctate, outer inter- vals not impressed apically. Lower surface virtually impunctate; abdomen not pubes- cent. Leg^s: 4th hind-tarsal segment mod- erately emarginate at apex. Measurements: length 11.0 mm; width 3.6 mm.
Type. Holotype 6 (South Australian Mus. ) from Great Cave, Telcfomin, ( Victor Emmanuel Rge.), N-E. N. G., (4700 ft. = 1435 m) May 31, 1965 (T. Hayllar); the type is unique.
Notes. See Supplementary Key following Genus Altagonum for differential charac- ters.
Altagonum avium n. sp.
Description. With characters of genus; form as in Figure 45, large, subparallel, normally convex; black, femora and tibiae dark, tarsi and antennae browner; shining, elytra and ( less so ) pronotum moderately iridescent, reticulate microsculpture c. iso- diametric on front, imperfect ( probably fine and strongly transverse) on pronotum and elytra. Head 0.67 width prothorax. Pro- tJiorax: width/length 1.15; base/apex c. 1.00 (posterior angles too rounded to mea- sure accurately); lateral margins arcuate for entire length to broadly rounded poste- rior angles, moderately wide, slightly wider posteriorly, rather strongly reflexed, each with seta on inner face ( not edge ) of mar- gin well before base; base and apex finely margined; disc with middle line light but c. entire, posterior transverse impression w-ell marked, anterior impression scarcely indicated, deep baso- lateral impressions formed by union of posterior transverse im- pression with basal-marginal channels, im- punctate. Elytra: width elytra/prothorax 1.30; margins at humeri c. right (slightly obtuse, slightly blunted); apices each with short spine in line of 3rd interval; striae impressed, not punctulate; intervals slightly
The carabid beetles of New Guinea • Darlington
299
convex, 3rd 3-punctate; outer intervals not impressed apically. Lower surface virtually impunctate; abdomen not pubescent. Legs: 4th hind-tarsal segment deeply emarginate, lobed, with outer lobe longer than inner. Measurements: length 14.5 mm; width 5.5 mm.
Type. Holotype 5 (MCZ, Type No. 31S35) from Mt. Elandora (Kratke Rge.), N-E. N. G., Oct. 17, 1965 (collector not given); the type is unique.
Notes. See Supplementary Key follow- ing Genus Altagonum for differential char- acters. This species may actually be related to (ancestral to?) certain Fortagonum but has fully developed inner wings and two strong seta-bearing punctures over each eye.
Altagonum erugatum n. sp.
Description. With characters of genus; form as in Figure 46, CalatJuis-like, sub- fusiform, rather strongly convex; dark brown, appendages somewhat paler; reticu- late microsculpture c. isodiametric on front, transverse on pronotum and elytra. Head 0.66 width prothorax. Prothorax: width length 1.00; base/apex c. 1.58 (anterior angles not distinct); sides weakly arcuate, c. straight but not sinuate to slightly obtuse, narrowly rounded basal angles; lateral mar- gins very narrow anteriorly, wader and weakly reflexed posteriorly, each with seta- bearing puncture on innner face ( not edge ) near base; base and apex finely margined; disc with middle line distinct but light and short, transverse impressions almost obsolete, baso-lateral impressions scarcely impressed, impunctate (slightly roughened). Elytra: width elytra /prothorax 1.48; mar- gins acutely angulate at humeri; apices each with short spine (acute tooth) in line of 3rd interval; striae impressed, scarcely punctulate; intervals c. flat, 3rd 2- punctate ( anterior puncture missing ) ; outer intervals not distinctly impressed apically. Lower surface impunctate; abdomen not pubescent. Legs: 4th hind-tarsal segment
emarginate, scarcely lobed. Measurements: length 11.0 mm; width 3.9 mm.
Type. Holotype 2 (Leiden Mus.), from Bivak 39A, Star Rge., West N. G., 1500 m, July 23, 1959 (Neth. N. G. Exp.); the type is unique.
Notes. See Supplementary Key following Genus Altagonum for differential charac- ters. However, this species does not seem closely related to any other.
Altagonum bigenum n. sp.
Description. With characters of genus, except anterior supraocular setae lacking; form as in Figure 47, subfusiform (broad Agonum-like); black, elytra weakly irides- cent or purplish in some lights, appendages dark; reticulate microsculpture very light, isodiametric on front, trans\'erse on pronotal disc, probably more transverse but scarcely distinct on elytra. Head 0.51 and 0.54 width prothorax; posterior supraocular setae pres- ent, anterior absent; mentum tooth narrow, not emarginate. Prothorax: wddth/length 1.44 and 1.46; base apex 1.63 and 1.62; lateral margins obsolete anteriorly, broad and slightly explanate posteriorly, without seta-bearing punctures; apex finely mar- gined, base not margined; disc with middle line distinct, transxerse impressions very weak, baso-lateral impressions almost obso- lete ( except for very obtuse depression be- tween margin and disc), irregularly faintly subpunctate. Elytra: width elytra/pro- thorax 1.32 and 1.39; base margined, mar- gins c. right or slightly obtuse at humeri; subapical sinuations obsolete or nearly so, apices each with very short spine c. in line of 3rd interval; striae lightly impressed, slightly irregular but scarcely punctulate; interxals nearly flat or slightly convex, 3rd 2 -punctate (anterior puncture missing), outer intervals not impressed toward apex, no 10th intervals. Inner wings fully devel- oped. Lower surface virtually impunctate except mesepistcrna faintly punctate; ab- domen not pubescent. Legs: hind tarsi slender, sulcate each side aboxe, with 4th segment rather long, deeply emarginate,
300
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
slightly lobed, \\'ith outer lobe longer than inner; 5th hind-tarsal segments without ac- cessory setae. Secondary sexual characters normal. Measurements: length c. 10.5-11.5 mm; width 4.3-5.0 mm.
Types. Holotype i (MCZ, Type No. 31837) from Okapa, Purosa, N-E. N. G., Feb. 27, 1965 (Hornabrook), and 5 para- types as follows, all 5 $. N-E. N. G.: 1, E Highlands, Purosa (same locality as holo- type), 1700 m. May 17-25, 1966 (Gressitt & Tawi); 1, Aiyura, 1550 m. May 13, 1966 (Gressitt), light trap; 1, Wau, Morobe Dist., 1200 m, June 25, 1961 (Scdlaceks); 1, Pindiu, Huon Pen., 870-1300 m, Apr. 21-22, 1963 (Sedlacek). Papua: 1, Owen Stanley Rge., Goilala: Loloipa, Feb. 1-15, 1958 (W. W. Brandt, Bishop Mus.).
MeaMired specimens. The S holotype and the 9 paratype from Wau.
Notes. If it were not fully winged (and the capture of one in a light trap suggests that it really flies), I would refer this spe- cies to Fortagonum because of the absence of anterior supraocular setae, and in de- scribing it I have noted characters of value in Fortagonumhut not (as a rule) useful in Altagonum. This species may actually be something like the ancestor from which Fortagonum has been derived. However, in my (practical rather than phylogenetic) ge- neric classification, it fits best in Altagonum, from all other known species of which ( ex- cept the following) it differs in lacking an- terior supraocular setae. In form, it is some- thing like a rather stout A. erugatum (described above), but it differs in so many details that I am not sure there is real rela- tionship. The present new species is repre- sented in extreme western New Guinea by the following presumably geographic iso- late.
Altagonum subconicolle n. sp.
Description. With characters of genus, except anterior supraocular setae lacking; fonu as in Figure 48, c. as in preceding (higenum) except that prothorax is differ- ently formed, relatively slightly wider, with
straighter (but converging) sides and more conspicuous anterior angles. Characters otherwise c. as in higenum, including pro- thoracic margins without seta-bearing punc- tures and 3rd elytral intervals each 2-punc- tate. Proportions: \\'idth of head /prothorax 0.50; prothoracic width length 1.61, base/ apex 1.58; width elytra/ prothorax 1.27. Measurements: length c. 11.0 mm; width 4.7 mm.
Type. Holotype $ (Bishop Mus.) from Fak Fak, S coast of Bomberai, Vogelkop, West N. G., 100-700 m, June 9, 1959'(T. C. Maa); the type is unique.
Notes. Although this is apparently a rep- resentative fomi of A. higenum, I prefer to consider it a species rather than a subspecies until more material showing distribution of both forms is available. The single speci- men of the present form was taken at a substantially lower altitude than any of the specimens of higenum.
Genus MACU/.AGONL/M Darlington Darlington 1952, Part II, p. 213.
Diagnosis (revised). Rather small (5.7- 10.5 mm); narrow Ag07H/;/i-like or more or less fusiform; elytra always blotched or mot- tled with dark and pale; wing-and-seta for- mula (see Part II, pp. 107ff) 4- w, ++, (-)( + )5 -( + )( + ) ( t'xcept M. seripox has 3rd intervals with more than 3 punctures); last ventral abdominal segment of i more or less emarginate at apex { S $ oi all spe- cies except scaphipox now known).
Notes. Six species of this genus were recognized in 1952; 7 more arc described below. The following Key and list of the 13 species constitutes an informal revision of the whole genus. The number of species is surprising, in view of the fact that they are all winged and that some of them fly (to light). These insects, with their mottled or blotched colorations, may be adapted to live in montane grassland, although the only specimen specifically recorded from such a habitat is the type of altipox Darlington (Part II, p. 218), taken in tussock grass above timberline on Mt. Wilhelm.
The carabid beetles of New Guinea
Darlington
301
Key (Revised) to the Species of Maculagonum
1. Eyes normal 2
- Eyes small but abnormally abruptly promi- nent 1 1
2. Third intervals of elytra each with several
( more than 3 ) punctures seripox
- Third intervals 2- or 1 -punctate 3
3. Third intervals 2-punctate 4
- Third intervals 1-punctate 7
4. Elytra with pale marks including relatively large pale spots before middle and near su- ture c. Vi from apex plagipox
- EKtra more finely mottled with dark and pale 5
5. Prothorax rounded at sides; length over 8 mm pox
- Prothorax c. straight at sides (trapezoidal); length under 7 mm 6
6. Less fusiform; head more than 0.80 width prothorax daymanpox
- More fusiform; head c. 0.64 width pro- diorax scaphipox
7. Elytra mainly yellow (see also description) tafapox
- Elytra mainly brown or c. equally mixed black or brown and yellow 8
8. Small, c. 7 mm 9
- Larger 10
9. Reticulate microsculptme distinct on prono- tum and \ery heavy on elytra altipox
- Reticulate microsculpture indistinct on pronotum, \isible but not heavy on elytra canipox
10. Posterior-lateral setae of pronotum present;
pronotum and base of elytra black
atropox
- Posterior-lateral setae absent; pronotum and base of elytra in part yellow depilapox
11. Median-lateral setae of pronotum present .. setipox
- Median-lateral setae absent 12
12. Tliird intervals of elytra 2-punctate _-
kaindipox
- Third intervals 1-punctate ivanpox
Maculagonum seripox n. sp.
Description. With characters of genus; form as in Figure 49, slender, convex; cas- taneous, lateral margins and base of pro- thorax paler, elytra with several rows of testaceous spots sometimes in part irregu- larly coalescent; reticulate microsculpture "aint on head and pronotum, light and some- vvhat transverse on elytra. Head 0.70 and ).71 width prothorax; eyes normal, moder-
ately large. Prothorax: width length 1.28 and 1.38; base/apex 1.77 and 1.74; margins rather wide (in genus) each with seta- bearing puncture at base, without median- lateral puncture; posterior angles right or sHghtly obtuse, narrowly rounded. Elytra: width elytra /prothorax 1.46 and 1.48; mar- gins c. rectangular at humeri; apices rounded, faintly lobed; striae impressed; in- tervals flat or slightly impressed, 3rd multi- punctate, the punctures often in pale spots and sometimes reduced in number ( present posteriorly but not anteriorly). Secondary sexual characters: normal; last \entral seg- ment of i moderately emarginate, of 9 entire. Measurements: length c. 8.5-10.5 mm; width 2.8-3.5 mm.
Types. Holotype $ (Bishop Mus. ) from Wau, "Wan Ck.,'" Morobe Dist., N-E. N. G., 1200-1500 m. Mar. 28, 1963 (Sedlacek). Paratypes: 7 (some in MCZ, Type No. 31839) from Wau and vicinity including Wau Ck. and Mt. Missim, 1200, 1250, 1200- 1500, 1450, 1700, 1600-2000 m, dates in Jan., Feb., Mar., June, Sept., 1961-1964 (Sed- laceks), 1 in light trap, 1 in Malaise trap; 1, Edie Ck. (nr. Wau), Bulldog Rd., "Stn. No. 13," 9700 ft. (2960 m), Sept. 20, 1964 (Bacchus, British Mus.).
Additional material (doubtfully identi- fied). N-E. N. G.: 1, Wau, Edie Ck., 2100 m, Oct. 7, 1961; 1, Sinofi, 30 km S of Kainantu, 1590 m, Sept. 30, 1959 (T. C. Maa, Bishop Mus.); 1, Daulo Pass (Asaro- Chimbu Div.), 2400 m, June 16, 1955 (Gressitt). West N. G.: 1. Star Rge., Bivak 39A, 1500 m, June 30, 1959 ( Leiden Mus., Neth. N. G. Exp.).
Measured specimens. The 6 holotype and 1 2 paratype from Wau.
Notes. The extra seta-bearing punctures of the 3rd elytral intervals distinguish this from all other knowii species of the genus. The specimens listed under additional mate- rial vary in width of prothoracic margins and in form of elytral apices. More material is needed to show whether more than one species is involved.
302 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Maculagonum pox Darlington
Darlington 1952, Part II, p. 215.
Additional material. West N. G. : 1, Star Rge., 1500 m, Bivak 39A, June 30, 1952 (Leiden Mus., Neth. N. G. Exp.).
Notes. The types (including holotype in Leiden Mus.) were from the Snow Mts., West N. G.
Maculagonum p/og/'pox Darlington Darlington 1952, Part II, p. 216.
Description (supplementary). See origi- nal description, based on single teneral c^ . Additional material shows color primarily castaneous with sides of prothorax pale- translucent and elytra each with 2 principal yellow spots slightly before middle and c. V-i from apex, each pair of spots sometimes meeting across suture, and with some addi- tional usually smaller yellow spots espe- cially on (but not confined to) intervals 4, 5, and 6. Frothorax with posterior angles right or slightly obtuse, slightly blunted. Elytra with apices strongly angulate, some- times sul:)denticulate.
Additional material. N-E. N. G.: 4, Tsenga, Upper Jimmi Valley, 1200 m, July 14, 15, 1955 (Gressitt); 1, Kumur, Upper Jimmi Valley, 1000 m, July 13, 1955 ( Gres- sitt); 3, Karimui, S of Goroka, 1000, 1080 m, June 7, 1961, July 13, 14, 1963 (Sed- laceks), 1 labeled also "Rattan"; 1, Adelbert Mts., Wanuma, 800-1000 m, Oct. 25, 1958 (Gressitt); 1, Pindiu, Huon Pen., 870-1300 m, Apr. 21-22, 1963 (Sedlacek). West N. G.: 1, Japen Is., Camp 2 Mt. Eiori, 2000 ft. (610 m), Sept. 1938 (Cheesman).
Notes. The unique S type ( British Mus. ) was from the Cyclops Mts., West N. G.
Maculagonum daymanpox n. sp.
Description. With characters of genus; form as in Figure 50; color ( both individuals slightly teneral) of head, pronotum, and base of elytra brown with prothoracic mar- gins yellowish, elytra except at base nearly evenly mottled brown and testaceous, the
testaceous color forming many spots in a brown reticulation, the pattern becoming slightly irregular dorsally c. Va from apex; reticulate microsculpture faint on front, light and somewhat transverse on pronotum, heavier and c. isodiametric on elytra. Head 0.84 and 0.83 width prothorax; eyes normal, rather large, not abruptly prominent. Pro- thorax subquadrate or trapezoidal with base c. wide as or slightly wider than width at middle; width/length 1.14 and 1.11 (width measured at middle); base/apex 1.57 and 1.71; lateral margins rather narrow, each with seta at basal angle but without median- lateral setae. Elytra ample, slightly nar- rowed toward base; width elytra/prothorax 1.86 and 1.95; humeral margins slightly (very obtusely) subangulate; apices rounded (faintly lobed); striae entire, groovelike; in- tervals flat, 3rd 2-punctate. Secondary sex- ual characters normal: last ventral segment of i shallowly but distinctly emarginate, of 9 entire. Measurements: length 6.5-7.0 mm; width 2.5-2.8 mm.
Types. Holotype <5 (AMNH) and 1 ? paratype (MCZ, Type No. 31840) both from Mt. Dayman, Maneau Rge., Papua, 22.30 m, "N. Slope No. 4," May 19-June 19, 1953 (Geoffrey M. Tate).
Notes. See preceding Key for differential characters of this species.
Maculagonum scaphipox Darlington
Darlington 1952, Part II, pp. 215, 220.
Notes. The unique 2 type (British Mus.), from Orrori, Papua, is still the only in- di\'idual of this species known.
Maculagonum fafapox Darlington
Darlington 1952, Part II, p. 219.
Additional material. Papua: 1, Mt. Day- man, NLmeau Rge., 2230 m, N Slope No. 4, May 19-June 19, 1953 (Geoffrey M. Tate, AMNH).
Notes. The unique type, from Mt. Tafa, Papua, is a 9. The present specimen, which I compared with the type at the British Museum in April, 1968, is a (5 , with
The carabid beetles of New Guinea
Darlington
303
last \cntral segment distinctly but weakly emarginate. Proportions and measurements of this specimen: head/pro thorax 0.76; pro- thoracic width/length 1.29, base/apex 1.39; width elytra prothorax 1.76; length 6.8 mm, width 2.3 mm.
32 km S of Wau, Morobe Dist., N-E. N. G., Bulldog Rd., 2850 m, May 29-30, 1962 ( Sed- lacek), Malaise trap; the type is unique.
Notes. See preceding Key for differential characters of this small, narrow, convex spe- cies.
Maculagonum altipox Darlington
Darlington 1952, Part II, pp. 215, 218.
Notes. This high-altitude species is still known only from the 9 type (in MCZ) from over 3000 m on Mt. Wilhelm, N-E. N. G., and from subspecies pallipox ( 6 holotype in Leiden Mus., 9 paratype in MCZ) from 2800 m on the Snow Mts., West N. G.
Maculagonum con/pox n. sp.
Description. With characters of genus; form as in Figure 51, small, slender, convex; head blackish, prothorax brown with base broadly and sides narrowly yellowish, elytra irregularly mottled dark brown and yellow- ish with larger yellowish spot on each ely- tron behind apical -/i (these spots nearly meeting across suture) and darker marks coalescent around the pale ones and in ir- regular zones centering on 3rd intervals; reticulate microsculpture hght (faint) on front and pronotum, distinct and c. iso- diametric on elytra. Head 0.74 width pro- thorax; eyes normal, moderately large. Pro- tliorax subquadrate, much narrowed in front, scarcely so behind; width length 1.22; base/ apex 1.56; margins narrow anteriorly, wider posteriorly, each with seta at base, without median-lateral setae. Elytra long-oval; width elytra prothorax 1.68; margins obtuse at humeri; apices independently broadly rounded; striae impressed, slightly irregular; intervals flat or slightlv convex, 3rd 1- punctate (puncture in subapical yellow spot). Secondary sexual characters: $ last ventral segment weakly emarginate; 9 un- known. Measurements: length c. 7.0 mm; width 2.4 mm.
Type. Holotype c^ (Bishop Mus.) from
Maculagonum atropox n. sp.
Description. With characters of genus; form as in Figure 52, suboval, convex; head, prothorax, and base of elytra black except base and lateral margins of prothorax slightly rufescent, elytra finely irregularly mottled dark brown and yellowish with slightly larger yellowish spot c. % from apex (and actual apices yellow) and dark color predominant on 3rd and 4th intervals; retic- ulate microsculpture light on front and pro- notum, heavier and c. isodiametric on elytra. Head 0.74 width prothorax; eyes nonnal, rather prominent but not abruptly so. Pro- tlwrax arcuately narrowed anteriorly; width/ length 1.17; base/apex 1.79; margins narrow anteriorly, slightly broader posteri- orly, each with seta at base but without median-lateral setae; posterior angles ob- tuse, slightly blunted. Elytra ample; width elytra prothorax 1.79; margins c. rectangu- lar at humeri; apices independently broadly rounded; striae fine, impressed; intervals flat, 3rd 1-punctate (puncture in subapical yellow spot). Secondary sexual characters: i last ventral segment deeply emarginate, angle of emargination c. rectangular; ? un- knowai. Measurements: length 8.7 mm; width 3.4 mm.
Type. Holotvpe <^ (Bishop Mus.) c. 14 km S Edie CL (near Wau), N-E. N. G., Bulldog Rd., 2405 m, July 4-10, 1966 (G. A. Samuelson), in light trap; the type is unique.
Notes. This rather large, dark, subo^'al species too is adequately differentiated in the preceding Key.
Maculagonum depilapox n. sp.
Description. With characters of genus; form as in Figure 53; brown, base and (more narrowly) lateral margins of prothorax yel-
304 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
lowish, elytra finely mottled with brown and yellow, the yellow color forming a larger common spot less than % from apex, and the dark color coalescent around it; re- ticulate microsculpture faint on front and pronotum, light and slightly diagonally transverse on elytra. Head 0.63 and 0.67 width prothorax; eyes rather large, less prominent than usual. Proihorax narrowed anteriorly, scarcely so posteriorly; width/ length 1.34 and 1.30; base apex 1.60 and 1.59; lateral margins moderately wide, with- out lateral setae; posterior angles slightly obtuse, narrowly rounded. Elytra ample; width elytra/prothorax 1.69 and 1.65; hu- meri right or slightly obtuse; apices irregu- larly broadly rounded; striae lightly im- pressed, slightly irregular; intervals c. flat, 3rd 1-punctate (puncture in subapical yel- low spot ) . Secondary sexual characters nor- mal; c? last ventral segment moderately emarginate. Measurements: length 8.0-S.4 mm; width 3.2-3.5 mm.
Types. Holotype S (MCZ, Type No. 31843) from Okapa, N-E. N. G., Apr. 27, 1965 (Hornabrook). Paratypes: N-E. N. G.: 1, (without head), Okapa, Sept. 16, 1964 (Hornabrook), Umi Simi, 7500 ft. (c. 2285 m), "among moss & epiphytes"; 1, Mt. Chapman, 2000 m, May 5, 1966 (Gressitt); 1 (broken), Wau, Edie Ck., 2000 m, Aug. 14, 1963 (P. Shanahan, Bishop Mus.). Papua: 1, Mt. Giluwe, 2550 m. May 27, 1963 (Sedlacek). West N. G.: 1, Star Rge., 1500 m, Bivak 39A, June 30, 1959 (Leiden Mus., Neth. N. G. Exp.).
Measured specimens: the 5 holotype and ? paratype from Mt. Chapman.
Notes. All specimens agree in lacking posterior- as well as median-lateral pro- thoracic setae, and they agree well also in other characters given in the Key and pre- ceding Description.
Maculagonum setipox Darlington Darlington 1952, Part II, pp. 214, 217.
Notes. Still known only from the uniciue
S type (British Mus.) from Mt. Tafa, Papua, 8500 ft. (c. 2590 m).
Maculagonum waupox n. sp.
Description. With characters of genus,, form as in Figure 54; head, pronotum, and elytra anteriorly black, prothorax only slightly reddish at base and lateral margins, elytra behind base finely mottk-d blackish and reddish yellow with larger common red- dish yellow spot c. % from apex; reticulate microsculpture light or faint on front and pronotum, impressed, fine, isodiametrie on elytra. Head 0.73 and 0.74 width prothorax;; eyes moderate in size, abruptly prominent. Frothorax widest at or near base, much nar- rowed anteriorly; width/length 1.18 and 1.16; base/apex 1.92 and 1.82; margins nar- row, slightly broader posteriorly, each with seta at base, without median-lateral setae. Elytra ample, slightly narrowed anteriorly;; width elytra/prothorax 1.74 and 1.74; mar- gins obtusely angulate at humeri; apices independently rounded; striae finely im- pressed; inter\'als flat, 3rd 1-punctate (punc- ture in yellow spot ) . Secondary sexual char- acters normal; 6 last ventral segment strongly (c. rectangularly) emarginate. Mea- surements: length 7.4-8.5 mm; width 3.1- 3.4 mm.
Types. Holotype <^ ( Bishop Mus. ) from. Wau, Morobe Dist., N-E. N. G., 2400 m, Jan. 9-12, 1962 (Sedlaceks and G. Monteith). Paratypes: 3(2 in MCZ, Type No. 31844) with same data as type; 1, Mt. Kaindi, 16 km SW of Wau, 2200 m, June 8-9, 1962 (Sedlacek), light trap; 1, same locality, 2350 m, Jan. 10, 1962 (Sedlaceks), m. v. light trap. (It seems likely that all speci- mens including the holotype actually came from Mt. Kaindi, near Wau, above 2000 m ) .
Measured specimens. The 6 holotype and a 9 paratype from Mt. Kaindi.
Notes. See preceding Key for this species' distinguishing characters.
Maculagonum kaindipox n. sp.
Description. With characters of genus;
The carabid beetles of New Guinea • Darlington 305
form as in Figure 55; head, pronotum, and base of elytra black, with base of prothorax faintly reddish, elytra behind base irregu- larly mottled black and reddish with two much larger common yellow spots meeting across suture slightly before middle and c. Vi from apex, the black marks coalescent es- pecially around the posterior yellow spot; re- ticulate microsculpture absent on front, faint on pronotum, distinct and c. isodiametric on elytra. Head 0.76 width prothorax; eyes abruptly prominent. Prothorax trapezoidal, widest at base; width/length 1.15 (width measured at middle); base/apex 1.74; lat- eral margins narrow, slightly broader near base, each with seta at base, without median- lateral setae. Elytra ample, scarcely nar- rowed toward base; width elytra /prothorax 1.67; margins at humeri obtuse; apices inde- pendently weakly rounded; striae finely im- pressed; intervals flat, 3rd 2-punctate (punc- tures in yellow spots). Secondary sexual characters: $ last ventral segment moder- ately emarginate; 9 unknown. Measure- ments: length 6.8 mm; width 2.8 mm.
Type. Holotype $ (Bishop Mus. ) from Mt. Kaindi (SW of Wau), N-E. N. G., 2350 m, Apr. 30, 1966 (Gressitt); the type is unique.
Notes. Although much like the preceding \{waupox) and sympatric with it, the pres- ent new species seems clearly distinct, being smaller, slightly different in proportions, with 2 rather than 1 conspicuous yellow ely- tral spots, and with 3rd intervals 2-punctate rather than 1-punctate.
Genus POTAMAGONUM Darlington
Darlington 1952, Part II, p. 221.
Diagnosis (revised). Large, with long, lender appendages; wing-and-seta formula (see Part II, pp. 107ff) +w, ++, -(-), -I- + + ; prothoracic margins wide, translu- :-ent; elytral striae conspicuously inter- rupted; 5th hind-tarsal segments usually with conspicuous accessory setae, but latter sometimes missing ( apparently broken off ) especially in specimens taken in light traps.
Generic distribution (revised). High mountains of much or all of New Guinea.
Notes. The 4 species covered by this diagnosis are all listed in the following pages, which constitute an informal revi- sion of the genus. The type species, P. diaphanum, was found among wet stones and other cover on the spray-drenched banks of turbulent mountain streams, and this is probably the habitat of the other fonns here described.
Key to the Species of Potamagonum
1. Posterior-lateral prothoracic setae present; ely- tra distinctly aeneous postsetosum
- Posterior-lateral prothoracic setae absent; ely- tra brownish black, not distinctly aeneous .._. 2
2. Elytral apices less produced (N-E N. G.) _—
_ diaphanum
- Elytral apices more produced, more acute -__. 3
3. Fragments of ( interrupted ) elytral striae more often linear (Papua) brandti
- Fragments of elytral striae more often puncti- fonn (West N. G. ) julianae
Potamagonum postsetosum n. sp.
Description. With characters of genus and of type species (below) except as in- dicated; form as in Figure 56; color as usual except elytra distinctly aeneous. Head 0.71 and 0.69 width prothorax; eyes slightly larger and genae slightly shorter than in diaphanum. Prothorax more evenly rounded and with more broadly rounded posterior angles than in diaphanum: width length 1.28 and 1.32; base/apex not determined because of rounding of angles; lateral mar- gins each with a seta-bearing puncture near outer edge of margin near base. Elytra: width elytra prothorax 1.42 and 1.41; apices slightly acute but not much produced; frag- ments of ( interrupted ) elytral striae mostly linear. Legs: 5th hind-tarsal segments prob- ably formerly with conspicuous accessory setae, but latter mostly broken off. Mea- surements: length c. 13.5-14.5 mm; width 4.7-5.0 mm.
Types. Holotype i (AMNH) and 1 $ paratype (MCZ, Type No. 31846) both
306 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
from Mt. Dayman, Maneau Rge., Papua, 2230 m, "N. Slope No. 4," May 19-June 19, 1953 (Geoffrey M. Tate).
Notes. The presence of posterior-lateral prothoracic setae distinguishes this from all other kno\\'n members of the genus.
Potamagonum diaphanum Darlington Darlington 1952, Part II, p. 221.
Additional material N-E. N. G.: 1 Wau, Morobe District, Mt. Missim, 1800 m, Apr. 22, 1966 (Gressitt, Wilkes), Malaise trap; 1, Wau, Nami Ck., 1700 m, June 15, 1962 (Sed- laceks); 1, E Highlands, Purosa, 1700 m. May 17-25, 1966 (Gressitt & Tawi).
Notes. The types ( including holotype $ , MCZ) were from the Chimbu Vy. and Mt. Wilhelm, Bismarck Rge., N-E. N. G. The Additional material recorded above indi- cates that the species is widely distributed in the high mountains of N-E. N. G. south of the Ramu-Markham Valley.
Fotamagonum brandti n. sp.
Description. With characters of genus and ( except as indicated in Notes below ) of P. diaphanum. Golor brownish with pro- thoracic margins testaceous-translucent. Head 0.70 and 0.69 width prothorax; details c. as in diapJianum. Prothorax: width/ length 1.19 and 1.22; posterior angles a. as in diaphanum, more narrowly rounded than in postsetosum; lateral margins without setae. Elytra: width elytra/pro thorax 1.40 and 1.38; apices produced as short spines; ( interiaipted ) striae with fragments short but mostly linear. Legs: 5th hind-tarsal segments with well-developed accessory setae. Measurements: length c. 14.5-15.5 mm; width 4.8-5.0 mm.
Types. Holotype ? (Bishop Mus.) and 2 9 9 (1 in MCZ, Type No. 31847) all from Goilala, Bome, Papua, Owen Stanley Rge., 1950 m, Apr. 30-May 2 (holotype), Mar. 16-31, Apr. 1-15, 1958 (W. W. Brandt).
Measured specimens. The 9 holotype and 1 9 paratype.
Notes. This is very close to P. diaphanum
(above), differing mainly in having the apical angulations of the elytra more pro- duced, forming short spines. It may prove to be only a geographic subspecies of diaphaniDU, but until more material is avail- able from more localities I prefer to con- sider it a slightly defined species. The exis- tence of a second, better defined species {postsetosum) also in Papua indicates that specific differentiation has occurred in this group in New Guinea.
Potamagonum julianae n. sp.
Description. \\'ith characters of genus and ( except as indicated in Notes below ) ol P. diaphanum. Color brownish black, with prothoracic margins paler-translucent, as usual. Head 0.66 width prothorax; details c. as in diaphanum. Prothorax c. evenly rounded except broadly weakly emarginatt anteriorly (as usual); width/length 1.24: lateral margins without setae. Elytra: width elytra/ prothorax 1.36; apices pro- duced as short acute spines; fragments of (interrupted) striae relatively short, ofter punctiform ( but variable ) . Legs: 5th hind- tarsal segments with conspicuous accessory setae. Measurements: length c. 14.5 mm: width c. 4.9 mm. I
Type. Holotype 9 (Leiden Mus.) frorr Juliana Bivak, Snow Mts., West N. G., 180C m, Sept. 8, 1959 (Neth. N. G. Exp.); tk type is unique.
Notes. This single specimen probably represents a distinguishable geographic pop- ulation, but more material is needed to show its characters fully. Perhaps it toe (as suggested for brandti above) will prove to be a geographic subspecies of diapluinum.
Genus GASTRAGONUM Darlington
Darlington 1952, Part II, pp. 115, 222.
Notes. Besides additional material ol subrotundum and terrestre (listed below). I have seen three unique specimens in the British Museum that may represent unde- scribed species of this genus. However, third-stage taxonomic study of the whole
I
The carabid beetles of New Guinea • Darlington
307
genus, preferably with more material than is now available, will be necessary to estab- lish their distinctness and probable rela- tionships.
Gastragonum subrotundum Darlington
Darlington 1952, Part II, pp. 224, 225.
Additional material. N-E. N. G.: 1, Edie Ck. nr. Wau, 2000 m, Oct. 10, 1961 (Sed- lacek); 1, Adelbert Mts., Waniima, 800- 1000 m, Oct. 24, 1958 (Gressitt), light trap; 1, Sepalakambang, Salawaket Rge., 1920 ni, Sept. 12, 1956 ( E. J. Ford, Jr., Bishop Mus. ).
Gastragonum terrestre Darlington (and related forms)
Darlington 1952, Part II, pp. 224, 226.
Additional material. Twenty, from various localities mostly in N-E. N. G., at altitudes of 2000 to 2500-2790 m (and 6500, 8600 ft.); specimens in Bishop Mus., CSIRO, and Dept. Agr. Port Moresby.
Notes. These specimens show confusing variation and require third-stage taxonomic treatment, which I cannot give them now. Specimens from near Wau (Edie Ck., Mt. Kaindi, 2000 to 2400 m ) may be referable to terrestroides Darlington (1952: 227), already recorded from Mt. Mis(s)im.
Genus IDIAGONUM Darlington Darlington 1952, Part II, pp. 114, 229, fig. 10.
Generic distribution (revised). Known from high altitudes (probably usually or always in high-mountain forest) in West N. G. ( the Snow Mts. ) and in western N-E. N. G. south of the Ramu-Markham Valley (Bismarck Rge., Bulldog Rd.) and in the adjacent NW corner of Papua ( Mt. Giluwe).
Notes. This very distinct genus is perhaps not directly related to any other genus in New Guinea. Its origin is undetermined. The 6 known species are all now thoroughly flightless and confined to limited areas at high altitudes. However, the species are
closely interrelated, and the different mon- tane populations have perhaps not been long isolated from each other.
Idiagonum asperun) Darlington Darlington 1952, Part II, pp. 230, 231.
Additional material. N-E. N. G.: 1, Mt. Wilhelm, 3000 m, July 4, 1955 (Gressitt); 2, Wapenamanda (W Highlands), 2500- 2700 m, June 9, 1963 (Sedlacek).
Notes. The Wapenamanda specimens are assigned to this species doubtfully, but they are certainly very close and I do not care to describe them now.
Idiagonum giluwe n. sp.
Description. With characters of genus; fonn c. as in asperum Darlington except prothorax slightly broader at base and less angulate at sides; black; head and pronotum dull, elytra moderately shining, micro- sculpture as described below. Head 0.75 and 0.76 width prothorax; eyes small, abruptly prominent (as usual in genus); front finely densely granular. Prothorax c. as in asperutn; width/length 1.14 and 1.17; base/apex 1.05 and 1.05; margins at most weakly angulate near middle; apex mar- gined, base weakly or indistinctly so; disc finely granular (the granules tending to be longitudinal), less strongly (transversely) rugulose than in asperum, and with sur- face especially more finely ( longitudinally ) rugulose before anterior transverse impres- sion and behind posterior transverse impres- sion. Elytra: width elytra prothorax 1.33 and 1.35; details c. as in asperum. Measure- ments: length 13.2-14.0 mm; width 4.5- 4.9 mm.
Types. Holotype S (Bishop Mus.) from Mt. Giluwe, (NW corner of) Papua, 2500- 2750 m, May 30, 1963 (Sedlacek), and 3 paratypes ( c^ in MCZ, Type No. 31849, c^ ? in Bishop Mus.) from same localitv, 2500, 2800-3280 m. May 30, June 6, 1963 (Sed- lacek ) .
Measured specimens. The 6 holotype and 9 paratype.
308
Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Notes. Althougli close to asperum, this new species is distinguished by the finer microsculpture of the pronotum, as de- scribed above. The prothorax is also slightly wider at base in the new fonn, but I am not sure the difference is significant.
Idiagonum limatulum n. sp.
Description. With characters of genus; form as in Figure 57, with eyes very abruptly prominent and prothorax wider than in asperum and with better defined and more prominent anterior angles; black, relatively shining, microsculpture as de- scribed below. Head 0.75 and 0.78 width prothorax; front with very fine isodiametric microreticulation and sparse punctulation. Prothorax: width length 1.27 and 1.25; base apex 1.09 and 1.04; margins angulate before middle of length, strongly sinuate before c. right posterior angles; apex mar- gined, base irregularly so; disc rather coaj-sely transversely striguJose, coarsely longitudinally strigulose before anterior transverse impression, coarsely irregularly wrinkled behind posterior transverse im- pression, with surface of disc somewhat transversely microreticulate rather than finely roughened. Elytra: width elytra/ prothorax 1.35 and 1.41. Measurements: length 13.7-14.0 mm; width 4.0-4.1 mm.
Types. Holotvpe S (Bishop Mus.) from Wau, Bulldog Rd., N-E. N. G., 2450 m, May 31, 1962 (Sedlaceks), and 1 9 paratype (MCZ, Type No. 31850), c. same locality ("19-21 km S. of Wau, Bulldog Rd."), 2200- 2500 m, May 28, 1962 (Sedlaceks).
Notes. This presumably localized species differs from asperum, etc., in being more shining and in having more prominent ante- rior prothoracic angles.
Genus MONTAGONUM Darlington
Darlington 1952, Part II, pp. 116, 233.
Diagnosis (revised). Large or medium- sized, somewhat pterostichinelike Agonini; inner wings atrophied, elytra often connate; wing-and-seta formula (see Darlington,
Part II, pp. 107ff) -w, ++, ( + )( + ), ( + )( + )( + ) ( position of lateral prothoracic setae exceptionally variable); mentum toothed, tooth either emarginate (anax, anassa, nepos) or entire (other species); outer elytral margins not iterrupted; pro- stemal process without setae; 5th tarsal segments usually without accessoiy setae (but these present in fugitum); 6 copula- tory organs ( Figs. 60A, 63A ) typically ago- nine, with both parameres short and irregu- larly oval, the left one larger than the right.
Description. No new description re- quired; see preceding Diagnosis and (for some significant variations) following Key to Species.
Type . species. M. toxopeanum Darlington.
Generic distribution (revised). Widely distributed on high mountains of New Guinea.
Notes. This genus may prove to be polyphyletic and should perhaps be con- sidered temporarily a genus of convenience. The variation in form of mentum tooth and in number and position of prothoracic and elytral setae and punctures is striking, and may eventually give a basis for dividing the genus. However, to divide it now would probably require the recognition of several new monotypic genera, of which the inter- relationships would not be known, and which would confuse rather than clarify the situation. Many more species of this group of New Guinean agonines probably remain to be discovered, and making of new genera may well wait until more of them are known.
The following Key and list of species constitute an informal revision of all k-nown species of the genus.
Key to the Species of Montagonvm
1. Very large, c. 20 mm; basal margin of elytra olwolete; 3rcl interval of elytra not punctate
2 - Smaller; basal margin of elytra present; 3rcl interxal of elytra with at least one puncture 3
2. Pronotal margins each with only 1 (median- lateral) seta; elytral margins slightly raised at liumeri anax
The carabid beetles of New Guinea • Darlington
309
- Pronotal margins with additional setae; ely- tral margins not raised at humeri ana.ssa
3. Size medium, 13-17 mm; apex of prothorax margined, base not distinctly so 4
- Size smaller -._ 6
4. Third elytral interval with 1 seta-bearing puncture (near apex) __ nepos
- Third elytral interval 3-punctate 5
5. Sides of protliorax not sinuate sororcula
- Sides of prothorax strongly sinuate before base pancUim
6. Prothorax with both base and apex plainly margined filiolum
- Prothorax with apex plainly but base weakly or not margined 7
7. Sides of prothorax broadly (not strongly) reflexed fugitum
- Sides of prothorax scarcely reflexed
__ toxopeanum
Montagonum anax n. sp.
Description. With characters of genus; form as in Figure 58, very large, heavily built; black, appendages dark; moderately shining, reticulate microsculpture fine, iso- diametric on front, in part slightly trans- verse on pronotum, slightly irregular but nearly isodiametric on elytra. Head 0.84 width prothorax; eyes small, c. ^2 long as genae behind them and not breaking out- line of sides of head (diameter of head is measured across genae rather than across eyes in this case); mandibles very long, almost straight for much of length, moder- ately arcuate apically; mentum with deeply emarginatc tooth; palpi long, slender. Pro- thorax: width/length 1.27; base apex 0.79; lateral margins moderate, somewhat broader posteriorly, well reflexed, each with One seta-bearing puncture on inner face (not edge) slightly before middle, without basal punctures; base and apex not margined; disc with fine middle line, usual transverse impressions (the posterior deeper), and poorly defined baso-lateral impressions, im- punctatc. Elytra, connate; width elytra/ prothorax 1.27; margin obsolete at base except near humeri, raised and obtusely angulate at humeri, not interrupted before apex; subapical sinuations slight, apices rather narrowly irregularly rounded; posi- tions of striae irregularly impressed with
very fine incomplete irregular impressed lines at bottom; intervals slightly convex, 3rd without dorsal punctures, outer inter\'als not impressed apically. Lower surface vir- tually impunctate; abdomen not pubescent. Legs: 4th hind-tarsal segment broadly emarginatc, not lobed. Secondary sexual characters: 6 front tarsi with 3 segments (irregularly) 2-seriately squamulose, middle tarsi not squamulose; 6 with 1 seta-bearing puncture each side apex last ventral seg- ment; $ copulatory organs as described under genus (but not fully hardened and too distorted to draw); 9 unknown. Mea- surements: length 20.5 mm; width 7.3 mm.
Type. Holotype i (Bishop Mus. ) from Murray P(ass), Papua, 2800 m, Nov. 6, 1965 (Sedlaceks); the type is unique.
Notes. See preceding Key to Species and notes under following species for com- parison. This is one of the finest of all the agonines of the world.
Monfagonum anassa n. sp.
Description. With characters of genus; form (Fig. 59) virtually as in preceding species (anax) and characters the same except as follows. Head 0.85 width pro- thorax. Prothorax: width length 1.30; base/ apex 0.84; lateral margins each with seta slightly before middle of length (as in anax), an additional seta midway between median-lateral one and apex, and a seta on margin just before base (subbasal seta present on right side, but left side broken at this point). Elytra: width elytra/ prothorax 1.29; margin not raised at hinneri; otherwise c. as in anax. Secondary .sexual characters: 9 with 2 setae each side last ventral segment; 6 unknown. Measure- ments: length 20.0 mm; width 7.1 mm.
Type. Holotype $ (MCZ, Type No. 31852), from Mt. Yule, Central Dist., Papua, 10,700 ft. (c. 3260 m), Sept. 1963 (D. Hutton & M. Stevens), "alpine grass zone"; the type is unique.
Notes. This species may be a geographic representative of the preceding (anax),
310 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
but the difference in prothoracic setae is striking and unexpected.
Montagonum nepos n. sp.
Description. With characters of genus; form as in Figure 60, c. as in anax but with head relatively narrower and prothorax less narrowed behind; black, appendages dark; reticulate microsculpture fine, c. iso- diametric on head, slightly transverse on disc of pronotum and elytra. Head 0.75 and 0.71 width prothorax; form c. as in anax but eyes slightly larger, c. long as genae behind them; mentum tooth deeply emar- ginate. Frothorax: width/length 1.12 and 1.18; base/apex 0.90 and 0.96; lateral mar- gins narrow (somewhat variable), moder- ately explanate, each with seta-bearing puncture at or slightly before middle of length but none at base; apex finely mar- gined, base not margined; disc irregularly rather weakly impressed, with baso-lateral impressions shallow, irregular, not distinctly punctate. Elytra: width elytra/prothorax 1.22 and 1.26; base margined, margin ob- tusely angulate or narrowly rounded at humeri; subapical sinuations of margins subobsolete, apices narrowly rounded into suture; striae slightly impressed, impunctate; intervals slightly convex, 3rd with 1 seta- bearing puncture far back (on declivity), often on outer edge of interval (puncture present on both sides of all specimens but position slightly variable), outer intervals not modified apically. Lower surface vir- tually impunctate (abdominal segments in part longitudinally sHghtly rugulose); abdo- men not pubescent. Legs as described for aruix. Secondary sexual characters: i front as well as middle tarsi apparently without squamae; 6 probably usually with 1, ? 2 setae each side last ventral segment (in fact the single 6 has 1 puncture on the left but 2 on the right side); c5 copulatory organs as in Figure 60A. Measurements: length c. 17.5 mm; width 5.9-6.1 mm.
Types. Holotype <? (Bishop Mus. ) from Mt. Kaindi (near Wau, Morobe Dist. ), N-E. N. G., 2350 m, Jan. 12, 1962 (Sed-
laceks); 1 9 paratype (MCZ, Type No. 31853), Bulldog Rd., 19-29 km S of Wau, 2200-2500 m. May 31, 1962 (Sedlacek).
Additional material. One $ (Bishop Mus.), Mt. Kaindi, 2,350 m, July 12, 1963 ( Sedlacek ) , m. v. light trap.
Measured specimens. The c5 holotype and ? paratype.
Notes. This well-characterized species is adequately distinguished from others in the preceding Key to Species. The spec- imen listed under Additional material, though from the same locality as the holo- type, has wider prothoracic margins and differs slightly in other ways, and may prove to be a separate species.
Montagonum sororcula n. sp.
Description. With characters of genus; form as in Figure 61; black, prothoracic margins slightly reddish-translucent, ap- pendages dark reddish; reticulate micro- sculpture fine, isodiametric on front, some- what transverse on disc of pronotum, more transverse on elytra. Head 0.65 width pro- thorax; eyes c. long as and more prominent than genae; mandibles rather long, scarcely arcuate except at apex; mentum tooth tri- angular, entire. Prothorax: width /length 1.12; base/apex 1.13; lateral margins mod- erately wide, scarcely wider posteriorly, moderately reflexed, each with seta-bearing punctures before middle and well before base (position shown in Fig. 61); apex conspicuously margined, base vaguely so; disc with median impressed line distinct, transverse impressions subobsolete, baso- lateral impressions weak, poorly defined, vaguely subpunctate. Elytra: width elytra/ prothorax 1.35; base margined, margin ob- tuse at humeri, subapical sinuations weak, apices narrowly rounded; striae impressed, not distinctly punctulate; intervals slightly convex, 3rd 3-punctate, outer intervals not modified apically. Lower .surface impunc- tate; abdomen not pubescent. Legs: 4th hind-tarsal segments emarginate, not lobed. Secondary sexual characters: S unknown; 9 with 2 or 3 (unsymmetric) setae each
The carabid beetles of New Guinea • Darlington 311
side last ventral segment. Measurements: length 13.0 mm; width 4.8 mm.
Type. Holotype ? (Bishop Mus.) from Murray P(ass), Papua, 2400-2S00 m, Nov. 6, 1965 (Sedlaccks); the type is unique.
Notes. See preceding Key to Species for place of this among other Montagomim.
Montagonum pandum n. sp.
Description. With characters of genus; form as in Figure 62; black, margins of prothorax slightly reddish translucent, ap- pendages dark; reticulate microsculpture fine, isodiametric on front, slightly trans- verse on pronotum, more transverse on ely- tra. Ileod 0.67 and 0.68 width prothorax; eyes small, slightly shorter than genae be- hind them but more abruptly prominent than usual; mandibles rather long, weakly arcuate; mentum tooth entire, triangular. Prothorax: width/length 1.08 and 1.15; base/apex 1.07 and 1.05; sides abruptly sinuate c. % of length before base, then subparallel to narrowly rounded basal angles; lateral margins moderate, moder- ately reflexed, each with seta-bearing punc- ture between Vi and % from base (2 punctures close together at this point on one side of one specimen ) but without other lateral setae; apex conspicuously margined, base inconspicuously or weakly so; disc with fine middle line, weak transverse im- pressions, shallow poorly defined faintly punctulate baso-lateral impressions. Elytra: width elytra/prothorax 1.33 and 1.27; base margined, margin c. rectangular (slightly obtuse) at humeri, where outer margin slightly raised; subapical sinuations broad, slight; apices narrowly rounded; striae en- tire, well impressed, not or faintly punc- tulate; intervals slightly convex, 3rd 3- punctate ( position of punctures as usual in Agonini ) ; outer intervals not modified api- cally. Lower surface virtually impunctate; abdomen not pubescent. Legs c. as in anax. Secondary sexual characters: S front (not middle) tarsi slightly dilated, 3 segments biseriately squamulose; S with 1 seta each side last ventral segment; 9 unknown.
Measurements: length 15.5-16.8 mm; width 5.5-6.1 mm.
Types. Holotype S (Bishop Mus.) from Mt. Chapman, N-E. N. G., 2000 m, May 5, 1966 (collector not given); and 1 $ para- type (MCZ, Type No. 31855) from Bulldog Rd., 2200-2500 m. May 31, 1962, N-E. N. G., 19-29 km S of Wau (Sedlacek).
Notes. This species is unique in the genus in the abrupt sinuation of the sides of the prothorax and especially in the position of the lateral prothoracic setae. Other differ- ential characters are given in the Key to Species.
Montagonum fugitum n. sp.
Description. With characters of genus; form as in Figure 63, slender CaJatluis-like, with base of prothorax broadly emarginate at middle; dull black, margins of prothorax slightly reddish translucent, legs reddish, antennae brown; reticulate microsculpture fine, isodiametric ( in part ) on front, irregu- lar but not strongly transverse on pronotum and elytra. Head 0.58 width prothorax; eyes rather small but more abiaiptly prom- inent than usual; mandibles rather short, blunt, scarcely arcuate; mentum tooth narrow-triangular, entire. Frothorax: width/ length 1.12 (length measured at mid- dle as usual); base/apex 1.42; lateral mar- gins narrow anteriorly, running into broad slightly depressed areas i:)Osteriorly, each with seta-bearing puncture slightly before base just inside marginal bead, without median-lateral setae; apex conspicuously margined, base not or at most vaguely so (base partly broken); disc weakly convex, middle line distinct, transverse impressions vague, posterior-lateral impressions shallow, not well defined, impunctate. Ehjtra: width elytra/prothorax 1.29, base margined, mar- gin c. rectangular (slightly arcuate) at hu- meri, with humeral margins slightly raised in front of angle; subapical sinuations vir- tually obsolete, apices narrowly rounded; striae moderately impressed, entire, not punctate; intervals slightly convex, 3rd 3- punctate (punctures placed as usual in
312 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Figures 64—84 (see text, section [15j]; 64, Montagonum filiolum n. sp., i holotype; 65, M. toxopeanum Darl., S para- type; 66, Nebriagonum foedum n. sp., 9 para.; 67, Loevogonum pertenue n. sp., 9 para.; 68, L frustum n. sp., $ holo.; 69, L gi/uwe n. sp., $ holo.; 70, Fortogonum bufo Darl., 9, Swart Vy.; 71, F. antecessor n. sp., 9 holo.; 72, F. ooc^inum n. sp., 9 holo.; 73, F. okapa n. sp., S pore., Mt. Elandora; 74, F. tormiceps n. sp., 6 holo.; 75, F. horna- broofei n. sp., $ para., Wau; 76, F. d/stortum n. sp., 9 holo.; 77, Oodes terrestris n. sp., $ para., Dobodura; 78, Cfiydoeus hinnus n. sp., 9 holo.; 79, Physodero bocchusi n. sp., 6 holo.; 80, Demetrido parena n. sp., 9 holo.; 81, Hexogonio gress/fti n. sp., i holo.; 82, Colasidia papua n. sp., 9 holo.; 83, C. modong n. sp., 9 holo.; 84, Helluopa- puo c/ieesmani n. sp., c5 holo.
The carabid beetles of New Guinea • Darlington 313
Agonini), outer intervals not modified api- cally. Lower siwface virtually impunctate (except sides of mesosternum vaguely sub- punctate); abdomen not pubescent. Legs slender; hind tarsi with all segments sulcate each side above; 4th hind-tarsal segment long, shallowly emarginate, not lobed; 5th segment with c. 3 conspicuous accessory setae each side. Secondary sexual charac- ters: S with front (not middle) tarsi slightly dilated, biseriately squamulose; S with 1 seta each side last ventral segment; <5 cop- ulatory organs as in Figure 63A; 9 unknown. Measurements: length c. 11 mm; width 4.0 mm.
Tijpc. Holotype $ (MCZ, Type No. 31856) from Mt. Amagwiwa, nr. Wau, Morobe Dist., N-E. N. G., c. 11,000 ft. (c. 3350 m) Sept. 24, 1963 (Guy Rosenberg), in alpine grass zone; the type is unique.
Notes. Although I am tentatively placing this interesting species in Montagonum, in order to avoid premature multiplication of genera, it has nothing to do directly with any other species now placed in that genus. I think it may be an independent derivative of Altagonurn, although it is ruled out of that "genus of convenience" by atrophy of the wings and resultant shortening of mete- pistenia (which are slightly longer than wide) and superficial fusion of elytra. The presence of conspicuous accessory setae on the 5th tarsal segments differentiates fugi- tiim from all Altagonum, and indeed from all other Montagonum.
Montagonum filiolum n. sp.
Description. With characters of genus; form as in Figure 64; black, appendages dark brownish; reticulate microsculpture light and c. isodiametric on front, irregu- larly transverse on pronotum and elytra. Head 0.71 and 0.69 width prothorax; eyes moderate, slightly longer than and more prominent than genae; mandibles moder- ate (somewhat shorter and more arcuate than in most other species of genus); men- turn tooth triangular, blunted but not dis- tinctly emarginate. Prothorax: width/
length 1.20 and 1.27; base/apex 1.21 and 1.19; margins narrow, slightly wider basally, each with seta-bearing punctures slightly before middle and virtually at basal angle; base and apex conspicuously margined; disc with usual moderate impressions, baso- lateral impressions moderate, poorly de- fined, vaguely subpunctate. Ehjtra: width elytra/prothorax 1.32 and 1.28; base mar- gined, margin obtuse at humeri, not spe- cially elevated at humeri; subapical sinua- tions subobsolete, apices rather narrowly rounded to suture: striae lightly impressed, not distinctly punctate; intervals slightly convex, 3rd 3-punctate ( punctures placed as usual in Agonini); outer intervals not spe- cially modified apically. Lower surface virtually impunctate; abdomen not pubes- cent ( but segments 2-4 swollen at middle ) . Legs: 4th hind-tarsal segment short, deeply emarginate, scarcely lobed; 5th segments without obvious accessory setae. Secondary sexual characters: i front (not middle) tarsi slightly dilated, with 3 segments bi- seriately squamulose; S with 1, 9 2 setae each side last ventral segment. Mea.mre- ments: length c. 10-11 mm; width 3.8-4.0 mm.
Types. Holotype $ (MCZ, Type No. 31857), and 3 paratypes {$9 Dept. Agr. Port Moresby, 9 MCZ) all from Mt. Albert- Edward, Central Dist., Papua, 13,000 ft. (3660 m), Aug. 30, 1963 ( F. H. A. Kleckham & I. G. Pendergast), under stones in alpine grass zone.
Measured specimens. The $ holotype and 1 9 paratype.
Notes. Another very distinct species, differentiated in the preceding Key to Spe- cies.
Montagonum toxopeanum Darlington
Darlington 1952, Part II, p. 234.
Notes. Still kno\\-n only from the types (holotype S in Leiden Mus.) from Letter- box Camp, Snow Mts., West N. G., 3600 m.
Genus NEBRIAGONUM Darlington Darlington 1952, Part II, pp. 116, 235.
314 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Generic distribution (revised). Slopes of Mt. Wilhelm, Bismarck Range, N-E. N. G. (5 species), with 1 Mt. Wilhelm species ex- tending to Mt. Otto; 1 Mt. Wilhelm stock- represented by 1 or more endemic forms SE to the vicinity of Okapa; and 1 doubt- fully related species on the Snow Mts., West N. G.; all at rather high altitudes, 1 species reaching c. 4300 m on Mt. Wilhelm ( Darlington, Sedlacek ) .
Notes. The variation of characters and ecology of the species of this genus is discussed in Part II. The occurrence of a slightly differentiated endemic form, de- scribed below, in a cave on the Porol Range extends the ecologic as well as the geographic distribution of the genus.
Nebriagonum cephalum Darlington Darlington 1952, Part II, pp. 238, 239.
Additional material. N-E. N. G.: 1, Mt. Wilhelm, 4250 m, June 3, 1963 (Sedlacek).
Notes. Mt. Wilhelm is, of course, the type locality of the species, which ranges up to c. 4300 m on the mountain.
Nebriagonum foedum n. sp.
Description. With characters of genus; form as in Figure 66, virtually as in N. cephalum Darlington; dull black, append- ages dark; reticulate microsculpture fine, isodiametric on front, transverse on pro- notum, more transverse (scarcely distinct) on elytra where visible. Head 0.92 and 0.90 width prothorax; eyes small but rather abruptly prominent; 2 setae over each eye. Prothorax: width/length 1.00 and 0.99; base/apex 1.01 and 1.00; lateral margins narrow, somewhat wider posteriorly, ap- parently without lateral setae; base and apex margined; disc nonnally impressed, baso-lateral impressions slight, impunctate. Elytra oval; width elytra/prothorax 1.55 and 1.58; base margined, margins obtuse at humeri; apical sinuations slight or obsolete; apices independently rounded ( ci ) or angu- late or subdenticulate c. in line of 3rd inter- vals ( 9 ) ; striae deep, impunctate or faintly
punctulate; intervals moderately convex, ap- parently without dorsal punctures (see Notes below). Lower swface, inner wings, legs, and secondary sexual characters normal; 4th hind-tarsal segment lobed, outer lobe longer than inner; 5th segments without obvious accessory setae. Measurements: length c. 9.5 mm; width c. 3.3 mm.
Types. Holotype S (MCZ, Type No. 31858) and 2 9? paratypes all from Maimbobo Cave, North Dumam, Porol Rge., N-E. N. G., Sina Sina, Nov. 15, 1965 (F. Parker), in fruit bat droppings.
Additional material. N-E. N. G.: 1 S, Okapa, Purosa, 6500 ft. (c. 1980 m), 1965 (Hornabrook), under stone in stream bed.
Measured specimens. The S holotype and 1 9 paratype.
Notes. The 3 types are so coated with fruit bat guano, which I have been unable to remove satisfactorily, that I cannot be sure that the lateral prothoraeic setae and dorsal elytral punctures are lacking. The heads of all specimens are completely clean, so there can be no question about the supraocular setae being present on both sides in all specimens, but the pronota and elytra are completely coated except where I have scraped small areas bare with a pin point. Characters visible in spite of the coating, however, show that this new spe- cies is in general similar to Nebriagonum cephalum Darlington (above) but differ- ent in having deeper elytral striae and dif- ferently formed elytral apices, and probably also in lacking lateral prothoraeic and dorsal elytral seta-bearing punctures. The pro- thorax is slightly longer in foedum than in cephalum, but not as long as in perceph- alum Darlington (Part II, p. 240), from which the new species differs also in pos- sessing both pairs of supraocular setae (an- terior pair lacking in percephalum).
Although the i holotype has rounded and the 2 9 9 paratypes denticulate elytral apices, I think the difference is probably ' individual rather than sexual. The i from Okapa has ehtral apices strongly denticu- late on a slightly different pattern from the Maimbobo Cave 9 9 and may prove to
The carabid beetles of New Guinea • Darlington 315
represent a separate population, but ex- tent of variation should be established be- fore it is formally described.
Nebriagonum fransitum Darlington Darlington 1952, Part II, pp. 238, 241.
Additional material. N-E. N. G. : 20, Mt. Wilhelni (inch L. Aunde and L. Strunki), 2800-2900, 3400-3500 m, dates in June, July, 19&3 (Sedlacek); 10, Mt. Otto including summit (on Bismarck Rge. SE of Mt. Wil- helm), Nov. 1965 (collector not given, Dept. Agr. Port Moresby ) .
Notes. I found this species (the types) common on Mt. Wilhelm both in the highest mountain forest and above the forest on open grassy slopes.
3enus LAEVAGONUM Darlington Darlington 1952, Part II, pp. 116, 243.
Diagnosis (revised). Small (4.8-8.5 mm) Europhilus- or Calathus- or cistellidlike; body smooth in oudine (no sharp angles), smoothly convex; pronotum with baso- lateral fovea absent or weak; wing-and-seta formula (see Part II, p. 107) -w, ++, (-)- --(-); 4th hind-tarsal segments variable, simply emarginate or long-lobed in different species.
Generic distribution {revised). Known anly from the Bismarck Range, N-E. N. G., and Mt. Giluwe just across the border in NW Papua.
Notes. To this genus, previously known from 4 species, I am now adding 3 more. They differ mainly in characters of form and proportions best shown by drawings (Figs. 67-69) rather than by a new key. Besides those actually recorded, I have seen a single specimen from Mt. Giluwe, Papua (Bishop Mus.), w^hich might be re- ferred to L. citiim Darlington (Part II, p. 245) but is so much broader that it may prove to represent a separate species.
-aevagonum cisfelum Darlington Darlington 1952, Part II, pp. 245, 246.
Additional Jimterial. N-E. N. G. : 3, Daulo Pass (Asaro-Chimbu Div., Bismarck Rge.), 2400 m, June 16, 1955 (Gressitt).
Laevagonum pertenue n. sp.
Description. With characters of genus; form as in Figure 67, very slender; brown, head and pronotum blackish, appendages brown; reticulate microsculpture iso- diametric on front, transverse on pronotum and elytra. Head 0.78 and 0.77 width pro- thorax. ProtJwrax: width/length 0.91 and 0.96; base/apex 1.48 and 1.42; lateral mar- gins fine, slightly wider toward base, with- out setae; base inconspicuously margined, apex with margin interrupted at middle; disc with middle line and sHght transverse impressions and broad poorly defined baso- lateral areas scarcely impressed but slightly irregular or subpunctate. Elytra: width elytra /prothorax 1.54 and 1.55; base mar- gined, margin c. right (slightly acute) at humeri; subapical sinuations slight, apices independently rather narrowly rounded; striae light, slightly irregular but not dis- tinctly punctulate; intervals almost flat, 3rd with 1 seta-bearing puncture posteriorly (near top of declivity); outer intervals partly impressed and irregular toward apex. Legs: 4th hind-tarsal segment with 2 long nearly equal lobes. Measurements: length c. 7.5-8.5 mm; width 2.5-2.8 mm.
Types. Holotype S (Bishop Mus.) from Daulo Pass (Asaro-Chimbu Div., Bismarck Rge.), N-E. N. G., 2400 m, June 16, 1955 (Gressitt), and 1 $ paratype (MGZ, Tvpe No. 31859) from Mt. Wilhelm, Bismarck Rge., 3000 m, July 4, 1955 (Gressitt).
Notes. The presence of a (posterior) seta-bearing puncture on the 3rd interval, the long-lobed 4th hind-tarsal segments, and the very slender form distinguish this species from citum Darlington and all other previously known species of Laevagonum.
Laevagonum frustum n. sp.
Description. \\'ith characters of genus; form as in Figure 68, very small, with short,
316 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
rounded elytra; browaiish black, appendages yellowish browai; reticulate microsculpture light, isodiametric on front, somewhat trans- verse on pronotum, more transverse on ely- tra. Head 0.73 and 0.75 width prothorax. Pwthoiax: width/length 1.20 and 1.13; base/apex 1.19 and 1.20; lateral margins narrow, scarcely wider posteriorly, without setae; base and apex finely margined; disc with middle line, slight transverse impres- sions, baso-latcral impressions distinct but irregular, not or vaguely punctate. Elytra: width elytra/prothorax 1.58 and 1.61; base margined, margin very obtuse (virtually rounded) at humeri; subapical sinuations broad, apices independently rounded; striae slightly impressed, irregular but not dis- tinctly punctulate; intervals flat or very weakly convex, 3rd without punctures, outer intervals not impressed toward apex. Legs: 4th hind-tarsal segment with rather long lobes, outer lobe longer than inner. Sec- ondanj sexual characters of i unkno\vn, of 9 normal. Measurements: length c. 4.8- 5.0 mm; width 2.1-2.2 mm.
Types. Holotype 9 (Bishop Mus.) from Daulo Pass ( Asaro-Chimbu Div., Bismarck Rge.), N-E. N. G., 2400 m, June 16, 1955 (Gressitt); 1 9 paratype (Bishop Mus.) with same data except June 15; 1 9 para- type (MCZ, Type No. 31860), same data except 2500 m, June 12.
Measured specimens. The 9 holotype and first 9 paratype.
Notes. The small size and very stout form distinguish this from cistelum Darling- ton and from all other known species of the genus, and the long-lobed 4th hind- tarsal segments differentiate it also from all other species except peHenue ( above ) .
laevagonum giluwe n. sp.
Description. With characters of genus; form as in Figure 69; irregular dark brown, suture and appendages in part more rufous; reticulate microsculpture slightly irregu- lar, c. isodiametric on front and elytra, slightly transverse on pronotimi. Head 0.69 width prothorax. Prothorax rounded-
quadrate; width /length 1.17; base/apex c. 1.18 (angles too rounded for exact measure- ment); lateral margins fine, without setae; apex margined, base not margined at mid- dle; disc rather weakly c. evenly convex, with fine middle line, transverse impres- sions vague, baso-lateral impressions absent, surface slightly irregular baso-laterally but scarcely punctate. Elytra: width elytra/ prothorax 1.36; base margined, margins ob- tusely angulate at humeri; subapical sinua- tions absent; apices independently rounded; striae slightly impressed, faintly (hardly distinctly) punctulate; intervals c. flat, 3rd without dorsal pimctiues. Secondary sexual characters of c^ unknown, of 9 normal. Measurements: length 7.0 mm; width 2.6
mm.
I
Ti/pe. Holotype 9 (Bishop Mus.) from Mt. Giluwe, NW Papua, 3750 m. May 29, 1961 (J. L. & M. Gressitt), in forest patch: the type is unique.
Notes. This new species is intermediate in form between citum Darlington and cis-\ telu77i Darlington, being broader than citum (and without the denticulate elytra of that species) but less cistelid-like and with relatively wider head than cistelum.
Although the occurrence of this insect on Mt. Giluwe extends the range of Laeva- g,onu7n to Papua, the actual distance is not great.
Genus FORTAGONUM Darlington Darlington 1952, Part II, pp. 116, 247. I
Diagnosis (revised). Moderate-sized (8.8- p 13.5 mm), heavily built, broadly subparal- lel or fusiform Agonini; wing-and-seta for- mula (see Part II, p. 107) -w, (-)( + ), -(-)' (-)(-)(-)• (The only species of the genus in which anterior supraocular setae are present is distortum, under which see, below. )
Type species. Eortagonum fortellum Dar- lington.
Generic distribution {revised). High mountains of New Guinea.
Notes. Fortagonum may have been de- rived at high altitudes on New Guinea from
The carabid beetles of New Guinea • Darlington
317
an ancestor or ancestors (the genus may be polyphyletic ) something Hke ^"Altago- ninyf big,enum (p. 299), which is excluded from Fortagonum only by its full wings. Fortagonum is probably not directly re- lated to Montagonum (p. 308), which has probably had a separate origin ( or origins ) on the same mountain ranges. Both these two genera are heavily built and flightless, but they differ not only in presence (in Montagonum) and absence (in Fortago- num) of anterior supraocular setae but also in form, most Montagonum being Pterostichus-like, most Fortagonum more agonine in appearance. In fact, the phylog- eny of these and other high-mountain Ago- nini in New Guinea is likely to have been much more complex than second-stage tax- onomy can show.
Five species were originally included in Fortagonum, and 6 more are added now. The follo\\'ing Key and list of species con- stitute an informal revision of the whole genus. However, many more species prob- ably remain to be discovered on different mountain summits in New Guinea. Judging from their diverse form and the diverse modifications of the eyes and mandibles, some of the species probably have special habits or special foods, so that their biology is likely to be interesting.
Key (Revised) to the Species of Fortagonum
1. Both pairs supraocular setae lacking; form very l3roadly rounded-fusiform (Fig. 70) bufo
- At least posterior supraocular setae presenT:; form variable but never so broadly rounded 2
2. Margins of pronotum relatively narrow ante- riorly, although very broad posteriorly 3
- Protlioracic margins very broad anteriorly as well as posteriorly 9
3. Mandibles nomial, moderately long and ar- cuate 4
- Mandibles slender, virtually straight 7
4. Pronotum with posterior-lateral setae 5
- Pronotum without posterior-lateral setae .— 6
5. Form normal; pronotal margins moderately reflexed antecessor
- Fomi depressed; pronotal margins very wide but scarcely reflexed oodintim
6. Prothoracic width/length c. 1.40 or more;
elytra weakly iridescent fortellum
- Prothoracic width/length c. 1.25 or less; pronotum and especially elytra strongly iri- descent okapa
7. Form normal, subparallel; relatively wider (prothoracic width/length 1.28 and 1.35) forceps
- Form subfusiform, narrowed at least anteri- orly 8
8. More fusiform; prothoracic width/length 1.27 __ cychriceps
- Subfusifonn, more narrowed anteriorly than posteriorly; prothoracic width/length 1.11 formiceps
9. Eyes c. normal, front not swollen
hornahrooki
- Eyes abruptly prominent and/or front con- spicuously swollen 10
10. Eyes very abruptly prominent; front only slightly modified; only posterior supraocular setae present limiun
- Eyes less abruptly prominent; front con- spicuously swollen posteriorly; 2 pairs supra- ocular setae present (placed as indicated in Fig. 76) disiortuin
Fortagonum bufo Darlington
Darlington 1952, Part II, pp. 248, 252.
Additional material. West N. G.: 1 9, Swart Vy., W ridge, 1800-2000 m, Nov. 19, 1958(Gressitt).
Notes. This species was described from Mist Camp, Snow Mts., West N. G., at 1800 m. The Swart Valley specimen agrees well with the ? paratype in the MGZ.
Fortagonum antecessor n. sp.
Description. With characters of genus; form as in Figure 71; black, prothoracic margins and appendages slightly reddish; reticulate microsculpture fine, c. iso- diametric on front, irregular and slightly transverse on disc of pronotum and elytra. Head 0.60 width prothorax; eyes moderate; posterior supraocular setae present; mandi- bles rather short and arcuate (in genus); front almost evenlv convex. Prothorax: width/length 1.35; base/apex 1.51; lateral margins narrow anteriorly, much broader posteriorly, moderately explanate, each with seta-bearing puncture near posterior angle; apex margined and with anterior angles
318 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
only moderately advanced (in genus); base slightly broadly emarginate, not margined; disc with middle hne and weak transverse impressions; baso-lateral impressions formed mainly by troughs between disc and mar- gins, in part xaguely subpunctate. Elytra: width elytra prothorax 1.27; base margined, margin c. right (slightly obtuse) at humeri; subapical sinuations almost obsolete, apices independently rounded; striae impressed, not punctate; intervals slightly convex, 3rd normally 3-punctate, outer intervals not much modified toward apex, no 10th inter- vals. Lower surface in part slightly wrinkled or vaguely subpunctate. Leg^s: 4th hind- tarsal segment apparently emarginate but not lobed (in poor condition); 5th hind- tarsal segment with distinct but not large accessory setae. Secondary sexual char- acters: 9 with 2 seta-bearing punctures each side last ventral segment; S unknown. Measurements: length c. 9.0 mm; width c. 3.7 mm ( specimen not in condition to mea- sure accurately ) .
Type. Holotype 2 (MCZ, Type No. 31862) from Mt. Albert-Edward, Central Dist., Papua, 12,000 ft. (3660 m), Aug. 30, 1963 (F. H. A. Kleckham & I. G. Pendergast), in alpine grass zone, under stone; the type is unique.
Notes. This species has approximately the form of F. fortelhnn and may resemble the ancestral stock of that and of related species. It ( the present new species ) differs from fortelhnn in having posterior-lateral prothoracic setae present, anterior angles of prothorax less advanced, and humeral angles less acute, all these characters being perhaps relatively primitive or un- specialized.
Forfagonum oodinum n. sp.
Description. \Vith characters of genus; form as in Figure 72, with head very small and prothorax and anterior portion of elytra more depressed than usual; black, legs dark, tarsi and antennae more reddish; reticulate microsculpture isodiametric on front, trans- verse on pronotum, scarcely distinct on ely-
tra which faintly iridescent in some lights.; Head 0.46 and 0.46 width prothorax; eyeji moderate; posterior supraocular setae pres- ent; mandibles moderately long and luoder- ately arcuate (in genus). Prothorax large wide, and flatter than usual; width/length 1.50 and 1.51; base apex 1.87 and 1.73 (sic). lateral margins very wide especially poste- riorly but scarcely explanate, separatee from weakly convex disc by distinct bul shallow poorly defined channels ending posteriorly in poorly defined baso-latera impressions, each with seta-bearing punc ture at base somewhat in from edge of mar- gin; apex margined, base not margined a middle; disc with slightly impressed mid die line and weak transverse impressions with basal and lateral areas vaguely sub punctate. Elytra: width elytra/prothora^ 1.09 and — (right elytron missing); base margined, margin c. right (slightly arcu ate) at humeri; subapical sinuations sligh or obsolete; striae well impressed, not punc tulate; intervals slightly convex, 3rd finelj 3-punctate (punctures placed as usual ii Agonini), outer intervals (7, 8) deepl) impressed apically; narrow apparent lOtl interval present behind middle. Loner sur face not much punctate. Le<s,s: 4th hind tarsal segments long, shallowly emarginate not lobed; 5th segments without distinc accessory setae. Secondary sexual charac ters: S unknown; 9 with 2 setae each sidf last ventral segment. Measurements: lengtl c. 9.5-10.0 mm; width c. 4.0-4.2 mm.
Types. Holotype 9 (Bishop Mus.) fron Bulldog Rd., N-E. N. G., 2500 m. May 31 1962 (no collector given); and 1 (broken^ 9 paratype (MCZ, Type No. 31863) froir Bulldog Rd., 29 km SW of Wau, 2500 m Feb. 15, 1962.
Notes. This species is superficially some what similar to a terrestrial oodine. It i; probably related to Fortagonum fotiellun Darlington but differs in being broader anc more depressed, with smaller head, anc' with a distinct seta-bearing puncture or each side of base of pronotum. However this puncture may sometimes be present ir_ fortellum ( see below ) .
The c ARAB id beetles of New Guinea • Darlington
319
Fortagonum forfellum Darlington Darlington 1952, Part II, pp. 248, 251.
Additional material. Papua: 1 i , Murray P(ass), 2400-2800 m, Nov. 6, 1965 (Sed- laceks ) .
Notes. This specimen agrees in general and in proportions \\'ith the type series from the Bismarck Range, N-E. N. G., the propor- tions of the Murray Pass c^ being head 0.53 width prothorax; prothorax, width/length 1.42 and base/apex 1.69; and elytra, width of elytra prothorax 1.21. However, the Murray Pass individual differs from the types of fortellum in having a seta-bearing puncture on each side of the pronotum on the face ( not edge ) of the margin near the posterior angle — the actual setae are missing ( broken off ) but punctures mark their posi- tions. More material is needed to show whether these punctures characterize a distinguishable population, or \\'hether they vary individually. I have re-examined the 28 specimens of the type series of fortellum still at the MCZ and find no lateral f)ro- thoracic setae or punctures in any of them. However, the dorsal elytral punctures do vary remarkably in fortellum, as orginally noted.
Forfagonum okapa n. sp.
Description. With characters of genus; form as in Figure 73, nearly as in fortellum but with prothorax narrower, more nar- rowed anteriorly, and with anterior angles more produced; black, strongly iridescent especially on elytra, appendages dark, tarsi and outer segments of antennae paler brown; reticulate microsculpture faint or unresolved ( at 80x ) on most of upper surface. Head 0.52 and 0.50 width pro- thorax; eyes moderate; posterior supraocular setae present; mandibles moderate and moderately arcuate (in genus). Frothorax: width /length 1.18 and 1.25; base/apex 1.62 and 1.62; lateral margins wide, very wide posteriorly, moderately explanate, without seta-bearing punctures; apex margined, base not; disc with middle line impressed, trans-
verse impressions subobsoletc, baso-lateral impressions continuous with (obtuse) mar- ginal gutters, not distinctly punctate. Ely- tra: width elytra /prothorax 1.20 and 1.20; base margined (margin sometimes hidden under base of prothorax), margin acutely angulate at humeri (more acute than in fortellum); subapical sinuations obsolete, apices c. conjointly rounded except nar- rowly rounded into suture; striae impressed, not punctulate; interv^als slightly convex (more convex posteriorly), 3rd 3-punctate (punctures minute, difficult to see, but present on both sides in all specimens), outer intervals very convex but not com- pressed and not impressed toward apex, 10th intervals absent or indistinct. Lower su)-face in part slightly punctate. Legs: 4th hind-tarsal segments emarginate, scarcely lobed; 5th segments without accessory setae. Secondary sexual characters normal. Mea- surements: length c. 10.3-12.5 mm; width 4.2-4.8 mm.
Types. Holotype S (MCZ, Type No. 31864) and 2 paratypes (Hornabrook Coll.) from Okapa, Purosa, N-E. N. G., 7000 ft. (2135 m), Nov. 29, 1965 (Hornabrook); and 2 additional paratypes ( MCZ and Hornabrook Coll.) from Mt. Elandora, N-E. N. G., Oct. 15, 1965 (Hornabrook).
Measured specimens. The S holotype and 1 9 paratype from Okapa.
Notes. The present new species probably represents fortellum of the Bismarck Range but differs in form (as indicated at the beginning of the present description ) and is much more iridescent. The long type series of foiiellum showed remarkable var- iation in some characters including modifi- cation of the apices of the 7th and 8th striae, but this variation probably does not occur in okapa: the intervals in question are strongly convex in all the 5 specimens listed above, with no sign of impressions on any of them.
Forfagonum forceps Darlington Darlington 1952, Part II, pp. 248, 249.
320 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Notes. Still kno\\ii only from Moss Forest Camp, Snow Mts., West N. G., 2600-2800 m.
Fortagonum cychriceps Darlington Darlington 1952, Part II, pp. 248, 250.
Notes. The still-unique tvpe is from Mist Camp, Snow Mts, West N'. G., 1800 m.
Fortagonum formiceps n. sp.
Description. With characters of genus; form as in Figure 74, more narrowed ante- riorly than posteriorly, with exceptionally long slender mandibles; opalescent black, legs dark, antennae more reddish; reticu- late microsculpture light and c. isodiametric on front, not resolved at 80x on pronotum and elytra. Head 0.62 width prothorax; eyes small, rather abruptly prominent; pos- terior supraocular setae present, well be- hind level of eyes; mandibles long, slender, very weakly arcuate; front almost evenly convex except for shallow sublongitudinal frontal impressions. Frothorax: width/ length 1.11; base/ape\ 1.41; lateral margins narrow anteriorly, much broader posteriorly, weakly explanate, without seta-bearing punctures ( an apparent puncture near basal angle on left side is probably a nonsetose impression); base not margined, apex with margin interrupted at middle; disc with im- pressed middle line, transverse impressions virtually obsolete, baso-lateral impressions rather wide, weak, vaguely subpunctate. Elytra: width elytra/prothorax 1.31; base finely margined, margin acutely angulate at humeri; subapical sinuations obsolete; apices c. conjointly rounded; striae well im- pressed, not punctulate; intervals sHghtly convex, 3rd impunctate, outer intervals not modified apically, no 10th intervals. Lower swjace not distinctly punctate. Legs: 4th hind-tarsal segments shallowly emarginate, not lobed; 5th segments without distinct accessory setae. Secondary sexual charac- ters of 6 normal; 9 unknown. Measure- ments: length c. 11.0 mm; width 4.0 mm.
Type. Holotype 6 (Bishop Mus. ) from Swart Vy., West N. G., W ridge 1800-2000
m, Nov. 19, 1958 (Gressitt); the type is unique.
Notes. The long, slender, nearly straight mandibles relate this species to forceps and i cychriceps. Comparison with the paratype of the former in the MCZ shows that the present new species differs in having smaller eyes and narrower prothorax with narrower margins. Comparison with the description of cychriceps shows even greater differences, the new species having a rela- tively wider head (only 0.48 width pro- thorax in cychriceps) and narrower pro- thorax (width/length 1.27 in cychriceps). The unusual form of the mandibles in these 3 species suggests specialized feeding be- havior worth looking for in the field.
Fortagonum hornabrooki n. sp.
Description. With characters of genus; form as in Figure 75, with exceptionally wide, widely margined prothorax; black, elytra purplish, appendages dark, tarsi and ' antennae slightly browner; reticulate micro- sculpture fine, c. isodiametric on front, somewhat transverse on disc of pronotum, more transverse on elytra. Head 0.50 and 0.49 \\ idth prothorax; eyes moderately ab- ruptly prominent; mandibles moderately long, moderately arcuate; front slightly ir- regular but not strikingly convolved; pos- terior supraocular setae present. Prothorax: width /length 1.68 and 1.76; base/apex 1.48 and 1.47; lateral margins very wide anteri- orly and even wider posteriorly, rather weakly reflexed anteriorly, slightly more reflexed posteriorly, each with seta at ex- treme base on the denticulate angle; apex with anterior angles advanced, finely mar- gined; base subtruncate, finely margined; disc weakly convex, with usual impressions weak, baso-lateral impressions formed mainly by explanate margins, base and posterior-lateral impressions vaguely sub- punctate. Elytra: width elytra prothorax ' 1.10 and 1.06; base margined, margin rather narrowly rounded at humeri; subapical sin- uations moderate, apices each irregularly rather narrowly rounded, almost sub-
The carabid beetles of New Guinea • Darlington
321
angulate; striae impressed, not distinctly punctate; intervals slightly convex, 3rd 3- punctate, outer intervals not much modified apically, no 10th interval. Loner sui'foce not or not much punctate. Legs: 4th hind- tarsal segments moderately lobed, outer lobe longer than inner; 5th segments with- out accessory setae. Secondary sexual char- acters of c5 normal; of 9 unknown. Mea- surements: length 11.8-13.5 mm; width 4.6- 5.0 mm.
Types. Holotype S (MCZ, Type No. 31866) from Okapa, N-E. N. G., May 14, 1965 ( Hornabrook ) ; 1 S paratype, same locality, Aug. 18, 1965 (Hornabrook, in his collection); 1 S paratype (Bishop Mus.), Wau, Nami Ck., 1750 m, Aug. 6, 1963 ( Sed- laceks ) .
Measured specimens. The i holotype and the i paratype from Wau.
Notes. In width of prothorax and of pro- thoracic margins, this species is comparable only with limum and distort U7n (below), but the present species lacks striking modifi- cations of the eyes and front. It is a fine species, and I take pleasure in naming it for the collector, in recognition of his suc- cess in finding remarkable new Carabidae in New Guinea, especially in the moun- tains.
Fortagonum limum Darlington
Darlington 1952, Part II, p. 248.
Notes. The unique type, from Mt. Mis(s)im, Morobe Dist., N-E. N. G., is still the only individual known, but the follow- ing new species (distortum) is probably closely allied.
Fortagonum distorfum n. sp.
Description. With characters of genus; fonn as in Figure 76, very wide, convex; black, elytra faintly purplish, lateral margins of prothorax broadly reddish-translucent, legs dark, antennae brown; reticulate mi- crosculpture isodiametric on posterior jpart of head, transverse on pronotum, trans- verse (in part indistinct) on elytra. Head
0.50 and 0.50 width prothorax; eyes rather small, rather abruptly prominent ( but genae c. evenly long-oblique behind them), sep- arated from front by deep channels; mandi- bles moderate and moderately arcuate (in genus); front distorted, broadly and strongly swollen posteriorly, the swollen area deeply channeled at middle; hotli pairs supraocular setae present. Prothorax ver}^ broad, very broadly margined; width/length 1.74 and 1.71; base/apex 1.47 and 1.58; lateral mar- gins moderately explanate, each mth seta- bearing puncture at basal angle; apex con- spicuously margined, base finely so; disc with middle line well impressed, transverse impressions slight, baso-lateral impressions weak; disc vaguely subpunctate at base and sides. Elytra very wide and short; width elytra/prothorax 1.10 and — (1 elytron broken off, although mounted with speci- men); base margined, margin rounded at humeri; subapical sinuations broad, slight; apices narrowly independently rounded; striae rather lightly impressed, not or indis- tinctly punctulate; interv^als barely convex, 3rd apparently impunctate, outer intervals not much modified toward apex, 10th inter- vals indistinct. Lower sui^ace in part (not including proepisterna) more or less wrin- kled or subpunctate. Le^is: 4th hind-tarsal segments short, rather deeply emarginate, slightly lobed, outer lobe longer than inner; 5th segments without obvious accessory setae. Secoiidary sexual characters of S unknown, of 9 normal. Measurements: lensth c. 10.5-11.5 mm; width c. 5.0-5.5
mm.
Types. Holotvpe 9 (MCZ, Type No. 31867) from Offafina, Okapa, N-E. N. G., Nov. 12, 1964 (Hornabrook); and 1 broken 9 paratvpe (Bishop Mus.) from Kainantu, N-E. N. G., 2250 m, Jan. 8, 1965 (Sedlacek).
Notes. This remarkable species is gen- erally similar in fonn and appearance to limum but differs in structure of head: in the 9 type of limum, the eyes are more abruptly prominent but the front is not swollen, and litnum lacks the anterior supra- ocular setae which are present in disiortum. The relationships of these two species are
322 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
so obvious that I am placing distortxim in Foiiagoniim in spite of presence of anterior supraocular setae. In modification of head these two species of Porta goniim resemble the unrelated (lebiine) Dolichoctis disioiia Darlington (1968, Part III, p. 127), sug- gesting parallelism either of adaptations or of genetic processes.
Tribe PERIGONINI
Darlington 1968, Part III, p. 5.
This and the following tribes have been treated so recently by me that I have little or nothing to add in most cases.
Genus PERIGONA Castelnau
Darlington 1968, Part III, p. 6.
Notes. I have seen 39 additional speci- mens of Perigona from New Guinea. They include some new locality records (which cannot be given here) but no new species and no important new material of poorly known species.
Tribe LICININI
Genus DICROCHILE Guerin
Darlington 1968, Part III, p. 16.
Dicrochile gigas Castelnau
Castelnau 1867, Notes on Australian Coleop., p. 66. Sloane 1923, Proc. Linnean Soc. New South Wales, 48: 36.
Description. None required here; see fol- lowing Notes.
Types. Described from Rockhampton, Brisbane, and the Clarence R., all in south Queensland or northern New South Wales in eastern Australia. Since this species is primarily Australian, selection of a lectotype should await study of Australian material. I did not find types of this species in the museum at Melbourne in 1957.
Occurrence in New Guinea. Papua: 1, Rouku, Morehead R., (opposite the tip of
Cape York), Apr. 1962 (W. W. Brandt, CSIRO).
Notes. In my key to New Guinean Di- crochile ( 1968 ) the present species runs to acuta Darlington, which in fact may prove to be a small New Guinean form of gigas. Specimens of acuta measure from c. 12.5 to c. 15.5 mm; Australian specimens of gigas, c. 20 mm or more. The individual from Rouku measures 19.5 mm and may represent an extension of the Australian population to southern Papua.
Tribe CHLAENIINI
Genus CHLAENIUS Bonelli
Darlington 1968, Part III, p. 20.
Notes. I have seen 376 additional speci- mens of Chlaenius from New Guinea. They all represent more or less common species of which the occurrence and gross distribu- tion in New Guinea has been adequately stated. Only a few of the detailed locality records are worth listing here. No addi- tional species of the genus have been found near Wau.
Chlaenius maculiger Castelnau
Darlington 1968, Part III, p. 25.
Additional material. West N. G.: 4, Waigeo Is., Camp 2 (Buffelhorn), June 1938 (Cheesman), 1 of these specimens labeled also "at light."
Tribe OODINr Genus ANATRICHIS Leconte Anafrichis pusilla Sloane Darlington 1968, Part III, p. 32.
I
^ An endemic species of the primarily Australian genus Coptocarpus has been found at Dumun, N-E. N. G., 7500-8000 ft. (c. 2300-2400 m), Aug. 21, 1967 (Fred Parker), but was received too late to be described in die present paper. It is not included in my statistical analysis of the New Guinean carabid fauna. It has atrophied wings, as have all the Australian species, but the genus is presumably derived from a winged ancestor.
The carabid beetles of New Guinea • Darlington
323
Additional material. West N. G. : 3, River Tor (mouth), 4 km E of Hoi Maffen, July 19, 1959 (T. C. Maa, Bishop Mus.).
Genus OODES Bonelli
Oodes exiguus Andrewes
Darlington 1968, Part III, p. 33.
Additional material. West N. G. : 2, River Tor (mouth), 4 km E of Hoi Maffen, July 19, 1959 (T. C. Maa, Bishop Mus.).
Oodes terresfris n. sp.
laevissimus Andrewes 1924, Ann. Mag. Nat. Hist.
(9), 14: 588 (not lacvissimus Chaiidoir 1882). Darlington 1968, Part III, pp. 33, 34.
Description. Form as in Figure 77; parallel-sided, slightly more depressed than usual; black, appendages dark; shining, but upper surface very finely (irregularly iso- diametrically) reticulate and minutely punc- tulate. Head 0.51 and 0.51 width prothorax; clypeus not margined, without seta-bearing punctures; labrum with 6 separate setae, the inner ones smaller but not clumped; an- terior supraocular punctures absent, poste- rior present. Prothorax: width/length 1.70 and 1.70; base/apex 1.80 and 1.77; a seta on basal edge each side near angle; disc with middle line extremely fine, basal im- pressions very shallow and poorly defined. Elytra: width elytra/prothorax 1.07 and 1.06; base margined; striae veiy fine on disc, deeper laterally, finely punctulate especially laterally; intervals flat on disc, 3rd with 2 small punctures near middle and at or be- hind apical %. Inner tcinfis full in some, slightly reduced in other individuals. Lower surface : sides especially of meso- and meta- sterna finely closely punctate; prosternal process not margined between coxae. Sec- ondary sexual characters: $ front tarsi moderately dilated, 2nd segment slightly wider than long, 3 segments densely squamulose below; <^ with 1, 5 2 setae each side before apex last ventral segment. Mea- surements: length 11.5-13.3 mm; width 4.6-5.7 mm.
Types. Holotype $ (MCZ, Type No.
31868) and 21 paratypes from Dobodura, Papua, Mar.-July 1944 (Darlington); and additional paratypes as follows. Papua: 1, Kokoda, 1200 ft. (366 m), Aug. 1933 ( Chees- man). N-E. N. G.: 19, Aitape, Aug. 1944 (Darlington); 7, lower Busu R., Huon Pen., Apr. 4, May 13, 1955 (E. O. Wilson, MCZ); 2, Erima, Astrolabe Bay, 1897 (Biro); 1, Sattelberg (British Mus.); 2, Wareo, Finsch- hafen (Rev. L. Wagner, South Australian Mus.).
Measured, specimens. The $ holotype and 1 9 para type from Dobodura.
Notes. A specimen of this species in the Andrewes Collection (British Museum) is identified as lacvissimus Chaudoir and marked as compared with type. Compari- son with this specimen led me in 1968 to misidentify the present species and rede- scribe the true laevissimus as long^ior (see below). The present species is a very dis- tinct one, recognizable by parallel-sided, depressed form and by technical characters given in my previous key (Part III, pp. 32- 33). It differs from most other species of the genus in that it occurs in leaf litter on the floor of rain forest rather than in or beside swamps or pools. It is the base spe- cies of what must now be called the terrestris group of Oodes, characterized in my key (referred to above), and including O. rossi Darlington and O. wilsoni Darlington, both of which seem to be -w, flightless derivatives of terrestris-\]ke stock.
Oodes laevissimus Chaudoir
Chaudoir 1882, Ann. Soc. Ent. France, ser. 6, 2:
361. Andrewes 1930, Cat. Indian Carabidae, p. 313
(Simons), longior Darlington 1968, Part III, p. 38 (NEW
SYNONYMY).
Description. See under Oodes longior Darlington 1968.
Types. A large <i ( which I now designate as lectotype) and a smaller $ specimen both labeled "Nuova Guinea / Fly River / L. M. D'Albertis 1876-77," and "Ex Museo Chaudoir."; in Obcrthiir Coll. (Box 191), Paris Mus. ( seen ) .
324 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Occurrence in Neto Guinea. Known only from the 2 types (above), from Papua, and from the types of lona.ior from Hollandia, West N. G., and Ambunti, Sepik R., N-E. N. G.
Notes. For explanation of my ( 1968 ) misidentification of this species, see under Oodes terrestris ( above ) .
Tribe HARPALINI Darlington 1968, Part III, p. 38.
This tribe includes some of the com- monest Carabidae in New Guinea. Much additional material of some of the species has come to hand, but I shall record here only one obviously new species and a few important new records. The number of additional specimens in this tribe not re- corded below is 249.
Genus CHYDAEUS Chaudoir
Darlington 1968, Part III, p. 47.
Chydaeus hinnus n. sp.
Description. Doubtfully assigned to this genus (see following Notes); form as in Figure 78, strongly convex; black, epi- pleurae and appendages browner; dull, reticulate microsculpture close, isodiametric on front, slightly transverse on pronotum, slightly more transverse on elytra. Head 0.79 width prothorax; eyes rather small, separated from mouth below by c. ¥2 their diameter; mentum toothed; ligula free at apex, narrow, 2-setose; paraglossae curved, shghtly longer than ligula, separate from it; 2nd segment labial palpi plurisetose. Prothorax cordate; width/length 1.24; base/ apex 1.06; lateral margins fine, each with seta c. ^/4 from apex; base weakly finely mar- gined, apex not margined at middle; disc with short impressed middle line, very weak transverse impressions, virtually no baso- lateral impressions, irregularly inconspicu- ously punctulate across basal area. Elytra narrowed toward base; width elytra/pro- thorax 1.26; humeri subdcntate; apices rather strongly sinuate; striae entire, mod-
erately impressed, impunctate; 3rd intervals without dorsal punctures. Inner icings ves- tigial, reduced to scales that scarcely extend beyond metasternum. Lower sui'face and legs without obvious special characters. Secondary sexual characters: 9 with 2 setae each side last ventral segment; $ unknown. Measurements: length c. 8.7 mm; width 3.4 mm.
Type. Holotype 9 (MCZ, Tvpe No. 31869) from Okapa (Kamira), N-E. N. G., May 11, 1965 (R. Hornabrook); the type is unique.
Notes. In the absence of the c^ , I cannot be sure that this species is a Chydaeus. But the species is an interesting one, which should be described, so that collectors will look for it in the future. It is surely different from anything previously known in New Guinea. It differs from Clujdacus papua Darlington (Part III, p. 47) in being smaller and smoother, with relatively narrower pro- thorax and less punctate pronotum.
i
Genus TRICHOTICHNUS Morawitz Darlington 1968, Part III, p. 48.
Notes. Of this genus, I have seen 232 ad- ditional specimens from New Guinea. Be- sides a number of relatively common spe- cies, the new material includes some specimens that cannot be satisfactorily placed now and that may represent new forms, but which are not sufficiently dis- tinct to describe without third-stage study. No species, other than those previously recorded, have been found at or near Wau. |
Genus EGADROMA Motschulsky I
Darlington 1968, Part III, p. 69.
Notes. Of the common species of this \ genus, 118 additional specimens have been I seen. ]
Egadroma cyc/ops Darlington Darlington 1968, Part III, p. 70. Additional material. Papua: 3, Brown
The carabid beetles of New Guinea • Darlington 325
R., May 25, 1956 (E. J. Ford, Jr., Bishop Mus.), in light trap.
Notes. The unique type is from the Cyclops Mts., near Hollandia, West N. G.
I see now that this species is similar to and probably represents E. rectifrons Bates, which is recorded from SE Asia, Sumatra, and Borneo, and is represented in the Phil- ippines (Leyte, series in MCZ, possibly distinguishable from true rectifrons) but not in Australia. However, the New Guinean individuals have the elytral apices more strongly sinuate than in my few specimens of true rectifrons.
Tribe LEBIINI
Darlington 1968, Part III, p. 80.
Two new genera have been added to the New Guinean list since Part III went to press. The following insertions to my Key to Genera of Lebiini of New Guinea (Part III, pp. 81ff) should facilitate their iden- tification. The species concerned are, of course, treated below^ with a few additional important new records, including one of a Lehia previously unknown from New Guinea. Besides the individuals recorded under other headings below, I have seen 220 additional specimens of this tribe from New Guinea.
(Insertions for Key to Genera of Lebiini of New Guinea )
5. Form characteristic, either as in Fig. 42 (1968) or Fig. 79 (present paper); upper surface either coarsely rugose and pubes- cent or sparsely pubescent chiefly at sides of elytra and anterior angles of prothorax ..,_ 5a Not described in one or more details 6
5a. Upper surface coarsely rugose and pubes- cent; form as in Fig. 42 (1968)
Lachnodenna
- Upper surface not coarsely rugose, pubes- cence sparse and restricted; form as in Fig. 79 Phijsodera
11. [Characters of Somotrichus]
Not as above in one or more ways 11a
11a. Form and color (head and prothorax yel- low, elytra dark blue) of minute Brachiniis;
length c. 5 mm Omobiiis
Either form or color different .— 12
Genus LEBIA Latreille
Darhngton 1968, Part III, p. 85.
Notes. Two species of Lehia are to be added to the number listed from New Guinea by me in 1968. One has been re- corded before but was overlooked by me in 1968. The other has recently been dis- covered on the island. Fifteen additional specimens of previously recorded species have been seen but need not be listed in detail.
iebia papuensis Macleay
Macleay 1876, Proc. Linnean Soc. New South
Wales, 1 : 167. Sloane 1917, Proc. Linnean Soc. New South Wales,
42: 424.
Description ( from Macleay ) . Apparently a Lehia of typical fonn; reddish-testaceous becoming brown on elytra "which have an indistinct black fascia near the apex"; pro- thorax short-transverse, probably of usual Lehia form; elytra broad, flat, sinuate- truncate, strongly striate with intervals broad and convex; length 3 lines ( c. 6 mm ).
Type. From Hall Sound (south coast of Papua); presumably in Macleay Coll., Sydney (not seen).
Occurrence in Neio Guinea. Known with certainty only from the type.
Notes. This may pro\'e to be a senior synonym of Lehia papuella Darlington (Part III, p. 88), but papuella is usually smaller and does not have the indistinct black fascia called for in the description of pa- puensis.
Lebia melanonota Chaudoir
Chaudoir 1870, Bull. Soc. Nat. Moscow, 4.3, Part 2, p. 226, t. 1, f. 45.
Csiki 1932, Coleop. Cat., Carabidae, Harpalinae 7, p. 1325 (see for synonymy and additional refer- ences ) .
Louwerens 1956, Treubia, 23: 225.
Description. A large Lehia; reddish yellow, elytra with broad black median stripe covering 4 inner intervals each side of suture and extending from base to less than ^4 from apex; prothorax transverse-
326 Bulletin Museum of Comparative Zoology, Vol. 142, No.
subcordate ( not hemispheric ) ; outer angles of elytra rounded; length c. 8 mm.
Type. From Moreton Bay (Brisbane), Australia; type in Oberthiir Coll., Paris Mus. (not seen).
Occurrence in Neiv Guinea. Papua: Rouku, Morehead R., West Papua, Apr. 1962 (W. W. Brandt, CSIRO).
'Notes. This species is now known from eastern Australia, New Guinea, the Solo- mons (specimens in MCZ), Moluccas (Ilalmahera Is., Louwerens, 1956), Java (Andrewes Coll., British Mus.), and the Lesser Sundas (Adonare Is.).
The large size and broad black median elytral stripe immediately distinguish me- lanonota from all species of Lehia previously known from New Guinea. Regardless of its distribution elsewhere, the occurrence of melanonoto at Rouku in southern Papua suggests a recent extension from Australia.
Genus PHYSODERA Eschscholtz
Eschscholtz 1829, Zool. Atlas, p. 8.
Csiki 1932, Coleop. Cat., Carabidae, Harpalinae 7,
p. 1346 ( see for additional references and list of
species ) . Jedlicka 1963, Ent. Abhandlungen, 28: 296, 300.
Diagnosis. Fonn characteristic ( Fig. 79 ) ; surface in part very sparsely pubescent (at sides of elytra, and more conspicuously at front angles of prothorax ) ; 4th hind-tarsal segments deeply emarginate. See preceding Supplementary Key to Genera.
Description. None required here.
Type species. Phy socle ra dejeani Esch- scholtz.
Generic distrihution. The SE comer of Asia to the Philippines, Celebes, and New Guinea (not Australia).
Notes. The new species described below constitutes the first record for this genus from New Guinea, and sets the eastern limit of the generic distribution.
Physodera bacchusi n. sp.
Description. With characters of genus ( above ) ; form as in Figure 79, broad, elytra relatively convex; head and pronotum black,
slightly reddish in part, elytra black slightly aeneous each with small subapical red spot near suture, lower surface irregu- larly reddish black, appendages black; shin- ing, reticulate microsculpture indistinct. Head 0.75 width prothorax; front flattened, slightly irregularly impressed each side and middle between eyes. ProtJwrax: width/ length 2.06; base apex 1.63; lateral margins broad, broadly reflexed, each with group of hairs at basal angle (including 1 special seta near base) and more longer hairs an- teriorly; disc with deep middle line coarse to base, irregular transverse impressions, baso-lateral impressions deep but scarcely distinct from posterior ends of marginal troughs; surface irregularly punctate across base and on margins. Elytra: width elytra/ prothorax 1.67; striae entire, punctulate but not otherwise deeply impressed; intervals almost flat, 3rd with several inconspicuous seta-bearing punctures mostly on inner edge, and 5th and 7th each with a few such punc- tures in part near middle of width of inter- vals. Lower surface scarcely punctate but in part (especially abdomen) sparsely pu- bescent. Inner a/nf^.s' fully developed. Legs: 4th hind-tarsal segments broad, very deeply emarginate, with long broad lobes; claws each with c. 6 teeth. Secondary sexual char- acters: S front tarsi with 3 segments very narrowly biseriately scjuamulose; 6 middle tibiae slightly bent out toward apex but not otherwise modified; $ with 1 principal seta each side last ventral segment; 9 unknown. Measurements: length 12.0 mm (to apex elytra ) ; width 6.0 mm.
Type. Holotype i (British Mus.) from Finisterre Rge., N-E. N. G., Damanti, 3550 ft. (1083 m), "Stn. No. 30," Oct. 2-11, 1964 (M. E. Bacchus); the type is unique.
Notes. Of previously described species, this is probably nearest to P. cyanipennis V. d. Poll (known to me only from th(^ de- scription, 1889, Notes from Leyden Mus. 11, p. 253) of Celebes, but the color is dif- ferent {cyanipennis has elytra dark blue with violet reflections, and without red spots), and the median line of the pronotum
I
The carabid beetles of New Guinea • Darlington 327
is obsolete in cyanipennis but deeply im- pressed in bacchusi.
Genus MINUTHODES Andrewes
Darlington 1968, Part III, p. 95.
Notes. Besides the individual of sed- lacekorum recorded below, I have seen 70 additional specimens of commoner species of this genus from New Guinea.
Minuthodes sedlacekorum Darlington Darlington 1968, Part III, p. 97.
Additional material. N-E. N. G.: 1, Okapa (Okasa), July 8, 1965 ( Hornabrook ) , "pine forest."
Notes. The unique type is from Wau. The present specimen agrees with it struc- turally but has the pale elytral markings forming 3 more nearly continuous fasciae.
Genus CATASCOPUS Kirby Darlington 1968, Part III, p. 101.
Notes. Seventy-five additional New Guin- ean specimens of this genus have been examined, in addition to those recorded below.
Catascopus sidus Darlington Darlington 1968, Part III, p. 105.
Additional material. N-E. N. G. : 4, Okapa (Okasa), Sept. 29, 1964 (Hornabrook).
Notes. Variation in color of this species is noted under the original description. The present specimens are entirely green above, except for coppery areas behind the humeri. Color may prove to distinguish geographic populations of the species in different parts of New Guinea, but much more material is necessary to delimit them.
Genus COPTODERA Dejean
Darlington 1968, Part III, p. 110.
Notes. Additional New Guinean speci- mens of this genus examined total 108.
Genus DOLICHOCTIS Schmidt-Goebel
Darlington 1968, Part III, p. 124.
Notes. One hundred forty additional specimens of Dolichoctis have been exam- ined from New Guinea but need not be listed in detail.
Genus OMOBRUS Andrewes
Andrewes 1930, Zool. Mededeelingen Mus. Leiden
13: 199. Van Emden 1937, Stettiner Ent. Zeitsclirift, 98:
41. Jedlicka 1963, Ent. Abhandlungen, 28: 299, 431.
Diagnosis. See under tribe Lebiini (above), insertion for key to genera of tribe. Form of minute Brachinus but abdo- men with only 6 visible ventral segments and mandibles without setae in scrobes; 4th hind-tarsal segments only emarginate ( not lobed ) ; claws simple.
Description. None required here.
Type species. O. proetextus Andrewes ( below ) .
Generic distribution. See under following species.
Notes. Andrewes originally assigned this genus to the tribe Brachinini, but this was surely wrong. The insect does look like a small Brachinus, but its technical charac- ters are wholly different. It is in fact a member of the Lebiini, as stated by Van Emden. The latter's assignment of it to a place among the Dromii is at least reason- able, although further study is needed to decide its exact relationships.
Omobrus praetexfus Andrewes
Andrewes 1930, Zool. Mededeelingen Mus. Leiden
13: 200. Jedlicka 1963, Ent. Abhandlungen, 28: 431 PI. 4
Fig. 30.
Description. With characters of genus; form as in Jedlicka's colored figure; Brachinus-\ike- head and prothorax yellow, elytra dark blue, lower surface posteriorly dark, legs yellow, antennae dark with 3 basal segments yellow; upper surface ir- regularly pubescent, front and pronotal disc
328 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
otherwise nearly smooth, elytra roughened. Length of New Guinean individual 5.3 mm.
Type. From Java, in British Mus. ( seen ) .
Occurrence in New Guinea. N-E. N. G. : 1, Wareo, Finschhafen ("Finsch Haven") (Rev. L. Wagner, South Australian Mus.).
Notes. Specimens of OmoJ)rus which may all be assignable to praetextus have now been found in Java, Malaya, the Philip- pines including Luzon ( at Bagnio, Darling- ton), New Guinea, and the Solomons. The Philippine form seems to be a (slightly defined) subspecies, punctulatus Jedlicka, and the Solomon Is. form has been de- scribed as subspecies hrachinoides by Van Emden. The New Guinean form also may prove to be slightly differentiated, but I do not care to describe it on the basis of the single known specimen.
My Philippine specimens were found (near Bagnio) under stones where the ground was damp with seepage, at altitudes of c. 2000 m.
Genus PARENA Motschulsky
Parana p/'cea (Macleay)
Darlin,uton 1968, Part III, p. 139.
Additional material N-E. N. G.: 1, Markham R., 50 m, Jan. 20-25, 1962 (Sed- lacek); 1, Mt. Otto summit, Nov. 1965 (Dept. Agr. Port Moresby).
Genus DEMETRIDA White Darlington 1968, Part III, p. 140.
Notes. Besides the 3 new species de- scribed below and the specimens of a few previously described species recorded in detail, 177 additional New Guinean speci- mens of this genus have been received since Part III went to press. Some addi- tional new species may be represented in this material, but their discrimination would require more extensive study than I can undertake now.
For Demetrida in the Moluccas, see Dar- lington, 1968a.
Demefrida nigripes Darlington Darlington 1968, Part III, p. 171.
Additional material. West N. G.: 1, Waigeo Is., Mt. Nok, Camp 2 ( Buf felhorn ) , June 19.38 (Cheesman).
Notes. The two previously known speci- mens of this very distinct species were from localities in Papua and N-E. N. G. respec- ti\'ely. The present specimen extends the known range of the species to beyond the western tip of New Guinea.
Demefrida nigriceps Darlington Darlington 1968, Part III, p. 174.
Additional material. West N. G.: 3,
Wissel Lakes (Moanemani and Enarotadi), 1500, 1850, 1850-1900 m, dates in June, July, Aug., 1962 (Sedlacek). N-E. N. G.: 1, Mt. Missim, 1600-2000 m, Sept. 21-24, 1964 (M. Sedlacek).
Notes. The 2 types, from Sibil Valley, Star Mts., West N. G., were both S S. The 4 individuals recorded above are all 9 9 . They compare well with the types, but c^ $ are necessary to confirm the identification.
Demefrida karimui n. sp.
Description. With characters of genus; | rather slender, normally convex; head, pro- thorax, base and apex and (very narrowly) lateral margins of elytra black, disc of elytra broadly red, legs black, antennae and tarsi i brown; reticulate microsculptme indistinct. Head 1.02 and 1.05 width prothorax; eyes prominent, genae very short. Prothorax sub- quadrate except angles rounded to neck in front; width/length 1.00 and 0.99; base/ apex 1.43 and 1.54 (but apex can not bej measured exactly); base/head 0.92 and 0.95; sides rather weakly rounded anteriorly, sin- uate well before c. right but blunted basal angles; margins rather narrow, each with seta-bearing puncture slightly before mid- dle but none at base; disc with deep entire middle line, other impressions very weak, surface slightly irregular or subpunctate baso-laterally. Elytra: width elytra/pro-
The carabid beetles of New Guinea • Darlington
329
thorax 1.97 and 2.07; apices with short spines or very acute teeth, with outer angles right or acutely subdenticulate; striae slightly impressed, punctulate; intervals very slightly convex, 3rd 2-punctate. Claws with c. 5 teeth. Secondary sexual characters of 6 unknown; 9 with 4 or 5 apical ventral setae each side. Measurements: length 8.2- 9.6 mm; width 2.8-3.2 mm.
Types. Holotype 9 (Bishop Mus.) and 3 (all 9 9) paratypes (2 in MCZ, Type No. 31871) all from Karimui, N-E. N. G., 1080 m, July 13 (1 paratype July 14-15), 1963 (Sedlacek).
Notes. In my key to the New Guinean species of Demetrida (Part III, pp. 146ff), this runs to couplet 47, and falls with saidor. However, as compared with saidor, the present new species has a relatively nar- rower head and narrower elytra (compare proportions given) and differs in color, the elytra being entirely bordered with black (only humeri or small basal area black in saidor). The color recalls D. dorsalis (Part III, p. 171), but the latter has a much wider prothorax; the present species is probably not related to dorsalis. The present specie^ also resembles D. discoidalis (Part III, p. 181) but lacks the posterior-lateral pro- thoracic setae of the latter, and has the ely- tra more extensively red; again there is probably no direct relationship between these two species. The general similarity of color of D. dorsalis, discoidalis, and karimui is probably simply convergent, or possibly mimetic.
Demetrida parena n, sp.
Description. With characters of genus; foi-m as in Figure 80; black, elytra faintly greenish in some lights, appendages reddish brown; reticulate microsculpture iso- diametric on front, faint and somewhat transverse on pronotum, more distinct and and much more transverse on elytra. Head 0.86 width prothorax; eyes large, prominent, with genae short and oblique. Prothorax: width/length 1.34; base/apex 1.49; base/ head 1.05; sides rather strongly sinuate well
before c. right slightly blunted posterior angles; margins moderate, not strongly re- flexed, each with seta-bearing puncture at or slightly before middle but none at base; disc with middle line coarse and entire, transverse impressions very weak, surface subpunctate across base and in margins. Elytra short (in genus); width elytra/pro- thorax 1.81; outer angles obtuse but distinct, apices short-spined, sutural angles obtuse; striae moderately impressed, faintly punctulate; intervals very slightly convex, sparsely punctulate, 3rd 2-punctate. Claws with c. 4 teeth. Secondary sexual charac- ters: S unknown; 9 with 3 or 4 (un- symmeti-ic) seta-bearing punctures each side last ventral segment. Measurements: length 9.4 mm; width c. 4.0 mm.
Type. Holotype 9 (British Mus.) from Waigeo Is., West N. G., Camp Nok, 2500 ft. (c. 660 m), Apr. 1938 (Cheesman); the type is unique.
Notes. This new species is apparently close to D. imitatrix Darlington of New Guinea (1968, Part HI, p. 176) but differs in color, being black with elytra at most faintly greenish (not blue) and with ap- pendages reddish brown (not dark). It differs also slighdy in proportions, espe- cially in having a slightly broader pro- thoracic base than imitatrix.
In appearance, except for the spined ely- tra, this new species looks like an unmarked Parena, from which the specific name is derived.
Demetrida viridlpennis Darlington Darlington 1968, Part III, p. 177.
Additional material. N-E. N. G.: 1, Ilerzog Mts., Morobe Dist., Vagau, 4000 ft.' (1220 m), Jan. 4-17, 1965 (Bacchus, British Mus.), Station No. 144. West N. G.: 1, Japen Is., Gamp 2, Mt. Eiori, 2000 ft. (610 m), Sept. 1938 (Gheesman).
Notes. This species is known from a num- ber of localities in Pa[)iia and N-E. N. G, but the Japen Is. specimen is the first re- corded from West N. G.
330 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
Demetrida aiyura n. sp.
Description. With characters of genus; form c. as in D. scticollis (Part III, Fig. 109) but differing sHghtly in proportions (see following ratios), without extra pronotal setae, and with longer elytral spines; brown, head and prothorax castancous, elytra paler except slightly darker near base and sides, appendages slighdy paler; shining, reticu- late microsculpture absent on front and pro- notum and scarcely distinct (apparently c. isodiametric but very lightly impressed ) on elytra. Head 1.10 and 1.13 width prothorax; eyes normally prominent, genae shorter, oblique. Prothorax: width/length 0.99 and 0.99; base/apex 1.42 and 1.31; base/head 0.92 and 0.90; lateral margins rather narrow anteriorly, slightly broader posteriorly, each with strong seta at basal angle, strong seta at or slightly before middle, but no trace of other anterior setae or formerly-seta-bearing punctures; disc with strongly impressed middle line, weak transverse impressions, surface slightly irregular but not punctate basally. Elytra: width elytra/prothorax 2.00 and 2.04; apices with outer angles acute but not much produced, long-spined c. op- posite ends 3rd intervals, with sutural angles obtuse; striae slightly impressed, slightly punctulate; intervals slightly (scarcely) con- vex, 3rd 2-punctate. Claws with c. 5 teeth. Secondary sexual characters: $ tarsi as genus; S middle tibiae slightly bent out, tuberculate-serrate in distal third of inner edge (c. 4 tubercles); S with 3, ? 3 or 4 (unsymmetric) setae each side last ventral segment. Measurements: length 9.4-10.0 mm (including elytral spines); width 2.9- 3.5 mm.
Types. Holotype $ (British Mus.) from Aiyura (nr. Kainantu, E Highlands), N-E. N. G., (altitude probably c. 1500 m), Sept. 2.5, 1957 (J. Smart); 1 5 paratype (Bishop Mus.) from Pindiu, Huon Pen., N-E. N. G., 870-1300 m, Apr. 21-22, 1963 (Sedlacck).
Notes. I hesitate to describe another spe- cies in this genus from 2 specimens from different localities, but the specimens are clean and in good condition, and they agree
well in proportions and nonsexual details except that the 9 has the prothorax a little i less narrowed antt>riorly, with better de- fined anterior angles (reflected in the ratio of prothoracic base/apex). These speci- mens have characters which suggest the ex- istence of a distinct, possibly ancestral pop- ulation in a group of Demetrida in which several related species are localized. The group in question includes nonpubescent forms with posterior-lateral prothoracic se- tae present and with elytra spined. They are placed at the cmd of my key to the New Guinean species of Demetrida (Part III, pp. 146-149 ) , in couplets 57ff . From the species there named, the present new one is dis- tinguished from seticollis by lacking extra anterior-marginal prothoracic setae, and from the others {pallipes, etc.) by brown color, the others being wholly or mainly black, blue-black, or greenish-black, and often with dark legs. Although the present new species is based on only 2 individuals from different localities, its immediate rela-- tives (indicated above) are known from series which indicate that variation within single populations is not excessive.
If it were not for the presence of posterior- lateral prothoracic setae, this species would I run ( in the key referred to above ) to forma (Part III, p. 167), which may be a com- posite species. However, aiyura differs from typical forma ( the measured 5 9) in pro- portions of base of prothorax width of head^ and of width of elytra/prothorax, as well as in presence of the posterior prothoracic se- tae.
Tribe PENTAGONICINI
Genus PENTAGONICA Schmidt-Goebel
Dariington 1968, Part III, p. 192.
Notes. Nine additional New Guinean specimens seen.
Genus SCOPODES Erichson |
Darlington 1968, Part III, p. 197.
Notes. Five additional New Guinean specimens seen.
The carabid beetles of New Guinea • Darlington 331
Tribe HEXAGONIINI
Genus HEXAGONIA Kirby Darlington 1968, Part III, p. 202.
Notes. The following new species, the second of the genus to be found in New Guinea, is not directly related to the other (papua Darlington, Part III, p. 203) but represents a different Oriental stock.
Hexagonia gressitti n. sp.
Description. With characters of genus; form as in Figure 81; brownish black, legs yellow, antennae dark brown; shining, re- ticulate microsculpture absent on front and pronotum and indistinct on elytra. Head 0.95 width prothorax; antennae short; man- dibles long, nearly straight; front trans- versely grooved anteriorly, broadly im- pressed each side between eyes. Prothorax cordate; width/length 1.26; base/apex 1.29; lateral margins narrow but separated from disc by channels, each with seta-bearing puncture c. '■■. from apex but none at base; disc with deep middle groove, other im- pressions subobsolete, impunctate. Elytra: width elytra/prothorax 1.43; striae im- pressed, punctate; intervals slightly convex, 3rd 3-punctate ( 1 near base on outer edge, 2 behind middle), 5th with 1 puncture on outer edge c. V\ from apex. Lower surface: sides of prostenium (but not most of pro- epistema) with some coarse punctures; ab- domen somewhat roughened. Legs: mod- erate; tarsi wide; 4th segments deeply emarginatc; claws not toothed. Secondary sexual characters: i front tarsi apparently without sexual squamules; $ with 1 seta each side last ventral segment; 9 unknown. Measurements: length 6.0 mm; width 2.0 mm.
Type. Holotype S (Bishop Mus.) from Nabire, S of Geelvink Bay, West N. G., 1-20 m, July 8, 1962 (Gressitt); the type is unique.
Notes. This is much smaller than the single previously known New Guinean Hex- agonia (see under genus), differently col-
ored (not bicolored), and differently shaped. It has evidently been derived from the Orient independently. General form and plain brown color seem to relate it to several Oriental species including bowringi Schaum, but gressitti is distinct from all of them by its small size and relatively long, slender, nearly straight mandibles.
Tribe ODACANTHINI Genus DOBODURA Darlington Dobodura armata Darlington Darlington 1968, Part III, p. 21.5.
Additional tnaterial. N-E. N. G.: 1, nr. Okapa, Wanatabe Vy., E Highlands, c. 5000 ft. (1525 m), Feb. 5, 1965 (Bacchus, British Mus.).
(Genus ANDREWESIA Liebka)
Liebke 1938, Festschrift Embrik Strand, 4: 84 ( not Andreivesius Jedlicka 1932, Ent. Nachrich- tenblatt, 6: 74).
{Andrewesia apicalis (Chaudoir))
Chaudoir 1872, Bull. Soc. Nat. Moscow, 4.5 (1):
408 (Odacantha). Andrewes 1930, Cat. Indian Carabidae, p. 229
( Odacantha) . Csiki 1932, Coleop. Cat., Carabidae, Harpalinae,
8: 1536. Liebke 1938, Festschrift Embrik Strand, 4: 85. Louwerens 1967, Ent. Meddelelser, 3.5: 202.
Description. None needed here.
Types. From Bangkok and Singapore,
presumably in Oberthiir Coll., Paris Mus. (not seen).
Occurrence in New Guinea. Doubtful.
Notes. I did not refer to this species in Part III because I failed to note that New Guinea is included in its range by Csiki and (presumably following Csiki) by Louwer- ens. I have been unable to find the source of the New Guinean record. It may be a compiler's error. Or it may be based on a misidentified specimen of the species de- scribed by me (Part III, p. 214) as Eudalia anomala. There are taxonomic problems
332 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
here at both the generic and specific levels which I cannot solve now.
Tribe ZUPHIINI (LELEUPIDIINI)
Leleupidiini Basilewsky 1951, Re\iie Zool. Bot. Afr.,
44: 178.
1953, Revue Zool. Bot. Afr., 47: 264.
1954, Revile Frangai.se d'Ent., 21: 213.
1967, Bull. Soc. Ent. France, 72: 250.
Landin 1955, Arkix- f. Zoolosi, ser. 2, 8: 467 (Gun-
voiita, new genus from Sikkini, evidently a le- leupidiine ) .
This group of small-eyed, flightless, ant- like carabids differs from more ordinary Zuphiini in having the first antennal seg- ment not scaphifonn, the palpi remarkably modified, and in other details. Whether it should be recognized as a separate tribe or as a subgroup within the Zuphiini is a matter of point of view — of how many tribes one wishes to recognize within the family Carab- idae. The leleupidiines are surely a dis- tinct group worthy of at least subtribal rec- ognition.
Most leleupidiines are African (and 1, Madagascan), but 2 have been described from southern Asia, and the range of the group is now extended to New Guinea. As Basilewsky (1954) suggests, the group is probably old ( originally dispersed, I should think, by winged ancestors ) and now has a relict distribution, the (flightless) survivors occurring on certain mountains in Africa and at scattered localities in southern Asia, etc.
Genus COLASIDIA Basilewsky
Basilewskv 1954, Revue Francaise d'Ent., 21: 215, fig. 1.
Type species. C. malayica Basilewsky, of Singapore.
Dw^no.s?,? (for identification in New Guinea only ) . Anthke Zuphiini; eyes small; first antennal segment not scaphifonn; api- cal segment labial palpi greatly enlarged.
Description. None required here.
Generic distribution. At present knowm only from Singapore and New Guinea.
Notes. I am not prepared to discuss the
relationships or differential characters of this genus in comparison with other leleupidi- ines.
The two New Guinean species described below ( each known from a single ? ) re- semble malayica ( which also is known from a single 9 ) in general but differ in details, especially in shape of the head and pro- thorax. These three species are all flightless now ( unless they prove to be dimorphically winged), but their common ancestor that dispersed across the Malay Archipelago may have been winged.
Colasidia papua n. sp.
Description. With characters of genus; fonn as in Figure 82; brown, appendages paler; whole upper surface closely coarsely punctate and rather sparsely pubescent, but surface shining between punctures under the pubescence. Head 0.86 width prothorax; eyes moderate (large in group), enclosed behind by long genae; antennae moniliform, 1st segment scarcely longer than 3rd; man- dibles short, transverse; labrum 6-setose; front almost e\'enly conxex, with \'ery small frontal impressions almost above antennal bases; 2 setae each side (doubtfully dis- tinguishable from the general pubescence) over eye and at posterior corner of head; mentum with emarginate tooth at middle; labium subtruncate (slightly broadly emarginate), 1-setose each corner; para- glossae membranous, longer than labium. Prothorax: width length 0.91; base/apexi 1.32; posterior angles subbasal, dentiform; margins narrow, each with seta on denti- fonn posterior angle and c. V* from apex; disc strongly convex, scarcely impressed. Elytra: width elytra/ prothorax 1.82; punc- tures not fonning distinct striae, but inter- \als indistinctly indicated. Inner uing,s at- rophied. Lower surface extensively punc- tate, but punctation absent on proepistema and sparse on anterior-lateral portions of ventral segments. Legs moderately slender; tarsi rather long, segments not much lobed; claws simple. Secondary sexual characters: 9 with 1 seta-bearing puncture each side
The carabid beetles of New Guinea • Darlington 333
apex last ventral segment; i unknown. Measurements: length 4.5 mm (to apex elytra) (c. 5.0 to apex abdomen); width 1.7
mm.
Type. Holotvpe ? (sex determined by dissection) (MCZ, Type No. 31875) from Dobodura, Papua, Mar .-July 1944 (Dar- lington ) ; the t}'pe is unique.
Notes. This insect has a more o\'al head, more rounded posteriorly and with rela- tively larger eyes, than Cohisidia malayica Basilewsky, and is probably much more coarsely punctate above. It is evidently en- demic to New^ Guinea, and probably local- ized \\'ithin the island. It cannot fly and therefore is unlikely to be taken in light traps. The type was taken struggling in flood water in rain forest, where a cloud- burst had brought down a bank and blocked a small brook. The insect probably lives in or under leaf litter on the forest floor.
Colasidia madang n. sp.
Description. With characters of genus; form as in Figure 83; black, legs irregularly brownish, antennae browiiish yellow; an- swering description of preceding species ( papiio ) in detail ( except mouthparts not examined) except as follows. Punctation of upper surface slightly finer and of head slightly less dense. Head 0.89 width pro- thorax; eves smaller and more transverse than in papua. Prothorox: width length 0.95; base apex 1.40; posterior angles more prominent and more acute than in papua. Elytra slightly shorter and more rounded than in papua; width elytra pro thorax 2.07. Measurements: length 4.4 mm (to apex elytra) (5.0 to apex abdomen); width 1.75 mm.
Type. Holotype ? (sex determined by dissection) (British Mus. ) from Damanti, Madang Dist., Finisterre Rge., N-E. N. G., 3550 ft. (1083 m), "Stn. No. 34," Oct. 2-11, 1964 (M. E. Bacchus); the t>'pe is unique.
Notes. The differences between the t\vo New Guinean species of Colasidia are in- dicated in the preceding description. The most important are that madang,, as com-
pared with papua, is black rather than bro\A'n, with smaller eyes, prothorax with more prominent and acute posterior angles, and slightly finer dorsal punctation. As com- pared with (the original figure of) malayaca Basilewsky, madang has the head less wide posteriorly, the prothorax with much more prominent and more acute posterior angles, and the whole upper surface probably more coarsely punctate.
Tribe HELLUONINI
Genus HELLUONIDIUS Chaudoir
IHelluonidius lafipes Darlington Darlington 1968, Part III, p. 231.
Additional material. N-E. N. G.: 1, Finisterre Rge., Damanti, 3550 ft. ( 1083 m), Oct. 2-11, 1964 (Bacchus, British Mus.), Station No. 46; 1, Finisterre Rge., Budemu, c. 4000 ft. (1220 m), Oct. 15-24, 1964 ( Bacchus, now in MCZ ) .
Notes. The Damanti individual is a 5 with front tarsi with segments 2 and 3 nar- rowly 2-seriately squamulose; squamules are present also below segment 4 but may not be attached to it.
The unique type of latipes is from Rattan Camp, Snow Mts., West N. G.
Genus HELLUOPAPUA Darlington
Darlington 1968, Part III, p. 232.
Diagno.sis (revised). Characters as given ( 1968 ) except S front tarsi either without squamules or with 2 rows of minute squamules at middle of segments 2 and 3.
Generic distribution (revised). West N. G.: now known from the type species from Rattan Camp, Snow Mts., and from the following new species from Waigeo Is.
Notes. See Notes under the following species for discussion of this genus in rela- tion to Helluonidius.
Helluopapua cheesmani n. sp.
Description. With characters of genus; form as in Figure 84; black, appendages
334 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
slightl)' brow nish; shining, reticulate micro- sculpture absent or indistinct even on elytra. Head 0.89 and 0.93 width prothorax; genae rounded, slightly prominent; clypeus slightly sinuate-truncate, broadly ti-iangularly pro- duced at middle (slightly more than in papua), with several long setae each side but none near middle; front weakly convex, longitudinally impressed each side and transversely impressed anteriorly, sparsely irregularly punctate; mentum with acute ti'i- angular tooth and long pointed side lobes; ligula c. as in papiia (but see Notes below). Prothorax: width length 1.39 and 1.43; base/apex 0.75 and 0.72; base/head 0.66 and 0.61; lateral margins narrow, not inter- rupted; disc weakly convex, middle line and anterior transverse impression distinct, baso-lateral impressions small and weak, surface of disc irregularly coarsely punctate with impunctate areas each side of middle. Elytra: width elytra prothorax 1.35 and 1.34; striae deep, impunctate; intervals con- vex, each with a row of punctures each side. Measurements: length c. 23.0 mm; width 6.0 mm.
Types. Holotvpe $ (British Mus.) and 1 9 parat)'pe ( MCZ, Type No. 31877) both from Waigeo Is., West N. G., Mt. Nok, Camp 2 (Buffelhom), June 1938 (Chees- man ) .
Notes. I have been able to compare these specimens directly with the unique S type of HeUuopaptia toxopei (which has not yet been returned to the Leiden Museum). The two species agree well in most generic char- acters including foiTn of labrum, form of inner lobe of maxillae (but see below), and slender tarsi. However, while the i of H. toxopei completely lacks sexual squamae on the front tarsi, that of cheesniani has minute white scales, in two series, on segments 2 and 3 only. In addition to this, the present new species differs from toxopei in having the elytra without distinct microsculpture, and in having the elytra relatively narrower (or the head and prothorax wider).
The hook of the inner lobe of the maxillae is more nearly apical in the present species
than in toxopei, but still less strictly apical than in Helluouidius. The difference is actually due to the form of the outer apical angle of the maxillary lobe, which is con- siderably produced in Helluopapua toxopei, somewhat produced in the present new spe- cies, and not produced in Helluouidius.
The next reviser of this group will have to decide whether, in \'iew of the fact that the present new species is in some way transi- tional between Helluonidius and Helluo- papua (although much closer to the latter), the two genera should be kept separate.
Tribe BRACHININI
Genus BRACHINUS Weber
Brachinus papua Darlington
Darlington 1968, Part III, p. 239.
Additional material. N-E. N. G.: 1, Maprik (Sepik Dist.), Oct. 26, 1957 (J. Smart, British Mus.).
BIBLIOGRAPHY
( See also references under individual species. )
Andreaves, H. E. 1927. Carabidae. In Insects of Samoa, Part IV, Fasc. 1-Coleop., pp. 1-14. London, British Museum.
. 1929. Coleoptera, Carabidae, Vol. 1, Ca-
rabinae ["Table of Tribes" of Carabidae, pp. 4.3-46]. In Fauna of British India, . . . Lon- don, Taylor and Francis.
. 1930. Catalogue of Indian Insects. Part
18-Carabidae. Calcutta, Government of India.
(The) Australian Encyclopaedia (A. H. Chis- holm, ed.). 1958. 10 vols. Sydney, Angus & Robertson; East Lansing, Michigan, Michi- gan State University Press.
Ball, G. E. 1956. Notes ... on the classifica- tion of the tribe Broscini . . . Coleoptcrists' Bull., 10: 33-52.
. 1959. A taxonomic study of the North
American Licinini . . . Mem. American Ent. Soc. (Philadelphia), No. 16: 2.58 + iv pp., 15 pis.
. 1960. Carabidae. In Arnctt, R. II., The
Beedes of the United States, Part 4: .55-174. Ann Arbor, American Entomological Insti- tute.
Baxnlnger, M. 1950. The subtribe Pasimachina
The carabid beetles of New Guinea
Darlington
335
(Coleoptera, Carabidae, Scaritini). Rev. En- tomologia, 21: 481-511.
Basilewsky, p. 1953. Carabidae. In Explora- tion du Pare National de I'Upemba, Fase. 10, pp. 1-252, 10 pis., map.
Bishop Museum. 1966. List of New Guinea localities [mimeographed]. Honolulu, Bishop Museum.
BisKUP, P., B. Jinks, and H. Nelson. 1968. A Short History of New Guinea. Sydney, Angus and Robertson.
Brass, L. J. 1941. The 1938-39 expedition to the Snow Mountains, Netherlands New Guinea. J. Arnold Arboretum, 22: 271-342, 7 pis.
. 1964. Results of the Archbold Expedi- tions. No. 86. Summar>' of the Sixth Arch- bold Expedition to New Guinea (1959). Bull. American Mus. Nat. Hist. 127: 145-215, 12 pis.
Brookfield, H. C. 1966. The Chimbu: a high- land people in New Guinea. 7n Eyre, S. R., and G. R. J. Jones, (eds.). Geography as Himian Ecology / Methodology by Example: 174-198. New York, St. Martin's Press.
Brown, R. W. 1956. Composition of Scientific Words (revised ed.). Privately published; now available from Smithsonian Institution, Wash- ington.
Brown, W. L., Jr. 1958. General adaptation and evolution. Syst. ZooL, 7: 157-168.
Browne, W. R. 1958. Geology [of Papua and N-E New Guinea]. In Australian Enc\x'lo- paedia. Vol. 6: 452-453.
Bruntjin, L. 1965. On the real nature of trans- antarctic relationships. Evolution, 19: 496- 505.
. 1966. Transantarctic relationships and
their significance, as evidenced by chironomid midges, . . . Stockholm, Ahnquist & ^^'iksell.
Chees.\l\x, (L.) En'elyx. n. d. [Summary of her own expeditions.] In Barrett, C. (ed. ), The Pacific: 63-65. Melbourne, N. H. Sew- ard.
CsiKi, E. 1927-1933. Carabidae: Carabinae, (Mormolycinae), Harpalinae. 7/i Schenkling, S., Coleop. Cat., Pars 91, 92, 127 (Carab- inae), 97, 98, 104, 112, 115, 121, 124, 126 Harpalinae). (Berlin), The Hague, W. Junk.
Darlington, P. J., Jr. 1936. Variation and at- rophy of flying wings of some carabid beetles. Ann. Ent. Soc. America, 29: 136-176, 3 pis.
. 1937. West Indian Carabidae III: new
species and records from Cuba, with a brief discussion of the mountain fauna. Mem. Soc. Cubana Hist. Nat., 11: 115-136.
■ . 1943. Carabidae of mountains and islands:
data on the evolution of isolated faunas, and on atrophv of wings. Ecological Monographs, 13: 37-61.
1947. Two new tiger beetles from New Guinea. Psyche, 54: 241-245.
1948. The geographical distribution of cold-blooded \ertebrates. Quart. Rev. Biol., 23 : 1-26, 10.5-123.
1952. The carabid beetles of New Guinea. Part II. The Agonini. Bull. Mus. Comp. Zool., 107: 87-252, 4 pis.
-. 1956. Australian carabid beetles III. Notes on the Agonini. Psyche, 63 : 1-10.
1957. Zoogeography: The Geographical Distribution of Animals. New York, John Wiley & Sons.
1959. Area, climate, and evolution. Evo- lution, 13: 488-510.
-. 1959a. Thv Bemhidion And Trechus {Co\: Carabidae ) of the Malay Archipelago. Pacific Insects, 1 : 331-345.
1960. Australian carabid beetles IV. List of localities, 1956-1958. Psyche, 67: 111- 126.
1961. Australian carabid beetles V. Tran- sition of wet forest faunas from New Guinea to Tasmania. Psyche, 68: 1-24.
1961a. Australian carabid beetles VII. Trichosternus, especially the tropical species. Psyche, 68: 11.3-130.
1962. The carabid beetles of New Guinea. Part I. Cicindelinae, Carabinae, Har^ialinae through Pterostichini. Bull. Mus. Comp. Zool., 126: 321-564, 4 pis.
1962a. Australian carabid beetles X. Bem- bidion. Breviora (Mus. Comp. Zool.), No. 162: 1-12.
1963. Australian carabid beetles XIII. Further notes on Agonini, and a genus of Licinini new to Australia. Breviora ( Mus. Comp. Zool.), No. 183: 1-10.
1964. Australian carabid beetles XIV. Perigona. Ps> che, 71 : 125-129.
1965. Biogeography of the Southern End of The World. Distribution and history of far-southeni life and land, with an assessment of continental drift. Cambridge, Mass., Harvard University Press.
1967. A new Oniophron and a new Siagona from the Philippines. Psyche, 74: 314-319.
1968. The carabid beetles of New Guinea. Part HI. Harpalinae ( continued ) : Perigonini to Pseudomorphini. Bull. Mus. Comp. Zool., 137: 1-253, 7 + 3 pis.
-. 1968a. Dcmctrida (Coleoptera: Carabi- dae) in the Moluccas. Psyche, 75: 360-362.
-. 1970. Coleoptera, Carabidae, including Cicindelinae. Insects of Micronesia, 15 (1): 1-49. Honolulu, Bishop Museum,
1970a. A practical criticism of Hennig- Brundin 'Th\logenetic Systematics" and ant- arctic biogeography. Syst. Zool., 19: 1-18.
336 Bulletin Museum of Comparative Zoology, Vol. 142, No. 2
David, T. W. E. 1950. The Geology of the Commonwealth of Australia. London, Arnold.
Erwin, T. L. 1970. A reclassification of bom- bardier beetles and a taxonomic revision of the North and Middle American species (Carabi- dae: Brachinida). Quaestiones entomologicae, 6: 4-215.
Greenslade, p. J .M. 1967. Problems in study- ing the fauna of soil and allied habitats in Solomons Islands rain forest. J. Ecology, 36: 68P.
. 1968. Habitat and altitude distribution of
Carabidac (Coleoptera) in Argyll, Scotland. Trans. R. Ent. Soc. London, 120: 39-54.
Gressitt, J. L. 1953. The coconut rhinoceros beetle (Oryctes rhinoceros), with particular reference to the Palau Islands. Bull. Bishop Nhis., 212: 1-157 [Pseudozaena and Morion, pp. 95-96].
. 1959. Wallace's Line and insect distribu- tion. Proc. XVth International Congress Zool., sect. I, Paper 16; 66-68.
. 1961. Problems in the zoogeography of
Pacific and Antarctic insects. Pacific Insects Monograph, 2: 1-94.
Gressitt, J. L., and J. J. H. Szent-Ivany. 1968. Bibliography of New Guinea entomology. Pacific Insects Monograph, 18: 1-674.
, . In Preparation. [History of entomol- ogical exploration in New Guinea.]
, . In Preparation. [Environment of
New Guinea from an entomological stand- point.]
Hahu, a. 1967. Fauna Japonica, Carabidac, Truncatipennes group. Biogeographical Soc. Japan, Uyeno Park, Taito-ku, Tokyo.
Hennig, W. 1966. Phylogenetic Systematics. Urbana, University of Illinois Press.
Hlavac, T. F. 1969. A review of the species of Scarites (Antilliscaris) . . ., with notes on their morphology and evolution. Psyche, 76: 1-17.
Horn, W. 1926. Carabidac: Cicindelinae. In Schenkling, S., Coleop. Cat., Pars. 86. (Berlin), The Hague, W. Junk.
Jeannel, R. 1941-1942. Coleopteres carabiques. Faune de France, 39, 40. Paris, Lechevalier et fils.
Keast, a. 1968. Australian mammals: zoogeog- raphy and evolution. Quart. Rev. Biol., 43: 37.3-408.
Lea, D. A. M., and P. G. Irwin. 1967. New Guinea. The Territory and its People. Mel- bourne, etc., Oxford University Press.
Lindroth, C. H. 1956. A revision of the genus Synuckus Gyllenhal ... in the widest sense, . . . Trans. R. Ent. Soc. London, 108: 485- 576.
LouwEREXs, C. J. 1953. The Oriental species of Colpodes Macl. Treubia, 22: 75-151.
LovEiuDGE, A. 1948. New Guinean reptiles and 1 amphibians in the Museum of Comparative Zoology and United States National Museum. Bull. Mus. Comp. Zool., 101: 302-430 [list of localities, pp. 310-314].
MacArthur, R .H., and E. O. Wilson. 1967. The Theory of Island Biogeography. Prince- ton, N. J., Princeton University Press.
Mani, M. S. 1968. Ecology and biogeography of high altitude insects. Series Ent., Vol. 4. The Hague, W. Junk.
Mayr, E. 1944. Wallace's Line in the light of re- cent zoogeographic studies. Quart. Rev. Biol., 19: 1-14.
. 1963. Animal Species and Evolution.
Cambridge, Mass., Harvard University Press.
. 1969. Principles of Systematic Zoology.
New York, McGraw-Hill.
Moore, B. P. 1963. Studies on Australian Cara- bidac (Coleoptera) 3. The Psydrinae. Trans. R. Ent. Soc. London, 115: 277-290.
. 1965. Studies on Australian Carabidac
(Coleoptera) 4. The Pterostichinae. Trans. R. Ent. Soc. London, 117: 1-32.
National Academy of Sciences — National Re- search Council. 1969. Systematic Biology. Publ. 1692. Washington, National Acad. Sci.
Reichardt, H. 1967. A monographic revision of the American Galeritini. Arquivos Zool. (Sao Paulo), 15: 1-176.
Rivalier, E. 1950-1963. Demembrement du genre Cicindela L. Rev. franyaise d'Ent., 17 (1950): 217-244 ( Palearctic ) ; 21 (1954): 249-268 (American); 24 (1957): 312-342 (African); 28 (1961): 121-149 (Indo- malasian); 30 (1963): 30-48 (Australian).
Sharp, D. 1903. Coleoptera. II. Coleoptera Caraboidea. In Sharp, D. ( ed. ) , Fauna Ha- waiiensis ... 3, Part 3: 175-292, 2 pis. ["Flighdessness. Vestigial wings," pp. 177- 182]. Cambridge, England, Cambridge Uni- versity Press.
Slmpson, G. G. 1940. Mammals and land bridges. J. Washington Acad. Sci., 30: 137- 163.
. 1961. Historical zoogeography of Austra- lian mammals. Evolution, 15: 431-446.
Sloane, T. G. 1907. New Carabidac from Ger- man New Guinea and its dependencies; (and) Further Carabidac from . . . Deutsche Ent. Zeitschrift for 1907: 177-185, 467-474.
. 1923. The classification of the family
Carabidac. Trans. Ent. Soc. London for 1923: 234-250, and Suppl.: 250a-c.
Systeaiatic Zoology. 1952-datc. Society of Systematic Zool., place of publ. varies (now, Lawrence, Kansas).
SzENT-IvAXY, J. J. H. 1965. Notes on the verti- cal distribution of some beetles in New Guinea with new locality data and host plant records
The carabid beetles of New Guinea • Darlington
337
of some high altitude species. Trans. Papua and New Guinea Sci. Soc., 6: 20-36.
ToxoPEUs, L. J. 1940. Netherlands Indian- American Expedition to Netherlands New Guinea. (3rd Archbold Expedition to New Guinea 1938-39.) List of collecting stations. Treubia, 17: 277-279.
Umbgrove, J. H. F. 1949. Structural History of the East Indies. Cambridge, England, Cam- bridge University Press.
United States Am Force. Various recent dates. USAF Operational Navigation Charts: ONC M 12-14, N 14-15 (together covering all of New Guinea). St. Louis, Aeronautical Chart and Information Center, USAF.
United States Board on Geographic Names. 1955, 1956. Gazetteers No. 13, Indonesia, Netherlands New Guinea, and Portuguese Ti- mor (in 2 vols., section on Netherlands New Guinea [West N. G.] in Vol. II), and No. 29,
Southwest Pacific ( including separate sections
on Territory of New Guinea [NE N. G.] and
Territory of Papua). Washington, Superin- tendent of Documents. Webster's Geographical Dictionary. 1949.
Springfield, Mass., G. & C. Merriam Co. Wickler, W. 1968. Mimicry in Plants and
Animals. New York, McGraw-Hill. Wilson, E. O. 1958. Patchy distributions of ant
species in New Guinea rain forests. Psyche,
65: 26-38. . 1961. The nature of the taxon cycle in
the Melanesian ant fauna. American Nat., 95:
169-193. Zimmerman, E. C. 1948. Insects of Hawaii.
Vol. 1. Introduction. Honolulu, University of
Hawaii Press. Zoological Record. 1864-date. Insecta. Zool.
Soc. London and Commonwealth Inst. Ent.,
London.
IN LIEU OF INDEX
I had planned to add here an index covering all four parts of Tlie Carabid Beetles of Netv Guinea, but have decided against it. To make the inde.x would be very time-consuming; to print it would add ma- terially to the cost of publication; and it would probably be of only limited use. In lieu of an index, therefore, I offer the fol- lowing suggestions to users of this work.
The work should be arranged as sug- gested in the footnote on the first page of Part IV. The table of contents of Part IV will then come first, and will serve as a guide to the introduction and discussion of Part IV, which include references to important items in the shorter introductions
of the other parts. Parts I, II, and III will come next, and cover the Carabidae of New Guinea in taxonomic order. Each of these parts has a table of contents which lists the tribes and genera treated, and the order is, of course, that of the Coleopterorum Cata- logus (see Part IV, [4] ), which all persons working seriously on Carabidae of remote parts of the world must have. The Cata- logiis will serve as an indirect index or at least a guide to the taxonomic part of the present work. And the Zoological Record indexes the new species. The taxonomic supplement of Part IV, placed last, also follows the order of the Coleopterorum Catalogus. I suggest that users of my work make marginal notes in Parts I to III in- dicating the genera and species treated in this supplement.
uUetin OF THE
Museum of
A New Rhinoceros from the Late Miocene of Loperot, Turkana District, Kenya
D. A. HOOIJER
HARVARD UNIVERSITY VOLUME 142, NUMBER 3
CAMBRIDGE, MASSACHUSETTS, U.S.A. 18 OCTOBER 1971
PUBLICATIONS ISSUED
OR DISTRIBUTED BY THE
MUSEUM OF COMPARATIVE ZOOLOGY
HARVARD UNIVERSITY
Bulletin 1863-
Breviora 1952-
Memoirs 1864-1938
JoHNSONiA, Department of Mollusks, 1941-
OccASiONAL Papers on Mollusks, 1945-
Other Publications.
Bigelow, H. B, and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth.
Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth.
Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth.
Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth.
Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. ( Price list on request. )
Turner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae ( Mollusca: Bivalvia ) . $8.00 cloth.
Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth.
Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. )
Publications of the Boston Society of Natural History.
Authors preparing manuscripts for the Bulletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available fi'om Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A.
© The President and Fellows of Harvard College 1971.
A NEW RHINOCEROS FROM THE LATE MIOCENE OF LOPEROT, TURKANA DISTRICT, KENYA
D. A. HOOIJER'
CONTENTS
Introduction and acknowledgments 339
Skull and dentition of Chilotheridium patter-
soni gen. et sp. nov. 342
Chilotlicridiiim from East African sites other
than Loperot 358
Postcranial skeleton of Chilotheridium patter-
soni gen. et sp. nov. 365
Distinguishing characters of Chilotheriuvi,
Chilotheridium, and Dieeratherium 387
Literature Cited 389
Appendix 390
Abstract. Chilotheridium pattersoni, a new genus and species of Rhinocerotidae from the late Miocene, Vindobonian, Turkana Grit Formation of northwestern Kenya, is described and compared with its close relatives Chilotherium and Dieera- therium. The species also occurs at Ngorora. Fragments of Chilotheridium sp. from Bukwa II, Rusinga and Kirimun, of Aceratherium sp. or Dicerorhinus sp. from Kirimun and Ngorora, and of Brachijpotherium sp. from Ngorora are re- corded. Phalanges of a hippopotamid were mingled with tlie rhinoceros remains from the Turkana Grit; these constitute the earliest record of the family.
INTRODUCTION AND ACKNOWLEDGMENTS
In a paper on Miocene rhinocerose.s of East Africa (Hooijer, 1966), a single last upper molar from the Turkana Grit For- mation near Loperot, Turkana District, Kenya, collected in 1948 and preserved in the National Museum Centre for Prehistory
^ Rijksmuseum van Natuurlijke Historic, Leiden, Netherlands.
and Palaeontology in Nairobi, was referred to the genus Chilotherium Ringstrom. To the same genus, and likewise without spe- cific allocation, I referred two incomplete upper molars from Gumba and Wakondu on Rusinga Island. Teeth indistinguishable from those of Chilotherium have since been found at Bukwa II, Uganda (Walker, 1968), and at Ngorora, Kenya (collected by Dr. W. W. Bishop in 1968), early Miocene and early Pliocene, respectively. The Loperot rhinoceros has been cited as Chilotherium sp. by Leakey (1967: 15) and bv Maglio (1969: 2).
In the years 1964 and 1965 Professor Bryan Patterson led field parties of the Harvard Museum of Comparative Zoology to the Loperot area, which is at latitude 02° 20' N, and longitude 35° 50^ E, or 50 miles SSE of Lodwar and 45 miles SW of Lake Rudolf. The rhinoceroses collected were generously offered to me for description. The Loperot area has been geologically mapped by Joubert (1966), and three Potassium/Argon dates are available for the lava overlying the fossil-bearing Turkana Grit, 17.5 ± 0.9 m.y. for a sample five feet above the contact \\'ith the Turkana Grit at the rhinoceros quarry, 16.7 ± 0.8 m.y. for a sample approximately 200 feet above the contact with the Tur- kana Grit in the Auwerwer Hills, and 15.8 ± 1.2 m.y. from a basalt boulder in the Turkana Grit at the base of the Auwerwer
Bull. Mus. Comp. Zool, 142(3): 339-392, October, 1971
339
340 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
13 |
14 |
15 |
16 |
17 |
18 |
19 |
||||||
|
, |
X |
• ••* |
||||||||||||||||||||||
|
,• |
' |
•.X |
||||||||||||||||||||||
|
A |
•■:65B>'* |
"N |
Jx |
|||||||||||||||||||||
|
A" |
1 |
\65a: \ |
r |
i 1 t |
65C |
.' |
||||||||||||||||||
|
A' |
N |
-' |
y |
N |
1 1 1 |
\ \ \ * |
1 ! |
|||||||||||||||||
|
A |
^''" |
.y |
~^ |
\ |
J |
/slock ) |
t 1 1 1 1 |
/■ |
||||||||||||||||
|
B |
V |
o o |
\B J |
ock BL / |
B |
B / |
1 t. |
, • |
* |
|||||||||||||||
|
C |
X |
o 0 |
J |
1 |
||||||||||||||||||||
|
D |
V |
V |
X |
X |
X |
D, O y |
^°N |
D^O^ |
yo^ |
D X |
P o |
y |
||||||||||||
|
E |
x |
X |
X |
X |
X |
'*•— |
||||||||||||||||||
|
F |
X |
^ |
X |
X |
X |
|||||||||||||||||||
|
G |
V |
X |
X |
X |
V |
|||||||||||||||||||
|
H |
X |
X |
X |
X |
X |
|||||||||||||||||||
|
1 |
X |
V |
X |
X |
X |
|||||||||||||||||||
|
J |
V |
X |
X |
V |
V |
|||||||||||||||||||
|
K |
X |
)x( |
X |
X |
X |
X |
X |
|||||||||||||||||
|
L |
X |
X |
X |
X |
X |
X |
o o o weathered edge
XXX bone continuing beyond excavation
surface fragments found in squares so marked
' 1964 excavation 1965 excavation
K
Text-figure 1. Map of the Chilotheridium pattenoni quarry, 70-64 K. The squares measure two feet to a side. Redrawn from expedition field notes.
Hills, 40 feet below the contact with the lava (Patterson, personal communication). The fossil-bearing beds at Loperot may be taken as early Vindobonian, late Miocene. They are older than those at Fort Teman {ca. 14 m.y. ). The rhinoceros of Fort Ter- nan, Parodiceros mtikiiii Hooijer (1968b), is widely different from that of Loperot in being bicorn, without lower canines, and with brachyodont cheek teeth, but its metapodials, so far as available, show a remarkable resemblance to those from the Loperot locality, as will be remarked upon in the proper places in the present mono- graph. A new genus and species of rhinoc-
eros should ideally be based on skulls and teeth, as well as bones: I think we have such an ideal situation with the Loperot collection of Harvard. Rhinoceros remains' make up the great bulk of the material collected from the Turkana Grit by the museum expeditions. All of them were found at the same level in the formation and in essentially the same spot. This locality is three and one-eighth miles north of the Kamuthia waterhole near the head of a dry wash known as Laminkwais (see map in Joubert, 1966), and the level is 55 feet below the overlying basalts of the Tvbi series. The great majority of the
I
Rhinoceros from the Miocene of Kenya • Hooijer 341
rhinoceros remains come from one quarry in a bed of light brownish pink, jointed mudstone, in which the other specimens were also found. The state of preservation is poor: most of the teeth and bones are crushed and broken and the broken surfaces are not clean and sharp, being abraded as a result of postdepositional movements in the sediment. Nearly all of them were entirely dissociated and piled against or upon each other. The one ex- ception is a right pes from the quarry, which is represented by most of its ele- ments. Remains of at least twelve (and probably many more) individuals are represented, eight of them in the quarry. This mass occurrence with very little in the way of other animals is reminiscent of con- ditions at the famous Agate Springs Diceratherium quarry in the Miocene of Nebraska. Numbers of rhinoceroses evi- dently perished at these localities, perhaps along the courses of streams and rivers that were drying up during a prolonged dry season, the bones being subsequently swept by floods into a catchment area. Professor Patterson infomis me that the Chilothcrid- ium quarry was not exhausted when exca- vation of it was stopped in 1965. Parties working the area in the future should be able to collect additional material there.
The associated fauna of the Turkana Grit has as yet been mentioned only in part. Maglio (1969) records a tusk fragment of a shovel-tusked gomphothere, a very early member of the group, which suggests that Africa may have been the continent of origin of the amebelodontines. A similar conclusion may be drawn as to the hip- popotamids: serendipitously, during the study of the Loperot collection it was found that there are a number of phalanges in 68-64K and 70-64K not or hardly dis- tinguishable from those of the modern Hippopotamus' amphihiiis: As the oldest remains of hippopotamids known to date are from the early Pliocene (Pontian) of Sicily and Spain (Hooijer, 1946; Aguirre, 1963), the Loperot hippopotamus is the
earliest in the world. Maglio (1969) cites as elements of the Loperot fauna Deino- therium hohleiji Andrews, CMotherium sp. (now ChUotheridium), Brachijodus (?) sp., Dorcathcrium cf. pigofti Whitworth, and a hyracoid aff. Prohyrax.
As I was studying the collection, it be- came increasingly evident that the cranial and postcranial skeletal remains of this rhinoceros differed rather markedly from those of the genus CMotherium, no matter how closely the dentition resembled that of this genus. In fact, had cranial and post- cranial material not been found in associ- ation with the teeth, the East African form of rhinoceros described in the present paper would still have been called Chilotherium. As the material other than dental cannot be placed in any genus of rhinoceroses at present known, the Loperot rhinoceros is here referred to a new genus and species, ChUotheridium pattersoni gen. et sp. nov. It has been necessary to use the original field numbers in this paper. Thus, 6&-64K means the sixty-eighth specimen or lot collected in Kenya by the 1964 expedition of the museum. The quarry bears the col- lective number 70-64K and combinations following this number, such as BB and A17, denote the position of a bone in the quarry (see Fig. 1). In addition, the var- ious elements, skull, mandible, scapula, humerus, etc., have been consecutively numbered for each kind. All specimens are the property of the National Museum of Kenya and will in due course receive the permanent catalogue numbers of that in- stitution.
I am greatly indebted to Professor Bryan Patterson for offering me the Loperot rhinoceros remains for study and report. I am likewise grateful to Dr. L. S. B. Leakey for allowing me to describe the Kirimun tusk of ChUotheridium, to Dr. W. W. Bishop for permission to record the Ngorora ChUotheridium, and to Dr. Alan Walker for sending me casts and data on the Bukwa II ChUotheridium. Professor Patterson's field work was supported by
342 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
National Science Foundation Grant No. G.P. 1188.
Family Rhinocerotidae Owen, 1845 Chilofheridium gen. nov.
Dia^nosi.s. Small single nasal horn in both sexes; premaxillaries weak, no upper I; frontals and parietals pneumatized; orbit not placed so near upper contour of skull as in Chilotherium; cranium and occiput rather narrow; parietal crests not widely separated; inferior squamosal processes not united below; symphysial portion of mandible narrow, slightly expanding an- teriorly. Cheek teeth fully hypsodont as in Chilotherium and with the same pattern: uppers with paracone style fading away basally and posterior portion of ectoloph flattened; protocone well set off by folds and flattened internally; anterior fold in metaloph, marking off hypocone; antecro- chet prominent basally, curving inward to medisinus entrance; crochet usually well developed, and crista weak or absent; metacone bulge at base in M'^ anterior cingulum strong, internal cingulum weak and usually fonning cusp at medisinus entrance. Lower canine subtriangular in cross section, depressed dorsoventrally, in- ternal edge sharpened by wear, outer lower edge rounded, and outer upper edge ridged. Scapula low and wide; limb and foot bones not much shortened; radius and ulna, and tibia and fibula not ankylosed; radius with cuneiform facet; lunar without facet for ulna; metacarpal V present, three-fifths the length of metacarpal IV; lateral meta- podials somewhat divergent posteriorly; femur with small third trochanter; cal- caneum without tibia facet; navicular nearlv rectangular; cuboid wider than hieh; metatarsal III with small cuboid facet.
Type species. Chilotheridium pattersoni sp. nov.
Chilotheridium pattersoni sp. nov.
Diapwsis. As for the genus. Type. Skull No. 2 described and figured in the present paper (70-64K, B12).
Ilypodi^m. The type and numerous other elements (see Appendix, p. 390).
Horizon and locality. Turkana Grit; vicinity of Loperot, Turkana district, Kenya.
Aii,e. Late Miocene, Vindobonian.
Name. The specific name is given in honor of Professor Bryan Patterson, who let me have the Loperot material for study.
SKULL AND DENTITION OF CHILOTHERIDIUM PATTERSONI GEN. ET SP. NOV.
Two skulls from the Loperot rhinoceros quarry, with most of the dentition, estab- lish the uniqueness of the rhinoceros from this site; they will be described in the following pages.
Loperot skull No. 1 (70-64K, C9-10), four views of which are given, (Pi. 2, figs. 1-3, PI. 3, fig. 1 ) is a much deformed speci- men that is broken into innumerable small pieces. Plaster has been applied wherever needed to hold the skull parts together, evidently in the position in which they were found. Most of the right side of the skull is concealed by a thick mass of plaster, ex- posing only part of the occiput (both occip- ital condyles are there, but too close together and displaced to the right of the median line of the skull), part of the temporal fossa, the nasal, and the premolars and molars, which lack their outer portions. Of the skull base we find the body of the sphenoid embedded in plaster and lying obliquely to the right.
The left side of skull No. 1 is better preserved; it is, however, much depressed because of crushing in the middle, and the top of the occiput is missing. The fronto- parietal crest does not meet its fellow on the right side but remains a few centi- meters distant from it. The postglenoid process is heavy, and does not unite with the posttympanic process below the ex- ternal auditory meatus. The glenoid cavity is partially restored with plaster. The zygo- matic arch is pressed downward and has been restored from fragments that do not
Rhinoceros from the Miocene of Kenya • Hooijer 343
Plate 1. Chilotheridium pattersoni. Skull No. 2 (70-64K, B12), type. Fig. 1, top view; fig. 2, left view; fig. 3, right view. X 0.25.
344 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
fit very well. The orbitotemporal fossa is so crushed that the position of the orbit cannot be made out. Because of crushing, the anterior frontonasal region of the skull lies much higher than the middle part of the skull, and holds most of the nasals, which show a rugose area for a horn. The nasals, about 55 mm wide and only 25 mm high at a point about 10 cm in front of the nasomaxillary' notch, suddenly expand verticalh' to a height of 43 mm, where there begins a rugose horn boss 60 mm long and 35 mm wide, \^'ith a weak median groove. The nasals diminish to a width of 48 mm and a height of 30 mm at the front end of the horn boss, and are broken off 1 cm in front of the boss. The ventral surface of the nasal bones is flat (Pi. 5, figs. 1-2).
The depth of the nasomaxillary notch is considerable (the portion of bone em- bedded in the plaster above the P- on the left side does not belong there). As seen on the right side the nasals are free for about 10 cm behind the horn boss, that is, to above the P^-M^ junction.
The dentition of skull No. 1, at least that on the left side, is rather well preserved, considering the state of preservation of the cranium. The right toothrow lacks P- and M"* entirely and the outer parts of P'^-M-. The inner columns of these teeth are nearly all broken.
P- is worn to a height of 17 mm from the crown base externally, and has medi- sinus as well as postsinus closed off as fossettes. Tlie entrance to the medisinus fonns an indentation. There is a very weak internal cingulum. Tlie ectoloph is regu- larly convex with no styles showing.
P'^, the worn crown of which is 28 mm high externally, has the same two fossettes, and a trace of a cingulum at the base of the internal indentation representing the entrance to the medisinus. On the ectoloph there is only one style, the paracone style, more distinct above than at the base of the crown.
P^, 45 mm high extemalK', as worn, has the antecrochet touching the metaloph, just
about to close off the medisinus, in which a weak crista and a bifurcated crochet are seen. The postsinus is still open behind as the level of the posterior cingulum has not yet been reached by wear. The internal cingulum is manifest as a weak ridge along the bases of proto- and metaloph, and at the medisinus entrance. On the ectoloph, the paracone style, again, is seen to flatten out basally. while there is no metacone style. At this stage of wear, the anterior and posterior protocone folds, and the an- terior h\'pocone fold, can be seen distinctly.
M^ about 40 mm high as worn at the ectoloph (part of it is plaster), is not very well preserved: most of the metaloph is missing. The crochet, however, is there; it is well de\eloped but does not block the medisinus. In the protoloph, the constric- tion of the protocone is very marked, and' the antecrochet can be seen distinctly. The internal cingulum is barely indicated.
M-, worn externally to a height of 60 mm, has the metaloph displaced upward and forward, making the medisinus too narrow. It has the same characters as M^, but shows in addition that the paracone style disappears in the basal part of the crown, which is depressed only between the roots.
M"' is unfortunately broken at the junc- tion of proto- and ectoloph; the protoloph is displaced somewhat inward, with the cleft filled \\'ith plaster, so that the antero- transverse diameter cannot be gi\'en. The top of the ecto-metaloph (outer surface) internal to the large crochet is broken off. The crown is worn to a height of 70 mm, and there has not been very much wear, as seen from the narrow worn edges of the lophs. The unworn crown of M' would not have been more than some 5-10 mm higher. As the basal length of the outer surface is 62 mm, this is a decidedly hypso- dont crown. At 50 mm above the base the, length of the outer surface still amounts to 52 mm.
The M'' of ^"Chilotherium spec." from Loperot described earlier (Hooijer, 1966:
Rhinoceros from the Miocene of Kenya • Hooijer 345
Plate 2. Chilotherid'wm pattersonl. Skull No. 1 (70-64K, C9-10). Fig. 1, top view; fig. 2, left view; fig. 3, right viev X 0.17.
346 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
150-152) is only a trifle smaller, and more worn, but resembles that in skull No. 1 very closely indeed. The paracone style, fading away basally; the internally flattened, con- stricted protocone; the basally prominent antecrochet (the medisinus base is broken and filled with plaster); the metacone bulging out basally; and, the posterior cingulum forming a point some 20 mm high, are all very much as in the 1948 Loperot specimen.
Loperot skull No. 2 (70-64K, B12) is better preserved than skull No. 1, and is the holotype of Chilotheridium pottersoni gen. et sp. nov. Four views of the specimen are given (PI. 1, figs. 1-3; PI. 3, fig. 2). Although this specimen, too, is broken into numerous small fragments held together by matrix, plastic, or plaster, there is not as much distortion. Most of the right side of the skull is there; the nasals and the pre- molar-bearing part of the palate are broken off but are preserved separately. On the left side the palate, zygomatic arch and occiput are missing, and the temporal fossa is pushed inward. This side of the skull is much fortified with plaster.
Seen from the right side, then (PI. 1, fig. 3), the dorsal surface of skull No. 2 is weakly concave anteroposteriorly and flat transversely, with no ti'ace of a horn boss on the frontals. The postorbital processes of the frontals are damaged, but the width over these can be given approximately. The two frontoparietal crests converge behind the orbit to a least distance of 25 mm, and then diverge into the temporal crests, of which only that on the right side is pre- served. The occiput is notched in the median line above, and projects backward slightly beyond the occipital condyle. The occipital surface, of which only the right half (without the paroccipital process) re- mains, has been restored with plaster just above the beginning of the depression for the nuchal ligament. The zygomatic arch bears a shght postorbital process, behind which it is heavily restored with plaster. As it is, the arch is much extended along
the fractures, and it ends below the glenoid cavity, which is distorted, too. The post- glenoid process is, however, well preserved, and does not unite \\'ith the posttympanic process but remains a few millimeters distant from it below the external auditory meatus. The anterior border of the orbit is placed above the anterior border of M-. Because of superficial damage the infra- orbital foramina cannot be located. The nasomaxillary notch extends backward to above the anterior border of M^ The nasals have broken off a few centimeters from the deepest point of the notch. Fortu- nately, however, there were many frag- ments of the nasal bones, and it has been possible to restore them; although they do not fit on to the skull, they doubtless be- long to the same individual.
The portion of the nasals preserved (PI. 4, figs. 2-3) is 14 cm long, and shows the weak median horn boss, 55 mm long and 35 mm wide, grooved in the middle. The height of the nasals from the top of the boss is 42 mm behind, and over 30 mm in front. Anterior to the horn boss the nasals form a projection about 45 mm long and 33 mm wide, bluntly pointed.
Tlie premolars (in the maxillary portion: PI. 4, fig. 1 ) and the molars are more worn than those in skull No. 1. Whether or not there was a persistent DM^ cannot be made out in this specimen. Very little is preserved of the premaxillaries, which seem rather weak and were in all probability edentulous.
P-, worn down to 8 mm from the crown base, shows only two small enamel pits of the medisinus and the postsinus, and a weak internal cingulum.
P-' shows the same two pits, and an inner cingulum forming a point at the indentation representing the entrance to the medisinus. Its crown is worn down to 15 mm from the base.
In P^, of which the outer portion is miss- ing, the cro\\'n is still 20 mm high internally. The deep grooves delimiting the protocone (which is split vertically, the cleft being
Rhinoceros from the Miocene of Kenya • Hooijcr 347
'late 3. Chilotheridium pat/ersonj. Fig. 1, LP--M", RP-M" of skull No. ;M' (part)-RM" of skull No. 2 (70-64K, B12), type, crown view. X 0.70.
(70-64K, C9-10), crown view. X 0.50. Fig. 2,
348 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
filled with matrix) are well shown, as is also the antecrochet next to it, which ex- tends across the medisinus and joins the metaloph, cutting off the medisinus as a fossette. There is no trace of a crista or a crochet. The postsinus is closed off, too. The inner cingulum is continuous and well de\ eloped; it forms a conspicuous ridge at the medisinus entrance.
There is an anterior-internal fragment of the j\r attached to the maxillary portion, showing neatly the anterior protocone fold. This portion belongs to the M^ in the skull, but it cannot be replaced because of dis- tortion of the bone. M^, the outer surface of which is restored \\ith plaster, is poorly preserved, having the metaloph with the crochet broken and distorted.
In M- the ectoloph (height as worn ca. 40 mm) is broken. Its structure is well shown: the strong antecrochet. the con- stricted protocone (split again, as in P^), as well as the crochet, which extends for- ward externallv of the antecrochet. There is no crista. The paracone style is weak, and fades out in the basal portion of the crown. The internal cingulum is con- tinuous. There is an anterior fold in the metaloph opposite the protocone.
M", worn to 55 mm above the base, has the portion of the outer surface internal to the crochet broken and displaced, so that the length of the outer surface cannot be given. The protocone is flattened internally and well marked off by folds; the ante- crochet is prominent basally and curves inward to the medisinus entrance. The outer surface is flattened especially toward the base, where the paracone style fades away. The metacone forais a bulge at the base, near the internal angle. The internal cingulum is present along the protocone, and, as a prominent cusp, at the medisinus entrance; it joins the posterior cingulum, which fomns a point 28 mm high.
Apart from the more developed cingula and the absence of a (weak) crista in all the teeth, there is no difference between
the dentition of skull No. 2 and that of skull No. 1.
There is further in the Loperot collection a right maxillaiy holding D\F, P-^ DM^ and M^ (70-64K, 65B), representing a third individual (PI. 7, fig. 3). The an- teriormost tooth in this specimen is small, much worn down, and subtriangular, evi- dently a persisting anterior milk molar, DM^ Its dimensions are ca. 25 mm an- teroposteriorly, and ca. 20 mm transversely. In the middle of its broken worn surface it shows the base of the medisinus.
P- is broken, and the anterior part of its ectoloph is displaced forward, flanking the crown of the DM^ It is 33 mm high exter- nally, and not much worn; the protocone constriction can be seen clearly, but the metaloph (in part restored with plaster) is badly preserved.
P'' is 42 mm high at the worn ectoloph, which is split vertically in the middle and distended along the fracture. A very small crista and a crochet are present, and the protocone constriction is very marked. The anterotransverse diameter of P"" is 41 mm (less than that in skulls 1 and 2: Table 2), and the posterior width cannot be taken, as the metaloph is incomplete internally.
DM^, the last milk molar, is rather worn but not broken: its greatest crown height is 25 mm. It shows all the characters of the first and second molars in skulls Nos. 1 and 2: the prominent antecrochet external to the constricted protocone, the anterior metaloph fold, the well-developed crochet, a trace of a crista, and tlie weak inner cingulum. The enamel is, of course, thinner, and the size less (anterotransverse 49 mm, posterotransverse 46 mm).
M^ in the maxillary fragment is broken and incomplete internally. The external height of the worn crown is just about 60 mm.
The left maxillary belonging to the same individual as the right (70-64K, 65B) has a l)roken P'', a DM' the ectoloph of which is displaced anteriorly but which is other-
Rhinoceros from the Miocene of Kenya • Hooijer 349
Plate 4. Chilotheridium pattersoni. Fig. 1, RP--M' (part) of skull No. 2 170-64K, B12), type, crown view. X 0.80. Figs. 2-3, nasals of some skull in righit and top views. X 0.70.
350 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 1. Measurements of the skull from Loperot (in mm)
Loperot No. 2
Chilothcrium China (RinRstrom, 1924)
Greatest length from occipital to tip of nasals
From occipital crest to front of orbit
Least distance between parietal crests
Width over postorbital processes of frontals
Distance from nasal notch to front of orbit
Widtlr of nasals at 3 cm from tip
Height of occiput from lower border of foramen magnum
Greatest width of upper portion of occiput
|
ca. |
520 |
ca. 445-ca. 520 |
|
360 |
290-322 |
|
|
25 |
45-63 |
|
|
ca. |
125 |
129-169 |
|
65 |
65-78 |
|
|
32 |
35-52 |
|
|
ca. |
190 |
160-ca. 205 |
|
ca. |
115 |
ca. 135-175 |
wise a mirror image of RDM^, and M^"- both transversely compressed. The M- is unworn and the eotoloph height of this molar is exactly 71 mm by a greatest anteroposterior ectoloph length of 62 mm, demonstrating the marked hypsodonty of the Loperot form.
Among the surface finds at 70-64K, C9- 10, there are a number of tooth fragments making up a considerable part of an RM"', similar to those described above. Its worn ectoloph is 63 mm high.
There is also a nasal portion in 7(>-64K, A" 18, very much like those of skulls Nos. 1 and 2. The height of the nasals at the highest ( posterior ) portion of the horn boss is 52 mm, and the basal \\'idth at that level is 51 mm. The boss is shorter and wider than the others: length 50 mm, and width 38 mm. In front of it the nasals are only 34 mm high and wide; they taper to their blunt tip for a length of about 50 mm (Pi. 5, figs. .3-4).
Now that we have the skull as well as the upper dentition of the Loperot rhinoceros, it is easy to see that this form cannot be referred to Chilotherium as defined by Ringstrom (1924). Chilotherium h^shom- less, straight nasals, frontals and parietals not pneumatized, and the orbit placed just below the upper contour of the skull. The Loperot fonn, as we have seen, has a single, weak nasal horn boss, and the nasals are straight only as far as the ventral surface is concerned. The frontals and parietals are pneumatized: many air cells are seen
on the broken surfaces. In keeping with this condition, found in most rhinoceroses ex- cept in TeJeoccras, the orbit is not placed as high in the Loperot rhinoceros as in Chilotherium. As far as the homlessness of Chilotherium is concerned, Bohlin (1937: 92) points to an indistinct, rugose structure on the nasal tips of a skull of Chilothcrium hahereri var. laticeps from Shansi that may perhaps be interpreted as a horn boss. Ringstrom also states in his diagnosis of Chilotherium that the frontal region is depressed, but this is not a con- stant character among the Chilotherium species. Among the Chinese Pontian Chilo- therium species there is one, Chilothcrium planifrons Ringstrom (1924: 47), in which the frontal region is flat, not depressed. The parietal crests are farther apart in the Chinese chilotheres than in the Loperot fonn, and the occiput is wider above (see Table 1 ) . The premaxillaries of the Loperot rhinoceros are rather weak, and there are no remains of upper tusks in the collection, so that they were apparently edentulous, as is also the case in Chilothcrium.
The Loperot skidls agree with those of'| the Chinese Chilotherium in the small dis- tance between, and the position relative to the molars of, the orbit and nasomaxil- lary notch. Further they agree with Chilo- therium in their separation of the inferior squamosal processes, and, above all, in the details of their dental structure, such as the hypsodonty combined with flattening of the ectolophs, the marked constriction
Rhinoceros from the Miocene of Kenya • Hooijer 351
.(^^
'late 5. Chilotheridium patterson'i. Figs. 1-2, nasals of skull No. 1 (70-64K, C9-10); figs. 3-4, isolated nasals (70-64K). ?ight and top views. X 0.70.
352 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 2. Measurements of the upper dentitions from Loperot (in mm)
|
Skull No. 1 |
Skull No. 2 |
|||
|
P^ a. -p. |
31 |
|||
|
ti-., ant. |
31 |
ca. 35 |
||
|
tr., post. |
35 |
ca. 40 |
||
|
F, a. -p. |
33 |
— |
||
|
tr., ant. |
45 |
ca. 45 |
||
|
tr., post. |
48 |
|||
|
P', a. -p. |
42 |
|||
|
tr., ant. |
58 |
|||
|
tr., post. |
56 |
|||
|
M\ a. -p. |
||||
|
tr., ant. |
64 |
ca. 60 |
||
|
tr., post. |
||||
|
M-, a. -p. |
57 |
ca. 55 |
||
|
tr., ant. |
ca. 70 |
|||
|
tr., post. |
ca. 65 |
Loperot 1948 |
||
|
M'', a. -p. ( internally ) |
CO. 57 |
ca. 55 |
56 |
|
|
tr., ant. |
ca. 60 |
60 |
||
|
length outer |
surface |
62 |
61 |
of the protocone, the antecrochet develop- ment, the weakness of the crista, if any, and the metacone bulge in M".
The great length from occipital crest to front of orbit, as seen in Table 1, would seem to differentiate the Loperot skull from those recorded by Ringstrom. How- ever, the relative length in the Loperot fomi is not greater tlian that in all of the Chinese species. It is true that in two skulls of Chilothcrium anderssoni Ringstrom fully as long as the Loperot skull {ca. 510-c«. 520 mm in occipitonasal length), the dis- tance from occipital crest to front of orbit is only 310-322 mm, as opposed to 360 mm in Loperot skull No. 2. However, in the skull of Chilotherium pJanijrons the occip- itonasal length is ca. 445 mm, and the length from occipital crest to front of orbit is 300 mm (Ringstrom, 1924: 54), that is, two-thirds the occipitonasal length {ca. 0.67), equal to that in the Loperot skull {ca. 0.69).
There remain, therefore, several impor- tant cranial differences between the Loperot rhinoceros and the Chinese species of Clulotherhim. In the Loperot form a weak median nasal horn is present in three
out of three specimens, whereas in the chilotheres, nasal horns, if any at all, are the exception rather than the rule. The unpneumatized frontals and parietals, and the wider occiput and greater distance be- tween the parietal crests set Chilotherium off from the Loperot form.
It is of interest to observe that the Loperot rhinoceros, with respect to the narrowness of the skull, rather resembles the Chinese forms referred to the genus Diceratherium { palaeosinense Ringstrom, 1924, and tsaidamense Bohlin, 1937). The width of the upper portion of the occiput is 98-129 mm in Diceratherium, and the least width between the parietal crests 10-31 mm (Bohlin, 1937: 64-65), both ranges that include the observations on the Loperot form (cf. Table 1). However, the Loperot rhinoceros cannot be referred to DiccratJicriiim because it is not horn- less (? female), nor does it have a trans- verse pair of horns on its nasals (? male). The dentition of the Loperot form is fully as hypsodont as in Chilothcrium, and noti subhypsodont as in Diceratherium. In the latter genus, moreover, the inferior squa- mosal processes enclose the subaural chan-
Rhinoceros from the Miocene of Kenya • Hooijer 353
C\J
|
n |
CN |
|
>o |
|
|
6 |
o |
|
Z |
|
|
<u |
X |
|
J3 |
^ |
|
to |
|
|
T3 |
? |
|
c |
_Q> |
|
O |
|
|
E |
'> |
|
OJ |
|
|
0 |
|
|
3 |
|
|
lit: |
o |
|
o |
|
|
-C |
-D |
|
c |
|
|
Jc |
o |
|
O) |
a |
|
^ |
o |
|
LO" |
^ |
|
Os |
|
|
d) |
■<3- |
|
Li_ |
CN |
|
' |
|
|
CO |
CO |
|
rv. |
n |
|
o |
|
|
o |
|
|
X |
c |
»,"" i>^
a. o
vO CO
|
c |
D |
|
D |
>- |
|
U |
C |
|
<u |
|
|
<U |
^ |
|
$ |
^ |
|
o |
c |
|
D |
|
|
, |
E |
|
OJ |
|
|
^ |
|
|
CN |
^ |
|
1 |
c |
|
o |
|
|
i/i |
12 |
|
O) |
0) |
|
i-U |
o |
|
a |
|
|
c o |
<J |
|
1;2 |
.^ • |
|
QJ |
u |
|
0 a |
Q_ |
|
E |
E |
|
3 |
3 |
|
-D |
j6 |
|
c |
|
|
o |
QJ |
|
-£ |
-C |
|
Jo |
o |
|
-c |
-c |
|
U |
O |
|
^6 |
OJ |
|
CN |
|
|
<u |
o |
|
Jo a. |
X |
354 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
nel. This all apart from the fact that the reference of the Chinese forms to Diccra- therium is provisional and subject to revision (Ringstrom, 1924: 120; Bohlin, 1937: 98).
There are no mandibles in the Loperot collection associated ^^'ith the crania de- scribed, but there are three isolated ones, all broken in the symphysial region, as well as two halves and two isolated lower canines. The mandibles, like the skulls, are extremely fragmented and distorted; plaster and plastic have been used to hold the specimens together in one piece. Some of the specimens of the lower jaw have been crushed sideways, and the symphysis is so deformed that width measurements cannot be given. Only in two specimens is enough of the symphysis preserved to per- mit measurements to be taken.
Mandible No. 1, labelled 7{)^64K. has been crushed from above downward; the ascending rami lack the coronoid process, and the condyle has been pressed down into the fragmented ramus so that its height above the lower border of the angle of the mandible is only some 185 mm, or roughly two-thirds that in the other mandibles, in which the height has not been so visibly reduced. The right canine of the mandible is lost, but its alveolus remains, while the left is broken off just inside its alveolar border. The cross section seen is a trans- verse oval, approximately 22 by 17 mm in diameter. The two canines are placed quite laterally in the symphysis, and there are no incisors or traces of alveoli between them. The symphysis widens to the front, but exact measurements cannot be given. The premolars and molars are all broken. An inner view of the left ramus with the distorted symphysis is given in Plate 7, figure 2.
Mandible No. 2 (70-64K, 65) has the symphysis laterally compressed, and shows parts of the two canine alveoli, although it is impossible to measure them. The space between the two rami is only a centimeter or two, and the ascending portions, re-
stored from fragments, are very unequal: the right is a full 7 cm higher (from condyle to angle ) than the left. The cheek- teeth are characterized by the smallness of P2, the external groove between the lophids of P.-,-M,{ being well defined, and the ab- sence of an external cingulum.
Mandible No. 3 (70-6'4K, Bll) lacks the ascending ramus on the left side. The sym- physial region is deformed, but a few mea- surements can be given. The svimphysis widens slightly to the front, as it does in mandible No. 1 as well. The left ramus with the symphysis is presented in Plate 7, figure 1; the anterior two premolars from the right side are attached to this portion. An inner view of the right half of this mandible is given in Plate 6, figure 5.
Mandible No. 4 (70-64K, 65C) is quite complete on the right side, liut it lacks the condyle. Of the left half of the same speci- men only the portion bearing P.-> and P4 is preserved. Tlie forwardly expanding sym- physis is incomplete in front, but the least width, at P-, can be given.
Mandible No. 5 (70-64K, A18) consists of part of the left ramus, with Po and two complete molars.
The lower canine marked 7Q-64K, 65-? is well preserved (PI. 6, figs. 1-2). It is of the left side, and the crown, worn to a height of 44 mm, is subtriangular in section. The internal edge is very shaqo because of wear, the outer lower edge is rounded, and the upper outer edge marked by a longitu- dinal ridge. The base of the crown is slightly swollen lateroventrally. The dimen- sions at the crown base are 30 mm hori- zontally and 18 mm \'ertically. The enamel is very thin, especially on its upper surface. The root, a transverse oval 25 by 18 mm in cross section below the crown, becomes nearly round in section at the (broken) apex (15 by 14 mm); its length as preserved is 70 mm. This is just about the size of the smallest three lower canines of Chilo- therium andcrssoni as recorded by Ring- strom (1924: 37: 28-30 by 18-19 mm).
The other isolated lower canine ( 70-64K,
S
Rhinoceros from the Miocene of Kenya • Hooijer 355
ro
o. o X
I o
o
Z
c o
E
cn" 6>
o X
o
o
Z
(U CO
15 o
X
-^ 0)
^ X
^ i
356 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 3. Measurements of tlie mandible from Loperot (in mm)
|
No. of |
specimen |
||||
|
1 |
0 |
3 |
4 |
(Ringstr5ni, 1924) |
|
|
Heiglit from condyle to lower border of angle |
ca. 265 |
ca. 250 |
215-231 |
||
|
Lengtli from posterior border of C to that of angle |
ca. 530 |
ca. 520 |
ca. 490 |
ca. 480 |
415-485 |
|
Height of ramus at Ms |
ca. 90 |
ca. 95 |
ca. 100 |
76-ca. 90 |
|
|
Width of ramus at Mi |
ca. 40 |
ca. 40 |
40-50 |
||
|
Median length of symphysis |
ca. 150 |
ca. 130 |
104-137 |
||
|
Width of symphysis at P2 |
— |
ca. 80 |
98-128 |
||
|
Anterior width of symphysis |
ca. 100 |
— |
ca. 110 |
— |
130-cc. 190 |
|
Distance between C at alveoli |
ca. 45 |
ca. 60 |
75-93 |
||
|
Diastema C-P2 |
ca. 50 |
55-96 |
A16) is not as well preserved. It is from the right side, measures 30 by 15 mm at the crown base, and is, therefore, more depressed from above downward than the left canine. Tlie worn crown is 55 mm high, and there is a basal cingulum and a ridge along the dorsolateral edge. The inner edge of the cro\\ai is, again, sharp because of wear on the upper surface.
Table 4. Measurements of the lower dentitions from Loperot (in mm)
|
No. |
of specime |
n |
|||
|
1 |
2 |
3 |
4 |
5 |
|
|
P., a. -p. |
|
22 |
25 |
23 |
|
|
tr. |
15 |
14 |
|||
|
Po, a. -p. |
34 |
||||
|
tr., ant. |
— |
20 |
— |
18 |
|
|
tr., post. |
— |
21 |
21 |
— |
|
|
P., a. -p. |
— |
38 |
|||
|
tr., ant. |
26 |
— |
25 |
||
|
tr., post. |
30 |
26 |
— |
28 |
|
|
Ml, a. -p. |
— |
— |
46 |
||
|
tr., ant. |
ca. 26 |
28 |
— |
||
|
tr., post. |
29 |
26 |
|||
|
Mn, a. -p. |
58 |
57 |
53 |
||
|
tr., ant. |
29 |
28 |
28 |
— |
|
|
tr., post. |
32 |
31 |
32 |
30 |
— |
|
Ms, a. -p. |
57 |
60 |
53 |
— |
|
|
tr., ant. |
29 |
30 |
27 |
||
|
tr., post. |
29 |
28 |
28 |
27 |
It will be observed that, in keeping with its narrow cranium, the Loperot rhinoc- eros has a mandible that is narrower than that in the Chinese chilotheres. Fiu-ther, although the height and the lengtli of the jaw, as well as the symphysial length, may be greater in tlie Loperot form than in Chiloflierinm from China, there is no dif- ference in proportions. The ratio of the height to the length of the mandible in Nos. 2 and 3 (the only ones in which both of these dimensions can be given approxi- mately) is ca. 0.51; two mandibles of Chilo- therium anderssoni give 0.48 (218:445) and 0.52 (231:443) respectively (Ringstrom, 1924: 54). The length of the symphysis in mandibles No. 1 and 3 is 0.27-0.28 (ap- proximately) of the total length; in the Chinese chilotheres tliis ratio varies from 0.25 (in Chilotherium hahereri var, laticeps 104:415) to 0.29 (in Chilotherium anders- soni 128:443) (Ringstrom, 1924: 54). In the narrow s>anphysis the Loperot rhinoc- eros approaches the Chinese Dicera- therittm (distance between C at alveoli 24-45 mm; width of symphysis 75-95 mm: Bohlin, 1937: 70), but in these Chinese forms the symphysis does not widen to the front, P2 is relatively larger, and the coro- iioid process is stronger (cf. Ringstrom, 1924: 109-110; Bohlin, 1937: 71).
I
Rhinoceros from the Miocene of Kenya • Hooijer 357
o Z
o
3
^:^^m^
o
z
-a
D
CN
M
o
0)
a
358 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
CHILOTHERIDIUM FROM EAST AFRICAN SITES OTHER THAN LOPEROT
Rusinga Island: Giimba and Wakondu
Two incomplete upper molars in the National Museum Centre for Prehistory and Palaeontology, Nairobi, originating from Gumba and Wakondu, respectively, have been described as Chilofherium sp. (Hooijer, 1966: 151, pi. 6, figs. 10 and 11), an identification that in the light of the Loperot disco\'eries may now be changed to Clulothcridium sp. Whether the Rusinga molars are specificalh' the same as those from Loperot must remain uncertain. While most of the vertebrate fossils from Rusinga come from strata about 18 m.y. old, age estimations of the formations on the Gumba Peninsula must be deferred until the completion of the study by Van Couvering and Miller (1969).
Kirimun, Kenya
The tip of a lower left canine from Kiri- mun in the collection at the National Museum Centre for Prehistory and Palaeon- tology, Nairobi (no. 33, 1949), is heavily worn and rather flattened horizontally (PI. 6, figs 3-4). The vertical diameter at crown base is 25 mm, the horizontal diam- eter at least 40 mm. At the inner edge wear has produced a sharp angle. The enamel is thin but is present externally and ven- trally. The tip is broken; the crown length as far as preserved is 60 mm. Very little more than the crown is preserved, but the root seems to assume a round cross section. The shape of the crown is as in Chilo- theridium from Loperot, but the Kirimun specimen is larger; in size it is larger than all but one of the lower canines of Chilo- thcrium andcrssoni recorded by Ringstrom (1924: 37), which measures 47 by 26 mm.
Chilotheridium is not the only genus of rhinoceroses present at Kirimun. Among the bits of teeth from this site, collected during the Harvard Kenya Expedition of 1963 and sent to me for identification by Professor Bryan Patterson, there are part
of an M;., and part of a DM" or DM^ refer- able to either Aceratherium or Dicerorhimis. The posterior half of an RM, (39- 63K) from Kirimun, 27 mm wide, is worn to a height of 24 mm. Direct comparison with M:{ of Acemtherium ocutiwstratnm (unworn height 30 mm) shows the same marked cro\Miward taper of the sides of the postero-internal column and the same marked postero-external angle of the crown. In the Loperot M.? the crown is higher ( un- worn height 50 mm), and, consequently, the cro^^'n^^'ard taper is less; the postero- external crown angle is less angular, too. The antero-external portion of a DM'^ or LDM^ from Kirimun (25-63K), having thin enamel and showing the parastyle fold and paracone style, can be matched in the homologous teeth of Dicerorhimis leakeyi and Aceratherium acutirostratum from Rusinga described before (Hooijer, 1966: 134 and 142). Whether the second species of rhinocerotids from Kirimun represents Aceratherium or Dicerorhimis cannot be made out on the basis of this meagre ma- terial.
The Kirimun locality, at latitude 00° 43'N, and longitude 36° 54'E, is considered either late Miocene or early Pliocene by Leakey {in Bishop, 1967: 47).
Bukwa II, Uganda
Early in 1969 Dr. Alan Walker sent me casts of a number of teeth in the Uganda Museum, Kampala, excellently prepared by him and identified as Chilotherium sp. nov. (Walker, 1968, 1969). The specimens originate from the site Buk\\'a II on the northeast slopes of Mt. Elgon (Masaba), at latitude 01° 17'N, longitude 34° 47'E, and the capping lava has been dated at 22 m.v. The teeth, illustrated in Walker (1968), do agree with their homologues in the Loperot collection in all their diagnostic characters. There are teeth evidently of a single individual: a RP- incomplete inter- nally and a LP- lacking the posterior outer corner and an inner portion of the proto- loph; a RP^ \\i\h an external height of 25
I
Rhinoceros from the Miocene of Kenya • Hooijer 359
^ X
X
lO
O
X 3
5 D
0)
> R"
"5 <
£ ^'
a XT
— ^o . I
— o
— r\
Q
o
Z
o = if
CM O
d o
g 1
o -^
^ !5
^<£
-r, ^~ ° C so ^
o o
^ CM JZ
VO —
o o -
_ — ' c
o
Z
3 ^ X
~ >
D D
'>- O
CN o :^
I CO
^ X -
> ^
> NO
en
o
E '".
x d
2 Z
_. >
O Q.
D u O
3 " I
u
o
D
O
z :h
5 u:
360
Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 5. Measurements of teetli of Chilotheri- diitm from Buk\\a II (in mm)
|
F, a. -p. |
29 |
M', |
a. -p. |
52 |
Ps, a. -p. |
34 |
|
tr., ant. |
34 |
tr. |
, ant. |
67 |
tr., ant. |
20 |
|
tr., post. |
ca. 39 |
tr. |
, post. |
64 |
ti-., post. |
23 |
|
P', a. -p. |
40 |
M\ |
a. -p. |
61 |
P4, a. -p. |
37 |
|
tr., ant. |
58 |
tr. |
, ant. |
74 |
tr., ant. |
25 |
|
tr., post. |
56 |
tr. |
, post. |
63 |
tr., post. |
28 |
mm as worn, with the medisinus just closed off internally, very marked protocone folds as well as an anterior fold in the metaloph, an inner cingulum just a little less de- veloped than that in P^ of Loperot skull No. 2; an RM^ worn down to 20 mm from the base externally, and an RM- nearly 30 mm high at the ectoloph, both sho\\'ing an in- ternally flattened, constricted protocone and an anterior metaloph fold, the powerful antecrochet, and the flattened ectoloph be- hind the (only) style, the paracone style, very weak so near the base, and the internal cingulum barely indicated at protoloph and metaloph but present at the medisinus entrance as a cusp. The P-, as observed by Walker (1968: 155), has an anterior con- tact facet indicating the presence of a tooth, which must have been DM^ also demonstrable in Loperot skull No. 1. Both P^s (illustrated in the position of P'^ in Walker, 1968, plate) and the M^ show what appears to be a very weak external cingulum, almost invisible on the casts. Of the mandible there are RP3-4 and the posterior portion of LM3. Measurements of the Bukwa II teeth (Table 5) are veiy close indeed to those of the Loperot teeth (cf. Tables 2 and 4). Tlie M.^ fragment has a posterior width of 31 mm, slightly greater than that in Loperot specimens (27-29 mm). The Bukwa site may be taken as earlv Miocene, even very early at that (Walker, 1968: 155).
The Bukwa II material described above is indistinguishable from that of Loperot, but although the generic position is cer- tain (Chilothcridium) the specific identity of the two fonns is a matter of conjecture.
There is a second genus of rhinoceros at Bukwa II, identified by Walker (1968, 1969) as Dicerorhinus sp. (I have not seen this material). An incomplete right astrag- alus is all we have of postcranial material of rhinocerotids at Bukwa II, and a cast of it has been kindlv sent to me bv Dr. Walker. It lacks the medial ridge of the trochlea as well as the medial portion of the facet for the navicular, so that the medial height, the total width, and the widths of the trochlea and of the distal facets cannot be taken. Tlie lateral height of the Bukwa II astragalus is 63 mm. It is impossible, even by direct comparison with the astragali of Chilotheridium of Loperot (this paper, p. 377) and with those of Dicero- rhinus and Acerathcrium (Hooijer, 1966: 173 ) , to detemiine to which of these genera the Bukwa II bone should be referred. So, pending the discovery of postcranial ma- terial at Bukwa II that will prove to be identical with that of Loperot, the specific identity of the Bukwa II Chilotheridium with that of Loperot must remain un- certain.
Ngorora, Kenya
Late in 1968 Dr. W. W. Bishop entrusted to me the rhinoceros remains collected by him that year in the Ngorora Fonnation, Kenya, at latitude 00^ 53'N, longitude 35° 51'E, approximately 10 m.y. old, i.e., early Pliocene. The rhinoceros remains were all picked up from the surface and are rather fragmentary. However, there is material of Chilotheridium again in this lot, dental as well as postcranial, which justifies the inclusion of the Ngorora material in the present paper. f
To begin with, there is a right maxilla ^ with DM\ P- ' and M^ marked in the field' 2/13.S. The tooth crowns are much worn and damaged internally as well as exter- nally. No measurements can be given of either DM^ or P-. P"' is ca. 45 mm antero- tnmsversely, and P^ is ca. 55 mm wide anteriorly, and ca. 52 mm wide behind, close to the Loperot teeth (Table 2). In
Rhinoceros from the Miocene of Kenya • Hooijer
361
ro
OJ
|
j: |
|
|
03 |
|
|
■^ |
|
|
^ |
|
|
n |
|
|
_D> |
|
|
in |
|
|
CN |
|
|
CO |
|
|
o |
|
|
X |
|
|
^ |
|
|
» |
|
|
a |
|
|
'> |
|
|
.4_ |
|
|
C |
|
|
O |
|
|
L. |
|
|
*4- |
|
|
^ |
|
|
•o |
|
|
< |
|
|
iiT |
|
|
T |
|
|
f^ |
|
|
h^ |
|
|
— |
|
|
d |
|
|
Z |
|
|
^ |
|
|
D |
|
|
In |
CN |
|
CO |
|
|
_^ |
o |
|
X |
|
|
'C |
|
|
CN |
^ o |
|
O) |
■> |
|
4— |
|
|
iZ |
c |
|
o |
|
|
k. |
|
|
-«r |
**- |
|
CN |
- |
|
o |
>o |
|
X |
1 IT) |
|
^ |
OQ |
|
s |
|
|
_o> |
i^" |
|
'> |
|
|
*? |
|
|
c |
|
|
o |
Jv |
|
^ |
' — |
|
V |
6 |
|
N3 |
z |
|
i |
tn |
|
rx |
O |
|
' — " |
D. |
|
,^ |
.t_ |
|
_c |
|
|
O) |
|
|
6 |
^ CN
~ o
^■" X
§■5 !2 4)
D
a
u
O NO
-2 ^
c> ci Z
362 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
P^ the medisinus remains only as an enamel island, the postsinus having been worn off completely. In P^ both the medisinus, showing a weak crochet, and the postsinus remain as enamel islands. The posterior portion of the ectoloph is preserved in P', and it sho\\'s the flatness characteristic of Chilotheridium molars, there being no metacone style. The entrance to the medi- sinus has a cingular cusp slightly less de- veloped than that in Loperot skull No. 1; the internal cingulum is weakly developed along the protoloph, too. The posterior protocone fold as well as the anterior meta- loph fold can be seen distinctly; the ante- crochet extends all across the medisinus. The M^ is so much worn down and incom- plete externally and internally that no measurements can be given; it shows, how- ever, the anterior protocone fold (PI. 11, fig. 1).
In the lot labelled 2/13.S there are further a number of fragments of an upper dentition, some of which are more char- acteristic than others. The RP^ is the most complete specimen; it comprises most of the ectoloph and the external portion of the protoloph, and further, the inner por- tion of the metaloph not fitting on to the remainder of the crown. The ectoloph of P^ is worn to a height of 49 mm and measures 42 mm anteroposteriorly. The paracone style is there, but effaced near the crown base, and there is no metacone style, the posterior half of the outer surface being flat, just as in the P^ of Loperot skull No. 1 that is slighth' more worn down. The portion of the protoloph preserved bears a well-developed cingulum. The anterior metaloph or hypocone fold is seen in the detached fragment; the protocone is not preserved in this specimen.
To the same individual appear to belong the posterior portions of the ectolophs or RM^ and RM-, both showing the absence of the metacone style. This makes the posterior portion of the ectoloph flat or even concave apically. The antero-extemal portion of an unworn LM-' fortmiately is
present in the collection as well. The Ngorora cheek teeth show the hypsodonty by which Chilotheridium is characterized, notably the M- (PI. 11, fig. 4). This denti- tion as a whole is a little less worn down than that of Loperot skull No. 1, the (worn) heights of P^ M^ and M- (49, 43, and 68 mm, respectively) being somewhat greater than those in Loperot skull No. 1 (45, 40, and 60 mm, respectively). The portion of LM"^ lacks the base, so that the full height cannot be deteiTnined; it is broken off anteriorly along the cingulum, which is highest in the depression into ^^'hich fits the metastyle of M-, 20 mm be- low (rootward of) the unworn edge. In the slightly worn M'^ of Loperot skull No. 1 the anterior cingulum is about 15 mm below the worn edge so that some 5 mm may be added to get the full crown height, which may be, then, 75 mm. Among the smaller fragments in lot 2/13. S there is one showing the posterior protocone fold being curved inward toward the base, as is characteristic of Chilotheridium molars. The other bits presei-ved are not character- istic one way or the other. In the lot 2/ll.S there is an internal fragment of LM^, radier worn, with the characteristic antecrochet, limited by the posterior protocone fold curving inward basally. In lot 2/llA.S there is a protoloph portion of a left upper molar with the strong anterior cingulum as well as the anterior protocone fold, and the inner surface of the protocone shows the characteristic flattening.
Although at the moment of writing we do not have any better preserved uppei molars from Ngorora, the marked hypso- donty as seen in M-, the flattened posterioi ectoloph portions, the strong anterior cingu- lum, tlie inwardly curving posterior proto- cone folds, and the internal flattening ol the protocone are absolutely diagnostic ol Chilotheridium. In Aceratherium we find constricted protocones, too, but these are not flattened internally, and the molars are low crowned, the height of the outer sur- face of M^' (unworn) in Aceratheriun
Rhinoceros from the Miocene of Kenya • Hooijer 363
late 11. Chilotheridium paftersoni. Fig. 1, right maxillary wifh dm', P"-M' (2 13.S), crown view, X 0.75. Fig. 4, RM", losterior portion of ectoloph (2/13.5), outer view, X 0.60. 6rachypofher/um sp. Figs. 2-3, L dm:; (2/2. S), outer and crown iews, X 0.67. Acerafher/um c. q. Dicerorhinus sp. Fig. 5, left ramus with Pi-M.i (2,'ll.S), outer view, X 0.55. All from Jgorora, Kenya.
.364 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 6. Measiu-ements of lower teetli of Acera- therium and Diccrorhiniis (in mm)
|
Ngorora 2/11. S |
Ngorora 2/14. S |
Ngorora 2/1. SL |
Rusinga, Karungii, and Moniorot |
|
|
P., a. -p. tr. |
25 18 |
26 18 |
26-28 17-22 |
|
|
P^, tr., post. |
25 |
— |
22-24 |
|
|
P4, a. -p. tr., ant. tr., post. |
38 26 29 |
37 25 28 |
24 27 |
35-39 24-26 25-30 |
|
M\ a. -p. tr., ant. tr., post. |
28 .30 |
30 |
— |
41-44 26-29 28-31 |
|
M2, a. -p. tr., ant. tr., post. |
45 29 30 |
49 29 31 |
— |
41-50 26-32 27-31 |
|
Mo, a. -p. tr., ant. tr., post. |
47 28 28 |
28 |
44-53 27-31 26-29 |
aciitirostratum being only 49 mm by a length of 65 mm (Hooijer, 1963: 43).
Lower teeth in the Ngorora collection at present available do not show the hypso- donty by which Chilotheridium is char- acterized: in the lot 2/14.S an miwom posterior lophid of RM3 is only 30 mm high, against 50 mm in mandible No. 3 from Loperot. In lot 2/11. S there are a left ramus of the mandible with P4-M.S and a right ramus with Mo.a of the same in- dividual (PI. 11, fig. 5). All the teeth are worn, and there are external cingula in the groove between the lophids of P4, Mi and M2, which are short ridges placed 7-8 mm above the crowii bases. The external grooves are well marked as in Acemtheriimi or Dicewrhinus, unlike what we find in advanced brachypotheres, where the ex- ternal grooves are flattened out. In lot 2/14. S there are, all isolated but evidently belonging to one individual, P2 from both sides, RP3 incomplete in front, RP4, the posterior part of RMi, RMo, and the poste- rior portion of RM.;. In this lot there are no external cingula except in Mo, a few cusp- lets down near the base of the external groove. There are further a RP^ and a LP4
in lot 2/1. SL, the external part of a lower P in 2/1. SU, and the posterior portion of a left lower molar, either Mi or Mo, in lot 2/ll.A.S., the latter unworn and with a full height of 40 mm. In Loperot mandible No. 3 a worn M2 has just this height, and therefore was higher \\'hen un\\'om. Mea- surements (Table 6) show the Ngorora lower teeth to be similar to tliose of Loperot ( Table 4 ) in size, except for the lengths of Mo and M3, which are greater in the Loperot mandibles. The Ngorora teeth, of course, are also less high crowned than those of Loperot, as stated above. As shown in Table 6 the Ngorora teeth are well within the variation limits of the mandib- ular teeth of Aceratheriiim and Dicew- rhinus previously recorded from Rusinga. Karungu, and Moruorot (after Hooijer, 1966: 131, 133 and 141; 1968a: 234).
The lower teeth of Dicewrhinus leokeij, and Aceratherium acutirostratum are in- distinguishable, and there are no differ- ences between these and the Ngorora low- ers. The lower teeth from Loperot are more hypsodont, as we have seen, and M^ and Ms are somewhat longer than those from Ngorora.
There is a lower milk molar in the Ngorora collection, a left DM3 markec 2/2.S (PI. 11, figs. 2-3) characterized b> its thin enamel and the presence of a weal but continuous cingulum externally. Ex- ternal cingula may develop in Brachij- potherium lower molars (see Roman anc Viret, 1934: pi. X). Our Brachypothcriun lieinzelini from Congo, Kenya, and Uganda does not show a cingulum on its lowei molars so far as biown, and lower mill molars of this species have not yet beer found. In size the Ngorora DM3 ex ceeds that of Brochy potherium brachypm (Lartet) from La Grive-Saint-Alban; den tally there is no great difference betsveer this species and B. heinzelini (Hooijer, 1966: 144), and therefore the NgoroK milk molar would seem to be too large t( be referred to the latter species. In a col lection from Lothagam Hill, Kenya, shortl)
Rhinoceros from the Miocene of Kenya • Hooijer
365
Table 7. Measurements of DMs of Brachypo- therium sp. (in mm)
Ngorora
Lothagam Hill
DM., a. -p. tr., ant. tr., post.
46 21
24
43 21 23
to be published upon by Professor Patter- son and myself, and dating back approxi- mately 5 million years, there is a large species of Braclujpothcrium, and its DM3 is rather similar in dimensions and has a vveak external cingulum as well. Measure- ments are given in Table 7.
The postcranial material from Ngorora, scanty as it is, belongs to Chilotheiidiiim )nly. There are the proximal and distal oarts of a right radius and a phalanx, both abelled 2/11. S. Tlie radius is ca. 90 mm vvide proximally by a shaft width of ca. 46 nm, and a width of the distal facets of 80 nm. These dimensions are as in Chilo- ^heridium from Loperot (Table 8), and the presence of a small, lateral facet for the cuneiform unequivocally shows the Ngorora radius to belong to this genus. The phalanx is the first of a median digit, with a length 3f 33 mm and a proximal width of 46 mm, Df the same size and proportions as in the Loperot Chilotheiidiiim (p. 385).
Since we have both dental and post- cranial material from Ngorora that is in- listinguishable from that of Loperot, it vvould seem justified to accept not only generic but also specific identity of the hinoceros from the two localities.
^OSTCRANIAL SKELETON OF ZHILOTHERIDIUM PATTERSONI 3EN. ET SP. NOV.
There are two specimens of the atlas in he Loperot collection, one (70-64K, 65B) learly complete, the other (70-64K, CI) acking the dorsal arch and much distorted. The greatest width of the first specimen is '85 mm, the width across the occipital irticular facets 130 mm, the distance be-
tween the intervertebral foramina in the dorsal arch 93 mm, and the mid-ventral length (including the median posterior tubercle ca. 20 mm long and wide) ca. 65 mm. These data do not differ much from those of the atlas of Chilotlieriiim anders- soni (Ringstrom, 1924: 55; Bohhn, 1937: 72), but the atlas of Aceratherium acuti- rostrahim (Hooijer, 1966: 158) is not so very different either.
Of the scapula we have a series of five specimens, two of which are rather com- plete although the)^ are fragmented (70- 64K, A18, and 70-64K, BB, from the left and from the right side, respectively), and three specimens all from the right side (70-64K, BL, 70-64K, 65B, and 70-64K, BB ) lacking most of the borders and of the spine; the last specimen is a proximal por- tion only.
The thickened \ertebral or upper border is best preserved in scapula No. 2; it is highest at the point where the spina scapulae ends and is regularly convex. It forms an angle behind, at two-thirds of the height from the anterior border of the glenoid ca\ity, where it passes into the thin posterior border, \\'hich is conca\'e through- out. The anterior border of the scapula is likewise thin. It is straight for the most part in the reconstructed specimen No. 2 but Mas probably weakly convex in its upper three-fourths, the basal part being concave, forming the "neck," and becoming very thick where it ends in the massive tuber scapulae. Tlie spina scapulae, run- ning from the neck to the upper border, gives off a large, triangular, posteriorly directed tuber spinae, which extends just beyond the posterior border with its thick, rough extremity a little distance above the middle of the height of the bone. It is broken into fragments that are held to- gether with plastic and plaster and is pressed against the infraspinous fossa, but it originally extended outward as well as backward. Its anteroposterior extent is 130-140 mm (the upper portion of the
366
Bulletin Museum of Comparative Zoology, Vol. 142, \o. 3
Table S. Measurements of radius from Loperot (in mm)
Xo. of specimen
1
S
9
10
11
14
Median length 315
Proximal width 94
Middle width co. 50
Greatest distal width 95
Width distal facets 87
|
310 |
280 |
|
96 |
90 |
|
ca. 45 |
ca. 45 |
|
95 |
|
|
86 |
88 |
290 ______
90 — 92 86 — — -
ca. 45 — — — — — -
cfl. 100 — — — — —
ca. 90 82 — — 85 80 SO
tiiber spinae, preserved in scapula No. 1, is slighth- longer than that in specimen Xo. 2). The total height, from the anterior border of the glenoid ca\it>- to the end of the spine at the upper border, is 440 mm in No. 2; in No. 1 it is co. 470 mm, but this measurement is too great because of the spaces between the fragments, filled with plaster. Likewise, the anteroposterior diam- eter of the neck. o\"er 120 mm in No. 2, is too larcfe; in No. 1 this diameter is 100 mm, which must be very nearh" correct as there are no spaces bet\veen the bone fragments in this portion. The anteroposterior diam- eter over the glenoid cavit\- and tuber scapulae reads 120 mm in No. 1, and the anteroposterior and transverse diameters of the glenoid ca\"it>" are SO mm and TO mm in No. 2. The greatest anteroposterior diameter of the scapula is o\'er 240 mm, as seen in No. 2. No more exact measure- ments can be given.
The scapula of CJulothcrium from China. originalK" stated to be co. 400 mm high and 93 mm wide at the neck (Ringstrom. 1924: 60 61), as a more complete specimen showed, is 47S mm high and SS mm wide at the neck (BohliiL 1937: S0\ Our Loperot scapulae are less high, and w ider at the neck (see above), while the tuber spinae is more developed than in CJulo- thcrium, longer than the widUi of the neck. The scapula of Diccrorliinus Jcokcyi is hke- wise higher and slenderer than tliose of Loperot (Hooijer, 1966: 158), with a less de\eloped tuber spinae, only 75 mm long and not extending to the posterior margin
of the bone. A lateral \iew of Loperot scapula No. 2 is gi\en in Plate 8, figure 1. The proportions of the Loperot scapula are \"ery similar to those in the slightly smaller Diccrotlwriiim from China, of which the greatest height is 404 mm, the neck width 77-? 85 mm. and the greatest antero- posterior diameter 200-204 mm (Bohlin, 1937: SO and fig. 120).
Humeri in the Loperot collection are \er\' poorly preser\ed. There are three bones from the right side: 70-64K, A18, 70-64K, BB, and 70-64K. In No. 1 both^ ends are preserved, but no exact measure- ments can be gi\en because of the crushing; of the specimen. Number 2 lacks the proxi- mal end, and in No. 3 the distal end is shattered. There is, in addition. No. 4, the distal trochlea of a right humerus, 70-64K.
The humerus of the Chinese ChiJo- tlicrium (Ringstrom, 1924: 55 and 61) is shorter than that in fossil Diccrorhinus (Hooijer, 1966: 160), but less broad as well except in the greatest width at the distal end. which is proportionally greater in Chilotlicrium than in Diccrorliinus. It is unfortunate that the Loperot bones cannot be exactl\- measured; the length from caput to medial cond\le in No. 1 would have been about 340 mm, like that in Chilo- tJwrium (345-353 mm), but the greatest distal w idth would not have exceeded 130 mm. which is less than that in the ChiJo- therium humeri (150 mm) but proportion- all\- as great as that in D. oricntolis and D. primocL'us witli a length of 370-400 mm and a greatest distal width of 160-167 mm.
I
Rhinoceros from the Miocene of Kenya • Hooijer
367
Table 9. Measurements of ulna from Loperot (in mm I
No. of specimen
Greatest length
Length from proc. anconaeus ("beak")
Length of olecranon from "beak"
Width at semihmar notch
Greatest distal diameter
|
420 |
ca. 400 |
|
355 |
ca. 350 |
|
145 |
ca. 125 |
|
90 |
80 |
|
62 |
60 |
ca. 390
ca. 125
66
135 ca. 80
60
The width of the distal trochlea is 90 mm in Loperot humerus No. 4, not greater than that in ChilotJierium.
The following specimens of the radius are in the Loperot collection: 1) right radius, 70-64K, BB; 2) left radius, 70-64K, A16; 3) and 4) right and left radius, 70- 64K, C14; 5) right radius without distal end, 70-64K, BB; 6) left radius without distal end, 70-64K, A17; 7 ) right radius in three parts, 70-64K, BB; 8) proximal por- tion of right radius, 70-64K; 9) proximal portion of left radius, 7a-64K, ElO; 10) dis- tal portion of right radius, 70-64K; 11) distal portion of right radius, 70-64K, B16; 12) idem, 57-64K; 13) idem, 7a-64K, C12; 14) distal portion of left radius, 70-64K, BB; and, 15) distal fragment of right radius comprising only part of the facet for the scaphoid, 70-64K. Measurements are given in Table 8.
These radii, especially Nos. 3 and 4, agree very well with those of Chilotherium anderssoni, Dice rather him paJacosinense, and D. tsaiclamcn.se in length and proximal and distal width (Ringstrom, 1924: 55 and 113; Bohlin, 1937: 82)."The mid-shaft width is greater in ChilotJierium (55-57 mm) than in DiceratJierium (41-46 mm); in this respect the Loperot radii are nearer to DiceratJierium. All the specimens in which the lateral portion of the distal end is well preserved show^ a very small facet for the cuneiform, set off at an obtuse angle from that for the lunar ( Nos. 1-4, 7, 10-12, and 14). Such a facet, only 1 cm wide and ly-) to 2 cm anteroposteriorly, is found in
CJiiJotJierium as well as in the Chinese DiceratJierium and in the American dicera- theres (Ringstrom, 1924: 46 and 111; Bohlin, 1937: 82). It does not show in AceratJicrium or Dicerorhinus (which have longer radii: Hooijer, 1966: 161), and not in the recent rhinoceroses either. Radius No. 1 is shown in Plate 8, figure 2.
Tlie ulna is represented in the Loperot collection by the following specimens : 1 ) left ulna, 70^64K, A17; 2) right ulna broken at mid-shaft, 70-64K; 3) right ulna broken at mid-shaft, 70-64K, BB; 4) right ulna without distal end, 70-64K, BB?; 5) left ulna without distal end, 70-64K, C14; 6) left ulna, much broken, distal end missing, 70-64K, C14; 7) left ulna, olecranon and distal end missing, 70-64K, BB; 8) right ulna, olecranon and distal epiphyses miss- ing, 70-64K, A17; 9) distal portion of right ulna, 70-64K, BB; 10) distal portion and part of shaft of right ulna, 57-64K; and, 11) distal portion of left ulna, 70-64K, BB. Few of these bones can be measured ex- actly.
Entire ulnae are not a\ailable either in the Chinese CJiiJotJierium or in the Chinese DiceratJierium; Ringstrom (1924: 55) gives the length of the uhia of CJiiJotJierium an- derssoni as 370-390 mm, with a least \\ idth of 33 mm. Our most complete specimen ( No. 1 ) has a least width of ca. 45 mm; the specimen is figured in Plate 8, figure 3. It should be remarked that among the Loperot material there is no case of anky- losis of radius and ulna as we see in ChiJo- therium (Ringstrom, 1924: 56).
368
Bulletin Museum of Comparative Zoo/o^//, Vol. 142, No. 3
Table 10. Measurements of scaphoid from Loperot (in mm)
|
No. of |
specimen |
|||||
|
1 |
2 |
3 |
4 |
5 |
6 |
|
|
Posterior height Anterior height Proximal width Distal width Greatest distal diameter |
59 47 43 52 71 |
55 43 40 43 ca. 67 |
55 46 |
44 |
39 |
60 47 43 52 70 |
Six specimens of the scaplioid' are in the Loperot collection, Nos. 1-5 are from the right side (marked consecutively 7G-64K, A16; 70-64K; 7(>-64K, BB; 70-64K, Dll; and 70-64K, BB), and No. 6 is from the left side (70-64K). Number 3 is incomplete laterally, No. 4 lacks the posterior half, and No. 5 is incomplete distally. Numbers 1 and 6 may well have belonged to one and the same individual. Measurements are in Table 10.
The Loperot scaphoids differ from those of Aceratherium and Dicerorhinus (Hooijer, 1966: 162) in that the distal outward pro- jection, which bears on its distal surface the facet for the magnum, extends beyond the proximal radius facet, thus making the distal width greater than the proximal. This is most marked in Nos. 1 and 6, and less in No. 2. Further, the proximal projection be- hind the radius facet is much developed in Nos. 1 and 6, and not so in Nos. 2 and 3. In none of the specimens does the distal projection behind the trapezium facet ex- tend downward beyond this facet, as it
' In the nomenclature of the carpal bones British usage is adopted. The terms used are in the first column, below, and those used by Ring- strom (1924) and Bohlin (1937, 1946) appear in the second colunm.
scaphoid
lunar
cuneiform
pisiform
trapezium
trapezoid
magnum
unciform
Radiale Intermediiun Ulnare Pisiforme Carpale I Carpale II Carpale III Carpale \\ & V
does in Aceratherium and Dicerorhinus to a certain extent. In all the fossil specimens the posterior height exceeds the anterior height (the latter measured over the con- vex anterior portion of the radius facet and the ridge between the facets for the trape- zoid and the magnum ) , while these heights are nearly equal in Dicerorhinus sumatren- .sis. In this recent species the ratio of an- terior height (55 mm) to greatest distal diameter (79 mm) is 70, which shows that the scaphoid is relatively higher than that in Dicerorhinus rin<istroemi (see Hooijer, 1966: 162). in which this ratio is 66. In a- specimen of Dice rathe rium pa]aeosinense\ this ratio is 62, and in a specimen of Chilo- therium it is only 51 (Bohlin, 1946: 222). Our Loperot specimens give ca. 64-67 for this ratio, and therefore are not as low as the scaphoid in ChilotJierium but appear to agree better with Diceratherium and^ Dicerorhinus in this respect.
The lunar is represented in the Loperot collection by four specimens, but none of these is entire, unfortunately. All lack the posterior do\\'nward projection. In No. 1, a right lunar (70-64K), there are at least the greater parts of the upper and lower lateral facets for the cuneiform, of the distal facet for the unciform, and of the medial facets for the scaphoid ( anteriorly ) and the mag- num (posteriorly). The medial part of the proximal facet for the radius is broken off, and hence the proximal width cannot be given. Number 2, a right lunar (70-64K) is damaged anteriorly as well, but its proxi- mal width can be given. Number 3, a right lunar (7()-64K) consists merely of a lateral portion, and is injured posteriorly, lacking! most of the radius facet. Number 4, a left lunar (70-64K), is nearly complete proxi- mally but lacks the distal unciform facet. In all of these specimens there is no facet for the ulna, the proximal lateral facet for the ulna that we find in Aceratherium andj Dicerorhinus (Hooijer, 1966: 162). As stated above, in these genera the radius does not show a distal lateral facet for the cuneiform, the ulna articulating (for a veryi
Rhinoceros from the Miocene of Kenya • Hooijer
369
Tahle 11. Measurements of lunar from Loperot (in mm)
|
No. of |
specimen |
|||
|
1 |
2 |
3 |
4 |
|
|
Anterior height Proximal width |
38 |
ca. 35 36 |
30 + |
40 |
small part) with the lunar, cutting off the radius from contact with the cuneiform. The condition seen in the Loperot lunars is like that in Chilotherhim and Dicera- thcrium (Ringstrom, 1924: 56 and 111; Bohlin, 1937: 82).
Of the cuneiform we have twelve speci- mens in the Loperot collection; Nos. 1-7 are from the right side, and Nos. 8-12 from the left: 1) 57-64K; 2) 57-64K; 3) 70-64K, damaged proximally; 4) 68-64K, posterior half missing; 5) 6S-64K, incomplete proxi- mally and laterally; 6) 70-64K, incomplete distally; 7) 70-64K, A16, incomplete be- hind; 8) 68-64K; 9) 57-64K, incomplete behind; 10) 70-64K, BB; 11) 70-64K, in- complete behind; and, 12) 70-64K, incom- plete in front and distally.
The cuncifomi of CJiilotherium as de- scribed by Bohhn (1946: 224) does not appear to have very distinctive characters: the depressions and swellings on the non- articular surfaces vary individually, and so do the heights of the two facets for the lunar and the shape of the distal unciform facet. It is stated that the unciform facet is quadrangular rather than triangular be- cause the anterolateral side of the bone is so strongly convex, but this varies individu- ally, too: in No. 1 the facet is subtriangular, and in No. 2 it is rather quadrangular. In
Dicewrhinus the cuneiform varies in these respects also.
Three proximal ends of pisifonns, marked 57-64K, 70-64K, and 70-64K, display the two small facets for the ulna and the cunei- form, set at right angles to each other, or slightly less than 90°. Bohlin (1946: 226) states that the angle between these two facets on the pisiform is distinctly obtuse in Chilotherhim, whereas it is less than 90° in the American Diceratherhim. Since the main part is missing in the Loperot pisi- forms, the diameter over the two proximal facets may be given, which is 26 mm, 23 mm, and 24 mm in Nos. 1-3.
No trapezium has been recognized in the collection.
The trapezoid is represented in the Loperot collection by seven specimens, as follows: 1) right trapezoid, 70-64K, HU, incomplete posteriorly; 2) right trapezoid, posterior half only, 70-64K; 3) right trape- zoid, 70-64K, damaged at both ends; 4) left trapezoid, 57-64K; 5) left trapezoid, 70-64K, A16; 6) left trapezoid, 57-64K; and, 7) left trapezoid, 70-64K, incomplete behind. Ringstrom (1924: 57) mentions the trapezoid of CJiilotherium to be of the ordinary type; it is mentioned by Bohlin (1937: 82) to differ from that of the Chinese Diceratheriiim in the markedly obhque posterior surface. This Chilo- therium feature does not show in the Loperot trapezoids. As seen in Table 13 the anterior width and height are nearly equal in Nos. 1 and 7, and very different in Nos. 4-6; the two trapezoids of Dicera- theriiim (Bohlin, 1937: 84) vary in the same way. In Cluh)tlieriiim (2 specimens:
Table 12. Measurements of cimeiform from Loperot ( in nun )
|
No. of |
specimen |
||||||||||||
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
11 |
12 |
||
|
Anterior height |
45 |
42 |
|
47 |
43 |
45 |
42 |
43 |
42 |
|
|||
|
Distal width |
43 |
34 |
36 |
— |
34 |
— |
34 |
39 |
35 |
33 |
38 |
34 |
|
|
Proximal ant. post. |
diam. |
33 |
31 |
— |
— |
29 |
32 |
— |
34 |
— |
30 |
— |
— |
|
Greatest horizontal diam. |
51 |
44 |
49 |
— |
47 |
— |
48 |
— |
46 |
— |
370 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 13. Measurements of trapezoid from Lo- perot ( in mm )
|
N( |
}. of specimen |
|||||
|
1 |
2 |
4 |
5 |
6 |
7 |
|
|
Anterior height |
ca. 26 |
30 |
29 |
30 |
26 |
|
|
Anterior width |
28 |
— |
24 |
22 |
25 |
25 |
|
Greatest ant. post. |
||||||
|
diam. |
— |
41 |
39 |
46 |
— |
|
|
Posterior height |
27 |
28 |
27 |
33 |
||
|
Posterior width |
— |
18 |
16 |
16 |
17 |
Bohlin, 1937: 84) anterior width and height of the trapezoid are nearly the same, and in one specimen the width exceeds the height, as in Loperot trapezoid No. 1.
In the earher Miocene East African rhinoceros collection (Hooijer, 1966) the trapezoid was not represented; the trape- zoid of Dicewihiniis primaeviis Arambourg (1959: 67) has the same anterior height and anteroposterior diameter as Loperot No. 5, while the trapezoid of Dicerorhinus siimatrensis (same skeleton as that nsed in the 1966 paper) in anterior height and width (34 mm and 28 mm) as well as in posterior height and width (33 mm and 21 mm) exceeds all the fossil specimens, even thongh its anteroposterior diameter is only 38 mm.
The magnum is represented in the Loperot collection by no less than ten specimens, but none of these is entire. The posterior downward process is missing in all the specimens; it is separately preserved in Nos. 1 and 2 but cannot be fitted on to the remainder of the bone. The convex posterior facet for the lunar is broken off in all specimens except Nos. 1, 2 and 8, leaving just the front portion of the bone.
In No. 10 this portion is incomplete later- ally. Numbers 1-7 are from the right side, Nos. 8-10 from the left. They are marked as follows: 1) 70-64K, BB; 2) 70-64K, B14; 3) 70-64K; 4) 70-64K; 5) 57-64K; 6) 70-64K; 7) 70-64K; 8) 70^4K; 9) 68-64K; and, 10) 70-64K, Hll. Measurements are given in Table 14.
The magnum of CJnlotJierium is de- scribed by Ringstrom ( 1924 : 57 ) as large and wide with a small height; the magnum of Diceratherium palueosinense is wider still (45 mm against 36-37 mm) by nearly the same height (23 mm against 20-22 mm), and that of Diceratherium tsai- damense is 36 mm wide and 26 mm high (Bohlin, 1937: 84). The ratio of height to width in the Loperot magnum series varies from 0.63 (in No. 1) to 0.75 (in No. 9); in D. tsaidamen,se and D. palaeosinense these ratios are 0.72, and 0.51, respectively, and in two ChiJotherium specimens these values are 0.61 and 0.54. Tlius, the Chilotherium magnum appears to be relatively wider than those from Loperot, and so is that of Diceratherium palaeosinense, while D. tsaidamense resembles the Loperot bones in relative height of the magnum. It is further worthy of note that two Rusinga magna (Hooijer, 1966: 164) are very close to that of D. tsaidamense in height (25-27 mm) and width (36-38 mm) but exceed it in greatest length (75-82 mm against only 67 mm ) .
Twelve spcK'imens of the unciform arc m the Loperot collection; all except No. 1 and No. 9 lack the posterior process. Numbers 2, 4 and 11 are incomplete laterally, and No. 8 is merely a mediodistal fragment. Numbers 1-8 are from the right side, and
Table 14. Measurements of magnum from Loperot (in mm)
|
No. of |
specimen |
|||||||||
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
|
|
Greatest anterior height |
31 |
32 |
32 |
33 |
32 |
33 |
30 |
30 |
33 |
32 |
|
Greatest anterior width |
49 |
44 |
48 |
47 |
47 |
45 |
44 |
44 |
44 |
— |
|
Proximal ant. post. diam. |
59 |
59 |
— |
- |
57 |
— |
Rhinoceros from the Miocene of Kenya • Hooijer 371
Table 15. Measurements of unciform from Loperot ( in mm ]
|
No. of |
specimen |
||||||||||
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
9 |
10 |
11 |
12 |
|
|
Greatest anterior height |
47 |
45 |
38 |
43 |
43 |
42 |
42 |
42 |
45 |
43 |
39 |
|
Greatest width |
57 |
47 |
47 |
47 |
51 |
46 |
51 |
— |
54 |
||
|
Greatest ant. post, diameter |
83 |
— |
— |
— |
71 |
— |
— |
— |
Nos. 9-12 from the left. Numbers 5 and 9 almost certainly belong to a single individ- ual. The markings on the bones are as follows: 1) 70-64K, A17; 2) 57-64K; 3) 70-64K, J7; 4) 70-64K, A16, 17; 5) 70^4K; 6) 57-64K; 7) 57-64K; 8) 70-64K; 9) 70- 64K, A16; 10) 70-64K, Ell; 11) 70-64K; and, 12) 70-64K, DIO.
The measurements in Table 15 indicate the variability in the Loperot series. In comparing these data with those of Chilo- therium and Dicerorhmns as given by Bohlin (1946: 225), it should be remarked that the height anteriorly as given by this author and as shown in his illustrations is taken at right angles to the proximal facet for the lunar, which corresponds to what I use as greatest width. I take the greatest anterior height perpendicular to the straight portion of the distal surface articulating with metacarpal III and the medial por- tion of metacarpal IV, and from there to the top of the bone, which is the lateral end of the facet for the cuneiform. As can be seen from the front views of the bones in Bohlin (1946: 225, fig. 81), the greatest height is only ca. 35 mm in the largest, of the two Chilotheiium unciforms with a greatest width of 46 mm and a greatest anteroposterior diameter of 86 mm. The greatest height in an unciform of Dicero- rhiniis- riuiistroemi Arambourg (1959: 73; for D. orientali.s Ringstrom, non Schlosser) from the Chinese Pontian is ca. 65 mm by a greatest width of 78 mm and a greatest anteroposterior diameter of 108 mm.
The ratio of anterior height to anterior width is ca. 0.76 in the CJiilotlierium, and ca. 0.83 in the DicerorJiinus specimen, a difference of no significance; in our Loperot
series this height/width ratio varies from 0.72 in No. 12 to 0.92 in No. 5. The differ- ence between ChiJotlierium and Dicero- rhimis unciforms can be demonstrated in ratio of anterior width to greatest antero- posterior diameter; this was already shown by Bohlin (1946: 225, table). In Chilo- theriiim (two specimens) the ratio is 0.53 and 0.57, whereas in Dicewrhinus ring- stroemi this ratio is 0.71 and 0.72; in Recent Dicerorhinus sumatrensis (Hooijer, 1966: 164) the ratio is even 0.79. In their ratio of greatest width to greatest anteroposterior diameter the two entire Loperot unciforms are intermediate and even nearer to Dicero- rhinus' than to CJiilotlierium, the ratio being 0.65 in No. 9, and 0.69 in No. 1. It should finally be noted that the two Rusinga unci- forms previously recorded (Hooijer, 1966: 164) are within the variation limits of the Dicerorhinus specimens, their ratios being 0.72 and 0.75.
In the Loperot collection there are a number of specimens of all four meta- carpals, as follows:
Metacarpal II, 9 specimens: 1) right Mc. II, 70-64K, B13; 2) left Mc. II, 70-64K, A17; 3) right Mc. II, proximal portion, 70- 64K, BB; 4) right Mc. II, proximal portion, incomplete laterally, 70-64K; 5) right Mc. II, proximal portion, incomplete behind, 70-64K; 6) right Mc. II, proximal portion, incomplete laterally, 57-64K; 7) left Mc. II, proximal portion, incomplete laterally, 70-64K; 8) left Mc. II, proximal portion, 70-64K, B14; and, 9) left Mc. II, proximal portion, 70-64K.
Metacarpal III, 5 specimens: 1) right Mc. Ill, 70-64K; 2) right Mc. Ill, proximal portion, incomplete behind, 57-64K; 3)
372 Bulletin Museum of Comparative Zoolof^y, Vol. 142, No. 3
Table 16. Measurements of Metacarpals II-\^ from Loperot (in mm)
|
No. of specimen |
|||||
|
Me. II |
1 |
2 |
3 |
7 |
8 |
|
Median length |
123 |
123 |
|||
|
Proximal width |
41 |
42 |
43 |
43 |
47 |
|
Proximal ant. post, diameter |
36 |
37 |
43 |
42 |
39 |
|
Middle widtli |
33 |
ca. 33 |
36 |
— |
|
|
Middle ant. post, diameter |
17 |
15 |
18 |
18 |
|
|
Greatest distal width |
43 |
37 |
- — |
— |
|
|
Width of distal trochlea |
37 |
32 |
— |
||
|
Distal ant. post, diameter |
37 |
37 |
|||
|
Ratio middle width/length |
0.27 |
ca. 0.27 |
|||
|
No. of specimen |
|||||
|
Mc. Ill |
1 |
2 |
3 |
4 |
5 |
|
Median length |
140 |
143 |
154 |
||
|
Proximal width |
49 |
61 |
57 |
52 |
61 |
|
Proximal ant. post, diameter |
38 |
40 |
44 |
||
|
Middle width |
39 |
40 |
43 |
||
|
Middle ant. post, diameter |
17 |
— |
— |
18 |
21 |
|
Greatest distal width |
52 |
51 |
61 |
||
|
Width of distal trochlea |
45 |
45 |
49 |
||
|
Distal ant. post, diameter |
39 |
. — - |
— ■ |
38 |
40 |
|
Ratio middle width/length |
0.28 |
= |
0.28 |
0.28 |
|
|
No. of specimen |
|||||
|
Mc. IV |
1 |
2 |
3 |
4 |
5 |
|
Median length |
115 |
116 |
. |
126 |
118 |
|
Proximal width |
37 |
34 |
39 |
— |
35 |
|
Proximal ant. post, diameter |
44 |
41 |
43 |
42 |
|
|
Middle width |
26 |
— |
— |
31 |
23 |
|
Middle ant. post, diameter |
18 |
16 |
— |
18 |
17 |
|
Greatest distal width |
41 |
40 |
— |
46 |
39 |
|
Width of distal trochlea |
36 |
34 |
— |
38 |
34 |
|
Distal ant. post, diameter |
33 |
37 |
— • |
36 |
36 |
|
Ratio middle width/length |
0.23 |
^^"~ |
" |
0.25 |
0.19 |
|
No. of specimen |
|||||
|
Mc. V |
1 |
2 |
3 |
4 |
5 |
|
Median length |
71 |
71 |
70 |
82 |
|
|
Proximal width |
18 |
20 |
17 |
22 |
18 |
|
Proximal ant. post, diameter |
35 |
34 |
28 |
35 |
ca. 30 |
|
Middle width |
17 |
18 |
15 |
19 |
— |
|
Middle ant. post, diameter |
12 |
13 |
12 |
15 |
|
|
Greatest distal width |
28 |
28 |
26 |
32 |
29 |
|
Width of distal trochlea |
25 |
25 |
21 |
31 |
24 |
|
Distal ant. post, diameter |
25 |
24 |
23 |
25 |
23 |
|
Ratio middle width/length |
0.24 |
0.25 |
0.21 |
0.23 |
— |
Rhinoceros from the Miocene of Kenya • Hooijer 373
right Mc. Ill, proximal portion, incomplete behind, 70-64K, B14; 4) riaht Mc. Ill, 70-64K, A17; and, 5) left McC III, 70-64K, A17.
Metacarpal IV, 5 specimens: I) right Mc. IV, 70-64K, B14; 2) right Mc. IV, 70-64K, C14; 3) right Mc. IV, proximal portion, 57-64K; 4) left Mc. IV, damaged proximally, 70-64K, A17 (proximal half), and A16,'l7 (distal half); and, 5) left Mc.
IV, facet for Mc. V damaged, 70-64K, C14. Metacarpal V, 5 specimens: 1) right Mc.
V, 70-64K, B14; 2) left Mc. V, 70-64K, BE; 3) left Mc. V, 70-64K, C14; 4) left Mc. V, BB; and, 5) left Mc. V, portion at mid- shaft missing, 70-64K.
Of the Fort Ternan rhinoceros, Paradi- ceros mukirii (Hooijer, 1968b: 87), the only metacarpal available, Mc. Ill, is hardly distinguishable from its homologne in the Loperot Chilotheiidium pattersoni, show- ing that metapodials alone are unreliable for specific (or generic) differentiation (see Hooijer, 1966: 153/54, and above, p. 340).
The metacaipals of the Loperot rhinoc- eros, as shown in Table 16, are not at all as short and wide at mid-shaft as are those of the Chinese Cliilotherium; the measure- ments of a set of metacarpals of Chilo- therium aiulerssoni as given by Ringstrom (1924: 58) give the following data for the ratio middle width/length: Mc. II, 0.34; Mc. Ill, 0.34, and Mc. IV, 0.28. We shall find the same difference in the metatarsals, of which more material of ChiJotherium- is available for comparison. Moreover, in Cliilothcriwn the fifth metacarpal is re- duced to the same extent as in Dicer o^ bicornis, resembling a rounded sesamoid bone about 25 mm in diameter ( Ringstrom, 1924: 57). In a Recent skeleton of this species in the Leiden Museum ( Reg. No. 5738) the rudiment of Mc. V is 35 mm long and pointed distally; the proximal facet for the uncifonn is convex antero- posteriorly and measures 25 by 20 mm in diameter, that for Mc. IV is much smaller, 20 by 7 mm. The fifth mctacaipal of the
Loperot rhinoceros is small, but has a fully developed distal articular surface. Meta- carpal V No. 1 doubtless belonged to the same individual as Mc. IV No. 1, and its median length is three-fifths that of Mc. IV. When the interproximal facets are placed on each other, the Mc. V is seen to be directed backward, its shaft forming an angle of 45° with that of Mc. IV. The proximal end of Mc. V is much extended anteroposteriorly, and bears a large convex facet for the unciform that projects much behind the shaft. The proximal medial facet for Mc. IV is placed along the poste- rior half of the unciform facet, at right angles to it, and measures only 20 by 10 mm against 30 by 17 mm for the uncifonn facet. The anterior projection of the proxi- mal end of Mc. V is fonned by a protuber- ance below the uncifonn facet, which brings the bone on a level with Mc. IV. The shaft of Mc. V, then, diverges distally from that of Mc. IV at an angle of 45°. The distal end of Mc. V with the trochlea is turned outward (away from Mc. IV): the rather asymmetrical trochliea has its median posterior ridge set at an angle of 35° to the anteroposterior long axis of the proxi- mal end. One of the specimens of Mc. V (No. 4) is decidedly longer than the others; unfortunately this specimen cannot be as- sociated with any other metacarpal. In its width/length ratio this bone is within the limits of the three shorter Mc. V Nos. 1-3. Undoubtedly the small Mc. V in the Loperot rhinoceros carried some phalanges, and some of these have been found.
There are very few associations among the metacarpals, but Mc. II No. 2 belonged to the same individual as Mc. Ill No. 4, and when these bones are held together \\'ith their interproximal facets on each other, the Mc. II is seen to be not parallel to Mc. Ill but directed backward from it at an angle of 15-20°. In the same way, Mc. IV was probably directed backward relative to Mc. Ill, but there are no associated bones to prove this. The backward divergence of the lateral metacarpals relative to the
374 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 17. Measurements of metacarpal II in various genera ( in mm )
Br(ichy}}othcrium Chilotherium
Diceratherium
Chilotheridium
Diccrorhinus
|
Median length |
125 |
106 |
122-123 |
123 |
136 |
|
Proximal width |
52 |
(44) |
43-46 |
41-42 |
35 |
|
Proximal ant. post, diameter |
47 |
39 |
38-39 |
36-37 |
37 |
|
Middle width |
45 |
36 |
34 |
33 |
34 |
|
Middle ant. post, diameter |
25 |
13 |
13 |
15-17 |
18 |
|
Greatest distal width |
50 |
37 |
33-34 |
37-43 |
47 |
|
Ratio middle width/length |
0.36 |
0.34 |
0.28 |
0.27 |
0.25 |
median metacarpal is seen in Chilotherium as well (Ringstrom, 1924: 57). The meta- carpals discussed above are presented in Plate 9, figures 4-6.
Of the metacarpals of the Chinese Diceratherium only Mc. II is known by entire specimens (Bohlin, 19.37: 84). Bohlin gives the greatest length of two Mc. II of Diceratlierium tsaidamense as 127-129 mm, which makes for a median length of 122 or 123 mm (the greatest length of Mc. II Nos. 1 and 2, both 123 mm in median length, is 129-130 mm). Therefore, the Loperot Mc. II are closely comparable in length to those of D. tsaidamense. Bohlin (1937: 85) gives also the dimensions of an Mc. II of Onlothcrium: the median length of this specimen is only 106 mm by a middle width of 36 mm (Ringstrom, 1924: 58). In Table 17 the comparisons of Mc. II are extended to include Brachijpotherium heinzelini Hooijer (1966: 147) from Rusinga, the Mc. II of which, again closely com- parable to those of Diceratherium tsai- damense in median length, is much more massive at mid-shaft, surpassing Chilo- therium in this respect. Unfortunately, there are as yet no entire specimens of Mc. II of the African Aceratherium or Diccro- rhinus, but it is conceivable that these would not differ much in proportions from the Mc. II of Diccrorhinus sumatrensis (Hooijer, 1966: 166), the measurements of which are given in the last column of Table 17.
Although in the slenderness of the meta- carpal the Loperot rhinoceros (Chilo- tJieridium) is nearest to Diceratherium and
Diccrorhinus, and far removed from Chilo- therium and Brachijpotherium in the pe- culiar contact of the radius with the cuneiform, not seen in the Recent genera, Chilotheridium agrees only with Chilo- tJierium and Diceratherium. Whether the Chinese Diceratherium had an Mc. V is not known; in the American diceratheres it is represented only by a rudiment, as in the living species (Peterson, 1920: 445, pi. LXIII, fig. 1). A small Mc. V, about three-fifths the length of the adjoining Mc. IV, is found in various species of Aceratherium (see references in Hooijer, 1966: 153). In a typical Aceratherium^ lemanense of the Aquitanian (Roman, 1924: 52, figs. 23-24) Mc. V is 85 mm long, and Mc. IV 125 mm. In the old illustration of the manus of Aceratherium tetradactij- lum in Duvemoy (1853, pi. VII, fig. la), the fifth metacarpal has two phalanges as- signed to it, one as wide as the metacarpal itself and squarish, the terminal phalanx narrow and pointed.
Isolated phalanges abound in the Loperot collection, but, with the exception of those of the pes marked 70-64K, B15, 16, they cannot be assigned to any meta- podial in particular, and the only categories that can be made are phalanges I, II, and III of either a median or a lateral digit, i Some few of the isolated phalanges are decidedly smaller than those of digits II or IV, and these are the ones that I regard as belonging to metacarpal V. The speci- mens are marked as follows: 1) phalanx I and II (associated), 57-64K; 2) phalanx I, 57-64K; 3) phalanx I, 70-64K; 4)
Rhinoceros from the Miocene of Kenya • Hooijer 375
Table 18. Measurements of phalanges of Mc. V from Loperot (in mm)
No. of specimen
Phalanx I, length 19 19 21 21 19 18 18
Proximal width 25 24 23 23 21 21 20
Phalanx II, length 14 — — — — — 11
Proximal width 24 — — — — — 17
phalanx I, 70-64K, BB; 5) phalanx II, 70- 64K, D7; 6) phalanx I, 70-64K, A16; and, 7) phalanx I and II (associated), 70-64K. Measurements are given in Table 18.
Among the rib fragments there is one, marked 70-64K, 65A, preserving the verte- bral end and measuring 60 cm along the curve. The greatest width of the rib is ca. 50 mm. In the configuration of the head and tubercle it agrees best with the 5th to 7th rib, right side, in Recent skeletons.
The greater part of a left os coxae, marked 70-64K, A16, has the acetabulum and the shaft and most of the wing of the ilium. The specimen is broken into nu- merous small fragments that have been somewhat forced apart; the spaces between them are filled with plastic and plaster. The ilium is flattened to such an extent that the natural curvations of the gluteal and pelvic surfaces are almost gone. Pubis and ischium are broken off along the borders of the acetabulum, the diameter of which can be given only as 7-8 cm. The naturally three-sided shaft of the ilium is flattened, and measures about 9 cm in least width between the medial and the lateral borders.
The concave lateral border of the ilium, up the tuber coxae, is relatively well preserved. Of the anterior border only the convex and thickened lateral half of the iliac crest is there. The tuber sacrale is preserved, but the concave medial border, from there on to the acetabulum, is rather damaged. The diameter of the ilium from acetabulum to the middle of the iliac crest is about 50 cm, and the greatest diameter of the wing from tuber coxae to tuber sacrale is about 55 cm. Both measurements are at least 5 cm too large, considering the filled-in cracks of the bone.
The following specimens of the femur are in the Loperot collection: 1) left femur, 70-64K; 2) left femur, incomplete distally, 70-64K; 3) right femur, immature shaft only, 70-64K, BB; 4) distal epiphysis of right femur, possibly belonging to No. 3, 70-64K; 5) distal end of left femur, 7a-64K, A16; and, 6) proximal part of shaft of left femur, 70-64K, BB. Because of the crush- ing of the specimens very few measure- ments can be given (Table 19).
The most striking character of the Loperot femora is the small size of the third trochanter. This is shown in No. 1 (PI. 10, fig. 1) as well as in Nos. 2, 3, and 6. The trochanter tertius is placed just at the middle of the height, is not more than 50 mm vertically at base, and does not pro- ject outward and forward for more than 20 mm. The femur of Dicewrhiniis Icakcyi from Rusinga (Hooijer, 1966: 169, pi. 13, fig. 1), 545 mm in greatest length, has a trochanter tertius in the right (unfigured)
Table 19. Measurements of femur from Loperot (in mm)
|
No. of specimen |
|||||
|
1 |
2 |
3 |
4 |
6 |
|
|
Greatest length |
ca. 470 |
ca. 470 |
|||
|
Proximal width |
ca. 180 |
||||
|
Width over third trochanter |
90 |
90 |
75 |
ca. 80 |
|
|
Least width of shaft |
ca. 60 |
55 |
— |
||
|
Greatest distal width |
ca. 120 |
110 |
|||
|
Distal ant. post, diameter, |
|||||
|
medial side |
145 |
155 |
376 BiiUetin Museum of Comparative Zoology, Vol. 142, No. 3
Table 20.
Measurements of patella from Loperot ( in mm )
|
No |
of specimen |
|||
|
1 |
2 |
3 |
||
|
Length Width |
89 90 |
89 93 |
72 77 |
Table 2L Mcasiuements of tibia from Loperot ( in mm )
No. of specimen
specimen that measures 100 mm high vertically at its base and projects outward and forward, extending the width across this process to 130 mm in contrast to the least width of the shaft below it of 75 mm. In Recent D. suiiuiirensis the third tro- chanter is likewise well developed; a femur not very much shorter than the Loperot specimens (greatest length 423 mm) has a width over the third trochanter of 115 mm by a least shaft width of 56 mm. The femora of Chilotherium do show a large and prominent third trochanter; Ringstrom (1924: 62, pi. IX, fig. 4) figures a speci- men of Chilotherium from China only 430 mm in greatest length but with a width over the third trochanter of 128 mm, and a least width of shaft (estimated) of ca. 60 mm, while the Chilotherium femur from the Middle Siwaliks figured by Colbert (1935: 211, fig. 96) is likewise twice as wide over the third trochanter as its least shaft width.
It is in Diceratherium tsaidamen.se that we find a third trochanter on the femur that is just as small as that in the Loperot rhinoceros (Bohlin, 1937: 87, pi. IX, fig. 4); the width over the third trochanter is only one-half greater than the least shaft width. However, in the other Chinese Diceratherium, D. palacosineme, the femur is nearer to that of Chilotherium, and its third trochanter is even somewhat larger than that in the latter genus (Bohlin, 1937: 87).
Of the patella there are three entire specimens in the Loperot collection as well as parts of five others. Number 1 is a right patella, 70-^4K; No. 2 a right patella, 57- 64K; No. 3 a left patella, 70-64K, ElO;
|
1 |
2 |
3 |
4 |
||
|
Greatest length |
320 |
ca. 360 |
— |
||
|
Medial length |
285 |
ca. 320 |
— |
||
|
Proximal width |
118 |
— |
|||
|
Middl(> width |
45 |
ca. 55 |
— |
||
|
Distal width |
88 |
95 |
93 |
96 |
|
|
Distal ant. post, diamt |
>tei |
69 |
CO. 75 |
— |
71 |
whereas the fragments come from 70-64K, Dll (surface), bl2, and BB. The Loperot patellae are slightly wider than long ( Table 20), the Rusinga patellae longer than wide (Hooijer, 1966: 170), but this difference is most probably insignificant. A patella of Chilotherium amlerssoni (Ringstrom, 1924: 58) is 90 mm long and 87 mm wide; a specimen of Diceratherium tsaidamense (Bohlin, 1937: 88) measures 79 mm in. length.
The following specimens of the tibia are in the Loperot collection: 1) right tibia, 70-64K, A16; 2) right tibia, damaged at both ends, 70-64K; 3) right tibia, proximal part flattened, 70-64, BB; 4) distal end of right tibia, 57-64K; 5) proximal end of shaft of left tibia, 70-64K, BB?; and, 6) lateral distal fragment of left tibia, 70-64K, ElO. Measurements (Table 21) show that the most complete specimen is the smallest; yet it is longer than the tibia in Chilo- therium by the same middle and distal widths (Ringstrom, 1924: 58 and 63; length 275-cfl. 280 mm, middle width 47-48 mm, distal width 84-86 mm). The greatest length of the Middle Siwalik Chilotherium tibia is onlv 245 mm bv a middle width of 43 mm (Colbert, 1935- 212).
The difference between proximal and distal width is less in Chilotherium (105 and 92 mm, respectively) than in Diceratherium tsaidamense (95 and 69 mm, respectively); in the latter species (measurements taken from Bohlin, 1937: 89) the proximal width is one-third greater than the distal width,
Rhinoceros from the Miocene of Kenya • Hooijer 377
Table 22. Measurements of fibula from Loperot (in mm)
|
No |
. of specimen |
||||||
|
1 |
2 |
3 |
5 |
6 |
7 |
||
|
Length Greatest proximal diameter Mid-shaft diameters Greatest distal diameter |
290 44 21 X 14 48 |
ca. 300 45 |
255 40 19 X 14 41 |
45 |
38 |
38 |
as it is in Loperot tibia No. 1 (see PI. 10, fig. 2).
Although there are seven specimens of the fibula in the Loperot collection none of these appears to belong to any of the tibiae. Nor is there any case of ankylosis of these bones, in contradistinction to what we find in ChiJotheriiim (Ringstrom, 1924: 58). The specimens are as follows: 1) right fibula, 70-64K, 65C; 2) left fibula, 70-64K, 65C; 3) right fibula, 70-64K, B15, 16; 4) proximal portion of right fibula, 70-64K; 5) distal end of left fibula, 70-64K; 6) dis- tal end of right fibula, 70-64K; and, 7) distal end of left fibula, 70-64K.
Bohlin (1937: 89) found a fibula of Diceratherium tsaidamense to be more rounded in section at mid-shaft ( 17 X 17 mm) than one of ChilotJwiium (24x14 mm). If this is a good distinguishing char- acter the Loperot bones are closer to Chilo- thcrium than to Diceratherium (Table 22). The best preserved Loperot fibula is figured in Plate 10, figure 3.
There is most of a right pes in the Lope- rot collection; all bones are marked 70-64K, B15, 16, and fit together so well that there
Tahle 23. Measurements of astragalus from Lo- perot ( in mm )
No. of specimen
(Lateral height 65 71
Medial height 65 —
Total width 89 — Ratio medial height/
total width 0.73 —
Trochlea width 70 75
•Width of distal facets 73 — —
|
— |
72 |
63 |
|
64 |
72 |
67 |
|
89 |
93 |
89 |
|
.72 |
0.77 |
0.75 |
|
67 |
76 |
71 |
|
— |
79 |
73 |
is no doubt as to their belonging to a single indi\idual. There are the astragalus, calcaneum, navicular, cuboid, ectocunei- form, and all three metatarsals with their three phalanges each except for the third phalanx of the fourth digit; there is even one sesamoid. Thus, the tarsals missing are the mesocuneiform and the entocunei- form, but of these there are several speci- mens of other individuals in the collection. The bones in this right pes (PI. 10, fig. 4) are all No. 1 in their series.
The series of Loperot astragaH is as follows: 1) right astragalus, 70-64K, B15, 16; 2) right astragalus, 70-64K; 3) right astragalus, 70-64K, A16, L7; 4) left astrag- alus, 70-64K, A18; and, 5) left astragalus, 70-64K, C12. Numbers 2-5 are incomplete distally. In addition there are seven frag- ments of right, and four fragments of left astragali; of these no measurements can be given.
In the Loperot astragali (Table 23) the ratio of medial height to total width ( 0.72- 0.77) is intermediate between that in B r achy pother ium (0.64-0.73) on the one hand and that in Dicerorhinus and Acera- theriinn (0.80-0.97) on the other (Hooijer, 1966: 148 and 173). The trochlea width is slightly greater than the medial height, as may be the case in Dicerorhinus and Acera- therium (Hooijer, 1966: 174); in Brachij- potherium the difference between these two measurements is greater. Ringstrom (1924: 58) mentions that in a large number of entire Chilotherium astragali the three calcaneum facets are separate, whereas in the astragalus of Diceratherium (Ring- strom, 1924: 111) the medial and the distal
378 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 24. Measurements of calcaneum from Loperot ( in mm )
|
No. of |
specimen |
|||||||||
|
1 |
2 |
3 |
4 |
5 |
6 |
7 |
8 |
9 |
10 |
|
|
Lateral height |
112 |
126 |
132 |
113 |
|
|
||||
|
Greatest width |
74 |
ca. 75 |
— |
— |
68 |
— |
— |
— |
||
|
Ratio width height |
0.66 |
ca. 0.57 |
— |
0.60 |
— |
— |
— |
— |
||
|
Ant. post. diam. cuboid facet |
42 |
— |
40 |
— |
— |
42 |
— |
— |
— |
|
|
Transverse diam. of same |
25 |
— |
22 |
— |
21 |
22 |
— |
— |
— |
|
|
Greatest diameter of tuber |
59 |
62 |
72 |
61 |
59 |
70 |
— |
59 |
ca. 60 |
51 |
|
Transverse diam. of same |
42 |
42 |
48 |
40 |
45 |
43 |
47 |
42 |
41 |
40 |
calcaneum facets are confluent. I do not consider this difference to be of any sig- nificance (cf. Hooijer, 1966: 174, footnote). As follows from Ringstrom's illustrations the astragalus of Diceratherium is relatively higher than that of Chilotherium; measure- ments are given only by Bohlin (1946: 228), and they show that in Diceratherium palaeosinense the medial height exceeds the trochlea width, whereas in Chilotherium the trochlea width slightly exceeds the medial height. Both conditions are found in Dicerorhinus and Aceratheriuin. The total width of the Chilotherium and Dicera- therium astragali is not recorded, but I have measured an astragalus of the Middle Siwalik Chilotherium recorded by Colbert ( 1935: 212) and found the medial height to be 60 mm and the total width 77 mm, giving a ratio of 0.78, very much as in the Loperot astragali and in those of Dicero- rhinus and Aceratherium. It is clear that not only are the Loperot astragali not as much shortened as in Brachypotherium but they are also not as long as in Dicero- rhinus and Acertherium on the whole; they are nearer to Chilotherium than to Dicera- therium in that the trochlea width slightly exceeds the medial height, admittedly a variable feature. The relative height of the Loperot astragali is more like that in Chilo- therium than in Diceratherium (cf. Ring- strom, 1924: text-figs. 40 and 71-72), the approximate medial height/total width ratio as taken from the figures being 0.75 in the former against 0.81 in the latter. The calcaneum is again well represented
in the collection from Loperot, although in the majority of the specimens the trans- verse process, the sustentaculum tali, has broken off, and the proximal portion with the cuboid facet is missing in one-half the number of specimens (Nos. 4 and 7-10). In No. 1 the sustentaculum tali is preserved separately and there has been some loss of substance so that it does not fit on to the remainder of the calcaneum, but the as- sociated astragalus fits the calcaneum so perfectly that the greatest width can never- theless be taken. The series is as follows: 1) right calcaneum, 70-64K, B15, 16; 2) right calcaneum, 70-64K, BL; 3) right cal- caneum, 70-64K, A18; 4) right calcaneum,, 70-64K, A16, 17; 5) right calcaneum, 68- 64K, tuber portion and proximal portion separate; 6) left calcaneum, 70-64K, BB; 7) left calcaneum, 70-64K, A16; 8) left cal- caneum, 70-64K, ElO; 9) left calcaneum, 70-64K, E12; and, 10) left calcaneum, 70-64K.
The calcaneum of Chilotherium is rather short and massive (Ringstrom, 1924: 58, pi. VIII, fig. 7), and has the three astragalar facets separate, whereas in Diceratherium palaeosinense the two lower astragalar facets are confluent. In the few Loperot calcanea in which this can be checked there is no fusion of the two lower facets for the astragalus (and neither is there any fusion of the two corresponding facets for the calcaneum on the astragali of Loperot) This feature is variable in Diceratherium (Bohlin, 1937: 89), and is evidently not ii very reliable character. In the Lopero)
Rhinoceros from the Miocene of Kenya • Hooijer
379
Table 25. Measurements of navicular from Lo- perot (in mm)
No. of specimen
1
Greatest anterior height
Total width
A.nt. post, diameter
22 20 19 ca. 19 49 50 ca. 50 — 61—63 —
astragali there is no trace of a facet for the tibia behind and lateral to the upper facet for the astragalus; this facet is mentioned by Bohlin as most characteristic of Chilo- '^herhim. There remains a slight difference in relative height of the calcaneum: in the Loperot calcanea (Table 24) the ratio vvidth/length is ca. 0.57-0.66, while in two ZMotherium calcanea this ratio is 0.67 and ).74, but in four specimens of Dicera- .heriiim tsaidamense and D. palueosinense he width/length ratio is 0.61-0.71 (cf. Bohlin, 1937: 90). The development of the ruber calcanei is too variable to be of any /alue for specific distinction. Thus, the Loperot calcanea are slightly longer than :hose in Chilotherium, but differ in not Having a facet for the tibia; on the whole :hey are nearer to Diceratherium from !^hina.
The navicular of the right pes from 70- 34K, B15, 16, is not complete; it has a cut m the anterior face and lacks a portion posterolaterally. The other naviculars are not complete either. The series is: 1) right navicular, 70-64K, B15, 16; 2) right navic- ular, incomplete posteromedially, 57-64K; 3) right navicular, lacking the postero- lateral portion, 70-64K; and, 4) left navic- ilar, all borders except the lateral ncomplete, 70-64K, Cll. In Chilotherium mclerssoni this bone (called Centrale) is vvider behind than in front; it has an obtuse mteromedial angle (Bohlin, 1937: 90, fig. 155). That of Diceratherium tsaidamense (Bohlin, 1937: 90, fig. 156) is not as wide behind and is more nearly rectangular (it should be noted that in the upper [proxi- mal] views of the navicular given by
Bohlin [1937: figs. 155 and 156] the an- terior side is above, and the medial to the right). Our Loperot naviculars (Table 25) are decidedly more rectangular than is that of Chilotherium (the width of the latter, given as 59 mm by Ringstrom [1924: 60], as Bohlin's figure 155 shows, is only 50 mm behind and 30 mm in front), and agree with the navicular in Diceratherium tsai- damense in that anteroposterior diameters are about one-fifth greater than the width (46 mm, and 38 mm); in Chilo- therium the anteroposterior diameter is very nearly equal to the (posterior) width (52 mm, and 53 mm: Bohlin, 1937: 90).
There are two facets for the cuboid laterally on the navicular, a small and low anterior one, and a larger posterior facet that is not vertical but oblique, facing downward and outward. Between these two facets there is a nonarticular groove or fossa. We find, of course, the correspond- ing facets on the cuboid, the posterior facet facing upward and inward. The latter facet is bordered below by a nearly vertical facet that articulates with the ectocunei- fonn, for which there is also an anterior medial facet on the cuboid, placed below the anterior navicular facet and separated from it by a nonarticular groove. A third element that articulates \\'ith the medial surface of the cuboid is metatarsal III, situated, of course, below the ectocunei- form: there is a very small but yet distinct facet proximally and anteriorly on the lateral surface of metatarsal III, placed between the large proximal ectocuneiform facet and the anterior of the two lateral metatarsal IV facets. On the cuboid itself this little facet is practically indistinguish- able; in the articulated pes, the cuboid facet on metatarsal III forms just a small downward extension of the cuboid facet on the ectocuneiform.
The relations of the contact facets be- tween cuboid on the one hand, and navic- ular, ectocuneiform, and metatarsal III on the other, described in the preceding para- graph, exist in the Loperot rhinoceros and
380
Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 26. Measurements of cuboid from Loperot (in mm)
|
No. of |
specimen |
|
|
1 |
2 |
3 4 |
|
Anterior lieight 32 Anterior width 41 Greatest ant. post, diameter 60 |
35 41 63 |
32 35 36 39 — 56 |
in Recent Diceros hicornis and Dicero- rhinus (Recent and fossil) as well. Ad- mittedly the facet for the third metatarsal on the cuboid is hardly discernible, but when the associated metatarsal III is ex- amined, we find the little cuboid facet, which truncates the edge between the ectocuneiform facet and the metatarsal IV facet on the middle metatarsal. In Diccro- rliinus stimatrensis as well as in both Dicerorhinus Icakeyi (Hooijer, 1966: pi. 13, fig. 3) and Diceros hicornis there is a con- tact between cuboid and metatarsal III as well as a contact between cuboid and ecto- cuneifonn. I am mentioning this specially because Ringstrom (1924: 59) states that in the Recent forms there is a distinct facet on the cuboid (Tarsale IV & V) for meta- tarsal III, but none or only a very indistinct one for the ectocuneiform (Tarsale III). This is not in accord with my own ob- servations; on the contrary, the ectocunei- fonn facets are much more clearly seen than the metatarsal III facets. This reverse situation is the one that obtains in Chilo- therium; Ringstrom states that in that genus the ectocuneiform facet on the cu- boid is large, whereas there is no metatarsal III facet on the cuboid. As far as I know this is the condition in the nonchilothere rhinoceroses as well, and no reliance can be placed on this for the distinction be- tween genera.
Ringstrom (1924: 59) further mentions that in Chilotlwrium the two proximal facets for astragalus and calcaneum meet at an obtuse angle in the middle of the cuboid and are separated by a ridge, a condition elsewhere found onlv in Tcleo-
cercis. The obtuse angle and a smooth ridge (not a very sharp one) is seen in the Recent forms as well as in the Loperot cuboids. Of these cuboids we have four specimens, as follows: 1) right cuboid, 70-64K, B15, 16; 2) right cuboid, 7()-64K, Dll; 3) right cuboid, incomplete behind, 7(>-64K; and, 4) right cuboid, idem, 70- 64K, A 16, 17. The measurements are given in Table 26.
All four of the Loperot cuboids are wider than high anteriorly; this is, however, much more marked in the cuboid of Chilo- therium anderssoni, with a height of 27 mm and a width of 46 mm ( Ringstrom, 1924: 60). The great width in the Loperot cuboids, it seems, is caused by the presence of a lateral outgrowth of bone that is wel separated from the proximal (calcaneum) and distal (metatarsal IV) facets. This out- growth is a very distinctive feature of the Loperot specimens, and does not show iri the cuboids of Dicerorhinus and Accra- therium (Hooijer, 1966: 176, pi. 13, figs. I and 6 ) . In the cuboids of these two generc anterior height and \\'idth are either nearl) equal, or the height exceeds the width. Thf cuboid of Chilotherium is seen to project laterally much beyond the fourth meta- tarsal (Ringstrom, 1924: pi. IX, fig. 3) it does not do so in the Loperot rhinocerof as the articulated pes ( Pi. 10, fig. 4 ) shows Unfortunately there is no cuboid of th( Chinese Diceratherium on record. |
Five ectocuneiform s are in the Lopero collection: 1) right ectocuneiform, 70-63K B15, 16; 2) right ectocuneiform, 70-64K 3) right ectocuneiform, 70-64K, A16; A\ left ectocuneifonn, 70-64K, incomplete medially; and, 5) left ectocuneiform, 70- 64K, oil. This bone, the cuboid facets o which have already been mentioned, ha: two medial proximal facets for metatarsa II, and a small, high-placed posterio medial facet for the mesocuneifonn. TIk width anteriorly is very nearly twice thi anterior height (Table 27), in which i contrasts with the ectocuneifonn of Chilo therium. with a width (44 mm) nearl;
I
I
Rhinoceros from the Miocene of Kenya • Hooijer
381
Table 27.
Measurements of ectocuneifonii from Loperot ( in mm )
Table 28. Measurements of mesocuneiform from Loperot (in mm)
No. of specimen
1
Anterior height 21 23 21 21
Anterior width 42 44 43 —
Ant. post, diameter 44 48 46 43
three times the height ( 15 mm ) ( Ring- strom, 1924: 60). The ectocuneiform of the Chinese Diceratherhim has not been de- scribed.
The mesocuneifonn, a bone missing in the right pes from 70-64K, B15, 16, fits on to metatarsal II and has a correspondingh' shaped distal facet, transversely convex an- teriorly, elongated anteroposteriorly, end- ing narrow behind. There are three speci- mens: 1) right mesocuneiform, 68-64K; 2) right mesocuneifonn, 70-64K, DU; and, 3) left mesocuneifonn, 70-64K. The bone has a facet for the ectocuneiform proxi- mally on the lateral side, and another one for the entocuneiform posteromedially. The latter facet is either limited to the proximal part and is then continuous with the ento- cuneiform facet on the navicular, or the facet on the mesocuneiform may extend along the full height and, in that case, it is continuous with both the facet on the navicular and that on metatarsal II. The First-mentioned condition is seen in meso- C'uneiforms 1 and 3, whereas the second condition obtains in mesocuneiform 2. The sntocuneifonn facets on mesocuneifonns 1 uid 3 differ much in size. Although No. 3 is larger than No. 1, the entocuneiform Facet is smaller in No. 3, in which it is confined to the proximal third of the height, than in No. 1, in which it occupies the proximal half of the height. The mea- surements (Table 28) indicate that the Loperot mesocuneifonns are not as wide "elative to their height as the mesocunei-
orm of Chilotherium, which has a height 3f 12 mm by a width of 23 mm ( Ringstrom, 1924: 60: Tarsale II). The difference is
ather small.
No. of specimen
1
23 Height 13 12
45 Width 21 20
48 Anteroposterior diameter 32 33
15 22 34
Of the entocuneiform we have three specimens in the Loperot collection: 1) right entocuneiform, 70-64K; 2) left ento- cuneiform, 70-64K, Dll; and, 3) left entocuneiform, 70-64K. The posterior tuberosity is missing in the last specimen. This bone, which is placed behind the mesocuneiform, has a large, nearly hori- zontal facet proximally for the navicular. At right angles to it (nearly vertical) is a small facet for the mesocuneifonn, which may, or may not, be continuous with the facet for metatarsal II. Ringstrom (1924: 59) and Bohlin (1937: 90), who refer to the entocuneiform as the large sesamoid bone, mention these three facets in Chilo- therium anderssoni and Diceratherium tsai- damense but do not mention whether the mesocuneifonn and metatarsal II facets are separate or united. In Loperot No. 1 these two facets are continuous, but in Nos. 2 and 3 the facets for mesocuneifonn and metatarsal II are separated by a non- articular fossa (among the mesocuneifonns treated above the same variation occurs. No. 2 showing the entocuneiform facet to be continuous with that on metatarsal II, Nos. 1 and 3 showing these to be separate). The proximal facet for the na\'icular is the largest of all facets, the facet for the meso- cuneiform is low and wide, and only in entocuneifonn No. 1 it is continuous with the vertical, narrow facet for metatarsal II. In Table 29 I give the measurements of the Loperot specimens as well as those of Chilotherium and Diceratherium of China; the anteroposterior diameter (width in the table of Bohlin, 1937: 90) is taken above, thus not including the posterior hook-
382 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
Table 29. Measurements of entocuneiform from Loperot (in mm'
|
No. |
of specimen |
Chilotheriitm |
||||
|
1 |
2 |
3 |
Diceratherium |
|||
|
Height Anteroposterior diameter Width |
42 33 18 |
43 31 17 |
31 29 14 |
50 32 15 |
40 26 18 |
shaped process, and the width (thickness ingless; lateral views of Loperot Nos. 1
in Bohlin's table) is transverse. and 2 are given in Plate 9, fignres 1-2.
The variability in the Loperot series is so Apart from the right Mt. II-IV of thft
great as to make the intergeneric differ- associated pes there are only two entir(»
ences in size and proportions appear mean- metatarsals, and, further, some proxima
Table 30. Measurements of metatarsals II-IV from Loperot (in mm)
|
No. of specimen |
Chilotherium |
||||
|
Mt. n |
1 |
2 |
3 |
Siwaliks |
|
|
Median length |
110 |
119 |
94 |
||
|
Proximal width |
28 |
31 |
ca. 24 |
ca. 23 |
|
|
Proximal ant. post, diameter |
39 |
ca. 33 |
|||
|
Middle width |
23 |
27 |
— |
23 |
|
|
Middle ant. post, diameter |
21 |
26 |
18 |
||
|
Greatest distal width |
35 |
42 |
ca. 30 |
||
|
Width of distal trochlea |
34 |
38 |
ca. 26 |
||
|
Distal ant. post, diameter |
38 |
43 |
32 |
||
|
Ratio middle widtli/length |
0.21 |
0.23 |
0.24 |
||
|
No. of specimen |
|||||
|
Mt. ni |
1 |
||||
|
Median length |
124 |
104 |
|||
|
Proximal width |
43 |
41 |
|||
|
Proximal ant. post, diameter |
40 |
ca. 40 |
|||
|
Middle width |
37 |
34 |
|||
|
Middle ant. post, diameter |
19 |
ca. 18 |
|||
|
Greatest distal width |
49 |
44 |
|||
|
Width of distal trochlea |
43 |
39 |
|||
|
Distal ant. post, diameter |
36 |
— |
|||
|
Ratio middle width/length |
0.30 |
0.33 1 |
|||
|
No. of specimen |
|||||
|
Mt. IV |
1 |
2 |
3 |
4 |
|
|
Median length |
109 |
107 |
|
87 |
|
|
Proximal width |
39 |
39 |
41 |
44 |
ca. 37 |
|
Proximal ant. post, diameter |
ca. 40 |
43 |
41 |
||
|
Middle width |
23 |
22 |
— |
24 |
|
|
Middle ant. post, diameter |
21 |
21 |
— |
16 |
|
|
Greatest disttU width |
33 |
34 |
— |
— |
|
|
Width of distal trochlea |
33 |
35 |
— |
|
ca. 28 |
|
Distal ant. post, diameter |
36 |
35 |
— |
— |
— |
|
Ratio middle width/length |
0.21 |
0.21 |
— |
0.28 |
Rhinoceros from the Miocene of Kenya • Hooijer
383
|
Table 31. Measurements of metatarsal IV |
in various |
species (in mm) |
||||
|
Chilotherium |
Diceratherium |
Diceratherium |
Chilotheridium |
Dicerorhinus |
||
|
anderssoni |
palaeosinense |
tsaidameiisc |
paftcrsoni |
leakeiji |
||
|
Median lengtli |
89-90 |
83 |
107-109 |
166 |
||
|
Proximal width |
36-37 |
34 |
29 |
39 |
44 |
|
|
Proximal ant. post, diametei |
39 |
41 |
31 |
ca. 40^3 |
46 |
|
|
Middle width |
25 |
29 |
20 |
22-23 |
29 |
|
|
Middle ant. post, diameter |
— |
|
21 |
|||
|
Greatest distal widtli |
— |
— |
33-34 |
38 |
||
|
Ratio middle width/length |
0.28 |
0.35 |
0.21 |
0.18 |
portions of metatarsals in the Loperot col- lection, as follows:
Metatarsal II, 3 specimens: 1) right Mt.
II, 70-64K, B15, 16; 2) right Mt. II, dam- aged proximally, 7(>-64K, BB; and, 3) right Mt. II, proximal portion only, 70-64K.
Metatarsal III, 1 specimen: 1) right Mt.
III, 70-64K, B15, 16.
Metatarsal IV, 5 specimens: 1) right Mt.
IV, damaged proximally, 70-64K, B15, 16; 2) left Mt. IV, 70-64K, 65B; 3) left Mt. IV, proximal portion, incomplete behind, 70-^4K, Dll; 4) left Mt. IV, proximal portion, incomplete medially, 70-64K, A16, 17; and, 5) right Mt. IV, proximal end, incomplete anteriorly and laterally, 70-64K, D12.
When the measurements and indices of the Loperot metatarsals are compared with those of Paradiceros miikini of Fort Ternan (Hooijer, 1968b: (S7), it is seen that the Loperot Mt. II is nearly identical with that of Fort Ternan, and that the single Loperot Mt. Ill is perfectly intermediate between the two Mt. Ill of Paradiceros mtikirii on record. I found the same to be true for Mc. III. Yet the two forms are widely different cranially and dentally (above, p. 340).
The metatarsals from Loperot are longer and relatively more slender than those of Chilotherium; the measurements in the last column of Table 30 are those of the pes from the Middle Siwaliks recorded bv Col- bert (1935: 212) and taken by me on a visit to the American Museum of Natural History in New York in September 1965. The difference in relative length is greatest
in metatarsal IV. The metatarsals of Chilo- therium anderssoni, the length and middle width of which are given by Ringstrom (1924: 60), are very similar to those of the Middle Siwalik Chilotherium: Mt. II, 24-25 mm; Mt. Ill, 36-37 mm, and Mt. IV, 27-29 mm. As already mentioned above, the Loperot Mt. HI has a small cuboid facet; according to Ringstrom (1924: 72) Chilotherium does not have a cuboid facet on its metatarsal HI.
According to measurements recorded by Bohlin (1937: 91), metatarsal IV of the Chinese Chilotherium (two specimens) has a middle width/length ratio of 0.28, but that of Diceratherium palaeosinense is in- complete but seems to be much slenderer. The measurements are given in Table 31, together with those of the Loperot Chilo- theridium and those of Dicerorhinus Icakei/i (Hooijer, 1966: 179), which has the slenderest shaft of all Mt. IV recorded here.
Metatarsal IV of the Loperot species is nearest to Dicerorhinus in relative shaft width (in Recent D. sumatrensis the ratio is 0.20). Tlie proximal facet (for the cuboid) is almost flat, as it is in Chilo- therium as well as in Diceratherium palaeosinense; in D. tsaidamense (accord- ing to the incomplete specimen referred to this species by Bohlin, 1937: 91, figs. 159 and 160) the cuboid facet is raised later- ally, and, further, the posterior of the two facets for metatarsal HI is placed lower than the anterior. In the Loperot form the posterior metatarsal II facet is placed slightly lower than the anterior, as in fossil
384 Bulletin Museum of Comparotive Zoolop,y, Vol. 142, No. 3
Table 32. A. Distal ends of median metapodials from Loperot (in mm)
Table 33. Measurements of posterior phalanges from Loperot ( in mm )
No.
Greatest width
Trochlea width
Ant. post, diameter
1 70-64K, ElO
2 70-64K, B16
4 70-64 K
5 70-64 K
6 68-64 K
7 70-64K, A 16, 17
8 7n-64K, Dll
B. Distal ends
1 57-64K
3 7()-64K
4 7n-64K, BB
5 70-64 K
6 70-64K
7 70-64K
8 57-64 K
9 70-64K
10 70-64K
1 1 70-64K
12 70-64K, A16, 17
13 70-64K
14 70-64K
15 70-64K, Dll
16 70-64K, Dll
59
47 ca. 54
52
of lateral 42 48 42 46
45 38
36 40 30 41
33
48 43 48
43
43
37
44
39 36
metapodials
37 38 38 35 ca. 30 38 37 32 34 35 31 35 35 31 33
37 39
36 35
38 36 39
41
38
and Recent Dicewrhinus. In the articulated pes (7(>-64K, B15, 16) there is only a slight posterior divergence of the lateral metatarsals relative to the median; this is more marked in Chilothciium ( Ringstrom, 1924: 60, pi. IX, fig. 3). What the position of the lateral metatarsals in the Chinese Diceratheriiim is I do not know.
As seen in the proximal views of Mt. IV of D. tsaidameme and Chilothciium (Bohlin, 1937: 91, figs. 159 and 161), the bone extends laterally beyond the cuboid facet, which is sharply marked off laterally. Such a collar of bone is also found lateral to the proximal cuboid facet in the Loperot specimens; it is particularly well developed in No. 3 (PI. 9, fig. 3), which is from the left side as are Bohlin's specimens. In the articulated pes this bone prominence is placed just below the lateral bone de- velopment on the cuboid, serving for at- tachment of ligaments. We do not find such a development in Recent Dicero-
Digit
n
ni
IV
Plialanx 1, length Proximal width Phalanx II, length Proximal widtli Phalanx III, length Greatest diameter
|
30 |
31 |
|
34 |
45 |
|
22 |
20 |
|
33 |
42 |
|
34 |
31 |
|
45+ |
70 |
30
32 18 31
rhinus, and in D. Icakciji it is present, but placed posteriorly rather than laterally.
Measurements of a number of distal ends of median as well as of lateral metapodials are given below.
The phalanges I-III of digits II and III, and phalanges I and II of digit IV of the right pes (70-64K, B15, 16) are available, and their measurements are given in Table 33.
The measurements of the first phalanx of digit III agree very well with those in Diceratherium tsaidamense (length 29 mm, width 45 mm: Bohlin, 1937: 86), and those of the first phalanx of digit II are the same as those in this species as well as in Chilotherium (length 30 mm, width 34 mm: Bohlin, 1937: 85). In Bmchypo- thcrium heinzelini the proximal phalanges of these digits are shorter and wider (length 28 mm, proximal width 55 mm in digit III, and length 28 mm, proximal width 43 mm in a lateral digit: Ilooijer, 1966: 149), while in Dicewrhinus Icakeyi the proximal phalanges of these digits are much longer (length 40 mm, proximal width 55 mm in digit III, and length 37 mm, proxi- mal width 40 mm in digit II: Ilooijer, 1966: 180).
There remain a fair number of isolated phalanges in the Loperot collection; whether they belong to the fore or to the hind foot is impossible to tell. These speci- mens are enumerated below.
Of the third phalanges of the median digit only one specimen is entire (70-64K. BB?). It is 31 mm high and the greatest
Rhinoceros from the Miocene of Kenya ♦ Hooijer
385
Table 34. Phalanx 1, median digit from Lopeiot ( in mm )
No. Length Proximal width
50 49 46
47 48 51 49 46 43 41 43 45 52 49 42
49 52 53 43 43 49
CO. 53 44
ca. 55 48 49
Table .34. (Continued) Phalanx 2, lateral digit
|
70-64K, ElO |
29 |
|
70-64K |
33 |
|
70-64 K |
30 |
|
7()-64K |
31 |
|
68-64K |
31 |
|
70-64K, BB |
29 |
|
68-64K |
31 |
|
70-64 K |
33 |
|
70-64 K, A 16, 17 |
33 |
|
70-64 K |
31 |
|
70^64K, A16, 17 |
31 |
|
68-64K |
31 |
|
57-64K |
28 |
|
57-64K |
30 |
|
70-64K |
29 |
|
Phalanx 2, |
median |
|
70-64K, E12 |
24 |
|
57-64K |
20 |
|
70-64K, A16 |
19 |
|
70-64K |
20 |
|
7()-64K |
21 |
|
70-64K |
19 |
|
70-64K |
21 |
|
70-64K, A16, 17 |
20 |
|
68-64K |
21 |
|
70-64K, EM |
21 |
|
57-64K |
23 |
Phalanx 1, lateral digit
Length
Proximal width
|
70-64K, ElO |
31 |
37 |
|
70-64K, A17 |
30 |
ca. 33 |
|
70-64K |
32 |
38 |
|
70-64K, A16 |
32 |
34 |
|
57-64 K |
30 |
40 |
|
70-64K, BB |
32 |
39 |
|
70-64 K |
32 |
36 |
|
70-64 K, HIO |
30 |
33 |
|
70-64K |
31 |
34 |
|
57-64K |
29 |
34 |
|
70-64K |
28 |
30 |
|
7()-64K |
33 |
41 |
|
70-64K |
30 |
32 |
|
70-64K |
30 |
35 |
|
70-64 K |
31 |
35 |
|
57-64K |
28 |
37 |
|
70-64 K, B14 |
30 |
33 |
|
70-64K, Dll |
30 |
37 |
|
70-64K |
29 |
33 |
1 70-64K 21
2 70-64K, Dll 20
3 70-64K, Dll 20
4 70-64K, D12 21
5 7()-64K 22
6 70-64K, A16, 17 21
7 70-64K, A16, 17 18
8 70-64K 28
9 .57-64K 19
10 70-64K, Hll 20
11 70-64K 20
12 70-64K 20
13 .57-64 K 21
14 70-64K, B14 18
15 .57-64K 19
16 68-64K 19
17 70-64K, A16, 17 21
18 70-64K 19
37 33 34 39 32 34 37 41 35 37 32 34 41 30 41 33 34 32
(transverse) diameter is 78 mm. Of the third phalanges of lateral digits there are only incomplete specimens varying in greatest length from 33 to 41 mm; the transverse diameter cannot be taken in any of the specimens. These terminal phalanges, which belong either to digit II or to digit IV, manus or pes, are as rough and porous, with many perforations on or near the distal border, as the end phalanges of digit III. I find this also in the living rhinoceroses. Ringstrom (1924: 63) found the terminal phalanx of a lateral digit of Clulotlwriiim to be much less rough on the surface, with few, small perforations showing, and states that this is probably because the lateral digits diverge backward and are functional only to a very slight extent.
To round off the account of the remains of the Loperot rhinoceros I have to mention the sesamoid bones. One, a proximal sesa- moid of the median digit, is associated with the right pes marked 70-64K, B15, 16; there are eleven more sesamoids of digit III (manus or pes), and there are twenty entire proximal sesamoids of lateral digits ( II or IV ) , as listed below.
386 Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
|
Table |
35. P |
'loximal sesamoids |
from Loperot |
||
|
(in mm) |
|||||
|
No. |
Length |
Width |
|||
|
median |
digit |
||||
|
1 |
70-64K, |
B15, |
16 |
38 |
19 |
|
2 |
57-64K |
46 |
24 |
||
|
3 |
70-64K, |
A16 |
38 |
18 |
|
|
4 |
70-64K |
45 |
22 |
||
|
5 |
57-64K |
37 |
19 |
||
|
6 |
70-64K |
43 |
22 |
||
|
7 |
70-64K |
45 |
24 |
||
|
8 |
57-64K |
48 |
22 |
||
|
9 |
7n-64K |
39 |
18 |
||
|
10 |
70-64K |
41 |
21 |
||
|
11 |
70-64K |
40 |
21 |
||
|
12 |
70-64K |
41 |
22 |
||
|
lateral |
digit |
||||
|
1 |
70^64K, |
Dll |
35 |
18 |
|
|
2 |
70-64K |
36 |
18 |
||
|
3 |
70-64K, |
A16 |
31 |
15 |
|
|
4 |
70-64K |
35 |
19 |
||
|
5 |
70-64K |
32 |
16 |
||
|
6 |
70-64K, |
Dll |
27 |
15 |
|
|
7 |
7()-64K |
34 |
17 |
||
|
8 |
70-64K |
27 |
16 |
||
|
9 |
7()-64K |
33 |
17 |
||
|
10 |
70-64K |
31 |
15 |
||
|
11 |
57-64K |
34 |
18 |
||
|
12 |
70-64K |
31 |
16 |
||
|
13 |
70-64 K |
34 |
17 |
||
|
14 |
70-64K |
29 |
15 |
||
|
15 |
70-64K |
29 |
14 |
||
|
16 |
70-64K |
30 |
16 |
||
|
17 |
70-64K |
32 |
15 |
||
|
18 |
70-64K |
29 |
16 |
||
|
19 |
70-64K |
32 |
17 |
||
|
20 |
70-64K |
34 |
16 |
Rhinoceros from the Miocene of Kenya • Hooijer
387
|
■p |
|||||||||||||||
|
1 |
c |
«, |
|||||||||||||
|
in |
o |
o |
H |
||||||||||||
|
r-t |
Ij |
>> |
•o |
3 |
|||||||||||
|
H fci |
CS |
a s |
tn M |
r-< |
o |
0 |
E |
||||||||
|
rt O |
■tJ |
o |
4J |
•M |
o |
•H C |
0) |
m |
•u |
0) |
|||||
|
UJ |
a; |
tn |
3 |
0 |
■o |
rH rH |
U) -H |
-o |
Q. |
■H |
<H x: |
||||
|
CO ^-\ |
■H |
a |
c |
(1) |
CO 0) |
>. c |
•H |
>> |
s |
u |
|||||
|
B |
C«0 |
*^ |
■o |
T3 |
■H |
5 |
-p |
Cfi 12 |
s: Q) ■*-> |
> |
x: |
• 3 |
|||
|
3 |
^-^ |
« |
Q) |
<u |
o |
0 |
(fl |
to |
O |
a-o c |
a |
T3 |
in B |
||
|
•H |
nH |
a |
N |
o |
o |
^ |
-»-> |
u |
e -a |
e H o |
o |
■o |
3 |
C |
3 |
|
U |
« tfl |
•H |
rt x; |
0 |
^ |
tn |
a |
CO 0) |
>. » 'h |
tfi |
(U |
tn |
CO |
•H JJ |
|
|
0) |
c c ^^ |
■o |
■t-> |
rH M |
tU |
QJ |
c |
3 ■»-■ |
l« ■+-( |
■t-> |
"O o |
||||
|
s: |
•H I, Of |
c |
rt |
a-H |
•u |
c |
^ |
0) |
U<r^ |
j-> |
■p |
CO |
j= |
> |
CO C "O |
|
■>-> |
e o^ |
« |
S |
£ |
c |
CJ |
U5 |
Cfi c |
Si O 0) |
0 |
•h |
-»-> |
o |
Ih 0) |
|
|
CS |
Ih jC |
3 |
+J |
« |
In |
3 |
cu c x: |
c |
CO |
(U |
rH |
CO C |
|||
|
u |
QJ Cfl |
(fl |
0) |
o >. |
0 |
■—1 |
>. |
M |
rH 4-> |
a |
0) |
* -H q; |
|||
|
Q) |
■l-> 'M [fl |
rH |
c |
C 'm |
s |
X |
CC |
nH |
0 tfi |
3 - |
o |
0 |
-t-" |
CO |
tfl J3 ■*-> |
|
O |
O QJ |
« |
a |
QJ |
3 |
■t-i |
*-> |
<D |
■H Q) |
£1 S 0 |
en |
rH |
3 -H :h |
||
|
■H |
rH n-( |
+J |
+-> > |
■iH |
rt |
V |
T3 |
in [fi |
■H 0 ■'-' |
-H |
^ |
3 |
tn ■(-) 0 |
||
|
a |
rH !h C |
c |
-H |
c |
!-( |
■H |
■H |
ID Ul |
T3 S-. |
QJ |
0) |
a |
0) x: |
||
|
ca H ti |
0 |
J3 |
« |
^ |
s |
■■H 0) |
C « |
5 |
0) |
CO |
E -o en |
||||
|
E « 0 |
S-i |
In |
u |
OJ |
C U |
CO CO |
^V |
x: |
u |
3 C |
|||||
|
w cs: |
t^ |
o |
U |
&, |
1— 1 |
S C |
J |
u |
1/1 |
X CO |
|
<1) |
|||||||||||||||||
|
1 s |
■o |
» |
|||||||||||||||
|
(0 |
o c |
c |
!-( |
||||||||||||||
|
rH |
Ih rH |
CO |
>> |
p |
3 |
||||||||||||
|
Ct |
a (D |
tn a |
rH |
c |
0) |
E |
|||||||||||
|
+J |
0 |
■w |
■(-> |
12 |
■H X |
tu |
c |
"O |
0) |
||||||||
|
rH |
tn |
(B |
tn |
3 |
o |
rH |
tn 0) |
■Q |
■o |
•H |
'H X |
||||||
|
CO |
0) |
■H |
a |
3 |
■o |
CO T) |
>> |
•H |
0 |
s |
u |
||||||
|
E |
tn |
X |
rH |
•o |
■o |
■(-( |
0) |
tn 0) |
X >> |
? |
tn |
- 3 |
|||||
|
3 |
CO |
0) |
CO |
0) |
0) |
u |
* |
tn |
+-> |
0 -P |
ttrH |
a |
■o |
tn E |
|||
|
■H |
c |
(fl |
a |
N |
o |
u |
0 |
-M |
CO |
E H |
E P >. |
o |
>> |
c |
3 |
||
|
•D |
•H |
CO |
x; |
c |
Ih |
tn |
^ |
CO C |
>,r. '-\ |
in |
■u |
X |
CO |
•H J-J |
|||
|
■H |
0) |
x: |
■D |
■M |
r~t |
bX) |
Sh |
0) |
CO |
3 3 |
tn iic in |
0) |
■o o |
||||
|
t. |
rH |
■(-> |
c |
CO |
a |
•H |
•o |
CO |
Ih |
a |
cy |
•H 0 |
H-> |
p |
X |
? |
CO C "O |
|
0) |
ic |
0 |
CO |
E |
X |
c |
c |
CJ |
OJ |
tn p |
in rH -H |
0 |
CO |
HJ |
0 |
U Q) |
|
|
s: |
c |
ja |
3 |
■t-> |
CO |
tn |
0 |
CO tn '-H |
c |
u |
0) |
r-{ |
CO c |
||||
|
4-> |
■H |
tn |
<D |
0 |
>> |
Jh |
r-{ |
u c |
rH Q) |
CO |
Q) |
-•H Q) |
|||||
|
0 |
cn |
c |
^^ |
C |
c |
-H |
E |
0) |
CO |
>. |
0 |
3 --P |
u |
a |
-P |
CO |
tn ij -p |
|
r— ( |
■H |
CO |
a |
QJ |
— > |
X |
4-) |
^^ |
•H tfj |
^ > C |
0) |
rH |
3 H Ih |
||||
|
■H |
rH |
+J |
■t-j |
> |
■H |
-(-> |
0) |
0 |
U Q) |
H C CO |
^ |
tn |
^ |
3 |
in P 0 |
||
|
s: |
rH |
C |
c |
■H |
C |
CO |
■H |
■0 |
0) tn |
■o '- |
0) |
0) |
a |
0) £ |
|||
|
u |
CO |
-H |
0 |
a |
ns |
u |
!h |
■H |
=H tn |
G 1h |
> |
V |
(0 |
E -U tfl |
|||
|
£ |
0 |
S-« |
Ih |
^ |
CO |
» |
C Q) |
CO CO |
o |
x: |
o |
3 C |
|||||
|
CO |
J= |
ClH |
o |
u |
Oh |
1-1 u |
s c |
J |
tj |
w |
X CO |
|
x: |
|||||||||||||||
|
u |
■o |
^1 |
|||||||||||||
|
•k |
•p |
f > |
3 |
0 |
V |
*. |
|||||||||
|
10 |
(0 |
X; rH |
o |
>> |
0 o |
E |
+-> |
T5 |
L, |
||||||
|
0) |
iH |
tl£ rH |
c |
f-i |
U r-i |
CQ |
■p |
C |
3 |
||||||
|
X |
C« |
■H 3 |
0 |
a 0 |
tn |
>> |
Sh |
c |
0 |
E T3 |
|||||
|
0) |
+J |
x; -a: |
+-> |
T3 |
£! |
■H |
rH |
CQ |
o |
rH |
0 0 |
||||
|
CO |
a> |
tn |
3 |
■H |
rH |
tn |
rH |
O. |
•o |
tn |
=P C |
||||
|
•H -O |
>. |
a |
» |
CO -D |
>> |
0 |
0 |
o |
0 |
||||||
|
£ |
£ |
Ih 0 |
u u |
-r^ |
tn 0 |
X |
■H |
tn |
in |
■D |
- +-> |
||||
|
3 |
■p |
CO N |
0 0 |
o |
tn |
O -P |
a |
^ |
c. |
C |
in Ih |
||||
|
•H |
0 |
a-H |
> a |
o |
•D |
■p |
■D |
E -H |
E |
0 |
>> |
>. |
CO |
3 0 |
|
|
Ih |
£i |
-p |
c |
in |
o |
0 |
tn |
0 |
CO C |
>> |
J-> |
rH |
X |
•H j;: |
|
|
<0 |
T3 CO |
■O 3 |
3 |
> |
0 |
■p |
3 3 |
tn |
c |
0 |
x; |
-o tn |
|||
|
x: |
c |
C E |
0 |
o |
■u |
O |
Ih |
CO |
C |
CO |
•o |
x: |
bc |
CO |
|
|
4-> |
•H |
CO 3 |
o S |
p |
c |
U |
u |
U |
tn -p |
ti |
•H |
•p |
•rH |
U CO |
|
|
O |
0 |
CO 0 |
c |
CO |
u |
CO |
0 |
CO |
T3 |
> |
0 |
X |
•H |
||
|
rH |
tn |
tn c |
r-i r^ |
0 |
CO |
rH |
a |
tn C |
rH |
0 |
0 |
-£J |
|||
|
•H |
to |
rH a |
a. 0 |
o |
E |
c |
CO |
0 |
0 |
3 |
C |
u |
■p |
CO |
in -H |
|
X |
(U |
« |
^ |
3 |
-p |
tn |
■H tn |
J2 |
0 |
rH |
3 -P |
||||
|
o |
rH |
-p -p |
■P |
■H |
0 |
tH 0 |
■H |
■o |
u |
^ |
3 |
Li |
|||
|
c |
c o |
•H +J |
c |
-H |
0 tn |
•D |
•H |
0 |
0 |
a |
0 T3 |
||||
|
^< |
0 c |
X2 tn |
CO |
Lh |
■■H tn |
C |
> |
s |
0 |
CO |
E C |
||||
|
o |
u |
U 3 |
Lh |
CO |
C 0 |
CO |
0 |
x: |
o |
3 CO |
|||||
|
s |
(x< |
O -o |
u |
D< |
P-H CJ |
S |
J |
o |
t/3 |
X |
388
Bulletin Museum of Comparative Zoology, Vol. 142, No. 3
|
• |
u |
||||||||||||
|
0) |
1— 1 |
0) |
|||||||||||
|
to |
>. |
he |
■(-> |
||||||||||
|
J= |
c |
■u |
+J |
c |
(1) |
||||||||
|
■M |
c |
0) |
u |
O |
o |
o |
1 |
||||||
|
x: |
•H |
•H |
e |
■H |
CO |
1— 1 |
ca |
o |
|||||
|
4-> |
? |
C |
ca |
£ |
■o |
=M |
0) |
||||||
|
•H |
■v |
O |
XJ |
■(-> |
•H |
+J |
Sh |
||||||
|
* |
•o |
0) |
•H |
■H |
Sh rH |
u |
ro |
CO |
|||||
|
0) |
•o |
^-> |
M |
1— 1 |
+J >. |
0) |
x: |
•r-( |
>. |
^-N |
|||
|
•a |
+-> |
c |
u |
m |
r— I |
u |
■fj |
hD |
> ja |
(— 1 |
C |
||
|
0) |
•H |
3 |
0) |
+-> |
rt |
0) |
>. u |
C |
0) |
■H |
t< |
> |
|
|
+j |
c |
o |
m |
6 |
■M |
r-i a |
o |
E |
■M |
c« |
u |
0 |
|
|
•H «3 |
3 |
u |
•, w |
c |
x> -o |
t-H |
O |
(U |
a |
c |
|||
|
C 3 |
■!-> |
Ql |
« |
re -H |
tn |
4-> |
c |
1— 1 |
^; |
||||
|
3 -H |
•fJ |
■o |
tf-l |
sz |
x; |
X) Si |
tfi |
3 |
3 |
c |
|||
|
TD |
o |
c |
RS |
C -M |
y |
0 ■I-' |
3 |
E |
0 |
u |
bJJ |
3 |
|
|
■(-> m |
c |
« |
JZ |
■H -H |
o |
u |
1— 1 |
3 |
£ |
ca |
G |
||
|
O i. |
Cfi |
^ =c |
In |
a tn |
« |
(1> |
-u |
1— 1 |
a |
XI |
|||
|
c |
a |
* |
1 |
■)-) |
(D |
M |
C |
■H |
3 |
-M |
■H |
||
|
t— 1 |
a |
■D |
Vt ^N |
tfi a |
Bl |
rt |
S |
U |
O |
||||
|
« |
3 |
■H |
■H |
3 >. |
3 |
Jh |
O |
■H |
^ |
||||
|
c |
XJ XI |
e |
e ^ |
C |
•M |
r-H |
> |
3 |
|||||
|
l-t |
■H |
■H |
(D C |
« |
01 |
W |
ca |
(J |
|||||
|
3 |
Pm |
4-> |
U-, |
0 |
Z |
<: |
o |
^: |
v_^ |
|
+J |
|
|
3 |
|
|
X3^ |
|
|
QJ |
|
|
~ tn |
+-> |
|
t, c |
Li |
|
01 0) |
O |
|
bC d |
X |
|
C •!-( |
tn |
|
o m |
|
|
rH o |
u |
|
Q) |
0) |
|
tn ca |
x: |
|
rH rH |
+j |
|
ca ca |
ca |
|
■H a |
u |
|
•o |
|
|
0 • > |
|
|
aQ|M |
|
|
ca |
|
|
■P c • |
|
|
0) -H +J |
|
|
s^ |
S |
>
o
c
c
3
tn u ca
CQ
■p v
T3
C
o
U
CO
pa
|
x: -o |
"O |
||||
|
•p |
(V |
u |
|||
|
x: |
•H |
•o |
c |
•H |
|
|
■p |
* |
c |
0 |
X |
|
|
■H |
3 |
■H |
■p |
||
|
> |
■o |
0 |
+J |
||
|
0) |
'-H |
CJ |
rH t, |
||
|
■o |
■p |
0 |
rH 0) |
||
|
0) |
■H |
4-) |
tn |
ca -P |
|
|
-t-> |
c |
o |
E C |
||
|
■H |
3 |
c |
■p |
tn ca |
|
|
c |
tn |
•*-( |
X |
||
|
3 |
3 |
p |
■D |
ca |
X u |
|
■H |
o |
c |
X |
■p o |
|
|
p |
■D |
c |
ca |
in |
•H t^ |
|
0 |
ca |
1 |
? -p |
||
|
c |
!-( |
ca |
- |
•o |
|
|
rH |
ca |
-H |
^H |
||
|
ca |
3 |
■H |
£ |
3 |
|
|
c |
x: |
XI |
E |
||
|
r-* |
•H |
■H |
C |
0 |
|
|
3 |
[IH |
•P |
■H |
tx< |
|
u |
|||
|
0) |
|||
|
tJC |
|||
|
tn |
C -P |
||
|
tu |
ts |
0 Q) |
|
|
a |
rH |
rH o |
|
|
0) |
ca |
||
|
•. |
+J |
■P fH |
|
|
rH |
ca |
ca |
|
|
>.rH |
Ih |
X ca |
|
|
■P >> |
OJ |
? -H |
|
|
O -P |
XJ |
0) XJ |
|
|
ca CJ |
o |
E H |
|
|
"D ca |
E |
0 -P |
|
|
ca TJ |
tn |
||
|
Ss -H |
in |
M |
-p |
|
■p t< |
3 |
C |
E 3 |
|
0) -P |
rH |
c |
3 0 |
|
p |
ca |
,—t |
0) £ |
|
bC |
C -P |
||
|
tn |
ca |
ca -H |
|
|
3 |
i^ |
u > |
|
|
C |
-p |
rH |
|
|
ca |
tn |
ca |
|
|
S |
< |
U |
(0 0)
|
•. |
I |
4J |
||||
|
X |
E |
■u |
ca |
|||
|
1 |
bO |
0 |
c |
r-i >. |
||
|
V |
•r( |
Lh |
0) ca |
-P rH |
||
|
0) |
X |
a |
bJ3 |
' a U |
||
|
tn |
XI tn |
In X O |
||||
|
c |
0) |
■H 3 |
0) it -H |
|||
|
>> |
ca |
c |
U r-i |
tn |
be 0) l< |
|
|
rH |
X |
0 |
ca |
■p |
CEO) |
|
|
t, |
•p |
XI |
X bC |
0 |
O 0 -P |
|
|
ca |
^ |
-p ca |
CJ |
rH tn en |
||
|
0) |
CC |
u |
,—t |
0 Sh |
ca |
o |
|
c |
r-i |
1 |
ca |
0 +-> |
m |
tn tn a |
|
3 |
X3 |
t^ |
E tn |
rH •-{ |
||
|
t, |
M |
■H |
(D |
tn ca |
E |
ca ca p |
|
ca |
C |
> |
■p |
3 |
■H -H C |
|
|
r-i |
ca |
ca |
- c |
0) |
X! XI 0) |
|
|
3 |
-p |
•o |
r-( |
<D (U |
c |
0 O bC |
|
o |
■r-1 |
CJ 0) |
ca |
D. & !h |
||
|
•r^ |
o |
X |
c > |
V |
ca ca 0) |
|
|
> |
X5 |
•P |
0) -P |
r-{ |
p -p > |
|
|
ca |
3 |
•H |
C (U |
ca |
Q) Q) -ri |
|
|
z |
U |
S |
•H J2 |
o |
S E -O |
ca
E tn
P X)
|
> |
0 |
|
X2 |
|
|
HH |
3 |
|
l-H |
CJ |
|
HH |
|
|
u |
|
|
rH |
0 |
|
ca |
«M |
|
in |
|
|
^ |
+-> |
|
ca |
0) |
|
-p |
CJ |
|
ca |
ca |
|
-p |
tp |
|
(1) |
|
|
^-< |
|
|
.* |
|
tn 3 |
i-t CQ |
||||||||||||||||
|
rH |
S |
U |
|||||||||||||||
|
r-i |
ca |
3 |
QJ |
||||||||||||||
|
T3 |
•o |
>, |
-u |
bc |
XI |
c |
Q) |
•p |
■p |
||||||||
|
X Q) |
t< |
P |
>> |
X c |
ca |
c |
ca |
C |
ca hc |
3 |
|||||||
|
■P X3 |
c |
•H |
CJ |
rH |
■P -H |
;h |
•H |
X |
c |
ca |
r-l c |
0 |
|||||
|
X |
•H C |
o |
X |
ca |
QJ |
■H x: |
•p |
X |
■p |
Q) |
CJ |
■H |
X |
||||
|
-p |
> 3 |
■H |
■p |
T3 |
■p |
^ Q) |
tn |
QJ |
Q) |
rH |
- bC |
■p |
|||||
|
•H |
0 |
•p |
■H |
ca |
x: |
ca |
X2 |
u |
> |
ca |
■P !h |
•H |
|||||
|
* |
-o u |
o |
(U |
u |
r( |
u |
Q> |
■p |
u |
ti Q) |
? XJ |
||||||
|
0) |
Q) |
bD |
Q) |
p |
Q) |
■p ca |
u |
t^ |
X> |
Q) |
o > |
•H |
|||||
|
•o |
tn -p |
tn |
t* |
■P |
XI |
U -rt |
o |
QJ |
■p |
■H |
^ |
-o |
X -H >> |
HH 0 |
|||
|
(1) |
0 o |
ca |
C |
tn |
0 |
O X2 |
fn |
X) |
c |
? |
c |
tn X) rH |
l-H J2 |
||||
|
tn tn |
,-i C |
^-> |
,-i |
ca |
QJ |
E |
X -n |
•H |
o |
x; |
ca |
tn |
u |
<-< 3 |
|||
|
O 3 |
>. |
'p |
X |
a |
tn -p |
■P |
> |
u |
X |
Q) |
-p |
tn tn o |
CJ |
||||
|
rH -H |
X n |
ca |
X |
CJ |
tn |
b£ |
Q) |
=4H |
o |
he |
tn |
Q) |
rH rH -rl |
rH |
|||
|
>>XI |
c c |
X |
-p |
o |
XI |
3 |
C |
E fcH |
CJ |
tl |
3 |
■V |
3 |
CJ |
ca ca t< |
ca u |
|
|
^ ca |
ca ca |
tn |
•H |
t< |
c |
rH |
o |
3 o |
ca |
ca |
c |
E |
•H |
rH |
ca |
•H -H Q) |
tn o |
|
c u |
1 |
S |
•p |
ca |
ca |
rH |
0) fH |
(p |
rH |
•rH |
Sh |
ta |
tM |
XI XI -P |
t, 4H |
||
|
ca |
ca - |
■V |
bC |
c |
3 |
•o |
hc |
o 0 tn |
ca |
||||||||
|
r-i a |
•H |
;h |
tn |
ca |
ca -p |
Lh |
CJ |
C |
■rH |
- |
ca |
a a o |
■p -p |
||||
|
CO |
3 -H |
E |
3 |
3 |
!-. |
O <D |
Q) |
■rH |
ca |
0 |
X |
u |
ca ca a |
ca Q) |
|||
|
B |
X3 X3 |
E |
C |
■P |
,-i V |
a |
> |
X |
X2 |
bD-P |
■p -p |
■P V |
|||||
|
r-i |
•H -H |
c |
0) |
ca |
tn |
CO m |
a |
ca |
■p |
3 |
•H |
tn |
OJ 0) |
0) ca |
|||
|
P |
(1, +J |
■r( |
i^ |
S. |
< |
CJ t«H |
3 |
Z |
CJ |
X |
ca |
s s |
s <p |
Rhinoceros from the Miocene of Kenya • Hooijer
389
LITERATURE CITED
Aguirre, E. de. 1963. Hippopotamus crusafonti n. sp. del Pliocene inferior de Arenas del Rey (Granada). Notas Comun. Inst. Geol. Minero Espana, No. 69: 215-230.
Arambourg, C. 1959. Vertebres continentaux du Miocene superieur de I'Afrique du Nord. Puhl. Serv. Carte Geol. Algerie, N. S., Paleontologie, Mem. 4: 1-161.
Bishop, W. W. 1967. The later Tertiary in East Africa — volcanics, sediments, and faunal in- ventory. In W. W. Bishop and J. D. Clark (eds.). Background to Evolution in Africa. Chicago and London, University of Chicago Press, pp. 31-56.
Bishop, W. W., and J. D. Clark (eds.). 1967. Background to Evolution in Africa. Chicago and London, University of Chicago Press, X + 935 pp.
BoHLiN, B. 1937. Eine Tertiiire Siiugetierfauna aus Tsaidam. Sino-Swedish Expedition VI-1; Paleontol. Sinica, sen C, 14(fasc. 1): 1-111.
. 1946. The fossil mammals from tlie
Tertiary Deposit of Talien-buluk, Western Kansu. Part II. Palaeontol. Sinica, N. S., C, No. 8B: 1-259.
Colbert, E. H. 1935. Siwalik mammals in the American Museum of Natural History. Trans. American Phil. Soc. Philadelphia, N. S., 26: X + 1-401.
COUVERING, J. A. VAN, AND J. A. MiLLER. 1969.
Miocene stratigraphy and age determinations,
Rusinga Island, Kenya. Nature (London),
221(5181, Feb. 15): 628-632. DuvERNOY, M. 1853. Nouvelles etudes sur les
Rhinoceros fossiles. Arch. Mus. Hist. Nat.
Paris, 7: 1-144. Hooijer, D. A. 1946. Notes on some Pontian
mammals from Sicily, figured by Seguenza.
Arch. Neerlandaises Zool., 7: .301-333. . 1963. Miocene Mammalia of Congo ( with
a chapter by A. Gautier and J. Lepersonne).
Ann. Mus. Roy. Afrique Centrale, Tervuren,
Belgique, Sci. Geol, No. 46: ix + 1-77. . 1966. Miocene rhinoceroses of East
Africa. Bull. British Mus. (Nat. Hist.), Geol.,
13(2): 117-190. . 1968a. A note on the mandible of Acera-
therium acutirostratum ( Deraniyagala ) from
Moruaret Hill, Turkana District, Kenya. Zool.
Meded. Rijksmus. Natuur. Leiden, 42(21):
231-235. . 1968b. A rhinoceros from the late Mio-
cene of Fort Ternan, Kenya. Zool. Meded.
Rijksmus. Natuur. Leiden, 43(6): 77-92. JouBERT, p. 1966. Geology of the Loperot area.
Geol. Surv. Kenya, Rept. No. 74: ii + 1-52. Leakey, L. B. S. 1967. Notes on the mammalian
faunas from the Miocene and Pleistocene of
East Africa. In W. W. Bishop and J- D.
Clark (eds.). Background to Evolution in
Africa. Chicago and London, University of
Chicago Press, pp. 7-29. Maglio, V. J. 1969. A shovel-tusked gomphothere
from the Miocene of Kenya. Breviora, No.
310: 1-10. Peterson, O. A. 1920. The American Dicera-
theres. Mem. Carnegie Mus., 7(6): 399-456. Ringstrom, T. 1924. Nashorner der Hipparion-
Fauna Nord-Chinas. Palaeontol. Sinica, ser.
C, l(fasc. 4): 1-156. Roman, F. 1924. Contribution a I'etude de la
faune de mammiferes des Littorinenkalk
(Ohgocene Superieur) du Bassin de Mayence.
Les Rhinoceros. Trav. Lab. Geol. Fac. Sci.
Lyon, fasc. VII - Mem. 6: 1-54. , AND J. ViRET. 1934. La faune de
mammiferes du Burdigalien de la Romieu
(Gers). Mem. Soc. Geol. France, N. S.,
9(21): 1-67. Walker, A. 1968. The Lower Miocene fossil site
of Bukwa, Sebei. Uganda Journal, 32: 149-
156. . 1969. Fossil mammal locality on Mount
Elgon, Eastern Uganda. Nature (London),
223(5206, Aug. 9): 591-593.
390
Bulletin Museum of Comporativc Zoology, Vol. 142, No. 3
APPENDIX
Hypodigm of Chilotheridium pattersoni Hooijer, gen. et sp. nov.
Field No. 70-64K, found by B. Patterson. Skull, C9-10
„ , TYPE, B12 Right and left maxilla, 65B Nasal bones. A" 18 Nf , right, C9-10 Mandible II , 65
, Bll II , right ramus, part of left, 65C II , part of left ramus, A18 Lower canine, left, 65
II II , right, A16
Atlas, 65B .1 , CI Left scapula, A18 Right -, , BB II II , BL
II II , 65B
II II , BB
Right humerus, A18 II II , BB
Right radius
Left
Right
Left
Right
Left
Right
Left Right Left Right
II
II Left
Right
Left
, distal end, — BB A16 C14 C14 BB A17 BB proximal end, —
II , ElO distal end, — I, , B16 I. , C12 M II , BB
ulna, A17
BB BB
C14 C14 BB A17
distal end, BB II II , BB
Right scaphoid, A16
(I It Left
BB
Dll
BB
(No. 1) 2
(No. 1)
II 2
,1 3
I, 4
II 5
(N
o.
(No. 1) 2 3 4
1)
2
3 4 5 6 7 8 9 10 11 13 14 15 (No. 1) 2 3 4 5 6 7
(N
9 11
1)
2
3 4 5 6
Right lunar, —
II II , —
M II ,
Left II , — Right cuneiform, —
Left
, A16 , BB
Pisifonn, proximal end, II , II II ,
Right trapezoid, Hll
Left
, A16
Right magnum, BB , B14
II II , —
Left II , —
II II , —
I, , Hll
Right unciform, A17
II II , J7
II II , A16
Left
A16 Ell
, DIO Right metacarpal II, B13
II II II , proximal part,
BB II II II , proximal part
Left
(No. 1)
II 2
.1 3
,1 4
(No. 3)
(N
(N
B14
Right metacarpal III, —
II II II , proximal part, B14
Left II II , A17
Right metacarpal IV, C14 Left II I, , A16, 17
II II II , C_j14 Left metacarpal \', BB
II II II , C<14
II II II , BB II 4
II II II , mid-shaft missing, — n 5
Left metacarpal II (No. 2) and left metacarpal
III (No. 4) of one individual, A17.
Right metacarpal IV ( No. 1 ) and the right meta- carpal V ( No. 1 ) of one individual, B14.
Rib, 65A
Partial sacrum and part of left os innominatum, A16.
3 5 7
o. 1) 2 3 4 6 7 8 9 10
o. 1) 3 4 5
(N
2 3 4 5 7 8 9
(No. 1) 3 5
Rhinoceros from the Miocene of Kenya • Hooijer
391
Left femur, — (No. 1)
M I' , II 2
Right 11 , shaft only, juv., BB h 3
11 11 , distal epiphysis, — h 4
Left 1. , distal end, A16 n 5
11 II , proximal part of shaft, BB n 6
Right patella, — (No. 1)
Left ,1 , ElO I, 3
Parts of five others, Dll, D12, BB
Right tibia, A16 (No. 1)
11 II , incomplete at end, — n 2
II It , DD II O
Left II , proximal end of shaft, BB h 5 11 11 , lateral distal fragment, ElO h 6
Right fibula, 65C (No. 1)
Left 1- , 65C II 2
Right II , B15, 16 ,1 3
II II , proximal part, — n 4
Left II , distal end, — n 5
Right II , II II , — II 6
Lett II , 11 II , — II 7
Most of right pes of one individual, including astragalus (No. 1), calcaneimi (No. 1), navi- cular (No. 1), cuboid (No. 1), ectocuneifonn (No. 1), mt. II (No. 1), mt. Ill (No. 1), mt. IV (No. 1), all phalanges except IV 3, and one sesamoid (No. 1); B15, 16. Right astragalus, incomplete distally, — (No. 2)
A16, 17 Left astragalus, incomplete distally,
A18 II II ,11 II ,
C12
|
igh |
t calcaneum. |
BL |
|
|
M |
A18 |
||
|
M |
A16, |
17 |
|
|
eft |
BB |
||
|
11 |
A16 |
||
|
II |
ElO |
||
|
II |
E12 |
11 3 II 4
II 5
(No. 2) II 3 11 4 II 6 II 7 II 8 II 9 II 10
(No. 3) ,1 ^4
(No. 2) II 3 ,1 4
(No. 2) II 3 II 4 II 5
(No. 2) II 3
(No. 2) I, 3
(No. 2) .1 3
Left metatarsal IV, 65B (No. 2)
ti II II , proximal part, Dll n 3
Right navicular, — Left .1 , Cll
Right cuboid, Dll
II II , —
II II , A16, 17
Right ectocuneiform, —
II II , A16
Left 1. , Dll
, Dll Right mesocuneifomi, Dll Left 1. , —
Right entocuneiform, Dll Left II , —
Right metatarsal II, BB
II II II , proximal part, —
A16, 17 Right II II , proximal end, D12
Median metapodial, distal end, ElO
B16
Lateral metapodial
Phalanx 1
II II
M
Phalanx 2
Phalanx 3 Phalanx 1
distal end
(N
A16, 17 Dll
— (No
BB
A16, 17
Dll Dll
median digit, ElO
BB
A16,
A16,
median digit, E12 II . A16
17 17
, A16, 17
median digit lateral digit, ElO , A17
A 16 BB
HIO
B14 Dll
(N
(No
(No.
4 5
1)
2
3 4 5
7 8
2)
3
4
5
6
7
9
10
11
12
13
14
15
16
1) 2
3
4
6
8
9
10
11
15
1) 3
4
5
6
7
8
10
1) 2
3
4
6
7
8
9
11
12
13
14
15
17
18
19
392 BuUciin Museum of Comparative Zoologij, Vol. 142, No. 3
Phalanx 2, lateral digit, —
Dll Dll D12
A16, 17 A16, 17
nil
B14 A16, 17
|
(No. |
1) |
|
2 |
|
|
3 |
|
|
4 |
|
|
5 |
|
|
6 |
|
|
7 |
|
|
8 |
|
|
10 |
|
|
11 |
|
|
12 |
|
|
14 |
|
|
17 |
|
|
18 |
lateral digit; several incomplete speci-
Phalanx 3, mens.
Proximal sesamoid, median digit, B15,
16 A16
Proximal sesamoid
lateral digit
Dll A16
Dll
(No. 1)
M 3 M 4
„ 6
„ 7
„ 9
-, 10
M 11
„ 12
(No. 1)
„ 2
,■ 3
M 4
M 5
n 6
11 7 ,1 8 11 9 1, 10 1, 12 1, 13 „ 14 1, 15 „ 16 n 17 11 18 ,1 19 1, 20
At least eight individuals are represented. Nu- merous other bones were also collected from this quarry but are not listed since they do not add to knowledge of the species.
Field No. 57-64K, approximately 50 yards south- east of 70-64K and at same level. Found by B. Patterson. (This may be a continuation of the 70- 64K quarr\'. )
Right radius, distal part Left ulna, distal part Right cuneifonn
11 II
Left
Pisiform, proximal end Left trapezoid
11 II
Right magnum Right unciform
Right metacarpal II, proximal part
Right metacarpal III, proximal part
Right metacarpal I\', proximal part
Right patella
Right tibia, distal end
Right navicular
Lateral metapodial, distal end
II II , II II
Phalanx 1, median digit
II , II II
Phalanx 2, median digit
II , II II
Phalanx 1, lateral digit
II , II II
II , II II
Phalanx 2, lateral digit
II , II II
II 5 II II
Proximal sesamoid, median digit
II II , II II
II II , II II
Proximal sesamoid, lateral digit At least two individuals are represented.
Field No. 64-64K, approximately 20 yards east of and at same level as 70-64K. Found In' C. T. Williams.
Incomplete mandible.
Field No. 68-64K, as for 64-64K. Found by W.
D. Sill. Right cimeiform Right calcaneum Right mesocuneifonn Median metapodial, distal end Phalanx 1, median digit
II , II II
II , II II
Phalanx 2, median digit Phalanx 2, lateral digit
|
(No. |
12) |
|
(No. |
10) |
|
(No. |
1) |
|
II |
2 |
|
M |
9 |
|
(No. |
4) |
|
6 |
|
|
(No. |
5) |
|
(No. |
2) |
|
M |
6 |
|
II |
7 |
|
(No. |
6) |
|
(No. |
2) |
|
(No. |
3) |
|
(No. |
2) |
|
(No. |
4) |
|
(No. |
2) |
|
(No. |
I) |
|
(No. |
8) |
|
(No. |
13) |
|
II |
14 |
|
(No. |
2) |
|
II |
11 |
|
(No. |
5) |
|
II |
10 |
|
II |
16 |
|
(No. |
9) |
|
1 1 |
13 |
|
II |
15 |
|
(No. |
2) |
|
II |
5 |
|
M |
8 |
|
(No. |
11) |
|
(No. |
4) |
|
(No. |
5) |
|
(No. |
1) |
|
(No. |
6) |
|
(No. |
5) |
|
12 |
|
|
(No. |
9) |
|
(No. |
16) |
eti/t OF THE
seum o
Comparative Zoology
Osteology of the Malaysian Phallostethoid
Fish Ceratostethus bicornis, with a Discussion
of the Evolution of Remarkable Structural
Novelties in its Jaws and External Genitalia
TYSON R. ROBERTS
HARVARD UNIVERSITY
CAMBRIDGE, MASSACHUSETTS, U.S.A.
VOLUME 142, NUMBER 4 29 DECEMBER 1971
PUBLICATIONS ISSUED
OR DISTRIBUTED BY THE
MUSEUM OF COMPARATIVE ZOOLOGY
HARVARD UNIVERSITY
Bulletin 1863-
Breviora 1952-
Memoirs 1864-1938
JoHNSONiA, Department of Mollusks, 1941-
OccASiONAL Papers on Mollusks, 1945-
Other Publications.
Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth.
Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth.
Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth.
Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth.
Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. )
Turner, R, D., 1966. A Survey and Illustrated Catalogue of the Teredinidae (Mollusca: Bivalvia). $8.00 cloth.
Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth.
Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only.)
Publications of the Boston Society of Natural History.
Authors preparing manuscripts for the Bulletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A.
© The President and Fellows of Harvard College 1971.
OSTEOLOGY OF THE MALAYSIAN PHALLOSTETHOID FISH CfRATOSTETHUS BICORNIS, WITH A DISCUSSION OF THE EVOLUTION OF REMARKABLE STRUCTURAL NOVELTIES IN ITS JAWS AND EXTERNAL GENITALIA^
TYSON R. Roberts-
Abstract. The osteology of the phallostethoid Ccratostethtis bicornis (Regan) is described and figured. Comparative observations on osteology of atherinoids, cyprinodontoids, and other phallo- stethoids are also given. Phallostethoids apparently originated from atherinids. The most closely related atherinids are Taeniomembrasinae. The osteological observations tend to support the idea that atherinoids and cyprinodontoids are related, as postulated ])y Rosen ( 1964 ) in proposing the order Atheriniformes. Atherinifomis exhil^it a widespread tendency to develop teeth with two and three cusps, especially on the pharyngeal bones. The trend is pronounced in cyprinodontoids, exocoetoids, and scomberesocids, practically absent in atherinoids, and completely absent in phallo- stethoids and belonids.
The Phallostethoidea can be divided into two families, Phallostethidae and Neostethidae. Neo- stetliidae comprises two subfamilies, Neostethinae and Gulaphallinae. These groups are distinguish- able on the basis of morphological differences in the jaws and external genitalia. The highly pro- tractile jaws of Neostethinae ( as exemplified" ])y Ceratostethus and Neostethtis) are remarkalile in having several pairs of bones without homologues in other fishes. These new bones, invohed mainly in protrusion of the jaws, evolved in soft structures already present in the jaws of atherinids. The functional anatomy of the jaws of phallostethoids is briefly discussed. Phallostethidae and Gula- phallinae lack the neomorphic jaw bones found in Neostethinae. Two ctenactinia formed from pelvic fin spines or rays occur only in CeratostetJius. The so-called "second ctenactinium" of Gulaphallus is
^ Published by a grant from tlie Wetmore Colles Fund.
- Museum of Comparative Zoology, Harvard Uni- versity, Cambridge, Massachusetts 02138
an externalized pelvic bone. Phallostetliidae ap- parently arose from Neostcthus. The toxactinium, the main externalized bony element in the priapium of Phallostethidae, is derived from the inner pulvinular bone, which is the anteriormost internal liony element in tlie priapium of Neostetliidae. The inner pulvinular bone of Neo- stethus bears a small lateral projection, the pulvinular spine, which may be a rudimentary toxactinium. The structure of the papillary ])one, intimately associated with the genital pore, is relatively simple in Gulaphallinae and Phallo- stethidae, ])ut in Neostethinae it divides into numerous thin processes each bearing a booklet at its tip. A comprehensive definition is given for the superfamily Phallostethoidea. All taxa used in this paper were proposed l^y previous authors. Oviparous Atheriniformes with internal fertiliza- tion have external genitalia far more complicated than the gonopodium of any of the viviparous .\theriniformes. The explanation of this difference is sought in terms of selection pressures for and against the evolution of liighly complicated ex- ternal genitalia in forms with internal fertilization. Neotenic characters probably played an important role in the origin of phallostethoids from Atherin- idae.
INTRODUCTION
Ichthyologists have marvelled at phallo- stethoids since thev \\'ere first made known by C. T. Regan (1913; 1916). Males of these delicate little fishes from Southeast Asia can be distinguished in an instant by their strange subcephalic copulatory organ, or priapium, as Regan designated it. The 19 species now known have been divided
Bull. Mus. Comp. Zool., 142(4): 393-418, December, 1971
393
394 Bulletin Museum of Comparative Zoology, Vol. 142, No. 4
,-K
■'K
,^
t;r:
a
> o
a c
D
.c U
J3 O
U
o
3
o n
U
— o
E
E E
CO
3
o
0 0)
0)
k_
into two families and ten genera, largely based on differences in the external arma- ture of the priapium (see Herre, 1942). H. M. Smith (1927) made the surprising an- nouncement that phallostethoids he ob- served in Thailand are oviparous. This discovery was confirmed in species from the Philippines by Villadolid and Manacop (1934) and Woltereck (1942a). Despite their outstanding interest, phallostethoids have remained virtually unknown to the general zoological public.
The present paper gives a relatively complete, illustrated account of the osteol- ogy of Cerotostethu.s ])icornis (Regan) (Fig. 1), with special attention to the hitherto undescribed jaws and their func- tional anatomy. Previous osteological work on Phallostethoidea (with the exception of brief observations by Regan [1913; 1916], who lacked adequate skeletal preparations ) has been almost totally restricted to the priapium (Bailey, 1936; Aurich, 1937; Woltereck, 1942a, b). The priapium is derived mainly from pelvic bones and fin rays, with contributions from the anterior- most ribs and pectoral girdle. The pelvic elements are so excessively modified that their homologies remain unresolved. This copulatory organ is perhaps even more specialized than the copulatory organs derived from the anal fin in the cyprin- odontoids Horaichthi/s and Tomeunis. Interestingly enough, Iloraichthys and Tomeunis are also oviparous, and they exhibit numerous morphological and eco- logical similarities to phallostethoids. Tlie morphological similarities, however, do not extend to the jaws of Ceratostethtis: al- though basically similar to the jaws of atherinoids, they are even more highly protractile and possess two major and two minor paired bones found so far in no other fishes except the closely related Neostethus (probably also present in SolenophaUm and Plectrostethus and possibly Phallo- stethus, but absent in PJienacostethus and Guhphallus). No observations have evei
Ceratostethus — Osteology and Structural Novelties • Roberts 395
been made on the feeding movements of Ceratostethus or any other phallostethoids; I have tried to detemiine the main features of their functional anatomy by manipula- tion of alizarin-stained specimens macerated in potassium hydroxide and cleared in 50 percent glycerine.
My initial objectives in studying the osteology of Ceratostethus were to provide information that might lead to a better understanding of 1 ) relationships of Phallo- stethoidea to other fish groups, and 2) relationships among phallostethoid genera. I chose Ceratostethus because, judging from priapial stiiicture, it seemed to repre- sent a relatively primitive phallostethoid, and I had an abundant supply of fresh material from Thailand. Regan (1913; 1916) regarded the phallostethoids as an aberrant subfamily of Cyprinodontidae, and noted that their osteology was "typically cyprin- odontid." He did not indicate to which cyprinodontids they might be most closely related. Subsequent to the discovery by Herre ( 1925 ) that phallostethoids possess a spiny (al]:)eit minute) first dorsal fin, most ichthyologists followed Myers ( 1928 ) in relating them to the Atherinidae. Myers intimated that Atherinidae and Cyprin- odontidae might actually be fairly closely related, a suggestion generally viewed with disfavor (cf. Hubbs, 1944) until Rosen (1964) pointed out a large number of anatomical, morphological, and reproduc- tive characters and trends commoj! to atherinoids, cyprinodontoids, and allied forms (including Phallostethoidea ) and united them in a new order, Atherinifonnes. While placing the superfamily Phallo- stethoidea in the suborder Atherinoidei {ibid.: 261), he suggested that it might be more closely related to cyprinodontoids than to atherinoids after all {ibid.: 242). During the course of the present study numerous supplemental observations were made on cyprinodontoids and atherinoids. Phallostethoids apparently are more closely related to atherinoids than to cyprinodon-
toids. The most closely related forms ap- pear to be Stenatherina and its Indo-Pacific allies (placed by Schultz, 1948, in the atherinid subfamily Taeniomembrasinae ) . If one considers only zoogeographic dis- tribution and priapial morphology (which has been worked out in considerable detail for almost all of the genera by Bailey [1936] and especially by Aurich [1937]), a rel- atively simple picture of phyletic relation- ships \\'ithin Phallostethoidea emerges. There are two families: Phallostethidae and Neostethidae. Phallostethidae, presumably most highly modified from the primitive or ancestral type of phallostethoid, com- prise two genera, Phallostethus and Fhena- costethus, confined to the Malay Peninsula and adjacent parts of the mainland of Southeast Asia. Neostethidae comprise six to eight genera (Herre, 1942, probably went too far in splitting genera) belonging to two subfamilies, Neostethinae and Gula- phallinae. Gulaphallinae are restricted to the Philippines. Neostethinae, which in- clude Ceratostethus, SoJenophaUus, Plectro- stethus, and Neostethus, are more widely distributed. Neostethus and Ceratostethus occur in the Philippines, Borneo, and both sides of the Malay Peninsula.^ The priapium of Neostethinae, in which the only externalized elements are derised from pelvic spines and rays, is evidently more primitive than are the priapia of Phallo- stethidae and Gulaphallinae. In phallo- stethid priapia the main externalized element is homologous witli the neostethid outer pulvinular (Bailey, 1936; Roberts, 1971), which is the anteriormost internal element in the priapium of Neostethidae. Neostethus, alone among Neostediidae, has a single spur near the base of its cten- actinium, resembling the series of spurs on the ctenactinium of PhaUostethus (the ctenactinium of Phenacostethus is absent or
^ There are two records of indeterminalile phallostethoids: from Sabang, northwestern Su- matra (Aurich, 1937: 282-284) and from Zamboanga (Banjagan, 1966: 46).
396
Bulletin Museum of Comparative Zoology, Vol. 142, No. 4
greatly reduced). In addition, the outer puK'inular bone of Neostethus has a small but distinct lateral projection, immediately anterior to the pulvinulus (absent in other neostethids), which might be an incipient or rudimentary toxactinium. Regan (1916: 6, fig. 3) referred to this projection as the pulvinular spine. It would be of great interest to compare the osteology of Neo- stethus and PhaUostethus. Unfortunately, Phallostethus is represented in collections by only four type specimens in poor con- dition. The priapial characteristics of Phenacostcthus have been treated in detail elsewhere (Roberts, 1971). Both Gulo- })luillus and Cerato.stethus have been de- scribed as having two ctenactinia. The second ctenactinium of Cerato.stethus, however, is evidently a modified pelvic spine or ray, whereas that of Gulaphallus is an externalized pelvic bone. In addition, ill (Tulaphallinae {Gulaphallus and Miro- phallus) the anterior end of the cten- actinium fits into a fleshy sheath in the anterior end of the priapium, whereas in Ncostethinae it remains uncovered (see Aurich, 1937: 266, figs. 1 and 2).
The discovery of neomorphic jaw bones in Ceratostethus and Neostethus adds a new dimension to the above sketch. Their presence in these genera and their absence in CAilaphallus indicate the distinctness of the subfamilies Ncostethinae and Gula- phallinae and make direct derivation of Gulaphallinae from Ncostethinae harder to conceive. Absence of the neomorphic ele- ments is perhaps to be expected in the minute species of Phcnacostethtis; they may well be present in PhaUo.stethus. Their absence in Gulaphallus, in which the jaws are distinctly less protractile than in Neo- stethus, Ceratostethus, and Phenacostethus, indicates that more diversity exists among Phallostcthoidea than might otherwise have been thought.
The intergeneric relationships of Phallo- stcthoidea, as currentlv understood, mav be represented in a diagram as follows:
Phenacostethus
Solenophallus
Phallostethus
Manacopus
Plectrostethus
/ Neostethus
Gulaphallus
As previously indicated (Roberts, 1971), I am inclined to believe that the order Atheriniformes, as conceived by Rosen (1964), constitutes a natural assemblage. Thus far my researches have not uncovered any facts that cast serious doubt on this concept. Although Rosen (1964: 255) stated that the circumorbital series in Atherinifonnes is represented only by lacrimal and dermosphenotic (lacrimal with separate ventral piece in a few cases), in some phallostethoids and atherinids there is a large, trough-shaped infraorbital bone immediately beneath the lacrimal, and in some atherinids {Melanotaenia and Telmatherina) there are two separate, troughlike or laminar infraorbitals below the lacrimal. Rosen (p. 288) found in melano'taeniids a small, spatulate element broadiv and firmlv joined to the ventral surface of the lacrimal, and noted that these two bones in melanotaeniids together re- semble the single elongate lacrimal of I Xenopoecilus. The second infraorbital bone (considering the lacrimal as the first infraorbital) in phallostethoids and ather- inids observed by me is quite separate from the lacrimal. A first pharyngobranchial is present in Cerato.stethus, Melanotaenia, andi Allanetta, although Rosen (p. 237) stated! that the first pharyngobranchial is lacking' in Atheriniformes (Melanotaenia indicatedl as a possible exception). The Atherini- formes are diverse, and it is understandable that as additional representatives are studied more thoroughly, definitions will
% i f f
■U. :,i^ .'« ^
.,i-eM»T, ./.
^-
■'?.'-
Ceratostethus — Osteology and Structural Novelties • Roberts 397
* ve to be revised and expanded. It is un- I niable that certain evokitionarv trends,
r instance those concerning reproduction, ; 3m to be more readily understandable if 1 3 groups placed in Atheriniformes are : deed related. Thus some observations on ■ ? widespread occurrence of bi- and tri-
• spid teeth in Atheriniformes, which seem
confirm their relationship, are reported
this paper even though the dentition of
allostethoids is entirely conical. My own
: iling, however, is that the concept of
. herinifonnes should be verified by
; jrougli osteological comparisons among
I' 3 primitive or amiectant representatives
I the suborders assigned to it.
V^illadolid and Manacop ( 19.34 ) analyzed
; tmach contents throughout one year in
' Id-caught GuJaphalJus mirahilis, and
I served courtship, copulation, and egg-
dng of this species in aquaria. These
Dects of behavior have not been observed
any other phallostethoids. No new spe-
I 's of Neostethidae have been described
: ice the genera and species were reviewed
1 Herre (1942). I reviewed the Phallo-
: 'thidae, described a new species of
i enocostethus from the Indian Ocean
< ast of Thailand, and remarked upon 1 3 ecology of Phallostethidae and Neo- ! 'thidae in Thailand. I also discussed the ilective advantage of internal fertihzation i Phallostethoidea as well as trends in the ] productive biology of atheriniforms that i-ght be conducive to the evolution of i:emal fertilization (Roberts, 1971). The <5Cussion section of the present paper
< als with evolutionary topics such as the <igin of the neomorphic jaw bones found i Ceratostethus and Neosfethus, neotenic " aracters in phallostethoids, and the : ture of selection pressures for and against
s development of highly complicated ' temal genitalia in atheriniforms.
IKNOWLEDGMENTS
The help rendered during my stay in lailand by Dean Jinda Thiemeedh, Mr.
Prajit Wongrat, and Mrs. Supap Monkol- prasit, of the College of Fisheries, Kasetsart University; by Miss Prachuab Sukcharean and Mr. Sopon Chantarat, of the Marine Fisheries Station of Songkhla; and by Dr. Vagn Hansen, of the Phuket Marine Bio- logical Center, has already been acknowl- edged (Roberts, 1971): without it the observations and collections of phallo- stethoids that provide the basis for my studies could not have been made. My fieldwork was approved by the National Research Council of Thailand.
Miss Deborah White and Dr. Elizabeth Deichmann kindly translated the papers by Aurich and Woltereck. Much of the in- spiration for the discussion section of this paper came from an essay on "The Emer- gence of Evolutionary Novelties," presented by Prof. Ernst Mayr at the Centennial Celebration of Darwin's Origin of Species in London (Mayr, 1960). Acknowledg- ment is gratefully made to Prof. George S. Myers. Prof. M>'ers acquainted me with phallostethoids when I was an undergradu- ate at Stanford University, and we have discussed them on numerous occasions.
OSTEOLOGY OF CERATOSTETHUS (Figures 2-12)
The following osteological account is based on 14 specimens (ten males and four females) 17.0-26.8 mm in standard length, part of a large series collected in a roadside ditch shaded by nipa palm about two miles from Tha Chalap on the road to Chantaburi Cit\', Chantaburi Province, Tliailand (MCZ' 47304). The only osteo- logical differences detected between males and females involve genitalia, and the anteriormost ribs and the cleithrum ( modified in males to fonn a suspensorium for the priapium). The figures are based on four specimens, 24.8-26.8 mm. A few observations are included on the osteology of Phenacosfethus, Neostethus, and Gula- pliaUus. Comments and comparative obser- vations on cyprinodontoids and atherinoids
396
Bulletin Museum of Comparative Zoology, Vol 142, No. 4
greatly reduced). In addition, the outer pulvinular bone of Neostetlms has a small but distinct lateral projection, immediately anterior to the pulvinulus (absent in other neostethids ) , which might be an incipient or rudimentary toxactinium. Regan (1916: 6, fig. 3) referred to this projection as the pulvinular spine. It would be of great interest to compare the osteology of Neo- stetlms and PhaUostethtis. Unfortunately, PhoUostethtis is represented in collections by only four type specimens in poor con- dition. The priapial characteristics of Plienacostethus have been treated in detail elsewhere (Roberts, 1971). Both Gula- pliallus and Ceratostethiis have been de- scribed as having two ctenactinia. Tlie second ctenactinium of Ceratostetlius, however, is evidently a modified pelvic spine or ray, whereas that of GidaphaUus is an externalized pelvic bone. In addition, in Gulaphallinae {GiilaphaUiis and Miro- pluiUus) the anterior end of the cten- actinium fits into a fleshy sheath in the anterior end of the priapium, whereas in Neostethinae it remains uncovered (see Aurich, 1937: 266, figs. 1 and 2).
The discovery of neomorphic jaw bones in Ceratostethiis and Neostetlms adds a new dimension to the above sketch. Their presence in these genera and their absence in GidaphaUus indicate the distinctness of the subfamilies Neostethinae and Gula- phallinae and make direct derivation of Gulaphallinae from Neostethinae harder to conceive. Absence of the neomorphic ele- ments is perhaps to be expected in the minute species of Phenacostethtis; they may well be present in Phallosiethus. Their absence in GidaphaUus, in which the jaws are distinctly less protractile than in Neo- stethus, Ceratostethiis, and Phenacostethtis, indicates that more diversity exists among Phallostethoidea than might otherwise have been thought.
The intergeneric relationships of Phallo- stethoidea, as currently understood, may be represented in a diagram as follows:
Plienacostethus
Phallostethus
Solenophallus
Manacopiis
Plectrostethus
Gulaphallus
Neostelhus
As previously indicated (Roberts, 1971 I am inclined to believe that the ord Atheriniformes, as conceived by Ros( ( 1964 ) , constitutes a natural assemblag Thus far my researches have not uncoverc any facts that cast serious doubt on tl concept. Although Rosen (1964: 25J stated that the circumorbital series Atheriniformes is represented only 1 lacrimal and dermosphenotic (lacrim with separate ventral piece in a few cases in some phallostethoids and atherini there is a large, trough-shaped infraorbit bone immediately beneath the lacrim. |i and in some atherinids {Melanotaenia a]| Telmatherina) there are two separa | troughlike or laminar infraorbitals belt ' the lacrimal. Rosen (p. 288) found |! melano'taeniids a small, spatulate elemel broadly and firmly joined to the ventif surface of the lacrimal, and noted that the | two bones in melanotaeniids together ij semble the single elongate lacrimal I Xenopoecihis. The second infraorbital bo | (considering the lacrimal as the fi infraorbital) in phallostethoids and ath» inids observed by me is quite separate frc the lacrimal. A first pharyngobranchial S present in Ceratostethiis, Melanotaenia, a I AU-anetta, although Rosen (p. 237) stati that the first pharyngobranchial is lacki| in Atheriniformes (Melanotaenia indicatJ as a possible exception). The Atheri- formes are diverse, and it is understandale that as additional representatives ie studied more thoroughly, definitions vU
M.^
« *•
Ceratostethus — Osteology and Structural Novelties • Rohctis 397
have to be revised and expanded. It is un- deniable that certain evolutionarv trends, for instance those concerning reproduction, seem to be more readily understandable if the groups placed in Atheriniformes are indeed related. Thus some observations on the widespread occurrence of bi- and tri- cuspid teeth in Atherinifonnes, which seem to confirm their relationship, are reported in this paper even though the dentition of phallostethoids is entirely conical. My own feeling, however, is that the concept of Atherinifonnes should be verified by thorough osteological comparisons among the primitive or annectant representatives of the suborders assigned to it.
Villadolid and Manacop (1934) analyzed stomach contents throughout one year in wild-caught GiilaphalJiis mirabilis, and observed courtship, copulation, and egg- laying of this species in aquaria. These aspects of behavior have not been observed in any other phallostethoids. No new spe- cies of Neostethidae have been described since the genera and species were reviewed by Herre (1942). I reviewed the Phallo- stethidae, described a new species of Phenocostethus from the Indian Ocean coast of Thailand, and remarked upon the ecology of Phallostethidae and Neo- stethidae in Thailand. I also discussed the selective advantage of internal fertilization in Phallostethoidea as well as trends in the reproductive biology of atheriniforms that might be conducive to the evolution of internal fertiUzation (Roberts, 1971). The discussion section of the present paper deals with evolutionary topics such as the origin of the neomorphic jaw bones found in Ceratostethus and Neostethus, neotenic characters in phallostethoids, and the nature of selection pressures for and against the development of highly complicated external genitalia in atherinifomis.
ACKNOWLEDGMENTS
The help rendered during my stay in Thailand by Dean Jinda Thiemeedh, Mr.
Prajit Wongrat, and Mrs. Supap Monkol- prasit, of the College of Fisheries, Kasetsart University; by Miss Prachuab Sukcharean and Mr. Sopon Chantarat, of the Marine Fisheries Station of Songkhla; and by Dr. Vagn Hansen, of the Phuket Marine Bio- logical Center, has already been acknowl- edged (Roberts, 1971): without it the observations and collections of phallo- stethoids that provide the basis for my studies could not have been made. My fieldwork was approved by the National Research Council of Thailand.
Miss Deborah White and Dr. Elizabeth Deichmann kindly translated the papers by Aurich and Woltereck. Much of the in- spiration for the discussion section of this paper came from an essay on "The Emer- gence of Evolutionary Novelties," presented by Prof. Ernst Mayr at the Centennial Celebration of Darwin's Origin of Species in London (Mayr, 1960). Acknowledg- ment is gratefully made to Prof. George S. Myers. Prof. M>'ers acquainted me with phallostethoids when I was an undergradu- ate at Stanford University, and we have discussed them on numerous occasions.
OSTEOLOGY OF CERATOSTETHUS (Figures 2-12)
The following osteological account is based on 14 specimens (ten males and four females) 17.0-26.8 mm in standard length, part of a large series collected in a roadside ditch shaded by nipa palm about two miles from Tha Chalap on the road to Chantaburi Citv, Chantaburi Province, Tliailand (MCZ' 47304). The only osteo- logical differences detected between males and females involve genitalia, and the anteriormost ribs and the cleithrum ( modified in males to form a suspensorium for the priapium). Tlie figures are based on four specimens, 24.8-26.8 mm. A few observations are included on the osteology of Phenacostethus, Neostethus, and CaiIq- phalhis. Comments and comparati\e obser- vations on cyprinodontoids and atherinoids
398 Bulletin Museum of Comparative Zoology, Vol. 142, No. 4
pararostral ^ OS max. -mand. dentary
maxillary
submaxillary palatine
rostral
frontal
sphenotic pterotic
exoccipital
first vertebr;
supraoccipita epiotic
dermosphenotic Figure 2. Ceratosfethui bicornis. Dorsal view of skull and first vertebra; jaws sligfitly protruded.
are inserted in numerous places. I have Foster) and atherinoids (including alizarii
examined alizarin preparations of a con- preparations in the Department of Ichthyol
siderable variety of cyprinodonts (a num- ogy of the American Museum of Natura
ber of them kindly provided by Neal History). The osteology of atherinids ii
nasal
pterosphenoid
parasphenoid prefrontal
dermosphenotic sphenotic
prtmaxillary
pterotic
exoccipital
basioccipital
first vertebra
posttemporal
prootic
supracleithrum
maxillary
hyomandibular
Figure 3. Cero/osfefhus b/corn/s. Ventral view of cranium and first vertebra; witfi portions of jaws, suspensorium, one sfioulder girdle of right side.
Ceratostethus — Osteology and Structural Novelties • Roberts 399
dermosplienotic
maxillary frontal pterosphenoid
rostral i „;isal
lacrimal
supraoccipital
epiotic exoccipital
paradentary
dentary premaxillary OS max. -mand.
maxillary
articular infraorbital
angular
quadrate
interopercle
preopercle
symplectic
interhyal
opercle
subopercle
Figure 4. Ceratostethus bicornis. Lateral view of si<uil, first vertebra, and part of pectoral girdle.
the subfamily Taeniomeinbrasinae shows much in common \\'ith that of Cerato- stethus. The parts of the skeleton are taken up in the following order: cranium, jaws, suspensorium, facial bones, hyoid and gill arches, pectoral girdle, median fins, vertebral column, and priapium.
Cranium (Figs. 2-4). Bones in cranium: prevomer, mesethmoid, prefrontals, front- als, parasphenoid, pterosphenoid, sphenotic, prootic, pterotic, epiotic, supraoccipital, exoccipitals, and basioccipital. Basisphe- noid, parietals, and intercalar absent.
Anterior end of each side of prevomer enlarged and rounded to form a joint sur- face with submaxillary bone. In some specimens the prevomer is a single ossifi- cation. In others ( including the ones in Figs. 2-4) it is ossified in two separate parts, one on each side of the mesethmoid. Mesethmoid a laminar bone of irregular hexagonal fonn. In some atherinids the mesethmoid is a thin lamina. (In cyprin- odontoids mesethmoid either cartilaginous or a bilaminar ossification.)
Roof of skull convex, with a deep V- shaped trough between main body of frontal and supraorbital lamina; frontal bones depressed anteriorly where they meet prefrontals to fomi posterior floor of rostral fossa. In these features the roof of the skull resembles that of various atherinids; cyprinodontoids usually (always ?) have a uniformly flattened or only very slightly convex roof, without a trough between main body of frontal and supraorbital lamina. Skull roof smooth, without bonv canals for cephalic laterosensory system; wings or crests absent excepting poorly developed (vestigial ?) supraoccipital crest. Supraoccipital spine undivided. Supra- occipital spine usually (invariably ?) un- divided in atherinoids, bifid in cyprin- odontoids (personal communication from D. E. Rosen). Dorsal border of foramen mag- num formed exclusixely by exoccipitals.
Jaws (Figs. 5-7). Although bearing strong morphological and functional re- semblance to the highly protractile jaws of certain atherinids, the jaws of Cerato-
400 Bulletin Museum of Comparative Zoology, Vol. 142, No. 4
Figure 5. Cera/os/efhus bicornis. Dorsal view of anterior part of skull witfi jaws greatly protracted. Compare Fig. 6C.
stethus are perhaps even more specialized, for they include a number of hitherto un- reported ossified elements, some of which are without knowai osseous homologues in atherinoids or cyprinodontoids or even analogues in other teleosts. These evidently neomorphic structures, apparently present only in phallostethoids, are functionally in- volved in the extreme protrusibility of the jaws (evidently including closure of the jaws in protruded position). The new bones, discussed below, are paired ele- ments, named as follows (in order of oc- currence anteroposteriorly ) : paradentary, maxillomandibular, pararostral, and sub- maxillary bones.
Bones in upper jaw: paired premaxil- laries, maxillomandibularies, maxillaries, pararostrals, and submaxillaries; in lower jaw: dentaries, paradentaries, articulars, angulars, and coronomeckelians. Premaxil- laries and dentaries with a single row of conical teeth; other jaw bones and palate toothless. Proximal third of premaxillary with eight to ten relatively small teeth, middle third with about ten enlarged teeth, distal third toothless. Dentaries with three to five very small (vestigial ?) teeth near symphysis; otherwise toothless. Ascending premaxillary processes slender, elongate
(their length almost equal to a third of cranial length).
Posteriorly directed, broad-based, roundec 1 processes ("articular processes") arising at midlength of prcniaxillaries; sucl processes (either rounded or pointed) characteristic of many atherinoids, are usu ally (invariably ?) absent in cyprin odontoids. Several atherinoids have slender elongate ascending processes as in Cerate stcthiis; ascending processes in cyprin- odontoids usually (invariably ?) relativel) short, broad-based, and wide. Maxillar) with well-developed, strongly curved in-j ternal hooks. As in atherinoids (but not iri cyprinodontoids) the maxillary has i\ laminar projection dorsal to the interna' hook, part of which projects dorsally to th( ascending premaxillary processes (wherj mouth is abducted). Lower (distal) limb of maxillary uniformly slender. A swelling (cranial condyle) at junction of interna hook with the main body of maxillar) provides articular surface for submaxillar}^ bone. Maxillary without ligamentous ati tachment to palatine. In atherinoids and ii some (all ?) cyprinodontoids maxillary articulates directly to prevomer or to the ethmoid by a submaxillary meniscus. Ii atherinoids and cyprinodontoids there i usually (always ?) a direct ligamentou
Ceratostethus — Osteology and Structural Novelties • Roberts 401
connection between palatine and maxillary In the following account I attempt to at or near cranial condyle. Head of maxil- give an idea of functional (mechanical) re- lary bound in place, probably by liga- lations of the jaw bones even though their mentous connections with nasal, lacrimal, movement was studied only in macerated and perhaps mesethmoid which, however, specimens. I have assumed that in live permit swinging and screwing movements, specimens the mouth is opened mainly by Lower (free) limb of maxillary unifonnly depression of the lower jaw; that return of slender. Internal hooks attached by dense the jaws to normal "resting" position in- connective tissue extending ventrally to volves two distinct, consecutive stages, rostral bone. Rostral connective tissue namely 1 ) closing of the mouth while it is (ethmopremaxillary ligaments ?) with a still protracted, accomplished by retraction pair of very small pararostral bones dorso- on the lower limb of the maxillary, drawing lateral to anterior part of rostral bone. Near free ends and premaxillaries back (while tip of lower limb of premaxillary, in the position of lower jaw is unaffected), and maxillomandibulary ligament lies the small, 2) inoving the entire assemblage back into rounded, dorsoventrally compressed maxil- the nearly vertical position it occupied be- lomandibulary bone. The maxillomandibu- fore, probably accomplished by retraction lary ligament connects the ends of the lower on the coronoid process of the dentary and or free limbs of the premaxillary and maxil- on the "articular process" of the premaxil- lary to the dentary. The mouth is opened lary. From the work of Alexander ( 1966, when movement originated by depressing 1967a, b ) it seems likely that in many fishes the lower jaw is transferred by this liga- with protractile jaws, the mouth is closed ment to the upper jaw. The attachment of while the jaws are protracted. It is the ligament to the dentary extends from unclear whether the mouth can be closed middle of dentary to symphysis of lower when it is protracted in atherinoids, but it jaws. Near symphysis main body of liga- can be in some cyprinodontoids, as well as ment very tough and round in cross sec- in cyprinoids, and acanthopterygians ( Alex- tion. In shape and extent of attachment the ander, 1967a, b).
ligament resembles strongly that in certain Wlien mouth is in resting position, as-
atherinoids, particularly members of the cending premaxillary processes fit snugly
subfamily Taeniomembrasinae (for defini- into rostral concavity, their distal (free)
tion and discussion of this taxon see ends lying between depressed anterioiTnost
Schultz, 1948). In other atherinids and in portion of frontal bones. When jaws are
cyprinodontoids the attachment of the liga- maximally protracted ( as estimated by
ment does not extend much anterior to the manipulation of KOH-macerated speci-
middle of the dentary, and is spread out or mens) premaxillaries have moved forward
sheetlike. The anterior part of the liga- ^bout 125 percent of their length, or about
ment in Ceratostethus differs strikingly 35 percent of the headlength.
rrom that in Taeniomembrasinae in that \t ■^^ ■ c . • n .1 • i r
1 . ,. 1 . , . . . . Maxillanes of typically athermoid tonn,
lymg entirely witnm it is a very conspicuous 1 i i r 1 ..1 n i
^„ 1 r .11 • r- . r- . • .1 excluded from gape and with well-de-
and tunctionally significant ossification, the 1 i 1
paradentary bone. Distal end of paraden- ^'^1^?^^ internal hooks. With mouth com- tary lies free in the ligament; its proximal P^^tely closed, straight, slender ventral end foi-ms the ball of a ball and socket joint ( distal) limb of maxiUaries Hes at an angle with the dentary at anterionnost point of of about 10 degrees posterior to a vertical; attachment of ligament, i.e., near symphysis with mouth fully protracted, ventral limb of lower jaw. Dentary with a high coronoid of maxillary at an angle of about 45 de- process, as in many atherinids; coronoid grees anterior to a vertical (see Fig. 6, A-C). processes usually absent in cyprinodontoids. Screw movement is slight or nonexistent
402 BuUetin Museum of Comparative Zoology, Vol. 142, No. 4
Figure 6. Cerofosfefhus bicornh. Lateral view of jaws in varying stages of protrusion (for explanation see text).
until after more than half of the swinging movement is completed and the premaxil- laries are considerably protracted. Up to this point movement of the premaxillaries is equal to that of the rostral; from this point on, however, the premaxillaries move rel- atively further forward than do the rostrals. Thus not only does the rostral slide forward relative to the cranium, but the premaxil- lary processes slide forward relative to the rostral; Alexander ( 1967b ) found a similar movement present in Atherina but absent in the cyprinodontids Aplocheilus and Fundulus. Alizarin preparations of various
cyprinodontids examined by me failed to exhibit this movement. The screw move- ment evidently also causes dilation of the lower limbs of the premaxillary, causing the lateral series of enlarged premaxillary teeth (which are directed sideways when the mouth is in resting position ) to point straight ahead.
As stated by Alexander (1967b: 241) protrusion of the jaws must be accompanied by screw movement of the maxillary, sc that the internal hooks and consequenth rostral bone are moved anteriorly. In Ceratostethiis, however, at least some pro- trusion occurs before the screw movemeni begins. Furthermore, it is possible to causf considerably more protrusion (about 21 percent more) after the rostral has ceasec to move forward by simply continuing t( depress the lower jaw. If this is done, the premaxillary tilts upwards posteriorly sc that the mouth opening is slightly down- wards instead of vertical. Wliile it i; doubtful that the jaws are normally pro truded so far, this observation indicate; that depression of the jaws can act un accompanied by screw movements of tht maxillary to bring the premaxillaries for ward.
In cyprinodontoids the screw movemen - of the maxillary causes not only anterio: movement of the internal hooks, but alsc lateral movement (Alexander, 1967b: 239 fig. 5). I have observed this in mam cyprinodontoids. In Ftindiihis the rostra cartilage is Y-shaped, with the interna hooks attached finnly to the ends of th( arms of the Y; as the internal hooks mov(! laterally, the amis of the Y spread apart In addition, in many (but not all) cyprin, odontoids, the internal hooks are so fimib bound to the ascending premaxillary tha^ when the hooks move laterally, the ascend ing processes diverge posteriorly (see Alexi ander, 1967b, fig. 6e of Futululus). h\ Ceratostethus there is either no latera movement or very slight lateral movemen of the internal hooks; the ascending pre
Ceratostethus — Osteology and Structural Novelties • Roberts 403
maxillary processes never diverge from each other.
Submaxillary process of maxillary articu- lated to prevomer by submaxillary bone (developed in place of the usual submaxil- lary meniscus). Submaxillary bone forms ja meniscus with submaxillary process of .maxillary and with anterior end of pre- v^omer. When mouth is closed, submaxillary bone lies with its upper end at an angle of ibout 15 or 20 degrees anterior to a vertical; when mouth is protracted, rotation Df maxillary depresses upper end of sub- naxillary, which finally lies at an angle of ibout 45 degrees from vertical with mouth ully protracted. The curved, ventrally di- rected internal hooks of the maxillaries are igamentously attached to a median ossified element or rostral bone lying ventrally to iscending processes of premaxillaries. ( This slement evidently is represented by a •QStral cartilage in atherinoids; it is similar n position, but not in movement, to the 3yprinoid kinethmoid. ) With mouth closed, interior tip of rostral is horizontal with the interior tips of the nasal bones and an- erior margins of maxillaries; when mouth s fully protracted, it moves forward ap- proximately 50 percent of its length. Above :he rostral, in thickened connective tissue attaching it to ascending premaxillary proc- esses and internal hooks of maxillary, lies a lair of small bones or pararostrals. Tliese ire evidently equivalent to accessory rostral :;artilages of some atherinoids. Movement )f pararostrals equal to movement of ostral.
Lower jaws similar in form to those in itherinoids: dentaries with very large cor- )nary process, and with ventrally opened lange forming a trough for mandib- ilar segment of cephalic laterosensory iystem. Maxillomandibulary ligament, link- ng tips of the ventral limb of maxillary and ,)remaxillary to dentary, is similar to that In atherinoids (but not to that in cyprin- )dontoids ) in that its connection to dentary extends anteriorly to ramus of lower jaws, n the maxillomandibulary ligaments of
Figure 7. Ceratostethus bicornis. Hypothetic closure of protracted jaws (for explanation see text).
Ceratostethus are two paired bones: the small maxillomandibulary bones, lying free in that part of the ligament connecting premaxillary and maxillary, and the elon- gate paradentaries, which are attached by a sort of ball and socket joint to the dentaries near ramus of lower jaw. Move- ment of the maxillomandibulary bones is dependent on movement of the premaxil- lary; their position with respect to lower limb of premaxillary remains unchanged. Maxillomandibulary bone forming conspic- uous rounded prominence in angle of rictus of jaws when mouth fully opened (Fig. 6c). Paradentaries, with longitudinal axis parallel to that of dentary, lie snugly against dentaries in grooves; base of groove fonned by dorsal surface of bony canal for mandibular segment of cephalic latero- sensory system. With mouth about half open, paradentaries remain snugly against dentary; as mouth opens further, their distal (free) end swings sideways and up- wards until, with mouth fully protracted, they are at right angles to dentary (Fig. 6c). When mouth closes in protracted position (closing accomplished in macerated specimens by pushing back on lower limb of maxillary), the paradentaries return to lie flush with side of dentaries, while only the lower limb of the premaxillary is retracted; the ascending premaxillary proc- esses remaining as far or almost as far forward as they are when mouth is fully
404 BuUetin Museum of Comparative Zoology, Vol. 142, No. 4
opened (Fig. 7). Tims closure is ac- in the neostethid Guhphalhis in which the
complished almost entirely by movement upper jaws do not protract very far, the
of the upper jaw. The lower jaw remains mandible must be depressed further before
depressed, the lower jaw bones, excepting the upper jaw moves fonvard at all than
the paradentary, undergoing little or no when the upper jaw is fully protracted in
movement. Ceratostethm. Second, the paradentaries
The paradentaries evidently are involved do not rotate outward until the upper jaw
in increased protrusibility of the upper jaw. is considerably protracted, and only slightly
In specimens in \\hich the maxillomandibu- more forward movement of the upper jaw
lary ligaments have been cut immediately is involved for the paradentaries to rotate
posterior to the distal end of the para- as far outward as in Figure 5. It seems tc
dentaries, depression of the lower jaw still me highly likely that the paradentaries dc
causes premaxillary protraction, but the in fact rotate this much. Tliird, when one
prcmaxillaries do not project as far an- depresses the lower jaw of a macerated
teriorly and they project upwards (with specimen, the entire jaw assemblage move^
leading margin considerably elevated) in- readily and smoothly as far forward as ir
stead of straight forward. The paradentaries Figure 5. In particular, the motion of the
also help spread the gape sidewards as the maxillary is unencumbered. This contrast.'
mouth is opened. with GuIaphoUus and with various ather
Comments. Two objections that might inoids in which movement of the maxillar) be raised to this description of jaw function is relatively restricted, are 1 ) whether it is normal for the jaws to Whether closure of the jaws occurs ir be so greatly protracted, and 2) whether the manner indicated is much less sure. II closure actually occurs in the manner sug- seems likely that the jaws are closed ir gested. There are some fishes (including protracted position. Granted this happens Morwcirrlms among Nandidae [Liem, perhaps stages 1 and 2 envisioned in re- 1970], Epibiihis- among Labridae, several turning the jaws to resting position act ir atherinoids) with protractile upper jaws concert rather than consecutively. It ma>! and elongate ascending premaxillary proc- be that the mouth never closes with the esses in which the ascending premaxillary upper jaw fully protracted (the positior processes are advanced no more than a indicated in Fig. 7). It should be noted; fraction (a half in Epibitlns: about a third however, that the jaws of macerated speci- or less in Monoc/r;7jt/5 and the atherinoids ) mens are readily manipulated into thi5 of their length when the jaws are fully position, and that from this position the> protracted. In Monocirrhus and Epihulus can be readily manipulated back to resting (in which the lower jaws as well as the position, i.e., closed and completely re- upper are protractile) the ascending pre- tracted.
maxillary processes are exceedingly elon- Liem (1970: 106), considering fishes
gate — as long or longer than the total ^^ith protractile jaws, stated that the salieni
cranial length. The ascending premaxillary features of feeding (and respiratory)
processes of Ceratostethm are relatively movements are 1) opening and closing ol
short compared to those of Monocirrhus the jaws, 2) protrusion of the jaws, 3)|
and Epihidus: their length is about one- volume and pressure variations in the'
third of cranial length. Three facts indicate buccopharyngeal cavity, 4) abduction and'
that the upper jaw is protracted as far or adduction of the opercular apparatus, and'
nearly as far forward as shown in Figure 5. 5) complex movements of the gill arches
^ irst, protraction of the upper jaws this far He pointed out that these functions an is achieved by only moderate depression of mutually interdependent and should b( the lower jaw. In various atherinoids and considered in conjunction. Assessment ol
1
Ceratostethus — Osteology and Structural Novelties • Roberts 405
the movements of the opercular apparatus and gill arches and their relationships in Cerotostethti.s has not been attempted, since "natural" movements of these parts are less readily achieved (or achieved with less confidence) by manipulation of macerated specimens. The main, or at least a major, advantage of protractile jaws may lie in increasing the buccal component of the buccopharyngeal pumping mechanism; one of the advantages in closing the mouth with the premaxillaries protracted probably lies in increasing the volume of water that can be sucked into the mouth without being forced back out as the mouth closes (see more detailed discussion in Alexander, 1967a: 59-62). It seems likely that these factors operate in Ceratostethus.
Suspensorhim (Fig. 4). Bones in suspensorivim : palatine, ectopterygoid, entopterygoid, quadrate, symplectic, and hyomandibular. Anterodorsal part of pala- tine slender, tubular, apparently connected by ligaments to prefrontal, but free of at- tachment to nasal, lacrimal, and maxillary. Movement of jaws fails to induce move- ment in palatine. Anterior margin of suspensorium (formed by palatine, ecto- pterygoid and quadrate) almost vertically inclined. Ventral portion of hyomandibular with processes contacting symplectic and preopercle. Symplectic with a ventrally directed laminar portion.
Facial bones (Figs. 2-4). Nasal well developed, almost as long as orbit, a dor- sally projecting flange on its medial margin immediately lateral to depressed anterior- most portion of frontal, and a ventrally directed flange at its anterior tip that is intimately associated with anterodorsal comer of lacrimal bone. Circumorbital series comprised of lacrimal, a single infra- orbital directly beneath it, and denno- sphenotic. Lacrimal a large lamellar plate with a flange in its dorsoposterior comer forming a gutter for a segment of cephalic laterosensory canal. Ventral to lacrimal a single, concave infraorbital, forming a bony trough for a segment of cephalic
i;ill rakers
liypobrunchial
basibranchial glossohyal
epibranchial
ceratobranchial interhyal
liypohyal ceratohyal
brancliioste't^al rays
Figure 8. Ceratostethu% bicornis. Lateral view of hyoid arch, urohyal, and first brancfiial arcfi. Hyoid and branchial arches slightly separated.
laterosensoiy canal. Dermosphenotic simi- larly troughlike, its attachment anterior to sphenotic (attachment of dermosphenotic posterior to sphenotic in Horaichthys and Onjzias). Preopercle with a ventrally open flange fonning a trough for a segment of cephalic laterosensory canal. Opercle with- out spiny projections, its posterior margin concave as in some atherinids; in many ( all ?) cyx^rinodontoids posterior margin of opercle broadly rounded. Posterior margin of gill cover formed by opercle, not by subopercle. Anterodorsal comer of opercle where it attaches to hyomandibular rel- atively unmodified, without a strongly reinforced socket to receive hyomandibular process. Interopercle and subopercle weakly ossified. Anterodorsal projection of subopercle weakly developed; posteriorly subopercle projects considerably beyond ventral margin of opercle.
Hyoid and gill arches (Figs. 8 and 9). Hyoid arch of typical atherinifonn structure, with a single liypohyal; five branchiostegal rays on each side, first four branchiostegal rays attached to ceratohyal, last one to epihyal; dorsal half of joint between ceratohyal and epihyal strongly ossified; ceratohyal with a concave antero- ventral margin. Gill arches with three median basibranchials, four pairs of h\^o-
406
BuUetin Museum of Compnrotwc Zoologij, Vol. 142, No. 4
glossohyal hypohyal
ceratohyal
hypobranchials
epihyal
interhyal
epibranchials
postteniporal
distal radials
basibranchials ccratobranchials
inlrapharyngeal
Figure 9. Ceratostethus bicornis. Dorsal view of hyoid arch (minus branchiostegal rays) and branchial arches. Up- permost branchial elements of right side removed.
branchials, ccratobranchials, epibranchials, and pharyngobranchials, and a pair of infrapharyngeals. Third epibranchial T- shaped when viewed from side (figured only in dorsal view). Second, third, and fourth pharyngobranchials and anterior half of infrapharyngeals covered with moder- ately sized, irregularly arranged, conical teeth. Infrapharyngeals separate. Gill rak- ers only present on leading margin of first gill arch, 12 or 13 rakers on each side. First epibranchial bearing a single smaller raker, i.e., all rakers except this one are borne on lower half of gill arch. Anteriormost eight gill rakers on lower limb with distal ends directed laterally and posteriormost four or five rakers with distal ends directed increasingly medially, arrangement of these uppermost rakers thus simulating nonnal arrangement of rakers on upper limb of arch found in many fishes.
Pectoral fin (Fig. 10). Posttemporal forked; upper fork tightly bound to epiotic, lower fork short and not reaching base of cranium. Supracleithiiim (as in other atheriniforms ) a small scalelike element completely interposed between posttempo- ral and cleithrum. Clcithiiun slender, onlv
Figure 10. Cera/ostethus bicornis. Lateral view of pectorc girdle.
moderately expanded where it meeft scapular to form scapular foramen, with out an expanded portion extending poste riorly to scapular. In mature males anterio limb of cleithrum elongated anteriorly, t almost double relative length of cleithrur in immatures and females, approximatin; , pulvinular bone of priapium. Cleithrum o one side elongated slightly more than tha of other side and ligamentously attached t puK'inular; at point of attachment sHghtl expanded to form a concave lamella fittin; snugly over a convexity or bump in dorss surface of pulvinular; anterior tip of op posite cleithrum ending in a fine poinl Relative development of anterior prolon gation of cleithra variable. In some sped mens tip of cleithrum unattached t pulvinular falls far short of it, in other spea mens equal in length cleithrum attached t pulvinular. Laterality of cleithrum attachei to pulvinular independent of laterality c priapium itself. Apart from this differenc in development of the cleithrum, the pec toral girdle is similar in males and female; Anterior end of coracoid not prolonged i males. Posterior end of coracoid fused wit ventrally expanded portion of scapular, an^ not associated with radials of pectoral fii Postcleithra absent. Proximal radials two i number, lowermost considerably enlargec their proximal ends fitting snugly into concavity in posterior border of scapula;
Ceratostethus — Osteology and Structural Novelties • Roberts 407
igure 11. Cerofosfefhus fa/corn/s. Axial skeleton.
\ series of about four small distal radials. \n unpaired splint attached to proximal iorsal surface of uppermost pectoral ray. Pectoral rays 10 or 11.
Median fins (Fig. 11). First dorsal fin tvith two very short, slender, refractile ipines; second spine even shorter and slen- derer than first, attached to a pterygium :onsisting of a single element. Second iorsal with five segmented rays. First ray imbranched; last ray divided to its base, interior division branched, posterior divi- sion simple. Second dorsal fin pterygials 3ach consisting of a single element, with slender proximal and expanded distal por- tions. Anal fin with 14 or 15 elements, first element simple, short, and unsegmented, second long and segmented; third through penultimate elements branched, segmented rays; last element divided to base, both divisions simple. Form of anal fin pteryg- ials as in second dorsal fin, except for enlarged anterionnost pterygial, which is prolonged anteriorly in ventral midline for a distance equal to eye diameter in front of anal fin origin. Caudal skeleton (Fig. 11) with two hypurals, both fused to hypural centrum. No epineural or urostyle. Two slender epurals. Penultimate vertebra with well-developed haemal spine but neural spine reduced to expanded basal portion only. Neural and haemal spines of vertebrae preceding penultimate vertebra equally slender and elongate. No accessory haemal spine or separate slender element
interposed between haemal spines of antepenultimate and penultimate vertebrae (such an element present in Dermogenijs, Onjzias, and Xenopoecihis: see Rosen, 1964, fig. 21). Caudal fin with 7 upper and 7 or 8 lower procurrent rays, and 7 upper and 7 lower principal rays, the outermost principal ray in each lobe unbranched. Posterior margins of hypurals straight and vertically aligned.
Vertebral column (Fig. 11). Vertebrae 35 or 36 excluding hy^Dural; precaudal plus caudal either 18 + 17, 17 + 18 or 18 + 18. Distal ends of all except a few of the anteriormost and posteriormost ribs (which are associated with anterionnost pterygial elements of anal fin) meet opposite mem- bers in ventral midline of body. In females, anteriormost ribs are borne on third verte- bra: first pair short and slender, without modified parapophyses; first three or four pairs of ribs gradually increasing in length. In males, first ribs borne on fourth vertebra. These ribs greatly expanded, their distal ends entering priapium just anterior to pulvinulus; parapophyses of fourth vertebra greatly enlarged, projecting anteriorly be- neath \ertebral column to a point below first and second vertebrae (Fig. 11). First three vertebrae without parapophyses. Ribs of fifth vertebra short and slender, their parapophyses unenlarged and posteriorly directed. Ribs of sixth vertebra of full length, reaching ventral midline of abdo- men. Haemal arches not expanded (swim-
408
BuUcfin Museum of Comparative Zoology, Vol. 142, No. 4
Figure 12. Cerafosfefhus bicorn/s. Laferal view of shoulder girdle, urohyal, and priapial skeleton of dextrol male; ctenot tinio abducted.
bladder fitting snugly against posterior wall of abdominal cavity, without posterior pro- jection into haemal arches). Intermuscular bones absent. First four vertebrae with distal ends of neural spines longitudinally expanded. Supraneurals absent.
Priapium (Fig. 12). The priapial skeleton of Ceratosteihus is very similar to that of Solenophalhis (Aurich, 1937) and Neo- stethus. The first and second ctenactinia of Ceratostethtis evidently are homologous, respectively, with the ctenactinium and short, hooklike "Priapklaue" of Soleno- phalhis. The anterior bundle of tliree seg- mented branched rays and posterior bundle of three segmented branched rays also are present. These rays lie in the wall of a sac that becomes filled with spennatophores. Tlie branched ray tips protrude slightly
from this sac, providing its posterior mar gin with a "fringe" or "comb."
As in Solenophalhis, the main element in the axial skeleton of the priapium ar( the two pelvic bones. The proctal pclvi< bone (i.e., the one opposite the ctenactina" is considerably enlarged and acts as th( suspensorimn of the entire priapium. It i: suspended anteriorly by the outer pulvin ular and inner pulvinular ("Pulvinulus trager" of Aurich), which are in tun suspended by the attachment of the pulvin ulus to the cleithrum and perhaps also th{ urohyal. Tlie inner pulvinular is absent ir Giihphalhis and Mirophallus (of. Bailey 1936; Aurich, 1937). Posteriorly it is sus pended from the vertebral column by it: ligamentous attachment to the antcpleura cartilage and anterionnost modified rib;
Ceratostethus — Osteology and Structural Novelties • Roberts 409
(see Aurich, 1937, fig. 3 of SolenophaUus) . rhe proximal ends of rays in posterior ray bundle firmly articulated to side of pos- teriormost part of proctal pelvic bone. \proctal pelvic bone movably articulated to proctal pelvic bone by a series of complicated bones, the morphology of vvhich is difficult to make out in intact Driapia (see also Aurich, 1937). Two :-tenactinia movably articulated to aproctal Delvic bone. Lying between anterior end )f aproctal pelvic bone and accessory xilvinular is a small bone, the pulvinular >sselet, evidently homologous with the pulvinular cartilage in Solenophalhis. At he posterior end of the priapium, in the ixis of the ctenactinia, is a series of ex- pressively modified bones surrounding or partly surrounding the genital pore. These elements represent the "dorsal," "middle," md "ventral penisbones" of Aurich. In 'Jeratostethiis and Neostethus the "ventral penisbone" gives rise to a large number of straight, thin bony projections, constituting 1 sort of flap over the genital pore. Many, f not all, of these projections are doubled back at the tip, and end as a booklet. In Neostethtis the projections, about 80 in lumber, are parallel to one another. In Ceratostethus the projections are much nore numerous, some are considerably en- larged, and at each end of the genital pore 1 large number of the projections are oriented away from the main body of pro- jections. The "ventral penisbone" of Solenp- nhaUus ctenophortis is similarly modified, [t evidently is homologous with the greatly enlarged "ventraler Penisknochen" of Gula- ohalhis (called papillary by Bailey, 1936), vvith the slender "ventraler Penisknochen" >f MirophoUus, and perhaps with the en- larged papillary bone in the base of the oenis in Phenacostethiis. In GulaphaUus, MirophaUus, and Fhenacos-tethus, however, "iie parallel projections and booklets are absent.
Plectrostethus Myers (1935) is said to be closely related to Neostethus, "from which t differs trenchantly in the presence of the
spine-bearing process of the priapium and in the wing-like margin of the cten- actinium." The ctenactinium has a broad membranous margin along the lower side of its proximal half. At the base of the ctenactinium is a flat fleshy process armed on its upper and posterior border with a row of nine or ten short sharp recurved spines and on its anterior border by two longer spines directed forward (Myers, 1935: 5-6). The flat fleshy process is presumably homologous with the "ventral Penisbone" (= papiilaiy bone ?) of Cerato- stethus and Neostethus. Tlie only known specimens of Plectrostethus are 12 type specimens of P. palawanensis Myers (U. S. National Museum nos. 93421-93423).
CHARACTERIZATION OF THE SUPERFAMILY PHALLOSTETHOIDEA
This characterization of Phallostethoidea is based on a survey of the literature and my own observations on Phenacostethus (both species), Ceratostethus hicornis, Neostethus siamensis (probably equals N. lankesteri) and GulaphaUus mirahilis. My observations have been relatively complete only for Ceratostethus. Many, if not most, of the characters cited below can be found scattered in the literature; virtually all of these characters have been verified by me, and erroneous statements in the literature rectified (e.g., concerning the pectoral girdle and first pleural ribs). Tlie char- acters are presented in the following order:
1. general body features, size, habitat;
2. reproduction; 3. sensory organs; 4. squa- mation; 5. fins; 6. skull and visceral arches; 7. dentition; 8. pectoral girdle; 9. vertebral column (including ribs and intramuscular bones ) .
1. Largely translucent, bilaterally com- pressed and moderately elongate, tiny or very small, atherinifonn fishes, adults from 14 {Phenacostethus smithi) to 37 mm (GulaphaUus eximus) in standard length (females usually slightly larger than males). Inhabiting fresh to brackish, usu-
410 Bulletin Museum of Comporativc Zoology, Vol. 142, No. 4
ally turbid, coastal streams of mainland and insular Southeast Asia, usually within range of the tides.
2. Oviparous; gonad single; fertilization internal, males with a clasping and intro- mittant organ, die priapium, derived mainly from the pelvic fins; pelvics evanes- cent, vestigial or absent in females. In males anus opens on side of priapium op- posite genital pore, in females anal and genital openings close together on throat. Eggs demersal, chorionated, with adhesive filament (not verified in Phallostethidae ) ; sperm transferred in large masses held to- gether by a mucoid substance (sper- mozeugma), at least in Neostethidae. Newly hatched young are miniatures of adults, probably immediately capable of active feeding and swimming.
3. Eyes large, laterally directed, free from orbital rim. Nostrils, if present, minute (a single opening which may be a nostril lies on the side of the snout near the upper jaw in Neostethiis but is apparently absent in Ceratostethus; olfactorv lamellae are not evident beneath the skin of tlie snout). A large pore midway between tip of snout and anterior margin of eye in Neostethidae is evidently a pore opening into supra- orbital canal of cephalic laterosensory system. Cephalic laterosensory system well developed, supraorbital, preorbital, post- orbital, preopercular, and mandibular canals with large pores in Neostethidae. In Phallostethidae laterosensory organs on top of head housed in a membranous dome, evidendy without pores (see Roberts, 1971). Lateral line canal absent on body. Otoliths absent (dissolved by fonnahn ?). Contact organs absent.
4. Scales cycloid; body except "neck" completely scaled; scales in lateral series 31-58. Head scaleless or sparsely scaled, scales on head confined to posterior border of skull roof and preopercle.
5. Pectoral fin set high on sides of body, its shape slightly falcate, with ten or eleven rays. First dorsal fin usually present (ab- sent in Mirophalhi.^, SolenophaUus thessa
and perhaps absent in Fhallostethus), con- sisting of one or two tiny, weak (nonerectile ?) spines movably articulated to an elon- gate pterygium. The second spine, when present, usually even thinner and shorter than the first. First dorsal completely separated from second dorsal, but much closer to it than in atherinids, its origin posterior to a vertical through anal fin origin. Second dorsal with one or twc simple, unbranched, segmented rays (nc spines), and three to nine branched rays Origin of second dorsal well behind ana' fin origin, or even posterior to anal fin Anal fin with a short flexible spine, fol- lowed by an unbranched, segmented ray and ten to 28 branched rays. Initial pteryg- iophore of anal fin an enlarged backward.' L-shaped bone (Fig. 11). A translucent median, abdominal, membranous keel in- variably present. Caudal fin forked; in at least some species the upper lobe is slightly longer than the lower lobe and is pointed, while the lower lobe is rounded (Fig. 1: Roberts, 1971, figs. 2 and 3 of Phenaco- stethus ) ; principal caudal rays from 5 + 7 or 6 + 7 to 7 + 8 or 8 + 8.
6. Upper jaw usually highly protractile. Mesethmoid a single, irregularly hexagona lamina; intercalar absent; parietals absent; infraorbital series represented by lacrimal, second infraorbital and dermosphenotic; dermopalatine absent. Lower arm of maxilla separated from mandible by ex- panded lower arm of premaxilla; maxillo- mandibulary ligament round in cross sec- tion where it attaches to dentary, the attachment near symphysis of lower jaws. Hind border of opercle concave; opercle and preopercle without spines or serrations; interoperculum not extending posterior to vertical from preopercle or overlapping subopercle. Distal half of ceratohyal abruptly expanded, a dorsal bridge of bone joining it to epihyal; branchiostegal rays four or five; gill rakers present only on leading edge of first gill arch, other arches rakerless.
7. Teeth strictly conical, confined to
Ceratostethus — Osteology and Structural Novelties • Roberts 411
jremaxillary, dentary, second through ourth pharyngobranchials, and infra- )haryngeals. Teeth on premaxillary and ientary in a single row (Ceratostethus, Seostethus, Phenocostethiis) or in two to 'our rows (Giila phallus). Dentition of ower jaw well developed in Gulaphalhts, )oorly developed (vestigial ?) in Cerato- ;tethus and Neostethus.
8. Supracleithrum a small bone sand- viched between posttemporal and dorsal ip of cleithrum; cleithrum without wing- ike dorsal expansion; scapula and coracoid used together; only two proximal pectoral adials or actinosts, the lowermost en- arged; cleithra extended anteriorly in nales, one or both of them attached to )ulvinular bone of priapium.
9. Vertebrae 34 to 38. In Neostethidae he numbers of precaudal and caudal verte- brae are about equal. In Phallostethidae ?audal vertebrae more numerous than )recaudal: Phenacostethus with 14 to 16 Drecaudal and 19 or 20 caudal, Phallo- itetlius with about 10 precaudal and 28 ?audal vertebrae. First pair of ribs borne 3n third vertebra in females and on fourth /ertebra in males of Ceratostethus, Neo- itetlius, and PlienocostetJius. In GuhiphaUus mirahiUs (MCZ 33904) a 23.1-mm. immature nale and a 22.0-mm specimen that appears :o be an immature female both have the first ribs on the fourth vertebra. First pair 3f ribs in males greatly enlarged and at- tached to axial support of priapium; par- xpophyses of aproctal priapial rib enlarged ind directed anteriorly so that its distal i^nd, to which the rib is attached, lies be- aeath second vertebra. Intramuscular bones ibsent.
Comments. The Phallostethoidea con- stitute a distinctive and well-defined taxon. ■Vlmost all of their characters are in reason- ible agreement with Rosen's concept of the Atheriniformes, with the (probably minor) exception of number of pectoral radials and relationship of the scapula and coracoid. Within the Atheriniformes it appears to be most closely allied to the
atherinoids, and within atherinoids, to the subfamily Taeniomembrasinae. The sug- gestion by Rosen (1964: 242) that phallo- stethoids might be more closely related to cyprinodontoids than to atherinoids was based solely on the erroneous idea that the first pair of pleural ribs in phallostethoids originated on the second vertebra. In phallostethoids, as in many (all ?) ather- inoids, the first pair of ribs is borne on either the third or the fourth vertebra, whereas in cyprinodonts the first pair of ribs is usually (invariably ?) borne on the second vertebra. It is noteworthy that the lowest vertebral count in atherinoids is 31, only three less than the lowest count in phallostethoids, whereas a number of cyprinodontoids have as few as 26. Cyprin- odontoids usually (invariably ?) have intramuscular bones, while at least some (but not all) atherinids agree with phallo- stethoids in lacking them. Various other points in which phallostethoids agree with atherinoids more than with cyjDrinodontoids have been pointed out in my osteological description of Ceratostethus. Among the most important of these are the basically similar morphology of the jaws and at- tachment of the maxillomandibulary liga- ment to the dentary.
While phallostethoids are the only Atherinifonnes k^^o^^^l to me in which the first pair of ribs is ligamentously attached to the pelvic girdle, a ligamentous connec- tion between the pelvic girdle and the distal ends of a pair of ribs evidently is a primitive atherinifonn character. Which pair of ribs is attached depends mainly on the position of the pelvic girdle. Thus the ribs attached to the pelvic girdle are the third pair in Melanotaenia, the fourth in Telmatherina, the fifth in Fhiciphyhx, the sixth in Menidia, etc.
According to Bailey (1936: 464), in GuhphaUus mirahilis the distal ends of the first pair of ribs are "embedded in a heavy triangular mass of fibrocartilage, the ante- pleural cartilage, which is attached by fibers to the posterior crest of the axial
412 Bulletin Museum of Comparative Zoology, Vol. 142, No. 4
bone." In The nacoste thus (ibid.: 472) the tips of the priapial ribs are embedded in the antepleiual cartilage, "which is ossified throughout most of its length and fonns a distinct rod-like antepleural bone." An ossified antepleural is present in Cerato- stethus and Neostethus. In atherinoids and cyprinodontoids a "secondary postclei- thrum" is intimately associated with the distal end of each of the first pair of ribs. There is no trace of this element in female phallostethoids, and none in males either, unless it corresponds with the antepleural. The antepleural cartilage or bone of phallo- stethoids is perhaps neomorphic. The inner puK'inular bone of Ccratostethus and 'Neo- stethus, which apparently gave rise to the phallostethid toxactinium, probably is neo- moq^hic. It apparently has no homologue in atherinoids. For a discussion of the homologies of the priapial elements see Woltereck (1942b: 343-347).
The haemal spines of the anterionnost caudal vertebrae are expanded to perniit extension of the swimbladder posterior to the abdominal cavity in various exocoetoids (only in the family Exocoetidae ?), cyprin- odontoids, and atherinoids. In phallo- stethoids the swimbladder extends to, but not beyond, the posterior wall of the ab- dominal cavity (as in taeniomembrasines ) , and the haemal spines are unexpanded.
The single gonad of Phallostethoidea (observed in Phallostethus and Neostethus by Regan [1916], in Phenacostethus and Guhphallus by TeWinkel [1939]) is prob- ably a primitive character widespread in atheriniforms. The gonads are usually (in- variably ?) single in cyprinodonts, includ- ing Oryzias, Rivulus, Fundiihis, and others.
The teeth of phallostethoids, both on the jaws and in the pharynx, are evidently invariably simple conical teeth, as one would expect in fishes of their tiny size. Hence, they fail to exhibit one of the most persistent trends of the atheriniforms, namely possession of tricuspid teeth.
Wide.s^jiead occurrence of tricuspid teeth in atheriniforms. While most atherini-
forms have conical jaw teeth, bi- and tri- cuspid teeth occur in the jaws of diverst representatives; this is especially tiiie ir cyprinodontoids but is not confined tt them. Miller (1956: 8-9) hypothesized thai tricuspid teeth evolved independently ai least three times in New World cyprin odontids, and suggested that the Old Worlc cyprinodontid genera with tricuspid teetl also originated independently. Tricuspic jaw teeth occur in some genera of Poc ciliidae; the outer row of jaw teeth o Jenynsiidae has tricuspid teeth. Bicuspid jav teeth characterize the atherinid genu: Atherinops, and tricuspid jaw teeth occu in some genera of Exocoetidae anc Hemiramphidae. The significance of thes( facts has not been fully appreciated, be cause the remarkably widespread occur rence of bi- and tricuspid pharyngeal teetl in atheriniforms has passed virtualh' un noticed. Gannan (1895, pis. I-V) showec that practically all cyprinodontoid group: exhibit a considerable variety of tootl forms in the jaws and pharyngeals; in ; number of instances the pharyngeal teetl exhibit considerably more variation in foni than the jaw teeth. Anahleps, for example has conical jaw teeth, but the pharyngea teeth exliibit a \\'ide variety of fonns in eluding tricuspid. Many cyprinodonts wit) conical teeth in the jaws nevertheless ha\'( some pharyngeal teeth of bi- or tricuspic fonn. Pantanodon podo.xys, with no trace of jaw teeth, has tricuspid teeth on tli( pharyngeals, even in specimens only 17 mn in standard length (Whitehead, 1962, fig 8). As one might expect, however, the smallest cyprinodontids and poeciliids usu ally have simple conical teeth in both jaw; and pharynx. This does not seriously dc tract from the generalization that most, i not all, groups of cyprinodontoids have th( tendency to develop bi- and tricuspid teeth and that this is seldom completely sup' pressed. This generalization can be ex tended to include most groups of atherini forms. Thus tricuspid teeth occur on th( pharyngeals in exocoetids and hemiram
Ceratostethus — Osteology and Structural Novelties • Roheiis 413
)hicls (some of which also have tricuspid aw teeth) and in scomberesocoids (jaw eeth usually or invariably conical; verified n Scomberesox saiirus). Belonidae ex- imined by me have conical pharyngeal eeth; the conical pharyngeal teeth of everal belonids are figured by Collette 1966, figs. 2 and 3). The pharyngeal teeth )f adrianichthyoids are conical (verified in (.enopoecihis poptae and X. samsinoriim). rhroughout the atheriniforms, the smallest onns are likely to have simple conical eeth (e.g., phallostethoids, On/zias); again, his does not seriously detract from the ;eneralization that atheriniforms have a emarkable predisposition to develop bi- nd tricuspid teeth.
)ISCUSSION
TJw protractile jaws of phallostethoids. "he diversity of jaw mechanisms in teleosts ^ becoming ever more apparent. Probably nore neomorphic joints and bones have irisen in the jaws than in any other of the najor teleostean functional anatomical mits (as defined by Liem, 1967). A great nany of these structural innovations are ssociated with mobility of the upper jaw. t has been estimated that the jaws are )rotractile in about half of the living eleosts. While the majority of fonns with )rotractile upper jaws are acanthopteryg- ans, paracanthopterygians, or cyprinoids, )rotractile upper jaws also occur in Vtherinifonnes, in the gonorhynchoid Vnactolaemus, and the characoid Bivi- ^ranchia. Tlie functional mechanism of irotractility is completely different in .canthopterygians, cyprinoids, Phracto- aemus and Bivihranchia (for acanthopte- ygians, see Alexander, 1967a; Liem, 1970; )sse, 1969; Schaeffer and Rosen, 1961; for yprinoids, Matthes, 1963; Alexander, 1966; or Phractohiemtts and Bivihranchia, Gery, 963). The functional mechanisms in itheriniforms are relatively similar to those n acanthopterygians. Although both per- nit protraction of the upper jaws, the
attachments and movements of the acanthopterygian rostral and cyprinoid rostral or kmethmoid are radicallv dif- ferent. As fishes with protractile jaws are studied more closely, additional evo- lutionary novelties are bound to be dis- covered. For example, in Epihidus insidiator (family Labridae) the lower jaw is just as protractile as the upper jaw. As might be expected, this extraordinary adaptation involved major changes in the suspensorium. The outstanding jaw mech- anisms of many teleosts remain equally uninvestigated. Some of the most peculiar jaws occur in tropical inshore fishes for which no observations on feeding have been recorded, and suggestions as to how the jaws are used would necessarily be fairly speculative.
To my knowledge, no other fishes have bony stiTictures comparable to the paraden- taries and submaxillaries of Neostethidae. The rostral bone of neostethids is probably homologous to the rostral cartilage of acanthopterygians, and cartilaginous ele- ments similar to or homologous with the parardstrals occur in atherinoids ( accessory rostral cartilages). Analogous rostral car- tilages or bones and accessory rostral cartilages occur in cyprinoids. Bones anal- ogous to the submaxillaries occur in Cobitidae. The maxillomandibulary bones are small bones, and similar elements in other fishes, if they occur, are likely to have been overlooked. A small bone similar to the maxillomandibulary bone, but lying dorsal to the lower limb of the maxillary instead of ventral to it, occurs in the characoid Chilodus. Such "supramaxil- lary" bones probably have evolved inde- pendently in numerous lines. The presence of such an element in CdnJodiis cannot be construed as a primitive character.
The paradentaries and submaxillaries are large elements, obviously of considerable functional significance. They are evidently involved in increasing the extent to which the ja\\'s can be protracted. The paraden- taries probably are also important in keep-
414 Bulletin Museum of Comparative Zoology, Vol. 142, No. 4
ing the mouth wide open when it is fully the submaxillary bone and maxillary in neo
protracted and perhaps in pennitting the stethids is similar but of greater magnitude
upper jaw to close dowTJ over the lower Tlie condition of the prevomer should b(
jaw while the lower jaw remains fully noted. In almost all fishes this dermal bon(
depressed. The paradentaries are joined by is a single, median element. In Cerate
a ball and socket joint to the dentaries stethus, however, it sometimes ossifies ii
near the symphysis of the lower jaw, and two pieces that remain separate, each witl
are capable of rotating outward 90 degrees its anterior end greatly expanded where i
and upward 45 degrees from their resting is joined by a meniscus to the submaxillar
position in a trough on the side of the bone. While dermal bones usually maintaii
dentary. The submaxillaries are tubular their integrity, it is not impossible for then
bones with a meniscus at either end : the to give rise to separately ossifying elements
anterior meniscus with the maxillary bone, Frazzetta (1970) described and discusse(
the posterior meniscus with the prevomer. the subdivision of the maxillary bone b
Although novel as bony elements, the form two maxillary bones united by a:
paradentaries and submaxillaries arise in intramaxillary joint in bolyerine snakes,
structures that are characteristic of atherini- Neotenic characters of plmllostetJwich
forms, and particularly of atherinoids, the In teleosts it is extremely unusual for th
group to which they seem most closely anus and urogenital openings to be placci
related. Furthermore, the paradentary anterior to the pelvic fins. In most teleost
bones and submaxillary bones can be seen in which these openings lie at the throii
as the ultimate consequences of clearly the pelvics are either absent or are eve
defined evolutionary trends exhibited in farther forward than the openings. In thi
atherinoids. Thus the anterionnost part of respect phallostethoids are exceptional. I
the maxillomandibulary ligament in some males and in females in which the pelvic
atherinoids (notably in the Taenio- are vestigial or evanescent, anal am
membrasinae) has the peculiarities of the urogenital openings are morphologicall
anterior part of the maxillomandibulary anterior to pelvic fins. In adult atheriij
ligament in phallostethoids except that no oids the position of the openings i
paradentary bone develops in it. That is, variable, but they always lie posterior t
in these atherinids the attachment of the the pelvic origin. It may be immediatel
ligament to the dentary is extremely far in front of the anal fin, distinctly in front c
forward, near the symphysis; the body of it, or even between the pelvic fins. Woltt
this portion of the ligament is tough and reck (1942a: 256) made the importan
round in cross section, and when the mouth discovery that the openings lie anterior t
is closed it lies in a trough on the side of the pelvic fins in larval Atherina. This sm
the dentary identical with the trough in gests that the condition in phallostethoid
which the paradentary bone lies in neo- may be neotenic.
stethids. Concerning the submaxillary bone. It may well be that the anterior positio
it evidently serves much the same function of the pelvic fins, and perhaps even thei
as the submaxillary meniscus in atherinids. attachment to the shoulder girdle, ar
Alexander (1967b: 234) found the sub- neotenic, at least in part. In atherinoid
maxillary meniscus much thicker in thought to be primitive, Rheocles ani
Atherina than in Melanotaenia; in Atherina Bedotia, the pelvics are relatively far foi
"the submaxillary meniscus can swing an- ward. The ancestors of atherinoids ma
teriorly and somewhat medially about its have had the acanthopterygian charactei
articulation with the cranium, carrying the istic of thoracic pelvic fins connected wit
cranial condyle of the maxilla anteriorly the shoulder girdle,
and ventrally" (ibid.: 236). Movement of The subcephalic position of the priapiur
Ceratostethus — Osteology and Structural Novelties • Roberts 415
cannot, however, be ascribed entirely to leoteny. No atherinoids or atheriniforms >f any kind have the pelvics anywhere near IS far fonvard as they are in phallo- itethoids. The anteriad prolongation of the ?leithrum and of the parapophyses of the ourth vertebra in male phallostethoids ndicates that the priapium has "migrated or\\'ard ' ( in an evolutionary sense ) . The elective advantage of an anteriorly located, .e., subcephalic, priapium may lie in greater visual coordination during mating.
The phallostethoid "neck" may also be onsidcrcd a neotenic character, for it ap- )ears to result in large part from absorj^tion if the yolk. In larval phallostethoids the ompact yolk sac lies immediately posterior 3 the head; it occupies considerable space /here the neck eventually forms.
Almost certainly neotenic is the phallo- tethoid abdominal keel, derived from the ledian embryonic fin fold. In larval phallo- tethoids the fin fold extends continuously rom its origin at a point on the dorsum bout midway between snout and caudal lase (where the first dorsal fin arises), round the caudal lobe and then on the entrum uninterruptedly until it ends at he posterior margin of the yolk sac (see ^illadolid and Manacop, 1934, pi. 2, fig. -2, pi. 3, figs. 1 and 2 of larval GuhphaUus vmhilis). In adult phallostethoids the bdominal keel consists of actinotrichia the fibrous rays characteristic of embry- nic fins ) .
The abdominal keel presumably plays le most important of the hydrodynamic oles formerly played by the pelvic fins. If, Is may have been, acquisition of the ab- ominal keel permitted phallostethoids to tilize the pelvic fins for an entirely new jnction, it was a key step in their evo- ition.
Evolution of complicated external ^cni- ilia. Internal fertilization is a regular 'ature in the reproductive biology of lany groups of atheriniforms (excepting hallostethoidea, these groups belong to le Cyprinodontoidea). In three groups —
Poeciliidae, Goodeidae, and Jenynsiidae- Anablepidae — it has led to viviparity. Gonopodial morphology in these viviparous fonns ranges from relatively simple in Goodeidae and in the poeciliid tribe Poe- ciliini to remarkably complex in the poeciliid tribe Cnesterodontini (see Rosen and Bailey, 1963, for figures of the poe- ciliid gonopodia). It is a striking fact that in the three groups that evolved internal fertilization and are oviparous — Tomeur- idae, Horaichthyidae, and Phallostethoidea — the primary copulatory organ is far more complex than in any of the \'iviparous forms. ^ The gonopodium of Tomeunis can be compared directly \^'ith that in Poe- ciliidae, since Tomeunis is evidentlv related to the poeciliids and its gonopodium is derived mainly from the same rays in the anal fin as the poeciliid gonopodium. Per- haps the most complicated gonopodium in any poeciliid is that of the cnesterodontin Pliallotoiynus, which bears an enlarged scooplike element at its tip ( ibid., fig. 31a, b). The gonopodium of Tomeunis is much more complex: its tip bears a pair of greatly enlarged antlerlike processes, a far more complicated scooplike arrangement, and the distal end of the fifth anal ray, which is modified in a manner recalling the even more greatly modified papillary bone of neostethids {ibid., figs. 7, 13-16). The gonopodimn of Horoichthijs (Kulkarni, 1940), which evolved independently of the gonopodium of Poeciliidae and Tomeur- idae, is as complicated as that of Tomeunis. The neostethid priapium is moq:)hologically as complex as or even more complex than the gonopodia of Tomeunis or Horaichthys, and far more complicated than the gono- podium of any of the viviparous forms. The question may now be raised, \vhy are these structures in oviparous fonns more com- plex? Conversely, why have viviparous
^ The situation is admittedly complicated liy the development in \ iviparous forms of modified pelvic or pectoral fin structures that fimction as acces- sory copulatory structures, but this does not seriouslv alter the main theme under consideration.
416 Bulletin Museum of Comparative Zoologij, Vol. 142, No. 4
forms failed to evolve gonopodia as complicated as those in oviparous forms? The puzzle increases \\'hen we recall that the gonopodium of Hoiaichthys is not an intromittant organ but serves only to con- vey spennatophores with dartlike attach- ments that are implanted externally near the female's vent.
It would seem that perfection of the organs of fertilization would be just as im- portant to viviparous forms as to oviparous ones; and yet nature has favored the more complicated structures in oviparous forms, and selected against their development in the viviparous forms.
The explanation of this seemingly con- tradictory state of affairs perhaps lies in two considerations: first, effects on mating behavior of differences in the habitats occupied by oviparous and viviparous atheriniforms, and second, deleterious ef- fects of excessively complicated external genitalia. Evidence concerning these fac- tors was gathered during my fieldwork on phallostethoids in Thailand.
In Tliailand neostethids and phallo- stethids occur only a short distance inland, in waters strongly influenced by tides. Usually the current is strong and the water muddy. They are swift swimmers, generally stay in midwater and can maintain them- selves in the current. During flood waters they are likely to be widely dispersed. Thus several hours of collecting at Khlong Langu in Satul Province during high tides com- bined with heavy rain yielded only a half- dozen immature Neostethus, all taken singly. Whenever either Ceratostethus or Neostethus were found during low water, they were usually aggregated and some- times exceedingly numerous (as in the estuary of Chantaburi River, where Neo- stethus and Ceratostethus were found to- gether in large numbers ) . When the tide is out, Ceratostethus are likely to be confined to well-defined creeks and channels; when the tide is in, they are probably dispersed over a wide area of partially submerged mangroves and nipa palm. Horaichthys and
Tomeurus apparently occur in comparable habitats, Horaiclithys along the Bombay- Kerala coast of India, Tomeurus along the northeastern coast of South America, in- cluding the mouth of the Amazon. The selective advantage of internal fertilization in these fishes may lie in pennitting tem- poral separation of mating and egg-laying corresponding, respectively, with periodj of low water and high water ( Roberts 1971 ) . It seems likely that mating in sue! habitats is difficult, because of current anc low visibility. The waters in which these fishes live are probably especially turbic during the rainy season, when mating prob ably reaches its peak. In this connection it is noteworthy that copulation in Gula phallus mirahilis is prolonged, the mak clasping the female for one to two minute (Villadolid and Manacop, 1934: 200).
In contrast to phallostethoids, poeciliid tend to occiu- in habitats in which mating can occur much of the time and under rel atively favorable conditions. Poeciliid gono podial thrusts are relatively short in dur ation, some almost instantaneous; they ma; be repeated several times without spem transfer and often strike wide of the mark perhaps even more in some forms wit] less specialized gonopodia. Yet such male are undoubtedly quite efficient at fertiliz ing females; the number of females ii populations of poeciliids is general! greatly in excess of that of males.
In all my collections of neostethids, male outnumbered females. In places where the; were particularly abundant, it was some times possible to catch 25 or 50 at once, anc in these instances the ratio of males to fe males was sometimes about equal. Fron some places where individuals were reli atively scarce or collecting was particular! difficult, collections of half-a-dozen or dozen specimens of Ceratostethus are en tirely males. During fieldwork it appeared that 1 ) males of neostethids are at least a abundant, if not more abundant, tha females, and 2) males are more readil caught because the priapium prevents theri
Ceratostethus — Osteology and Structural Novelties • Roberts 417
from swimming away as quickly as females. This suggests that males are more suscep- tible to predation than females. It seems probable that the exceptionally large gonopodia of Horaichthys and Tomeurus would encumber their swimming.
Another disadvantage of highly compli- cated genitalia may lie in an increased proportion of developmental anomalies. Developmental anomalies in the priapium of phallostethoids are probably relatively frequent. Woltereck (1942a) mentioned, but did not describe, developmental anom- alies in the priapia of neostethids from the Philippines. I find a number of anomalies in my material, including a Neostethiis with a peculiarly bowed ctenactinium (which could not be properly swung into resting position beneath the head) and additional externalized elements or "super- numerary ctenactinia" and a Ceratostethus in which the priapium is equally developed (but not fully foiTned) on both sides.
Finally, the development of a large ex- ternal genitalium reqviires a considerable diversion of energy and materials, and such "expenditures" will always be selected against unless they confer a selective ad- vantage. The highly complicated copula- tory organs presumably ensure successful sperm transfer with each copulatory epi- sode. In fishes in which copulatory episodes are rare this will have great selective ad- vantage.
Provided environmental conditions are such that internal fertilization has a selec- tive advantage, there is little to prevent its development in fishes. There are numerous records of internal fertilization occurring in fishes, such as trout, which normally have external fertilization; sperm simply enter the oviduct and fertilize eggs within the female. Such occurrences do not lead to the evolution of internal fertilization, but are selected against, when the life history strategy of the fishes involved overwhelm- ingly favors external fertilization. Ovi- parous atheriniform fishes, on the other hand, which tend to have relatively small
numbers of eggs that are expelled in small batches or even singly, and in which em- bryonic development or hatching can be deferred, may be preadapted to the evo- lution of internal fertilization (Roberts, 1971). If internal fertilization is favored, then every modification of pelvic or anal fins (no matter how slight) that increases the probability of sperm entering the ovi- duct will confer a selective advantage. Thus the stage is set for intensification of func- tion (Mayr, 1960), which eventually can lead to structures as complicated as the priapium and papillary bone of Cerato- stethus bicornis.
LITERATURE CITED
Alexander, R. McN. 1966. The functions and mechanisms of the protrusible upper jaws of two species of cyprinid fish. J. Zool., London, 149: 288-296.
. 1967a. The functions and mechanisms
of tlie protrusible upper jaws of some acanthopterygian fish. J. Zool., London, 151: 13-64, 6 pis.
1967b. Mechanisms of the jaws of some
atheriniform fish. ]. Zool., London, 151;
233-255, 1 pi. AuRiCH, H. 1937. Die Phallostethiden ( Unterord-
nung Phallostethoidea Myers). Intnatl. Rev.
Ges. Hydrobiol. Hydrogr., 34: 263-286. Bailey, R. J. 1936. The osteology and relation- ships of the phallostethoid fishes. J. Morph.,
59(3): 453-483, 4 pis. Banjagan, M. C. 1966. [Untitled letter]. Dopeia
(Miami), June, 1966: 46. CoLLETTE, B. B. 1966. Bclonion, a new genus of
fresh-water needlefishes from South America.
American Mus. Novitates, No. 2274: 1-22. Frazzetta, T. H. 1970. From hopeful monsters
to Bolyerine snakes? Amer. Nat., 104(935):
5.5-72. Garman, S. 1895. The cyprinodonts. Mem. Mus.
Comp. Zool., 19(1): 1-179, 12 pis. Gery, J. 1963. L'appareil protracteur buccal de
Bivibmnchia (Characoidei) (1) avec une
note SUV Pliractolacmiis (Chanoidei) (Pisces).
Vie et Milieu, 13(4): 729-740. Herre, a. W. 1925. Two strange new fishes from
Luzon. Philippine J. Sci., 27(4): 507-514,
2 pis. . 1942. New and little known phallo-
stethids, with keys to the genera and Philip- pine .species. Stanford Ichth. Bull., 2(5):
137-156.
418 Bulletin Museum of Comparative Zoology, Vol. 142, No. 4
HuBBS, C. L. 1944. Fin structure and relation- ships of phallostethid fishes. Copeia, 1944 (2): 69-79.
KuLKARNi, C. V. 1940. On the systematic position, structural modifications, bionomics and de- \elopnient of a remarkable new family of cyprinodont fishes from the province of Bom- bay. Rec. Ind. Mus. Calcutta, 42: 379-423.
LiEM, K. F. 1967. Functional moiphology of the head of the anabantoid teleost fish, Helostoma tcmmincki. J- Morph., 121: 135-158.
. 1970. Comparative fimctional anatomy of
the Nandidae (Pisces: Teleostei). Fieldiana (Zool.) 56: iii-f 156.
Matthes, H. 1963. A comparative study of tlie feeding mechanisms of some African Cyprin- idae. Bijdr. Dierk., 33 : 1-36.
Mayr, E. 1960. The emergence of evolutionary novelties. In S. Tax (ed. ), The Evolution of Life. Chicago, University of Chicago Press, pp. 349-380.
Miller, R. R. 1956. A new genus and species of cyprinodontid fish from San Luis Potosi, Mexico, with remarks on the subfamily Cyprinodontinae. Occ. Papers Mus. Zool. Univ. Michigan, No. 581, 17 pp., 2 pis.
Myers, G. S. 1928. The systematic position of the phallostethid fishes, witli diagnosis of a new genus from Siam. American Mus. Novi- tates, No. 295, 12 pp.
. 1935. A new phallostethid fish from
Palawan. Proc. Biol. Soc. Washington, 48: 5-6.
OssE, J. W. M. 1969. Functional moiphology of the head of tlie perch ( Perca fluviatilis L. ) : an electromyographic study. Neth. J. Zoology, 19: 289-392.
Regan, C. T. 1913. Phallostethus dunckeri, a remarkable new cyprinodont fish from Johore. Ann. Mag. Nat. Hist., 12: 548-555.
. 1916. The morphology of the cyprinodont
fishes of the subfamily Phallostethinae, with descriptions of a new genus and two new species. Proc. London Zool. Soc, 1916: 1-26, pis. 1-4.
Roberts, T. R. 1971. The fishes of the Malaysian family Phallostethidae ( Atheriniformes ) . Bre- viora. No. 374: 1-27.
RosEX, D. E. 1964. The relationships and taxo- nomic position of the halflseaks, killifishes, silversides and their relatives. Bull. American Mus. Nat. Hist, 127(art. 5): 217-267, pls.j 14 and 15.
Rosen, D. E., and R. M. Bailey. 1963. The poe- ciliid fishes (Cyprinodontiformes), their struc- ture, zoogeography, and systematics. Bull. American Mus. Nat. Hist., 126(art. 1): 1-176.
Schaeffer, B., and D. E. Rosen. 1961. Major adaptive levels in the evolution of the actinopterygian feeding mechanism. Amer- ican Zool., 1(2): 187-204.
ScHULTZ, L. P. 1948. A revision of six subfamilies of atherine fishes, with descriptions of new genera and species. Proc. U. S. Nat. Mus., 98(3220): 1-48, 2 pis.
Smith, H. M. 1927. The fish Neostethiis in Siam. Science (n. s.), 6S: 353-355.
TeWixkle, L. E. 1939. The internal anatomy of two phallostethid fishes. Biol. Bull. Woods Hole, 76(1): 59-69.
ViLLADOLiD, D. v., AND P. R. Manacop. 1934. The Philippine Phallostethidae, a description of a new species, and a report on the biology of GithiphaUus mimhilis Herre. Philippine J. Sci., 55(3): 193-220, 5 pis.
Whitehead, P. J. P. 1962. The Pantanodontinae, edentulous toothcarps from East Africa. Bull. British Mus. (Nat. Hist.), Zool., 9(3): 103-137.
WoLTERECK, R. 1942a. Stufen der Ontogenese und der E\olution von Kopulationsorganen bei Neostethiden (Percesoces, Teleostei). Internatl. Rev. Ges. Hydrobiol. Hvdrogr., 42: 253-268.
: 19421). Neue Organe, durch postembryo-
nale Umkonstruction aus Fischflossen entste- hend. Internatl. Rev. Ges. Hydrobiol. Hydrogr., 42: 317-355.
^^^^WpWli^'l^^^:^^
SuLLetln OF THE
Museum of
Comparative
Zoology
Revision of the North American Ciidae
(Coleoptera)
JOHN F. LAWRENCE
HARVARD UNIVERSITY VOLUME 142, NUMBER 5
CAMBRIDGE, MASSACHUSETTS, U.S.A. 29 DECEMBER 1971
PUBLICATIONS ISSUED
OR DISTRIBUTED BY THE
MUSEUM OF COMPARATIVE ZOOLOGY
HARVARD UNIVERSITY
Bulletin 1863-
Breviora 1952-
Memoirs 1864-1938
JoHNsoNiA, Department of Mollusks, 1941-
OccAsioNAL Papers on Mollusks, 1945-
Other Publications.
Bigelow, H. B., and W. C. Schroeder, 1953. Fishes of the Gulf of Maine. Reprint, $6.50 cloth.
Brues, C. T., A. L. Melander, and F. M. Carpenter, 1954. Classification of Insects. $9.00 cloth.
Creighton, W. S., 1950. The Ants of North America. Reprint, $10.00 cloth.
Lyman, C. P., and A. R. Dawe (eds.), 1960. Symposium on Natural Mam- malian Hibernation. $3.00 paper, $4.50 cloth.
Peters' Check-list of Birds of the World, vols. 2-7, 9, 10, 12-15. (Price list on request. )
Tm-ner, R. D., 1966. A Survey and Illustrated Catalogue of the Teredinidae (Mollusca: Bivalvia). $8.00 cloth.
Whittington, H. B., and W. D. I. Rolfe (eds.), 1963. Phylogeny and Evolution of Crustacea. $6.75 cloth.
Proceedings of the New England Zoological Club 1899-1948. ( Complete sets only. )
Publications of the Boston Society of Natural History.
Authors preparing manuscripts for the Bulletin of the Museum of Comparative Zoology or Breviora should send for the current Information and Instruction Sheet, available from Mrs. Penelope Lasnik, Editor, Publications Office, Museum of Comparative Zoology, Harvard University, Cambridge, Massachusetts 02138, U.S.A.
© The President and Fellows of Harvard College 1971.
REVISION OF THE NORTH AMERICAN CIIDAE (COLEOPTERA)
JOHN F. LAWRENCE
CONTENTS
Introduction 419
Acknowledgments 420
Methods and Tenninology 421
Taxonomic Characters 421
Secondary Sexual Characters 423
Measurements and Ratios 424
Observations and Drawings 424
Synonymies and Descriptions 424
Variation 425
Label Data and Collections 425
Systematics of the Family Ciidae 425
Historical Review 425
Family Limits 426
Characterization of the Family Ciidae 426
Phylogenetic Relationships 428
Fimiily Name 432
The Nortli American Ciidae 432
Origin and Distribution 432
Host Preference 435
Key to the Subfamilies and Genera of
North American Ciidae 436
Subfamily Ciinae 437
Genus Cis Latreille 437
Key to the North American Species of Cis .. 439
Genus Enncarthron Melhe 480
Genus Dolichocis Dury — 482
Genus Orthocis Casey 484
Genus Strigocis Dury 488
Genus Hadmule Thomson 491
Genus Plesiocis Casey 493
Genus Ceracis Melhe — - 494
Key to the Nortli American Species of
Ceracis 494
Crenus Sulcacis Dury 502
Genus Malacocis Gorham 504
Subfamily Orophiinae 506
Genus RJioi)(iIodontus Mellie 506
Genus Octotenimis Mellie 508
Literature Cited 509
Figures — 516
Bull. Mus. Comp.
Tables in Text
Table 1. Differences between Sphindocis
and the Ciidae 427
Table 2. Distributional Patterns and Faunal
Affinities of Nortli American Ciidae 430
Table 3. Comparison of Faunal Sectors — 432 Table 4. Differences between Strigocis and S ulcacis 489
ABSTRACT
A taxonomic revision is made of the 12 genera and 84 species of Ciidae known to occur or likely to occur within the confines of tlie continental United States and Canada. Discussions of tlie family limits, phylogenetic relationships, zoogeo- graphic origins, and host relationships are also included. Keys are given to subfamihes, genera, and species. For each North American species, an informal diagnosis, distributional summary, and summary of host records are presented. The fol- lowing generic synonymies are proposed: Cis Mellie ( = Eridaulus Thomson syn., = Macrocis Reitter syn.); Hadraule Thomson (= Maphoca Casey syn.); Malacocis Gorham {= Brachtjcis Casey syn.). Type species are designated for the following genera: Eridaulus Thomson (Anobium nitidum Fabricius); Macrocis Reitter (M. taurus Reitter). The following new species are de- scribed: Cis acritus, Cis cayensis, Cis cornelli, Cis crinitus, Cis discolor, Cis festivulus, Cis megastic- tus, Cis niedhauki, Cis robiniophilus, Cis rotun- dulus, Cis stercophilus, Enncarthron aurisqua- niosuni, Hadraule cxplanata, Ceracis magister, Ceracis pecki, and Rhopalodontus americanus. In addition, 31 new species synonymies are proposed and 12 new combinations are made.
INTRODUCTION
The Ciidae is a relatively small, cosmo- politan family of small to minute beetles,
Zool. 142(5): 419-522, December, 1971 419
420 Bulletin Museum of Comparative Zoology, Vol 142, No. 5
which feed in both lanal and adult stages on the myceha and fruiting bodies of wood-rotting fungi, such as Polyporaceae. At present, the family includes about 40 genera and 550 species, but the generic concepts badly need revision and at least 400 more species remain to be described. The following paper consists of a short, general section on the taxonomy of the group and a more detailed account of the 12 genera and 84 species known to occur or likely to occur within the confines of the continental United States and Canada. Throughout the text, this area will be re- ferred to as North America, even though Mexico has been excluded. The keys and discussions presented should permit the identification of the vast majority of ciid species encountered north of the Mexican border, but there are probably a number of undescribed forms yet to be found in southern Florida, along the Gulf Coast of Texas, and in the mountains of the South- west.
Although the scope of this work is limited geographically, the concepts pre- sented are derived from an examination of large numbers of specimens from various parts of the world. The generic treatment is relatively conservative, and with the exception of a few obvious synonymies {Cis-Macrocis, Hadraulc-Maplwca, Mala- cocis-Brachycis), no drastic changes have been made in the classification currently used in the United States. The recognition of Casey's Orthocis and Plesiocis and Dury's Dolichocis and Strigocis, however, is contrary to the concepts of European workers. In Lohse's work (1967), Orthocis and Strigocis would be included in Cis and Siilcacis, respectively, while species of Dolichocis and Plesiocis would be placed in Ennearthron.
Brief sections are included on the geo- graphic distribution and host preferences of North American Ciidae, but these topics (particularly the latter) will be covered in more detail in a future publication (Law- rence, in preparation).
ACKNOWLEDGMENTS
I am indebted to the following in- dividuals and institutions for making their collections available to me: Academy of Natural Sciences, Philadelphia, Pa. (H. J. Grant); American Museum of Natural His- tory, New York, N. Y. (J. Rozen, P. Vaurie, L. Hemian); Bemice P. Bishop Museum, Honolulu, Hawaii (J. L. Gressitt, G. A. Samuelson); Brigham Young University, Provo, Utah (V. M. Tanner, S. L. Wood); British Museum (Natural History), Lon- don, England (J. Balfour-Browne, R. Pope,
C. von Hayek); Bureau of Entomology, California State Department of Agricul- ture, Sacramento, Calif. (M. Wasbauer); Burke Memorial Museum, University of Washington, Seattle, Wash. (M. H. Hatch); California Academy of Sciences, San Fran- cisco, Calif. (E. S. Ross, H. B. Leech); Cahfornia Insect Survey, University of Cahfomia, Berkeley, Calif. (J. Powell, P.
D. Hurd); Carnegie Museum, Pittsburgh, Pa. (E. Wallace); Cincinnati Museum of Natural History, Cincinnati, Ohio (C. Oehler); J. F. Cornell, Boone, N. C; Cor- nell University, Ithaca, N. Y. ( W. L. Brown, H. Dietrich); Entomology Research Insti- tute, Canada Department of Agriculture, Ottawa, Canada (W. J. Brown, E. C. Becker); Field Museum of Natural History, Chicago, 111. (R. L. Wenzel, H. S. Dybas); Hope Department of Entomology, Oxford' University, Oxford, England ( E. Taylor, M. Graham); H. F. Howden, Ottawa, Canada; Illinois Natural History Survey, Urbana, 111. (H. H. Ross, L. K. Gloyd); Institute Royale des Sciences de Belgique, Brussels, Belgium (R. Damoiseau); V. M. Kirk,: Brookings, S. D.; Museu de Zoologia da Universidade de Sao Paulo, Sao Paulo, Brazil (H. Reichardt, U. Martins); Mu- seum d'Histoire Naturelle, Geneva, Swit-| zerland (C. Besuchet); Museum National d'Histoire Naturelle, Paris, France (A. M.j Villiers); Museum of Comparative Zool-f ogy, Cambridge, Mass. (P. J. Darlington, Jr.); Museum Zoologicum Universitatis, Helsinki, Finland (M. Meinander); Natur-
North American Ciidae • Lawrence 421
historische Museums Wien, Vienna, Austria (F. Janczik); G. H. Nelson, Kansas City, Mo.; Oregon State University, Corvallis, Ore. (J. D. Lattin); F. Plaumann, Nova Teutonia, Santa Catarina, Brazil; Purdue University, Lafayette, Ind. (R. H. Arnett, L. Chandler); San Diego Natural History Museum, San Diego, Calif. (C. F. Harbi- son); J. Schuh, Klamath Falls, Ore.; Snow Entomological Museum, University of Kan- sas, Lawrence, Kans. (G. W. Byers); W. R. Suter, Kenosha, Wis.; United States National Museum, Smithsonian Institution, Washington, D. C. (J. M. Kingsolver, P. J. Spangler, T. J. Spilman, O. L. Cartwright, D. M. Anderson); Universitets Zoologiske Museum, Copenhagen, Denmark (S. G. Larsson); University of Alberta, Edmon- ton, Alberta (G. E. Ball, D. Whitehead); University of Arizona, Tucson, Ariz. ( F. G. Werner); University of British Columbia, Vancouver, B. C. (G. G. E. Scudder); Uni- versity of California, Davis, Calif. (A. T. McClay, R. O. Schuster); University of Wisconsin, Madison, Wise. (R. D. Shene- felt, J. K. Ackerman); Utah State Univer- sity, Logan, Utah (G. Knowlton).
Thanks are extended to the following specialists for their continued cooperation in the identification of fungus specimens and for advice pertaining to the system- atics of the Basidiomycetes: L. Bonar, O. Fidalgo, R. L. Gilbertson, P. Lentz, J. Lowe, O. K. Miller, M. Nobles, R. Singer, J. A. Stevenson, L Taveres.
Tlianks are also due to the following In- dividuals who have made a special effort to collect Ciidae and their host fungi: R. Andrews, J. T. Doyen, R. C. Graves, D. H. Janzen, A. Laska, H. B. Leech, E. Lind- quist, M. Lundgren, C. W. and L. O'Brien, C. Parsons, S. B. and J. Peck, J. Powell, J. R. Powers, A. G. Raske, W. Shear.
I am grateful to R. A. Crowson, H. V. Daly, H. S. Dybas, R. C. Graves, H. B. Leech, E. G. Linsley, G. Lohse, C. D. MacNeill, E. Mayr, J.' Powell, O. W. Rich- ards, R. F. Smith, K. Southern (Paviour- Smith), R. L. Usinger, and P. Wygodzinsky
for their help and advice during portions of this study, and to P. J. Darlington, Jr., for the critical reading of this manuscript.
I am indebted to R. Archbold, M. Cazier, and the American Museum of Natural His- tory for allowing me to use the facilities of the Archbold Biological Station, Lake Placid, Florida, and the Southwest Re- search Station, Portal, Arizona, and to C. R. Benjamin and the U. S. Department of Agriculture for pennitting me to examine the National Fungus Collections at Belts- ville, Maryland. Thanks are also due to E. O. Wilson and Mr. and Mrs. R. Niedhauk for their hospitality during my stay in the Florida Keys.
I also wish to thank J. Campbell, M. Catron, K. Davis, S. Duncan, K. Heise, S. Loth, W. Peairs, and K. Pearson for techni- cal assistance. I am especially gratefully to Mrs. Penelope Lasnik for the difficult and often thankless job of seeing this manu- script through to publication.
Part of this research was conducted with the aid of a National Science Foundation Grant (GB-4743), National Science Foun- dation Summer Fellowships, a Woodro\\' Wilson Fellowship, and grants-in-aid from the University of California Committee on Research and the Society of the Sigma Xi.
METHODS AND TERMINOLOGY
Taxonomic Characters
Color is of limited diagnostic value in the family, except when large samples are available for comparison, and it has been used to distinguish only a few forms with obvious elytral markings. When color is given in descriptions, simple English words are used and Latin terms are avoided. Vestiture has been used extensively in this work, especially at the species level, since it usually is subject to little geographic variation (except in Cis creherrimus, p. 452). The vestiture is said to be single if it consists of one type of element (bristles, hairs) and dual if there are two distinct t\'pes ( bristles and hairs, erect and inclined
422 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
bristles, Figs. 29 and 30). Tlie individual elements varv from short, fine hairs or short, stout scales to long, recurved hairs or long, erect bristles. In certain genera {Ceracis, Octotemnus) and species groups (Cis tricornis group, Cis pacificus group) the vestiture remains fairly constant, whereas in others (Cis nitidus group, Cis taiinis group) it is subject to considerable variation, even among closely related spe- cies. The lengths of individual elements are usually compared to the basal width of the scutellum.
Head characters are used primarily for males (see below). The vertex is defined as the entire area between the eyes from the frontoclypeal ridge to the concealed occiput. In most ciids, the area in the vicinity of the frontoclypeal or epistomal suture forms a ridge extending from eye to eye (Fig. 3); this is called the fronto- clypeal ridge and is composed of both frons and clypeus.
Tlie antennae may be 8-, 9-, or 10-seg- mented, with a 3-segmented club, although the first club segment is reduced in Had- raule hlaisdelli. The number of segments is diagnostic at various levels and their relative lengths may be of value at the specific level (ratio of III to IV is used in descriptions). The antennal club segments in the Ciidae are characterized by having at least four large, sensilla-bearing proc- esses, here called scnsilUfcrs (Figs. 1 and 2). These structures, which have been re- ferred to as "sensory pores" (Casey, 1898; Scott, 1926) or "ampoules a trichoides" ( Lesne, 1935 ) , appear to be homologous to the hygroreceptors of Triholiiim ( Roth and Willis, 1951a, 1951b) and the "organe sen- soriel" of Tijphlophlocus and Hypopldoeus {= Corticeus) (Jeannel and Paulian, 1945). I have seen similar structures in various tenebrionids (Gnathidiini, Diaperini, Ulo- mini, Strongyliini), Myrmechixenus (Coly- diidae), and Cryptophihis ( Languriidae ) . These organs may be useful at the generic level but are here restricted to the family diagnosis. Maxillary palps may also prove
useful, but they are not used extensively in the present treatment.
The pronotum varies considerably in relative size and shape, with rounded to subparallel sides. The lateral margins may be very narrow to broad and explanate (Figs. 24 and 22), and in some species there is a raised lip at the edge (Figs. 18 and 43). The edge itself may be smooth (Fig. 43) or crenulate (Fig. 44). The an- terior angles (where the dorsal and lateral edges meet) may be rounded (Fig. 15) or produced and angulate (Figs. 14, 18, and 22). The pronotal disc varies in sur- face sculpture and punctation. The surface is smooth and shiny or granulate and dull, and the punctures vary in both diameter and density. The average diameter is com- pared to either an eye facet or the base ol the scutellum, and as an index of density, the spaces between punctures are ex- pressed in terms of puncture diameters.
The elytra also vary in relative size and shape, with parallel to rounded sides and blunt to subacute apices. In Orthocis and; Strigocis, the elytral suture bears an in- flexed apical margin (Fig. 38). Elytra' punctation may be single (punctures rel- atively uniform in size and depth and al bearing bristles or hairs) or dual (punc tures falling into two classes differing ir size and depth, with only one bearing bris ties or hairs). The punctures vary in dis tribution from confused to seriate, anc also in coarseness and density.
The prosternum has been used exten sively, especially at the generic level. The subfamilies Ciinae and Orophiinae are dis tinguished mainly by the structure of the prosternum and procoxae (see generic key, p. 436), and the genera of Ciinae are usu- ally distinguished on the basis of prostema characters. The portion of the prosternun in front of the coxae may be long ( Fig. 28 , to very .short and striplike (Figs. 25 ano 27), while the intercoxal process may b(| broad (Fig. 26) or laminate (Fig. 24), anc' parallel-sided (Fig. 23) or tapering behinc (Fig. 27). In the Orophiinae, the inter
North American Ciidae • Lawrence 423
coxal process is very short (Fig. 25) and the procoxae extend well behind and below it (Fig. 16). The anterior part of the prosternum also varies in cross section, being concave or biconcave (Figs. 9-10) and on a different level than the intercoxal process (Fig. 15), or flat to carinate (Figs. 11-13) and on the same level (Fig. 14). The procoxae are almost always open be- hind (Figs. 22-25; 27-28), but in members of the Cis vitidus group the postcoxal bridges meet the intercoxal process to form a posterior closure ( Fig. 26 ) . In all Ciidae, the procoxae are open internally.
Tibiae have been used extensively in this group, and the apex of the protibia is particularly diagnostic. The outer apical angle may be narrowly rounded (Fig. 45), dentate (Fig. 50), or expanded and bear- ing several stout spines (Fig. 58). In ad- dition, the outer edge of the protibia may be spinose (Fig. 60) or serrate (Fig. 54). Spines may also occur at the inner apex (Fig. 54), but they are found in most spe- cies and are of little diagnostic value (these are excluded from most illustrations). Al- though the protibial apex is a useful char- acter at generic and specific levels, it must be treated with caution in certain groups because of sexual differences.
The metasternum varies in shape and convexity, while the metasternal suture varies in length and may be absent (Fig. 33). The hindwing is relatively simple with reduced anal region and a subcubital fleck; it exhibits little variation within the family, but some exotic ciids may be brachypterous or apterous. Genitalic characters have been studied in the male only and consist of the eighth abdominal sternitc and the aedea- gus. The latter consists of a small basal piece, a ventral tegmen, and a dorsal median lobe (Figs. 85 and 86). The latter two structures and sternitc VIII have been included in most descriptions, but these genitalic characters are rarely used in the keys. The terminology is taken from Shaip and Muir (1912), but Lindroth (1957) should be consulted for alternative names.
Secondary Sexual Characters
In the great majority of ciid species, males have a pubescent, glandular struc- ture in the middle of the first visible ab- dominal stemite ( III ) , which is here called a pubescent fovea or abdominal fovea, but which has been variously referred to as a "tubercule velu" (Abeille de Perrin, 1874b), "setigerous fovea" (Casey, 1898), "behaartes Grlibchen" (Reitter, 1902a), "setiferous pit" (Miyatake, 1954), "median depression" (Kevan, 1967), "setigerous pore" (Lawrence, 1967b), and "Auszeich- nung" (Lohse, 1967). Tlie structure may be simple (Fig. 32), margined, or raised and tuberclelike, and in the Orophiinae it may be covered by a triangular flap (Fig. 31). Similar abdominal structures are found in males of many different beetles, including Erotylidae (Delkeskamp, 1959), Sphindocis, Dermestes, and Blaps (Meix- ner, 1934). In the Ciidae, it varies in size and shape and is useful mainly at the spe- cies level. The greatest diameter or length of the fovea may be compared to the "body" of stemite III or that portion of the sternite behind the intercoxal process. Among the North American ciids, the fol- lowing species lack the fovea in the male: Cis con^estus, C. horrididus, C. hijstriculus, C. huachucae, C. subtilis, and C. vitidus.
Pubescent foveae may also occur on the vertex of the head of the male. This is the case in DoUchocis manitoba, the Cis paci- ficus group, and the "Ennearthron" filum group. Median and/or lateral tubercles occur on the male vertex in some species, such as Cis niedhauki (Fig. 5) and certain Xylo<i.rapJius and Octotemnus.
The frontoclypeal area is relatively simple in most Orophiinae, but in the Ciinae the development of a frontoclypeal ridge provides a good taxonomic character for species and species groups. This ridge may bear two teeth or tubercles ( Fig. 40 ) , two triangular plates (Fig. 39), two long horns (Fig. 42), a median horn (Fig. 6), four teeth (Fig. 4), or a raised, trisinuate
424 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
plate (Fig. 3). The anterior edge of the pronotiim is also modified in some species, forming a median process ( Fig. 41 ) or two homs (Fig. 7). There are many different types of frontoclypeal and pronotal orna- ments, and a particular type is often char- acteristic of a species group. The major difficulty in using these characters is that they tend to vary allometrically, so that those of larger males may differ greatly from those of smaller ones and the latter may be similar to those of females or small males of related species.
Modifications of the mandibles occur only in the Orophiinae and in uo North American species. Xylographus males may have a tooth on tlie left mandible, while in certain Old World Octotemmis both man- dibles are enlarged.
Measurements and Ratios
Pronotal length (PL) is measured along the midline and in males includes homs or laminae. Pronotal width (PW) is the greatest width. Elytral length (EL) is taken just to one side of the midline and from the base of the scutellum to the ely- tral apex. Elytral width (EW) is the greatest combined width of both elytra. Greatest depth (GD) is taken through the elytra and metasternum. The total length (TL) is the sum of PL and EL and does not include the head; it is given in milli- meters. For one sample of each species described, the range, mean, and standard error of the mean are given for TL and for the following ratios: TL/EW, PL/PW, EL/EW, EL/PL, and GD/EW. In the descriptions, these ratios are given for the holotype and allotype and in the keys they are used only when there is little or no overlap between alternatives. They have proven very useful in distinguishing be- tween species and even higher groups of Ciidae, because of the large amount of variation in general body fonn occvirring in the family. Other ratios included in de- scriptions are: antennal segment III/IV;
prostemal intercoxal process width/pro- coxal cavity width; length/width of meta- sternum; length of metastemal suture/ median length of metasternum (including its anterior process ) ; and, length of abdom- inal fovea/median length of sternite III ( body only, excluding its anterior process ) .
Observations and Drav/ings
Observations were made of dried, fluid- preserved, and slide-mounted material with the aid of a Leitz stereoscopic microscope (12.5X and 18x oculars, Ix, 2x, 4x, 8x, and 12x objectives) and a Bausch and Lomb compound microscope ( lOx oculars, lOOx and 440x objectives). Measurements were made with an ocular linear microm- eter, and drawings were made with the aid of ocular grids in the eyepieces of both microscopes. Initial drawings were trans- ferred from graph paper to white board and inked with a series of rapidograph pens. In most of the drawings, no attempt has been made to show relief, and stippling or solid shading has been used to set off specific areas (coxal cowling, abdominal fovea, eyes) or to indicate pigmentation (aedeagus, sternite VIII). Dotted lines have been used to indicate a sharp bend in a surface (raised frontoclyT^eal ridge, lateral pronotal margin) or a hidden out- line (protibial apex, sensillifers, overlap- ping abdominal sternites).
Synonymies and Descriptions
Synonymies have been reduced to in- clude only author, date, and page, l)ut complete references are cited in the ter- minal bibliography. The species synonymies are relatively complete, but generic synony- mies include only major works. Complete descriptions are given for new species only; for previously described fomis and for all genera a short diagnosis is included in the discussion. The species description is usu- ally based on the holotype and is followed by a brief indication of differences in the
North American Ciidae • Lawrence 425
allotype. The male genitalia figured are not those of the holotype but have been dissected from one of the male paratypes. The protibiae illustrated have also been removed from paratypes.
Variation
Each species description is followed by a section presenting the range of observed variation in size, form, color, vestiture, and secondary sexual characters. One sample of each sex is treated statistically in the manner described above. Geographic vari- ation is not treated in depth in the present paper, although there are several groups that obviously require further study at this level (Cis omericanus, page 444; Cis cre- berrimiis, page 452; Ortlwcis punctatus, page 486).
Label Data and Collections
Because of the large numbers of speci- mens examined, complete label data are given only for the type series. In all other cases, localities are listed (alphabetically by states and provinces), collecting dates are excluded, and host data is summarized. Maps are also included for the majority of North American species. The summari- zation of host data is discussed on page 435. Sources of material (institutions and private collections) are also listed only for types. The following abbreviations are used in the text to refer to locations of types and paratypes: BMNII, British Mu- seum (Natural History); BRUS, Institute Royale des Sciences de Belgique; CAS, California Academy of Sciences; CIN, Cin- cinnati Museum of Natural History; FMNH, Field Museum of Natural His- tory; GEN, Museum d'Histoire Naturelle, Geneva; JFC, J. F. Cornell Collection; JFL, J. F. Lawrence Collection; MCZ, Museum of Comparative Zoology; MNHN, Museum National d'Histoire Naturelle, Paris; MZUH, Museum Zoologicum Uni- versitatis Helsinki; PURD, Purdue Univer- sity; UAZ, University of Arizona; USNM, United States National Museum; UW, Uni-
versity of Washington; UWS, University of Wisconsin.
SYSTEMATICS OF THE FAMILY CIIDAE Historical Review
The earliest recognized name in the Ciidae is Scopoli's Dermestes holeti, but the genus Cis was not described by La- treille until 1796 and the family Ciidae (Cisidae) was proposed by Leach in 1819. The first world monograph of the family was that of Mellie (1848), which included 106 species in eight generic groups, and no other world study has been attempted since. Contributions made within the next eighty years consisted mainly of European faunal works (Thomson, 1863; Abeille de Perrin, 1874b; Kiesenwetter, 1877; Reitter, 1902a) and exotic faunal surveys resulting from foreign expeditions (Gorham, 1883, 1886; Scott" 1926). The Junk catalogue for the family (Dalla Torre, 1911) included 19 genera and 233 species. Outside of North America, most of the work on the family in the last forty years has concerned the Japanese fauna (Chujo, Miyatake, Nakane, Nobuchi) and the Pacific fauna (Blair and Zimmerman). In addition. Pic described almost 100 species from all over the world, and recently European workers, such as Lohse (1964-1969), have renewed interest in the Palaearctic Ciidae.
Although Mellie's monograph included several North American species and Man- nerheim described a few from expeditions to Russian America, LeConte and Hom almost ignored the group, and the first major treatment in this country was that of Casey (1898), which included 8 genera and 44 species. Kraus ( 1908 ) and Blatch- ley (1910) added a few more fonns, and in 1914 Dury published his North American synopsis that was meant to serve as a supplement to the earlier paper by Casey. Very little has been published since on the North American fauna, with the exception of Hatch's coverage in "Beetles of the Pacific Northwest" ( 1962 ) and my recent papers (Lawrence, 1965, 1967a, 1967b).
426 Bulletin Museum of Comparative Zoologij, Vol. 142, No. 5
Family Limits
The composition of the family has varied o\'(n- the years, and several forms have been removed to other groups. Two genera that are obviously not allied to the Ciidae are Hendecatomtis Mellie and Rliipidandms LeConte. The former was included in the family by most authors until Lesne ( 1934, 1935) presented considerable evidence for its removal to the Bostrichidae. The tme relationships of Hendecatomtis had been recognized much earlier, however, by Jacquelin du Val (1861) and LeConte (1861), and the latter had proposed the tribe Hendecatomini within the Bostrichi- dae. The genus RJiipidandnis was placed in the Ciidae by LeConte and Horn ( 1883), although it was originally described as a tenebrionid (LeConte, 1862) and has been treated as such by most workers (see Bar- ber, 1913). In spite of the reduced tarsi (4-4-4) and pectinate antennae, there is little doubt that the group belongs in the Tenebrionidae and is probably related to the Eledona Latreille.
The genus Pterogeniu.^ Candeze is not as easily dispensed with. It was included in the family Ciidae until Crowson ( 1955 ) transferred it to a new family (Pterogenii- dae) along with llistanocerus Motschulsky ( = Lahidocem Gebien). Tliese two genera belong among the primitive Heteromera as defined by Cro^^'son (1955, 1960, 1966, 1967) and are thus more or less closely related to the Ciidae; they differ from all ciid genera, however, in a number of char- acters, including the following: 1) anten- nae 11-segmented, filifonn or gradually enlarged apically, without sensillifers; 2) maxillary palps securiform; 3) tarsal formula 5-5-4; 4) procoxal cavities closed internally; 5) mesotrochantins visible; 6) abdominal stemites III and IV connate; 7 ) anal region of hindwing with four veins, the anterior one running through the sub- cubital fleck, and a wedge cell. It is un- likely that this family represents the sister group of the Ciidae, but it must be taken
into consideration in any study of the primitive heteromerous Coleoptera (Crow- son, 1966).
Another genus doubtfully included in the Ciidae is Sphindocis Fall (1917), described on the basis of a single species, S. deiiti- coUis, from coastal California. The beetle has been collected at various localities from Alameda County to northern Mendocino County and is usually found breeding in the fruiting bodies of Trametes sepium growing on dead and fallen branches of Madrone (Arbutus Menziesii). The adult of Sphindocis resembles a large Orthocis and exhibits the following similarities to members of the Ciidae: 1) procoxae with- out lateral extensions, trochantin hidden; 2) procoxal cavities open internally and posteriorly; 3) mesepimera reaching meso- coxal cavities; 4) hind\^'ing with subcubital fleck; 5) abdominal sternite III in male with pubescent fovea; 6) tarsal fonnula 4-A-A in both sexes; 7) aedeagus of in- verted heteromeroid type. Tire Sphindocis larva is similar to a ciid lan'a in general form (see below), and the mouthparts do not differ from those of a ciid larva in any significant manner. Tergite IX is modified to form a sclerotized, concave disc, similar to that found in Cis ynelliei Coquerel (1849) and in the tenebrionid Meracantha contracta (Beauvois) (Hyslop, 1915). A number of these adult and lan^al features may be found in other Cucujoidea, how- ever, and the differences presented in Table 1 argue for the exclusion of the genus from the Ciidae. Crowson (in Utt.) has suggested that Sphindocis may repr(>- sent the sister group of the Ciidae proper and that both groups might be included in one family. I think it is just as likely that the former is allied to the Tetratomidae or Prostomidae (see below) and I prefer to exclude it in the present treatment.
Characterization of the Family Ciidae
The following description will serve to distinguish members of this familv from
North American Ciidae • Lawrence 427
Table 1. Differences between Sphindocis ant) the Ciidae
Characters
Sphindocis
Ciidae
Antenna
Maxilla
Mesotrochantin Metendosternite
Anal region of hindwing
Abdominal sternites III and IV
Trochanters
Tibial spurs
Basal piece of aedeagus
Median lobe of aedeagns
Larval spiracle
Larval antenna
Larval stemite IX
no sensillifers 11-segmented
with 2 well- developed lobes
visible
With median stalk
with 4 veins and wedge cell; subciibital fleck undivided
connate
heteromeroid type
present
large witli 2 condyles
membranous with lateral struts
biforous
3-segmented
with row of asperites
8- to 10-segmented with sensillifers
with reduced lacinia
not visible
without median stalk
with 1 vein;
subcubital fleck divided
free
normal type
absent
small, without condyles
sclerotized
annular- 2-segmented without asperites
all other Coleoptera. A more complete comparative study of adults and larvae will be the subject of a paper in prepa- ration.
With the general characters of the Po- lyphaga: Cucujoidea.
Adult. Form variable, usually oval to elongate, convex. Size 0.5-6.0 mm.
Head globular, without neck, declined, often strongly so, partly concealed by pro- notum. Eye somewhat protuberant, oval, entire, fairly coarsely faceted. Frontoclyp- eal area with distinct suture, often raised in males to form a ridge. Antennal in- sertion in fossa formed by genal ridge and eye, concealed from above by frons. An- tenna 8- to 10-segmented, with large pedi- cel and scape and 2- or 3-segmented club.
each club segment bearing at least 4 sen- sillifers at apex. Mandible bidentate, with simple molar area. Maxilla with galea and lacinia reduced, palp 4-segmented with terminal segment subconical, not securi- form. Labium with ligula absent, palp 3-segmented.
Pronotum margined laterally and pos- teriorly, anterior edge usually produced forward. Prostemum variable, long or short, concave to carinate, coxae globose or transverse, sometimes projecting, contigu- ous to broadly separated, without internal- ized lateral extensions, trochantin hidden. Procoxal cavities open internally, narrowly open or closed posteriorly.
Elytra not striate, humeri tuberculate, epipleurae very narrow, extending almost
428 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
to apex. Sciitellum small and siibtriangular. Wing venation reduced, \\'ith 1 anal vein (or none); subcubital fleck present, di- vided.
Mesostemum short, coxae globose and narrowly separated, coxal cavities not closed outward!)' by sterna, trochantins hidden. Metasternum subquadrate, with or without median suture, coxae narrow, transverse, subcontiguous. Metendostemite consisting of a pair of diverging anns with anterior tendons near apices.
Tarsal formula in both sexes 4-4-4 (oc- casionally 3-3-3). Tarsi simple, first 3 segments small and subequal, temiinal seg- ment elongate, claws simple. Trochanters oblique, normal type (completely separat- ing coxa from femur). Tibiae without api- cal spurs, outer edge of protibia often expanded and modified at apex.
Abdominal segments all freely articu- lated, first segment without coxal lines, often with a median pubescent fovea m male. Aedeagus of inverted heteromeroid type, with small basal piece, ventral teg- men, and dorsal median lobe.
Larva. Orthosomatic, without lateral ex- tensions, lightly and evenly sclerotized except for head, thoracic tergite I, and abdominal tergite IX ( occasionally VIII ) . Head with Y-shaped epicranial suture, 5 ocelli or less, and short gula. Antenna 2- segmented, second segment bearing a long, ventral, sensory appendage near base and a long seta at apex. Mandible bidentate, with or without "retinaculum," with or without molar area. Maxilla with obliquely obtuse mala and a small, dorsal, subapical lobelike lacinia. Spiracles small, annular. Two setae on claw. Tergite IX variously armed but usually bearing 2 "urogomphi"; sternite IX without asperites. Segment X pygopodlike.
Phylogenetic Relationships
The placement of the Ciidae \\'ithin the order Coleoptera has a complex history, and the group has been associated at various times with the Bostrichoidea,
Cleroidea, Clavicornia, and Heteromera. In the 19th Century, the family was com- monly placed with the Bostiichidae and their allies (TerediHa, Xylophages, Bostri- choidea) primarily on the basis of the cylindrical form, declined head, expanded and often spinosc tibiae, and other fcatiu-es associated with the boring habit. Casey (1890) considered the group to comprise a subfamily of the Cryptophagidae, which was used in a very broad sense to include the Mycetophagidae, Sphindidae, and Bi- phyllidae as well.
Forbes ( 1926 ) united the Ciidae with the Lathridiidae, Corylophidae, and Mur- midiidae (= Cerylonidae) on the basis of wing venation and folding, and he derived this group of families from the Endomychi- dae and Colydiidae partly on the basis of the "double chitini/ation" in the anal region of the wing. This anal chitinization refers to the subcubital fleck of Crowson (1955), which is found in a number of cucujifomi families and which may be divided by the first anal vein and remains divided even after the vein has been lost. A divided subcubital fleck occurs in several families, including the Endomychidae, Lathridiidae, Biphyllidae, Byturidae, Colydiidae, Ptero- geniidac, and Ciidae. An undivided fleck occurs in the Mycetophagidae, Tetratomi- dae, Spliindocis, Salpingidae, and several other groups ( Bernet-Kempers, 1923; Crowson, 1955; Forbes, 1926; Miyatake, I960; Wilson, 1930). A further study of this character ( whose function is at present unknown ) may shed some light on relation- ships among the Cucujoidea.
Boving and Craighead (1930, 1931) in- cluded the Ciidae in the Cleroidea on the basis of larval characters, such as the lack of a mandibular mola (actually present in some Ciidae) and the reduction of the maxillary articulating area (also occurring in some cucujoids), while Jeannel and Paulian ( 1944 ) related the family to the Colydiidae, Byturidae, Boridae, and several other families on the basis of male geni- talia.
North American Ciidae • Lawrence 429
Crowson ( 1955 ) presented convincing evidence for excluding die Ciidae from the Bostrichoidea and Cleroidea but expressed doubt as to the affinities of the group within the Cucujoidea. In later works (1960, 1966, 1967), Crowson placed the Ciidae within the section Heteromera and indicated possible relationships to several of the more primitive families, such as the Biphyllidae, Byturidae, Mycetophagidae, Pterogeniidae, Tetratomidae, Prostomidae, and Colydiidae.
I agree basically with Crowson's view that the Ciidae are primitive members of the Heteromera, but the affinities of the group to other heteromerous families are not at all clear, and the possibility still exists that they have been derived inde- pendently from some clavicom stock, such as the Languriidae. The prothoracic struc- ture resembles that of the Mycetophagidae, Tetratomidae, and S})]iindocis, in that the coxae are not internalized (lacking lateral extensions) and the cavities are open internally and posteriorly. Other primitive heteromeran characters include the free abdominal stemites, nonheteromeroid tro- chanters, and annular larval spiracles. Specialized features, such as the reduction of antennal segments, maxillary lacinia, tarsi, and anal region of hindwing, and the loss of tibial spurs and the median stalk of the metendosternite all may be correlated with size decrease. The pygopodlike 10th abdominal segment in the larva occurs in both the Clavicomia and Heteromera and may be associated with the habit of boring into fimgi.
As mentioned above, the Ciidae, Sphin- docis, and the Pterogeniidae are similar in several respects. In all three, the procoxae are not internalized, the aedeagus is of the inverted heteromeroid type, and the larval mouthparts are similar, the maxilla, for instance, possessing a distinct laciniar lobe. In Spliindocis, however, the 9th larval ventrite bears a row of asperites, the antennae are 11-segmented without
sensillifers on the club, the trochanters are heteromeroid, and the subcubital fleck is not divided; in the Pterogeniidae, the pro- coxal cavities are internally closed, the antennae are filiform, the tarsal formula is 5-5-4, and the maxillary palps are strongly securiform; while in both of the latter groups, the lanal spiracles are biforous, the first two abdominal stemites are connate, tibial spurs are present, and the mesotro- chantins are visible. Sphindocis appears to be more closely related to the Tetratomidae and Mycetophagidae, while the Pterogenii- dae may have affinities with the Neo- tropical genus IscJujomius or perhaps to the byturid-biphyllid group.
Other primitive heteromerous families with which the Ciidae might be associated are the Byturidae, Biphyllidae, Prostomi- dae, Colydiidae, and perhaps the Tene- brionidae and their close allies. It is also possible that the Heteromera, as it is now constituted, does not represent a monophy- letic group, in which case certain of die more primitive forms might be derived independently from different clavicom an- cestors. The wing venation in the ciids is similar to that of certain cerylonoid groups, such as the Lathridiidae and Endomv- cliidae, while antennal sensillifers are found in the languriid genus Cnjptophilus and the overall adult structure in the Ciidae is closely approached in another languriid genus Setariola. Similar wing venation (with divided subcubital fleck) may also be found in the heteromerous Colydiidae, however, and antennal sensilli- fers occur in Mijrmechixenus (Colydiidae?) and Szekessya (Prostomidae?), as well as in a variety of true Tenebrionidae.
Tlie position of the family Ciidae must remain in doubt for the present, until a thorough phylogenetic study of the primi- tive Heteromera is completed. The phylo- genetic relationships within the family Ciidae are also unclear and will not be dis- cussed in detail here. The basic division into subfamilies appears to be sound, and die prothoracic differences used in the
430 Bulletin Museum of Coiiiparative Zoology, Vol. 142, No. 5
Table 2. Distributional patterns of faunae affinities of North American Ciidae.
Ciid species
NW
NE
SE
SW
Faunal affinities
CIS
acritus
americanus
angvistus
biannatus
castlei
cayensis
congestiis
comelli
cornutus
creberrimus
crinitus
discolor
dunedinensis
duplex
ephippiatiis
festivuliis
florid ae
fuscipes
hirsutus
horridulus
liuachucae
hystriculus
krausi
laniinatus
levettei
maritimiis
megastictus
miles
niedhaiiki
pistoria
qviadridentatiis
robiniophilus
rotundulus
stereophilus
striolatiis
subfuscus
subtilis
tetracentrum
tridentatus
tristis
ursulinus
versicolor
vitiilus
ENNEARTHRON aurisquamosum spenceri
DOLICHOCIS
indistinctus manitoba
ORTHOCIS
hiiesanus
longulus
pulcher
X XX XX XX
XX
(xx)
XX XX
X X
XX XX XX XX XX XX XX
XX
XX
XX
XX
|
XX |
X XX |
XX |
|
|
X |
X |
X XX |
|
|
X |
XX |
X |
X |
|
X |
XX |
||
|
XX |
(xx) |
XX |
|
|
X |
XX |
X |
|
|
XX |
X |
||
|
XX |
|||
|
X |
XX XX |
||
|
X |
XX |
||
|
X |
XX |
||
|
XX |
XX |
||
|
X |
XX |
||
|
X |
XX |
X XX |
|
|
X |
XX |
XX |
|
|
XX |
|||
|
X |
XX |
||
|
X |
XX |
||
|
X |
XX |
||
|
XX |
X |
(1)
(2)
(1)
Neotropical?
Palaearctic?
Palaearctic
Palaearctic
Neotropical
Neotropical ( 1 )
Neotropical
Neotropical
Neotropical
Neotropical
Neotropical
Neotropical (2)
Neotropical ( 1 )
Neotropical
Palaearctic
Palaearctic
Neotropical
Palaearctic
Neotropical ( 1 )
Palaeaictic
Neotropical
Palaearctic
Neotropical ( 1 )
Palaearctic
Palaearctic
Palaearctic
Palaearctic
Neotropical
Neotropical
Palaearctic
Oriental?
Palaearctic
Neotropical
Neotropical
Palaearctic
Neotropical
Neotropical?
Neotropical
Palaearctic?
Neotropical?
Neotropical ( 1 )
Neotropical
Neotropical
Oriental? Palaearctic
Palaearctic
(1)
(1)
XX XX XX
Neotropical Neotropical Neotropical ( 1 )
North American Citdae • Laivrence 431
Table 2. (Continued)
Ciid species
NW
NE
SE
SW
Faunal affinities
ORTHOCIS (Continued) punctatus transversatus
STRIGOCIS bilimeki opacicoUis opalescens
HADRAULE blaisdelli elongatula explanata
PLESIOCIS cribrum
CERACIS calitornicus curtus dixiensis magister minutissimus minutus monocerus multipunctatus nigropunctatus obrieni pecki powelli pullulus punctulatus quadricornis sallei schaefteri similis singularis thoracicornis
SULCACIS curtulus lengi
MALACOCIS brevicoUis
RHOPALODONTUS
americanus
OCTOTEMNUS
laevis
XX
X X
XX
XX
X X
X
(xx)
XX
XX
XX
X
XX
XX
XX
XX
(xx)
XX
|
X |
XX |
|
|
X |
XX XX |
XX |
|
x |
XX XX XX XX XX |
XX |
|
X |
XX XX |
XX |
|
X |
XX XX |
|
|
X |
XX XX |
(XX |
|
X |
XX |
|
|
X |
XX |
|
|
X |
X |
X |
|
X |
XX |
Palaearctic
Neotropical ( 1 )
Neotropical (2)
Neotropical (2)
Neotropical (2)
Palaearctic Palaearctic
Neotropical Neotropical ( 1 ) Neotropical Neotropical
Neotropical ( 1 ) Neotropical Neotropical ( 1 ) Neotropical Neotropical Neotropical Neotropical (2) Neotropical ( 1 ) Neotropical ( 1 ) Neotropical
Neotropical Neotropical Neotropical
Palaearctic
Neotropical
Palaearctic
Palaearctic
key (p. 436) are correlated with a number revision. It is hoped that an improved
of other characters. The Orophiinae con- generic classification, based on lai-val and
tains several distinct genera, while the adult characters, will result from a study
much larger Ciinae requires considerable now in progress.
432 Bulletin Museuin of Comparative Zoology, Vol 142, No. 5
Table 3. Comparison of faunal sectors.
NW NE SE SW NW
|
species |
26 |
|
Indigenous |
|
|
species |
15 |
|
Endemic |
|
|
species |
6 |
|
Shared |
|
|
species |
|
|
S. C. |
26 45 53 20 26
10 44 13 15
6 21 7 6
16 29 5 11
61.5 64.4 25.0 55.0
Family Name
The spelling of the family name has been a subject of controversy for many years. The family was originally named Cisidae by Leach (1819), and Wollaston (1854) used Cissidae. Gistel (1856) ap- pears to be the first to use Cioidae, and this spelling was used by most German authors (Kiesen wetter, 1877; Reitter, 1902a) until recently. Ciidae was first used by Marseul ( 1887 ) and has been adopted by a number of recent workers. The type genus Cis is derived from the Greek mascu- line noun kis, kios, meaning a wonn which bores into wood. The genitive stem is ki-, so that the correct family name should be Ciidae. There is no doubt that Ciidae is the correct spelling, but Grensted ( 1940, 1947) has suggested that this name be re- placed by one that is more euphonious and more clearly linked to the genus name Cis. Among recent authors, Amett (1962), Crowson ( 1955 ) , and Lohse ( 1967 ) have all used Cisidae, while I have continued to use Ciidae, as have Miyatake ( 1954 ) and other Japanese workers. This is a relatively unimportant matter and is included here only to clarify the inconsistent spellings found in the literature.
THE NORTH AMERICAN CIIDAE
The family Ciidae in North America in- cludes 12 genera and 84 species, 16 of which are here described as new. The
majority of species fall into the two genera Cis (43) and Ceracis (20), while most of the genera are represented by one or two species only.
Origin and Distribution
As would be expected the major affinities of the North American ciids are with the Neotropical and Palacarctic faunas. One or two species may be related to Oriental forms, but there is no evidence for Ethiopian or Australian affinities. A gen- eral faunistic analysis of the North Amer- ican species is presented in Tables 2 and 3. The continent has been divided into four sectors, as shown in Figure 87, and in Table 2 the occurrence of a ciid species within a sector is indicated by an "x," while an "xx" is used for the sector in which the species is assumed to be indig- enous ( or at least most common and wide- spread relative to the remainder of the area considered). Doubtful occurrences or possible introductions are indicated by "(xx)." Finally, the last column gives the zoogeographic region in which the closest relatives of the species occur. The number ( 1 ) after Neotropical indicates West Indies, while the number (2) stands for the Mexican Plateau. Table 3 gives the total species, indigenous species, and en- demic species for each sector and an ex- pression of faunal similarity between each adjacent sector, using Simpson's Coefficient (Simpson, 1947; S. C. = lOOC/nj, where C is the number of species common to the two faunas and W\ the number of species in the smaller fauna ) . This is a very coarse analysis, since relatively broad and arbi- trary sectors have been chosen, but it will serve to emphasize some of the major features of ciid distribution in North America.
A more relevant type of distributional study is one which takes into consideration geographical origins and phylogenetic affinities, such as that of Linsley ( 1958 ) on the Cerambycidae. Since the Ciidae are
North American Ciidae • Lawrence 433
associated with woody plants, as are the cerambycids, it would be useful to compare Linsley's faunal elements with those based on the Ciidae. The North American ciids are most easily grouped into three major faunas, one of which may be further di- vided into three subfaunas. These are dis- cussed below.
Nortlwm fauna. This includes 26 species which occur, for the most part, in the northern forests and mountainous regions of North America, and are usually allied to forms from northern Eurasia.
Cis amcricanus Cis auLiusttis Cis biarmatiis Cis ephippiatus Cis festivuhis Cis fuscipes Cis honidiilus Cis hystriculus Cis laminattis Cis levettei Cis maritimus Cis mcgasticius Cis pistoria
Cis robiniopJiilus Cis striolatus Cis tridentatus Enn. spenceri Dol. indistinchis Dol. iuanito])a Ortli. piinctatus Hadr. elongatida Iladr. explanata Pies, cribnim Side, curtidtis Rhop. amcricanus Oct. laevis
Some of these species are equally com- mon in the Northeast and Northwest (O. laevis. Fig. 109), others may be common in the western forests but rare in the North- east (P. crihrum. Fig. 89), and still others are restricted to the Northwest Coast (C. hiarmatus. Fig. 102). Some forms occur in association with hardwoods (S. ciirtidiis, C. fuscipes, O. laevis, and C. pistoria), and others are found on conifers (C. hiarmatus, P. crihrum, and C. lu/stricuhis), but the majority may be found on either. Most of the species comprising this fauna would form part of the modern Holarctic element of Linsley and are related to or even synonymous with northern Palaearctic forms. The following species pairs (Nearctic-Palaearctic) are very closely al- lied and some may be conspecific: Cis horriduluS'C. tomentosus Mellie; Cis hijstri- cuhis-C. punctulatus Gyllenhal; Cis levet- tei-C. glcihratus Mellie; Cis pistoria-C.
micans (Fabricius); Cis striolatus-C. stri- atidus Mellie; Dolichocis indistinctus-D. laricimis (MeHie); Orthocis punctatus-0. alni (Cyllenhal); Sulcacis curttdus-S. hi- denttdus (Rosenhauer); Rhopalodontus americanus-R. strandi Lohse; Octotemnus laevis-O. g,lahricuhis (Gyllenhal). Three of these northern species (Cis laminatus, Ennearthron spenceri, and Hadraule elon- gatula) may represent recent introductions. Two species, Cis festivuhis and C. rohini- ophdus, also have Palaearctic counterparts — C. festiviis (Panzer) and C. castanetis Mellie, respectively — but are more south- em in distribution and may represent an older element which Linsley called the Alleghenian. In Western North America, there are several coastal forms (Cis ani:.us- tus, C. hiarmatus, C. ephippiatus, C. mari- timus, C. megastictus, and C. tridentatus) which do not have Palaearctic near rel- atives, but which belong to Holarctic spe- cies groups. These may be part of the Vancouveran of Linsley, in that their Old World affinities are not as close. Finally the two species Dolichocis manitoha and Plesiocis crihrum appear to be the most isolated fomis with no obviously related species in the Old World or in the Neo- tropical Region.
Southwestern fauna. This group includes 15 species ( one of which is also mentioned in the next section), which would fonn part of the Sonoran and Califomian faunas of Linsley. Their ranges extend from northern California through the Southwest into Mexico.
Cis acritus Cis creherrimus Cis discolor Cis duplex Cis huachucae Cis tetracentrum Cis versicolor
Cis vituhis Str. hUimeki Iladr. hlaisdelli Cer. californicus Cer. dixiensis Cer. ohrieni Cer. poicelli Cer. similis
Only four of the above species occin- in central and northern California, while the
434 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
remainder are restricted to the Soutliwest. Cis vitulus is practically restricted to Cali- fornia, occurs in mesic and humid habitats, and is most closely related to Cis congesttis of the Southeast and to several Neotropical species; this is the only species that I would consider part of Linsley's Californian sub- fauna. Cis versicolor, Hadraiile bluisdelli, and Ceracis calif ornictis extend north along the Pacific Coast, but are also common in the Southwest at low and intermediate elevations. Cis acritus, C. creherrimus, C. discolor, C. duplex, and Ceracis powelli are usually found in coniferous forests at higher elevations, while Cis tetracentrum occurs in mountain canyons, along with Cis versicolor, Ceracis californicus, and Ceracis dixiensis. The last three species also inhabit the arid Ceracis obrieni and similis.
Southeastern fauna. great majority of North American Ciidae occurring east of the 100th Meridian and corresponds, for the most part, to Linsley's Neotropical Fauna. It may be subdivided into three groups: 1) the main group of species ranging from New England and the Great Lakes south to Florida and Texas, 2) the Antillean group occurring mainly in Florida and adjacent states, and 3) the Mexican group extending into southern Texas and along the Gulf Coast to Louisi- ana. These will be discussed separately be- low.
lowlands, as do probably Ceracis
This includes the
1) Main group ( Cis castlei Cis congestus Cis cornelli Cis cornutus Cis creherrimus Cis miles
Cis quadridentatus Cis rotundulus Cis stereophilus Cis sidy t His Cis tristis Cis ursulinus
24 species). Enn. aurisquamosum Orth. longulus Str. opacicollis Sir. opalescens Cer. minutissimus Cer. pecki Cer. punctulatus Cer. sallei Cer. singtdaris Cer. thoracicornis Sulc. lengi Mai. brevicollis
Most of the species in this group have Neotropical affinities and the ranges of many extend into northern Mexico. Those with continuous distributions into Mexico (usually somewhat broken up in the more arid regions) are probably more modem Neotropical elements, while those with more restricted southeastern distributions may represent older Alleghenian forms. Cis congestus (Fig. 105), with its counter- part C. vitulus in California, is probably one of the latter, while Ennearthron auri- squamosum represents an Alleghenian spe- cies with Old World affinities, its closest relative occurring in southwestern China. Cis quadridentatus is a rather widespread form and its relationships are obscure; it may be related to an Oriental group of species. Cis rotundulus and C. ursulinus belong to a West Indian species group, but they appear to have evolved on the main- land and are more widespread than the Antillean fonns discussed below (Figs. 89 and 92).
2 ) Antillean group ( 16 species ) .
Cis cayensis Cis crinitus Cis dunedinensis Cis floridae Cis hirsutus Cis krausi Cis 7iiedhauki Orth. huesanus
Orth. pulcher Orth. transversatus Cer. curtus Cer. magister Cer. minutus Cer. monocerus Cer. multipunctatus Cer. pullulus
Most of the species in this group occur only in southern Florida and the majority are also found in the Greater Antilles. Cis krausi and Orthocis transversatus also oc- cur in Texas and might be included in the next group, but both have close relatives in the West Indies.
3) Mexican group (4 species). Cis suh- fuscus, Ceracis nigropunctatus, Ceracis quadricornis, and Ceracis schaefferi are. Neotropical species that occur in Texas and Mexico but are not found in the W(>st Indies. There are probably several more undescribed forms that extend into southern!
North American Ciidae • Lawrence 435
Texas, but the fauna of that region is not well known at present.
Host Preference
The subject of host preference or speci- ficity in the North American Ciidae has been discussed briefly in a pre\1ous paper (Lawrence, 1967b) and will be covered in detail in a future publication (Lawrence, in press).
More than 100 species of fungi have been recorded as ciid hosts in North America, and the great majority of these belong to the basidiomycete family Poly- poraceae. In the following text, all host records are listed for each ciid species. For each fungus, the total number of records and the number of "breeding" records (in parentheses ) are given. A particular record is considered to be a breeding record if it consists of any one of the following: 1) ten or more fully pigmented adults, 2) two or more tenerals only, 3) one teneral and two or more fully pigmented adults, or 4) one or more larvae and/or pupae. Species of fungi are listed according to the number of records, so that preferred hosts are first. An example is as follows: Fohjporus anceps [6(5)]; Fomes officinalis [2(2)]; Fomes pinicola [2(2)]; Ganoderma oregonense [1(1)]. The classification of fungi used in the presentation of host data is a rather conservative one, following Lowe (1957, 1966), Lowe and Gilbertson (1961a, 1961b ) , and Overholts ( 1953 ) .
Patterns of host preference have been demonstrated for British Ciidae (Paviour- Smith, 1960, 1969) and for North American Ccracis (Lawrence, 1967b). The following is a preliminary grouping of 58 North American ciids into the four host prefer- ence groups discussed in the Ceracis paper. A detailed analysis of host data will be presented at a later date (Lawrence, in preparation ) .
PoJyponis versicolor group. Fungi with thin, whitish, coriaceous fruiting bodies and trimitic hyphal system {Polyponis versi-
color, P. hirsutus, Lenzites betulina, Trametes hispida, etc.).
|
Cis congestiis |
Str. bilimeki |
|
Cis corntitus |
Str. opacicollis |
|
Cis fuscipes |
Str. opalescens |
|
Cis miles |
Cer. dixiensis |
|
Cis pistoria |
Cer. mimitus |
|
Cis siibfuscus |
Cer. qiiadricornis |
|
Cis tet race nt rum |
Cer. monocerus |
|
Cis tristis |
Side, cmiulus |
|
Cis versicolor |
Side, lengi |
|
Cis vituhis |
Oct. laevis |
Polyponis pargamemis group. Fungi similar to the above but with brownish or purple pore surface {Polyponis ahietimis, P. pargamemis, P. sector, Daedalea uni- color).
Cis acritus Cis horridtdiis Cis hystricidiis Cis striohtiis
Cis siibtilis Cer. minutissimiis Cer. powelli Cer. thoracicornis
Polyponis gdviis group. Fungi with woody or fibrous fruiting bodies, brownish color and tinning black in potassium
m
hydroxide, dimitic or monomitic hyphal system (Polyponis gilviis, P. licnoides, Fomes igniariiis, F. robiniae, etc.).
|
Cis |
cayensis |
Cer. |
pecki |
|
Cis |
maritimus |
Cer. |
pulluhis |
|
Cis |
niedhaiiki |
Cer. |
punctulatus |
|
Cer |
. magister |
Cer. |
singularis |
|
Cer |
. obrieni |
Mai. |
brevicolUs |
Ganoderma applanatum group. Diverse fungi \\'ith dark or light tissues and dimitic or trimitic hyphal systems {Ganoderma applanatum, G. lucidum, Fomes fomen- tariiis, F. pinicola, Polyponis betulinus, P. adustus, P. hydnoides, etc.).
Cis americamis Cis an gust us Cis biiirmatus Cis castle i Cis creberrimus Cis ephippiatiis Cis hirsutus
Dol. indistinctus Dol. manitoba Cer. californicus Cer. curtus Cer. multipunctatus Cer. nigropunctatus Cer. sallei
436 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
Cis levettei Cis megastictus Cis tridentattis
Cer. schaefferi Cer. similis Rhop. americanus
Some Ciidae are restricted to fungi not included in the above groups. Plesiocis cribnim, for instance, is almost always found in the fruiting bodies of Poltjporns volvatus, while Cis stcreophilus prefers those of thelephoraceous fungi in the genus Stereum. Within the above groups (espe- cially the last) the preferences of individual species vary considerably and some (Cis americanus) have a much wider host range than others {Cis ephippiatus, Ceracis sal- lei). Further details on host specificity will be included in species discussions.
Key to the Subfamilies and Genera of North American Ciidae
1. Procoxae subconical, strongly projecting below intercoxal process, which does not extend to middle of coxae (Figs. 16 and 25); metasternal suture absent (Fig. 33); first visible abdominal stemite (III) in male with posteriorly projecting, triangular flap, which partly conceals pubescent fovea (Fig. 31) _... Orophiinae 2
— Procoxae transverse or globular, not pro- jecting below intercoxal process, which extends beyond middle of coxae (Figs. 17, 22-24, 26-28); metasternal suture present (Fig. 34); first visible abdominal stemite in male simple or foveate, but witliout tri- angular flap (Fig. 32) Ciinae 3
2. Outer edges of all tibiae spinose for more than one-third of their lengtlis (Fig. 60); antennae 8-segmented; body form oval; vestiture consisting of minute hairs, which are not visible under 10 X magnification, and a few scattered long, fine hairs; pronotal punctation finer and sparser, the punctures much smaller than those on elytra and separated by 1.5 diameters or more Octotemmis (p. 508)
— Outer edges of tibiae with spines at apex only (Fig. 59); antennae 10-segmented; body form cylindrical; vestiture consisting of long, fine hairs; pronotal punctation coarser and denser, the punctures slighdy smaller than those on elytra and usually
separated by less than 1.0 diameter
Rhopalodontus (p. 506)
3. Prosternum in front of coxae almost twice as long as intercoxal process (Fig. 28);
sides of pronotmn subparallel or slightly diverging towards apex (Figs. 20-21, 28); body small, elongate, and extremely flat- tened, TL usually less tlian 1.4 mm.; EL/ EW greater than 1.60, and GD/EW usually less than 0.65; antennae 9-segmented with a 2- or 3-segmented club; elytral punctation dual and distincdy seriate, niicropunctures
bearing short, fine hairs or bristles
Hadraule (p. 491)
— Prosternum in front of coxae not or slightly longer than intercoxal process; witliout other characters in combination 4
4. Outer apical angle of protil)ia expanded, rounded, and bearing several spines (Figs. 57-58 ) 5
— Outer apical angle of protibia not as above, usually produced and dentate (Figs. 49-51, 54-56), blunt and angulate (Figs. 52-53), or narrowly rounded (Figs. 45-47); if ex- panded and rounded, then not spinose (Fig. 48) 8
5. Prosternum in front of coxae carinate and on same plane as intercoxal process (Figs. 13-14); lateral edges of pronotum visible for their entire lengths from above; anterior pronotal angles slightly produced (Fig. 14); elytral suture with inflexed margin near apex (Fig. 38); antennae 10-seg- mented Strigocis (p. 488)
— Prosternum in front of coxae concave or liiconcave and on different plane than intercoxal process (Figs. 9-10, 15); lateral edges of pronotum not visible for their entire lengths from above; anterior pronotal angles rounded or obtusely angulate, not produced (Fig. 15); elytral suture without inflexed margin near apex 6
6. Intercoxal process of prosternum laminate, less than 0.15 X as wide as a procoxal cavity (Fig. 24); vestiture consisting ot very short, fine hairs; pronotum of male usually bearing tubercles, horns, or laminae
at apex; antennae 8- to 10-segmented
Ceracis (p. 494)
— Intercoxal process of prosternum not lami- nate, at least 0.20 X as wide as a procoxal cavity (Fig. 27); vestiture consisting of short, stout bristles; pronotum of male al- ways simple 7
7. Pronotum very short and broad, PL/PW less than 0.73; prosternum in front of coxae only half as long as intercoxal process ( Fig. 27); metasternum at midline less than 0.40 X as long as broad; antennae 10-segmented
in North American species
— Malacocis (p. 504)
— Pronotmn more elongate, PL/PW more than 0.73; prosternum in front of coxae not or only slightly shorter than intercoxal proc-
North American Ciidae • Lawrence 437
ess; metasternum at midline more than 0.40 X as long as broad; antennae 9- or 10-segniented Sulcacis (p. 502)
8. Outer apical angle of protibia narrowly rounded (Fig. 45); elytral suture with an inflexed margin near apex (Fig. 38); head and pronotum in both sexes without tuber- cles or horns, male sometimes with densely pubescent area on clypeus (Fig. 37); elytral punctation single and uniform; vesti- ture consisting of \ery short, fine hairs; body elongate and parallel-sided, antennae 9- or 10-segmented Orthocis (p. 484)
— Outer apical angle of protibia usually pro- duced and dentate or blunt and angulate, if somewhat rounded, tlien vestiture con- sisting of short, stout bristles or elytral punctation dual and head of male bearing horns or tubercles; elytral suture without inflexed margin 9
9. Antennae 10-segmented Cis (p. 437)
— Antennae 9-segmented 10
10. Outer apical angle of protibia rounded (Fig. 46); body more elongate, EL/EW usually more than 1.50; apex of pronotum simple in both sexes Dolichocis ( p. 482 )
— Outer apical angle of protibia produced and dentate; body shorter and broader, EL/EW usually less than 1.50; apex of pronotum in male produced and emarginate forming
2 horns or tubercles (Figs. 4 and 8) 11
11. Intercoxal process of prostemum less than 0.25 X as wide as a procoxal cavity; meta- sternal suture less than 0.25 X as long as median length of metasternum; frontoclyp- eal ridge in male bearing 4 sharp teetli (Fig. 4) Plesiocis (p. 493)
— Intercoxal process of prostemum more tlian 0.25 X as wide as a procoxal cavity; meta- sternal suture more tlian 0.25 X as long as median length of metasternum; frontoclyp- eal ridge in male bearing 2 subtriangular plates with a distinct notch between them (Fig. 8) Ennearthron (p. 480)
Subfamily Ciinae
Cisidae Leach, 1819: 206. Cissidae Wollaston, 1854: 279. Cioidae Gistel, 1856: 143. Ciidae Marseul, 1887: 293.
Included genera. All those not placed in the Orophiinae (p. 506). About 30 genera have been described, but a number of these will have to be synonymized and several remain to be described.
Genus Cis Latreille
Cis Latreille, 1796: 50; Latreille, 1802: 205; Gyllenhal, 1813: 377; Gyllenhal, 1827: 624; Redtenbacher, 1847: 348; Mellie, 1848: 236; Lacordaire, 1857: 551; Jacquelin du Val, 1861: 237; Thomson, 1863: 183; Abeille de Perrin, 1874b: 19; Kiesenwetter, 1877: 173; LeConte and Horn, 1883: 232; Casey, 1898: 78; Reitter, 1902a: 47; Blatchley, 1910: 897; Dalla Torre, 1911: 5; Leng, 1920: 246; Arnett, 1962: 829. Type species, by subsequent monotypy, Der- mestes boleti ScopoH, 1763: 17 (Latreille, 1802: 205).
Eridaulus Thomson, 1863: 191; Lawrence, 1965: 282 (complete synonymy); Lawrence, 1967b: 98. Type species, by present designation, Anohium nitidum Fabricius, 1792: 238.
Xestocis Casey, 1898: 85; Lawrence, 1965: 282. Type species, by subsequent designation, Xesto- cis levettei Casey, 1898: 85 (Lawrence, 1965: 282).
Macrocis Reitter, 187Sc: 34; Gorham, 1883: 219. Type species, by present designation, Macrocis taunts Reitter, 1878c: 34. NEW SYNONYMY.
Included species and species groups. This genus contains about 350 named species, which will not be listed here, and numerous undescribed forms as well. Some distantly related forms with 10-segmented antennae have been described as Cis, but even when these are eventually removed, the genus will be by far the largest in the family. Within the genus Cis there are a number of well-defined subgroups of varying size, the species of which usually have similar body form, male genitalia, secondary sexual characters, larval urogomphi, and food preferences. Several of the more obvious species groups are Hsted below, witli the distribution, a few representative species, and all of the North American members. Of the 24 species groups listed, 6 are en- tirely exotic and the other 18 contain 34 of the 43 North American Cis.
C. hilamellatus group. Oriental-Austra- lian. C. hilamellatus Wood, C. australis Blackburn, C. clarki Blair. North America: none.
C. holeti group. Holarctic. C. holeti (Scopoli), C. rugulosus Mellie, C. micans (Fabricius), C. villosulus (Marsham). North America: C. pistoria Casey.
438 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
C. cayensis group. Neotropical. North America: C. cayensis, n. sp., C. nicdhauki, n. sp.
C. compressicornis group. Australian. C compressicornis Fairmaire, C. cerviis Blair. North America: none.
C. comptiis group. Holarctic. C. comp- tus Gyllenhal, C. striatuhis Mellie. North America: C. striolotiis Casey, C. versicolor Casey?
C. creherrimus group. Nearctic-Neo- tropical. North America: C. creherrimus MelHe.
C. fagi group. Holarctic. C. fagi Waltl, C. castaneus MelHe. North America: C. angustiis Hatch, C. rohiniophilus, n. sp.
C. festivus group. Holarctic. C. festivus (Panzer), C. pygmaeus (Marsham), C. vestitus Mellie. North America: C. festivu- lus, n, sp.
C ftiscipes group. Holarctic-Oriental. C. seriatopilosiis Motschulsky, C. seriatiilus Kiesenwetter, C. taiwamis Chujo. North America: C. ftiscipes Mellie.
C. huachiicae group. Nearctic-Neotropi- cal. North America: C. discolor, n. sp., C. hiiachticae Dury.
C. kraiisi group. Neotropical. C. afro- mactdatiis Pic, C. supcrhus Kraus. North America: C. krausi Dalla Torre.
C. litterattis group. Oriental-Australian. C. insignis Scott, C. litteratus Fauvel. North America: none.
C. melliei group. Neotropical. C. melliei Coquerel. North America: C. crinitus, n. sp., C. hirsutus Casey, C rotiindidus, n. sp., C. ursidimis Casey.
C. nitidtis group (= Eridauhis). Hol- arctic. C. jacquemarti Mellie, C. lineato- crihratus Mellie, C. nitidus (Fabricius). North America: C. americanus Manner- heim?, C. hiarmatus Mannerheim, C. ephip- piatiis Mannerheim, C. levettei (Casey), C maritimus (Hatch), C. megastictus, n. sp., C. tridentatus Nhmnerheim?
C. pacificus group. Oriental-Australian. C. agariconae Zimmerman, C. nuirquesaniis Blair, C. pacificus Sharp. North America: none.
C. pallidus group. Nearctic-Neotropical. C. corticimis Gorham, C. pallidus Mellie. North America: C. tetracentrum Gorham.
C. pilosus group. Nearctic-Neotropical. C. pilosus Gorham. North America: C. cornuttis Blatchley.
C. punctulatus group. Holarctic. C. punctulatiis Gyllenhal, C. tomentosus Mel- lie. North America: C. horrididus Casey, C. hystricidiis Casey.
C. setarius group (including Apterocis?). Australian (Hawaii). C. hicolor Sharp, C chloroticus Sharp, C. setarius Shaq?, C. tahidus Sharp. North America: none.
C. signatus group. Australian (Hawaii). C. kauaiensis Perkins, C. nigrofasciatus Blackburn, C. roridus Sharp, C signatus Sharp. North America: none.
C. suhtilis group. Nearctic-Neotropical. North America: C. acritus, n. sp., C. subtilis Mellie.
C. taurus group (= Macrocis). Nearctic- Neotropical. C. Z?i5on (Reitter), C. di- aholicus (Reitter), C. grandicornis (Pic), C. setifer (Gorham), C. taurus (Reitter), C. testaceus (Pic). North America: C. cornelli, n. sp.
C. tricornis group. Nearctic-Neotropical. C. delicatulus (Jacquelin du Val), C. fri- i cornis (Gorham). North America: C. ;?ji/e5' (Casey).
C. vitulus group. Nearctic-Neotropical. C. huhalus Reitter, C. hishidens Gorham, C. fasciatus Gorham, C. granarius Mellie. North America: C. congestus Casey, C. vitulus Mannerheim.
Members of this genus may be distin- guished from other North American Ciidae by the 10-segmented antennae, rounded or angulate to dentate protibial apex without spines at the outer angle, flat to carinate prosternum without a laminate intercoxal process, angulate or produced anterior pronotal angles, simple elytral suture, andi sexual modifications almost always present' on the head and/or pronotum of the male. Species of Strigocis, Sidcacis, Malacocls, and Ceracis may have 10-segmented anten- nae, but always have spinose protibial
North American Ciidae • Lawrence 439
apices. Species of Strigocis and Orthocis have an inflexed margin at the apex of the elytral suture, and members of the latter group never have horns or tubercles on the head or pronotum of the male. Species of DoUchocis, Ennearthion, and PJesiocis re- semble Cis in several respects and have rounded or dentate protibial apices, but the antennae are always 9-segmented.
In an earlier paper (Lawrence, 1965), I treated Thomson's Eridaulus as a full genus, characterized by the carinate pro- sternum, oval body form, dual elytral punctation, and triangular plates on the head of the male. Further study led me to abandon this concept (Lawrence, 1967b: 98) and to place the included species in at least two different species groups (Cis nitidus group and C. pacificus group). Macrocis Gorham is another genus whose species are characterized by having 10- segmented antennae, carinate prostemum, dual elytral punctation, and two plates on the head of the male. In this case, however, the body is very short and stout, and the frontoclypeal plates are produced to form long, narrow homs. I can see little reason for maintaining a separate genus for this group, although it contains a large number of Neotropical species. The prostcmal structure is not uncommon in the genus Cis, and the long, lateral horns on the head of the male may be found in members of the C. compressicornis group and the C. melliei group.
Key to the North American Species of Cis
1. Vestiture of elytra distinctly dual, consist- ing of 2 classes of liairs or bristles, which differ in length, thickness, color, form, or angle of inclination (Figs. 29-30) 2
— Vestiture of elytra not dual, tlie hairs or bristles varying slightly in length, but not falling into 2 distinct classes 8
2. Vestiture of elytra consisting of short, erect bristles, which are seriate, and very fine, inclined hairs, which are not visible under 10 X magnification; head of male with 2
tubercles on vertex (Fig. 5)
C. caijcnsis {see 22; p. 448)
— Both classes of hairs or bristles easily visible under 10 X magnification 3
3. Body shorter and broader, EL/EW usually less than 1.50 mm; if slightly more, then male with 2 subtriangular frontoclypeal plates; lateral margins of pronotum visible for their entire lengths from above, the anterior angles produced forward 4
— Body longer and narrower, EL/EW more than 1.50 mm; male with 4 frontoclypeal teeth or tubercles; lateral margins of pro- notum not visible for their entire lengths from above, the anterior angles not pro- duced 7
4. Erect bristles onlv slightly longer than in- clined ones; abdominal fovea in male absent or located on posterior part of ster- nite III 5
— Erect bristles at least 1.50 X as long as inclined bristles or hairs; abdominal fovea in male located in center of sternite III . 6
5. Size smaller, TL usually less tlian 1.75 mm; elytral punctation subseriate; male with 2 subtriangular, frontoclypeal plates and an
abdominal fovea
C. floridae {sec 36; p. 460)
— Size larger, TL usually more than 1.75 mm; elytral punctation confused; male with 4 frontoclypeal teeth and no abdominal fovea C. huachucae {see 36; p. 462)
6. Vestiture consisting of longer and shorter, fine, yellow bristles, which are subseriate on the elytra; EL/EW less than 1.33; fully
pigmented adult black ___
C. cormitus (p. 451)
— Vestiture consisting of longer, stiff and erect, dark bristles and shorter, inclined, pale hairs, which are uniformly distributed on elytra (Fig. 29); EL/EW more than 1.33; fully pigmented adult reddish brown C. crinitus (p. 453)
7. Body flattened, GD/EW usually less than
0.73; male with abdominal fovea
C. cieheninms {see 31, 46; p. 452)
— Body subcylindrical, GD/EW usually more than 0.73; male without abdominal fovea - C. JioniduJiis (p. 462)
8. Elytral punctation dual, consisting of larger, shallow megapunctures and smaller, deeper micropunctures, which bear bristles or hairs 9
— Elytral punctation single, the punctures fairly uniform in size and all or most of them bearing hairs or bristles 29
9. Vestiture consisting of fine hairs 10
— Vestitine consisting of short, stout bristles 17
10. Elytral hairs longer, more than 0.20 X as long as scutellar base and visible under lOX magnification, decumbent —..11
440
Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
— Elytral hairs very short, less than 0.15 X as long as scutellar base and not visible under 10 X magnification, erect or in- clined 12
11. Pronotal punctation finer and sparser, punctures subequal to eye facets and separated by 1.0 to 1.5 diameters; body usually smaller, narrower, and bicolored, brownish with black pronotum and trans- verse elytral macula
C ephippiatus (p. 458)
— Pronotal punctation coarser and denser, punctin-es larger than eye facets and separated by 0.33 to 0.66 diameter; body usually larger, broader, and more unifonnly
pigmented, brownish
C hiarmatus (p. 446)
12. Body more elongate, EL/EW more than 1.40; prosternum in front of coxae slightly timiid but not carinate (Figs. 11-12) 13
— Body shorter and broader, EL/EW less dian 1.40; prosternum in front of coxae carinate (Fig. 13) 14
13. Elytral punctation distinctly seriate; sides of pronotmn subparallel; pronotal pimcta- tion coarser and denser, the punctures usually separated by less than 0.50 diam- eter; head of the male with 2 widely
spaced, frontoclypeal teeth
C. dunedinensis (p. 457)
— Elytral punctation not seriate; sides of pronotum weakly rounded; pronotal punc- tation finer and sparser, the punctures usu- ally separated by more than 0.50 diameter; head of male with 4 frontoclypeal teeth ( the outer 2 sometimes obsolete ) and 2
horns on the vertex (Fig. 5)
- C. niedhauki (p. 467)
14. Pronotal punctures about as large as elytral megapunctures and usually separated by more than 0.75 diameter; TL usually less than 1.40 mm; male with 2 distant, lateral horns on pronotal apex and a single, median, forked horn on frontoclypeal ridge (Fig. 6) C. miles (p. 467)
— Pronotal punctures, if separated by more than 0.75 diameter, much smaller than elytral megapunctures; TL usually more than 1.40 mm; pronotal apex in male simple or bituberculate and frontoclypeal ridge bearing 2 tiiangular plates ( Fig. 39 ) 15
15. Outer edge of protibia irregularly notched or serrate for part of its length (Fig. 54); pronotal punctation finer and sparser, punc- tures smaller than eye facets and separated by more than 1 diameter; lateral edges of pronotimi smooth; elytral punctation not distinctly .seriate C. levettei (p. 464)
— Outer edge of protibia simple; pronotal
punctation coarser and denser, punctures usually larger than eye facets and separated by less than 1 diameter; lateral edges of pronotum coarsely crenulate; elytral punc- tation distinctly seriate — 16
16. Anterior angles of pronotum broadly rounded; pronotal pimctures \ery dense, separated by 0.33 diameter or less, inter- spaces smooth; elytral megapunctures 4 X as large as micropunctures; abdominal fovea in male circular C mcgastictus (p. 465)
— Anterior angles of pronotum subacute; pronotal punctures not as dense, separated by 0.50 to 0.66 diameter, interspaces granu- late; elytral megapunctures less than 4 X as large as micropunctures; abdominal
fovea in male longitudinally oval
C. maritimus (p. 465)
17. Elytra short and broad, EL/EW less than 1.28, apices broadly rounded; elytral punctation distinctly seriate; prosternum strongly carinate; TL less than 1.75 mm; frontoclypeal ridge in male with 2 long
and narrow, lateral horns (Fig. 42) i
C. cornelli (p. 450)
— Elytra longer and narrower, EL/EW more than 1.28, or if slightly less, apices narrowly rounded; male never with long, narrow horns; without other characters in combi- nation 18
18. Anterior angles of pronotum distinctly pro- duced forward and broadly rounded ( Fig. 43); pronotum shorter and broader, PL/ PW usually less than 0.80, the disc im- pressed anteriorly in male; lateral margins of pronotum broader, easily visible for their entire lengtlis from above; prosternum tumid but not carinate; size larger, TL usually more than 2.00 mm 19
— Anterior angles of pronotmn not or barely produced forward and angulate; pronotum longer and narrower, PL PW usually more than 0.80, the disc not impressed anteriorly in male; lateral margins of pronotimi nar- rower or prosternum carinate; size smaller, TL usually less than 2.00 mm 20
19. Elytral punctation seriate; elytral bristles longer and narrower, more tlian 0.25 X as long as scutellar base and more than 3 X as long as wide; latei'al edges of pronotum weakly and finely crenulate; pronotal and
elytral discs fairly even —
._.._ .- C. fuscipes (p. 460)
— Elytral punctation confused; elytral l^ristles shorter and broader, less than 0.20 X as long as scutellar liase and less than 3 X as long as wide; lateral edges of pronotum strongly and coai"sely crenulate; pronotal
North American Ciidae • Lawrence 441
and elytral discs irregularly impressed
C. pistoria (p. 469)
Prosternum distinctly carinate; elytra shorter and somewhat ovate, EL/EW usually less tlian 1.42; apex of pronotum in male produced and emarginate, forming a subtriangular process or two approximate
tubercles 21
Prosternum flat or slightly tumid, not cari- nate; elytra more elongate and parallel- sided, EL/EW usually more than 1.42;
apex of pronotum in male simple 22
Elytral punctation obscurely dual, mega- punctures barely larger than micropunc-
tures C. tridentatus (p. 477)
Elytral pimctation distinctly dual, mega- punctures 1.50 to 3.0 X as large as micro- punctures C. americanus (p. 444)
Elytral punctation distinctly seriate 23
Elytral punctation not distinctly seriate ._..27 Outer apical angle of protibia not produced, roimded, or obtusely angulate; lateral mar- gins of pronotum broader, visible for their entire lengths from above; fully pigmented adults bicolored, widi pronotmn reddish
and elytra 1)lack or red and black 24
Outer apical angle of protibia at least slightly produced, usually forming a dis- tinct tooth; lateral margins of pronotum narrower, not or barely visible for their entire lengths from above; fidly pigmented adults unifonnly brownish or black in color
or with pronotum darker than elytra 25
Body larger and more elongate, TL usually more than 1.75 mm and EL/EW usually more than 1.45; elytral punctation finer and sparser; frontoclyioeal ridge in male
simple C. versicolor (p. 478)
Body smaller and broader, TL usually less than 1.75 mm and EL/EW usually less than 1.45; elytral punctation coarser and denser; frontoclypeal ridge in male with 2
weak tubercles C. siihfuscus (p. 476)
Elytral vestiture dual, consisting of short, stout bristles and very fine hairs, which are not visible under 10 X magnification; TL usually less than 1.40 mm; head of male with 4 frontoclypeal teeth and 2 tubercles
on vertex C. cayensis {see 2; p. 448)
Elytral vestiture not dual; TL usually more than 1.40 mm; male without tubercles on
vertex 26
Pronotal punctation sparser, punctures usu- ally separated by more than 0.50 diameter, interspaces coarsely grantilate and dull; elytral bristles yellowish; abdominal fovea in male about 0.20 X as long as body of
sternite III C. striolattis (p. 475)
Pronotal punctation denser, punctures usu- ally separated by less than 0.50 diameter,
interspaces finely granulate and shiny; elytral bristles colorless; abdominal fovea in male more than 0.25 X as long as body of sternite III C. tristis (p. 477)
27. Body more elongate, EL/EW more than 1.60; elytra yellowish with black maculae
C. krausi (p. 463)
— Body less elongate, EL/EW less than 1.60; elytra uniformly brownish or black 28
28. Frontoclypeal ridge in male simple; ab- dominal fovea in male present; southwest- ern U. S. C. acritus (p. 443)
— Frontoclypeal ridge in male with 2 tubercles; abdominal fovea absent; eastern U. S C. siibtilis (p. 476)
29. Vestiture consisting of very short, fine hairs, which are not or barely visible under 10 X magnification; outer apical angle of protibia blunt and rounded; body elongate, EL/EW more than 1.50; elytral suture with an in- flexed margin near apex (Fig. 38); head
and pronotum in male simple
Orthocis (p. 484)
— Vestiture consisting of long hairs or short, stout bristles 30
30. Vestiture of long, fine hairs 31
— Vestiture of shorter, stout bristles 34
31. Body flattened and elongate, EL/EW more than 1.50; frontoclypeal ridge in male with
4 tubercles
C. creherrimus {see 7, 46; p. 452)
— Body not flattened, shorter and broader, EL/EW less than 1.50; frontoclypeal ridge in male with 2 narrow, lateral horns (Fig. 41) 32
32. Prosternum in front of coxae carinate; metastemiun short and broad, the length at midline less than 0.40 X the width; sides of elytra strongly rounded, the apices acute C. rotundulus (p. 471)
— Prosternum in front of coxae strongly tumid, but not carinate; metasternum more elongate, die length at midhne more than 0.40 X the width; sides of elytra weakly rounded or sul^parallel, the apices blunt 33
33. Lateral edges of pronotum strongly crenu- late; size smaller, TL usually less than 1.60 mm; fully pigmented adults usually reddish in color; abdominal fovea in male less than
0.20 X as long as body of sternite III
C ursulintis (p. 478)
— Lateral edges of pronotum not or weakly crenulate; size larger, TL usually greater than 1.60 mm; fully pigmented adults usually blackish in color; al^dominal fovea in male more than 0.20 X as long as body of sternite III C. hirsutus (p. 461)
34. Lateral margins of pronotum broader, usu- ally visible for tlieir entire lengths from
442 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
35.
36.
37.
38.
39.
40.
above; anterior angles of pronotum dis-
tincdy produced and rounded 35
Lateral edges of pronotum narrower, usu- ally not visible for their entire lengths from above; anterior angles of pronotum not or
barely produced and subacute 41
Vestiture indistinctly dual, the bristles dif- fering in lengdi and angle of inclination ,.36
Vestiture single 37
Elytral punctation subseriate; size smaller, TL usually less than 1.75 mm; male with 2 subtiiangvdar frontoclypeal plates; ab- dominal fovea present
C. floridae (see 5; p. 460)
Elytral punctuation confused; size larger, TL usually more than 1.75 mm; male with 4 frontoclypeal teeth; abdominal fovea ab- sent C. hiiachucae (see 5; p. 462)
Prosternum in front of coxae carinate; size smaller, TL less than 2.20 mm; male with 2 pronotal horns, 4 frontoclypeal teeth, and
a pubescent fovea
C. quadridentatns (p. 469)
Presternum in front of coxae flat to strongly tumid, but not carinate; size larger, TL more than 2.20 mm, or if slightly less, male without above coml^ination of characters —38 Lateral margins of pronotimi without raised lip, edges strongly crenulate (Fig. 44); body more elongate, EL/EW more than 1.50; elytra witir a transverse impression at anterior fourth; apex of pronotum in male weakly emarginate, but never with horns _.
__ C. discolor (p. 455)
Lateral margins of pronotum with a nar- row, raised Hp, edges not or barely crenu- late (Fig. 43); body shorter and broader, EL/EW less than 1.50; elytra without transverse impression; apex of pronotum in
male with 2 triangular horns ...39
Pronotal punctures only slightly smaller than elytral punctures; elytral bristles longer and finer, more than 6 X as long as wide, acute at apex; frontoclypeal ridge in male with 2 triangular plates; male widi abdominal fovea; elytra in female more
than 1.90 X as long as pronotimi
C. tetracentrum (p. 476)
Pronotal punctures much smaller than ely- tral punctures; elytral bristles shorter and stouter, less tlian 6 X as long as wide, blunt at apex; frontoclypeal ridge in male ele- vated and trisinuate, so that 4 teeth are formed (Fig. 3); male widiout abdominal fovea; elytra in female less dian 1.90 X as
long as pronotum 40
Pronotal punctation coarser and denser, the punctures usually separated by less than 0.75 diameter, interspaces smooth and
shiny; antennal segment III less than 1.50 X as long as IV; body somewhat smaller, shorter and stouter, TL usually less than 2.45 mm; EL/EW usually less than 1.31, and TL/EW in female usually less than
2.06; southeastern U. S.
C. congcsfus (p. 449)
— Pronotal punctation finer and sparser, the punctures usually separated by more than 0.75 diameter, the interspaces granulate and dull; antennal segment III more than 1.50 X as long as IV; body somewhat larger and more elongate, TL usually more tlian 2.45 mm, EL/EW usually more than 1.31, and TL/EW in female usually more than 2.06; California and northern Arizona C. vituhis (p. 479)
41. Body shorter and broader, EL/EW less than 1.40, or if slightly more, apex of pronotum in male with 2 triangular horns; elytral punctation confused ._ .42
— Body more elongate, EL/EW more than 1.40, or if slightly less, elytral bristles sub- seriate; apex of pronotum in male simple, emarginate, or with 2 small tubercles .44
42. Base of scutellum less than 0.10 X as wide as pronotum; body somewhat shorter and broader, EL/EW usually less than 1.30; prosternum in front of coxae slightly timiid but not carinate; apex of pronotum in male simple, frontoclypeal ridge elevated and weakly trisinuate C. latniuatus (p. 464)
— Base of scutellum more than 0.10 X as wide as pronotum; body somewhat more elongate, EL/EW usually more than 1.30; prosternum in front of coxae weakly carinate; apex of pronotum and frontoclyp- eal ridge in male each bearing 2 triangular plates or horns (Fig. 7) 43
43. Lateral edges of pronotum distinctly crenu- late; pronotal punctation coarser and denser, the punctiu-es usually separated by less than 0.75 diameter; pronotal surface granulate and dull; TL usually more than 1.40 mm C. duplex (p. 457)
— Lateral edges of pronotum not crenulate; pronotal punctation finer and sparser, the punctures usually separated by more than 0.75 diameter; pronotal surface smooth and
shiny; TL usually less than 1.40 mm
C. castlei (p. 447)
44. Elytral bristles shorter and broader, usually less than 2.5 X as long as wide, subseriate, blunt at apex; elytra usually bicolored,
darker anteriorly than posteriorly
C. stereophilus (p. 472)
— Elytral brisdes longer and narrower, usually more than 2.5 X as long as wide, confused or acute at apex; elytra concolorous 45
North American Ciidae • Lawrence 443
45. Outer apical angle of protibia roimdecl (Fig. 47); elytial bristles vaiying in length, blunt, and yellowish __„ C. festivulus (p. 459)
— Outer apical angle of protibia produced, dentate or angulate; elytral bristles more unifonn in lengtli and acute or colorless 46
46. Body somewhat flattened, GD/EW usually less than 0.73; lateral edges of pronotum not or barely crenulate; elytral bristles
subseriate and acute at apex
C. crchcrrimus {see 7, 31; p. 452)
— Body not flattened, GD/EW usually more than 0.73; lateral edges of pronotum dis- tinctly crenulate; elytral bristles confused and blunt at apex 47
47. Elytral bristles shorter and broader, usually less tlian 3.5 X as long as wide; elytral punctation finer and sparser, the punctures usually less tlian 0.30 X as wide as scutellar base and separated by 0.50 diameter or more C. robiniophilus (p. 470)
— Elytral bristles longer and narrower, usually more than 3.5 X as long as wide; elytral punctation coarser and denser, tire punc- tures usually more than 0.30 X as wide as scutellar base and separated by 0.33 diam- eter or less 48
48. Elytral bristles colorless; apex of pronotum in male usually bituberculate; male without abdominal fovea C. hystiiciihts (p. 463)
— Elytral Ijristles bright yellow in color; apex of pronotum in male simple; male with pubescent fovea on abdominal sternite III C. angustus (p. 446)
Cis acrifus NEW SPECIES
Holofype. S, ARIZONA: Rustler Park, 8 mi. W Portal, Cochise Co., Aug. 3, 1961, Lot 892 J. F. Lawrence, ex Polyporiis abietimis on conifer [CAS]. Allotype, ?, same data, [JFL].
Male. Length 1.67 mm. Body 2.23 X as long as broad, slightly convex. Head and pronotum reddish orange, elytra dark yel- lowish browai. Vestiture consisting of short, stout, blunt, colorless brisdes. Vertex flat- tened \\'ith a slight median impression; frontoclypeal ridge simple. Antennal seg- ment III 1.40 X as long as IV. Pronotum 0.85 X as long as broad, widest at middle; anterior edge moderately rounded, simple; sides weakly rounded, the margins narrow and distinctly crenulate, not visible for their entire lengths from above; anterior
angles not produced fonvard, almost right; disc weakly convex, even; surface distinctly granulate, slightly shiny; punctures 0.20 X as large as scutellar base and separated by 0.25 to 0.75 diameter. Elytra 1.50 X as long as broad and 2.04 X as long as pronotum; sides subparallel, apices blunt; punctation dual and confused; megapunctures some- what coarser and denser than those on pronotum, shallow and nude; each micro- puncture bearing a stout, blunt, colorless bristle, which is about 0.33 X as long as scutellar base. Prosternum slightly tumid; intercoxal process 0.30 X as wide as a pro- coxal cavity, parallel-sided. Protibial apex slightly produced, outer apical angle almost right (Fig. 52). Metastemum 0.54 X as long as wide; suture 0.40 X as long as median length of sternum. Abdominal stemite III bearing a median, oval, pubes- cent fovea, wliich is almost 2.0 X as long as wide, 0.30 X as long as body of sternite, distinctly margined, and located anterad of center. Sternite VIII as in Figure 63. Aedeagus as in Figures 73 and 80.
Female. Length 1.75 mm. Body 2.33 X as long as broad. Vertex and frontoclypeal ridge as in male. Pronotum 0.85 X as long as broad; anterior edge as in male. Elytra 1.57 X as long as broad and 2.04 X as long as pronotum. Protibial apex as in male. Sternite III without pubescent fovea.
Variation. Pronotum yellowish to dark reddish brown, usually reddish orange or reddish brown. Elytra yellowish to brown- ish black, usually brownish. Pronotum usu- ally lighter in color than elytra. Size and dimensions vary as follows in a mixed series of seven males and nine females from Cali- fornia, Arizona, and New Mexico:
TL mm: i 1.55-1.80 ( 1.70 ± 0.031),
2 1.47-1.82 (1.66 + 0.040); TL/EW: S 2.20-2.38 (2.27 ± 0.022),
5 2.18-2.44 (2.33 ±0.030); PL/PW: c5 0.82-0.88 (0.85 ± 0.007),
9 0.79-0.95 (0.88 ±0.017); EL/EW: 6 1.47-1.61 (1.53 ± 0.017),
9 1.48-1.67 (1.58 ±0.021);
444 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
EL/PL: S 1.96-2.13 ( 2.05 ± 0.024 ) , 9 2.00-2.14 (2.09 ±0.017);
GD/EW: S 0.73-0.81 (0.76 ± 0.010), 9 0.73-0.81 (0.77 ±0.009).
Paratypes. ARIZONA: 1, East Turkey Creek, 6 mi. W Portal, Cochise Co., July 24, 1963, Lot 1240 JFL (A. Raske, colk), ex Polyponis abietimis [JFL]; 4, Hitch- cock Hwy. Mi. 27, VI-22-1957 [MCZ, UAZ]; 2, Madera Canyon, Santa Rita Mts., Santa Cruz Co., Jan. 30, 1964, Lot 1285 JFL, ex Polyponis abietimis [JFL]; 1, 6 mi. SW Portal, Cochise Co., Aug. 2, 1961, Lot 879 JFL, ex Polyponis ahieiinus [JFL]; 3, Rustler Park, 8 mi. W Portal, Cochise Co., Aug. 3, 1961, Lot 892 JFL, ex Polyponis abietimis [JFL]; 1, Southwest Research Station, 5 mi. SW Portal, Cochise Co., Aug. 1, 1961, Lot 882 JFL, ex Polyponis abietimis [JFL]. CALIFORNIA: 1, 1 mi. SE Alta Sierra, Kern Co., July 5, 1962, Lot 1043 JFL, ex Polyponis abietimis [JFL]; 1, 3 mi. E Shaver Lake, Fresno Co., Sept. 1, 1960, Lot 685 JFL, ex Polyponis abietimis [JFL]. NEW MEXICO: 2, Las Vegas H[ot] S[prings], 5.8, Barber & Schwarz Coll. [USNM]; 1, same locality, 9.8, Barber & Schwarz Coll. [USNM]; 1, same locaHty, 3.8, Barber & Schwarz Coll. [USNM].
Distribution. Montane regions of south- em California, Arizona, and New Mexico (Fig. 106).
Host fung,i. Polyponis abietimis [7(1)].
Discussion. Tliis species may be dis- tinguished from most other North American Cis by the elongate body form, dual and confused elytral punctation, and vestiture of short, blunt, colorless bristles. It differs from C. siibtilis in having a pubescent fovea on abdominal sternite III of the male and the frontoclypeal ridge simple in both sexes. The pronotum in C. acritus is usually more reddish in color and more coarsely punctate than in C. siibtilis. Of those spe- cies occurring within the range of C. ac- ritus, C. versicolor and C. striolatiis have similar body form, dual elytral punctation, and short bristles; in both of these, how-
ever, the elytral punctation is distinctly seriate. C. discolor has similar pale, blunt bristles, but the size is usually larger, the punctation is single, the pronotal margins are broader, and the protibial apex is dis- tinctly dentate (Fig. 56). Smaller speci- mens of C. ftiscipes may be confused with C. acritus, but they differ in the broader body form and the produced and rounded anterior pronotal angles (Fig. 23).
The ranges of C. acritus and C. siibtilis are allopatric and the two species are closely related, even sharing the rmique type of aedeagus (Figs. 73 and 80). They are considered to be distinct species in this treatment mainly because of the lack of an abdominal fovea in the male of the eastern form. The two species do not appear to belong to any Palaearctic group and their closest relatives are to be sought among the diverse and little known Neotropical fauna.
Cis acritus is found in association with conifers at intermediate and high elevations in the mountains of the Southwest. It has been collected only on Polyporus abietimis, where it occurs with Cis hystriculus in Cali- fornia and with C. horridulus and Ceracis poivelli in Arizona.
The name acritus is derived from the Greek akritos, meaning confused, dis- orderly, mixed (referring to the confused elytral punctation ) .
Cis americanus Mannerheim
Cis americanus Mannerheim, 1852: 360; Pielou and Verma, 1968: 1184 (dist., biol.). Type locality: "Insulae Sitkhae." Lectotype, $ , Man- nerheim Coll., MZUH.
Xestocis insolens Casey, 1898: 86. Type locality: "Pennsylvania." Holotvpe, S , Casey Coll.,
usnm'. new synonymy.
Cis frosti Dury, 1917: 9; Brown, 1929: 153 (dist.); Frost, 1930: 41 (biol.). Type locality: "Orono, Maine." Holotype, $ , Durv Coll., CIN (type lost). NEW SYNONYMY.
Cis serricollis Dmy, 1917: 9; Weiss and West, 1921a: 61 (dist., biol.). Type locality: "Linn Co., Oregon." Svntypes, $ 9 , Dury Coll., CIN. NEW SYNONYMY.
Xestocis minor Hatch, 1962: 232. Type locality:
North American Ciidae • Lawrence 445
"Wheatland [Yamhill Co.], Ore." Holotype, $,
Hatch Coll., UW. NEW SYNONYMY. Xestocis oweni Hatch, 1962: 232. Type locahty:
"Seattle, Wash." Holotype, <5 , Hatch Coll.,
UW. NEW SYNONYMY. Xestocis strigulosus Hatch, 1962: 231. Type lo- cality: "Seattle, Wash." Holotvpe, £ , Hatch
Coll., UW. NEW synonymy! Cis hatchi, NEW NAME for Xestocis nitidus
Hatch, 1962 (not Anobium nitidum Fabricius,
1792). NEW SYNONYMY. Xestocis nitidus Hatch, 1962: 232. Type locality:
"Wheatland [Yamhill Co.], Ore." Holotype, S,
Hatch CoU., UW.
Distribution. Widespread across the northern part of North America from Alaslca to Nova Scotia, south in California to Monterey County on the coast and Fresno County in the Sierra Nevada, in the Rocky Mountains south to northern Utah and Colorado, and in the Appala- chian chain as far south as western North Carolina (Fig. 93). Marginal records: ALASKA: Kenai. BRITISH COLUMBIA: Terrace. ALBERTA: Cypress Hills. WIS- CONSIN: Bayfield Co. QUEBEC: Dupar- quet. MAINE: Orono, Washington Co. NEW BRUNSWICK: NE of Ludlow. NOVA SCOTIA: Portaupique. NORTH CAROLINA: Highlands, Macon Co. SOUTH DAKOTA: Hill City, Custer Co. COLORADO: Steamboat Springs, Routt Co. UTAH: Provo, Utah Co. NEVADA: 8 mi. SE Lamoille, Elko Co. CALI- FORNIA: Huntington Lake, Fresno Co.; Big Sur, Monterey Co.
Host fungi. Polyporus achistws [7(5)]; Pohjporus betuJimis [7(4)]; Stereum hirsu- tiim [6(4)]; Fomes pinicola [5(2)]; Poly- porus versicolor [5(1)]; Fomes annosus [4(1)]; Poria versipora [3(2)]; Steccheri- num ochraceum [3(2)]; Pleurotus ostreatus [3(1)]; Polyporus sulphiireus [3(1)]; Gan- oderma tsugae [3(1)]; Polyporus abiefinus [3]; Trametes mollis [2(2)]; Ganoderma applanatum [2(1)]; Polyporus resinosus [1(1)]; Polyporus hirsiitus [1(1)]; Poly- porus pargamenus [1(1)]; Polyporus squa- mosus [1(1)]; Polyporus tidipifcrae [1]; Poria nigresccns [1]; Fomes fomentarius
[1]; Polyporus biformis [1]; Phlebia meris- moidcs [1].
Discussion. This species is characterized by the relatively short and stout body form, strongly carinate prostemum, dual and con- fused or indistinctly seriate elytral punc- tation, and vestiture of short bristles. Several other sympatric forms resemble C. americanus in general body fonn, carinate prostemum, dual punctation, and second- ary sexual characters (Fig. 39); of these, C. levettei, C. megastictus, and C. mari- timus are clothed with very short, fine hairs, while C. biarmatus and C. ephippi- atus have longer, decumbent, fine hairs. Cis castlei and C. duplex are similar in body fonn and vestiture, but both differ from C. americanus by having the pro- sternum only weakly carinate and the elytral punctation single. Cis tridcntatus is very closely related to C. americanus and the two are often difficult to separate; in the latter the elytral punctation is more distinctly dual, the megapunctures being 1.5 to 3.0 X as large as the micropunctures.
This is an extremely variable, wide- spread, and polyphagous species, and it may represent a species complex. Manner- heim ( 1852 ) first described it on the basis of material from Sitka, Alaska, but the name was forgotten and Casey (1898) gave the name Xestocis insolens to specimens from Pennsylvania. Dury (1917) described Cis frosti from Maine and C. serricoUis from Oregon; the types of both of these, as well as that of insolens, fall within the range of variation observed in western populations of C. americanus. In 1962, Hatch attempted to unravel this complex in the Pacific Northwest and described four more species, which I have synonymizcd above. One of these, C. oweni, may repre- sent a distinct species (see below).
Cis americanus occurs across the north- ern part of the continent and is particularly common in the Pacific Northwest. Here t\\o fomis can be recognized that may proxe to be different species. In the first, the surface of the pronotum and elytra are
446 Bulletin Muscinn of Comparative Zoology, Vol. 142, No. 5
(
relatively shiny, the lateral pronotal mar- gins are nsually broader, the elytral mega- punetures are usually larger than the pro- notal punctures, separated by 1 to 3 diameters, and almost as numerous as the micropunctures, so that the vestiture is sparser, the elytral bristles are longer, about 0.50 X as long as the scutellar base, and the size is usually smaller; this form is similar to the typical C. americanus occur- ring in the Northeast. In the second form, the pronotum and elytra are relatively dull, the lateral pronotal margins usually nar- rower, the elytral megapunctures smaller, separated by 2 to 5 diameters and much less numerous than micropunctures, so that the vestiture is denser, the elytral bristles are only about 0.33 X as long as the scutel- lar base, and the size is usually larger; this corresponds to Hatch's oweni. In Cali- fornia, the first form is found on several fungi, including Stereum hirsntiim, Stcc- cheriniim ochracetim, and Pohjporii.s .«//- phnreiis, while the second is found most often on Poltjponis adustiis and Fames pinicola. The differences between the two forms are slight and intermediates may be found. The two have been collected at the same localities on several occasions and may represent syinpatric sibling species. Cis tridentatus occurs in the same general area and is difficult to distinguish from the two forms described above; it is treated here as a distinct species because the differ- ences, however slight, appear to be con- sistent. A thorough analysis of this group in the Pacific Northwest would require more field study.
This species and its sibling, C. tridenta- tus, are most closely related to members of the Cis nitidiis group, from which they differ in the vestiture and in the form of the aedeagus (Fig. 75; cf. Fig. 74).
Cis angusfus Hatch
Cis an<iustus Hatch, 1962: 230. Type locality: "Stanley, B. C." Holotype, 9, CAS.
Distribution. Mountains of the Pacific
Coast, from south-central British Columbia to the southern Sierra Nevada (Fig. 94). Marginal records: BRITISH COLUMBIA: Stanley. WASHINGTON: Paradise Park, 6000',' Mt. Rainier, Pierce Co. CALI- FORNIA: 7 mi. N Mineral, 6000', Tehama Co.; Huntington Lake, Fresno Co.
Host fungi. Fomes pinicola [2(2)]; Fomes annosus [1]; Fomes officinalis [1].
Discussion. This species is characterized by the long and narrow, somewhat cylindri- cal, body form; single, coarse, and confused elytral punctation; weakly dentate or angu- late protil)ial apex (Fig. 52); and moder- ately long and fine, yellowish bristles. Cis creberrimtis differs in being flattened and in having much finer and denser elytral punctation. C. fcstivulus has shorter and stouter, yellowish bristles, \\'hich are un- even in length, and a rounded protibial apex (Fig. 47). C. rohiniophilus and C. hysfricuhts are both quite similar, but the foiTner has shorter and stouter bristles and finer elytral punctures, while the latter has colorless bristles and no abdominal fovea in the male.
Cis angiistus appears to be most closely related to C. rohiniophilus from the eastern United States and C. fagi Waltl and C. castaneus Mellie from the Palaearctic region. The species has been collected only on the Pacific Coast, but further field work may reveal a broader distribution. It ap- pears to be restricted to coniferous forests at higher elevations and may be considered part of the Holarctic faunal element. It has been found on three related species of Fomes.
Cis biarmafus Mannerheim
Cis hianudtus Mannerheim, 1852: 360. T>^e locality: "... Insulae Sitkhae." Lectotype, S, Mannerheim Coll., MZUH.
Xestocis hiarmata (Mannerheim), — Casey, 1898: 86 (dlst.); Fall, 1926: 200 (dist.); Hatch, 1962: 233, pi. 48, fig. 4, 4a (dist.).
Eridatilus hiarmatus (Mannerheim), — Lawrence, 1965: 281.
Cis hicarinatus LeConte, 1867: 58. Incorrect sub- sequent spelling.
North American Ciidae • Lawrence 447
Distribution. Pacific Coast of North America, from southern Alaska to Marin Co., Cahfomia (Fig. 102). Marginal records: ALASKA: Sitka; BRITISH CO- LUMBIA: Peachland; WASHINGTON: Carbon R., Mt. Rainier, Pierce Co.; ORE- GON: Sandy, Clackamus Co.; CALIFOR- NIA: 2 mi. SW Inverness, Marin Co.
Host fungi. Fomes pinicola [10(6)].
Discussion. Cis hiarmatus may be dis- tinguished by the short and stout body form, distinctly carinate prostemum, dual and confused elytral punctation, and vesti- ture of moderately long, decumbent hairs. Cis americonus, C. tridentatus, C. megastic- tus, C. maritimus, and C. levettei all differ in vestiture, having either short, fine hairs or short, stout bristles. In the closely re- lated C. ephippiatus, the body is somewhat smaller and less unifonnly pigmented, the pronotal punctation is finer and sparser, the elytral punctation is more distinctly dual and subseriate, and the anterior edge of the pronotum in the male is less strongly produced forward.
This species is relatively restricted in its distribution, occurring only along the Pacific Coast. Throughout its entire range it is sympatric with C. ephippiatus, but the range of the latter extends across the north- ern part of the continent to the Atlantic Coast. Cis hiarmatus has been taken al- most exclusively in Fomes pinicola growing on conifers in the coastal forests, while its sibling, C. ephip))iatus, is most commonly collected in Ganoderma applanatum on hardwoods, at least on the Pacific Coast (see further discussion under Cis epliip- piatus, p. 458). Other Ciidae usually col- lected in association with C. hiarmatus are Cis tridentatus, Dolichocis indistinctus and D. manitoha.
Cis castlei (Dury), NEW COMBINATION
Xestocis castlei Dury, 1917: 17. Type locality: "Cincinnati, Ohio." Holotype, $, Dury Coll., CIN.
Distribution. Eastern North America, from extreme southern Ontario and Penn-
sylvania south to central Florida, west to Iowa, and south through eastern Mexico to Costa Rica (Fig. 97). Marginal records: ONTARIO: Tilbury, Essex Co.; PENN- SYLVANIA: Chestnut Hill, Philadelphia Co.; FLORIDA: Highlands Hammock, 6 mi. W Sebring, Highlands Co.; LOUISI- ANA: 5 mi. S Livingston, Livingston Co.; MISSISSIPPI: 5 mi. N Ackerman, Choc- taw Co.; IOWA: Maquoketa Caves State Park, Jackson Co.; SAN LUIS POTOSI: Huichihuayan; VERACRUZ: El Fortin; COSTA RICA: San Jose.
Host fungi. Folyponis adustus [9(4)]; Polyporus pargamemis [5]; Folyponis hyd- noides [2]; Stereum ostrea [2]; Folyporus mutahilis [1(1)]; Folyporus vinosus [1(1)]; Trametes corrugata [1(1)]; Ganoderma applanatum [1]; Lentinus crinitiis [1]; Polyporus lignosiis [1]; Folyponis zonalis [1]; Foria latemarginata [1]; Trametes pleheja [1].
Discussion. This species differs from most North American Cis by the small size (TL usually less than 1.4 mm), short and broad body form, weakly carinate pro- stemum, single and confused elytral punc- tation, vestiture of short, stout bristles, pronotum with narrow lateral margins that are not crenulatc, and uniquely curved, sub triangular, frontoclypeal plates in the male (Fig. 7). Cis duplex differs in being larger (TL rarely less than 1.4 mm) and in having the pronotum dull with coarser and denser punctation and smooth lateral edges. C. laniinatus is much larger, with a relatively smaller scutellum, slightly tumid prosternum, and simple pronotal apex in the male. Cis vitulus, C. congestus, and C. quadridenfatus differ in having the anterior pronotal angles distincth' produced forward, the lateral pronotal margins visible from above, and tlie frontoclypeal ridge quadridentate in the male; in addi- tion, the first two species are much larger in size.
Cis castlei occurs throughout the eastern United States and extends into tropical Mexico and Central America. It does not
448 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
appear to be closely related to any other North American species. It has been found in a number of fungi, but Polyponis aclus- tus may be the preferred host.
Cis cayensis NEW SPECIES
Holoti/pe. S, FLORIDA: Big Pine Key, Monroe" Co., June 4, 1968, Lot 2612 J. F. Lawrence, ex Poh/porus porrectus ? [MCZ No. 31684]. Allotype, 9, same data [MCZ].
Male. Length 1.27 mm. Body 2.12 X as long as broad, moderately convex. Head and pronotum blackish brown, elytra yel- lowish brown. Vestiture dual, consisting of moderately short, stout, blunt, yellowish bristles, and very short, fine, recurved, pale hairs. Vertex with deep, circular, median impression and two sharp, lateral tubercles; frontoclypeal ridge bearing two rounded tubercles on each side, the mesal two sepa- rated by 2.33 basal widths. Antennal seg- ment III 2.00 X as long as IV. Pronotum 0.78 X as long as broad, widest at posterior third; anterior edge weakly rounded, slightly flattened at midline; sides moder- ately rounded, the margins narrow and distinctly crenulate, not visible for their entire lengths from above; anterior angles not produced forward, obtuse; disc moder- ately convex, even; surface coarsely granu- late and dull; punctures shallow, shiny, 0.25 X as large as scutellar base, and separated by 0.50 to 1.25 diameters. Elytra 1.37 X as long as broad and 1.83 X as long as pronotum; sides weakly rounded, apices blunt; punctation dual and subseriate; megapunctures finer and denser than those on pronotum, sometimes confluent within a row, shallow and nude; micropunctures bearing either a fine, recurved, pale hair, which is about 0.25 X as long as scutellar base, or a stout, erect, blunt, yellowish bristle, which is 0.60 X as long as scutellar base. Prosternum moderately tumid and subcarinate; intercoxal process 0.45 X as wide as a procoxal cavity, parallel-sided. Protibia with outer apical angle expanded and rounded (Fig. 48). Metasternum 0.56 X as long as wide; suture 0.30 X as long as
median length of sternum. Abdominal stemite III bearing a median, circular, pubescent fovea, which is 0.23 X as long as body of stemite, indistinctly margined, and located slightly anterad of center.
Female. Length 1.32 mm. Body 2.17 X as body of sternite, indistinctly margined, frontoclypeal ridge simple. Pronotum 0.77 X as long as broad; anterior edge moder- ately rounded. Elytra 1.46 X as long as broad and 2.06 X as long as pronotum. Protibia as in male. Stemite III without pubescent fovea.
Variation. Pronotum yellowish orange to black, usually blackish brown. Elytra yel- lowish to dark brown, usually yellowish brown. Tubercles on vertex and frontoclyp- eal ridge in male vary somewhat in size and shape. Size and dimensions vary as follows in a series of 25 males and 24 fe- males from Big Pine Key, Florida (Lots 2612 and 2613):
TL mm: S 1.02-1.37 (1.19 ± 0.019),
9 1.0,5-1.35 (1.19 ±0.018); TL/EW: $ 2.05-2.25 (2.17 ± 0.010),
9 2.12-2.29 (2.19 ±0.009); PL/PW: c^ 0.75-0.85 (0.80 ± 0.005),
9 0.75-0.85 (0.80 ±0.005); EL EW: $ 1.3.5-1.50 ( 1.42 ± 0.008),
9 1.41-1.52 (1.46 ±0.007); EL/PL: $ 1.71-2.00 ( 1.89 ± 0.013),
9 1.82-2.14 (2.01 ±0.015); GD/EW: c^ 0.75-0.81 (0.77 ± 0.004),
9 0.75-0.84 (0.79 ±0.005). Total size range in material examined: 0.90 -1.40 mm.
Parati/pes. CUBA: 2, Buenos Aires, Trinidad Mts., VI. '39, Parsons [MCZ]. FLORIDA: Monroe Co.: 1, Big Pine Key, June 4, 1968, Lot 2611 JFL, ex Poh/porus porrectus ? [JFL]; 29, same locality and date. Lot 2612 JFL, ex Polijporus porrectus ? [FMNH, JFL, MCZ, USNM]; 23, same locality and date. Lot 2613 JFL, ex Poly-' j)orus porrectus ? [JFL, MCZ]; 8, same locality and date, Lot 2615 JFL, ex Poly- porus hijdnoides [JFL]; 6, same locality and date, Lot 2616 JFL, ex Polyporus
North American Ciidae • Lawrence 449
hydnoicJes [JFL]; 4, Key Largo, N end, Feb. 10, 1968, Lot 2550 JFL (S. B. Peck, coll.), ex Polijporus corrosus [JFL]; 6, Lig- num Vitae Key, May 28, 1968, Lot 2577 JFL, ex Fomes whiniae [JFL, MCZ]; 1, same locality and date. Lot 2580 JFL, ex Poli/ponts hydnoides [JFL]; 1, same lo- cality, June 1, 1968, Lot 2596 JFL, ex Steretim papyrimtm [JFL]; 3, same locality and date, Lot 2600 JFL, ex Fomes robiniae [JFL]; 2, same locality and date, Lot 2601 JFL, ex Fomes robiniae [JFL]; 1, same locality, June 5, 1968, Lot 2622 JFL, ex Fomes robiniae [JFL]; 2, same locality and date. Lot 2623 JFL, ex Fomes robiniae [JFL]; 3, Marathon, Mar. 15, 1968 (S. B. Peck, coll.), berlese sample B-111, litter and soil in sciTib forest [MCZ]; 5, Pennekamp State Park, Key Largo, June 28, 1965, Lot 1525 JFL, ex Polijporus porrecttis ? [JFL]; 13, same locality and date, Lot 1526 JFL, ex Polijporus porrectus ? [JFL, MCZ]; 2, same locality. May 29, 1968, Lot 2584 JFL, ex Polijponis porrectus ? [JFL].
Distribution. Florida Keys and Cuba.
Host fungi. Polijporus porreetus ? [6(4)]; Fomes robiniae [5]; Polijporus hijdnoides [3(2)]; Polijporus corrosus [1(1)]; Stereum papyrimtm [1].
Discussion. This species is unique among the North American Ciidae in the type of elytral vestiture, expanded and rounded protibial apex, and tubercles on the vertex of the male. The elytra bear short, erect, seriate bristles, alternating with very short and fine, inclined hairs, but the latter are not visible under lower magnifications, so that the vestiture may appear single. The expanded and rounded protibial apex and the male amiature are both found in Cis niedhauki, but that species differs in being subglabrous, the vestiture consisting of only short, fine hairs. In general body form and seriate elytral bristles, C. cayensis re- sembles C. striolatus and C. tristis, but it differs from both in the smaller size, pro- tibial apex, and head of the male.
This is a West Indian species and prob- ably occurs throughout the Greater An-
tilles. It appears to prefer melanic conks, as does the closely related Cis niedhauki.
The name cayensis is derived from the Spanish cayo, meaning reef or small island
(key).
Cis congestus Casey
Cis congesta Casey, 1898: 82. Type locality: "Louisiana." Holotype, 9, Casey Coll., USNM.
Cis lodingi Dury, 1917: 6; Blatchley, 1923: 19 (dist. ). Type locality: "Mobile, Ala." Syntypes, 9 2, Dury Coll., CIN. NEW SYNONYMY.
Distribution. Southeastern United States, from Maryland south to northern Florida and west to southern Illinois and eastern Texas (Fig. 105). Marginal records: MARYLAND: Baltimore, Baltimore Co.; FLORIDA: Ormond, Volusia Co.; TEXAS: San Felipe, Austin Co.; ILLINOIS: Foun- tain Bluff, Jackson Co.
Host fungi. Poly poms hirsutus [3(2)]; Polijporus pinisitus [1].
Discussion. This species may be dis- tinguished from most North American Cis by the very short and stout body fonn, single and confused elytral punctation, strongly tumid prostemum, distinctly pro- duced anterior pronotal angles (Fig. 26), and vestiture of short bristles. In addition, the male has a trisinuate frontoclypeal ridge (Fig. 3), two stout horns on the pronotum, and no pubescent fovea on the abdomen. It differs from the closely re- lated C. vitulus by having coarser and denser pronotal punctation, shiny pronotal surface, and a shorter and stouter body form. Cis laminatus has a similar build and a trisinuate frontoclypeal ridge in the male, but in that species the lateral pro- notal margins are narrower, distinctly crenulate, and lack a raised lip. Cis huachucae is also similar and lacks the abdominal fovea in the male, but the vesti- ture is dual, consisting of longer and shorter bristles. In Cis quadridentatus, the prosternum is carinate, the size is smaller, and the anterior pronotal angles are not as strongly produced fonvard.
Cis congestus is restricted to the south-
450 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
eastern United States, while C. vitulus is fomid on the Pacific Coast. The two spe- cies belong to a group consisting primarily of Neotropical forms, such as Cis huhalus Reitter, C. fasciatus Gorham, C. hishidens Gorham, C. gronariiis Mellie, and C. (i^ros- siis Mellie. The few host records indicate that Cis congestiis, like the otlier members of this group, prefers the whitish conks of Pohjporns hirsutus and its relatives.
This species may be parthenogenetic, like Cis fuscipes, which is discussed below (p. 460) and in another paper (Lawrence, 1967a). Of all of the collections examined so far, only two contained males. In a series of 50 specimens from Mobile, Ala- bama, no males were found, 3 males were seen in a series of 29 specimens from Ten- nessee, and a single male turned up in a series of 20 specimens from North Carolina. A total of 4 males were found in the total sample of 136 specimens. Further collect- ing and laboratory rearing will be required to verify the possible existence of partheno- genesis.
Cis cornelli NEW SPECIES Figure 42
Holotype. S, NORTH CAROLINA: At- lantic Beach, Carteret Co., May 19, 1966, Lot 1890 J. F. Lawrence (Carl Parsons, coll.), ex Poh/porus sector on dead hard- wood [MCZ" No. 31685]. Allotype, 9, same data [MCZ].
Male. Length 1.52 mm. Body 1.74 X as long as broad, strongly convex. Head and pronotum dark reddish brown, elytra dark yellowish brown. Vestiture of moder- ately short, fine, blunt, yellowish bristles. Vertex with a broad, shallow, median im- pression; frontoclypeal ridge bearing 2 lateral homs, which are about 4.0 X as long as wide at base, 0.86 X as long as pronotum, and separated by 1.5 basal widths, with mesal edges slightly and lateral edges strongly converging. Antennal segment III 1.70 X as long as IV. Pronotum 0.68 X as long as broad, widest at posterior
third; anterior edge weakly rounded, flat- tened at middle; sides weakly rounded, the margins broad and very weakly crenulate, easily visible for their entire lengths from above; anterior angles distinctly produced forward, rounded; disc strongly convex, even; surface finely granulate and shiny; punctures 0.33 X as large as scutellar base and separated by 0.50 to 0.75 diameter. Elytra 1.14 X as long as broad and L90 X as long as pronotum; sides moderately rounded, apices blunt; punctation dual and seriate; megapunctures about as coarse as those on pronotum, very dense and almost confluent within rows, shallow and nude, micropunctures located mainly between rows, each bearing a moderately fine, blunt, yellowish bristle, which is 0.67 X as long as scutellar base. Prosternum strongly tumid and carinate; intercoxal process 0.40 X as wide as a procoxal cavity, parallel- sided. Protibia with outer apical angle produced and dentate (Fig. 50). Metaster- num 0.38 X as long as wide; suture 0.35 X as long as median length of sternum. Abdominal sternite III bearing a median, circular, pubescent fovea, which is 0.27 X as long as body of sternite, indistinctly margined, and located anterad of center. Sternite VIII as in Figure 67. Aedeagus as in Figures 76 and 79.
Female. Length 1.57 mm. Body 1.80 X as long as broad. Vertex flattened; fronto- clypeal ridge simple. Pronotum 0.69 X as long as broad; anterior edge strongly rounded. Elytra 1.17 X as long as broad and 1.86 X as long as pronotum. Protibia as in male. Sternite III without pubescent fovea.
Variation. Pronotum yellowish orange to blackish brown, usually reddish brown. Elytra yellowish to blackish brown, usually dark yellowish brown. Frontoclypeal homs in smaller males shorter and subtriangular,. as little as 0.25 X as long as pronotum; in ' larger specimens the horns are longer and narrower and may be as much as 0.90 X as long as pronotum. Size and dimensions
North American Ciidae • Lawrence 451
vary as follows in a mixed series of 23 males and 25 females:
TL mm: S 1.32-1.55
9 1.45-1.72 (1.58± TL/EW: 6 1.71-1.88
9 1.75-1.89 (1.80 + PL/PW: i 0.64-0.72
9 0.64-0.71 (0.67 ± EL EW: i 1.12-1.25
9 1.14-1.26 (1.19± EL PL: S 1.77-2.12 (1.93±
9 1.82-2.00 (1.93 ±0.012); GD'EW: c^ 0.75-0.88 (0.80
9 0.76-0.84 (0.80 + 0.004).
(1.49 ±0.011),
0.013);
(1.79 ±0.011), 0.007); (0.68 ±0.004),
0.003);
(1.18 ±0.008),
0.006);
0.016), t 0.006)
Pairihipes. FLORIDA: 5, 5 mi. W Gainesville, Alachua Co., Nov. 29, 1963, H. S. Dvbas, No. 63-86 (S. B. Peck, coll.), live oak litter [FMNH, JFL]. NORTH CARO- LINA: 12, Atlantic Beach, Carteret Co., May 19, 1966, Lot 1890 JFL (Carl Parsons, coll.), ex Pohjponis sector [JFL, MCZ, USNM]; 4, Randolf Co., XII-18-63, J. F. Cornell [JFC, JFL]. SOUTH CAROLINA: 35, Yemassee, Beaufort Co., XII-28-63 (J. F. Cornell, coll.), berlese from litter under log in palmetto-cypress bog [JFC, FMNH, JFL, USNM].
Distribution. North and South Carolina and Florida.
Host fungi. Poh/porus sector [1(1)].
Discussion. This species is characterized by the very short and broad body form (EL/EW less than 1.25), strongly carinate prostemum, protibial apex with a stout tooth and several spines, dual and seriate elytral punctation, vestiture of short, fine bristles, and male with a simple pronotal apex and two long and narrow, lateral horns on the clypeus. In Cis fuscipes, the elytra are longer and narrower, the pro- sternum is not carinate, and the frontoclyp- eal horns are absent in the male.
Cis cornelU is the only North American member of a Neotropical species group formerly included in the genus Macrocis {see p. 439). It resembles the South American species Cis testaceus (Pic), C. grandicornis (Pic), and C. testaceimemhris
(Pic), but the first two are larger (TL more than 2 mm) and the last two have broader, blunt, frontoclypeal horns in the male. C. setifer (Gorham), known from Mexico and Centi-al America, is much smaller (less than 1.3 mm) with shorter and stouter, colorless bristles.
The species is probably distributed throughout the Southern Coastal Plain. It has been named after Dr. J. F. Cornell, who collected the first series.
Cis cornutus Blatchley
Cis cornutus Blatchley, 1910: 898, fig. 353. Type locality: "Grand Chain, Posey Co." [Indiana]. Holotype, S , Blatchley Coll., PURD.
Cis hirsuta Casey, — Weiss and West, 1920: 8 (biol., dist. ). Misidentification.
Distribution. Eastern United States, from New York south to northern Florida and west to Illinois and Arkansas. Marginal records: NEW YORK: West Point, Orange Co.; FLORIDA: Jacksonville, Duval Co.; ARKANSAS: Washington Co.; ILLINOIS: Fountain Bluff, Jackson Co.
Host fungi. Poly porus versicolor [4(3)]; Polij.porus suhectus [1(1)].
Discussion. Cis cornutus may be dis- tinguished by the dual vestiture, consisting of longer and shorter, yellowish bristles, which are seriate on the elytra, and the short and broad body form with fairl>- broad lateral pronotal margins. Tlie male bears two subtriangular plates on the frontoclypeal ridge, a single, raised, rounded process on the anterior edge of the pronotum, and a small, pubescent fovea on the abdomen. The elytral bristles are much finer than those in C. huachucae and fall into two distinct size classes; in addition, they are yellowish and seriate. In C. crini- tus, the longer bristles are dark in color, the vestiture is confused, and the elytra are longer and narrower. In C. fuscipes, the elytral bristles may be seriate, but they are relatively uniform in length and not as fine, while the elytra are more elongate and the male armature is different.
Cis cornutus is most closely related to
452 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
Cis pilosus Gorham, described from Guate- mala and occurring also in Costa Rica and Mexico. In C. pilosus the vestiture is longer and finer, the male fovea is absent, and the aedeagus is slightly different.
Cis creberrimus Mel lie
Cis creberrimus Mellie, 1848: 357, pi. 12, fig. 5; Gorham, 1883: 222 (dist.); Casey, 1898: 80 (dist); Blatchley, 1910: 897 (dist., biol.); Weiss and West, 1921b: 169 (dist, biol.); Blackman and Stage, 1924: 85 (bid.). Type locality: "Nouvelle Orleans." Lectotype, 5, Pic CoU. (Chevrolat Coll.), MNHN.
Cis puberulus Mellie, 1848: 358; Gorham, 1883: 222 (dist.). Type locaUty: "Saint-Thomas." Holotype, 5, Marseul Coll., MNHN. NEW SYNONYMY.
Cis mibillus Gorham, 1898: 331. Type locality: "St. Vincent: south end." Holotype, S , BMNH. NEW SYNONYMY.
Cis luibilus Dalla Torre, 1911: 15. Incorrect sub- sequent spelling.
Distrihiition. Eastern North America, from Vermont south to Florida and west to eastern Nebraska, Kansas, and Texas; mon- tane regions of the Southwest; throughout the West Indies; scattered localities in Mexico, Central America, and South Amer- ica as far south as Santa Catarina, Brazil; Galapagos Islands (Fig. 98). Marginal records: VERMONT: Manchester, Ben- nington Co. FLORIDA: Paradise Key, Dade Co. NEBRASKA: Central City, Merrick Co. KANSAS: Lawrence, Douglas Co. TEXAS: Kerrville, Kerr Co.; Browns- ville, Cameron Co. NEW MEXICO: Albu- querque, Bernalillo Co. ARIZONA: Hitch- cock Highway, Mile 25, Santa Catalina Mts., Pima Co.; Rustler Park, 8 mi. W Portal, Cochise Co. CALIFORNIA: 4 mi. E Running Springs, San Bernardino Co.; 1.5 mi. NW Mt. Laguna, San Diego Co. BAHAMAS: Pine Ridge, Grand Bahama Is.; 2 mi. E Conch Shell Hill, Great Inagua Is. CUBA: Pinar del Rio. JAMAICA: Windsor, 10 mi. S Falmouth, Trelawny Par. GRENADA: Mount Gay Est. (leeward side). SINALOA: 22 mi. S Espinol. GUATEMALA: El Reposo, SOC; Zapote. CANAL ZONE: Barro Colorado Is.
COLOMBIA: Medellin, Antioquia Prov. VENEZUELA: Yacua, Sucre Prov. BRA- ZIL: Nova Teutonia, Santa Catarina. GALAPAGOS IS.: 6 mi. N Academy Bay, Santa Cruz Is. This is a composite distri- buti(jn in that several closely related species are probably involved {see discussion be- low).
Host fungi. Fames sclerodermeus [6(1)]; Polyporus adustus [5(1)]; Ganoderma hici- dum [4(2)]; Ganoderma applanatum [4]; Polyporus Jnjdnoides [3(1)]; Fames affi- cinalis [2(2)]; Paria latemarginata [2(2)]; Fames pinicala [2(1)]; Polyporus par- gamenus [2(1)]; Trametes corrugata [2 (1)]; Polyporus gilvus [2]; Fames annosus [1(1)]; Ganoderma fuhellum [1(1)]; Poly- porus sanguineus [1(1)]; Paria corticala [1(1)]; Daedalea elegans [1]; Lenzites ■striata [1]; Pleurattis sp. [1]; Polyporus anceps [1]; Polyporus hirsutus [1]; Poly- porus sulphureus [1]; Polyporus .supinus [1]; Schizophyllum commune [1]; Trametes hispida [1].
Discussion. The material on which the above distributional summary is based probably represents a complex of species, which are here included under the single name Cis creberrimus, until a more thorough analysis is possible. In North America, the .species may be distinguished from other ciids by the elongate and flattened fonn, narrow lateral pronotal margins, dentate protibial apex, flat pro- sternum, fairly coarse and dense, single, subseriate, elytral punctation, and four rounded tubercles on the frontoclypeal ridge in the male.
The surface of the pronotum is subject to considerable variation. Punctation may be fine and sparse to coarse and dense, while the surface may be granulate and dull to smooth and shiny. In North America, southwestern populations have the pronotum smooth and shiny with coarse and dense punctation, while eastern popu- lations are characterized by having a dull pronotum with finer and sparser punc- tation.
North American Ciidae • Lawrence 453
The vestiture is highly variable iii this complex. The elytra may be clothed with relatively short and stout bristles, which are not obviously dual, longer and finer bristles, which are erect and inclined, or long, fine hairs. Long and fine hairs are characteristic of specimens from the Chiri- cahua Mountains of Arizona, while the short bristles occur in midwestem and eastern populations. The dual vestiture, consisting of erect and inclined, long bristles, occurs in various populations from the Southwest and Southeast. Some speci- mens with long, fine hairs are known from the Southeast as well.
Using these two character complexes, the North American specimens fall into four geographic segregates: 1) typical Ci.s creberrimtis with fine and sparse pronotal punctation, dull surface, and fairly short elytral bristles; 2) Floridian specimens with similar pronotal surface but with dual vestiture of erect and inclined, long bristles;
3) specimens from the mountains of south- ern California, Arizona, and New Mexico with coarse and dense pronotal punctation, shiny surface, and dual, long vestiture; and
4) Chiricahua Mountain specimens with similar pronotal surface and long, fine hairs. A very few specimens from Ohio, Louisiana, Mississippi, and Florida fall into the last category also.
There is some intergradation between the two eastern forms, and the increase in bristle length in the East may be clinal. The dual nature of the vestiture in southeastern populations is more obvious because of the greater length of the individual elements, and the shorter bristles of "typical" C. creberrimus appear to be dual upon closer examination. The situation becomes much more complex if the Mexican and West Indian forms are considered. All types oc- cur south of the United States, but material is not abundant and patterns are difficult to discern. Two different types are rarely found together, but when this does occur it would be difficult to justify lumping the forms together as one species. Both of the
described species from the West Indies (C. puberulus from St. Tliomas and C. mibillus from St. Vincent) are similar to populations from the southeastern United States and are the least likely to be specifi- cally distinct.
A preliminary analysis of the male geni- talia in this complex revealed a certain amount of variation in the form of the tegmen that may be correlated with dif- ferences in vestiture described above. A more detailed analysis must be postponed until more material can be examined.
Cis creberrimus is not an uncommon species in eastern North America, but it is rarely taken in large numbers. It may be found in association with various fungi but it is usually not the dominant species in a fruiting body. It is also commonly en- countered under bark. The species or com- plex is widespread in the New World and is one of the two ciids to occur on the Galapagos Islands, the other being Ceracis cucuUatus (MelHe), an equally ubiquitous form {see Lawrence, 1967b).
Cis crinitus NEW SPECIES
HoJotype. S, FLORIDA: Lignum Vitae Key, Monroe Co., June 5, 1968, Lot 2624 J. F. Lawrence, ex Stcreum papijrimim [MCZ No. 31686]. Allotype, ?, same data [MCZ].
Male. Length 1.72 mm. Body 2..30 X as long as broad, moderately convex. Head and pronotum reddish browm, el\i:ra yel- lowish brown. Vestiture distinctly dual, consisting of shorter, fine, recurved, yel- lowish hairs, and longer, stiff, erect, darker bristles (Fig. 29). Vertex flattened; fronto- clypeal ridge bearing 2 lateral horns, which are about 2 X as long as wide at base, 0.16 X as long as pronotum, and separated by 3.5 basal widths, \\\i\\ mesal edges subparallel and lateral edges strongly converging. An- tennal segment III 1.66 X as long as IV. Pronotum 0.90 X as long as broad, widest at posterior fourth; anterior edge strongly produced and emarginate, forming 2 ap- proximate, sharp, subtriangular homs.
454 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
which are 0.14 X as long as pronotum; sides subparallel, the margins broad and weakly crenulate, easily visible for their entire lengths from above; anterior angles dis- tinctly produced forward, subacute; disc moderately convex, even; surface smooth and shiny; punctures 0.30 X as large as scutellar base and separated by 0.25 to 0.50 diameter. Elytra 1.43 X as long as broad and 1.65 X as long as pronotum; sides subparallel, apices blunt; punctation ap- parently single and confused; punctures about as large as, but sparser than, those on pronotum, each puncture bearing either a fine, recurved, yellowish hair, which is about as long as scutellar base, or a stiff, erect, dark bristle, which is 1.50 X as long as scutellar base, the hairs being more abundant than the bristles. Prostenium slightly tumid; intercoxal process 0.40 X as wide as a procoxal cavity, parallel-sided. Protibia with outer apical angle strongly produced and dentate. Metasternum 0.50 X as long as wide; suture 0.33 X as long as median length of sternum. Abdominal sternite III bearing a median, circular, pubescent fovea, which is 0.36 X as long as body of sternite, distinctly margined, and located about in center.
Female. Length 1.77 mm. Body 2.15 X as long as broad. Vertex as in male; frontoclypeal ridge simple. Pronotum 0.77 X as long as broad; anterior edge strongly rounded, simple. Elytra 1.45 X as long as broad and 2.09 X as long as pronotum. Protibial apex only slightly produced, angulate or weakly dentate. Sternite III without a pubescent fovea.
Variation. Pronotum yellowish orange to dark brown, usually reddish browai. Elytra yellowish to dark brovvn, usually yellowish brown. Frontoclypeal ridge in smaller males bearing 2 short, subtriangular proc- esses; in larger specimens these are repre- sented by longer, narrow horns, which may be 0.20 X as long as pronotum. Anterior edge of pronotum in smaller males barely produced and emarginate, forming 2 small tubercles; pronotal horns in larger speci-
mens may be 0.20 X as long as pronotum. In West Indian specimens, the erect bristles are finer and lighter in color, so that the vestiture is less obviously dual. These insular specimens are almost surely con- specific with those on the mainland, but they have not been included in the para- type scries. Size and dimensions vary as follows in a mixed series of 10 males and 16 females from Florida:
TL mm: S 1.52-2.12 (1.81 ± 0.054),
? 1.67-2.32 (1.89 ±0.037); TL/EW: $ 2.16-2.33 (2.26 ± 0.018),
9 2.10-2.26 (2.17 + 0.010); PL/PW: 6 0.79-0.94 (0.88 ± 0.015),
9 0.74-0.82 (0.79 ±0.006); EL EW: S 1.38-1.47 ( 1.43 ± 0.008),
9 1.39-1.53 (1.45 ±0.010); EL PL: 6 1.61-2.00 ( 1.72 ± 0.040),
? 1.91-2.21 (2.03 ±0.023); GD EW: c^ 0.72-0.83 (0.77 ± 0.010),
9 0.74-0.80 (0.76 ±0.004).
Parati/pe.s. ALABAMA: 1, Mobile, 3.4 [GIN]. FLORIDA: 1, Big Pine Kev, Mon- roe Go., June 4, 1968, Lot 2605 JFL, ex Fohjporus sp. [JFL]; 1, same locaHty and date. Lot 2616 JFL, ex Folijporus hijdnoides [JFL]; 1, Grawl Key, Monroe Go., June 6, 1968, Lot 2630 JFL, ex Foh/porm fidvo- cincrem- [JFL]; 1, Dunedin," 12-7-21, W. S. B.[latchley] [PURD]; 1, same locality, 1-8-26, W. S. B. [PURD]; 1, same locality, 3-20-16, W. S. B. [PURD]; 1, same lo- cality, ;3-5-20, W. S. B. [PURD]; 2, Grassy Key, Monroe Go., June 28, 1965, Lot 1522 JFL, ex Folijponis fidvocinereu.s [JFL]; 1, Indian River, IV-7-30, J. R. Barass, Florida Fruit Fly Trap Surv. [USNM]; 3, Lignum Vitae Key, Monroe Go., June 5, 1968, Lot 2624 JFL, ex Sfereiirn papyriniim [FMNH, JFL]; 1, same locaHty, June 1, 1968, Lot 2595 JFL, ex Fohjporus fidvocinereus [JFL]; 1, same locality and date. Lot 2596 JFL, ex Stereuni papyrinum [JFL]; 1, same locality and date, Lot 2597 JFL, ex Poly- porus hydnoides [JFL]; 1, Marathon, Mon- roe Go."; Mar. 10, 1968, Lot 2545 JFL (S. B. Peck, coll.), ex Ganoderma zonatum
North American Ciidae • Lawrence
455
[JFL]; 1, Punta Gorda, II-7-40, Van Dyke Collection [CAS]; 5, St. Augustine, III-5- 1940, Van Dyke Collection [CAS]; 1, Tampa, 5.4, J. L. LeConte Collection [MCZ]. GEORGIA: 1, Tvbee Is., VI-23, Coll. by H. A. Wenzel [CIN]. LOUISI- ANA: 2, Covington, 28, V, Collection H. Soltau [USNM]. NORTH CAROLINA: 1, Longbeach, VI-12-53, G. H. Nelson, Beat- ing Qiierciis virginiana [GHN]. TEXAS: 1, Borden, 18.6 [MCZ].
Additional material. BAHAMAS: 1, 2 mi. E Conch Shell Hill, Great Inagua, Feb., 1967, Lot 2058 JFL (A. Laska, coll.), ex Polypurus hydnoides [JFL]. CUBA: 1, Cavamas, 29.5, E. A. Schwarz, Collector [USNM]; 1, same locality, 29.12, E. A. Schwarz [USNM]; 1, same locaHty, 9.3, E. A. Schwarz [USNM]; 1, same locality, 4.2, E. A. Schwarz [USNM]; 4, same locality, 10.6, E. A. Schwarz [USNM]; 1, same lo- cality, 12.3, E. A. Schwarz [USNM]. JA- MAICA: 1, Try, nr. Falmouth, VIII-9-1966, mangrove swamp, A. T. Howden [HH]. PUERTO RICO: 18, Caja de Muertos Is., Dec. 10, 1961, Lot 1660 JFL ( H. Heatwole, coll.), ex Lenzites striata [JFL].
Distribution. Southeastern United States, from North Carolina to Florida and west to eastern Texas, the Bahamas, and the Greater Antilles (Fig. 96).
Host fungi. Polyporus hydnoides [3(1)]; Polyporiis fidvocinereus [3(1)]; Stereum papyrinum [2(2)]; Lenzites striata [1(1)]; Ganoderma zomitum [1].
Discussion. Individuals of this species are easily recognized by the unique vesti- ture of short, recurved hairs and long, erect bristles. In most other species with dual vestiture, there are two classes of bristles, which differ mainly in length and angle of inclination (Fig. 30), but in C. crinitus the large bristles are much stouter and darker in color than the short hairs (Fig. 29). In Cis cayensis the bristles are short, pale, and seriate, while the hairs are very small and barely visible at lower magnifications. In C. cornutus the two types of vestitur(^ are not as distinct (longer, erect and shorter,
inclined, yellowish bristles) and are sub- seriate on the elytra.
Cis crinitus belongs to the Cis hirsutus group, but it is the only member with distinctly dual vestiture. In Cis hirsutus. C. rotundulus, and C. ursulinus the vesti- ture consists of long, recurved, fine hairs. Individuals of C. crinitus are usually larger than those of C. rotundulus or C. ur.mlinus and smaller and more elongate than those of C. hirsutus.
The name crinitus is taken directly from the Latin \\'ord meaning hairy.
Cis discolor NEW SPECIES
Holotype. S, ARIZONA: Rusder Park, 8 mi. W Portal, Cochise Co., Aug. 8, 1961, Lot 922 J. F. Lawrence, ex Fomes cajanderi on Pimis sp. [CAS]. Allotype, 2, same lo- cality and date. Lot 918 J. F. Lawrence, ex Fomes cajanderi on Pinus sp. [CAS].
Male. Length 2.45 mm. Body 2.45 X as long as broad, moderately convex. Head blackish posteriorly, reddish orange ante- riorly; pronotum reddish orange, with t^vo broad, longitudinal, black fasciae extending from the posterior edge to the anterior fifth; elytra dark browaiish posterolaterally and anteriorly along the suture, reddish orange at apices, each elytron with a vague, median, reddish orange fascia that broad- ens posteriorly. Vestiture of short, stout, somewhat flattened, colorless bristles. Ver- tex with a slight median elevation; fronto- clypeal ridge bearing two rounded tubercles on each side, the mesal two separated by 1.0 basal width. Antennal segment III 1.80 X as long as IV. Pronotum 0.86 X as long as broad, widest at posterior third; anterior edge strongly rounded, slightly emarginate at midline; sides weakly rounded, the mar- gins broad and weakly crcnulate, easily visible for their entire lengths from above (Fig. 44); anterior angles distincUy pro- duced fonvard, rounded; disc moderately convex, broadh' impressed anteriorly; sur- face finely granulate and shin\-; punctures varying considerabK' in size, 0.10 to 0.25 X as large as scutellar base, and separated by
456 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
0.20 to 0.50 diameter. Elytra 1.67 X as long as broad and 2.16 X as long as pro- notmn; sides subparallel, apices blunt; disc with a distinct transverse impression at anterior fourth; punctation single and con- fused; punctures more uniform in size and somewhat smaller than the larger of the pronotal punctures, each bearing a stout, somewhat flattened, colorless bristle, which is about 0.33 X as long as scutellar base. Prostemum slightly tumid, intercoxal proc- ess 0.43 X as wide as a procoxal cavity, broadened posteriorly. Protibia with outer apical angle produced and dentate (Fig. 56). Metastemum 0.58 X as long as wide; suture 0.26 X as long as median length of sternum. Abdominal stemite III bearing a median, oval, pubescent fovea, which is 1.40 X as long as wide, 0.17 X as long as body of sternite, distinctly margined, and located anterad of center. Stemite VIII as in Figure 62. Aedeagus as in Figures 68 and 81.
Female. Length 2.70 mm. Body 2.35 X as long as broad. Vertex as in male; fronto- clypeal ridge simple. Pronotum 0.85 X as long as broad; anterior edge barely emar- ginate at midline. Elytra 1.61 X as long as broad and 2.04 X as long as pronotum. Protibia as in male. Stemite III without a pubescent fovea.
Variation. Pronotum yellowish orange to reddish brown with variable brownish or blackish markings laterally, usually red- dish orange with 2 broad, blackish brown, lateral fasciae. Elytra yellowish to black with variable reddish orange markings, usually reddish orange with vague, dark brownish fasciae along the suture and lateral edges. Specimens from the Chiri- cahua Mountains are darker in color than those from the Santa Rita Mountains, but the latter sample consists mainly of tenerals from a single lot. The Chiricahua speci- mens are also significantly larger in size. Male specimens vary slightly in the de- velopment of secondary sexual characters; in some males the mesal frontoclypeal tubercles are sharper than the lateral ones.
Size and dimensions vary as follows in a series of seven males and four females from Rustler Park (Chiricahua Mts. ), Cochise Co., Arizona (Lots 918, 922, and 1294):
TL mm: <^ 2.45-2.85 (2.61 ± 0.052),
9 2.32-2.80 (2.65 ±0.110); TL/EW: 6 2.40-2.49 (2.44 ± 0.012),
9 2.35-2.41 (2.38 ±0.013); PL PW: i 0.81-0.87 (0.85 ± 0.008),
9 0.83-0.90 (0.86 ±0.014); EL/EW: $ 1.64-1.72 ( 1.66 ± 0.012),
9 1.61-1.67 (1.64 ±0.012); EL/PL: 6 2.08-2.30 (2.14 ± 0.029),
9 2.17-2.32 (2.22 ±0.035); GD/EW: c? 0.74-0.78 (0.76 ± 0.004),
9 0.72-0.76 (0.74 ±0.009).
A second series of 27 males and 19 fe- males from Madera Canyon (Santa Rita Mts.), Santa Cmz Co., Ariz. (Lot 1282) shows the following variation:
TL mm: S 1.80-2.52 (2.19 ± 0.037),
9 1.95-2.57 (2.28 ±0.043); TL/EW: £ 2.27-2.45 (2.38 ± 0.008),
9 2.31-2.54 (2.44 ±0.016); PL/PW: S 0.82-0.90 (0.87 ±0.004),
9 0.80-0.89 (0.85 ±0.005); EL/EW: S 1.50-1.66 (1.59 ± 0.007),
9 1.57-1.74 (1.68 ±0.012); EL/PL: S 1.85-2.12 (2.01 ± 0.013),
9 2.06-2.27 (2.19 ±0.014); GD/EW: S 0.68-0.82 (0.75 ± 0.006),
9 0.6S-0.80 (0.76 ±0.006). Total size range in material examined: 1.80-2.85 mm.
Parafypes. ARIZONA: 46, Madera Can- yon, Santa Rita Mts., Santa Cmz Co., Jan. 30, 1964, Lot 1282 JFL, ex Trametes sepium [CAS, FMNH, JFL, MCZ, UAZ, USNM]; 3, Rustler Park, 8 mi. W Portal, Cochise Co., Aug. 8, 1961, Lot 918 JFL, ex Fames cajanderi [JFL]; 1, same locality and date. Lot 922 JFL, ex Fames cajanderi [JFL]; 5, same loeality, Apr. 9, 1964, Lot 1294 JFL (A. Raske, coll.), ex Fames cajanderi [JFL].
Distributian. Montane regions of .south- ern Arizona.
North American Ciidae • Lawrence 457
Host fungi. Fomes cajanderi [3(3)]; Tramctes sepium [1(1)].
Discussion. Individuals of this species are characterized by the large size; elongate body form; coarse and dense punotation, which is single and confused on the elytra; vestiture of flattened, color- less bristles; broad, crenulate, lateral pro- notal margins; and, elytra with a transverse impression. The two populations compris- ing the type series differ both in size and color pattern, but they are almost certainly conspecific. Two other specimens from the Chiricahua Mountains have been excluded from the type series although they may be- long to this species. In one the TL is less than the average for the Madera Canyon series and the elytra are yellow with two distinct black spots, while in the other the size is still smaller, the elytra are unifonnly brownish, and the body is more elongate than any specimen of C. discolor ( EL/EW = 1.79). Also in this complex are speci- mens from northwest of Gomez Farias, Tamaulipas, Mexico, in which the elytra are maculate, and a series from Desierto de los Leones, Mexico D. F., in which the size and color are comparable to that in the Rustler Park population, but the bristles are much flatter, blunt, and wedgelike. It is obvious that further collecting will be necessary to understand this species com- plex.
Cis discolor is most closely related to C. huachticae, described from the Huachuca Mountains and known also from Texas and northern Mexico, but the latter is shorter and broader, with dual vestiture, four sharp tubercles on the frontoclypeal ridge of the male, and no abdominal fovea.
The name discolor is derived from the Latin word meaning variegated.
Cis dunedinensis Leng
Cis dunedinensis Leng, 1918: 207, replacement
name for Cis pitsiUiis Dury, 1917 (not Gorham,
1898). Cis pusiUiis Dury, 1917: 10. Type locality:
"Dunedin, Fla." Lectotype, $, Dury Coll.,
CIN.
Distribution. Florida and Cuba. Marginal records: FLORIDA: Dunedin, Pinellas Co.; Crescent City, Putnam Co.; Paradise Key, Dade Co.; Biscayne, Dade Co. CUBA: Cay am as.
Host fungi. Unknown.
Discussion. This species may be dis- tinguished by the elongate form, dual and seriate elytral punctation, coarse and dense pronotal punctation, granulate and dull pronotal surface, dentate protibial apex, and vestiture of very short and fine hairs. Individvials of Cis niedhauki differ in hav- ing confused elytral punctation, finer and sparser pronotal punctation, expanded and rounded protibial apex, and the presence of tubercles on the vertex of the male. Specimens of various Ortlwcis may be confused with C. dunedinensis, but the protibial apex is narrowly rounded, the elytral suture is inf lexed apically ( Fig. 38 ) , the antennae may be 9-segmented, and the frontoclypeal region is simple in the male. All other Cis with a vestiture of short, fine hairs have a much shorter and broader body form.
Cis dunedinensis appears to have no close affinities with other North American species, and its relatives are to be sought among the West Indian and Central Amer- ican faunas.
Cis duplex Casey
Cis duplex Casey, 1898: 82. Type locality: "Cali- fornia." Holotype, $, Casey Coll., USNM.
Distribution. Mountains of the south- western United States and Mexico, from southern California east to north-central New Mexico and south as far as Morelos and the southern tip of Baja California (Fig. 107). Marginal records: CALIFOR- NIA: 1.5 mi. NW Mt. Laguna, San Diego Co.; ARIZONA: Williams, Coconino Co.; NEW MEXICO: El Porvenir, San Miguel Co.; MORELOS: Cuautla; DURANGO: 37 mi. W El Salto; BAJA CALIFORNIA SUR. La Laguna, Sierra Laguna.
Host fungi. Polyporus anceps [6(5)];
458 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
Fomes officinalis [2(2)]; Fomes pinicola [2(2)]; Ganoderma oregonense [1(1)].
Discussion. This species may be dis- tinguished by the moderately short and broad body form; single and confused elytral punctation; weakly carinate pro- sternum; narrow, crenulate lateral pronotal margins; coarse and dense pronotal punc- tation; and vestiture of short, stout bristles. The male bears two subtriangular plates on the frontoclypeal ridge and two more on the anterior edge of the pronotvun. Cis cmtlei is similar in most of these characters, but is smaller in size (TL less than 1.4 mm), somewhat shorter and broader, with shinier pronotal surface, finer and sparser pronotal punctation, smooth lateral pro- notal margins, and uniquely curved fronto- clypeal plates in the male (Fig. 7). The species may be confused with Flesiocis cribrum, which is usually larger, with 9- segmented antennae and 4 teeth on the frontoclypeal ridge of the male.
Cis duplex is a distinct species that can- not be placed in any group at this time. It occurs at higher elevations in various mountain chains from southern California and New Mexico to central Mexico. It is associated with fungi on conifers and ap- pears to prefer Polyponis anceps. Speci- mens collected at the southern tip of Baja California are from the Sierra Laguna, where Finns cemhroides is abundant, ac- cording to Goldman (1916). C. duplex has been collected with Flesiocis cribrum in southern California and New Mexico and appears to replace the latter in southern Arizona.
Cis ephippiatus Mannerheim
Cis ephippiatus Mannerheim, 1853: 234. Type locality: "... insula Sitkha." Types, Manner- heim Coll., MZUH ? (not located).
Cis ephippium Dalla Torre, 1911: 10. Incorrect subsequent spelling.
Xestocis moznettei Dury, 1917: 16. Type locality: "Corvallis, Oregon." Svntvpes, S S $ $ , Dury Coll., CIN. NEW SYNONYMY.
Distribution. Western North America, from southern Alaska south in California to
Alameda Comity and the southern Sierra Nevada, and in the Rocky Mountain Region to northern Nevada and Colorado. Also known from Vermont, New Hampshire, and the Gaspe Penninsula of Quebec (Fig. 90). Marginal records: ALASKA: Sitka. BRITISH COLUMBIA: 5 mi. NE Field. WYOMING: Jenny Lake, 6800^ Grand Teton Nat. Park, Teton Co. COLORADO: 28 mi. NW Kremmling, Grand Co. NEVADA: 8 mi. SE Lamoille, Elko Co. CALIFORNIA: Huntington Lake, Fresno Co.; Berkeley, Alameda Co. NEW HAMP- SHIRE: 1 mi. W Mt. Lafayette, Grafton Co. VERMONT: Big Equinox Mta., Ben- nington Co. QUEBEC: Gaspe Co.
Host fungi. Ganoderma applanatum [10(8)]; Fomes pinicola [7(4)]; Gano- derma brownii [2(2)]; Fomes annosus [1]; Folijporiis betidinus [1]; Folyponis resino- sus [1]; Folyponis sulphureus [1]; Foria versipora [1].
Discu.ssion. This species may be sepa- rated from most North American Cis by the short, stout body fonn, dual and con- fused or subseriate elytral punctation, dis- tinctly carinate prostemum, and vestiture of moderately long, decumbent hairs. It differs from the closely related C. biar- mattis in the finer and sparser pronotal punctation, smaller size, dark transverse macula on the elytra, and weakly produced pronotal apex in the male. C. ephippiatus was described by Mannerheim from Sitka, Alaska, and although the type has not been seen, it is fairly apparent from the de- scription that it is the same species that Dury named Xestocis moznettei on the basis of Oregon specimens.
Like Cis biarmatus, C. ephippiutus is distributed along the Pacific Coast, but the latter also occurs in the Rocky Moun- tains and in the northeastern part of the continent (Figs. 90 and 102). Where the two species occur together, the fonner is found in conks of Fomes pinicola, while the latter is usual]\' taken in Ganoderma applanatum. Outside of the range of Cis
North American Ciidae • Lawrence
459
hiannatus, however, C. cphippiatus com- monly inhabits Fomes pinicola.
Cis festivuius NEW SPECIES
Hohtijpe. S, DISTRICT OF COLUM- BIA: Wash[in]gt[o]n, 15.8, Coll. Hubbard & Schwarz [USNM]. AlloUioe, 9, MARY- LAND: Blad[e]ns-b[ur]g, 13.7, Coll. Hub- bard & Schwarz [USNM].
Male. Length 1.62 mm. Body 2.41 X as long as broad, strongly convex. Head and pronotum dark reddish brown. Elytra dark brown. Vestiture of short, stout, blunt, yellowish bristles, which vary in length. Vertex slightly convex; frontoclypeal ridge bearing 2 blunt, setiferous tubercles, which are separated by 1 basal width. Antennal segment III 1.40 X as long as IV. Pronotum 0.84 X as long as broad, widest at posterior third; anterior edge strongly rounded, simple; sides weakly rounded, the margins narrow and weakly crenulate, barely visible for their entire lengths from above; anterior angles not produced fonvard, almost right; disc strongly convex, e\'en; surface finely granulate to smooth and shiny; punctures 0.22 X as large as scutellar base and sepa- rated by 0.33 to 0.50 diameter. Elytra 1.63 X as long as broad and 2.09 X as long as pronotum; sides subparallel, apices blunt; punctuation single and unifonn; punctures coarser and denser than those on pronotinn, each bearing a stout, blunt, yellowish bristle, which may be from 0.25 to 0.33 X as long as scutellar base. Prostemum slightly tumid; intercoxal process 0.30 X as wide as procoxal cavity, subparallel. Pro- tibial apex narrowly rounded, with slight indication of outer tooth (Fig. 47). Meta- stemum 0.50 X as long as wide; suture 0.52 X as long as median length of sternum. Abdominal stemite III bearing a median, circular, pubescent fovea, which is 0.42 X as long as body of stemite, indistinctly margined, and located anterad of center. Stemite VIII as in Figure 65. Aedeagus as in Figures 72 and 84.
Female. Length 1.52 mm. Body 2.44 X as long as broad. Vertex as in male;
frontoclypeal ridge simple. Pronotum 0.87 X as long as broad; anterior edge as in male. EMra 1.64 X as long as broad and 2.05 X as long as pronotum. Protibia as in male. Stemite III without pubescent fovea.
Variation. Pronotum Ncllowish orange to dark brown, usually dark reddish brown. Elytra yellowish to dark brown, usually dark brown or yellowish brown. Size and dimensions vary as follows in a mixed series of three males and seven females:
(1.55 ±0.047), 0.058); 2.45 ±0.032),
0.015);
(0.84 ±0.011), 0.007); (1.67 ±0.025),
TL mm: 6 1.50-1.62
9 1.42-1.87 (1.57 TL EW: S 2.41-2.50
9 2.33-2.44 (2.39 PL PVV: i 0.83-0.86
5 0.82-0.87 (0.86± EL/EW: S 1.63-1.70
91.59-1.65 (1.62 ±0.009); EL PL: 6 2.09-2.21 (2.15 ± 0.043),
9 2.00-2.17 (2.09 ±0.027); GD/EW: $ 0.79-0.81 (0.80 ± 0.007),
9 0.79-0.84 (0.81 ±0.008).
Paratype. MARYLAND: 4, Blad[e]ns- b[ur]g, 13.7, Coll. Hubbard & Schwarz [JFL,''USNM]; 1, same locality, 20.7, Coll. Hubbard & Schwarz [USNM]; 1, Plummers I[sland], 15.7.07, E. A. Schwarz Collector [USNM]; 2, Sparrows Pt., VII-4-36, J. W. Green [CAS, JFL]. VIRGINIA: 1, Pen- nington Gap, 30.6 [USNM].
Distribution. Known only from Mary- land, Virginia, and the District of Colum- bia.
Host fungi. Unknown.
Discu.ssion. This species may be dis- tinguished by the long and narrow, cylindri- cal body form, single and confused elytral punctation, vestiture of short, stout, yel- lowish bristles vaiying somewhat in length, and protibial apex narrowly rounded with only a slight indication of an outer tooth. In Cis stereophilus the elytral bristles are subseriate and more uniform in length, and the outer apical angle of the protibia is dis- tinctly produced and angulate (Fig. 53). Cis robiniophihis differs from C. festivuius
460 Bulletin Muaeuni of Comparative Zoology, Vol. 142, No. 5
in the somewhat finer punctures and dis- tinctly dentate protibial apex (Fig. 49). In Cis hystriculus and C. angiistiis the bristles are longer and finer and the protibial apex is dentate or angulate.
Cis festiviihis appears to be most closely related to the Palaearctic Cis festious (Panzer) and its allies C. vestittis Mellie and C. pyg,maeus (Marsham).
The name festiviiliis is derived from the Latin festivus, meaning joyous or merry, and the Latin diminutive suffix -ulus (re- fering to the similarity to Cis festivus and the smaller size).
Cis florid a e Dury
Cis floridae Dury, 1917: 9. Type locality: "Key West, Florida." Lectotype, $, Dury Coll., CIN.
Distiihtition. Southern Georgia, Florida, and Cuba. Marginal records: GEORGIA: Savannah, Chatham Co. FLORIDA: Cres- cent City, Putnam Co.; Lignum Vitae Key, Monroe Co.; Key West, Monroe Co. CUBA: Cayamas.
Host fungi. Poly poms gilvus [1]; Stereum papyrinum [2].
Discussion. This species is characterized by the moderately short and broad, parallel- sided body fonn, coarse and dense punota- tion that is subseriate on the elytra, slightly tumid prosternum, dentate protibial apex, pronotum with distinct side margins and projecting anterior angles, and vestiture indistinctly dual, consisting of longer and shorter pale yellow bristles. The male bears two frontoclypeal plates, two pronotal horns, and an abdominal fovea that is located at the posterior end of sternite III. Cis huachucae and C. comiitus have dual vestiture, but both are larger (TL more than 2 mm) and somewhat shorter and broader (with the sides more rounded); in the former species the male lacks an abdominal fovea, while in the latter, the fovea is much smaller and more centrally located than in C. floridae. In Cis crinitus the form is similar but the vestiture is more distinctly dual and confused. Cis quadri-
dentatus and C. castlei are also somewhat similar in general fonn, but the vestiture is single (consisting of short bristles), the prosternum is carinate, the elytral puncta- tion is confused, and the abdominal fovea in the male is relatively smaller and more centrally located; C. castlei is also smaller (TL less than 1.4 mm) and the male of C. quodridentatus bears four teeth on the frontoclypeal ridge.
This is another West Indian species that does not extend further north than Florida and southern Georgia. It is not obviously related to any other species or group. There are too few records to speculate on host preference.
Cis fuscipes Mellie
Cis fuscipes Mellie, 1848: 271; Lawrence, 1967a: 1-14 (syn., dist., biol. ). See Lawrence (1967a) for complete synonymy.
Distrihiition. Widespread and abundant throughout most of northern and eastern North America, ranging from northern British Columbia, south to Los Angeles County on the Pacific Coast, cast across Canada to Nova Scotia, and south through- out eastern and midwestern United States (east of the 100th meridian) to southern Texas and Florida. Also known from Cuba, Madeira, and Hawaii. Marginal records in North America: BRITISH COLUMBIA: Terrace; NORTHWEST TERRITORY: Fort Smith; MANITOBA: Lake Dauphin; QUEBEC: Duparquet; NOVA SCOTIA: Truro; FLORIDA: Dunedin, Pinellas Co.; TEXAS: Brownsville, Cameron Co.; CALI- FORNIA: Los Angeles Co. See Lawrence ( 1967a ) for further remarks on distri- bution.
Host fungi. Polyportis versicolor [91 (48)]; Polyporus hirsiitus [14(10)]; Len- zites heiulina [12(3)]; Polyporus puhescens [5(1)]; Polyporus conchifer [3(2)]; Gano- derrna hrotonii [1(1)]; Polyporus adustus [1(1)]; Polyporus squamosus [1(1)]; Dacdalea and)igua [1]; Fomcs fraxinophilus [1]; Fomcs pinicolor [1]; Ganoderrna ap- planatum [1]; Polyporus subectypns [1].
North American Ciidae • Lawrence 461
Discussion. This species is easily dis- tinguished from other North American forms by the dual and subseriate elytral punctation, vestiture of short, erect bristles, large and stout body form with a shortened pronotum, wide lateral pronotal margins with strongly produced anterior angles, and impressed pronotal disc in the male. Most other species with dual and seriate elytral punctation are smaller, longer, and narrower, with different sexual characters and a relatively longer pronotum. Cis cornelli has a very short and broad body form with carinate prostemum and two long lateral horns on the clypeus of the male. C. cornutus is somewhat similar to C. fuscipes but is shorter and broader with finer, yellowish elytral bristles that fall into two size classes.
The distribution and biology of Cis fuscipes have been discussed in an earlier paper (Lawrence, 1967a). This is a very widespread and common species that is usually found in the fruiting bodies of Polyponis versicolor and its relatives. It is most closely related to Cis seriatopilosus Motschulsky and its allies from Siberia and Japan. Parthenogenesis in Cis fuscipes is discussed in detail in the paper cited above.
Cis hirsutus Casey
Cis hirsitta Casey, 1898: 83. Type locality: "Florida (Lake Worth)." Holotype, S, Casey Coll., USNM.
Distribution. Florida, the Bahamas, and the Greater Antilles (Fig. 95). Marginal records: FLORIDA: Enterprise, Volusia Co.; BAHAMAS: Matthew Town, Great Inagua Is.; PUERTO RICO: Arecibo; HAITI: Port au Prince; JAMAICA: Kings- ton; CUBA: Baiios de San Vicente, Pinar del Rio.
Host fungi. Polyporus hijdnoides [11 (3)]; Fomes sclerodermeus [5(1)]; Gano- derma zonatum [4(3)]; Ganoderma sp. [3(3)]; Trometes corntgata [3(2)]; Pohj- porus fuhocinereus [2(1)]; Polyporus maximus [2]; Polyporus pinisitus [2]; Ganoderma lucidum [1(1)]; Polyporus
porgomenus [1(1)]; Polyporus sector [1 (1)]; Auricularia polytricha [1]; Daedalea amhigua [1]; Ganoderma opplanaturn [1]; Polyporus iodinus [1]; Polyporus rigidus [1]; Polyporus supinus [1].
Discussion. This species is fairly easy to recognize because of the short, broad form and vestiture of long and fine hairs, which are recurved at the apices. The male is characterized by two narrow, lateral horns on the frontoclypeal ridge and two approxi- mate, flattened horns (deeply emarginate median process) on the apex of the pro- notiun. It is similar in general foiTn and size to Cis cornutus, but in that species the vestiture is dual and seriate and consists of stouter bristles that are not recurved at the apices. In C. crinitus the vestiture is dual and the elytra are longer and nar- rower. Cis hirsutus differs from C. ro- tundulus in the larger size, more elongate metastemum, blunt elytral apices, and strongly tumid, but not carinate, pro- stemum. It differs from C. ursulinus in being somewhat larger, darker in color, with weakly crenulate lateral pronotal margins and a relatively larger abdominal fovea in the male.
Cis hirsutus belongs to a Neotropical species group that includes C. crinitus, C. rotundulus, C. ursulinus, C. melliei Coquerel, and a number of undescribed West Indian forms. The males of all of these species have two narrow, lateral, frontoclypeal horns and a median pronotal process that is usually emarginate (Fig. 41). It is quite possible that C. hirsutus is synonymous with C. hiiiellus Jacquelin du Val (1857), described from Cuba. The type of the latter, however, could not be located, and the description is not sufficient for identification.
The distribution pattern suggests that Cis hirsutus evolved in the Greater Antilles and spread onto the mainland in relatively recent times. In Florida, the beetle breeds in several diverse fungi, but the preferred hosts throughout the range are Polyporus
462 Bulletin Museum of Cumpamtive Zoology, Vol. 142, No. 5
hydnoides. Fames sclerodenueus, and Ganodenna zonatum.
Cis horridulus Casey
Cis horridula Casey, 1898: 81. Type locality:
"Pennsylvania (Westmoreland Co.)." Holotype,
S, Casey Coll., USNM. Cis mormonica Casey, 1898: 81. Type locality:
"Utah (southwestern)." Holot>'pe, $, Casey
Coll., USNM. NEW SYNONYMY.
Distribution. Northern and montane regions of North America, except Pacific Coast, ranging from south-central British Cohimbia to the Atlantic Coast, south in the Rocky Mountain Region as far as the Chiricahua Mountains of southern Arizona, and along the Appalachian chain as far as western North Carolina (Fig. 91). Marginal records: BRITISH COLUMBIA: Trinity Valley; ONTARIO: near Kenora; MAINE: Paris, Oxford Co.; NORTH CAROLINA: Highlands, Macon Co.; NEW MEXICO: El Porvenir, San Miguel Co.; ARIZONA: Rustler Park, 2 mi. W Portal, Cochise Co.; UTAH: southwestern.
Hosi fungi. Polyponis pargamenus [13 (6)]; Polyporus ahietinus [9(4)].
Discussion. This species may be dis- tinguished by the long and narrow, subcylindrical body fonn; distinctly dual vestiture consisting of longer and shorter, erect and suberect, colorless bristles; den- tate protibial apex; and, lack of an ab- dominal fovea in the male. Most other species with similar fonn have single vesti- ture, but if the bristles vary somewhat in size, they are much shorter than those of C. horridulus.
This is closely related to Cis hystriculus and to Cis punctulatus and its relatives in Eurasia. The Old World species most closely resembling C. horridulus is C. tomentosus Mellie, which is known from eastern Europe and the Caucasus region. Cis horridulus is distributed throughout most of the northern part of the continent (extending south at higher altitudes) where it breeds in the Iruiting bodies of Polyporus pargamenus and the related P.
ahietinus. It is replaced on the Pacific Coast by C. hystriculus {see discussion on p. 463).
Cis huachucae Dury
Cis luiacJiucae Dury, 1917: 8. Type locality: "Huacluica Mountains, Arizona. Miller Canyon." Holotype, $, Dury Coll., CIN.
Distribution. Southern Arizona, Texas, and northeastern Mexico. Marginal records: ARIZONA: Miller Canyon, Huachuca Mts., Cochise Co.; TEXAS: San Antonio, Bexar Co.; NUEVO LEON: Chipinciue Mesa, Monterrey.
Host fungi. Unknown.
Discussion. Individuals of this species are relatively large (more than 2 mm), short and broad, with dual vestiture, con- sisting of shorter and longer, colorless bristles, pronotuni with broad lateral mar- gins and produced anterior angles, and male with two pronotal horns, four fronto- clypeal teeth or tubercles, and no fovea on the abdomen. Cis floridae has similarly dual vestiture but is smaller and more parallel-sided, with subseriate, yellowish bristles, and with two frontoclypeal plates and an abdominal fovea in the male. In Cis cornutus, the elytral bristles are longer, finer, more distinctly dual, and yellowish, while the male possesses an abdominal fovea and subtriangular frontoclypeal plates. Cis quadridcntatus, C. vitulus, and C. congest us are all similar with respect to general form and sexual ornaments on the head and pronotum of the male; in the first the size is smaller, the vestiture single, the prosternum carinate, and the abdomen foveate in the male, while in the last two forms the \'estiture is single, the pro- sternum strongly tumid, and the body somewhat stouter and more convex.
Cis huachucae appears to be most closely related to Cis discolor, which differs in general form, being more elongate and parallel-sided, with vestiture that is not obviously dual. Tlie frontoclypeal ridge in the male of C. discolor bears four rounded tubercles, an abdominal fovea is present
North American Ciidae • Lawrence 463
but small, and the aedeagus (Figs. 68 and 81) is similar to that of C. liiiacJuicae. Both species are part of a complex of mide- scribed forms from the momitains of Mexico.
Cis hysfriculus Casey
Cis htjstricula Casey, 1898: 82; Hatch, 1962: 250 (tlist. )• Type locality: "California (Lake Tahoe)." Holotype, $, Casey Coll., USNM.
Distrihution. Western British Columbia, Washington, and Oregon, south through the Sierra Nevada and coastal California to the Transverse Ranges in the southern part of the state (Fig. 91). Marginal records: BRITISH COLUMBIA: Terrace; ORE- GON: Base of Mt. Pitt, Klamath Co.; CALIFORNIA: Ebbett's Pass, 873(r, Alpine Co.; 2 mi. NE Idyllwild, Riverside Co.
Host fungi. Pohjponis abietinus [27(16)]; Poria cinerascens [1]; Poria versipora [1].
Discussion. Tliis species is characterized by the long and narrow, subcylindrical body form; single, coarse and confused elytral punctation; vestitiu-e of short, color- less bristles; distinctly dentate pro tibia; and, lack of an abdominal fovea in the male. The similar C. angustus has yellowish bristles, a blunt and angulate protibial apex, and an abdominal fovea in the male. Cis horriduhis resembles this species, but differs in having longer elytral bristles that fall into two size classes (dual vestiture).
Cis Jujstricuhis is most closely related to Cis punctulatus Gyllenlial from northern and central Europe, and the two may rep- resent a single Holarctic species. The adults of C. punctulatus that I have ex- amined are practically indistinguishable from those of the North American species and the larva illustrated by Saalas (1923) has urogomphi similar to those of C. hijstricuhis larvae. In addition, both species breed in the same fungi {Polyporus abie- tinus). The situation becomes more compli- cated, however, if one considers the other species in the complex, such as Cis luir- ridulus in North America, C. tomentosus in
Europe and probably several more species from Europe and Asia.
Cis hijstricuhts is restricted to the Pacific Coast. It appears to be distinct from C. horridulus throughout most of its range, but a population from Riverside County, California (San Jacinto Mountains), ex- hibits slightly longer bristles and a more convex body form, approaching the con- dition of C. horridulus specimens from the mountains of southern Arizona. In southern British Columbia, however, the two species occur within 120 miles of one another and remain quite distinct. Larvae and male genitalia of the two are similar and the host fungi are the same, so that it is quite pos- sible that they represent geographic races. If so, a zone of iutergradation may be found in British Columbia in the vicinity of the Lillooet and Eraser Rivers.
C/'s krausi Dalla Torre
Cis krausi Dalla Torre, 1911: 13; Blatchley, 1928:
68 (dist. ). Replacement name for Cis hiiuacu-
latits Krans, 1908 (not Sharp, in Blackburn and
Sharp, 1885; not Germain, 1855). Cis bimacuJatiis Kraus, 1908: 76. Type locality:
"Victoria, Tex." Holotype, ?, USNM. Cis dunji Leng, 1918: 207. Replacement name for
Cis bicolor Dury, 1917 (not Shaip, 1879). NEW
SYNONYMY. Cis bicolor Dury, 1917: 7. Type locality: "Tybee
Island, Ga." Lectotype, c5 , Dury Coll., CIN.
Distribution. Georgia, Florida, and south- ern Texas. Marginal records: GEORGIA: Tybee Is., Chatham Co. FLORIDA: Dune- din, Pinellas Co.; Coconut Grove (Bis- cayne), Dade Co.; Key West, Monroe Co. TEXAS: Columbus, Colorado Co.; Victoria, Victoria Co.; Brownsville, Cameron Co.
Host fungi. Unknown.
Discussion. This species is fairly easily distinguished by the very long and narrow body form (EL/EW greater than 1.60), dual and subseriatc elytral punctation, pronotum \\ith narrow lateral margins and coarse, dense punctures, yellowish color with black markings, and vestiture of short, colorless bristles. The male bears two seti- ferous tubercles on the clypeus and an abdominal fovea that is located at the
464 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
anterior end of stemite III. Among North American species, Cis stereopliihis is similar in several respects, but in that species the form is not as elongate, the punctation is not dual, and the elytra, although bi- colored, are not maculate. The coloration in C. kraiisi is usually as follows: pronotum yellow with a large blackish patch anterad of center; elytra yellow with an anterior black triangle and two round, posterior, black spots.
This species apparently belongs to a group of West Indian forms, including Cis superhus Kraus (Cuba), C. atromociilatus Pic (Guadeloupe), and several unde- scribed. C. superhus is smaller than C. kmusi with coarser and denser pronotal punctation, more prominent frontoclypeal tubercles, and a more centrally located abdominal fovea in the male. C. atro- maculatus differs mainly in color pattern.
Cis J<rau9i has been collected beneath the bark of a grapefruit tree and has been beaten from oak limbs in Florida.
Cis laminatus Mellie
Cis laminatus Mellie, 1848: 318, pi. 11, fig. 16. Type locality: "Montrieux, pres de Toulon" [France]. Lectotype, S, Maiseul Coll., MNHN.
Distribution. In North America, known only from Albany, New York. In Europe, recorded from France, Italy, Gemiany, Poland, Czechoslovakia, and Hungary.
Host fungi. Recorded from Poh/poriis volvatus. In Europe the species has been collected in fungi growing on pine and spruce.
Discussion. This species may be dis- tinguished by the short and broad body form; coarse and dense punctation, which is single and confused on the elytra; vesti- ture of short bristles; relatively small scutel- lum; slighth' tumid prostenium; and, the simple pronotal apex and well-developed, trisinuate, clypeal plate in the male. Cis vitulus and C. congestus are similar, but both have broader lateral pronotal margins \\ith produced anterior angles, strongly tumid prostemum, and two pronotal horns
and no abdominal fovea in the male. Plesiocis cri])riim resembles this species and also feeds on Pohjporus volvatus, but the antennae in P. cribrum are 9-segmented and the clypcus of the male bears four sharp teeth.
Cw laminatus has been collected only once at Albany, New York, in 1920. It was probably introduced from Europe and may not be established in this country. The species appears to be most closely related to Cis fissicornis Mellie, which occurs throughout northern Eurasia.
Cis levettei (Casey), NEW COMBINATION
Xestocis levettei Casey, 1898: 85; Dury, 1917: 17 (dist.); Gibson, 1918: 113 (dist.); Weiss and West, 1920: 8 (dist, biol.); Hatch, 1924: 305 (dist, biol.); Peterson, 1957: 94-95, fig. C5-F, 192-193, fig. C54-G (larva); Hatch, 1962: 2.33 (dist.). Type locality: "... Indiana . . ."? Holotype, S, Casey Coll., USNM.
Xestocis levellei Dalla Torre, 1911: 20. Incorrect subsequent spelling.
Eridaitlus levettei (Casey), — Lawrence, 1965: 281; Pielou and Matthewman, 1966: 1310 (dist., biol.); Pielou and Verma, 1968: 1184 (dist, biol.).
Distribution. Widespread in North Amer- ica east of the 100th meridian, from New- foundland south to Alabama and west to Manitoba, Kansas, and Texas. In the west- em part of the continent, recorded from Alberta, nortlnvestem Colorado, eastern British Columbia and Washington, and California (Fig. 88). Marginal records: WASHINGTON: Palouse, Whitman Co.; BRITISH COLUMBIA: Trinity Valley; ALBERTA: Edmonton; MANITOBA: Aw- eme; NEWFOUNDLAND: Bay of Islands; SOUTH CAROLINA: Yemassee, Beaufort Co.; ALABAMA: Mobile, Mobile Co.; KANSAS: Salina, SaHne Co.; COLORADO: Steamboat Springs, R(nitt Co.; CALIFOR- NIA: (no specific locality). The California specimens, if they are not mislabeled, were probably collected in the northern part of the state or in the Sierra Nevada.
Ilo.st fungi. Ganoderma applanatum [67(36)]; Fomes fomentarius [17(11)];
North American Ciidae • Lawrence 465
Fomes pinicola [16(13)]; Ganoderma tsugae [9(4)]; Fomes connatiis [4(3)]; Poli/poms versicolor [3]; Ganoderma lucidum [2(1)]; Polyponis bettdinus [2(1)]; Polijporus pargamenus [2]; Pohjporus piibescens [2]; Steretim ostrea [2]; DaedaJea confragosa [1(1)]; Polyponis resinostis [1(1)]; Porta nigrescens [1(1)]; Fomes robinioe [ 1 ] ; Lenzites betiditm [1]; Polyponis adustus [1]; Pohjporus squamosus [1]; Poria vitrea
[1].
Discussion. This species differs from most North American Cis by the short and broad body fonn, distinctly carinate pro- sternum, dual and confused elytral punc- tation, vestiture of very short, fine hairs, and by the protibia (Fig. 54), which is irregularly serrate along the outer edge and bears a stout tooth at the apex. Cis maritimus and C. megastictiis are similar in most of the above characters, but the outer edge of the protibia is simple, the pronotal punctation is coarser and denser, the lateral edges of the pronotum are crenu- late, and the elytral punctation is seriate.
Cis levettei belongs to the Cis nitidus group, which has recently received some attention by European workers (Lohse, 1964; Strand, 1965) and includes the fol- lowing Palaearctic species: C. glabratus Mellie, C. hanseni Strand, C. jacqiiemarti Mellie, C. lineatocribratus Mellie, and C. nitidus (Fabricius). The serrations on the outer edge of the protibia are definitely present in C. glabratus and are weakly indicated in C jacqiiemarti.
This species is fairly common throughout the eastern part of the continent but has been collected only a few times in the Northwest. It breeds in several fungi but appears to prefer Ganoderma applanattim and is often found in association with Ceracis sallei and the tenebrionid beetle Bolitothenis cornutus ( Panzer ) ( Heatwole, 1968; Lawrence, 1967b; and Pace, 1967).
C(s marifimus (Hatch), NEW COMBINATION
Xestocis martimiis Hatch, 1962: 233. Type lo-
cality: "Ocean Park, Ore." Holotype, S, Hatch Coll., UW.
Distribution. Pacific Coast from extreme northwestern California to southwestern British Columbia. Also known from south- central Manitoba (Fig. 88). Marginal records: BRITISH COLUMBIA: Bowser; MANITOBA: Dauphin; OREGON: Blue River, Lane Co.; CALIFORNIA: Crescent City, Del Norte Co.
Host fungi. Polyporus .schweinitzii [2 (1)]; Fomes pini [1(1)].
Discussion. Cis maritimus differs from most other North American Cis in the short and broad body form, distinctly cari- nate prostemum, protibial apex with an outer tooth, dual elytral punctation, and subtriangular frontoclypcal plates in the male. It may be distinguished from related species in the Cis nitidus group (see p. 438) by the vestiture of very short, fine hairs, the finely granulate and shiny pronotum, which is coarsely and densely punctate, the pro- duced and acute anterior pronotal angles, and the longitudinally oval abdominal fovea in the male. It is most similar to C. levettei of eastern North America, but that species has much finer and sparser pro- notal punctation, duller pronotal surface, and serrate outer protibial edge.
Except for the single Manitoba record, the species is known only from the narrow coastal strip extending from northwestern California to southern British Columbia. It has been found in two species of fungi hav- ing reddish brown fruiting bodies. In all of North America, this is the only ciid species with a northern distribution that breeds in this type of sporophore; other Ciidae inhabiting these fungi are southern species with Neotropical affinities.
Cis megasfictus NEW SPECIES Figure 39
Holotype. $, CALIFORNIA: Bucks Lake, Plumas Co., July 26, 1964, Lot 1307 J. F. Lawrence (J. Doyen, coll.), ex Fomes
466
Bulletin Museum of Coiupaiative Zoology, Vol. 142, No. 5
annosus on Abies concolor [CAS]. Allotype, 2, same data [CAS].
Male. Length 1.70 mm. Body 1.94 X as long as broad, strongly convex. Head and pronotimi reddish bro\\ai, elytra yellowish brown. Vestiture of very short and fine yellowish hairs. Vertex with a slight median impression; frontoclypeal ridge bearing two broad, snbtriangular plates, which are separated by 0.80 basal width. Antennal segment III 1.60 X as long as IV. Pronotnm 0.75 X as long as broad, widest at posterior two-fifths; anterior edge strongly ronnded, barely flattened at mid- dle; sides weakly rounded, the margins broad and weakly crenulate, easily visible for their entire lengths from above; an- terior angles distinctly produced forsvard, rounded; disc strongly convex, with a nar- row, median furrow extending from poste- rior edge to anterior fifth; surface smooth and shiny; punctures 0.25 X as large as scutellar base and separated by 0.20 to 0.25 diameter. Elytra 1.26 X as long as broad and 1.83 X as long as pronotnm; sides moderately rounded, apices subacute; punctation dual and seriate; megapunc- tures much coarser than those on pronotnm, very dense and almost confluent within rows, shallow and nude; micropunctures located within and between rows, each bearing a fine, yellowish hair, which is about 0.17 X as long as scutellar base. Prosternum strongly tumid and carinate; intercoxal process 0.38 X as wide as a procoxal cavity, narrowing posteriorly. Protibia with outer apical angle produced and dentate (Fig. 55). Metasternum 0.50 X as long as wide; suture 0.28 X as long as median length of sternmri. Abdominal sternite III bearing a median, circular, pubescent fovea, which is 0.43 X as long as body of sternite, indistinctly margined, and located about in center. Sternite VIII as in Figure 64. Aedeagus as in Figures 74 and 78.
Female. Length 1.92 mm. Body 2.03 X as long as broad. Vertex slightly convex; frontoclypeal ridge simple. Pronotnm 0.77
X as long as broad; anterior edge strongly roimdcd, simple. Elytra 1.32 X as long as broad and 1.85 X as long as pronotnm. Protibia as in male. Sternite III without pubescent fovea.
Variation. Pronotnm yellowish orange to dark reddish brown, usuallv reddish brown. Elytra yellowish to dark brownish, usually yellowish brown. Frontoclypeal plates in smaller males shorter and more rounded. Size and dimensions vary as follows in a mixed series of six males and thirteen fe- males from California:
TL mm: S 1.70-2.07
2 1.72-2.10 (1.94± TL EW: c^ 1.89-2.06
9 1.97-2.03 (2.00± PL/PW: 6 0.71-0.81
9 0.76-0.84 (0.80 ± EL EW: 6 1.24-1.36
9 1.26-1.37 (1.29± EL PL: S 1.76-1.92 i
9 1.69-1.89 (1.78± GD/EW: $ 0.78-0.82
9 0.79-0.83 (0.81 ±
(1.80 + 0.057), 0.030);
(1.96 ±0.028), 0.006);
(0.76 ±0.014), 0.007); (1.30 ±0.018),
0.011);
:i.83± 0.021), 0.018);
(0.79 ±0.007), 0.004).
Paratypes. CALIFOBNIA: 4, (no spe- cific locality) [MCZ]; 8, Bucks Lake, Plumas Co., July 26, 1964, Lot 1307 JFL (J. Doyen, coll.), ex Fames annosus on Abies concolor [CAS, JFL, USNM]; 1, same lo- cahty and date. Lot 1305 JFL (J. Doyen, coll.), ex Polyponis sulphureus [JFL]; 4, Calaveras [CIN]; 1, Plaskett Meadows, 6200', Glenn Co., July 3, 1960, Lot 631 JFL, ex Fames pinicolor [JFL].
Distribution. Known only from montane regions in northern California.
Host fungi. Fames annosus [1(1)]; Fames pinicola [1]; Polijparus sulphureus
[1].
Discussion. This species is distinguished
by the short and broad body form, carinate prosternum, distinctly dual and seriate elytral punctation, vestiture of short, yel- lowish hairs, broad lateral pronotal margins with produced and rounded anterior angles, ctnuse and dense pronotal punctation, and dentate protibial apex. Cis cornelli, C.
North American Ciidae • Lawrence 467
americanus, and C. tridentatiis all differ in the vestiture of short, stout bristles. In C. levettei, the pronotal punctation is much finer and sparser, the lateral pronotal mar- gins are narrower, and the outer edge of the protibia is serrate (Fig. 54). In C. maritimiis, the anterior pronotal angles are acute, the elytral megapunctures are not as large, and the abdominal fovea in the male is oval, rather than circular.
This is another of the localized Pacific species in tlie Cis nitidus group ( the others being C. tridentatus, C. hiarmatus, and C. maritimiis). It occurs at higher elevations in the conifer forests of northern California but is apparently absent from tlie immedi- ate coast.
The name megastictiis is derived from the Greek meg,as, meaning large, and the Greek stiktos, meaning punctured (refer- ring to the size of the elytral megapunc- tures ) .
Cis miles (Casey), NEW COMBINATION
Xestocis miles Casey, 1898: 85; Blatcliley, 1928: 68 (dist., biol.); Lawrence, 1967b: 98. Type locality: "Pennsylvania (Westmoreland Co.)." [St. Vincent]. Holotype, $, Casey Coll., USNM.
Distribution. Eastern North America, from New York south to central Florida and west to Arkansas and Louisiana. Mar- ginal records: NEW YORK: Pompey, Onondaga Co.; FLORIDA: Highlands Hammock State Park, Highlands Co.; LOUISIANA: (no specific locality); AR- KANSAS: Washington Co.
Host fungi. Pohjporiis versicolor [7(1)]; Lenzitcs hetulimi [3(2)]; Pohjporus siih- ectijpus [1(1)]; Polijporiis supinus [1]; Stereiim ostrea [1].
Discussion. This species is characterized by the small size (TL usvially less than 1.4 mm), short and broad body fonn, carinate presternum, very fine and sparse pmicta- tion, vestiture of very short and fine hairs, and unique male armature consisting of two lateral bonis on the pronotal apex and a single, median, forked horn on the fronto- clypeal ridge (Fig. 6). Cis levettei and C.
maritimiis are both larger, with coarser and denser elytral punctation and different sexual characters.
Cis miles is the only North American member of a Neotropical species group which includes C. tricornis Gorham, C. delicatulus ( Jacquelin du Val), and a num- ber of undescribed forms. This species, like most other members of the group, feeds primarily on Polyporiis versicolor and its relatives.
Cis niedhouki NEW SPECIES
Holotype. £, FLORIDA: Lignum Vitae Key, Monroe Co., May 28, 1968, Lot 2577 J. F. Lawrence, ex Fomes robiniae [MCZ No. 31690]. Allotype, 9, same data [MCZ].
Male. Length 1.30 mm. Body 2.17 X as long as broad, moderately convex. Head reddish brown, pronotum and elytra black- ish. Vestiture of very short and fine, pale hairs. Vertex with a deep circular median impression, a sharp conical posteromedian tubercle, and two diverging, anterolateral blunt boms, which are about 2.0 X as long as wide at base, 0.22 x as long as pronotum and separated by 3.5 basal widths; fronto- clypeal ridge bearing two tubercles on each side, the lateral t\vo smaller and rounded, the mesal two larger, conical, and sepa- rated by 3.0 basal widths (Fig. 5). Antennal segment III 1.70 X as long as IV. Pronotum 0.78 X as long as broad, widest at anterior two-fifths; anterior edge produced and emarginate, forming two small approximate tubercles; sides moderately rounded, the margins narrow and weakly crenulatc, not visible for their entire lengths from above; anterior angles not produced forward, al- most right; disc moderately convex, slightly impressed anteriorly; surface granulate and slightly shiny; punctures 0.20 X as large as scutellar base and separated by 0.75 to 1.25 diameters. Elytra 1.42 X as long as broad and 1.89 X as long as pronotum; sides weakly rounded, apices blunt; punc- tation dual and confused; megapunctures coarser and denser than those on pronotum; each micropuncture bearing a very fine.
468 Bulletin Museum of Comparative Zoology, Vol 142, No. 5
erect pale hair, which is about 0.10 X as long as scutellar base. Prostemum moder- ately tumid and subcarinate; intercoxal process 0.37 X as wide as a procoxal cavity, narrowed posteriorly. Protibia with outer apical angle expanded and rounded. Meta- stemum 0.58 X as long as wide; suture 0.23 X as long as median length of sternum. Abdominal stemite III bearing a median, circular, pubescent fovea which is 0.25 X as long as body of stemite, indistincdy margined, and located slightly anterad of center.
Female. Length 1.15 mm. Body 2.09 X as long as broad. Vertex slightly convex; frontoclypeal ridge simple. Pronotum 0.79 X as long as broad; anterior edge moder- ately rounded. Elytra 1.41 X as long as broad and 2.07 X as long as pronotum. Protibia as in male. Stemite III without a pubescent fovea.
Variation. Pronotirm yellowish orange to black, usually dark reddish browai or black. Elytra yellowish to black, usually black, occasionally somewhat reddish posteriorly. Median tubercle of vertex and frontoclypeal tubercles in smaller males obscure or absent and lateral horns short and rounded; in larger males the lateral horns are strongly diverging and may be 0.25 X as long as pronotum. Size and dimensions vary as follows in a series of 25 males and 22 fe- males from Lignum Vitae Key, Monroe Co., Florida (Lots 2547, 2577, 2601, 2622):
TL mm: $ 1.00-1.35 ( 1.15 ± 0.016),
9 0.97-1.32 (1.18 ±0.018); TL/EW: i 2.09-2.26 (2.17 ± 0.010),
9 2.04-2.30 (2.15 ±0.013); PL/PW: $ 0.75-0.83 (0.79 ±0.004),
5 0.76-0.87 (0.80 ±0.005); EL/EW: $ 1.35-1.53 (1.42 ± 0.008),
9 1.35-1.53 (1.43 ±0.009); EL/PL: S 1.81-2.08 ( 1.90 ± 0.014),
9 1.88-2.13 (2.00 ±0.015); GD/EW: S 0.71-0.81 (0.77 ± 0.004),
9 0.73-0.83 (0.77 ±0.006).
Paratypes. FLORIDA: 3, Lignum Vitae Key, Monroe Co., Mar. 15, 1968, Lot 2547
JFL (S. B. Peck, coll.), ex Fames robiniae [MCZ]; 19, same locaHty, May 28, 1968, Lot 2577 JFL, ex Fomes robiniae [FMNH, JFL, USNM]; 14, same locality, June 1, 1968, Lot 2601 JFL, ex Fomes robiniae [JFL, MCZ]; 9, same locahty, June 5, 1968, Lot 2622 JFL, ex Fomes robiniae [JFL, MCZ].
Distribution. Known only from Lignum Vitae Key, Florida.
Hosi fungi. Fomes robiniae [4(2)].
Discussion. This species, like C. caijensis, is unique in having the protibial apex ex- panded and rounded (but not spinose), and the head of the male bearing annature on the vertex as well as on the frontoclypeal ridge. Cis caijensis males lack the median tubercle and the lateral tubercles on the vertex are not as long; in addition the vesti- ture is entirely different. The elongate body form, dual elytral punctation, and vestitm-e of short, fine hairs distinguish this species from all North American Cis, with the exception of C. dunedinensis, in which the elytral punctation is seriate, the pro- notum more parallel-sided with coarser and denser punctation, and the head of the male with two frontoclypeal teeth only. Smaller specimens of Orthocis species may resemble C. niedhauki, but they will usu- ally differ in the narrowly rounded protibial apex (Fig. 45), the nature of the elytral apices (Fig. 38), the lack of head orna- ments in the male, and often in the smaller size, maculate elytra, and 9-segmented an- tennae.
Cis niedhauki is most closely related to C. caijensis, also knowoi from the Florida Keys. It has been collected only on Lignum Vitae Key but probably occurs in the Greater Antilles. A single specimen from Cayamas, Cuba, may belong to this species, but the pronotum is somewhat shinier with coarser and denser punctation, and the color is reddish. Further specimens must be examined. Fomes robiniae is the only known host, but the species may occur on other melanic conks.
The species has been named in honor of
North American Ciidae • Laiorence 469
Russell and Charlotte Niedhauk, caretakers of Lignum Vitae Key, to whom I am thank- ful for their generous hospitality.
Cis pistoria Casey
Cis pistoria Casey, 1898: 79; Gibson, 1915: 137 (dist.); Hatch, 1924: 305 (dist.)- Type lo- cality: "Rhode Island (Boston Neck)." Holo- type, c5 , Casey Coll., USNM.
Distribution. Northeastern North Amer- ica from central Alberta to southern Minne- sota and southern New England (Fig. 104). Marginal records: ALBERTA: Lake George, near Busby; NORTHWEST TERRITORY: Fort Smith; MANITOBA: Dauphin Lake; QUEBEC: Duparquet; MAINE: Weld, Franklin Co.; RHODE ISLAND: Boston Neck, Newport Co.; MINNESOTA: Cedar Creek Forest, Anoka Co.
Host fungi. Polyporus versicolor [7(4)]; Polyporus puhescens [2(2)]; Pohjporus adustiis [1(1)]; Polyporus hirsutus [1(1)].
Discussion. Individuals of this species are relatively large (usually more than 2.2 mm ) with dual and confused elytral punc- tation, very short, scalelike bristles and slightly tumid prosternum. Most fonns with dual punctation are narrower and more elongate and do not have the broad lateral pronotal margins and somewhat uneven pronotal disc characteristic of this species. In Cis americanus and C. triden- tatus, the size is smaller, the fonn more oval, the prosternum carinate, and ^he bristles longer. Cis fuscipes and C. tetra- centrum are comparable in size and general form, but in the former the elytral punc- tation is subseriate, in the latter it is not dual, and in both the vestiture is longer.
Cis pistoria is the only New World mem- ber of a Palaearctic group, which includes Cis boleti (Scopoli), C. rugulosus Mellie, C. micans (Fabricius), C. hispidus Gyllen- hal, and C. villosidus (Marsham). Like most of its Old World relatives, the North American species occurs primarily on Polyporus versicolor and its relatives. In the Northeast it is usually found in associ-
ation with Cis fuscipes and Octotemyius laevis.
Cis quadridentatus (Dury), NEW COMBINATION
Xestocis quadridentatus Dury, 1917: 17. Type
locaUty: "Framingham, Mass." Syntypes, Duty
Coll., CIN. Cis blatchleiji Dury, 1917: 7. Type locahty:
"Dunedin, Fla." Lectotype, $, Blatchley Coll.,
PURD. NEW SYNONYMY.
Distribution. Eastern North America from \^ermont to Florida and west as far as south-central Texas. Marginal records: VERMONT: Pawlet, Rutland Co.; FLOR- IDA: Dunedin, Pinellas Co.; TEXAS: San Antonio, Bexar Co.; ILLINOIS: Sayer Bog, Volo, Lake Co.
Host fungi. Unknown.
Discussion. This species may be dis- tinguished by the short and broad body form; carinate prosternum; shiny surface; coarse and dense punctation, which is single and confused on the elytra; vestiture of short bristles; moderately broad, crenu- late, lateral pronotal margins; and, male with two pronotal horns, four frontoclypeal teeth, and an abdominal fovea. Cis vitulus and C. congestus are larger in size (TL usually more than 2.2 mm) with a raised lip on the lateral pronotal margins, the prosternum not carinate, and the male without an abdominal fovea. In C. castlei and C. duplex, the lateral pronotal margins are narrower, without produced anterior angles, and the clypeus of the male bears two subtriangular plates. Plesiocis cribrum is similar with respect to general form, vestiture, punctation, and male armature, but in that species the antennae are 9-seg- mented and the pronotal margins are much narrower.
This species does not appear to be closely related to any other New World species that I have studied, but it closely resembles Cis indicus Pic, Cis subsqua- mosus Scott, and several undescribed forms from the Oriental Region.
Cis quadridentatus has been collected from fungus fruiting bodies and has been
470 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
found under bark on several occasions, ]:)ut it has not yet been associated with a par- ticular fungus species.
Os robiniophilus NEW SPECIES
Holotype. S, OHIO: Preston, Hamilton Co., Lot 1960 JFL (A. P. Morgan, coll.), ex Poh/ponis rohiniophilus (herbarium specimen, C. G. Lloyd No. 41039) [USNM]. Allotype, 9, same data [USNM].
Male. 1.70 mm. Body 2.43 x as long as broad, moderately convex. Head and pronotum reddish brovvm, elytra yellowish brown. Vestiture of short, stout, blunt, yellowish bristles. Vertex slightly concave; frontoclypeal ridge bearing two blunt seti- ferous tubercles that are separated by 2.5 basal widths. Antennal segment III 1.50 X as long as IV. Pronotum 0.88 X as long as broad, widest at middle; anterior edge strongly romided, simple; sides strongly rounded, the margins narrow and distinctly crenulate, not visible for their entire lengths from above; anterior angles not produced forward, almost right; disc moderately convex, even; surface finely granulate and somewhat shiny; punctures 0.17 X as large as scutellar base and separated by 0.50 to 0.75 diameter. Elytra 1.61 X as long as broad and 1.96 X as long as pronotum; sides subparallel, apices blunt; punctation single and uniform; punctures slightly coarser and denser than those on pronotum, each bearing a stout, blunt, yellowish bristle, which is about 0.33 X as long as scutellar base. Prosternum slightly tumid; intercoxal process 0.30 X as wide as a procoxal cavity, parallel-sided. Protibia with outer apical angle strongly produced and dentate (Fig. 49). Metastemum 0.60 X as long as wide; suture 0.39 X as long as median length of sternum. Abdominal stemite III bearing a median, oval, pubes- cent fovea, which is 1.33 X as long as wide, 0.25 X as long as body of stemite, distinctly margined, and located anterad of center. Stemite VIII as in Figure 61. Aedeagus as in Figures 71 and 83.
Female. Length 1.67 mm. Body 2.48 X
as long as broad. Vertex flattened; fronto- clypeal ridge simple. Pronotum 0.88 X as long as broad; anterior edge as in male. Elytra 1.67 X as long as broad and 2.04 X as long as pronotum. Protibia with outer apical angle weakly dentate. Stemite III without pubescent fovea.
Variation. Pronotimi yellowish orange to dark reddish brown, usually reddish brown. Elytra yellowish to yellowish brown, usu- ally yellowish brown. Apex of pronotum in larger males weakly emarginate at midline. Size and dimensions vary as follows in a series of 13 males and 13 females from Preston, Ohio (Lot 1960 JFL):
TL mm: 6 1.50-1.75(1.64 + 0.021),
5 1.50-2.07 (1.73 ±0.051); TL/EW: S 2.37-2.52 (2.45 ± 0.011),
9 2.44-2.61 (2.51 ±0.012); PL/PW: S 0.84-0.92 (0.88 ± 0.008),
9 0.84-0.89 (0.87 ±0.004); EL EW: S 1.59-1.67 ( 1.62 ± 0.007),
9 1.65-1.78 (1.70 ±0.010); EL/PL: S 1.91-2.05 (1.97 ± 0.014),
9 2.00-2.16 (2.09 ±0.017); GD/EW: 6 0.73-0.79 (0.76 ± 0.005),
9 0.76-0.82 (0.78 ±0.005). Total size range in material examined: 1.30-2.10 mm.
Parati/pe.s. KENTUCKY: 3, Crittenden, Grant Co., Aug. 3, 1907, Lot 1962 JFL (C. G. Lloyd, coll.), ex Polyporus rohiniophi- lus (herbarium specimen, C. G. Lloyd No. 41030) [JFL]; MARYLAND: 3, Plummers Island, Montgomery Co., April 29, 1905, Lot 1961 JFL (H. S. Barber, coll.), ex Polyporus rohiniophilus (herbarium specimen, C. G. Lloyd No. 39777) [JFL, USNM]; OHIO: 65, Preston, Hamilton Co., Lot 1960 JFL (A. P. Morgan, coll.), ex Polyporus rohinio- philus (herbarium specimen, C. G. Lloyd No. 41039 [FMNH, JFL, MCZ, USNM].
Distrihution. Known only from Mary- land, Kentucky, and Ohio.
Host fungi. Polyporus rohiniophilus
[3(1)].
Discussion. This species is characterized by the long and narrow body form; vesti-
North American CiiDAE • Lawrence 471
ture of short, stout bristles; single and con- fused elytral punctation, which is relatively fine and sparse; narrow, crenulate, lateral pronotal margins; and, distinctly dentate protibial apex. In Cis festivuhis, the protib- ial apex is rounded and the elytral bristles are uneven in length. C. stereophiJtis is some- what shorter and broader with subseriate elytral punctation. In Cis cre]}errimiis the body is more flattened and the elytral bris- tles longer and acute. Individuals of Cis htjstricuhis and C. angiistus have coarser and denser punctation and longer bristles. In Dolichocis imlistinctus, the body is long and narrow and the bristles short and stout, but in that species the antennae are 9-segmented and the protibial apex rounded.
Cis robiniophilus is a member of the Cis fagi group, which includes the North American C. angiistus and the Palaearctic Cis fagi Waltl and C. castaneus Mellie.
The name rohiniophiliis is derived from Robinia, the generic name for certain of the locust trees, and the Greek phileo, to be fond of (referring to the occurrence of this beetle in fruiting bodies of Pohjporus rohiniophiliis) .
Cis rotundulus NEW SPECIES Figure 41
Holotype. S, SOUTH CAROLINA: Walterboro, Colleton Co., 11-15-64, H. Blocker [MCZ No. 31692]. Allotype, ?, same data [MCZ].
Male. Length 1.50 mm. Body 1.93 X as long as broad, strongly convex. Head and pronotum dark reddish, elytra black. Vesti- ture of long, fine, recurved, yellowish hairs. Vertex slightly convex; frontoclypeal ridge bearing 2 lateral horns, which are about 3.0 X as long as wide at base, 0.20 X as long as pronotum, and separated by 4.0 basal widths, with lateral and mesal edges subparallel and apices slightly converging. Antennal segment III 1.60 X as long as IV. Pronotum 0.83 X as long as broad, widest at posterior third; anterior edge strongly produced and emarginate, forming 2 ap-
proximate, sharp, subtriangular horns, which are 0.10 X as long as pronotum; sides strongly rounded, the margins narrow and distinctly crenulate, not visible for their entire lengths from above; anterior angles barely produced forward, subacute; disc strongly convex, even; surface smooth and shiny; punctures 0.30 X as large as scutellar base and separated by 0.25 to 0.50 diameter. Elytra 1.16 X as long as broad and 1.50 X as long as pronotum; sides strongly rounded, apices acute; punc- tation single and confused; punctures coarser than and not as dense as those on pronotum, each puncture bearing a fine, recurved, yellowish hair, which is about 2.0 X as long as scutellar base. Prostemum strongly tumid and carinate; intercoxal process 0.40 X as wide as a procoxal cavity, parallel-sided. Protibia with outer apical angle strongly produced and dentate (Fig. 51). Metasternum 0.35 X as long as wide; suture 0.20 X as long as median length of sternum. Abdominal stemite III bearing a median, raised, circular, pubescent fovea, which is 0.33 X as long as body of stemite, distinctly margined, and located posterad of center.
Female. Length 1.50 mm. Body 1.82 X as long as broad. Vertex as in male; fronto- clypeal ridge simple. Pronotum 0.73 X as long as broad; anterior edge strongly rounded, simple. Elytra 1.15 X as long as broad and 1.73 X as long as pronotum. Protibial apex only slightly produced, angu- late or weakly dentate. Stemite III with- out a pubescent fovea.
Variation. Pronotum yellowish orange to dark reddish, usually reddish. Elytra yel- lowish to black, usually dark brown or black. Frontoclypeal ridge on smaller males bearing 2 short, subtriangular processes; in larger males these are represented by longer, narrow homs, which may be 0.33 X as long as pronotum. Anterior edge of pronotum in small males barely produced and emarginate, fonning 2 small tubercles; pronotal homs in larger specimens may be 0.20 X as long as pronotum. Size and di-
472 Bulletin Museum of Comporotive Zoology, Vol. 142, No. 5
mensions vary as follows in a series of 18 males and 14 females from Walterboro, South Carolina:
TL mm: $ 1.27-1.77 ( 1.43 ± 0.028),
9 1.12-1.57 (1.37 ±0.034); TL EW: S 1.76-1.97 ( 1.87 ± 0.013),
9 1.78-1.93 (1.84 ±0.011); PL/PW: S 0.74-0.91 (0.82 ±0.011),
9 0.73-0.81 (0.78 ±0.006); EL/EW: S 0.07-1.20 (1.14 ± 0.007),
9 1.12-1.21 (1.17 ±0.008); EL/PL: $ 1.44-1.75 (1.55 ± 0.021),
9 1.67-1.87 (1.77 ±0.016); GD/EW: c^ 0.71-0.79 (0.75 ± 0.005),
9 0.70-0.77 (0.74 ±0.006). Total size range in material examined: 1.10-1.78 mm.
Farati/pcs. ALABAMA: 1, Mobile, II- 17-12, "h. p. Loding [GIN]; 1, same locality, 3-4-12 [GIN]; 3, Oak Grove, 17.6, Gollection H. Soltau [USNM]. ARKANSAS: 1, Pine Bluff, 23.11, Gollec- tion H. Soltau [USNM]. FLORIDA: 3, Orescent Gity, Goll. Hubbard & Schwarz [USNM]; 2, Enterprise, May 27 [MGZ]; 2, same locality, 27.5, Hubbard & Schwarz [USNM]; 1,'same locaHty, 28.5, Hubbard & Schwarz [USNM]; 2, 4 mi. SE Lake Placid, Highlands Go., June 25, 1965, Lot 1516 JFL, ex Pohjponis iodinus [JFL]; 2, St. Nicholas [USNM]. GEORGIA: 1, Savannah, Ghatham Go., Apr. 6, 1966, Lot 1810 JFL (H. and A. Howden, coll.), ex Pohjporus ^ilvus [JFL]. LOUISIANA: 2, Fontainebleau State Park, St. Tam- many Par., June 19, 1965, Lot 1456 JFL, ex Ganoderma lucidum [JFL]. MIS- SISSIPPI: 2, Hancock Go., 23.8, Hubbard & Schwarz [USNM]; 2, "Rose's Bluff," Natchez Trace, Madison Go., 25-III-1959, cortical [USNM]. NORTH GAROLINA: 5, Atlantic Beach, Garteret Go., May 19, 1966, Lot 1885 JFL (Gari Parsons, coll.), ex Stereum ostrea [JFL]; 2, Bladen Go., 2-V-1964, Jim F. Gornell [JFG]; 2, Moore Go., VI-1-64, J. F. Gornell [JFG]; 2, Raleigh, 25-IV-1964, Jim F. Goniell [JFG]; 2, same locality, IX-8-64, J. F. Gornell
[JFG]; 15, Sampson Go., 2-V-1964, J. F. Gornell [JFG, JFL]; 1, Statesville, Tredell Go., May 25, 1966, Lot 1911 JFL (Gari Parsons, coll.), ex Ganoderma cttrtisii [JFL]; 2, same locaHty and date. Lot 1912 JFL (Gari Parsons, coll.), ex Ganoderma curtmi [JFL]. SOUTH GAROLINA: 30, Walterboro, Golleton Go., II-1.5-64, H. D. Blocker [FMNH, JFG, JFL, MGZ, USNM].
Distribution. Southeastern United States, from North Garolina to Florida and west to Arkansas and Louisiana (Fig. 89).
Hostfuni!,i. Ganoderma curtisii [2]; Gano- derma lucidum [1]; Polyporus gilvus [1]; Pohjporus iodinus [1]; Stereum ostrea [1].
Discussion. This species resembles Cis hirsutus and C. urstdinus in the type of vestiture and frontoclypeal horns of the male. It differs from both in the shorter and broader form with narrowed elytral apices, reduced metastemum, and sharply carinate prosternum. In addition, C. iirsu- linus is usually more reddish in color with somewhat finer punctation, and C. hirsutus is larger with somewhat denser punctation.
Cis rotundulus is restricted to the south- eastern United States and is not known from the West Indies. The narrowed elytral apices and shortened metastemum indicate that the species may be evolving in the direction of flightlessness; the hindwings are slightly reduced with a tnmcate apex. Tlie distribution pattern suggests that C. rotundulus, like C. tirsulinus, may have evolved on the mainland rather than in the Greater Antilles.
Cis stereophilus NEW SPECIES Figure 40
Ilolotype. 6 , NORTH GAROLINA: At- lantic Beach, Garteret Go., May 19, 1986, Lot 1887 J. F. Lawrence ( Gari Parsons, coll. ), ex Stereum sp. on hardwood branch [MGZ No. 31693]. Allotype, 9, same data [MGZ].
Male. Length 1.42 mm. Body 2.28 X as long as broad, moderately convex. Head and pronotum reddish orange, elytra black- ish brown anteriorly, yellowish brown
North American Ciidae • Lawrence 473
posteriorly. Vestiture of short, stout, blunt, yellowish bristles. Vertex flattened, with slight median impression; frontoclypeal ridge bearing 2 sharp tubercles that are separated by about 2 basal widths. Anten- nal segment III 1.70 X as long as IV. Pronotum 0.87 X as long as broad, widest at posterior third; anterior edge strongly rounded, barely emarginate at midline; sides weakly rounded, the margins narrow and distinctly crenulate, not visible for their entire lengths from above; anterior angles barely produced forward, subacute; disc sti'ongly convex, even; surface finely granulate and shiny; punctmes 0.50 X as large as scutellar base and separated by 0.20 to 0.33 diameter. Elytra 1.44 X as long as broad and 1.71 X as long as pronotum; sides weakly rounded, apices blunt; punctation single and subseriate; punctures about as large and dense as those on pronotum, each bearing a stout, blunt, yellowish bristle, which is about 0.20 X as long as scutellar base. Prostemum slightly tumid; intercoxal process 0.22 X as wide as a procoxal cavity, narrowing posteriorly. Protibia with outer apical angle slightly produced and weakly dentate (Fig. 53). Metastemum 0.50 X as long as wide; suture 0.30 X as long as median length of sternum. Abdominal sternite III bearing a median, circular, pubescent fovea, which is 0.30 X as long as body of sternite, distinctly margined, and located anterad of center. Sternite VIII as^ in Figure 66. Aedeagus as in Figures 77 and 82.
Female. Length 1.30 mm. Body 2.17 X as long as broad. Vertex as in male; frontoclypeal ridge simple. Pronotiun 0.84 X as long as broad; anterior edge strongly rounded, simple. Elytra 1.42 X as long as broad and 1.89 X as long as pronotum. Protibia with outer apical angle slightly produced and angulate. Steniite III with- out pubescent fovea.
Variation. Pronotum yellowish orange to black, usually reddish orange. Elytra yel- lowish to black, usually blackish brown
anteriorly and yellowish brown posteriorly. Almost all specimens examined were either bicolored with a paler pronotum or tri- colored with the elytra yellowish poste- riorly; in only one specimen was the color- ation uniform. Frontoclypeal tubercles in males vary somewhat in size. Anterior edge of pronotum simple and rounded in smaller males, distinctly emarginate in larger speci- mens. Size and dimensions vary as follows in a mixed series of 13 males and 21 fe- males from various parts of the Atlantic Coast:
TL mm: S 1.20-1.50 (1.34 ± 0.025),
2 1.17-1.57 (1.37 ±0.026); TL/EW: S 2.22-2.32 (2.27 ± 0.010),
9 2.17-2.42 (2.27 ±0.014); PL/PW: $ 0.84-0.92 (0.87 ± 0.006),
9 0.83-0.95 (0.87 ±0.006); EL/EW: S 1.43-1.52 ( 1.46 ± 0.007),
5 1.42-1.58 (1.50 ±0.010); EL/PL: $ 1.71-2.00 (1.82 ±0.021),
9 1.75-2.06 (1.93 ±0.016); GD/EW: S 0.74-0.84 (0.77 ± 0.008), 9 0.71-0.81 (0.76 ±0.006). Total size range in material examined: 1.05-1.60 mm.
Specimens from northern Mexico differ from those of the eastern United States in being consistently smaller and darker in color and in having less well-developed frontoclypeal tubercles in the male.
Paratijpes. DISTRICT OF COLUMBIA:
3, Washington, 24.5, Coll. Hubbard & Schwarz [USNM]. FLORIDA: 2, Torreya State Park, Liberty Co., Apr. 8, 1969, Lot 2712 JFL (S. B. Peck, coll.), ex Stereum ostrea [JFL]. MARYLAND: 1, Blad[e]- nsb[er]g, 13.7, Coll. Hubbard & Schwarz [USNM]; 1, same locality, 20.7, Coll. Hub- bard & Schwarz [USNM]. MASSACHU- SETTS: 3, Naushon Is., Elizabeth Islands, May 25, 1965, Lot 1671 JFL ( Carl Parsons, coll.), ex Stereum ostrea [JFL, MCZ]; 2, Vineyard Haven, Martha's Vineyard, Dukes Co., May 20, 1965, Lot 1684 JFL (Carl Parsons, coll.), ex Stereiwi ostrea [MCZ];
474 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
1, Woods Hole, Barnstable Co., June 5,
1966, Lot 1812 JFL, ex Stereum ostrea [JFL]; 1, same locality and date. Lot 1813 JFL, ex Stereum ostrea [JFL]. NEW JERSEY: 2, Anglesea, 7.3, Coll. Hubbard & Schwarz [USNM]; 3, same locality, 24.7, Liebeck Collection [MCZ]. NORTH CAROLINA: 6, Atlantic Beach, Carteret Co., May 19, 1966, Lot 1887 JFL (Carl Parsons, coll.), ex Stereum sp. [JFL, MCZ]; 1, Highlands, Macon Co., June 16, 1962, Lot 146 R. C. Graves, ex Stereum fasciutum [ = S. ostrea] [JFL]; 1, nr. Magnolia, Duplin Co., V-26-64, J. Cornell & P. Mampe [JFC]; 5, Raleigh, IX-27-64, J. F. Cornell, coll. [JFC, JFL]; 1, Thompson River Corge, Transylvania Co., May 17,
1967, Lot 2498 JFL'(Carl Parsons, coll.), ex Stereum ostrea [JFL]; 24, same locality. May 19, 1967, Lot 2511 JFL (Carl Parsons, coll.), ex Stereum ostrea [CAS, FMNH, JFL, MCZ]; 1, Tiyon, 1784c Hopk. U.S., Quercus, W. F. Fiske [USNM]. NUEVO LEON: Chipinque Mesa, 5400', Monterrey, Jmie 23, 1969, Lot 3050 JFL (S. & J. Peck, coll.), ex Stereum sp. [JFL, MCZ]. PENN- SYLVANIA: 1, Wisahick[o]n Cr., 7.24, Liebeck Collection [MCZ]; 1, Ches[t]- n[u]t H[i]ll, VII.31, Liebeck Collection [MCZ].
D'lstrihution. Eastern coast of the United States, from southern Massachusetts to northern Florida, and northeastern Mexico.
Host fimgi. Stereum ostrea [10(5)]; Stereum sp. [2(2)].
Discussion. This species is characterized by the small size, moderately elongate form, single, subseriate, elytral punctation with very short and broad, blunt bristles, dentate or angulate protibial apex, coarsely and densely punctate pronotum \\'ith nar- row lateral edges, and the normally light colored pronotum and bicolored elytra. The male has two tubercles on the fronto- clypeal ridge. The species resembles C. krausi in several respects, but may be distinguished from it by the shorter body form (EL/EW less than 1.60), the single and more regular elytral punctation, and
the coloration. C. suhtilis differs from C. stereophilus in having distinctly dual elytral punctation, colorless bristles, finer and sparser pronotal pmictation, uniform color- ation, and in the absence of a pubescent fo- vea on the abdomen of the male. C. festi- vulus has a similar body form and blunt yel- lowish elytral bristles, but the punctation is confused, the bristles uneven, and the pro- tibial apex romided. C. rohiniophilus also has similar body fonn and vestiture, but the pronotal punctation is much finer and sparser and the elytral pimctures are uni- form. Elytral bristles in C. hijstriculus and C. angustus are not as short and broad and are more sparsely distributed; in addition, the former is much larger in size and lacks an abdominal fovea in the male, and the latter species is characterized by having sparser pronotal punctation.
In collections, individuals have been identified as C. julichi Dury, but the type of julichi is conspecific with C. suhtilis Mellie.
This is the only species of North American Ciidae that is restricted to fungi of the genus Stereum, particularly S. ostrea. Fruiting bodies of this fungus are very thin and become woody with age, yet several in- sects are able to feed within the tissue. C stereophilus definitely breeds in the fungus, and larvae have been taken on several occasions. It is rather rare in collections, but this is almost certainly due to the pe- culiar habitat. Specimens are known only from the Atlantic Coast and parts of north- eastern Mexico, but it is likely that the species occurs throughout the eastern part of North America.
At the present time this species cannot be placed in any group, although it may be related to C. krausi and its relatives. It does not appear to have any relatives in the Palaearctic Region, and may be a northern representative of a Neotropical complex.
The name stereo))hilus is derived from the the basidiomycete genus Stereu7n and the Greek pJiileo, meaning to be fond of
North American Ciidae • Lawrence 475
( referring to the preference of these beetles for Sterewn fruiting bodies).
C/s striolatus Casey
Cis striolata Casey, 1898: 79. Type locality: "Colorado (Salida)." Holotype, 9, Casey Coll., USNM.
Cis frateina Casey, 1898: 80. Type locality: "Utah (southwestern)." Holotype, $, Casey Coll., USNM. NEW SYNONYMY.
Cis macilenta Casey, 1898: 80. Type locality: "California (Lake Tahoe)." Holotype, ?, Casey Coll., USNM. NEW SYNONYMY.
Cis versicolor Casey. — Hatch, 1962: 231. Mis- identification.
Distribution. Widespread across the northern part of North America from the northern part of the Mackenzie District, Northwest Territoiy (not on map), to Nova Scotia, south into the Sierra Nevada and Rocky Mountains, in the Midwest as far as Kansas, and on the Atlantic coast as far as northern Florida (Fig. 101). Marginal records: NORTHWEST TERRITORY: Aklavik, Mackenzie Dis- trict; BRITISH COLUMBIA: Midday Val- ley, Merritt; QUEBEC: Gaspe; NOVA SCOTIA: Portaupique; VIRGINIA: Mt. Vemon, Fairfax Co.; FLORIDA: Panama City, Bav Co.; KANSAS: Lawrence, Doug- las Co.; COLORADO: Salida, Chaffee Co.; UTAH: Southwestern; CALIFORNIA: 21 mi. NE Strawberry, Toulumne Co.
Host fungi. Polyporus abictimis [6(2)]; Folijporus pargamenus [4(3)]; DaeclaJca unicolor [1(1)]; Polyporus hirsutus [1(1)].
Discussion. Individuals of this species may be distinguished from most other Cis by the narrow and elongate, somewhat de- pressed form, dual and distinctly seriate elytral punctation, ^'estiture of short, stout bristles, angular or dentate protibial apex, and slightly produced anterior pronotal angles. Cw fuscipes is larger and broader, with the anterior pronotal angles distinctly produced and the pronotal disc impressed anteriorly in the male. C. cayensis is smaller in size, with dual vestiture (visible only under higher magnification) and two tubercles on the vertex in the male. C.
striolatus differs from C. versicolor in hav- ing a darker, duller, and more distinctly granulate pronotum with narrower lateral margins, the frontoclypeal ridge bituber- culate in the male, and the abdominal fovea smaller. The species is most similar to Cis tristis, from which it differs by having a more sparsely pmictate and coarsely granu- late pronotmn, yellowish elytral bristles, a much smaller abdominal fovea and differ- ent male genitalia (Fig. 70, cf. Fig. 69).
Cis striolatus appears to be most closely related to Cis striatuJus Mellie from the Palaearctic Region and to C. versicolor from western North America. It is possible that C. striatal us and C. striolatus represent a single Holarctic species.
Cis striolatus, C. fraterna, and C. maci- lenta, all described by Casey, were based on a single female from Colorado, a pair from southwestern Utah, and a female from Lake Tahoe, California, respectively. Characters used to separate the three (im- pressions at the apical angles of the pro- notum and the natru-e of the outer apical angle of the protibia) vaiy considerably within one series. The extent to which the apex of the protibia is produced, thereby forming a tooth, varies between the sexes, and the impressions at the an- terior pronotal angles are found only in some larger males, in which the pronotum is laterally expanded.
The distribution pattern is a typically northern one, and the species is often as- sociated with conifer forests. In the south- eastern part of its range, C. striolatus is sympatric with C. tristis. C. striolatus normally feeds on the fruiting bodies of Polyporus abietinus and the related P. pargamenus, whereas C. tristis usually oc- curs on P. versicolor and its relatives. In western North America, C. striolatus may be sympatric with the related C. versicolor; the latter is associated \\'ith hardwoods and feeds on fruiting bodies of the P. versicolor group, while the fonner occurs on P. abietinu.s on conifers.
476
Bulletin Miiscun} of Coiiiparafivc
ZooJooy, Vol. 142, No. 5
Cis subfuscus Gorham
Cis suhfusciis Gorham, 1886: 357. Type locality: "Mexico, Cordova, Veracruz." Holotype, 2, BMNH.
Distribution. Central Texas south along the eastern coast of Mexico to Veracmz. Marginal records: TEXAS: College Station, Brazos Co.; HIDALGO: 7 mi. SW Jacala; PUEBLA: 29 mi. E Xicotepec; VERA- CRUZ: Veracruz.
Host fungi. Tohjporus hiisutus [2(1)]; Paiiellns sttjpticus [1(1)]; Daedalea ele- gans [1]. Also recorded from Poly poms sanguineus.
Discussion. Tliis species is similar to C. versicolor, but the body is shorter and broader, the size is usually smaller, the elytral punctation is coarser and denser, and the male bears two tubercles on the frontoclypeal ridge. It also resembles Cis pusillus Gorham, described from the West Indies but occmring also in Brazil.
Cis subtilis Mel lie
Cis subtilis Mellie, 1848: 353. Type locality: "Amerique boreale." Holotype, $ , Marseul Coll., MNHN.
Cis conftisus Blatcliley, 1910: 899. Type locality: ". . . Perry . . ." [Co., Indiana]. Lectotype, Blatchley Coll., PURD. NEW SYNONYMY.
Cis julichi Dury, 1917: 7. Type locality: "New York City." Holotype, $ , ' Dury Coll., CIN. NEW SYNONYMY.
Cis wenzeli Dury, 1917: 8; Weiss and West, 1920: 8 (dist., biol.). Type locality: "Del.- [aware] Co. Penn." Holotype, Dury Coll., CIN. NEW SYNONYMY.
Distribution. Eastern North America, from New Hampshire south to Florida and west to Illinois, Arkansas, and eastern Texas (Fig. 106). Marginal records: NEW HAMPSHIRE: 5 mi. N Wilton, Hillsboro Co.; FLORIDA: 4 mi. NW Copeland, Collier Co.; TEXAS: Houston, Harris Co.; ARKANSAS: (southwest); ILLINOIS: Olive Branch, Alexander Co.
Host fungi. Polyporus pargamenus [33 (11)]; Polyporus abietinus [7(1)]; Poly- porus sector [5(4)]; Ganodernia applana-
tum [1]; Polyporus adustus [1]; Polyporus gilvtis [1]; Steretim ostrea [1].
Discussion. This species may be dis- tinguished from other North American Cis by the relatively long and narrow body form, dual and confused elytral punctation, and vestiture of short, stout, colorless bristles. It differs from the closely related C. acritus by having two frontoclypeal tubercles and no abdominal fovea in the male. Within its range it may be confused with C. tristis, which differs in having the elytral punctation seriate. In Cis stereophi- lus, the elytral punctation is single and the bristles yellowish, while in C. krausi the elytra are more elongate (EL/EW greater than 1.60) and maculate.
Cis subtilis is fairly common throughout the eastern United States, where it breeds on the fmiting bodies of Polyporus par- gamenus and its relatives. If one ignores secondary sexual characters, the species is barely distinguishable from the western C. acritus (see p. 443). The species has no close relatives in Europe and similar fonns occur in the West Indies and Mexico.
Cis tetracentrum Gorham
Cis tetracentrum Gorham, 1886: 357. Type lo- cality: "... Northern Sonora." Lectotype, S , BMNH.
Cis arizonae Dury, 1917: 5. Type locality: "Madera Canyon, Santa Rita Mountains, Ari- zona." Holotype, Wenzel Coll.? Paratypes, $ 9 , Dury Coll., CIN. NEW SYNONYMY.
Distribution. Mountains of southern California and Arizona and south through the Mexican highlands as far as central Veracruz (Fig. 94). Marginal records: CALIFORNIA: (no specific locality); ARIZONA: Graham Mt., Graham Co.; VERACRUZ: Orizaba; MEXICO: Toluca; DURANGO: 14 mi. SW El Salto; SO- NORA: (northern).
Host fungi. Polyporus versicolor [5(4)].
Discussion. This is the largest of the North American Ciidae, individuals averag- ing 3 mm in length. The fonn is somewhat elongate, the elytral punctation is moder-
North American Ciidae • Lawrence 477
ately coarse, dense, single, and confused, the vestiture consists of relatively long and fine bristles, the lateral pronotal margins are broad with a raised lip and barely crenulate edge, the anterior pronotal angles are produced and rounded, the prosternum is slightly tumid, and the male bears two stout pronotal horns, two subtriangular frontoclypeal plates, and a fairly small abdominal fovea ( Fig. 32 ) . In Cis discolor the body is more elongate and parallel- sided and the pronotal margins lack the raised lip and are distinctly crenulate. In Cis vituhis and C. congesttis the body is shorter and broader, the bristles shorter and thicker, and the male lacks the ab- dominal fovea and bears an elevated, trisinuate ridge on the clypeus. In Cis pistoiia the elytral punctation is dual and the bristles are very short and scalelike.
Cis tetracentrum is closely related to Cis corticimis Gorham, from the highlands of Mexico and Guatemala, and the two may not be specifically distinct. Individuals of C. corticimis are usually larger than those of C. tetracentrum, the elytral punctation is somewhat finer and denser, so that the vestitLu-e is not as sparsely distributed, and the pronotal and clypeal homs in the male are more prominent. These two fonns and Cis pallidtts Mellie, from Brazil and Argen- tina, may constitute a species group, which, in turn, is related to the Cis vitidus com- plex, via Cis bisbidens Gorham.
Cis tridentatus Mannerheim
Cis tridentatus Mannerheim, 1852: 360. Type locality: "Insulae Sitkhae." Syntypes, Manner- heim Coll., MZUH.
Xestocis cdnae Hatch, 1962: 232. Type locality: "Seattle, Wash." Holotype, $, Hatch Coll., UVV. NEW SYNONYMY.
Xestocis reflexus Hatch, 1962: 232. Type locality: "Coupeville (Simnyside), Wash." Holotype, c? , Hatch Coll., UW. NEW SYNONYMY.
Distribution. Pacific Coast from southern Alaska, through British Columbia, Wash- ington, and Oregon, to Monterey Co., California. Marginal records: ALASKA: Sitka: BRITISH COLUMBIA: Massett,
Queen Charlotte Is.; WASHINGTON: White River, Mt. Rainier, Pierce Co.; CALIFORNIA: Big Sur, Monterey Co.
Host fungi. Fomes pinicola [10(7)]; Poria cinerascens [2(1)]; Ganoderma ore- gonense [1(1)]; Pohjporus sidphureus [1(1)]; Foria corbonica [1(1)]; Trametes sepitrm [1(1)]; Ganoderma applanatum [1]; Fleiirotus ostreatus [1]; Folyporus gilvus [1].
Discussion. This species is identical with Cis americanus in most respects but differs in having the elytral punctation obscurely dual, with the megapunctm-es barely larger than the micropunctures. C. tridentatus appears to be restricted to the immediate coast and no specimens have been taken inland. Tliroughout the range, it is sym- patric with C. americanus, but the t\vo have not been taken on the same host in any one area. Cis tridentatus is commonly found breeding in the fruiting bodies of Fomes pinicola, along with Cis biarmatus and DolicJiocis indistinctus. Cis americanus {"oweni' type) also frequents Fomes pini- cola but always at localities some distance from the coast and outside of the range of C. tridentatus. Coastal populations of C. americanus commonly breed in Folyporus adustus, Stereum hirsutum, and several other fungi (see p. 444).
Cis trisfis Mellie
Cis tristis Mellie, 1848: 343. Type locality:
"NouveUe-Orleans." Holotype, $ , Pic Coll.
(Chevrolat Coll.), MNHN. Cis setulosus Mellie, 1848: 257. Type locality:
"Amerique boreale." Holotype, S , Melly Coll..
GEN. NEW SYNONYMY. Cis falli Blatchley, 1910: 898; Dury, 1917: 9
(dist. ). Type locahty: "Marion ..." [Co.,
Indiana]. Lectotype, $, Blatchley Coll., PURD.
NEW SYNONYMY.
Distribution. Eastern North America, from New York and Massachusetts south to Florida and west as far as southeastern Colorado and northeastern Mexico (Fig. 103). Marginal records: WISCONSIN: (no specific localitv); MICHIGAN: Saugatuck, Allegan Co.; NEW YORK: (no specific
478 Bulletin Museum of Cotuparafivc ZooJop,y, Vol. 142, No. 5
locality); MASSACHUSETTS: (no specific locality); FLORIDA: Panama Citv, Bav Co.; LOUISIANA: New Oilcans, Orleans Par.; TEXAS: Browaisville, Cameron Co.; NUEVO LEON: Chorros de Agua, 13 mi. W Montemorelos; COLORADO: Pueblo, Pueblo Co.; NEBRASKA: Central City, Merrick Co.
Host fungi. PoJyponis versicolor [3(3)]; Lenzites hetulina [2(1)]; Polyportis hirsu- tus [2]; Vohiporus supinus [2]; Daedalca amhigiui [1(1)]; Poly poms maxiryiiis [1 (1)]; Trametes hispida [1(1)]; Pleurotus sp. [1]; Polyportis cinnaharinus [1].
Discussion. Tliis species closely resembles Cis striolatus but may be distinguished from that species by the shiny pronotum with denser punctation, colorless elytral bristles, and larger abdominal fovea in the male. Cis subtilis is similar to C. tristis in the elongate form, dual elytral puncta- tion, and colorless bristles, but the elytral punctation is confused and the abdominal fovea is absent.
The species is not very common, but it occurs throughout eastern North America, mainly south of New England and the Great Lakes, where it is usually found in association \\ith Polyporus versicolor and its relatives. Although Cis tristis resembles the northern C. striolatus, as well as the European species C. .striatulus Mellie and C. compttis Gyllenhal, the male genitalia are different and the true affinities may be with members of the Neotropical fauna.
Cis ursuimus Casey
Cis uisulina Casey, 1898: 83; Blatchley, 1910: 899 (dist); Blatchley, 1918: 54 (dist.). Type locality: "Alabama." Holotype, 9 , Casey Coll., USNM.
Distribution. Southeastern United States, from North Carolina south to Florida and west to Arkansas and Louisiana (Fig. 92). Marginal records: NORTH CAROLINA: Knott's Island, Cumtuck Co. FLORIDA: Dunedin, Pinellas Co.; Archibold Biological Station, Highlands Co. LOUISIANA: Audubon State Park, West Feliciana Par.
ARKANSAS: Hope, Hampstead Co. INDI- ANA: Crawford Co.
Host funfii. Ganoderma tsugae [1]; Polyporus adustus [1]; Polyporus gilvus [1]; Polyporus sulphiireus [1]; Polyporus versatilis [1].
Discussion. This species differs from most North American Cis in having a vesti- ture of long, fine, recurved hairs, as in C. hirsutus and C. rotund ulus. The form is more elongate than that of C. rotundulus, which also differs by virtue of the sharply carinate prostemum, narrowed elytral apices, shortened metasternum, and sparser punctation. Individuals of C hirsutus are larger (TL more than 1.6 mm) and darker in color, mth coarser punctation, smoother lateral pronotal margins, and a larger ab- dominal fovea in the male.
Like C. rotundidus, this species appears to be restricted to the southeastern United States. No specimens have been seen from the West Indies and the distribution in- dicates that the species may have evolved on the mainland.
Cis versicolor Casey
Cis versicolor Casey, 1898: SO. Type locality:
"California (Calaveras . . . Co.)." Holotype,
9, Casey Coll., USNM. Cis dichrous LeConte, 1867: 58; Blaisdell, 1892:
34 (biol.); Weiss and West, 1921b: 169 (dist.,
liiol.). Nomen nudum.
Distribution. Extreme southern Oregon, south throughout most of California, west of the Sierran crest, into Baja California Norte and east through Arizona into New Mexico and western Texas (Fig. 101). Marginal records: ORECON: 3 mi. W Dead Indian Springs, Jackson Co. CALI- FORNIA: Dutch Flat, Placer Co.; Ash Mt. R., Sequoia National Park, Tulare Co. ARIZONA: 5 mi. SE Wickenburg, Mari- copa Co. NEW MEXICO: San Juan Valley, Taos Co. TEXAS: Boquillas Camp, Big Bend National Monument, Brewster Co. BAJA CALIFORNIA NORTE: 12 mi. SE Maneandero.
Host fungi. Polyporus versicolor [22
North American Ciidae • Laurence 479
(14)]; Trametcs hispida [7(4)]; Pohjporiis hirsuttis [5(2)]; Lenzites betuUna [2(2)]; Ganoderma brownii [1]; Polyporus cinna- barinus [1]; Schizoplujllum commune [1].
Discussion. This species is characterized by the elongate and somewhat depressed form, dual and distinctly seriate elytral punctation, vestiture of short bristles, shiny pronotum with fairly broad lateral margins, rounded or angulate protibial apex, reddish pronotum (in contrast to the dark elytra), and lack of frontoclypeal tubercles in the male. Cis subfuscus is somewhat smaller and not as elongate, with coarser and denser elytral punctation and distinct frontoclypeal tubercles in the male. Cis striohtus and C. tristis are similar, but in both, the outer apical angle of the protibia is more prominent, the lateral pronotal margins are narrower, and the clypeus of the male is tuberculate. Cis stibtilis and C acritus differ in having confused elytral punctation, while individuals of C fuscipes are larger and broader with distinctly pro- duced and rounded anterior pronotal angles and dentate protibial apices.
Cis versicolor appears to be related on the one hand to C. striolatus and C. striatu- lus, of the northern Nearctic and Palac- arctic respectively, and on the other to C. subfuscus and C. pusiUus of the Neo- tropical Region. The distribution is south- western and the species probably extends into northern Mexico.
The species occurs on fungi of the Pohj- porus versicolor group, which are usually associated with hardwoods rather than conifers. In the wetter areas of central and northern California, it may be found with C. vitidus, C. fuscipes, Sulcacis curtu- his, Ceracis calif ornicus, and Octotemnus laevis, while in drier areas of the state it often occurs alone. In southeastern Cali- fornia and Arizona, Cis versicolor occupies two distinct habitats: along streams and river beds in the desert, where it feeds on fruiting bodies of Trametes hispida grow- ing on cottonwoods, and in mountain canyons, where it may be found in Poly-
porus versicolor on various hardwoods. It has been collected in association with Ceracis dixiensis in the former and with Cis ietracentrum in the latter situation.
Cis vitulus Mannerheim
Cis vitulus Mannerheim, 1843: 299; Weiss, 1920b: 133-134 (biol.); Weiss, 1923: 199 (biol.); Weiss and West, 1921a: 61 (dist., biol.); Weiss and West, 1921b: 169 (dist., biol). Type locality: "California." Types, Mannerheim Coll., MZUH ?
Cis caseiji Dalla Torre, 1911: 8, replacement name for Cis illustiis Casey, 1898 (not Broun, 1880). NEW SYNONYMY.
Cis iUustris Casey, 1898: 81. Type locality: "California ( Humboldt Co. ) ." Holotype, $ , Casey Coll., USNM.
Distribution. California, from Del Norte Co. to San Diego Co., and north central Arizona (Fig. 105). Marginal records: CALIFORNIA: Patrick Cr., Del Norte Co.; Dutch Flat, Placer Co.; Dorset Camp, Sequoia National Park, Tulare Co.; Palm Springs, Riverside Co.; San Diego, San Diego Co. ARIZONA: Oak Creek Canyon, Coconino Co.
Host fungi. Polyporus versicolor [29 (IS)]; Lenzites betulina [4(3)]; Polyporus adustus [2(1)]; Scliizopliyllum commune
[1].
Discussion. Individuals of this species are large and stout with single and uniform elytral punctation, vestiture of moderately short bristles, strongly tumid prostenium, broad lateral pronotal margins with a raised lip, produced and rounded anterior pronotal angles, and male with two pro- notal horns, an elevated, trisinuate, fronto- clypeal ridge (Fig. 3), and no abdominal fovea. Cis laminatus differs in the much narrower pronotal margins, shorter and broader form, coarser and denser pronotal punctation, and simple pronotal apex in the male. In Cis ietracentrum, the body is more elongate, the size is usually larger, the bristles are longer and finer, and the male bears two subtriangular frontoclypeal plates and an abdominal fovea. Cis con- gestus is quite similar to C vitulus but is
480 Bulletm Museum of Comparative Zoology, Vol. 142, No. 5
characterized by the smaller size, shorter and broader form, and coarser and denser pronotal punctation
Cis vituliis belongs to a group of species occiming mainly in the New World tropics and including the southeastern C. con- gestits and the Mexican species C. huhalus Reitter and C. fasckitus Gorham. The spe- cies has a relatively restricted range oc- curring primarily in mesic situations along the California coast and in the foothills of the Sierra Nevada.
Genus Ennearfhron Mellie
Ennearthron Mellie, 1847: 110; Mellie, 1848: 360; Lacordaire, 1857: 552; Jacquelin dii Val, 1861: 238; Thomson, 1863: 190; Abeille de Perrin, 18741): 80; Reitter, 1902a: 59; Dalla Torre, 1911: 23; Winkler, 1927: 794; Miyatake, 1954: 55; Miyatake, 1959: 27; Lawrence, 1967b: 91- 92; Lohse, 1965: 294. Type species, by sub- sequent designation, Cis corniitus Gyllenhal, 1827: 626 (Desmarest, 1860: 261).
Emiearthrum Bach, 1852: 111. Incorrect sub- sequent spelling.
Plesiocis Casey (in part). — Hatch, 1962: 233.
Inchided species. Ennearthron amamense Miyatake 1959: 27 [Ryukyu Is.]; E. auris- qtiamosum Lawrence, n. sp. [southeastern U. S., see p. 481]; E. cJnijui Nakane and Nobuchi, 1955: 49 [Japan]; Cis cornutus Gyllenhal, 1827: 626 [Eurasia]; E. ishiharai Miyatake, 1954: 57 [Japan]; E. mohrii Miyatake, 1954: 56 [Japan]; Plesiocis spen- ceri Hatch [northwestern North America, see p. 482]. Total: 7 species.
Doubtfully included species. Ennearthron aheillei Caillol, 1914: 160 [southern Europe]; Cis filum Abeille de Perrin, 1874a: 53 [southern Europe]; E7mearthron hayashii Nobuchi, 1955: 108 [Japan]; E. mussauense Ghujo, 1966: 529 [Bismark Is.]; E. ondreji Roubal, 1919: 63 [southeastern Europe]; E. palmi Lohse, 1966: 28 [north- em Europe]; E. poi'iae Nakane and No- buchi, 1955: 49 [Japan]; Cis pruino.stdiis Ferris, 1864: 291 [Europe]; E. pulchellum Scott, 1926: 36 [Seychelles]; E. reichei Abeille de Perrin, 1874b: 89 [Egypt]; Cis
reitten Flach, 1882: 249 [Europe]. See discussion below.
Excluded species. Ennearthron argen- timim Pic, 1916: 19 [Argentina]; E. breve- hirsutum Pic, 1922: 8 [Cameroons]; E. biroi Pic, 1956: 77 [New Guinea]; E. boettgeri Reitter, 1880: 181 [New Zealand]; E. longepilosum Pic, 1922: 8 [India]; E. multidentatum Pic, 1917: 4 [China]; E. obsoletum Reitter, 1880: 182 [New Zea- land]; E. sinense Pic, 1917: 4 [China]; E. vianai Pic, 1940: 12 [Argentina]. See dis- cussion below.
Almost 60 species have been described in the genus Ennearthron on the basis of an- tennal segmentation (9 segments), and about half of these have already been trans- ferred to other genera, such as Ceracis, DipJujllocis, Dolichocis, Uadraule, Orthocis, Sidcacis, and Wagaicis. The remainder may be divided into three groups: those definitely included in the genus as nar- rowly defined here, those doubtfully in- cluded for various reasons discussed below, and those excluded and transferred to the genus Cis.
The genus Ennearthron is used here in a restricted sense to include the seven species listed above and three or four others from India and China that are relatively small in size with a characteristic clypeal notch in the male (Fig. 8). They may be dis- tinguished from Plesiocis crib rum by the broader intercoxal process of the pro- sternum, relatively longer metastemal suture, and the clypeal notch.
The doubtful species include four (E. abeillei Caillol, E. mussauense Ghujo, E. ondreji Roubal, and E. reichei Abeille de Perrin) that have not been examined and are inadequately described, and seven more that seem to represent four separate lines :
1 ) E. pulchellum Scott. This is a broad and somewhat flattened form with bi- colored elytra and vestiture of long, fine hairs. The intercoxal process of the pro- sternum is broad, the protibial apices are angulate, and there are no sexual oma-
North American Ciidae • Laivrencc 481
ments on the head or pronotiim of the male. It probably should be included in a distinct genus along with Cis bifasciattis Reitter from Japan and several undescribed Indo-Pacific species.
2) E. filum (Abeille de Pcrrin), E. palmi Lohse, and E. hayashii Nobuchi. These three species are all quite small (1.4 mm or less), narrow, and cylindrical, with dentate protibial apices and two small foveae on the head of the male.
3) E. pniinosiilum (Ferris) and E. poriae Nakane and Nobuchi. These two may be related to the last group, but they are larger in size and lack the dentate protibiae and pores on the head of the male.
4) E. reitteri (Flach). Tliis species is short and broad with dual and indistinctly seriate elytral vestiture, carinate pro- sternum, broad lateral pronotal margins, and dentate protibial apices. It does not appear to be related to any of the above.
Nine species of Ennearthron were in- cluded on the basis of wrong antennal counts and should be transferred to the genus Cis (at least for the present). Ennearthron biroi Pic, E. boettgeri Reitter, and E. obsoletum Reitter belong to the Cis pacificus group {see Lawrence, 1967b: 98); E. sinense Pic and E. brevehirsutiim Pic are probably related to Cis fnscipes Mellie; E. uianai Pic will eventually fonn part of a new genus; and each of the remaining forms belongs to a different group of Cis.
Key to the North American Species of Ennearthron
1. Elytral bristles longer, more than 0.33 X as long as scutellar base, subseriate; lateral pro- notal margins visible for their entire lengths from above, the anterior angles produced and acute; apex of pronotum in male with 2 weak, approximate horns joined by an
impressed ridge; British Columbia
E. spenceri (p. 482)
— Elytral bristles shorter, less than 0.33 X as long as scutellar base, confused; lateral pro- notal margins not or Ijarely visible for their entire lengths from above, the anterior angles not or barely produced, almost right; apex of pronotum in male with 2 sharp, widely
spaced horns; southeastern U. S.
E. aurisqiiamosum (p. 481)
Ennearthron aurisquamosum NEW SPECIES
Holotype. £, NORTH CAROLINA: "N. C." "aurisqiiamosus Zimm." LeConte Col- lection [MCZ No. 31904]. Allotype, ?, "N.C." [MCZ].
Male. Length 1.72 mm. Body 2.23 X as long as broad, strongly convex. Head and pronotum reddish, elytra dark reddish brown anteriorly, yellowish brown poste- riorly. Vestiture of short, stout, blunt, yel- lowish bristles. Vertex with a deep, transverse impression, preceded by a median elevation; frontoclypeal ridge bear- ing 2 subtriangular plates with a median notch between them (Fig. 8). Antennal segment III 1.50 X as long as IV. Pronotum 0.87 X as long as broad, widest at posterior third; anterior edge produced and emargi- nate, forming 2 triangular horns that are separated by 1.75 basal widths; sides weakly rounded, the margins narrow and slightly crenulate, not visible for theii* entire lengths from above; anterior angles barely produced forward, almost right; disc strongly convex, even; surface finely granu- late and shiny; punctures 0.20 X as large as scutellar base and separated by 0.25 to 0.35 diameter. Elytra 1.39 X as long as broad and 1.65 X as long as pronotum; sides weakly rounded, apices blunt; punc- tation single and confused; punctures slightly larger than and about as dense as those on pronotum, each bearing a stout, blunt, yellowish bristle, which is about 0.33 X as long as scutellar base. Presternum biconcave; intercoxal process 0.30 X as wide as a procoxal cavity, parallel-sided. Protibia with outer apical angle strongly produced and dentate. Metasternum 0.46 X as long as wide; suture 0.33 X as long as median length of sternum. Abdominal sternite III bearing a median, circular, pubescent fovea, which is 0.27 X as long as body of sternite, distinctly margined, and located slightly posterad of center.
Female. Length 1.65 mm. Body 2.20 X
482 BuUetin Mttseiiiu of Comparative Zoology, Vol. 142, No. 5
as long as broad. Vertex with a slight median impression; frontoclypeal ridge simple. Pronotum 0.81 X as long as broad; anterior edge strongly rounded, simple. Elytra 1.47 X as long as broad and 2.00 X as long as pronotum. Pro tibia with outer apical angle produced and weakly dentate. Stemite III without pubescent fovea.
Variation. Pronotum reddish orange to reddish brown. Elytra yellowish brown to almost black anteriorly, yellowish brown posteriorly, usually some\\'hat lighter at apices. Size and dimensions vaiy as follows in type series:
TL mm: c^ 1.72, 1.82, 1.85, ? 1.65, 1.87;
TL/EW: S 2.23, 2.28, 2.31, ? 2.20, 2.20;
PL/PW: S 0.87, 0.90, 0.93, 9 0.81, 0.81;
EL/EW: £ 1.39, 1.44, 1.44, 2 1.44, 1.47;
EL/PL: S 1.64, 1.65, 1.70, 9 1.88, 2.00;
GD/EW: S 0.81, 0.84, 0.87, 9 0.82, 0.83.
Paratypes. KENTUCKY: 1, Mammoth Cave National Park, Edmonson Co., June 13, 1965, Lot 1430 JFL, ex Pohjponis adus- tus [JFL]; NORTH CAROLINA: 3, "N. C." [JFL, MCZ].
Distribution. Kentucky and North Caro- lina.
Host fungi. Polyporus adustus [1].
Discussion. This species may be dis- tinguished from E. spenceri by the nar- rower pronotal margins, shorter, confused elytral bristles, and more widely spaced pronotal horns. It resembles several spe- cies of Cis in general form and vestiture. Cis stereophilus has a more elongate form with subseriate elytral bristles, C. laminatus is somewhat larger with shorter and broader elytra, C. castlei is much smaller, and C. duplex has longer elytral bristles that are morc^ sparsely distributed. The species most closely resembles an unde- scribed Ennearthron collected in Polyporus voJvatus from western China.
Ennearthron spenceri (Hatch), NEW COMBINATION
Plesiocis spenceri Hatch, 1962: 233. Type lo- cality: "Vancouver, B. C." Holotype, $ , Hatch Coll., UW.
Disirihution. Known only from Van- couver, BRITISH COLUMBIA. Probably introduced from Japan.
Plost funiii. Recorded from Polyporus volvatus.
Discussion. This species may be dis- tinguished from E. aurisquamosum on the basis of vestiture, pronotal margins, and male armature, as indicated in the key. It is also similar to several species of Cis, including C. stereophilus, C. floridae, and C. duplex. Both Cis duplex and C. floridae resemble E. spenceri in general fonn, punc- tation vestitm'e, and male armature, but both have 10-segmented antennae and in neither species is the frontoclypeal ridge interrupted in the middle. In addition, the vestiture in C. floridae is indistinctly dual and that of C. duplex is composed of shorter bristles. Cis stereophilus is similar in general form and color, but the bristles are much shorter and the male armature is different. In Plesiocis cribrum, the pronotal punctation is coarser and denser, the lateral pronotal margins are narrower, the clypeus of the male is quadridentate, and the pro- sternum and metastemum differ as indi- cated in the generic key.
Ennearthron spenceri belongs to a group of species inhabiting eastern Asia and in- cluding E. amamense Miyatake, E. ishi- hiirai Miyatake, and E. mohrii Miyatake. Considering its limited distribution, the North American species may well be an introduction. Most of the individuals in the type series were taken from a herbarium specimen of Polyporus volvatus (origin not recorded), but at least one of the types in the University of British Columbia collec- tion was taken in Vancouver at a later date than the herbariimi series. Further collect- ing is needed to detennine whether or not this species is established in southern British Columbia.
Genus Dolichocis Dury
Dolichocis Dury, 1919: 158; Lawrence, 1965: 289 (complete synonymy). Type species, by mono- typy, Dolichocis manitoha Dury, 1919: 158.
North American Ciidae • Lawrence 483
Included species. Dolichocis indistinctus Hatch [northern North America, see p. 483]; Cis Jaricinus MeUie, 1848: 355 [Europe]; Dolichocis manitoha Dury [northern North America, see p. 483]; Ennearthron ijuasai Chujo, 1941: 85 [Japan]. Totah 4 species.
Dolichocis represents a small aggregate of species with 9-segmented antennae, elongate, cylindrical form, and simple pro- tibial apices. Like Plesiocis, it should be included within the genus EnneaHhron as currently defined, but it is here retained as a distinct group, which has independently undergone antennal reduction. The genus is further discussed in a previous paper (Lawrence, 1965).
Key to the North American Species of Dolichocis
1. Pronotal punctation coarser and denser, the punctures more than 0.25 X as large as scutellar base and usually separated by less than 0.33 diameter; elytral punctures dis- tinctly larger tlian pronotal punctin-es; elytral bristles 3 to 4 X as long as wide and about 0.33 X as long as scutellar base; vertex of male with median, raised, pubescent fovea; aljdominal fovea of male margined and
located in center of sternite III
D. manitoha (p. 483)
— Pronotal punctation finer and sparser, the punctures less than 0.25 X as large as scu- tellar base and usually separated by more than 0.33 diameter; elytral and pronotal punctures subequal in size; elytral bristles 2 to 3 X as long as wide and about 0.17 X as long as scutellar base; vertex of male simple; abdominal fovea of male not mar- gined and located anterad of center
D. indistinctus (p. 483)
Dolichocis indistinctus Hatch
Dolichocis indistinctus Hatch, 1962: 234; Law- rence, 1965: 288. Type locahty: "Stanley, B. C." Holotype, S, CAS.
Distribution. Known from scattered localities throughout the northern and montane parts of the continent, from the northern coast of British Columbia to the Gaspe Peninsula of Quebec and south into the Sierra Nevada, Chiricahua Mountains
of southeastern Arizona, and the Green Mountains of Vennont. Marginal records: BRITISH COLUMBIA: Terrace; 36 mi. N Radium; QUEBEC: Mt. Lyall, 1500 ft.; VERMONT: Peru, Bennington Co.; COLORADO: 28 mi. NW Kremmling, Grand Co.; ARIZONA: Rustler Park, 8 mi. W Portal, Cochise Co.; CALIFORNIA: Caspar, Mendocino Co.; 6 mi. NW Fish Camp, Mariposa Co.
Host fungi. Fomes pinicola [6(3)]; Trametes serialis [2(2)]; Fomes officinalis
[1].
Discussion. This species may be dis- tinguished from D. maniioha by the finer pronotal punctation, shorter and stouter bristles, and simple vertex in the male. The only species of Cis likely to be confused with D. indistinctus is C. robiniophilus, in which the antennae are 10-segmented, the protibial apex produced and dentate, the bristles somewhat longer, and the surface shinier.
Dolichocis indistinctus is closely related to and probably conspecific with Cis lari- cinus Mellie, which is included in the genus Ennearthron by European workers {see p. 480). Its range is broadly sympatric with that of C. manitoha and it occurs on the same fungi, but C. indistinctus appears to be much rarer.
Dolichocis manitoba Dury
Dolichocis manitoha Dury, 1919: 158; Griddle, 1921: 80 (dist.); Weiss and West, 1921a: 61 (dist., biol.); Hatch, 1962: 233, pi. 48, fig. 6, 6a (dist.); Lawrence, 1965: 288; Pielou, 1966: 1235 (dist., biol.); Pielou and Matthewman, 1966: 1310 (dist., biol.); Pielou and Verma, 1968: 1284 (dist., biol). Type locality: "Aweme, Manitoba." Holotype, 9 , Dury Coll., CIN.
Distribution. Northern and montane re- gions of North America, from the northern coast of British Columbia to New Bruns- wick and south to the central California coast, the southern Sierra Nevada, south- eastern Utah, and northern Pennsylvania (Fig. 92). Marginal records: BRITISH CO-
484 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
LUMBIA: Terrace; ALBERTA: Edmon- ton; NORTHWEST TERRITORY: Fort Smith; MANITOBA: Victoria Beach; NEW BRUNSWICK: Matapedia; PENNSYL- VANIA: Twin Lakes, Pike Co.; WISCON- SIN: Minocqua, Oneida Co.; SOUTH DAKOTA: Black Hills, Hamey Nat. For., Pennington Co.; COLORADO: 28 mi. N Kremmling, Grant Co.; UTAH: Wickiup Pass, 9000', Henry Mts., Carfield Co.; CALIFORNIA: Huntington Lake, Fresno Co.; Ben Lomond, Santa Cniz Co.
Host fungi. Fomes pinicola [50(21)]; Polijporus betulinus [5(3)]; Polyponts pargamemis [4]; Fomes annosus [3(2)]; Daedalea unicolor [2(1)]; Fomes offici- nalis [1(1)]; Ganoderma apphnatum [1(1)]; Polijporus sulphureus [1(1)]; Fomes fomentarius [1]; Polypoi-tis adustus [1]; Polijporus tuUpiferae [1]; Polijporus volvatus [1].
Discussion. This species is easily dis- tinguished from D. indistinctus by the coarser punctation, longer and finer bristles, and fovea on the head of the male. It may be confused with Sulcacis cuHulus, which has similar punctation and vestiture, but in that species the fonn is somewhat less elongate (EL/EW less than 1.50), the antennae are 10-segmented, and the apex of the protibia is expanded, rounded, and spinulose. Cis hijstriculus and C. angustiis are also somewhat similar, but the pro- tibial apex is angulate or dentate, the an- tennae are 10-segmented, and the pronotum is not distinctly narrowed anteriorly.
Dolichocis manitoba is a common inhabi- tant of Fomes pinicola and its relatives in western North America, but in the eastern part of the continent it appears to have a broader host range.
Genus Orthocis Casey
Orthocis Casey, 1898: 84; Lawrence, 1965: 288 (complete synonymy). Type species, by sub- sequent designation, Orthocis aterrima Casey, 1898: 84 ( =Cis punctatus Mellie, 1848: 337) (Lawrence, 1965: 288).
Cis (Mellieicis) Lohse, 1964: 122; Lawrence,
1965: 288. Type species, by original desig- nation, Cis alni Cyllenhal, 1813: 386. Cis (Orthocis) Casey, — Lohse, 1967: 285.
Included species. Cis ahijssinicus Guerin- Meneville, 1847: 325 [Ethiopia]; Cis aequalis Blackburn, 1888: 268 [Austraha]; Cis alni Cyllenhal, 1813: 386 [Eurasia]; Cis alnoides Reitter, 1884: 120 [southeastern Europe]; Ennearthron annulatum Kraus, 1908: 80 [Cuba]; Cis apicipennis Pic, 1916: 5 [Brazil]; Cis assimilis Broun, 1880: 347 [New Zealand]; Cis collenettei Blair, 1927: 166 [Marquesas]; Cis coluber Abeille de Perrin, 1874a: 52 [Europe]; Cis- cijlindrus Gorham, 1886: 358 [Panama]; Cis discoi- dalis Pic, 1922: 1 [Cameroons]; Cis flavi- pennis Pic, 1923: 12 [Indo China]; Cis giiamae Zimmerman, 1942: 49 [Guam]; Orthocis huesanus Kraus, 1908: 77 [Florida, see p. 485]; Cis iminaturus Zimmennan, 1939: 346 [Hawaii]; Cis insularis Water- house, 1876: 177 [Rodriguez]; Cis juglandis Reitter, 1885: 208 [southeastern Europe]; Cis lacernatus Reitter, 1908: 121 [eastern Africa]; Cis hanus Blackburn, 1907: 285 [Australia]; Cis linearis Sahlberg, 1901: 10 [northern Europe]; Orthocis longula Dury, 1917: 13 [eastern U. S., see p. 486]; Cis lucasi Abeille de Perrin, 1874b: 62 [south- em Europe, northern Africa]; Cis m- nigrum Champion, 1913: 161 [Mexico]; Cis nigrosplendidus Nobuchi, 1955: 105 [Japan]; Cis ornatus Reitter, 1877: 381 [Japan]; Cis perrisi Abeille de Perrin, 1874a: 53 [southern Europe]; Orthocis platensis Brethes, 1922: 302 [Argentina]; Cis pseudolinearis Lohse, 1965: 179 [Europe]; Orthocis pulcher Kraus, 1908: 78 [Florida, see p. 486]; Cis punctatus Mellie, 1848: 337 [North America, see p. 486]; Cis schizophijlli Nakane and Noliuchi, 1955: 47 [Japan]; Cis sublacernatus Scott, 1926: 24 [Seychelles]; Cis sidmrnatus Wol- laston, 1861: 140 [southern Africa]; Cis testaceofasciatus Pic, 1922: 2 [Guade- loupe]; Ennearthron transversatum Kraus, 1908: 79 [southeastern U. S., see p. 488]; Cis undulatus Broun, 1880: .347 [New Zea-
North American Ciidae • Lawrence 485
land]; Cis icoUastonii Mellie, 1849: 86 [Madeira]; Cis zoufali Reitter, 1902b: 6 [southeastern Europe]. Total: 38 species.
The genus Orthocis was discussed at length in an earlier paper (Lawrence, 1965) and arguments were presented for affording the group generic rank. Lohse (1964, 1965, 1967) prefers to consider Orthocis as a subgenus of Cis and includes within it certain species that I would ex- clude, namely Cis festiviis (Panzer), C. pygmaeiis (Marsham), and C vestitus Mellie. As it is here delimited, the genus consists of 38 described species and at least 30 undescribed forms from various parts of the world.
Members of the genus Orthocis are usu- ally distinguished from other ciids by the elongate form, 10-segmented antennae, simple and rounded protibial apices (Fig. 45), vestiture of short and fine hairs, elytral suture with inflexed margin at apex (Fig. 38), single and confused elytral punctation, well-developed and somewhat flattened prosternum with a fairly broad intercoxal process (Fig. 22), and complete absence of horns or tubercles on the head or pro- notum of the male. Some species have 9- segmented antennae (O. anmilatns, O. pollidus ) , while others may be clothed with stouter bristles (O. abijssinicus, O. coluber, O. hicasi), and a few are broader (O. col- lenettei). Lack of male armature is uni- versal in the group, and the few species of Cis that have simple protibial apices and resemble Orthocis in other ways, are char- acterized by having teeth or tubercles on the male clypeus. There is another type of secondary sexual character in Orthocis, however, that is not found in other groups of ciids. In a number of species, the clypeus of the male is covered with setae or hairs that are longer and denser than those on the surroimding parts of the head ( Fig. 37 ) . The inflexed elytral margin also occurs in Strigocis, but members of that genus have a carinate prosternum and spinose protibial apices.
Key to the North American Species of Orthocis
1. Antennae 9-segmented; elytra bicolored, black or dark brown with median, yellow, transverse band; pronotal punctation coarse and dense, punctures 0.25 X as large as scutellar base and separated by 0.33 diameter or less; lateral pronotal margins narrow, not visible for their entire lengths from above;
TL 1.4 mm or less
O. transversatus (p. 488)
— Antennae 10-segmented; without otlier char- acters in combination 2
2. Size smaller, TL less than 1.25 mm; pro- notal punctation coarse and sparse, punc- tures about 0.50 X as large as scutellar base and separated by 0.50 diameter or more; lateral pronotal margins not visible for their entire lengths from above; elytra bicolored, brownish with median, yellow, transverse band O. piilcher (p. 486)
— Size larger, TL more than 1.25 mm; pronotal punctation finer or denser; elytra unifonnly pigmented or lateral pronotal margins easily visible for their entire lengtlis from above — . 3
3. Lateral pronotal margins narrow, not visible for their entire lengths from above, diverging anteriorly and abruptly converging near apex, so that they appear angulate (Fig. 19); body elongate, TL/EW more than 2.65
and EL/EW more tlian L80
O. longulus (p. 486)
— Lateral pronotal margins broader, easily visible for their entire lengths from above, weakly rounded or subparallel, not angulate, witli raised lip (Fig. 18); body shorter and broader, TL/EW less dian 2.65 and EL/EW less than 1.80 4
4. Elytra bicolored, each elytron bearing 2
large, yellow spots, Florida Keys
O. huesanus (p. 485)
— Elytra uniformly pigmented
O. punctatus (p. 486)
Orthocis huesanus Kraus
Orthocis huesanus Kraus, 1908: 77, pi. 3, fig. 3; Lawrence, 1965: 283. Tvpe locality: "Key West, Fla." Holotype, $, USNM.
Distribution. Known only from Key West, Monroe Co., FLORIDA.
Host fungi. Unknown.
Discussion. This species is quite similar to O. punctatus as it is here defined, differ- ing mainly in the slightly flatter body and bicolored elytra. The general form and
486 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
coloration are also similar to O. m-nigrum (Champion) described from Mexico.
Orthocis longulus Dury
Orthocis longula Dury, 1917: 13; Lawrence, 1965: 283. Type locality: "Pennsylvania..." [Ling- lestown]. Holotype, 9, Dury Coll., CIN.
Distribution. Knowai from scattered lo- calities in the southeastern part of North America, from southern Pennsylvania to Florida and west to Kansas and the Gulf Coast of Texas and northeastern Mexico. Localities: PENNSYLVANIA: Linglestown, Dauphin Co. FLORIDA: Gainesville, Alachua Co.; Crescent City, Putnam Co.; Enterprise, Volusia Co.; Lakeland, Polk Co.; Dunedin, Pinellas Co. ALABAMA: Spring Hill, Mobile Co. MISSISSIPPI: Lucedale, George Co. LOUISIANA: Hara- han, Orleans Par.; Belle Chasse, Plaque- mines Par. KANSAS: (no specific locality). TEXAS: Columbus, Colorado Co.; Vic- toria, Victoria Co. TAMAULIPAS: Tam- pico.
Host fungi. Unknown.
Discussion. This species is unique among the North American Ciidae in the very long and narrow body form and the narrow and somewhat angulate lateral pronotal margins. Some specimens of O. punctatus have fairly narrow pronotal margins, l)ut these are never angulate and the body is always shorter and broader than that of O. longulus. The species is most closely related to O. cijlindrus (Gorham) from Panama.
Orthocis pulcher Kraus
Orthocis pulcher Kraus, 1908: 78, pi. 3, fig. 4; Lawrence, 1965: 283. Type locality: "Key West, Fla." Holotype, $, USNM.
Distribution. Southern Florida. Locali- ties: FLORIDA: Punta Gorda, Charlotte Co.; Coconut Grove, Dade Co.; Key West, Monroe Co.
Host fungi. Unknown.
Discussion. This species is easily dis- tinguished by the small size, elongate body
form, narrow lateral pronotal margins, very large pronotal punctures, and bicolored elytra. It is very similar to O. testaceo- fasciatus (Pic) described from Guadeloupe, and the two species may be synonymous.
Orthocis punctatus (Mellie)
Cis punctatus Mellie, 1848: 337. Type locality: "Amerique boreale." Holotype, i , Marseul Coll., MNHN.
Orthocis punctata (Mellie), Casey, 1898: 84 (dist): Cockerell and Fall, 1907: 185 (dist.); Blatchley, 1910: 899 (biol.); Dury, 1917: 13; Brown, 1931: 90 (dist.); Hatch, 1962: 231 (dist., biol.); Lawrence, 1965: 283.
Orthocis aterrima Casey, 1898: 84; Dury, 1917: 13; Lawrence, 1965: 283. Type locality: "Cali- fornia ( Alameda Co. ) ." Holotype, $ , Casey Coll., USNM. NEW SYNONYMY.
Distribution. Widespread across North America from the edge of the Beaufort Sea (Mackenzie District) to the Island of Newfoundland, south on the Pacific Coast to Santa Barbara, California, through the Rocky Mountains to northeastern New Mexico, into the Black Hills of South Dakota, and through the eastern and mid- westem states from New England to Florida and west to Kansas and Texas. Marginal records: NORTHWEST TERRI- TORY: Reindeer Depot, Mackenzie Dis- trict. ALBERTA: McMurray. MANITOBA: Winnipeg. ONTARIO: Biscotasing. NEW- FOUNDLAND: Hannon Field. FLOR- IDA: Biscayne, Dade Co. TEXAS: Kerrville, Kerr Co.; Victoria, Victoria Co. KANSAS: Riley Co. SOUTH DAKOTA: Hill City, Custer Co. NEW MEXICO: Las Vegas Hot Springs, San Miquel Co. CALIFORNIA: Santa Barbara, Santa Barbara Co. This is probably a composite distribution, since two or more species may be involved (see discussion below).
Host fungi. Auricularia auricula [1(1)]. Collected beneath the bark of various trees, including species of Finns, Abies, Pruntis, Quercus, Ricinus, and Fagus. Also taken by beating branches and in "powdery fmigus" and "shelf fungus."
Discussion. Orthocis punctatus, as it is
North American Ciidae • Lawrence 487
here delimited, is extremely variable and Surface sculpture may be coarsely
may represent a complex of two or more granulate and dull to smooth and shiny.
related species. It may be distinguished In specimens with coarser and denser
from other North American Orthocis by punctation, the interspaces are usually
the 10-segmented antennae, miiform elytral smoother and shiny.
coloration, and relatively broad lateral pro- Tlie length of the pronotum relative to
notal margins, which are raised at the edge that of the elytra exhibits considerable
to fonn a narrow lip (Fig. 18). The actual variation. EL/PL varies from 1.80 to 2.60.
widths of the pronotal margins vary con- The larger pronotum is usually found in
siderably, but they are never angulate as southern populations.
in O. low^ulus (Fig. 19). The total size is also variable in O.
Characters exhibiting the most noticeable ptinctatus, TL ranging from 1.30 to almost
variation are: 1) size and density of pro- 3.00 mm. Larger specimens are usually
notal punctures, 2) widths of lateral found in northern and western populations
pronotal margins, 3 ) shape and prominence and the smallest individuals are from
of anterior pronotal angles, 4 ) type of Florida and Texas.
pronotal surface sculpture, and 5) relative There is little variation in vestiture, all
size of pronotum. A comparative study of specimens examined being clothed with
the male genitalia has not been made and very short and fine hairs. The coloration
must await a more complete analysis of is also fairly uniform throughout the range,
the genus Orthocis. It is likely that two species are included
The pronotal punctation varies from fine among this material and a third may even and sparse to fairly coarse and dense. The be found in the southeastern collections, punctures in some specimens are 0.08 X In specimens from the western and north- as large as the scutellar base and separated em parts of the continent the size is larger, by more than 1 diameter, while in others the pronotum smaller, dull, finely and they may be as much as 0.25 X as large as sparsely punctate, and the lateral pronotal the scutellar base and separated by 0.20 margins are broad and somewhat rounded, diameter. Finer and sparser punctation is with prominent anterior angles. The type characteristic of western and northern specimens of both O. punctatus and O. populations. aterrimus are of this form. The second
Lateral pronotal margins may be rel- form may be found throughout the eastern atively narrow and only barely visible from and midwestern part of the continent and above or they may be quite broad and south to Florida and the Gulf Coast. This explanate. From above they may .be is more variable but usually has a larger rounded and convex, subparallel, or even prothorax, with a shiny, coarsely and slightly sinuate at middle. The anterior densely punctate pronotum, narrower, sub- angles in some specimens are strongly pro- ^^^^n^^ j^^j^,^.^^j ^^^,^^^^.^^^ ^^,.^j^ 1^^^ prominent
lecting and rounded, while in others they ^ . ^ r^ ^ i. r /n
' 1^ 1 . ,. 11 . . 1 . anterior angles. Casey s concept ot U.
are barely proiectmg and almost right. iii -iiii o-
c • c a 4.1 4- TT ^ J punctatus would be inchided here, bmce
Specimens rrom the southeastern United ;
States and the Gulf Coast of Texas usually l^^tle is known of the biology of either form
have narrower pronotal margins and less and since they cannot be consistently dis-
prominent anterior angles. In some south- tinguished from one another, I prefer to
em populations, the lateral margins may l"mp them under one name for the present,
diverge slightly towards the apex of pro- Orthocis punctatus is most closely related
notum. Narrower lateral margins are usu- to the Palaearctic species O. alni (Gyllen-
ally correlated with coarser and denser hal), O. linearis (Sahlberg), and O. pseudo-
punctation. linearis (Lohse), and the latter two differ
488 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
from O. alni in much the same way as the two forms above differ from one another (see Lohse, 1965, 1967).
Orthocis fransversatus (Kraus), NEW COMBINATION
Emicaiiliwn transversatum Kraus, 1908: 79, pi. 3, fig. 7; Lawrence, 1965: 283, 286. Type lo- cality: "Crescent City, Fla." Holotype, $ , USNM.
?Ennearthwn pallidum Kraus, 1908: 79, pi. 3, fig. 6; Lawrence, 1965: 283. Type locality: "Haw Creek, Fla." Holotype, 9 , USNM ( type lost). NOMEN DUBIUM.
Orthocis pulcher Kraus, — Blatchley, 1923: 19 (dist, biol.); Blatchley, 1928: 68 (dist., biol.). Misidentification.
Distribution. Southeastern United States, from southern Virginia to Florida and west to the Gulf Coast of Texas. Marginal records: VIRGINIA: Virginia Beach, Prin- cess Anne Co.; FLORIDA: Dunedin, Pinellas Co.; TEXAS: Dallas, Dallas Co.; Columbus, Colorado Co.
Host fungi. Unknown. Bred from decay- ing rattan vines in Virginia. Collected by beating dead branches of Quercus and from a dead limb of "red-bay" in Florida.
Discussion. Tliis species may be dis- tinguished from other North American Orthocis by the 9-segmented antennae, small size, coarse and dense pronotal punc- tation, and bicolored elytra. It resembles certain small, bicolored Cis, such as C. krausi, C. superhus Kraus, and C. atro- maculatus Pic, but in all of these the vestiture consists of stouter bristles rather than fine hairs, the antennae are 10-seg- mented, and the protibial apices are angu- late or dentate. The type of Kraus's Ennearthron pallidum was lost and the description is inadequate: thus the name is doubtfully synonymized above and con- sidered a nomen duhium. O. transversatus is closely related to O. annulatus (Kraus) from Cuba, which also has 9-segmented antennae and may not be .specifically dis- tinct.
Genus Strigocis Dury
Strigocis Dury, 1917: 18; Leng, 1920: 247; Arnett, 1962: 829. Type species, by monotypy, Stri- gocis opacicollis Dury, 1917: 20.
Cis Latreille (in part), — Mellie, 1848: 356; Abeille de Perrin, 1874b: 75; Reitter, 1878c: 33; Gorham, 1883: 222; Reitter, 1902a: 57; Dalla Torre, 1911: 6; Winkler, 1927: 793; Kevan, 1967: 143.
Rhopalodontus Mellie (in part), — Fowler, 1890: 212; Peyerimhoff, 1915: 26; Horion, 1951: 321; Nobuchi, 1960: 39.
Xestocis Casey (in part), — Casey, 1898: 86; Dury, 1914: 18-19; Leng, 1920: 247; Arnett, 1962: 829.
Enttjpus (Entijpocis) Lohse (in part), — Lohse, 1964: 121.
Sulcacis {Entijpocis) Lohse (in part), — Lohse, 1967: 284.
Sulcacis Dury (in part), — Lawrence, 1965: 277.
Included species. Cis hicornis Mellie, 1848: 356 [Europe]; Cis bilimeki Reitter [Mexico, see p. 490]; Strigocis opocicollis Dury [eastern North America, see p. 490]; Xestocis opalescens Casey [eastern North America, see p. 490]; Rhopalodontus toku- nagai Nobuchi, 1960: 39 [Japan]. Total: 5 species.
In a previous paper (Lawrence, 1965), 1 included Cis hicornis Mellie in the genus Sulcacis, although specimens had not been seen at that time. Rhopalodontus tokunagai Nobuchi was tentatively placed in the same genus. Having examined specimens of C hicornis and reviewed the description of R. tokunagai, I would now place both species in the North American genus Stri- gocis, originally proposed by Dury (1917) for S. opacicollis. Members of both genera are characterized by having the protibial apex spinose, but in other respects the groups are quite distinct and probably dis- tantly related. The main differences be- tween Strigocis and Sulcacis are listed in the table below.
In addition to the species mentioned above, Strigocis includes the North ' American Xestocis opalescens Casey, the Mexican Cis bilimeki Reitter, and at least three more vmdescribed fonns from Central America. Members of this genus resemble
North American Ciidae • Lawrence
489
Table 4. Differences between Strigocis and Sulcacis.
Characters
Strigocis
Sulcacis
Vestiture
Antennae
Prostemum ( in front of coxae)
Intercoxal process of prostemum
Pronotal hypomera
Anterior angles of pronotiim
Lateral margins of pronotum
Anterior edge of pronotum in male
Frontoclypeal ridge in male
Maxillary palps
Elytral suture
Metasternal suture
single confused fine hairs or bristles
10-segmented
tumid to carinate and on same level as intercoxal process (Fig. 14)
parallel-sided
strigose
angulate or
produced (Fig. 14)
broader, usually visible from above
produced and emarginate, usually with 2 processes
produced and emarginate
short and stout
with inflexed margin at apex (Fig. 38)
longer
dual
confused or seriate
brisdes
9- or 10-segmented
concave or biconcave and on different level than intercoxal process (Fig. 15)
tapering behind
not strigose
rounded and not produced (Fig. 15)
narrower, not
visible from above
simple and rounded
bearing 2 teeth or tubercles
long and narrow
without inflexed margin
shorter
various Cis species having a strongly tumid or carinate prostemum ( Cis tricornis group or Cis nitidus group), but differ from them by virtue of the spinose protibial apex and inflexed elytral suture. Species of Ceracis have spinose protibial apices, but differ in having a concave prostemuin with laminate intercoxal process, rounded anterior pro- notal angles, and simple elytral suture.
Key to the North American Species of Strigocis
1. Vestiture of very short, stout bristles, which are le^s than 2.0 X as long as wide; pronotal punctation fine and dense, punctures less than 0.10 X as large as scutellar base and separated by less than 0.75 diameter; pro- notal surface distinctly granulate and dull;
abdominal fovea in male more tlian 0.60 X as long as body of stemite III; known from highlands of Mexico and probably oc- curring in southern Arizona
-- S. bilimeki (p. 490)
— Vestiture of finer hairs; pronotal punctures separated by more than 0.75 diameter or punctures more than 0.15 X as large as scu- tellar base; abdominal fovea less than 0.60 X as long as body of sternite III; eastern United States and northern Mexico 2
2. Body shorter and stouter, EL/EW less than 1.33; elytral hairs colorless, less than 0.15 X as long as scutellar base and not or barely visible under 10 X magnification; lateral margins of pronotum broader and with a raised lip; pronotal surface usually shiny . _ S. opalescens (p. 490)
— Body longer and narrower, EL/EW more than 1.33; elytral hairs yellow, more than 0.20 X as long as scutellar base and easily
I
490 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
visible under 10 X magnification; lateral margins of pronotnm narrower and without
raised lip; pronotal surface usually dull
S. opacicoUis ( p. 490 )
Strigocis bilimeki (Reitter), NEW COMBINATION
Cis bilimeki Reitter, 1878a: 33; Gorliam, 1883: 222. Type locality: "Mexico." Lectotype, $, Obertluir Coll., MNHN.
Distribution. Mountains of northern Me.xico; probably extending into southern Arizona. Marginal records: DURANGO: 11 mi. SW El Salto; HIDALGO: Tenango de Doria; TAMAULIPAS: Rancho del Cielo, 370(X, NW Gomez Farias.
Host fungi. Lenzites hetulina [1(1)]; Pohjporus versicolor [1].
Discussion. Individuals of this species are somewhat larger than those of S. opaci- coUis and S. opalesccns (TL usually greater than 1.5 mm) and are easily distinguished by the short, stout, yellowish bristles, fine and dense pronotal punctation, dull sur- face, and large abdominal fovea in the male. The predicted occurrence of S. bilimeki in the Southwest is based on a collection from southern Arizona that was discarded because it was thought to be a contaminant population of S. opacicoUis. Specimens and records were not retained, but I now think it is more likely that this "contaminant" collection was S. bilimeki, present in the original collection as a few larvae and thus overlooked. Further col- lecting in Arizona should verify this.
Strigocis opacicoUis Dury
Strigocis opacicoUis Dury, 1917: 20; Weiss and West, 1920: 8 (dist., biol.). Type locality: "Cincinnati, Ohio." Syntypes, Dury Coll., CIN.
Distribution. Eastern North America, from southern Vermont and New Hamp- shire to the Florida Keys, west as far as eastern Kansas, and south into Mexico (Fig. 100). \hirginal records: NEW YORK: N. Fairhaven, Cayuga Co. VER- MONT: Manchester, Bennington Co. NEW HAMPSHIRE: 7 mi. NW W ilton, Hillsboro
Co. FLORIDA: Chipola Park, Dead Lake, Calhoun Co.; Pkmtation Key, Monroe Co. ARKANSAS: Washington Co. KANSAS: Onaga, Pottawatomie Co. NUEVO LEON: Chorros de Agua, 13 mi. W Montemorelos. HIDALGO: Tenango de Doria. VERA- CRUZ: 20 mi. E Cordoba; Orizaba.
Host fungi. Polyporus versicolor [29 (14)]; Lenzites betulina [7(1)]; Polyporus hirsutus [5]; Polyporus maximus [1(1)]; Polyporus pinisitus [1(1)]; Polyporus sub- ectypus [1(1)]; Polyporus adustus [1]; Polyporus gilvus [1]; Polyporus sanguineus [1]; Polyporus tenuis [1].
Discussion. This is the most common and widespread species of Strigocis in North America. It differs from S. opalescens in the more elongate fonn, narrower lateral pronotal margins, and the size and color of the hairs. It is similar to S. bilimeki in general form and male armature (fronto- clypeal ridge and pronotal apex each form- ing an emarginate plate), but the vestiture is quite different and the pronotal puncta- tion is much denser.
In eastern North America, S. opacicoUis is sympatric with S. opalescens, and the two may feed on the same fungi; the latter species, however, appears to be much rarer. In northern Mexico, all three Strigocis oc- cur in the same areas along with two undescribed species. All appear to frequent the same fungi (Polyporus versicolor and its relatives) and as many as three have been taken on the same fniiting bodv.
Individuals from Plantation Key differ from those to the north in being smaller in size (TL = 1.0 to 1.3 mm) with the pro- notal surface shiny and reddish in color, contrasting with the blackish elytra. Tliis may be a dwarf race or a distinct species. Specimens from Veracruz are somewhat similar.
Strigocis opalescens (Casey), NEW COMBINATION
Xestoci.s opalesceius Casey, 1898: 86; Dury, 1917: 17 (dist.). Type locality: "Pennsylvania (West- moreland Co.)" [St. Vincent]. Holotype, $, Casey Coll., USNM.
North American Ciidae • Lawrence 491
Xestocis dacisi Duiy, 1917: 16. Type locality: ■'Staten Island, N. Y." Holotype, S, Dury Coll., CIN. NEW SYNONYMY.
Distiibution. Eastern North America, from southern Michigan and New York south and west as far as northeastern Mexico. Marginal records: MICHIGAN: Detroit, Wayne Co.; NEW YORK: Staten Is., Richmond Co.; NORTH CAROLINA: 2.5 mi. NW Highlands, Macon Co.; MIS- SISSIPPI: Meridian, Lauderdale Co.; TAMAULIPAS: Rancho del Cielo, 370(r, NW Gomez Farias.
Host fungi. Pohjporus versicolor [3(1)].
Discussion. Individuals of this species are shorter, broader, and more rounded than those of either S. opacicollis or S. bilinicki, the lateral pronotal margins are broader and somewhat raised, and the vestiture consists of vei-v short and fine, pale hairs. Although it occurs throughout eastern North America and into Mexico, S. opolescens appears to be rare and there are few specimens in collections.
Genus Hadraule Thomson
Hadraule Thomson, 1863: 182; Lawrence, 1965: 282 (complete synonymy); Lohse, 1967: 295. Type species, b)' monot\py, Cls elongatulus Gyllenhal, 1827: 627.
Maphoca Casey, 1900: 165. Type species, by monotvpy, Maphoca blaisdcUi Casey, 1898: 165. NEW ' SYNONYMY.
Mapheae Dalla Torre, 1911: 21. Incorrect sub- sequent spelling.
Diphyllocis Reitter, — Dury, 1917: 4; Leng, 1920: 247; Arnett, 1962: 829; Hatch, 1962: 235^ (not Reitter, 1885: 209).
Pityocis Peyerimhoff, 1918: 141. Type species, by monotypy, Pityocis coarctatus Peyerimhoff, 1918: 142 {=Cis elongatulus Gyllenhal).
EuneartJiwn {Knahlia) Roubal, 1936: 53. Type species, by monotypy, Cis elongatulus Gyllenhal.
Inchided species. Maphoca hlaisdcUi Casey [western North America, see p. 491]; Cis elongatulus Gyllenhal [Eurasia and nortlieastem North America, see p. 492]; Hadraule explanata Lawrence, n. sp. [northeastern North America, see p. 493]. Total: 3 species.
The only species currently included in
the genus Hadraule is the type, H. elonga- tula, although several others have been added at one time or other {see Lohse, 1964 and Lawrence, 1965). Casey's Maphoca hlaisdelli was placed in a separate genus because of the 2-segmented anten- nal club, and Dury (1917) transferred the species to Diphyllocis for the same reason. Actually, M. hlaisdelli is similar to H. elongatula in several respects, including prothoracic structure (Fig. 28), and the two appear to be congeneric in spite of the slight reduction of the fii^st club segment in the fonner. Reitter's genus Diphyllocis, on the other hand, is based on a distinct and unrelated European species, D. opacu- lus (Reitter). Members of Hadraule are easily recognized l:)y the small size, flat- tened fonn, 9-segmented antemiae, and elongate prostemum.
Key to the North American Species of Hadraule
1. Pronotum slightly narrowed apically, widest behind middle, the margins broad and ex- planate, easily visible for their entire lengths from above (Fig. 21); head strongly de- chned, only partly visible from above; size
larger, TL more than 1.5 mm
^^ H. explanata (p. 493)
— Pronotum slightly broader apically, widest at anterior fourth, the margins narrow, not or barely visible for their entire lengths from above (Fig. 20); head only weakly de- clined, \isible from above; size smaller, TL less than 1.5 mm 2
2. Antennal club apparently 2-segmented, the first club segment barely larger than the last funicular segment; pronotum reddish, much
lighter in color tlian elytra
H. hlaisdelli (p. 491)
— Antennal club 3-segmented; pronotum and
elytra uniformly dark in color
H. elongatula ( p. 492 )
Hadraule blaisdelli (Casey), NEW COMBINATION
Maphoca blaisdelli Casey, 1900: 165; Lawrence, 1965: 279. Type locality: "California ( Mokel- umne Hill, Calaveras Co.)." Holotype, 9, Casey Coll., USXM.
Diphyllocis blaisdcUi (Casey), — Leng, 1920: 247; Griddle, 1926: 98 (dist.); Hatch, 1962: 235, pi. 48, fig. 8 (dist., biol.).
492 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
Distribtition. Western North America, from southern British Cokimbia to south- em CaHfornia, east through Utah, Arizona, and New Mexico into Texas, and south into Mexico. Also known from Iowa, Ohio, Michigan, Massachusetts, Connecticut, and Florida. Marginal records: BRITISH COLUMBIA: Peachland; CALIFORNIA: L5 mi. NW Mt. Laguna, San Diego Co.; UTAH: Logan, Cache Co.; ARIZONA: 15 mi. E Douglas, Cochise Co.; NEW MEXI- CO: Deming, Luna Co.; TEXAS: San Antonio, Bexar Co.; VERACRUZ: San Juan de la Punta; IOWA: Iowa City, Johnson Co.; OHIO: Mt. Washington, Cincin- nati, Hamilton Co.; MICHIGAN: Sauga- tuck, Allegan Co.; MASSACHUSETTS: Amherst, Hampshire Co.; CONNECTI- CUT: New Haven, New Haven Co.; FLORIDA: (no specific locality). Only one of the northeastern records ( Saugatuck, Michigan) is known to be based on field collected specimens; the others are prob- ably herbarium infestations {see below).
Ho.st fungi. Lenzites saepiaria [3(2)]; Fohjporus adiistus [3(1)]; Tmmctes sepiuni [3]; Pohjponis munzii [2]; Pohjponis versi- color [2] ; Daedalea confragosa [1(1)]; Fames annosus [1(1)]; Fohjporus anceps [1(1)]; Fohjporus biformis' [1(1)]; Tra- metes hispida [1(1)]; Fames rahiniae [1]. Also recorded from Fleurotus sp. and Fohjporus cuticularis.
Discussion. This is the only widespread Hadraule in North America, the other two l)eing restricted to the extreme northeastern part of the continent. The species may be distinguished from both //. explunata and H. elongatida by the 2-segmented antennal club and reddish prothorax.
hi. bhisdeUi is fairly common in Cali- lomia, where it breeds on several different lungi, and it extends throughout the South- west and into southern Mexico. It has also been collected in Michigan and has been recorded as a herbarium pest from various localities in eastern North America. Speci- mens from Vancouver, British Columbia, were taken from a herbarium polypore and
the Amherst specimens are from a lichen collection. In the National Fungus Col- lections at Beltsville, Maryland, a large number of dead specimens were found in various boxes of fungi, but there is reason to believe that this is an old infestation originating at the Missouri Botanical Gar- den. Tliis is the only species of Ciidae that is kno\vn to be an herbarium pest.
Hadraule elongafula (Gyllenhal)
Cis dongatultis Cyllenhal, 1827: 627. Type lo- cality: "Spansatra Westrogotliiae" [Sweden]. Types, Gyllenhal Coll., Zool. Univ. Mus., Uppsala ?
Hadraule elongatida (Gyllenhal), — Thomson, 1863: 182.
Hadraida elongatida ( Gyllenhal ) , • — Pielou and Verma, 1968: 1184 (dist., biol.).
Ennearthron striatum J. Sahlberg, 1901: 11. Type locaHty: "Fennia austraU" [Finland]. Types, C. Sahlberg Coll., MZUH ?
Pitijocis coarctatus Peyerimhoff, 1918: 142. Type locality: "Ain-Haouas pres Djelfa." [Algeria]. Lectotype, $, Peyerimhoff Coll., MNHN.
Cis elongatus Schilsky, 1900: 59. Incorrect sub- sequent spelling.
Distribution. Widespread throughout Europe, Siberia, and North Africa. In North America, known only from Ludlow, NEW BRUNSWICK. Probably introduced from Europe.
Host fungi. Fohjporus betulinus [1(1)].
Discussion. This species may be dis- tinguished from //. blaisdelU by the 3-seg- mented antennal club and darker pro- notum. H. explunata is somewhat larger with broader pronotal margins. Smaller specimens of Cis strioJatus and its relatives may be confused with H. elongatida but they differ in having lO-seginented an- tennae and a shorter prostemum in front of the procoxae.
Hadraule elangatula has been recorded from various parts of the Palaearctic Region, but there is only a single record from North America. Although the New Brunswick population may represent an introduction, it is possible that this small and elusive beetle occurs throughout Canada. In the Old World, //. elongatula
North American Ciidae • Lawrence 493
has been taken in the galleries of Dorca- toma ( Anobiidae ) in Liriodendron, in bark beetle galleries (Scolytidae) in Ficea, in association with Finns halepensls in North Africa, and on a birch fungus in Scandi- navia (Reitter, 1878a; Peyerimhoff, 1918; Palm, 1946, 1952). This population was apparently breeding in the conks of Pohjporus betidimis.
Hadraule explanata NEW SPECIES
Holotype. ?, MAINE: E Machias, Wash- ington Co., June, Fenyes Collection [CAS].
Female. Length 1.55 mm. Body 2.48 X as long as broad, flattened, GD/EW 0.68. Head and pronotum reddish brown, elytra yellowish brown. Vestiture of very short, erect, colorless, fine bristles. Vertex flat- tened; frontoclypeal ridge simple. Anten- nal club 3-segmented. Pronotum (Fig. 21) 0.78 X as long as broad, widest at posterior fourth; anterior edge weakly rounded; sides subparallel, barely sinuate at middle, slightly converging anteriorly, the margins broad and distinctly crenulate, easily visible for their entire lengths from above; an- terior angles distinctly produced and rounded; disc slightly convex, even; surface distinctly granulate and dull; punctures 0.18 X as large as scutellar base and sepa- rated by 1.0 to 2.0 diameters. Elytra 1.76 X as long as broad and 2.44 X as long as pro- notum; sides subparallel, apices blunt; punctation dual and distinctly seriate; megapunctures at least 2.0 X as large as pronotal punctures, separated by less than 0.50 diameter between and within rows, subquadrate, shallow, and nude; micro- punctures located mainly between rows, each bearing a fine erect, colorless bristle, which is 0.20 X as long as scutellar base. Presternum slightly tumid; intercoxal proc- ess 0.62 X as wide as a jirocoxal cavity, parallel-sided. Protibial apex narrowly rounded.
DistrUmtion. Known only from the type locality.
Host fungi. Unknown.
Discussion. This species is easily dis- tinguished from the other two Hadraule by the broad and explanate lateral pronotal margins; it is also somewhat larger in size. In general form the species resembles Cis striolatus, but the latter differs in having 10-segmented antennae, angulate or den- tate protibia, and narrower lateral pronotal margins. At present only a single female specimen is known, but it is sufficiently distinct to warrant description at this time. It is hoped that further collecting in the Northeast will turn up the male and reveal something of the biology of this species.
The name explanatus is derived from the Latin word meaning to make level or flat (referring to the broad, explanate lateral pronotal margins).
Genus Plesiocis Casey
Plesiocis Casey, 1898: 87; Dalla Torre, 1911: 20; Dury, 1917: 22; Leng, 1920: 247; Arnett, 1962: 829; Hatch, 1962: 233. Type species, by monotypy, Plesiocis cribrum Casey, 1898: 87.
Ennearthron Mellie, — Hubbard, 1892: 254, 255.
Included species. Flesiocis cribrum Casey [western and northern North Amer- ica, see p. 494].
Excluded species. Flesiocis spenceri Hatch (see p. 482).
This genus contains the single species P. cribrum, which probably should be lumped together with the heterogeneous assemblage of foiTOs currently included in Ennearthron {see discussion on p. 480). I prefer to retain the genus Flesiocis, which has been consistently recognized by American authors, until a thorough generic revision is possible. The North American species of Ennearthron are easily distinguished from Flesiocis cribrum by the characters given in the generic key as well as by the smaller size and less robust body foiTn. The nar- row and parallel-sided intercoxal process of the prosternum is similar to that found in Ceracis, but the protibial apex bears a stout tooth as in species of Cis. Although the relationships of P. cribrum are obscure at present, there is little doubt that it repre-
494
Bulletin Museum of Coiupaiafive Zoology, Vol 142, No. 5
sents an independent line exhibiting reduc- tion in antennal segmentation.
P/es/'oc/s cribrum Casey
Plesiocis cribrum Casey, 1898: 87; Dury, 1917: 21 (dist.); Griddle, 1921: 80 (dist.); Weiss, 1923: 199 (biol.); Weiss and West, 1922: 199 (dist., biol.); Hatch, 1962: 234, pi. 48, fig. 5, 5a (dist, biol.). Type locality: "California (Mokel- iimne Hill, Calaveras Co.)." Holotype, S, Casey Coll., USNM.
Distribution. Northern and montane regions of North America, from British Cokimbia to the Gaspe Peninsula of Que- bec, south to the Laguna Mountains in California, the edge of the Colorado Plateau in Arizona and New Mexico, the Great Lakes Region, and the Shenandoah Moun- tains of Virginia (Fig. 89). Marginal records: BRITISH COLUMBIA: Salmon Arm; QUEBEC: Gaspe; VIRGINIA: Stokesville, Augusta Co.; MICHIGAN: East Tawas, Iosco Co.; WISCONSIN: (no specific locality); NEW MEXICO: 5 mi. W Luna, Catron Co.; ARIZONA: Williams, Coconino Co.; CALIFORNIA: 1.5 mi. NW Mt. Laguna, San Diego Co.
Host fungi. Polyporus volvatus [48 ( 30 ) ] ; Fohjpoms anceps [8(5)]; PoJt/poius albo- hiteiis [2].
Discussion. Tliis species may be dis- tinguished from most North American Ciidae by the 9-segmented antennae, fairly short and broad body fonn, dentate p)ro- tibial apex, coarse and dense punctation which is single and confused on the elytra, vestiture of short, stout bristles, and 4 sharp teeth on the clypeus of the male (Fig. 4). Cis laminatus is similar but has 10-seg- mented antennae and a raised, Insinuate plate on the clypeus of the male. Cis du- plex differs in antennal segmentation and male armature, and is usually smaller in size. Dolichocis monitoba has 9-segmented antennae and similar coarse punctation, but the body is elongate and cylindrical (EL/ EW usually more than 1.50), the protibial apices are simple, and the male sexual char- acters are different. EnneartJiron spenceri
and E. aurisquamosum are both similar to P. cribrum in antennal segmentation, pro- tibial apices, and general body form, but they differ from the latter in having a broader prostemal intercoxal process, longer metastemal suture, and distinctly notched clypeus in the male (Fig. 8).
Plesiocis cribrum is widespread and abundant throughout the coniferous forests of western North America and appears to be rare in the eastern part of the continent. It occurs primarily on the conks of Poly- porus volvatus but also breeds in P. anceps. At the southern end of its range, it shares the same habitat with C. duplex and the latter apparently replaces P. cribrum in southern Arizona.
Genus Ceracis Mellie
Ennearthron (Ceracis) Mellie, 1848: 375; Law- rence, 1967b: 95 (complete synonymy). Type species, by subsequent designation, Ennearth- ron (Ceracis) sallei Mellie, 1848: 377 (Law- rence, 1967b: 95).
Included species. See Lawrence (1967b: 97-98). Also included are Ceracis magister Lawrence, n. sp. [Florida, see p. 498] and C. pecki Lawrence, n. sp. [southeastern U. S., see p. 499]. Total: 42 species.
This genus has been treated in a previous paper ( Lawrence, 1967b ) and need not be discussed in detail here. In the following section, two new species are described, one species (C. similis) is added to the fauna (although not actually recorded from the United States), new keys to males and fe- males are presented, and significant addi- tions are made to distributions and host ranges based on recently accumulated data.
Key to the North American Species of Ceracis
Males (Abdominal sternite III with pubescent fovea)
1. Antennae 10-segmented - - 2
— Antennae with less than 10 segments 3
2. Elytral punctation distinctly seriate, the larger punctures forming relatively straight rows; pronotal apex l^earing 2 horns, each witli a rounded knob above it; abdominal
North American Ciidae • Lawrence 495
fovea slightly transverse; color of elytra usually reddish C singularis (p. 502)
— Elytral punctation not distinctly seriate; pronotal apex with a raised, emarginate lamina; abdominal fovea circular; color of
elytra usually black
C. magister (p. 498)
3. Antennae 9-segmented 4
— Antennae 8-segmented 12
4. Pronotal apex rounded or shallowly emar- ginate, without distinct tubercles, horns, or lamina 5
— Pronotal apex produced, forming a lamina or 2 horns or tubercles 7
5. Frontoclypeal ridge produced, forming a long, narrow, median horn; pronotal and elytral punctation very fine and sparse; metasternal suture more than 0.20 X as
long as median length of metasternum
C monocenis (p. 499)
— Frontyclypeal ridge simple or forming 2 rounded plates or tubercles; elytral puncta- tion much coarser and denser than pronotal punctation; metasternal suture less than 0.20 X as long as median lengtli of meta- sternum 6
6. Body longer and narrower, EL/EW more than 1.45; EL/PL more than 1.85; pronotal apex very shallowly emarginate; elytral punctation dual and confused; southern
Arizona and western Mexico
C. powelli ( p. 501 )
— Body shorter and broader, EL/EW less than 1.45; EL/PL less than 1.85; pronotal apex rounded; elytral punctation single and
uniform; Florida and tlie West Indies
C. rniiltipiinctattis (p. 499)
7. Elytral punctation single and uniform, very coarse and dense; EL/EW less than 1.35; surfaces of pronotum and elytra finely granulate or smooth and shiny 8
— Elytral punctation dual, the punctures fall- ing into 2 size classes; if obscurely du^l, then EL/EW more tlian 1.35 or pronotum coarsely granulate and dull in contrast to smooth and shiny elytra 9
8. Size larger, TL usually more than 1.55 mm; sides of elytra somewhat rounded; body somewhat shorter and broader; surfaces of pronotum and elytra smooth; pronotal apex
bearing 2 approximate tubercles
C. curtus (p. 497)
— Size smaller, TL usually less than 1.55 mm; sides of elytra subparallel; body somewhat longer and narrower; surfaces of pronotum and elytra finely granulate; pronotal apices
bearing 2 widely-spaced tubercles _
C. nigropunctatus (p. 499)
9. Elytral punctation distinctly seriate, the larger punctures forming relatively straight
rows; pronotal apex bearing 2 approximate,
flattened horns with rounded apices
C. pulhilus (p. 501)
— Elytral punctation not distinctly seriate; pronotal armature different 10
10. Body shorter and broader, EL/EW less than 1.37; pronotal apex with a short, broad, elevated, shallowly emarginate lamina C. pecki (p. 499)
— Body longer and narrower, EL/EW more than 1.37; pronotal apex with a deeply emarginate lamina or 2 horns 11
11. Pronotal apex with a deeply emarginate lamina or 2 slightly tumid, subtriangular horns; elytral punctation coarser and denser, the punctures usually separated by less than 0.75 diameter; size somewhat larger, TL usually more than 1.50 mm; elytra somewhat longer and narrower, EL/EW usually more than 1.44; western North America C californicus (p. 497)
— Pronotal apex with 2 distinctly trnnid, nar- row, diverging horns; elytral punctation finer and sparser, the punctures usually separated by more than 0.75 diameter; size somewhat smaller, TL usually less than 1.55 mm; elytra somewhat shorter and broader, EL/EW usually less than 1.48; eastern North America C thoracicornis (p. 502)
12. Pronotal apex rounded or shallowly emar- ginate, without distinct horns, tubercles, or lamina; alidominal fo\ea strongly trans- verse and at least 0.50 X as long as body of sternite III 13
— Pronotal apex produced, forming a lamina or 2 horns or tubercles; abdominal fovea not or slightly transverse and less tlian 0.50
X as long as body of sternite III 14
13. Size larger, TL more than 1.40 mm; fronto- clypeal ridge simple; pronotal disc strongly declined anteriorly, the apex rounded;
elytra expanded apically
C. obrieni (p. 499)
— Size smaller, TL less than 1.40 mm; fronto- clypeal ridge forming 2 tubercles; pronotal disc only weakly declined, the apex shal- lowly emarginate; elytra subparallel
C. dixiensis (p. 498)
14. Body longer and narrower, TL/EW more than 2.40; apex of pronotmn witli 2 nar- rowed, slightly diverging horns
C. qttadricornis (p. 501)
— Body shorter and broader, TL/EW less than 2.40; apex of pronotum with a short, broad lamina or 2 flattened, subtriangular horns — 15
15. Elytral punctation much coarser and denser tlian pronotal punctation 16
— Elytral punctation as fine and sparse as or
496 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
finer and sparser tlian pronotal puncta- tion 18
16. Size smaller, TL usually less than 1.30 mm; EL/PL usually more tlian 1.60; pro- notal punctation somewhat finer and sparser, tlie punctures usually less than 0.10 X as large as scutellar base and separated by more than 1.5 diameters; pronotal apex with 2 approximate, triangu- lar horns; color black or dark brown, with pronotal apex yellowish; eastern North America C. 7ninutissimus (p. 499)
— Size larger, TL usually more than 1.30 mm; EL/PL usually less than 1.60; pronotal punctation somewhat coarser and denser, the punctures usually more than 0.10 X as large as scutellar base and separated by less than 1.5 diameters; pronotal apex with a short, broad, ele\'ated lamina; color usually uniformly reddish; western Nortli America and Mexico 17
17. Elytra shorter and broader, EL/EW usually less than 1.32; pronotal lamina shorter, broader, less abrupt, and strongly elevated at apex (Fig. 36); southern Texas and eastern Mexico C. schaefferi (p. 502)
— Elytra longer and narrower, EL/EW usu- ally more than 1.32; pronotal lamina longer, narrower, more abrupt, and weakly ele- vated (Fig. 35); western Mexico and Baja Cahfornia C. similis (p. 502)
18. Size smaller, TL less tlian 1.5 mm; abdom- inal fovea less than 0.33 X as long as body of sternite III; pronotal apex with 2 sub- parallel or slightly diverging, narrow horns;
elytral punctation subseriate
C. minuttis (p. 499)
— Size larger, TL more than 1.5 mm; ab- dominal fovea more than 0.33 X as long as body of sternite III; pronotal apex witli a deeply emarginate lamina, giving the ap- pearance of 2 broad, subtriangular horns; elytral punctation confused 19
19. Pronotal punctation as fine and sparse as elytral punctation, the punctures usually separated by 1 diameter or more; abdom- inal fovea slightly transverse
'. C. sallei (p. 501)
— Pronotal punctation somewhat coarser and denser than elytral punctation, the punc- tures usually separated l^y less than 1 diam- eter; abdominal fovea circular „
C. punctulatus (p. 501)
Females (Abdominal sternite III simple)
1. Antennae 10-segmented 2
— Antennae with less than 10 segments 3
2. Elytral punctation distinctly seriate, the
larger punctures forming relati\ely straight rows; color of elytra usually reddish; east- ern North America, north of tlie Florida peninsula C. singularis (p. 502)
— Elytral punctation not distinctly seriate; color of elytra usually black; Florida Keys C. magister (p. 498)
3. Antennae 9-segmented 4
— Antennae 8-segmented 12
4. Elytral punctation single and confused, very coarse and dense, the punctures separated by 0.50 diameters or less; EL/EW less tlian 1.40 5
— Elytral punctation distinctly dual, the punc- tures falling into 2 size classes; if obscurely dual, EL/EW more than 1.40 7
5. Pronotal punctation finer and sparser, tlie punctures less than 0.10 X as large as scutellar base and separated by more than 0.75 diameter; pronotal disc strongly de- clined anteriorly, its surface distinctly granulate and dull, in contrast to the
smooth and shiny elytral surface
C. multipiinctatus (p. 499)
— Pronotal punctation coarser and denser, the punctures more than 0.10 X as large as scutellar base and separated by less tlian 0.75 diameter; pronotal disc not or weakly declined anteriorly, its surface similar in texture to that of elytra - 6
6. Size larger, TL usually more than 1.55 mm; sides of elytra somewhat rounded; pro- notum shorter and broader, PL/PW usually less than 0.88; surfaces of pronotum and elytra smooth C. cuiius (p. 497)
— Size smaller, TL usually less than 1.55 mm; sides of elytra subparallel; pronotum longer and narrower, PL/PW usually more tlian 0.88; surfaces of pronotum and elytra finely granulate C. nigropunctatus (p. 499)
7. Elytral punctation distinctly seriate, the larger punctiues forming relatixely straight rows; pronotal punctures separated by more than 1 diameter; southeastern U. S. and West Indies ..- C. pulltilus ( p. 501 )
— Elytral punctation not distinctly seriate; if subseriate, pronotal punctures separated by less than 1 diameter and distribution west- ern — 8
8. Elytral punctation finer and sparser, the punctures usually smaller tlian those on pronotum and separated by more than 1 diameter; EL/PL more than 1.50; eastern North America 9
— Elytral punctation coarser and denser, the megapunctures larger tlian pronotal punc- tures and separated by less than 1 diameter; EL/PL less than 1.50 or distribution western - 10
North American Ciidae • Lawrence 497
9. Pronotal punctation very fine and sparse, the punctures less than 0.10 X as large as scutellar base and separated by 1.5 diam- eters or more; pronotum somewhat shorter and broader, PL/PW usually less than 0.91; metasternal suture more tlian 0.20 X as long as median length of metastemum; pro- notum usually lighter in color than elytra
C. monocerus (p. 499)
— • Pronotal punctation coarser and denser, the punctures more tlian 0.10 X as large as scutellar base and separated by less than 1.5 diameters; pronotum somewhat longer and narrower, PL/PW usually more tlian 0.91; metasternal suture less than 0.20 X as long as median length of metasternum; pronotum and elytra similar in color or pronotum darker ..._ C thoracicornis (p. 502)
10. Pronotum strongly declined anteriorly and relatively long, EL/PL less than 1.50; elytra shorter and broader, EL/EW less than 1.37; color usually reddish; soutlieastern U. S C. pecki (p. 499)
— Pronotum not or weakly declined anteriorly and relatively shorter, EL/PL more than 1.50; elytra longer and narrower, EL/EW more than 1.37; color usually black or red and black; western North America 11
11. Pronotum distinctly narrowed anteriorly; EL/PL more than 1.85; antennal segment III 2.0 X as long as IV; size smaller, TL
usually less than 1.52 mm
C. powelli (p. 501)
— Pronotum not narrowed anteriorly; EL/PL less than 1.85; antennal segment III 3.0 X as long as IV; size larger, TL usually more tiian 1.45 mm C. californicus (p. 497)
12. Pronotal disc strongly declined anteriorly; elytra expanded apically; southern Arizona and northern Mexico C. ohiieni (p. 499)
— Pronotal disc not or only weakly declined; elytra not expanded apically .,-13
13. TL/EW more than 2.35; elytral punctures separated by 0.75 diameter or more, not subconfluent anteriorly; soutliern Texas,
eastern Mexico, and Central America
C quadricornis (p. 501)
— TL/EW less than 2.35 or elytral punctures separated by less than 0.75 diameter and becoming subconfluent anteriorly 14
14. Elytral punctation coarser and denser than that of pronotum, tlie elytral punctiues be- coming subconfluent anteriorly 15
— Elytral punctation finer and sparser than or as fine and sparse as that of pronotum, the elyti-al punctures not subconfluent an- teriorly 8
15. Size smaller, TL usually less than 1.30 mm;
EL/PL more than 1.60; color of elytra uni- formly black A6
— Size larger, TL usually more than 1.30 mm; EL/PL less than 1.60; color of elytra red- dish or black and red, rarely black __ — 17
16. Pronotum somewhat shorter and broader, PL/PW usually less than 0.90; antennal segment III 1.50 X as long as IV; eastern North Anerica C. 7iunutissimus (p. 499)
— Pronotimi somewhat longer and narrower, PL/PW usually more than 0.90; antennal segment III 2.0 X as long as IV; western Nortli America C. dixiensis (p. 498)
17. Elytra shorter and broader, EL/EW usually less than 1.32; southern Texas and eastern Mexico C. schaefferi (p. 502)
— Elytra longer and narrower, EL EW usu- ally more than 1.32; western Mexico and Baja California C. similis (p. 502)
18. Size smaller, TL less than 1.10 mm; elytral punctation subseriate; pronotal surface
lightly granulate and shiny
C. minutiis (p. 499)
— Size larger, TL usually more than 1.10 mm; elytral punctation confused; pronotal sur- face distinctly granulate and dull 19
19. Pronotal punctation about as fine and sparse as elytral punctation, the punctures usually separated by more than 1 diameter :.... C. sallei (p. 501)
— Pronotal punctation coarser and denser than elytral punctation, the punctures usually
separated by less than 1 diameter
C. punctulatus (p. 501)
Ceracis californicus (Casey)
Ennearthron californicitm Casey, 1884: 36. Ceracis caJiforiiicus (Casey), — Lawrence, 1967b:
107-110, figs. 15, 29. See Lawrence (1967b)
for complete synonymy.
Distribution. See Lawrence (1967b). Also known from Rio Florido, 50 mi. from Parral, CHIHUAHUA.
Host fungi. See Lawrence (1967b).
Ceracis curtus (Mellie)
Ennearthron curtum Melhe, 1848: 367, pi. 12, fig. 15.
Ceracis curiis (Mellie), — Lawrence, 1967b: 110- 112, fig. 30. See Lawrence (1967b) for com- plete synonymy.
Distribution. See Lawrence (1967b). Also known from Great Inagua, BA- HAMAS; Isla de Pinos, CUBA; Dunedin,
I
498 Bulletin Museum of Comparative Zoology, Vol 142, No. 5
i
Pinellas Co., FLORIDA; and Windsor, 10 mi. S Falmouth, Trelawny Par., JAMAICA. Host fungi. Pohjporus hijilnoides [2(1)]; Fames sclerodermeiis [1].
Ceracis dixiensis (Tanner)
Octotemmis dixiensis Tanner, 1934: 47. Ceracis dixiensis (Tanner), — Lawrence, 1967b: 112-113, fig. 33.
Distribution. See Lawrence (1967b). Also knowai from Rio Florido, 50 mi. from Parral, CHIHUAHUA.
Host fungi. Trametes hispida [10(8)]; Ganoderma sp. [1].
Ceroc/s magister NEW SPECIES
Holotifpe. i, FLORIDA: Lignum Vitae Key, Monroe Co., June 1, 1968, Lot 2600 J. F. Lawrence, ex Fomes rohiniae [MCZ No. 31694]. Allotype, 5, same data [MCZ].
Male. Length 1.80 mm. Body 2.18 X as long as broad. Head, pronotimi, and elytra black, elytral suture somewhat reddish. Vertex with a deep, semicircular, transverse impression, preceded by a median ele- vation; frontoclypeal ridge produced and deeply emarginate, forming 2 triangular plates. Antennae 10-segmented; segment III 1.12 X as long as IV. Pronotum 0.93 X as long as broad, widest at posterior third, strongly declined anteriorly; sides weakly rounded; anterior edge produced forming a flat, elevated lamina which is deeply emarginate, giving an appearance of 2 somewhat rounded plates; disc im- pressed anteriorly just behind lamina and bearing a short, oblique carina on each side of it; surface distinctly granulate; punctures 0.17 X as large as scutellar base and separated by 1.50 to 2.0 diameters. Elytra 1.30 X as long as broad and 1.48 X as long as pronotum; sides subparallel, apices blunt; punctation dual and con- fused, much coarser and denser than pro- notal punctation. Metasternum 0.48 X as long as wide; suture 0.15 X as long as median length of sternum. Abdominal
sternite III bearing a median, circular, pubescent fovea, which is 0.25 X as long as body of sternite, distinctly margined, and located posterad of center.
Female. Length 1.77 mm. Body 2.15 X as long as broad. Vertex slightly convex; frontoclypeal ridge simple. Pronotum 0.96 X as long as broad; anterior edge strongly rounded, simple. Elytra 1.30 X as long as broad and 1.54 X as long as pronotum. Sternite III without a pubescent fovea.
Variation. Pronotum yellowish orange to black, usually black, often somewhat red- dish anteriorly. Elytra yellowish to black, usually black and somewhat reddish along suture. In smaller males the frontoclypeal tubercles are rounded and the pronotal lamina is shallowly emarginate, while in some larger specimens the pronotum bears two distinct, rounded plates which are slightly divergent. Size and dimensions vary as follows in a series of 21 males and 20 females from Lignum Vitae Key, Mon- roe Co., Florida (Lot 2600):
TL mm: S 1.45-2.07 (1.74 ± 0.033),
9 1.55-1.92 (1.68 ±0.023); TL/EW: $ 2.14-2.32 (2.21 ± 0.010),
9 2.06-2.33 (2.17 ±0.010); PL/PW: S 0.88-0.97 (0.94 ± 0.005),
5 0.86-0.97 (0.92 ±0.007); EL/EW: S 1.29-1.38 (1.33 ± 0.005),
9 1.30-1.40 (1.34 ±0.005); EL/PL: 6 1.39-1.64 ( 1.50 ± 0.013),
9 1.50-1.75 (1.61 ±0.015); GD/EW: S 0.87-0.97 (0.92 ± 0.005),
9 0.90-0.97 (0.93 ±0.005). Total size range in material examined: 1.45-2.25 mm.
Paratypes. FLORIDA: 2, Lignum Vitae Key, Monroe Co., May 27, 1968, Lot 2577 JFL, ex Fomes rohiniae [JFL]; 48, same locahty, June 1, 1968, Lot 2600 JFL, ex Fomes rohiniae [FMNH, JFL, MCZ, USNM]; 4, same locality and date. Lot 2601 JFL, ex Fomes rohiniae [JFL]; 11, same locality, June 5, 1968, Lot 2623 JFL, ex Fomes rohiniae [JFL].
North American Ciidae • Lawrence 499
Distribution. Known only from Lignum Vitae Key, Florida.
Host fungi. Fomes rohiniae [4(4)].
Discussion. This is one of the two North American Ceracis with 10-segmented an- tennae, and it is easily distinguished from the otlier, C. si7iguhris, by the confused elytral punctation, darker color, and dif- ferent male armature. It appears to be most closely related to C. pecki, from which it differs in antennal segmentation, color, and relative length of pronotum. C. magister and C. pecki belong to a group of Central American and West Indian spe- cies (all undescribed) that breed in melanic conks and are usually fairly large, with 9- or 10-segmented antennae and a short pronotal lamina in the male.
Ceracis magister is knowm only from the Florida Keys and may occur in the Greater Antilles. On Lignum Vitae Key, it breeds in the conks of Fomes rohiniae, along with Cis niedhauki, Malacocis brevicollis, and Ceracis punctulatus.
Ceracis minufissinnus (Mellie)
Cis minutissimus Mellie, 1848: 334, pi. 11, fig. 12. Ceracis minutissimus (Mellie), — Lawrence, 1967b: 113-114, figs. 18, 32.
Distribution and host /wngi. See Law- rence (1967b).
Ceracis minutus Dury
Ceracis minuta Dury, 1917: 25; Lawrence, 1967b: 114-115, figs. 19, 33.
Distribution. See Lawrence ( 1967b ) . Also known from the BAHAMAS; Lignum Vitae Key, Big Pine Key, Plantation Key, and Key Largo, Monroe Co., FLORIDA; Kingston, Windsor, Mt. Diablo, and Ewar- ton, JAMAICA.
Host fungi. Pohjporus pinisitus [10(6)]; Polyporus hijdnoides [7(4)]; Stereum papyrinum [5(3)]; Polyporus maximus [3(3)]; Pohjporus occidentaUs [3(3)]; Pohjporus sanguineus [3(3)]; Daedalea ele- gans [1(1)]; Trametes corrugata [1(1)]; Fomes sclerodermeus [1].
Ceracis monocerus Lawrence
Ceracis monocerus Lawrence, 1967b: 115-116, figs. 20, 30. Replacement name for Ennearth- ron tinicorne Casey, 1898: 90 (not Ceracis uni- cornis Gorham, 1898: 332).
Distribution and liost fungi. See Law- rence ( 1967b ) .
Ceracis mulfipunctatus (Mellie)
Ennearthron multipunctatum Mellie, 1848: 368,
pi. 12, fig. 16. Ceracis mulfipunctatus (Mellie), — Lawrence
1967b: 116-118, figs. 24, 32.
Distribution. See Lawrence (1967b). Also known from Wakulla Springs State Park, Wakulla Co., FLORIDA; Adelphi, 3 mi. NW Ulster Spr., 10 mi. S Falmouth, and Ewarton, JAMAICA.
Host fungi. Ganoderma zonatum [6(5)]; Fomes sclerodermeus [4(4)]; Ganoderma applanatum [4(4)]; Ganoderma lucidum [1]; Pohjporus hydnoides [1]; Pohjporus supinus [1].
Ceracis nigropunctatus Lawrence
Ceracis nigropunctatus Lawrence, 1967b: 118- 119, figs. 16, 27.
Distribution. See Lawrence ( 1967b ) . Also known from Antigua Morelos, TAMAULIPAS, and Madden Dam, CANAL ZONE.
Ho.st fungi. Pohjporus hydnoides [9(4)]; Ganoderma sp. [2]; Pohjporus hirsutus [2]; Fomes sclerodermeus [1(1)].
Ceracis obr/en/ Lawrence
Ceracis obrieni Lawrence, 1967b: 119-120, figs. 25, 29.
Distribution. See Lawrence ( 1967b ) . Known also from Peria Blanca, Santa Cruz Co., ARIZONA.
Host fungi. See Lawrence ( 1967b ) .
Ceracis pecki NEW SPECIES
Holotype. S, FLORIDA: Florida Cav- erns State Park, Jackson Co., Sept. 6, 1968, Lot 2670 J. F. Lawrence (S. B. Peck, coll.)
500
Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
ex Poria nigra [MCZ No. 31695]. Allotype, 9, same data. [MCZ].
Male. Length 1.77 mm. Body 2.22 X as long as broad. Head and pronotum dark, reddish brown, elytra reddish. Vertex with a deep transverse impression, preceded by a median elevation; frontoclypeal ridge produced, forming a short, broad, slightly concave, elevated lamina, which is shal- lowly emarginate at apex. Antennae 9- segmented; segment III 2.0 X as long as IV. Pronotum 0.97 X as long as broad, widest at middle; sides weakly rounded; anterior edge produced, forming a flat, slightly elevated lamina which is shallowly emarginate; disc impressed anteriorly just behind lamina and bearing a short oblique carina on each side of it; surface distinctly granulate; punctures 0.20 X as large as scu- tellar base and separated by 0.50-1.0 diam- eter. Elytra 1.28 X as long as broad and 1.37 X as long as pronotum; sides sub- parallel, apices blunt; punctation indis- tinctly dual and confused; punctures coarser and denser than those on pro- notum. Metasternum 0.44 X as long as wide; suture 0.14 X as long as median length of sternum. Abdominal sternite III bearing a median, circular, pubescent fovea, which is 0.50 X as long as body of sternite, indistinctly margined and located about in center.
Female. Length 1.70 mm. Body 2.19 X as long as broad. Vertex slightly convex; frontoclypeal ridge simple. Pronotum 1.00 X as long as broad; anterior edge strongly rounded, simple. Elytra 1.29 X as long as broad and 1.43 X as long as pronotmn. Sternite III without a pubescent fovea.
Variation. Pronotum yellowish orange to bro\Mi, usually reddish bromi. Elytra yel- lowish to brown, usuallv reddish. Fronto- clypeal and pronotal laminae in smaller males barely developed. Size and dimen- sions vary as follows in a series of 25 males and 25 females from Florida Caverns State Park, Jackson Co., Florida (Lot 2760):
TL mm: 6 1.47-2.02 (1.68 ± 0.028), $ 1.32-1.95 (1.66 ±0.031);
TL/EW: c^ 2.11-2.33 (2.20 + 0.011),
9 2.12-2.25 (2.19 + 0.007); PL PW: S 0.93-1.03 (0.97 ± 0.005),
5 0.93-1.00 (0.98 ±0.004); EL EW: i 1.21-1.37 (1.28 ±0.007),
9 1.23-1.32 (1.29 ±0.005); EL PL: i 1.25-1.50 ( 1.38 ± 0.021),
2 1.35-1.48 (1.42 ±0.008); CD EW: $ 0.86-0.93 (0.90 ± 0.004),
9 0.86-0.94 (0.90 ±0.004). Total size range in material examined: 1.32-2.13 mm.
Paratypes. FLORIDA: 52, same data as holotype [FMNH, JFL, MCZ, USNM]; ILLINOIS: 1, Starved Rock State Park, LaSalle Co., X-24-1953, leaf mold, C. Ziol- kowski leg. [FMNH]; MARYLAND: 1, Oakland, 10.7 [USNM]; VIRGINIA: 1, (no specific locality) [CIN].
Distribution. Eastern United States, from Illinois and Maryland south to Florida.
Host fungi. Poria nigra [1(1)].
Discussion. This species is characterized by the relatively large size, 9-segmcnted antennae, short and broad elytra with coarse and dense, obscurely dual, confused punctation, relatively elongate pronotum with fine, sparse punctation and coarsely granulate surface, reddish color, and short, broad pronotal lamina in the male. C. nmgister and C. singiilaris are similar, but the former is darker in color with a rel- atively shorter pronotiun, the latter has distinctly seriate elytral punctation, and both have 10-segmented antennae. Other species with 9-segmented antennae have single elytral punctation (C. curtus and C. nigropunctatus), seriate elytral punctation (C. puUulus), finer and sparser elytral punctation (C. monocerus and C. thoraci- cornis) or longer and narrower elytra (C. poicelli and C. calif ornicus) . Larger speci- mens of C. saJlei resemble this species, but the antennae are 8-segmented, the pro- notum is relatively shorter, and the male armature is different.
Ceracis pecki is known from a few scat- tered localities but may be distributed
North American Ciidae • Lawrence 501
throughout the Southeast. It appears to be most closely related to C. ma<i,ister and to several undescribed forms from Central America. It has been taken only in Foria nigra, which is unusual in having a dark chocolate or violet-brown fruiting body.
Ceracis powelli Lawrence
Ceiacis powelli Lawrence, 1967b: 120-121, figs. 23, 27.
Distribution and host fungi. See Law- rence (1967b).
Ceracis pullulus (Casey)
Enncarthron puUulum Casey, 1898: 90. Ceracis pullulus (Casey), — Lawrence, 1967b: 121-123, figs. 22, 28.
Distribution. See Lawrence (1967b). Also known from Pine Mountain State Park, Harris Co., GEORGIA; Atlantic Beach, Carteret Co., NORTH CAROLINA; Cerro Dona Juana, Ponce, PUERTO RICO; West Union, Oconee Co., SOUTH CARO- LINA; 1 mi. NW Adams, Robertson Co., TENNESSEE; and Cinnamon Bay, St. John, VIRGIN ISLANDS.
Host fungi. Poh/porus gihus [20(8)]; Polyporus licnoides [7(6)]; Polyporus por- recttis ? [4(4)]; Ganoderma zonatum [4 (2)]; Polyporus iodinus [2(1)]; Polyporus hydnoides [2]; Polyporus corrosus [1(1)]; Poria nigra [1(1)]; Polyporus vinosus [1].
Ceracis punctulafus Casey
Ceracis punctulata Casey, 1898: 90; Lawrence,
1967b: 123-124, fig. 34. Ceracis pur\ctulatus ruhiculus Lawrence, 1967b:
124-127, figs. 13, 34.
Distribution. See Lawrence (1967b). Also known from Big Pine Key, Bill Find's Key, Crawl Key, Grassy Key, Key Largo, Lignum Vitae Key, Plantation Key, and Upper Matecumbe Key, Monroe Co., FLORIDA; and Windsor, 10 mi. S Fal- mouth, Trelawny Par., JAMAICA.
Host fungi. Polyporus gilvus [46(25)]; Polyporus hydnoides [18(9)]; Fomes robiniae [5(2)]; Polyporus fulvocinereus
[3(2)]; Ganoderma zonatum [3(1)]; Poly- porus licnoides [2(2)]; Stereum papyrinum [2(2)]; Polyporus sanguineus [2]; Poly- porus porrectus ? [1(1)]; Fomes fomen- tarius [1]; Ganoderma curtisii [1]; Gano- derma lucidum [1]; Ganoderma sp. [1]; Pohjporus adustus [1]; Polyporus par- gamenus [1]; Polyporus pinisitus [1]; Stereum ostrea [1]; Trametes corrugata
[1]. Discussion. A large series of specimens
collected in the Florida Keys shows a great deal of variation in size and color, and in- dividuals are, on the average, smaller and more reddish than those from central and southern Florida. These island populations seem to prefer a wider variety of fungi and were found breeding on Polyporus fulvo- cinereus and Stereum papyrinum, as well as Polyporus hydnoides and Fomes robiniae. The species appears to be common in the area and is the only ciid collected on Bill Find's Key, a small red mangrove island near Sugarloaf Key.
Ceracis quadricornis Gorham
Ceracis quadricornis Gorham, 1886: 359; Law- rence, 1967b: 127-128, figs. 17, 30.
Distribution. See Lawrence (1967b). Also known from Chorros de Agua, 13 mi. W Montemorelos, NUEVO LEON; and Antiguo Morelos, TAMAULIPAS.
Host fungi. Polyporus occidentalus [7 (4)]; Polyporus hydnoides [5(4)]; Pohj- porus maximus [4(3)]; Polyporus hirsutus [2(1)]; Trametes corrugata [1(1)]; Poly- porus crocatus [1]; Trametes cirrifer [1].
Ceracis sallei Mellie
Ennearthron (Ceracis) sallei Mellie, 1848: 377, pi.
12, fig. 22. Ceracis sallei Mellie, — Lawrence, 1967b: 128-
130, fig. 11, 26.
Distribution. See Lawrence (1967b).
Host fungi. Ganoderma applanatum [31(16)]; Ganoderma curtisii [3(1)]; Ganoderma zonatum [2(1)]; Polyporus hydnoides [2(1)]; Fomes connatus [1(1)]; Ganoderma lobatum [1(1)]; Fomes pini-
502 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
cola [1]; Fomes sclerodermeus [1]; Gano- derma sp. [1]; Pohjporus cinnabariniis [1].
Ceracis schaefferi Dury
Ceracis schaefferi Durv, 1917: 25; Lawrence, 1967b: 130-131, figs.' 14, 28.
Distribution and host fungi. See Law- rence (1967b).
Ceracis similis Horn
Ceracis similis Horn, 1894: 391; Casey, 1898: 90; Dmy, 1917: 27. Type locality: "Coral de Piedra, Sierra el Taste" [Baja California]. Types, CAS.
Distribution. Baja California Siir and Nayarit south to El Salvador. Marginal records: Porto Balandra, Carmen Is.; San Jose del Cabo, BAJA CALIFORNIA SUR; San Bias, NAYARIT; 45 mi. N Acadan, PUEBLA; 2 mi. S Quezaltepeque, EL SALVADOR.
Host fungi. Ganoderma zonatum [3(3)]; Ganoderma sp. [2(2)]; Ganoderma ap- planatum [1(1)].
Discussion. Although this species has not been recorded from the United States, it is included here because of its occurrence on several of the islands in the Gulf of Cali- fornia. It is possible diat further collecting in extreme southern California may reveal the presence of C. similis there. The spe- cies most closely resembles C. schaefferi from eastern Mexico and southern Texas.
Ceracis singularis (Dury)
Xesiocis singuknis Dury, 1917: 14. Ceracis singularis (Dury), — Lawrence, 1967b: 131-132, figs. 12, 33.
Distribution. See Lawrence (1967b). Also known from Chattahoochie State Park, Houston Co., ALABAMA; Florida Caverns State Park, Jackson Co., FLORIDA; Spring Mill State Park, Lawrence Co., INDIANA; Wildcat Den State Park, Muscatine Co., IOWA; and Madison, Dane Co., WISCON- SIN.
Host fungi. Pohjporus gilvus [12(7)]; Ganoderma applanatuni [8(2)]; Fomes
robiniae [4(1)]; Fomes igniarius [2(1)]; Ganoderma curtisii [2(1)]; Poria nigra [2 (1)]; Fomes sclerodermeus [2]; Pohjporus pargamenus [2]; Fomes rihis [1(1)]; Len- zites saepiaria [1(1)]; Daedalea confragosa [1]; Fomes annosus [1]; Ganoderma sp. [1]; Ganoderma tsugae [1]; Pohjporus licnoides [1]; Pohjporus supinus [1]; Pohj- porus versicolor [1]; Trametes hispida [1]; Trametes plebeja [1].
Ceracis thoracicornis (Ziegler)
Cis thoracicornis Ziegler, 1845: 270.
Ceracis thoracicornis (Ziegler), — Lawrence,
1967b: 132-136, figs. 21, 31. See Lawrence
( 1967b ) for complete synonymy.
Distribution. See Lawrence (1967b). Numerous additional records have made no significant changes in the range.
Host fungi. Pohjporus pargamenus [41 (15)]; Pohjporus supinus [25(15)]; Pohj- porus adust us [15(6)]; Pohjporus versicolor [11(3)]; Lenzites betulina [10(6)]; Daedalea ambigua [9(7)]; Ganoderma applanatum [7]; Pohjporus sector [6(2)]; Daedalea uni- color [6(2)]; Fomes ulmarius [6(1)]; Ganoderma lucidum [5(2)]; Ganoderma tsugae [5(1)]; Pohjporus hirsutus [4(1)]; Pohjporus squamosus [3(2)]; Pohjporus abietinus [3(1)]; Pohjporus gilvus [3]; Fomes sclerodermeus [2(1)]; Trametes corrugata [2(1)]; Stereum ostrea [2]; Trametes plebeja [2]; Ganoderma sp. [1 ( 1 ) ] ; Pohjporus spraguei [1(1)]; Pohjporus tulipi ferae [1(1)]; Poria vitrea [1(1)]; Trametes hispida [1(1)]; Trametes trogii [1(1)]; Boletus sp. [1]; Daedalea elegans [1]; Fomes fomentarius [1]; Fomes fraxi- neus [1]; Fomes pinicola [1]; Ganoderma curtisii [1]; Pohjporus cinnaharinus [1]; Pohjporus dichrous [1]; Pohjporus hijd- noides [1]; Pohjporus sulphureus [1]; Pohj- porus vinosus [1].
Genus Sulcacis Dury
Sulcacis Dury, 1917: 20; Lawrence, 1965: 278 (complete synonymy); Lohse, 1967: 284. Type species, by subsequent designation, Sulcacis lengi Dury, 1917: 21 (Lawrence, 1965: 278).
North American Ciidae • Lawrence 503
Entypus Redtenbacher, 1847: 350 (not Dahlbom, 1843: 35). Type species, by monotypy, Cis af finis Gyllenhal, 1827: 628 ( misidentif ied as Apate fronticornis Panzer).
Entijpus (Entypocis) Lohse, 1964: 121. Type spe- cies, by original designation, Cis bidentulus Rosenhauer, 1847: 58.
Sulcacis (Entypocis) Lohse, — Lohse, 1967: 284.
Include d species. Cis af finis Gvllenhal, 1827: 628 [Eurasia]; Cis hidentidm Rosen- hauer, 1847: 58 [southern Europe, northern Africa]; Cis curtuhis Casey [northern North America, see p. 503]; Apate fronticornis Panzer, 1809: 98: 7 [Eurasia]; Rhopalodon- tus japonicus Nobuchi, 1960: 39 [Japan]; Sulcacis lengi Dury [eastern North Amer- ica, see p. 504]. Total: 6 species.
Excluded species. Cis bicornis Rosen- hauer (see p. 488); Rhopalodontus tokuna- gai Nobuchi (see p. 488).
This genus was discussed in detail by Lohse ( 1964, under Entypus and Entypo- cis) and Lawrence (1965), but the concept is somewhat modified in the present treat- ment by the transfer of C. bicornis and R. tokunagai to Strigocis {see p. 488). Species of Sulcacis may be distinguished from most other North American ciids by the spinose protibial apices, biconcave prostemum with a relatively broad, tapering intercoxal process, rounded anterior pronotal angles, and indistinctly dual vestiture, consisting of longer and shorter bristles. The most closely related group is Malacocis, in which the body form is shorter and broader and the prostemum veiy short and straplike. .
Key to the North American Species of Sulcacis
1. Antennae 9-segmented; elytra! punctation finer, punctures usually less than 0.20 X as large as scutellar base; pronotal surface dull; male with frontoclypeal ridge simple and
abdominal fovea transversely oval
S. lengi (p. 504)
— Antennae 10-segmented; elytral punctation coarser, punctures usually more than 0.20 X as large as scutellar base; pronotal surface somewhat shiny; male with frontoclypeal ridge bidentate and abdominal fo\ ea circular S. curtuhis (p. 503)
Sulcacis curtulus (Casey), NEW COMBINATION
Cis curtula Casey, 1898: 83; Weiss and West, 1920: 8 (dist, biol.); Lawrence, 1965: 277. Type locality: "New York." Holotype, $ , Casey Coll., USNM.
Cis montana Casey, 1898: 82. Type locality: "Montana ( Missoula ) ." Holotvpe, $ , Casey Coll., USNM. NEW SYNONYMY.
Cis sorror Casey, 1898: 83; Hatch, 1962: 230, pi. 48, fig. 3, 3a (dist, biol.). Type locality: [Victoria] "Vancouver Island ..." [British Columbia]. Holotype, $, Casey Coll., USNM. NEW SYNONYMY.
Cis cylindricus Dury, 1917: 8; Weiss, 1920a: 110- 111 (biol); Weiss and West, 1920: 8 (biok); Weiss and West, 1921a: 61 (dist., biol.); Weiss and West, 1921b: 169 (dist., biol); Hatch, 1962: 230 (syn. with sorror). Type locahty: "Umatilla Co., Oregon." Syntypes, Dury Coll., CIN. NEW SYNONYMY.
Sulcacis niger Dury, 1917: 21. Type locality: "Southern Illinois." Holotype, 9 , Dury Coll., CIN. NEW SYNONYMY.
Cis criddlei Dury, 1919: 158. Type locality: "Aweme, Manitoba." Holotype, $ , Dury Coll., CIN. NEW SYNONYMY.
Cis hystriculus Casey, — Weiss and West, 1921a: 61 (biol.). Misidentification.
Distribution. Widespread throughout the northern and montane regions of North America, from the northern coast of British Columbia to southern Quebec and New England, south to San Diego County, Cali- fornia, the mountains of southern Arizona and northern Mexico, and the states of Nebraska, Illinois, and North Carolina (Fig. 110). Marginal records: BRITISH COLUMBIA: Terrace. ALBERTA: Mc- Murray. MANITOBA: Aweme. QUEBEC: Chelsea. NEW HAMPSHIRE: Intervale, Mt. Surprise, Carroll Co. NORTH CARO- LINA: Southern Pines, Moore Co. ILLI- NOIS: Cahokia, St. Clair Co. NEBRASKA: Central City, Merrick Co.; 15 mi. W Sydney, Cheyenne Co. COLORADO: Pueblo, Pueblo Co. NEW MEXICO: (no specific localit>). CHIHUAHUA: Rio Florido, 50 mi. from Parral. ARIZONA: Miller Canyon, 10 mi. W Hereford, Hua- chuca Mts., Cochise Co. CALIFORNIA: 10 mi. N Descanso, San Diego Co.
Host fungi. Polyporus versicolor [69
504
Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
(35)]; Polyporus hirsutiis [12(6)]; Poly- poriis adiistus [11(3)]; Lenzites bettilina [8(5)]; Tra metes h ispida [6(4)]; Pleurotns ostreatu.s [3]; Polyporus cinnabarimis [2 (1)]; Fomes i^,niarius [2]; Schizophylhnn commune [1(1)]; Ganoderma opplanatiim [1]; Polyporus gilvus [1]; Polyporus vul- pinus [1]; Steccherinum ochroceum [1]; Stereum hirsutum [1].
Discussion. This species is similar in general form and punctation to Dolichocis manitoha, which is more elongate (EL/EW more than 1.50), with 9-segmcnted anten- nae, narrowly rounded protibial apex, and a fovea on the vertex in the male. Sulcacis lengi is somewhat shorter and broader and differs in having 9-segmented antennae and finer punctation. Cis hystricidus resembles S. curtulus in several respects, but differs in having the protibial apex dentate, the body more elongate, the pronotal puncta- tion denser, the elytral bristles colorless (rather than yellow), and the abdominal fovea absent in the male. It appears to be most closely related to the Palaearctic Sulcacis bidentulus (Rosenhauer).
Sulcacis cuHulus breeds in a number of fungi but prefers Polyporus versicolor and its relatives, as do other Sulcacis whose feeding habits have been recorded. Al- though the species is absent from the southwestern deserts, it is fairly widespread throughout the western mountains and along the Pacific Coast, frequenting both dry and humid habitats. Along the coast of California, it breeds in the conks of Poly- porus versicolor and related fungi, in as- sociation with Cis fuscipes, Cis vitulus, Cis versicolor, and Octotemnus laevis. In the drier parts of the western mountains and at higher elevations, S. curtulus is often fomid alone in the same habitats. The species is much less common in eastern North America, although it is known from widely scattered localities.
Sulcacis lengi Dury
Sulcacis lengi Dury, 1917: 21; Weiss 1919b: 203- 204 (biol.); Weiss and West, 1920: 8 (dist..
biol. ); Lawrence, 1965: 277. Type locality: "Vermont" [Bennington Co.]. Syntypes, Dury Coll., CIN.
Distribution. Eastern North America, from Maine to the Carolinas and west to eastern Kansas and Texas (Fig. 99). Marginal records: MAINE: Weld, Frank- lin Co.; NORTH CAROLINA: Calypso, Duplin Co.; SOUTH CAROLINA: Moun- tain Rest, Oconee Co.; MISSISSIPPI: 4 mi. W Starkville, Oktibbeha Co.; TEXAS: Huntsville, Walker Co.; KANSAS: Salina, Saline Co.; IOWA: Estherville, Emmet Co.; MICHIGAN: (no specific locality).
Host fungi. Polyporus versicolor [11(3)]; Lenzites betulina [5(2)]; Polyporus pubes- cens [1(1)]; Polyporus hirsutus [1]; Bo- letus sp. [1].
Discussion. This species is similar to Malacocis brevicollis in punctation, vesti- ture, protibial apices, and secondary sexual characters, and it may represent a link between Sulcacis and Malacocis. Members of the genus Malacocis, ho\\'ever, are char- acterized by the shorter and broader body fonn, and shortened pronotum, prosternum, and metasternum; in addition, M. brevi- collis has 10-segmented antennae. S. curtu- lus differs in being more elongate and coarsely punctate, with 10-segmented an- tennae and different secondary sexual characters.
Sulcacis lengi occurs throughout the eastern part of the continent and breeds in the fruiting bodies of Polyporus versicolor and its relatives, in association with Cis fuscipes, Strigocis opacicollis, Octotemnus laevis, and several other ciid species.
Genus Malacocis Gorham
Malacocis Corhain, 1886: 356. Type species, by monotypv, Malacocis championi Gorham, 1883: .356.
Brachtjcis Casey, 1898: 86; Dmy, 1917: 21; Leng, 1920: 247; Aniett, 1962: 829. Type species, by monot>'py, Brachycis brevicollis Casey, 1898: 87. NEW SYNONYMY.
Included species. Brachycis brevicollis Casey [eastern North America, see p. 505];
North American Ciidae • Lawrence 505
Malacocis championi Gorham, 1883: .'356 [Guatemala]. Total: 2 species.
Excluded species. Malacocis hahiensis Pic, 1916: 6 [Brazil]. See discussion below.
Tlie genus Malacocis includes two de- scribed and several undescribed species from North and Central America and the West Indies. They are all characterized by the very short and broad body form, vesti- ture of short to moderately long bristles, straplike presternum with tapering inter- coxal process (Fig. 27), and spinose pro- tibial apex (Fig. 57). The antennae are 10-segmented in the North American M. hrevicoUis, but in the Central American M. championi they are 9-segmented, and in an undescribed Mexican form there are only 8 segments. In the male, the pronotal apex is always simple and the frontoclypeal region usually bears t\vo small teeth or tubercles. Some species of Ceracis are similar, but the prostenial intercoxal proc- ess is laminate and the vestiture consists of very short, fine hairs. The most closely related genus appears to be Sidcacis, in which the form is more elongate and the prosternum longer in front of the coxae.
Since Casey's species Biachycis hrevi- coUis differs from M. championi in little more than antennal segmentation, \\'hich has undergone reduction in several groups of Ciidae, I have synonymized the junior name Brachijcis above. Pic's species Mala- cocis hahiensis does not belong in this group at all, but rather is a member of the Cis tauriis group ( = Macrocis Reitter, see p. 439).
Malacocis brevicollis (Casey), NEW COMBINATION
Biacht/cis brevicollis Casey, 1898: 87; Diiry, 1917: 21 (dist); Gibson, 1918: 113 (dist.); Weiss, 1919a: 145-147 (biol.); Weiss and West, 1920: 8 (dist., biol.). Type locality: "New York (Ithaca)." Holotype, $, Casey Coll., USNM.
Distrihittion. Eastern North America, from northern Maine to the Florida Keys and west to southern Manitoba, eastern Kansas, and Texas (Fig. 108). Marginal
records: MAINE: Baxter State Park, Piscataquis Co. QUEBEC: Montreal; Knowlton. FLORIDA: Lignum Vitae Key, Monroe Co. TEXAS: Austin, Travis Co. OKLAHOMA: South of Broken Bow, Mc- Curtain Co. KANSAS: 5 mi. S Lawrence, Douglas Co. MANITOBA: Aweme. ON- TARIO: 10 mi. SE Upsala.
Host fungi. Pohjporus gilvtis [36(16)]; Fomes igniarius [13(8)]; Fomes rohiniae [6(3)]; Fames rohustus [4(2)]; Fomes pomaceus [3(2)]; Polyporus licnoides [3]; Fomes everharfii [2(1)]; Fomes pini [1(1)]; Polyporus corrosus [1(1)]; Fomes fotnentarius [1]; Fomes jolmsonianus [1]; Pohjporus hydnoides [1]; Pohjporus vinosus [1]; Porta ferruginosa [1]; Porta laevigata
[1].
Discussion. Malacocis brevicollis is easily distinguished by the very short and broad body form, the short and straplike pro- sternum, shortened metastemum, spinose protibial apices, vestiture of short, stout, bristles, and lack of sexual ornaments on the clypeus or pronotiun of the male. Sulcacis lengi is somewhat more elongate, with the prosternum and metastemum less reduced. In addition, M. hrevicoUis is usually larger in size and the elytra are usually brownish with reddish yellow brist- les, while the smaller S. lengi has blackish elytra with yellow bristles. x\nother unique character in M. hrevicoUis is the tendency for the elytral punctures to become trans- versely confluent on the anteromesal part of the disc, so that a series of indistinct transverse ridges is formed.
There is considerable variation in color and vestiture within this species. Most non- teneral specimens are browaiish and many have reddish browii elytra \\'ith black or dark brown prothorax. Tlie bristles are shorter in specimens from northern popu- lations, while those from the Southwest have longer bristles that fall into two size classes. One population from Key Largo is comprised entirely of small individuals that are darker in color and have shorter and more yellowish bristles than those of
506
Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
other southern populations. This may well be a distinct species, but it is not treated as such here. The genus Malacocis con- tains a number of forms (mostly unde- scribed) from the West Indies, Mexico, and Central America, which should be studied as a unit before any more new species are described.
Malacocis hrevicoUis is probably the most common and characteristic North American inhabitant of the orange-brown conks characteristic of Fames rohinae, F. igniarius, Polyporus gilvus and other fungi usually included in the genera PhelUmis and Inonotus. It may be associated with Ceracis singtilaris, Ceracis punctulatiis, or Ceracis pulhiJus, as well as the tenebrionid Platydcma ellipticu7n (Fabricius), but it usually occiu's alone in the northern part of the range and in the larger, woodier conks.
Subfamily Orophiinae
Orophiina Thomson, 1863: 195. Orophyidae Kiesenwetter, 1877: 194. Octotemnidae Reitter, 1878b: 21. Apatini (in part), — Seidlitz, 1872: 90 (in family Anobiidae ) .
Included genera. Octotemmis Mellie, Paratrichapus Scott, Rhopalodontus Mellie, Scolytocis Blair, and Xylographus MeUie.
Genus Rhopolodontus Mellie
Ropalodontus Mellie, 1847: 109; Mellie, 1848: 233. Type species by monotypy, Cis perforatus Gyllenhal, 1813: 385.
Rhopalodontus Mellie, — Gaubil, 1849: 123; La- cordaire, 1857: 550; Thomson, 1863: 195; Abeille de Perrin, 1874b: 76; Kiesenwetter, 1877: 194; Reitter, 1902a: 57; DaUa Torre, 1911: 21; Winkler, 1927: 794; Lohse, 1964: 117; Lawrence, 1965: 275; Lohse, 1967: 283; Lohse, 1969: 48-52. Justifiable emendation.
Rhopalodontus (Cedriniis) Abeille de Perrin, 1876: 312. Type species, by monotypy, Rhopalodontus (Cedrinus) camelus AbeiUe de Perrin, 1876: 312.
Cedrinus AbeiUe de Perrin, — Winkler, 1927: 791.
Cis Latreille, — Gyllenhal, 1813: 385.
Included species. Rhopalodontus amer- icanus Lawrence, n. sp. [Wisconsin, see p.
507]; R. armifrons Reitter, 1913: 655 [Algeria]; R. baudtieri Abeille de Perrin, 1874a: 52 [southern Europe]; R. camelus Abeille de Perrin, 1876: 312 [Lebanon]; R. harmandi Lesne, 1917: 191 [Japan]; R. novorossicus Reitter, 1902a: 58 [south- eastern Europe]; Cis perforatus Gyllenhal, 1813: 385 [Eurasia]; R. perrini Reitter, 1878d: 221 [southeastern Europe]; R. populi Brisout de Barneville, 1877: cvii [southern Europe]; R. strandi Lohse, 1969: 50 [Scan- dinavia]. Total: 10 species.
Doubtfully included species. Rhopalo- dontus gyllenhali Gistel, 1857: 59 [Europe]; R. sassaparilhe Motschulsky, 1852: 22 [Europe]. See discussion below.
Excluded species. Rhopalodontus japoni- cus Nobuchi (see p. 503); R. tokunagai Nobuchi (see p. 488).
Members of this genus may be dis- tinguished from species of Octotemmis by the lO-segmented antennae, spinose pro- tibial apices (Fig. 59), and vestiture of moderately long and fine, erect hairs. The male abdominal fovea in Rhopalodontus is covered (Fig. 31), as it is in Octotemnus, and this condition separates the two genera from Xylographus, Scolytocis, and Paratri- chapus. In Scolytocis the antennae are 9- segmented, with a more compact club, while in Paratrichapus the tarsi have three segments (instead of four). Tlie species of Xylographus are usually distinguished from those of Rhopalodontus by the form of the tibiae, which are spinose along the outer edge; this character may break down, however, in certain Indo-Pacific species. The male se.xual ornaments in this genus are not spectacular and usually consist of tubercles on the frontoclypeal ridge and occasionally tlie pronotal apex.
Rhopalodontus gyllenhali and R. sassa- parillae are doubtfully included in this genus, since types have not been seen and descriptions are completely inadequate. Neither name has been used since, and both probably should be rejected alto- gether. Rhopalodontus japonicus and R.
North American Ciidae • Lawrence 507
tokunagai have been transferred to Sulcacis and Strigocis, respectively {see p. 503 and 488).
Species of Rhopalodontus occur through- out Eurasia from Scandinavia to northern Africa, Burma, and Thailand (undescribed fomis), and a single species has recently been discovered in North America {see below).
Rhopalodontus americanus NEW SPECIES
Holotype. S, WISCONSIN: Woodruff, Oneida Co, July 26, 1968, C. H. Porter, No. 68-57, ex Pohjporiis hetitUnus [FMNH]. Allotype, 9, same data [FMNH].
Male. Length 1.82 mm. Body 2.21 X as long as broad, strongly convex. Head and pronotum reddish orange, elytra yellowish orange (teneral). Vestiture consisting of long, fine, erect, yello\vish hairs. Vertex slightly concave with a low, median ele- vation; frontoclypeal ridge bearing 2 rounded tubercles which are separated by 2.0 basal widths. Antennal segment III 1.40 X as long as IV. Pronotum 0.83 X as long as broad, widest at posterior third; anterior edge strongly romided, slightly flattened at middle; sides strongly rounded, the margins very narrow and weakly crenu- late, not visible from above; anterior angles not produced, rounded; disc strongly con- vex, even; surface finely granulate and shiny; punctures 0.20 X as large as scu- tellar base and separated by 0.50 to 1.0 diameter. Elytra 1.45 X as long as broad and 1.92 X as long as pronotum; sides sub- parallel, apices blunt; pmictation single and confused; pmictures coarser and denser than those on pronotimi, about 0.30 X as large as scutellar base and separated by 0.20 to 0.50 diameter, each punctiu-e bear- ing a fine, erect, yellowish hair, which is about 1.25 X as long as scutellar base. Prosternum (Fig. 25) short and concave, 0.33 X as long as a procoxal cavity; inter- coxal process short and subtriangular. Pro- tibia with outer apical angle produced, rounded, and bearing 9 stout spines (Fig.
59). Metasternum 0.46 X as long as wide; strongly convex, flattened in middle; suture absent. Abdominal stemite III bearing a median, transversely oval, pubescent fovea, which is partly covered by a posteriorly projecting, subtriangular flap (Fig. 31).
Female. Length 1.97 mm. Body 2.19 X as long as broad. Vertex slightly convex; frontoclypeal ridge simple. Pronotum 0.82 X as long as broad; anterior edge strongly rounded. Elytra 1.44 X as long as broad and 1.93 X as long as pronotum. Stemite III without a pubescent fovea.
Variation. Pronoitum yellowish orange to reddish browii, usually reddish brown. Elytra yellowish to reddish brown, usually reddish browm. Frontoclypeal tubercles in some males sul)acute. Anterior edge of pronotum in large males slightly emargmate at middle. Size and dimensions vary as follows in a series of 13 males and 13 females:
TL mm: <^ 1.65-2.05 (1.84 ± 0.030),
9 1.75-1.97 (1.87 + 0.022); TL/EW: c^ 2.00-2.21 (2.12 + 0.017),
5 2.11-2.26 (2.15 + 0.011); PL/PW: c^ 0.77-0.85 (0.81 + 0.008),
9 0.81-0.87 (0.83 + 0.005); EL/EW: $ 1.32-1.47 (1.39 + 0.012),
9 1.36-1.51 (1.42 + 0.010); EL/PL: S 1.78-2.00 (1.90 + 0.016),
9 1.81-2.04 (1.93 + 0.019); GD/EW: S 0.85-0.93 (0.88 + 0.006),
9 0.85-0.92 (0.89 + 0.006).
Paratypes. WISCONSIN: 26, Woodruff, Oneida Co., July 26, 1968, C. H. Porter, No. 68-57, ex Polyporm betiiliniis [JFL, MCZ, USNM, UWS]; 5, same locality and date, C. H. Porter, No. 68-59, ex Russida sp. [JFL, UWS]; 5, Vilas Co. (no specific lo- cality), July 26, 1968, C. H. Porter, No. 68-80, ex Polyponis fibrillosus [UWS]; 5, same locality and date, C. H. Porter, No. 68-8.3, ex Fames sp. [JFL, UWS].
Distribution. Known only from northern Wisconsin.
Host fungi. Polyporus betulimis [1(1)];
508 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
Pohjponis fihriUosiis [1]; Fornes sp. [1]; Russ'uh sp. [1].
Discussion. Tliis species may be dis- tinguished from all other North American Ciidae by the characters given in the generic key. It is similar to Sulcacis curtii- lus in general form, pmictation, antennal segmentation, and protibial stnictm-e, but in that species the vestiture consists of shorter and stouter bristles, the prostemal intercoxal process is much longer, and the abdominal fovea in tlie male is not covered by a subtriangular flap. R. americanus is apparently related to the Palaearctic spe- cies R. pei-fomtus (Gyllenhal), R. novoros- sictis Reitter, and R. strancU Lohse. It differs from R. peiioratiis in being more elongate (TL/EW more than 2.00), and having a shorter and broader prostemal intercoxal process. It may be distinguished from R. novorossicus by the longer hairs, coarser elytral punctation, and relatively simple pronotal apex. R. americanus closely resembles R. strand i, which was described from Scandinavia, but differs from the latter in being smaller and rel- atively shorter and broader, according to Lohse (in litt.), who has examined speci- mens of both species. According to Lohse's description, R. strandi is 1.7 to 2.2 mm long and 2.35 x as long as broad {see figures for R. americanus above). The aedeagus in R. americanus is similar to that of R. strandi (Lohse, 1969: 50, fig. 2b), except that the median lobe is sHghtly longer than the tegmen.
Although this new American species is known only from northern Wisconsin, it may be more widespread in the Boreal parts of the continent, where it may breed in Polyporus betulinus and perhaps Fom.es fomentariiis, both of which are common on birch. Host records are not available for R. strandi, but R. pei-foratus has been reported from Fomes fomentariiis and Polyporus betidinus in Scotland (Paviour- Smith, 1969) and from F. fomentarius in Scandinavia (Saalas, 1923: Palm, 1959).
Genus Octotemnus Mellie
Octotcmnus Mellie, 1847: 110; Mellie, 1848: 384; Lacordaire, 1857: 554; Jacquelin du Val, 1861: 239; Abeille de Penin, i874b: 90; Casey, 1898: 91; Reitter, 1902a: 60; Dalla Tone, 1911: 26; Duiy, 1917: 28; Leng, 1920: 247; Arnett, 1962: 830. Type species, by subsequent designation, Cis filabriculus Gyllenhal, 1827: 627 (Jacquelin du Val, 1861: 239).
Oropht/us (Octotemnus) Mellie, — Kiesenwetter, 1877: 197.
Orophyus Redtenbacher, 1847: 350; Lacordaire, 1857: 553; Thomson, 1863: 196; Dalla Torre, 1911: 26. Type species, by monotypy, Cis mandihularis Gyllenhal, 1813: 717.
Octotemnus ( Orophius ) Redtenbacher, — Reitter, 1878a: 21; Reitter, 1902a: 61.
Orophyus Kiesenwetter, 1877: 195. Incorrect sub- sequent spelling.
Orophinus Marschall, 1873: 226. Incorrect sub- sequent spelling.
Cis Latreille (in part), — Gyllenhal, 1813: 717; Gyllenhal, 1827: 627.
Included species. Orophius dilutipes Blackburn, 1891: 308 [Australia]; Orophius diaholicus Pic, 1916: 6 [India]; Cis gjabri- cuhis Gyllenhal, 1827: 629 [Eurasia]; Orophius hehridarum Blair, 1941: 178 [New Hebrides]; Octotemnus (Orophius) japonicus Miyatake, 1954: 64 [Japan]; Octotcmnus laevis Casey [North America, see p. 509]; Orophius laminifrons Motschul- sky, 1860: 17 [Japan]; Cis mandihularis Gyllenhal, 1813: 717 [Eurasia]; Octotem- nus mindanaonus Chujo, 1966: 530 [PhiUp- pines]; Octotemnus omogensis Miyatake, 1954: 61 [Japan]; Octotemnus opacus Mellie, 1848: 386 [Madeira]; Octotemnus palawanus Chujo, 1966: 531 [Philippines]; Octotemnus parvulus Miyatake, 1954: 62 [Japan]; Octotemnus {Orophius) puncti- dorsum Miyatake, 1954: 63 [Japan]; Oro- phius quadridentatus Pic, 1916: 6 [Indo- China]; Orophius testaceus Pic, 1916: 6 [India]; Octotemnus walkeri Blair, 1940: 136 [Austraha]. Total: 17 species.
This is a well-defined group of Orophii- nae in which the antennae are 8-segmented, the tibiae spinose along most of the outer edge (Fig. 60), and the abdominal fovea in the male is covered by a flap (Fig. 31).
North American Ciidae • Lawrence 509
In Xylographus and Scohjtocis the tibiae are similar, but the antennae are 10- or 9-segmented and the abdominal fovea is naked or absent. In Rhopalodontus and Paratrichapus, the tibiae are spinose at the apices only and the antennae are 10-seg- mented. Most species of Octotemnus are subglabrous with a few scattered, erect hairs; the Madeiran species, O. opaciis, however, has die entire surface covered with fine, decmnbent, but easily visible hairs. Secondary sexual characters in the genus are unique in that males of several species have enlarged mandibles resem- bling those of stag beetles (Lucanidae). Males of some species have lateral seti- ferous tubercles on the vertex (also in Xijlographus), but the pronotal apex is never modified. The genus occurs through- out the Palaearctic and Indo-Pacific re- gions, with a single species occm-ring in the northern part of North America.
Octotemnus laevis Casey
Octotemnus laevis Casey, 1898: 91; Blatchley, 1910: 901 (dist, biol.); Dury, 1917: 27 (dist., biol.); Gibson, 1917: 150 (dist.); Gibson, 1918: 113 (dist.); Weiss and West, 1920: 8 (dist, biol.); Blackman and Stage, 1924: 86 (biol.); Graves, 1960: 66 (bid.). Type locality: "Rhode Island." Holotype, S, Casey Coll., USNM.
Octotemnus denudatus Casey, 1898: 91; Gibson, 1915: 137 (dist); Dury, 1917: 27 (syn.); Weiss and West, 1920: 8 (dist., biol.); Fall, 1926: 200 (dist); Hatch, 1962: 235, pi. 48, fig. 9 (dist., biol.). Type locality: California. Holotype, $, Casey Coll., USNM.
Distribution. Widespread throughout the northern part of North America, from southern Alaska to Quebec and Nova Scotia, south along the Pacific Coast to Monterey County, California, in the Sierra Nevada to Tulare County, in the Midwest to southern Iowa and Kansas, and on the East Coast to Alabama ( Fig. 109 ) . Marginal records: ALASKA: Skagway. BRITISH COLUMBIA: Terrace. ALBERTA: Ed- monton. MANITOBA: Winnipeg. QUE- BEC: St. Jean; Laniel. NOVA SCOTIA: Truro. ALABAMA: (no specific locality).
MINNESOTA: Comiorant, Becker Co. MANITOBA: Aweme. ALBERTA: Cyi^ress Hills. WASHINGTON: Kooskooskie, Walla Walla Co. CALIFORNIA: Dorset Camp, Sequoia National Park, Tulare Co.; Big Sur, Monterey Co.
Host fungi. Polyportis versicolor [86 (49)]; Polyponts hirsutus [11(1)]; Pohj- porus puhesceivs [7(4)]; Lenzites betulina [6(2)]; Polyporus conchifer [5(4)]; Gano- derma apphnatum [5(1)]; Polyportis ahietinus [1]; Polyporus adustus [1]; Polyporus albellus [1]; Polyporus galactinus [1]; Polyporus pargamenus [1]; Polyporus squamosus [1]; Polyporus sulphureus [1]; Stereum sp. [1].
Discussion. This species is easily dis- tinguished by the ovoid body forai, spinose tibiae, 8-segmented antemiae, subconical procoxae with incomplete intercoxal proc- ess, vestiture of very short, fine hairs and scattered long fine hairs, and the triangular flap covering the abdominal fovea in the male. In Rhopalodontus americanus, the prostemum and male abdomen are similar, but the body is more elongate and parallel- sided, die tibiae spinose at the apices only, the antennae 10-segmented, and the vesti- ture of long, fine hairs only.
Octotemnus laevis is closely related to and possibly conspecific with Octotemnus glabricidus (Gyllenhal) of the Palaearctic region. The range is typically northern and the species is not veiy common in the Southeast. This is one of the most common species breeding in Polyporus versicolor and its relatives. Like Cis fuscipes, it ap- pears to be equally common in die North- east and on the Pacific Coast.
LITERATURE CITED
Abeille de Perrin, E. 1874a. [Note]. Bull. Soc.
Ent. France, 1874: 52-53. . 1874b. Essai Monograpliique sur les
Cisides Europeens et Circamediterraneens.
Marseille, Camoin. 100 pp. . 1876. Notes sur les Cisides europeens et
circamediterraneens. Ann. Soc. Ent. France,
45: 309-314.
510 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
Arnett, R. H. 1962. The Beetles of the United States (A Manual for Identification). Part V. Fascicle 98. Cisidae. Washington, D. C, Catholic University of America Press. Pp. 827-830, fig. 1.98.'
Bach, M. 1852. Kaferfauna fiir Nord- und Mit- teldeutschland mit besonder Riicksicht auf die preussischen Rheinlande. Band II [first part]. Coblenz, Holscher. 148 pp.
Barber, H. S. 1913. Notes on Rhipidandri ( Coleoptera ) . Proc. Ent. Soc. Washington, 15(4): 188-193.
Bernet-Kempers, K. J. W. 1923. Abbildungen von Fliigelgeader der Coleopteren. Ent. Mitt, Berlin-Dahlem, 12(2): 71-115.
Blackburn, T. 1888. Further notes on Australian Coleoptera, witli descriptions of new species. Trans. Proc. Rept. Roy. Soc. South Australia, 10: 52-71, 177-287.
. 1891. Further notes on Australian
Coleoptera, with descriptions of new genera and species. Trans. Proc. Rept. Roy. Soc. South Austraha, 14: 292-345.
. 1907. Further notes on Austrahan
Coleoptera, with descriptions of new genera and species. Trans. Proc. Rept. Roy. Soc. South Australia, 31: 231-299.
Blackburn, T., and D. Sharp. 1885. Memoirs on the Coleoptera of the Hawaiian Islands. Sci. Trans. Royal Dublin Soc, ser. 2, 3: 119- 300, pis. 4-5
Blackman, M. W., and H. H. Stage. 1924. On the succession of insects Hving in the bark and wood of dying, dead and decaying hick- ory. New York State Coll. Forestry, Syracuse, Tech. Publ., 17: 1-269.
Blair, K. G. 1927. The Heteromera and some other famihes of Coleoptera from Polynesia collected on the 'St. George' Expedition, 1925. Ann. Mag. Nat. Hist., ser. 9, 20: 161-174.
. 1940. Some new species of Cisidae ( Col. )
from Australia. Ent. Mon. Mag., 76: 131- 136.
1941. Cisidae (Col.) from the New
Hebrides. Ent. Mon. Mag., 77: 178-180.
Blaisdell, F. E. 1892. Notes on the habits of some species of Coleoptera observed in San Diego County, Cal. Insect Life, U. S. D. A., Div. Ent, Per. Bull., 5(1): 33-36.
Blatchley, W. S. 1910. An illustrated descrip- tive catalogue of the Coleoptera or beetles (exclusive of the Rhynchophora ) known to occur in Indiana. Bull. Indiana Dept. Geol. Nat. Res., 1, 1385 pp.
. 1918. On some new or noteworthy
Coleoptera from tlie west coast of Florida. — IV. Canadian Ent., 50: 52-59.
. 1923. Notes on the Coleoptera of south- ern Florida with descriptions of new species. Canadian Ent., 55: 13-20, 30-36.
. 1928. Notes on some Florida Coleoptera
with descriptions of new species. Canadian Ent., 60: 60-73.
BoviNG, A. G., and F. C. Craighead. 1930. An illustrated synopsis of the principal larval forms of the Coleoptera. [part]. Ent. Amer., N. S., 11(1-3): 1-257, pis. 1-85.
, and . 1931. An illustrated synopsis
of the principal larval forms of the order Coleoptera. [part]. Ent. Amer., N. S., 11(4): 258-351, pis. 86-125.
BuETHES, J. 1922. Descripcion de varios Coleop- teros de Buenos Aires. An. Soc. Cient. Ar- gentina, 94: 263-305.
Brisout de Barneville, C, and H. 1877. [Note]. Bull. Seances Soc. Ent. France, 1877: cvii.
Broun, T. 1880. Manual of tlie New Zealand Coleoptera. Part 1. New Zealand Mus. Geol. Surv. Dept., WelHngton. 652 pp.
Brown, W. J. 1929. The entomological record, 1929. Ann. Rep. Ent. Soc. Ontario, 60: 146-157.
. 1931. The entomological record, 1930.
Ann. Rep. Ent. Soc. Ontario, 61 : 86-95.
Caillol, H. 1914. Catalogue des Coleopteres de Provence, d'apres des documents recueillis et groupes. 3me Partie. Societe Linneenne de Provence, Marseille.
Casey, T. L. 1884. Contributions to the Descrip- tive and Systematic Coleopterology of North America. Part I. Philadelphia. 60 pp., 1 pi.
. 1890. Coleopterological notices. II. Ann.
New York Acad. Sci., 5: 307-504, pi. 4.
. 1898. Studies in the Ptinidae, Cioidae,
and Sphindidae of America. Jour. New York Ent. Soc, 6: 61-93.
1900. Review of the American Corylophi-
dae, Cryptophagidae, Tritomidae and Der-
mestidae, with other studies. Jour. New York
Ent. Soc, 8: 51-72. Champion, G. C. 1913. Notes on various Central
American Coleoptera, witli descriptions of
new genera and species. Trans. Ent. Soc.
London, 1913: 58-169, pis. 3-4. Chujo, M. 1941. Description of a new ciid-beetle
from Japan. Kontyu, 15(2): 85-86. . 1966. Some Bostrichidae, Languriidae,
and Ciidae (Col.) from the Philippine and
Bismarck Islands collected by tlie Noona Dan
Expedition. Ent. Medd., 34(6): 521-532. Clark, M. E. 1956. An annotated hst of the
Coleoptera taken at or near Terrace, British
Columbia. Part 3. Proc. Ent. Soc. British
Columbia, 52: 39-43. CocKERELL, T. D. A., AND H. C. Fall. 1907. The
Coleoptera of New Mexico. Trans. Amer.
Ent. Soc, 33: 145-272. Coquerel, C. 1849. Observations entomologi-
ques sur divers Coleopteres recueillis aux
North American Ciidae • Lawrence 511
Antilles. Ann. Soc. Ent. France, ser. 2, 7:
441-454, pi. 14. Criddle, N. 1921. The entomological record,
1920. Ann. Rep. Ent. Soc. Ontario, 51:
72-90. . 1926. The entomological record, 1925.
Ann. Rep. Ent. Soc. Ontario, 56: 94-107. Crowson, R. a. 1955. The natural classification
of the families of Coleoptera. London, Lloyd.
187 pp. . 1960. The phylogeny of Coleoptera. Ann.
Rev. Ent, 5: 111-134. . 1966. Observations on the constitution
and subfamilies of the family Melandryidae.
Eos, 41: 507-513. . 1967. The natural classification of the
families of Coleoptera. Addenda and corri- genda. Ent. Mon. Mag., 103: 209-214.
Dahlbom, a. G. 1843. Hymenoptera Europae praecipue borealis. Tomus [1]: Sphex in sensu Linneano. Lund, 528 pp.
Dalla Torre, K. W. von. 1911. Cioidae. Pars. 30. In W. Junk and S. Schenkling (eds.), Coleopterorum Catalogus. Berlin, Junk. 32 pp.
Delkeskamp, K. 1959. Sekundiire Geschlechts- merkmale bei Erotyliden. Wiss. Zeit. Martin- Luther-UniversitJit Halle-Wittenberg, Matli.- Nat, 8(6): 1089-1098.
Desmarest, E. 1860. Tome 17. Coleopteres. Partie 3. In J. C. Chenu (ed.). Encyclopedic d'Histoire Naturelle ou Traite Complet de cette Science. Paris, Didot. 360 pp., 48 pis.
Doxisthorpe, H. 1916. Cis dentatus, Mellie, not a British insect. Ent. Rec. Jour. Var., 28: 155-156.
DuRY, C. 1917. Synopsis of the coleopterous fam- ily Cisidae (Cioidae) of America north of Mexico. Jour. Cincinnati Soc. Nat. Hist., 22 (2): 1-27.
. 1919. A new ciside genus with new spe- cies from Manitoba. Canadian Ent, 51: 158.
Fabricius, J. C. 1792. Entomologia Systematica Emendata et Aucta. Tomus I [Pars 1]. Copenhagen, Prost. xx -|- 330 pp.
Fall, H. C. 1901. List of tire Coleoptera of southern California . . . San Francisco, Cali- fornia Academy of Science. 282 pp.
. 1917. New Coleoptera, VI + VII. Cana- dian Ent., 49: 163-171, 385-390.
1926. A list of the Coleoptera taken in
Alaska and adjacent parts of the Yukon Terri- tory in the summer of 1924. Pan-Pacific Ent., 2: 127-154, 191-208.
Flach, C. 1882. Zwei neue deutsche Cis-Arten. Deutsche Ent. Zeit., 26: 249-251.
Forbes, W. T. M. 1926. The wing-folding pat-
terns of the Coleoptera. Jour. New York Ent. Soc, 34: 42-68, 91-1.39^ pis. 7-18.
Fowler, W. W. 1890. The Coleoptera of the British Islands. Vol. IV. LameUicornia — Serricornia — Longicornia — Phytophaga. Lon- don, L. Reeve. 411 pp.
Frost, C. A. 1930. Cis frosti Dury. Bull. Brook- lyn Ent Soc, 25: 41.
Gaubil, J. 1849. Catalogue Synonymique des Coleopteres d'Europe et d'Algerie. Paris, Maison. 296 pp.
Germain, P. 1855. Descripcion de Coleopteres de diversas especies que no se hallan en la obra del Sefior Gay. An. Univ. Chile, Santi- ago, 12: 386-407.
Gibson, A. 1915. The entomological record, 1914. Ann. Rep. Ent. Soc. Ontario, 45: 123-150.
. 1917. The entomological record, 1916.
Ann. Rep. Ent. Soc. Ontario, 47: 137-171.
. 1918. The entomological record. 1917.
Ann. Rep. Ent Soc. Ontario, 48: 99-127. and N. Criddle. 1920. The entomo-
logical record, 1919. Ann. Rep. Ent. Soc. Ontario, 50: 112-134.
GiSTEL, J. 1856. Pleroma zu den Mysterien der europaischen Insectenwelt, mit ein syste- matischen Verzeichniss der Schmetterlinge und Kafer Europas. Straubing, Schorner. 250 pp.
. 1857. Achthundert und zwanzig neue
oder unbeschriebene wirliellose Thiere. Strau- bing, Schorner. 94 pp. [Extract from Vacuna, 2: 513-606.]
Goldman, E. A. 1916. Plant records of an expedition to Lower California. Contr. U. S. Nat Herb., 16: 309-371, pis. 104-133.
GoRHAM, H. S. 1883. Cleridae, Ptinidae, Bostry- chidae, Cioidae. In F. D. Godman and O. Salvin (eds.), Biologia Centrali-Americana. Insecta. Coleoptera. Vol. 3, Part 2. London, Porter. Pp. 169-224, pi. 10, figs. 23-27.
. 1886. Supplement to Malacodermata. In
F. D. Godman and O. Salvin (eds.), Biologia Centrali-Americana. Insecta. Coleoptera. Vol. 3, Part 2. London, Porter. Pp. 31:3-360, pi. 13, figs. 21-24. -. 1898. On the serricorn Coleoptera of St.
Vincent, Grenada, and the Grenadines (Mala- codermata, Ptinidae, Bostrychidae), with de- scriptions of new species. Proc Zool. Soc. London, 1898: 315-343, pi. 27.
Graves, R. C. 1960. Ecological obser^'ations on the insects and other inhal^itants of woody shelf fungi (Basidioniycetes: Polyporaceae ) in die Chicago Area. Ann. Ent. Soc. America, 53: 61-78.
Grensted, L. W. 1940. Cissidae, Cisidae, Cioidae or Ciidae? Ent. Mon. Mag., 76: 145-146.
512 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
\
. 1947. On die formation of family names.
Ent. Mon. Mag., 83: 137-141.
Guerin-Meneville, F. E. 1847. Insectes. In T. Lefebvre (ed.), Voyage en Abyssinie Exe- cute Pendent les Annees 1839, 1840, 1841, 1842, 1843. Partie 4. Histoire naturelle.- Zoologie. Tome 6. Paris, Bertrand. Pp. 243- 398, pis. (in Atlas).
Gyllenhal, L. 1813. Insecta Svecica. Classis I. Coleoptera sive Eleuterata. Tomus I, Pars III. Scaris, Leverentz. 730 pp.
. 1827. Insecta Svecica. Classis I. Coleop- tera sive Eleuterata. Tomus I, Pars IV. Leipzig, Fleischer. 761 pp.
Hatch, M. H. 1924. A preliminary list of the Coleoptera of tlie Cranberry Lake region, New York, exclusive of the Buprestidae, Cerambycidae, and Ipidae. New York State Coll. Forestry, Syracuse, Tech. Publ., 17: 273-312.
. 1962. The beetles of the Pacific Nortli-
west. Part III: Pselaphidae and Diversicomia I. Seattle, Univ. Washington Press, ix -j- 503 pp., 66 pis.
Heatwole, H. and a. 1968. Movements, host- fungus preferences, and longevity of Bolito- therus cornutus (Coleoptera: Tenebrioni- dae). Ann. Ent. Soc. America, 61: 18-23.
HomoN, A. 1951. Verzeichniss der Kafer Mittel- europas ( Deutschland, Osterreich, Tschecho- slovakei) mit kurzen faunistischen Angaben. Stuttgart, Kernen. 536 pp.
Horn, G. H. 1894. The Coleoptera of Baja Cali- fornia. Proc. California Acad. Sci., ser. 2, 4: 302-449.
Hubbard, H. G. 1892. The inhabitants of a fungus. Canadian Ent., 24: 250-256.
Hyslop, J. A. 1915. 01:)sei-vations on the life history of Meracantha contracta (Beauv. ). Psyche, 22 : 44-48, pi. 4.
Jacquelin du Val, p. N. C. 1857. Ordre des Coleopteres Lin., pp. 137-328 in . . . F. E. Guerin-Meneville, Animaux articules a pieds articules. In R. Sagra (ed. ), Histoire Physique, Politique, et Naturelle de I'lle de Cuba. Tome 7. Paris, Bertrand. 868 pp.
. 1861. Manuel Entomologique. Genera
des Coleopteres d'Europe. Tome 3 [part]. Paris, DeyroUe. Pp. 236-240, cat. p. 168.
Jeannel, R., and R. Paulian. 1944. Morphologic abdominale des Coleopteres et systematique de I'ordre. Rev. Frangaise Ent., 11: 64-110.
. 1945. Mission scientifique de I'Omo.
Faune des terriers des rats-taupes. IV. Coleopteres. Mem. Mus. Nat'n. Hist. Nat., Paris (N. S.), 19(1): 51-147.
Kevan, D. K. 1967. On the apparent conspeci- ficity of Cis pygmaeus (Marsh.) and C. rhododactylus (Marsh.) and on otlier closely
allied species (Col., Ciidae). Ent. Mon.
Mag., 102: 138-144. Kiesenwetter, H. Von. 1877. Anobiidae. Cioidae.
Aspidophoridae. In W. F. Erichson (ed. ),
Naturgeschichte der Insecten Deutschlands.
Abtheilung 1. Coleoptera. Band 5, Halfte 1.
Berlin, Moeser. Pp. 169-200. Kraus, E. J. 1908. New bicolored Cioidae (Co- leoptera). Proc. Ent. Soc. Washington, 10:
74-81. Lacordaire, J. T. 1857. Histoire Naturelle des
Insects. Genera des Coleopteres. Tome 4.
Paris, Libr. Encycl. Roret. 579 pp. Latkeille, P. A. 1796. Precis des Caracteres
Generiques des Insectes, Disposes dans vm
Ordre Naturel. Bordeaux, 201 pp. . 1802. Histoire Naturelle, General et
Particuhere des Crustaces et des Insectes.
Tome 3. Paris, Dufart. 467 pp. Lawrence, J. F. 1965. Comments on some recent
changes in the classification of the Ciidae
(Coleoptera). Bull. Mus. Comp. Zool., 133:
273-293. . 1967a. Biology of the parthenogenetic
fungus beetle Cis fuscipes MeUie (Coleoptera:
Ciidae). Breviora, No. 258, 14 pp. . 196713. Delimitation of the genus Ceracis
(Coleoptera: Ciidae) with a revision of North
American species. Bull. Mus. Comp. Zool.,
136: 91-144.
( In preparation ) . Host preference in ciid
beetles (Coleoptera: Ciidae) inhabiting fruit- ing bodies of Basidiomycetes in Nortli America.
Leach, W. E. 1819. In G. Samouelle (ed.). The Entomologists' Useful Compendium; or an Introduction to tlie Knowledge of Britisli Insects. London, Boys. 496 pp.
LeConte, J. L. 1861. Classification of the Coleoptera of North America, [part]. Smiths. Misc. Coll., Vol. 3, Art. 3, pp. 1-208.
. 1862. Classification of the Coleoptera of
North America, [part]. Smiths. Misc. Coll., Vol. 3, Art. 3, pp. 209-278.
. 1867. List of the Coleoptera of North
America. Smiths. Misc. Coll., Vol. 6, No. 140, 77 pp. -, AND G. H. Horn. 1883. Classification of
the Coleoptera of North America. Smiths.
Misc. Coll., No. 507, xxxvii + 567 pp. Len'G, C. W. 1918. Notes on some changes in
the list of Coleoptera. Jour. New York Ent.
Soc, 26: 201-211. . 1920. Catalogue of the Coleoptera of
America, North of Mexico. Mt. Vernon, New
York, Sherman, x -|- 470 pp. Lesne, P. 1917. Notes sur divers Cisides (Col.
Ciidae). Bull. Soc. Ent. France, 1917: 190-
192.
North American Ciid.\e • Lawrence 513
. 1934. Sur le geni-e Hendccatomus Mellie
(Col. Bostiychidae ) . Bull. Soc. Ent. France, 39: 174-175.
1935. La position systematique des
Coleopteres du genre Hendecatomus. Arch. Mus. Nat. Hist. Nat., Ser. 6, 12: 427-433.
LiNDROTH, C. 1957. Principal terms used for male and female genitalia in Coleoptera. Opusc. Ent., 22: 241-256.
LiNSLEY, E. G. 1958. Geographical origins and phylogenetic affinities of tlie cerambycid beetle fauna of Western North America, pp. 299-320. In C. L. Hubbs (ed.), Zoogeography. Amer. Assoc. Adv. Sci. Publ. No. 51, x + 509 pp.
LoDiNG, H. P. 1945. Catalogue of the beetles of Alabama. Geol. Surv. Alabama, Monograph 11, 172 pp.
LoHSE, G. A. 1964. Die in Mitteleuropa verti-e- tenen Gattimgen der Cisidae. (1. Beitrag zur Kenntnis der mitteleuropiiischen Cisidae). Ent. Blatter, 60: 116-122.
. 1965. iJber einige umstrittene Cts-Arten.
(2. Beitrag zur Kenntnis europiiischer Cisi- dae). Ent. Blatter, 60: 174-181.
. 1966. Cisiden-Studien (3. Beitrag zur
Kenntnis der mitteleuropaischen Cisiden) mit 4 Abbildungen. Ent. Blatter, 62: 25-29.
. 1967. Famihe: Cisidae, pp. 280-295. In
H. Freude, K. W. Harde, and G. A. Lohse, Die Kafer Mitteleuropas. Band 7. Krefeld, Goecke and Evers. 310 pp.
1969. Cisiden Studien IV. Ent. Blatter,
65: 48-52.
Lowe, J. L. 1957. Polyporaceae of North America. The genus Fames. State Univ. Coll. Forestry at Syracuse Univ., Tech. Publ. No. 80, 97 pp.
. 1966. Polyporaceae of North America.
The genus Porta. State Univ. Coll. Forestry at Syracuse Univ., Tech. Publ. No. 90, 183 pp.
AND R. L. GiLBERTSON. 1961a. Synopsis
of the Polyporaceae of tlie soutlreastern LTfiited States. Jour. Elisha Mitchell Sci. Soc, 77: 43-61.
AND . 1961b. Synopsis of the Poly-
poraceae of tlie western United States and Canada. Mycologia, 53: 474-511.
Mannerheum, C. G. Von. 1843. Beitrag zur Kaefer-Fauna der aleutischen Inseln, der Insel Sitkha und Neu-Californiens. Bull. Soc. Imp. Nat. Moscou, 16(2): 175-314.
. 1852. Zweiter Nachtrag zur Kaefei^-
Fauna der nord-amerikanischen Laender des russischen Reiches. Bull. Soc. Imp. Nat. Moscou, 25(2): 283-387.
. 1853. Dritter Nachtrag zur Kaefer-Fauna
der nord-amerikanischen Laender des rus- sischen Reiches. Bull. Soc. Imp. Nat. Moscou, 26(3): 95-273.
Marschall, a. 1873. Nomenclator Zoologicus.
Vindobonae, Ueberreuter. 482 pp. Marseul, S. a. de. 1887. Catalogue synonymique
et geographique des Coleopteres de I'ancien-
monde. Europe et contrees hmitrophes en
Afrique et en Asie: Clavicornes (suite) . . .
vesicants [part]. Abeille, 24: 193-360. Meixner, J. 1934. Achte tJberordnung der
Pterygogenea: Coleopteroidea [part]. 7ji W.
Kiikenthal and T. Krumbach (eds.), Hand-
buch der Zoologie. Vierter Band, zweite
Halfte. Insecta 2, zweite Lieferung. Berlin.
W. de Gruyter. Pp. 1037-1132. Mellie, J. 1847. Melanges et nouvelles. Rev.
Zool. Soc. Cuvierienne, 10: 108-110. . 1848. Monographic de I'ancien genre CLs
des auteurs. Ann. Soc. Ent. France, ser. 2, 6:
205-274, 313-396, pis. 9-12. . 1849. [Note]. Bull. Soc. Ent. France,
ser. 2, 7: kxxvi-lxxxvii. Miyatake, M. 1954. Studies on the Japanese
Ciidae, I. (Coleoptera). Sci. Rep. Matsuyama
Agr. Coll., 14: 40-67, 11 pis. . 1959. The Ciidae of Amami-Oshima
(Studies on the Japanese Ciidae, II). Mem.
Ehime Univ., Sect. 4 (Agriculture), 4: 119-
124.
1960. The genus Pisenus Casey and some
notes on the family Tetratomidae (Coleop- tera). Trans. Shikoku Ent. Soc, 6: 121-135.
MoTSCHULSKY, V. 1852. Notices. Etudes Ent., 1: 21-23.
. 1860. Entomologie speciale. Insectes
du Japon. Etudes Ent., 9: 4-39.
Nakane, T., and a. Nobuchi. 1955. On a new genus and six new species of ciid-beetles from [apan (Ciidae, Coleoptera). Sci. Rep. Saikyo Univ., ser. A, 2(1): 47-52.
Nobuchi, A. 1955. Studies on the ciid-beetles from Japan (II). Akitu, 4: 105-108.
. 1960. Ciidae from Hokkaido ( Coleop- tera). Ent. Rev. Japan, 11(2): 37-41.
OvERHOLTS, L. C. 1953. The Polyporaceae of the United States and Canada. Univ. Michigan Studies, Sci. Ser., Vol. 19, xiv -|- 466 pp., 132 pis.
Pace, A. E. 1967. Life history and behavior of a fungus beetle, Boliioihcrus cornutus (Teneb- rionidae). Occ Pap. Mus. Zool. L^^niv. Michigan, No. 653, 15 pp.
Palm, T. 1946. Anteckningar om Svenska skal- baggar. II. Ent. Tidskr., 67: 11-20.
. 1952. Die Holz- und Rinden-Kafer der
nordschwedischen Laubbiiume. Medd. Stat. Skogs-Forskingsinstitut, 40(2): 1-241.
. 1959. Die Holz- und Rinden-Kafer der
siid-und mittelschwedischen Laul^baume. Opusc. Ent. Suppl., 16. 374 pp.
514 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
Panzer, G. W. F. 1809. Fauna insectoriim Ger- manicae initia oder Deutschlands Insecten. Tahrg. 9, Hefte 97-108. Niimberg, Felssecker. 13 pp.
Paviour-Smith, K. 1960. The fruiting bodies of macrofungi as haliitats for l^eetles of tlie family Ciidae (Coleoptera. Oikos, 11: 43-71, 1 pi.
. 1969. An attempt to correct some mis- takes and misconceptions about some fungus beetles and their habitats. Entomologist, April: 86-96.
Ferris, E. 1864. Description de quelques especes nouvelles de Coleop teres et notes diverses. Ann. See. Ent. France, ser. 4, 4: 275-310.
Peterson, A. 1957. Larvae of Insects. An intro- duction to Nearctic species. Part II. Coleop- tera, Diptera, Neuroptera, Siphonaptera, Mecoptera, Trichoptera. Columbus, Ohio, Peterson, v -f 416 pp.
Peyerimhoff, p. de. 1915. Notes sur la iMologie de quelques Coleopteres phytophages du Nord-Africain. (Deu.\"ieme serie). Ann. Soc. Ent. France, 84: 19-61.
. 1918. Nouveau Coleopteres du Nord- Africain. ( Vingt-huitieme note: Faune du Pin d'Alep). Bull. Soc. Ent. France, 1918: 141-143.
Pic, M. 1916. Diagnoses specifiques. Melanges Exotico-Ent., 18: 1-20.
. 1917. Descriptions abregees diverses.
Melanges Exotico-ent., 26: 1-24.
. 1922. Nouveautes diverses. Melanges
Exotico-Ent., 35: 1-32
. 1923. Coleopteres exotiques en partie
nouveaiLX (suite). L'Echange, Rev. Lirmcenne,
39: 4, 7-8, 11-12, 15-16. . 1940. Diagnoses de Coleopteres exotiques.
L'Echange, Revue Linneenne, Moulins, 56:
10-12, 13-16.
1956. Nouveaux Coleopteres de diverse
families. Ann. Hist.-Nat. Mus. Nat. Hung, (ser. nov. ), Budapest, 7: 71-92. Pielou, D. P. 1966. The fauna of Folypoms betiiliniis (Bulhard) Fries (Basidiomycetes: Polyporaceae ) in Gatineau Park, Quebec. Canadian Ent., 98: 1233-1237.
AND W. G. MATTHE^\'MAN. 1966. The
fauna of Fames fomentarius ( Linnaeus ex Fries) Kickx growing on dead bircli in Gatineau Park, Quebec. Canadian Ent., 98: 1308-1312.
AND A. N. Verma. 1968. The arthropod
fauna associated with the birch bracket fungus, Polyponis hetulinus, in eastern Can- ada. Canadian Ent., 100: 1179-1199.
Redtenbacher, L. 1847. Fauna Austriaca. Die Kiifer. First edition. [Part 3]. Vienna, Ceroid. Pp. 321-480.
Reitter, E. 1877. Beitriige zur Kiiferfauna von
Japan (Drittes Stiick). Deutsche Ent. Zeit.,
21: 369-384. . 1878a. Beitrag zur Coleopteren-Fauna der
Carpathen. Deutsche Ent. Zeit., 22: 33-64. . 1878b. Nachtriige und Ergiinzungen zur
Bearbeitung der Cioiden von H. V. Kiesen-
wetter. Deutsche Ent. Zeit., 22: 21-30. . 1878c. Neue Cioidae. Mitt. Miinchener
Ent. Ver., 2: 32-37. . 1878d. In O. Schneider and H. Leder.
Beitriige zur Kenntniss der kaukasischen
Kfifer-Fauna. Verhandl. Naturforsch. Ver.
Brimn, Abhandl., 16: 1-258. . 1880. Beitriige zur Kiiferfauna von Neu-
Zeeland. Verhandl. Naturforsch. Ver. Briinn,
Abhandl., 18: 165-183. . 1913. Eine Serie neuer Coleopteren aus
der paliiarktischen Fauna. Detusche Ent.
Zeit., 1913: 649-666. . 1884. Resultate einer coleopterologischen
Sammel-Campagne wiihrend den Monaten
Februar bis April 1883 auf den jonischen
Inseln. Deutsche Ent. Zeit, 28: 101-122. . 1885. Coleopterologische Ergebnisse einer
Excursion nach Bosnien im Mai 1884. Deutsche Ent. Zeit, 29: 193-216.
. 1902a. Analytische Uebersicht der palae-
arctischen Gattungen und Arten der Coleop- teren-Familien: Byrrhidae (Anobiidae) und Cioidae. Verhandl. Naturforsch. Ver. Briinn, 40: 1-61.
. 1902b. Neue und seltene Coleopteren,
gesammelt im Jahre 1901, in der Herzego- winia, in Dalmatien und Bosnien. Wiener Ent. Zeit., 21: 1-9.
. 1908. Verzeichnis der von Dr. F. Eichel-
baum im Jahre 1903 in Deutsch-Ostafrika
gesammelten Cis- Arten. ( Coleoptera ) . Wiener
Ent Zeit, 27: 119-124. RosENHAUER, W. 1847. Beitriige zur Insekten-
Fauna Europas. Erstes Biindchen. Erlangen,
Blaesing. x -f- 159 pp. Roth, L. M., axd E. R. Willis. 1951a. Hygrore-
ceptors in adults of TrihoUum ( Coleoptera,
Tenel)rionidae). Jour. E.xp. Zool., 116:
527-570. AND . 1951b. Hygroreccptors in
Coleoptera. Jour. E.xp. Zool., 117: 451-487. RouBAL, J. 1919. Drei coleopterologische Nova.
Deutsche Ent Zeit., 1919: 63-64. . 1936. Beschreibimg einer neuen Unter-
gattvmg, einer Varietiit und einer Form der
Coleopteren. Festschr. 60. Gebuits. Embrik
Strand, 1: 53-58. Saalas, U. 1923. Die Fichtenkiifer Finnlands. II.
Spezieller Teil 2 und Larvenbestimmungstab.
Ann. Acad. Sci. Fennicae, ser. A, Vol. 22,
xii -f 746 pp., 28 pis.
North American Ciidae • Lawrence 515
Sahlberg, J. 1901. Coleoptera nova vel minus cognita faunae Fennicae. Acta Soc. Fauna Flora Fennica, 19(3): 1-23.
ScHiLSKY, J. 1900. Heft 37, In H. C. Kiister, G. Kraatz, and J. Schilsky (eds.), Die Kafer Europas. Niimberg, Bauer and Raspe. Pp. i-iv, 1-100, A-W.
ScopoLi, J. A. 1763. Entomologia Carniolica ex- hibens Insecta Carnioliae indigena et distri- buta in ordines, genera, species, varietates methodo Linnaeana. Vienna, Trattner. 420 pp.
Scott, H. 1926. Coleoptera, Ciidae. Reports of the Percy Sladen Trust Expedition to the Indian Ocean in 1905. Vol. 8, No. 1. Trans. Linnaean Soc. London, ser. 2. Zoology, 19 (1): 1-41.
Seidlitz, G. 1872. Fauna Baltica. Die Kiifer (Coleoptera) der Ostseeprovinzen Russlands. Dorpat, Laakmann. 142 pp.
Sharp, D. 1879. On some Coleoptera from the Hawaiian Islands. Trans. Ent. Soc. London, 1879: 77-105.
AND F. MuiR. 1912. The comparative
anatomy of the male genital tube in Coleop- tera. Trans. Ent. Soc. London, 1912: 477- 642, pis. 42-78.
Simpson, G. G. 1947. Evolution, interchange, and resemblance of the North American and Eurasian Cenozoic mammaUan faunas. Evo- lution, 1: 218-220.
Strand, A. 1965. Uber die nordischen Arten der Gattung Cis Latr., Untergattung Eridanhts Thoms., mit Beschreibung einer neuen Art, Cis hanseni n. sp. (Col., Cisidae). Norsk Ent. Tidsskr., 13: 61-65.
Tanner, V. M. 1934. Coleoptera of Zion National Park, No. 2. Ann. Ent. Soc. America, 27: 43-49.
Thomson, C. G. 1863. Skandinaviens Coleoptera, synoptiskt bearbetade. Tom. 5. Lund, Lund- berg. 340 pp.
Waterhouse, C. O. 1876. New species of Coleoptera from the island of Rodriguez,
collected by tlie naturalists accompanying the
Transit-of-Venus Expedition. Ann. Mag. Nat.
Hist., ser. 4, 18: 105-121. Weiss, H. B. 1919a. Notes on Ceracis sallei Mellie
and Brachycis hrevicollis Csy. bred from
fungi. Bull. Brooklyn Ent. Soc, 14: 144-147. . 1919b. Notes on Sulcacis lengi Dury and
Orchesia castanea Mels., breeding in fungi.
Canadian Ent., 51 : 203-204. . 1920a. Notes on tlie fungus beetle, Cis
cijlindricus Dury. Bull. Brooklyn Ent. Soc,
15: 110-111. . 1920b. Notes on the fungus-beetle, Cis
vitula Mann. Psyche, 27: 133-134.
1923. More notes on fungous insects.
Canadian Ent., 55: 199-201. Weiss, H. B., and E. West. 1920. Fungous in- sects and their hosts. Proc. Biol. Soc Wash- ington, 33: 1-20, 1 pi.
. 1921a. Additional fungous
AND
insects and their hosts. Proc. Biol. Soc. Wash- ington, 34: 59-62.
. 1921b. Additional notes on
AND
fungous insects. Proc. Biol. Soc. Washington, 34: 167-172, 1 pi. AND . 1922. Notes on fungous in-
sects. Canadian Ent., 54: 198-199. Wilson, J. W. 1930. The genitaha and wing
venation of the Cucujidae and related famihes.
Ann. Ent. Soc America, 23: 305-358. Winkler, A. (ed.) 1927. Catalogus Coleoptero-
nun Regionis Palaearcticae. Pars 7. Vienna,
A. Winkler. Pp. 753-880. ZiEGLER, D. 1845. Descriptions of new North
American Coleoptera. Proc. Acad. Nat. Sci.
Philadelphia, 2: 266-272. Zimmerman, E. C. 1939. Cis of Maui and
Lanai (Coleoptera, Ciidae). Proc. Hawaiian
Ent. Soc, 10: 343-348. . 1942. Ciidae of Guam. In Insects of
Guam, I. Bull. B. P. Bishop Mus., 172:
47-52.
516 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
14
I 1
15
18
17
20
21
Fig. 1. Cis vitulus Mannerheim, antennal club [.076 mm]. 2. Same, section of terminal club segment, showing sensillifer [.018 mm]. 3. C. vitulus Mannerheim, head and pronotal apex of male, dorsal view. 4. P/esioc/s cribrum Casey, same. 5. C. niedhauki, n. sp., same [.125 mm]. 6. C. miles (Casey), same [.125 mm]. 7. C. Castlei (Dury), same [.125 mm]. 8. En- nearfhron aurlsquamosum, n. sp., same [.125 mm]. 9-13. Diagrammatic cross-sections of prosternum and hypomero, showing concave, biconcave, flat, tumid, and carinate prosterna. 14. Strigocis opacicollis Dury, prothorax of mole, lateral view [.125 mm]. 15. Su/coc/s lengi Dury, same [.125 mm]. 16. Octofemnus laevis Casey, prothorax, anterior view [.125 mm]. 17. Cis fusc/pes Mellie, same. 18. Orthocis punctatus (Mellie), head and pronotum, dorsolateral view. 19. O. /ongu/us Dury, same. 20. Hadraule elongatula (Gyllenhal), head and pronotum, dorsal view. 21. H. explanata, n. sp., same. Unless otherwise indicated, 1 line = .250 mm.
North American Ciid.\e • Lawrence
517
' ' 28
26
Figs. 22-28. Prothorax of male, ventral view. 22. Orfhoc/s punctalus (Mellie). 23. Cis (uscipes Mellie. 24. Ceracis thoracicornis (Ziegler). 25. Rhopalodontus americanus, n. sp. 26. Cis vitulus Monnerheim. 27. Malacocis brevicollis (Casey). 28. Hadraule blaisdelli (Casey) [.125 mm). 29. Cis crinitus, n. sp., section of elytral surface, showing long, erect, dark bristles and short, decumbent, light hairs. 30. Cis horridulus Casey, same, showing longer, erect ond shorter, inclined bristles. 31. Rhopo/odontus omericanus, n. sp., abdomen of mole, ventral view. 32. Cis tetracentrum Gorham, same. 33. Octotemnus laevis Casey, pectus (meso- and metathorax), ventral view (.125 mm]. 34. C/'s vitulus Monnerheim, same. 35. Ceracis similis Horn, pronotal apex of male, dorsal view. 36. Ceracis schaefferi Dury, same. 37. Orfhocis huesanus Kraus, head of male, anterodorsai view (.125 mm). 38. Orthocis punctafus (Mellie), elytral apices, posterodorsal view. Unless otherwise indicated, 1 line ^ .250 mm.
518 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
43 I 1 \
Figs. 39-42. Outline of male, dorsal view [all drawn to some scale). 39. Cis megasf/cfus, n. sp. 40. C. stereophilus, n. sp. 41. C. rotundulus, n. sp. 42. C. cornel//, n. sp. 43. C. tetracentrum Gorham, anterior angle and lateral margin of pronotum, dorsal view [1 line =z .062 mm]. 44. C. discolor, n. sp., same. 45-60. Rigtit tibia of male, posterior view [all drawn to same scale]. 45. Orthocis punctatus (Mellie). 46. Do//choc/s manitoba Dury. 47. Cis lestivulus, n. sp. 48. C. coy- ensis, n. sp. 49. C. robiniophilus, n. sp. 50. C. corne///, n. sp. 51. C. rotundulus, n. sp. 52. C. acritus, n. sp. 53. C. sfereop/i/Zus, n. sp. 54. C. levettei (Casey), showing apical spines. 55. C. megost/ctus, n. sp. 56. C. discolor, n. sp. 57. Malacocis brevicollis (Casey). 58. Ceroc/s fhoroc/corn/s (Ziegler). 59. Rhopalodontus americanus, n. sp. 60. Octotemnus laevis Casey.
North American Ciidae • Lawrence 519
71
80^
Figs. 61-67. Abdominal sternite VIII of male [all drawn to same scale]. 61. C/s robiniophilus, n. sp. 62. C. discolor, n. sp. 63. C. acr/fus, n. sp. 64. C. megasf/cfus, n. sp. 65. C. festivulus, n. sp. 66. C. stereophilus, n. sp. 67. C. cornelli, n. sp. 68-77. Tegmen of oedeagus [all drawn to scale). 68. C. discolor, n. sp. 69. C. frisfis Mellie. 70. C. striolatus Casey. 71. C. rob/n/ophi/us, n. sp. 72. C. festiVu/us, n. sp. 73. C. acritus, n. sp. 74. C. megosf/cfus, n. sp. 75. C. amer- /canus Mannerfieim. 76. C. cornelli, n. sp. 77. C. stereop/i/7u;, n. sp. 78-84. Median lobe of oedeagus, outline only [all drawn to same scale]. 78. C. megastictus, n. sp. 79. C. cornelli, n. sp. 80. C. acritus, n. sp. 81. C. discolor, n. sp. 82. C. stereophi/us, n. sp. 83. C. rob/nioph/'/us, n. sp. 84. C. lestivulus, n. sp.
520
Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
85
Fig. 85. Diagram of terminal abdominal segments and aedeagus, ventral view. 86. Diagram of aedeagus, lateral view. 87. Map of four regions of North America discussed on p. 432 and Tables 2 and 3. 88-91. Distribution maps. 88. Cis levettei (Casey) [full dot] and C. morit/mus (Hatch) [half dot). 89. P/es/ocis cribrum Casey [full dot] and Cis rotundulus, n. sp. [half dot]. 90. Cis ephippiatus Mannerheim. 91. Cis borridulus Casey [full dot] and C. Iiystriculus Casey [half dot]. [S8 ;= sternite VIII; 9 = segment IX or genital ring; bp = basal piece; teg ^ tegmen; ml = median lobe.]
North American Ciidae • Lawrence 521
Figs. 92-100. Distribution maps. 92. Dolichocls manitoba Dury [full dot] and Cis ursulinus Casey [half dot). 93. Cis americanus Mannerheim. 94. C. fetrocenfrum Gorfiam [full dot] and C. ongusfus Hatchi [fiolf dot]. 95. C. hirsutus Casey. 96. C. crinitus, n. sp. 97. C. castlel (Dury). 98. C. creberrimus Mellie [Nortti American localities only]. 99. Su/coc/s lengi Dury. 100. Sfrigoc/s opacicollis Dury.
522 Bulletin Museum of Comparative Zoology, Vol. 142, No. 5
|
^ r |
.^^ |
1 1 |
|
|
''—^ ' |
• • |
• 1 |
|
|
IK:: |
•• |
||
|
I) |
\ |
||
|
Y |
\\ |
■y |
|
|
^^ |
\9 |
.•.%'v\ |
|
|
107 |
^^ |
V.
L
108
Figs. 101-110. Distribution maps. 101. Cis sfriolatus Casey [full dot] and C. versicolor Casey [half dot). 102. C. biarma- tus Mannerheim. 103. C. fristis Mellie. 104. C. pistoria Casey. 105. C. vltulus Mannerheim [full dot] and C. congesfus Casey [half dot]. 106. C. subf///s Mellie [full dot] and C. acritus, n. sp. [half dot]. 107. C. duplex Casey. 108. Mala- coc/s farev/co///s (Casey). 109. Ocfofemnus laevis Casey. 110. Sulcoc/'s curtulus (Casey).
Harvard MCZ Libra
2044 066 304 163
|
i « T ft 8 |
•^' it - ^^- , |
|
' '6 ,-* « t :l .t .i |
■ t Ht ^«! ■■ |
|
* « ? -^ '^ '■ ' " |
■ ^ ;« i^ S * |
|
^ ft' « * V- V vv ,. ., |
•" n t f ^^ |
|
iT' 't * .« "Jf ^ ^? ^ |
' ^ .« ;t Ji |
|
^ .i i :% |
|
|
* |
•« ^ if |
4 «
t'«r
f «r
'i « It
'1 f .:'«!;
%*'"l^'''^
V'^-S*-
^*:%- % P '^'
m- %■■
»: W
!' t
f , r, * I
u %\ *: f n »,
|
f E''' BMr*! |
||
|
" %F^^' ^ ■ ■ |
||
|
r, 'i |
||
|
^.^- .>v^. . |
./J^, i*..^,. |
|
|
k. * ^. |
.. |
■^' It'
*■ p
# *;. '^
^i ^:. -^•.. 'iL ¥.^K^
^^.^^^^
iftM. wh; Hk>.< <«£} "^ " llK>.- li'i:' liTKi Slir!
m"€
M ;t .4
n -^ ■€
w'-m
m'M
^ m
|
■ ■■ fr- |
■ |
|
; ,.1 ■ |
P 1 |
|
■ |
«:• J.' |
|
H |
« |
|
i |
m:''M
■v: JS;
#■ ■^•
# *,ll
■•■* ■*>,""
I