A Generic Revision--And Phylogenetic Study of the Family Kaloter- Mitidae (Isoptera)
Total Page:16
File Type:pdf, Size:1020Kb
A GENERIC REVISION--AND PHYLOGENETIC STUDY OF THE FAMILY KALOTER- MITIDAE (ISOPTERA) KUMAR KRISHNA BUL-LETIN OF THE, AMERICAN MUSEUM OF NATURAL HISTORY VOLUM'E 122 ARTICLE 4 NEWV. YORK: 1961 A GENERIC REVISION AND PHYLOGENETIC STUDY OF THE FAMILY KALOTERMITIDAE (ISOPTERA) A GENERIC REVISION AND PHYLOGE- NETIC STUDY OF THE FAMILY KALOTERMITIDAE (ISOPTERA) KUMAR KRISHNA Department of Zoology The University of Chicago THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY AT THE UNIVERSITY OF CHICAGO BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY VOLUME 122 : ARTICLE 4 NEW YORK :1961 BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY Volume 122, article 4, pages 303-408, figures 1-81, tables 1-6 Issued September 15, 1961 Price: $1.50 a copy CONTENTS INTRODUCTION. 309 Acknowledgments. 310 . Material and Technique . 310 . Terminology 311 SYSTEMATIC REViSIONS.IN............ 312 . Family Kalotermitidae Banks, 1919. ........ 312 . Key to the Genera of the Family Kalotermitidae . 315 Genus Proelectrotermes von Rosen, 1913 . 317 . Genus Electrotermes von Rosen, 1913. 318 . Postelectrotermes, New Genus. 319 . Genus Neotermes Holmgren, 1911 . * . .... .. 321 Genus Rugitermes Holmgren, 1911. 325 Genus Eucryptotermes Holmgren, 1911. ... 328 ? Genus Prokalotermes Emerson, 1933 . 331 . Genus Kalotermes Hagen, 1853 . 331 . Genus Paraneotermes Light, 1934 . 336 . Ceratokalotermes, New Genus. .338. Comatermes, New Genus. 341 . Genus Glyptotermes Froggatt, 1896 . 343 . Genus Calcaritermes Snyder, 1925. 348 . Genus Pterotermes Holmgren, 1911. 349 . Incisitermes, New Genus. 353 . Genus Allotermes Wasmann, 1910. 358 . Marginitermes, New Genus. 358 . ...... ..... .. Tauritermes, New Genus . 361 . Genus Proneotermes Holmgren, 1911. 363 . Bifiditermes, New Genus. 365 . Bicornitermes, New Genus . 370 . Genus Epicalotermes Silvestri, 1918 . 374 . Genus Procryptotermes Holmgren, 1910. 376 . Genus Cryptotermes Banks, 1906 .... 379 . PHYLOGENY. 383 . DISCUSSION.. ... 389 SUMMARY. 400 . BIBLIOGRAPHY. 400 . 307 INTRODUCTION THE FAMILY Kalotermitidae consists of 353 the traditional classical practice of weighting living and fossil species, grouped under 24 characters. They argue that the procedure of genera in the present revision. The family is assigning significance to either adaptive or primitive in its morphology, nesting behavior, non-adaptive characters brings in subjectiv- and social organization. The species found ity in classification. They have proposed what and maintain their colonies strictly in wood, they consider to be a more objective, quanti- without necessary soil connections, often in tative approach in grouping 97 species of sound damp or dry wood. They do not con- megachilid bees. Their basic procedure con- struct definite nests, as do the higher ter- sisted of measuring correlations in 122 char- mites, but instead excavate an irregular acters weighted equally. They thus arrived at system of galleries with constructed parti- a classification by grouping species according tions. Because of such an ecology, they have to the extent to which the characters are escaped competition with higher termites. correlated. They claim that the resulting The Kalotermitidae have intestinal flagellates classification is closely similar to the one pre- but do not have any staphylinid beetles as viously established by classical methods. guests. Inger (1958) attacked the quantitative The family Kalotermitidae consists of a methodology of Michener and Sokal and number of living genera, which have existed strongly objected to their practice of giving presumably since Cretaceous times or earlier, equal weight to all characters. He states that even though they have given rise to phylo- their method has no advantage over the more genetic series of advanced genera. Also, the traditional taxonomic approaches, as their transitional forms in the phylogenetic se- technique also rests on a subjective base. quence have persisted to a remarkable degree Emerson (in Schmidt and Emerson, 1960) to the present time, so that we have "living states: "The tendency to give equal impor- fossils" or connecting links between many tance to all comparative characters of organ- genera. Because of the presence of such isms often leads to bias and to errors of judg- transitional forms, it is possible to theorize ment. Because the order of taxonomic classi- with considerable assurance about the deriva- fication is based upon homology through tion of a particular living genus from another descent, principles of evolution are both living genus. For these reasons, the family is substantiated by taxonomic data, and in exceptionally satisfactory for the analysis of turn, may be used for taxonomic interpreta- evolutionary history. tion. For example, it is possible to distinguish The objectives of this study are: (1) the homology from analogy and both from fortui- regrouping of the formerly classified species tous resemblance, strongly adaptive from into several genera consistent with the mod- weakly adaptive or nonadaptive characters, ern concept of the genus unfolded by the primitive from derivative characters, rapidly studies of the imago mandibles and their asso- evolving from slowly evolving characters, ciated structural patterns in all castes and persistent conservative characters from labile redescription and illustration of the various changing characters, progressively functional new and old genera; (2) the formulation of a characters from regressed vestiges, relatively hypothetical phylogeny; and (3) a synthesis complex from relatively simple characters, of taxonomic, phylogenetic, and biogeograph- and individually acquired characters from ical data, in an attempt to elucidate the hereditary characters." All characters are not evolutionary history of the group. equivalent in their genetic control and in their The basic method of taxonomy is to com- development and ecological effects, and there- pare and evaluate the characteristics of the fore cannot be treated equally (Inger, 1958). structure by morphological comparisons and A competent taxonomist after long training interpret them in the light of comparative and experience can determine which charac- genetics, biochemistry, embryology, behav- ters are constant, which are variable, and ior, ecology, and geography. which are more important for generic classifi- Michener and Sokal (1957) have criticized cation. That character weighting has been 309 310 BULLETIN AMERICAN MUSEUM OF NATURAL HISTORY VOL. 122 used injudiciously by poor taxonomists is not different castes with the same basic genetic sufficient reason for rejecting the method. system corrects the errors due to the confu- Stroud (1953) used the technique of multi- sion of convergence with homology and pro- ple factor analysis in the evaluation of char- vides excellent data for generic classification acters, the elucidation of evolutionary proc- and phylogenetic analysis. In the present esses by detection of adaptive trends, the study, I have conformed to the traditional detection of phylogenies, and the definition of approach of character weighting in defining categories. By this quantitative technique he categories. I have studied a number of analyzed correlations of 14 characters for morphological characters-wing venation, soldiers of 48 species and imagoes of 43 species imago-nymph mandible, arolium, eye size, of the genus Kalotermes. The application of pigmentation, pilosity, tibial spines, and multiple factor analysis as used by Stroud in other characters in the imago, and head establishing categories has limitations, be- shape, size, and pronotum in the soldier- cause in his study of the genus Kalotermes he and have defined genera based on a constella- selected characters in such a way that homol- tion of conservative, adaptive, and regressed ogous dimensions of soldier and imago characters. anatomy could be measured. He limited him- The imago mandible is a conservative self to only 14 characters found in both character which has been found valuable in imagoes and soldiers and thus omitted impor- generic classification. In a number of super- tant characters such as wing venation, imago- generic groups within the family Kalotermiti- nymph mandible, arolium, anterior margin of dae, the genera show no differences in the soldier pronotum, and tibial spines. The imago mandible. Yet it has been possible to statistical approach used by Stroud failed to recognize genera by other sets of characters, correct the errors resulting from earlier in- both in the imago and the soldier caste. An vestigations. The phylogenetic analysis used animal can retain primitive characters and at in the present paper conforms more closely to the same time have obvious derivative modi- classical methods than did that of Stroud fications. Different portions of the same (1953). The method of multiple factor analy- individual may show conservative characters sis used by Stroud when applied to the with little evolutionary change and also have categories distinguished by the method used more rapidly evolving adaptive specializa- in this paper may indicate more subtle rela- tions. tionships and correlations of characters here ACKNOWLEDGMENTS treated as independent units. (Also see under I am greatly indebted to Dr. Alfred E. the genus Kalotermes.) Emerson for suggesting the problem and Inger (1958) suggests a biological approach taking keen interest in the progress of the to the delimitation of genera. He outlines a work. My wife, Valerie Krishna, has pro-