HARVARD UNIVERSITY Library of the Museum of Comparative Zoology LIBRARY DEC 2 01982 nARVAKD UNIV^RJ^ITY VOLUME 23 1981-1982 TULANE UNIVERSITY NEW ORLEANS TULANE STUDIES IN ZOOLOGY AND BOTANY, a publication of the Biology Department of Tulane University, is devoted primarily to the biology of the waters and adjacent land areas of the Gulf of Mexico and the Caribbean Sea, but manuscripts on areas outside this geographic area will be considered. Each number contains an indivi- dual monographic study or several minor studies. Normally two numbers plus an index and a table of contents are issued annually. Preferred citation of the journal is Tulane Stud. Zool. and Bot. INFORMATION FOR AUTHORS: Manuscripts submitted for publications are eval- uated by the editors and by an editorial committee selected for each paper. Contrib- utors need not be members of the Tulane faculty. Manuscripts of 20 or more pages, double-spaced, are preferred. We recommend conformance with the principles stated in CBE Style Manual, 4th ed., 1978. Manuscripts should be typewritten and double spaced. Two additional copies should accompany the original to expedite editing and publication. Legends for figures should appear on a separate page and in sequence. Illustrations should be proportioned for one or two column width corresponding to our printed page size, and should allow for insertion of the legend if occupying a whole page. Guidelines for letter and other extraneous markings should be done with a non-photo blue pencil such as Eagle Prismacolor. Photographs should be on glossy paper. Many tables, if carefully prepared with a carbon ribbon and electric typewriter, can be photographically reproduced, thus helping to reduce publication costs. Lettering in any illustrative or tabular material should be of such a size that no letter will be less than 1 Vi mm high when reduced for publication. An abstract not exceeding three percent of the length of the article must accompany the manuscript. Separates of published articles are available to authors at a nominal cost. Page charges, calculated at $45/page, are solicited from authors who have funds for this purpose through their institutions or grants. Acceptance of papers is not dependent on ability to underwrite costs but excessive illustrations and tabular matter may be charged to the author. EXCHANGES, SUBSCRIPTIONS, ORDERS FOR INDIVIDUAL COPIES: Ex- changes are invited from institutions publishing comparable series. Subscriptions are billed in advance. A price list of back issues is available on request. Individuals should send th^ir remittance, preferably money order, along with their orders. Remittances should be made payable to "Tulane University." Subscription rates: Volume 23. $8.50 domestic, $9.50 foreign. Copies of Tulane Studies in Zoology and Botany sent to regular recipients, if lost in the mails, will be replaced if the editorial offices are notified before the second subsequent issue is released. COMMUNICATIONS: Address all queries and orders to: Editor, TSZ&B, Depart- ment of Biology, Tulane University, New Orleans, Louisiana 701 18, U.S.A. Harold A. Dundee, Editor CONTENTS OF VOLUME 23 Number Page 1. BIOSYSTEMATICS OF THE KINOSTERNON HIRTIPES SPECIES GROUP (TESTUDINES: KINOSTERNIDAE) JohnB. Iverson 1 LIFE HISTORY OF ETHEOSTOMA COOSAE (PISCES: PERCIDAE) IN BARBAREE CREEK, ALABAMA Patrick E.O'Neil 75 THE TAXONOMIC RELATIONSHIP BETWEEN MALACLEMYS GRAY, 1844 AND GRAPTEMYS AGASSIZ, 1857 (TESTUDINES: EMYDIDAE) James L.Dobie 85 2. CHANGES IN MELANIN MIGRATION INDUCED BY NORADREN- ERGIC AND HISTAMINERGIC AGENTS IN THE FIDDLER CRAB, UCA PUGILATOR Mukund M. Hanuamante and Milton Fingerman 103 ADDITIONAL TREMATODES OF MAMMALS IN LOUISIANA WITH A COMPILATION OF ALL TREMATODES REPORTED FROM WILD AND DOMESTIC ANIMALS IN THE STATE Wesley L. Shoop and Kenneth C. Corkum 109 COMPARATIVE VISCERAL TOPOGRAPHY OF THE NEW WORLD SNAKE TRIBE THAMNOPHIINI (COLUBRIDAE, NATRI- CINAE) Nita J. Rossman, Douglas A. Rossman, and Nancy K. Keith 123 TULANE STUDIES IN ZOOLOGY AND BOTANY VOLUME 23 INDEX TO SCIENTIFIC NAMES (NEW TAXONOMIC ENTITIES IN BOLDFACE) JOL Ablates baliodeira, 129, 137 Acrochordidae, 129, 163 Acrochordis, 137 arafurae, 137, 163 granulatus, 129, 137, 163 javanicus, 129, 137, 163 A de lop his, 124 Agamodistomum marcianae, 112 Agkistrodon piscivorus, 129 A haetulla prasinus, 1 37 Alaria alahodes, 110-112, 115, 119, 120 americana, 112-113 canis, 112 da th rata, 112 marcianae. 111, 113, .115, 119-120 minnesotae, 112 mustelae, 110-111, 119 pseudoclathrata, 112 Ambloplites rupestris, 11 Amnicola, 118, 120 Integra, 118 longiqua, 114 Amphiesma vibakari, 164 Amphimerus, 114 caudalitestis, 113-114 interrupt us, 113-114 minimus, 113-114 /jeo tropicalis, 113-114 parciovatus, 113-114 /?r/ce/, 113-114 pseudofelinus, 113-114, 120 speciosus. 111, 114-115, 119-120 Ancistrodon rhodostoma, 137 Aniliidae, 129, 137 Aromochelys carinatus, 41 Baschkirovitrema incrassatum, 1 10-1 11, 117, 119-120 Boidae, 129, 137, 163 Boiga, 163 Bolyeria, 163 Brachylaima, virginiana, 110-111, 119 Bungarus candidus, 137 fasciatus, 137 Calamaria multipunctata, 129, 137 Calloselasma rhodostoma, 137 Cambarellus puer, 118 Campostoma anomalum, 11 Carneophallus basodactylophallus, 110-111, 119 Catonolus, 78 f label lore, 76 DEC 20' '^ species, 78 Catostomus commersoni, 116 Causus rhombeatus, 163-164 Cerberus rhynchops, 137 Cercaria marcianae, 1 1 2 Chamaetortus, 163 Chinosternum hirtipes, 46 Chrysemys, 88-90, 92-95, 97-99 insculpta, 95 p/cro, 8, 94-95, 97-98, 100 Cinosternon henrici, 41, 61 hippocrepis, 22 /7/W//7e5, 18, 23, 41, 44-45, 48, 65 pensylvanicum, 21-22, 45-46, 48-49 species, 46, 48 Cinosternum flavescens, 41, 61 integrum, 15, 46, 49, 51 sonoriense, 41, 49 Claudius angustatus, 23 Clemmys, 88-89, 92-95 gwr/a?fl, 90, 95, 98, 100 insculpta, 90, 95, 98, 100 marmorata, 72, 100 muhlenbergi, 100 Clonophis, 126-127, 134-135, 137, 139, 142, 144-146, 155, 157 kirtlandi, 125, 128-132, 134-136, 145-156, 159 Codoma ornatus, 1 Coluber melanurus, 129, 137 oxycephalus, 163-164 radiatus, 137 Colubridae, 129, 137, 163-164 Compsemys, 97 Corallus, 163 Cottus carolinae, 11 Crotaphopeltis hotamboeia, 163 Cryptocotyle, 116 concava. 111, 114-116, 119-120 echinata, 114 Cylindrophis rufus, 129, 137 Deirochelys, 87-90, 92-95 com. 98 reticularia, 98, 100 Dendrelaphis pictus, 129, 137 Dendrophis pictus, 137 Didelphis virginiana, 109, 118 Diplostomidae, 110 Diplostomum, 112 alaroides, 110, 112 fosteri, 1 1 2 Dipsadoboa, 163 Distoma concava, 114 Dryophis praslnus, 1 37 Echinochasmus schwartzi, 119 Echinocirrus metis, 116 Echinostoma revolutum, 117 Echinostomatidae, 116 Echmatemys, 98-99 pusilla, 98 Elaphe flavilineata, 129, 137 radio t us, 137 Elapidae, 129, 137 Elapoides fuscus, 137, 164 Emydoidea, 88-90, 92-95, 98 bland ingi, 98, 100 Enhydrina schistosa, 137 Enhydris alternans, 129, 137 enhydris, 137, 164 plumbea, 129, 137, 164 Enhydrodiploslomum, 1 12 alarioides, 110 Etheostoma, 76, 82 acuticeps, 82 barbouri, 81 blennioides, 82 coosae, 75-81 rfi/ry/, 75-76, 79 flabellare, 76 fonticola, 82 gracile, 82 jordani, 11 kennicotti, 81 nigrum, 82 proeliare, 81 radiosum cyanorum, 82 simoterum, 75, 79 species, 79 squamiceps, 78, 80 stigmaeum, 11 Eunectes, 163 Euparyph 'um, 116-117, 1 20 beaveri, 117, 120 A«e//5, 116-117 Exiliboa placata, 164 Easciola putori, 1 16 trigonocephala, 1 1 6 Felis domesticus, 109, 113 Fibricola cratera, 110-111, 119 /wc/rffl, 110-111, 119 Fordonia leucobalia, 129, 137 Fundulus stellifer, 11 Gasterosteus aculeatus, 1 1 6 Gongylosoma baliodeira, 129, 137 Gonysoma oxycephalus, 163-164 Graptemys, 85-88, 90-102 barbouri, 86, 91, 96, 98, 100 cag/e/, 86, 91, 98, 100 cordifera, 97 ftavimaculata, 86-87, 91, 98, 100 geographica, 89-91, 93, 98, 100 inornata, 97 nigrinoda, 86, 90-92, 96, 98, 100 oculifera, 86, 91, 98, 100 ouachitensis, 98, 100 ouachitensis ouachitensis, 91, 100 ouachitensis sabinensis, 86, 91, 100 pseudogeographica, 86, 90-94, 98, 100 pseudogeographica kohni, 101 psdudogeographica pseudogeographica, 100 pulchra, 86-87, 89, 91-94, 98, 100-101 ver^a, 86, 91, 98, 100, 101 Gyrauiis parvus, 119 Gyrosoma singulare, 110-111, 119 Hasstilesia texensis, 110-111, 119-120 Heterobilharzia americana, 110-111, 1 1 9- 1 20 Heterophyidae, 114 Homolopsinae, 129, 137, 164 Homalopsis buccata, 129, 137 Hydrophis fasciatus, 129, 137 Hypentelium etowanum, 11 Hypsirhina alternans, 129, 137 plumbea, 137, 164 Ictalurus natalis, 11 species, 8 Isonychia, 75 Isthmiophora, 116, 120 mefc. 111, 115-116, 119-120 Kinosternidae, 1,9, 11, 13 Kinosternon, 13, 15, 19, 21-22, 29-30, 35, 41, 43, 46, 55 acutum, 23, 35 alamosae, 1, 2, 7, 21, 23, 41 angustipons, 35 /70«/-/, 35, 55 cobanum, 23 dunni, 35 flavescens, 2, 4, 6, 7-11, 22, 40-41, 44, 46, 50, 55, 60 henrici, 40, 41, 61 herrerai, 22, 55 hertipes, 46, 50 hirtipes, 1-4, 6, 8-24, 26-31, 34-35, 38-41, 44, 46-56, 60, 64 hirtipes chapalense, subsp. novum, 46, 51, 54, 57, 65 hirtipes chapalense X K.h. murrayi, 65 /?/W//7e5 group, 1-2, 4-7, 11, 15, 20, 22, 24-28, 33-38, 42, 46, 54-55 hirtipes hirtipes, 37, 46, 48-52, 59, 63, 65 hirtipes magdalense, subsp. novum, 6, 33, 46, 53-54, 58, 64 hirtipes megacephalum, subsp. novum, 6, 33, 37, 46, 52, 54, 58, 64 hirtipes murrayi, 6, 21. 33, 34, 37, 39, 46, 48-52, 59,62 hirtipes tarascense, subsp. novum, 33-34, 37, 46, 52, 54, 58, 64 integrum, 1-5, 7, 12-18, 21-23, 40-41, 44-45, 48, 55, 64-65 leucostomum, 21, 23 leucostomum group, 35 murrayi, 21, 46, 50 oaxacae, 1, 2 oblongum, 44 pennsilvanicum, 11, 45 punctatum, 41 scorpiodes, 1, 2, 8, 20-21 scorpiodes group, 1, 2, 5, 22-23, 29, 35 sonorensis, 41-43, 46, 55, 61 sonoriense, 1-8, 12-14, 19-20, 22-31, 33-35, 37-38, 40-46, 54-55, 60, 63 sonoriense longifemorale, subsp. novum, 6, 42-44, 54, 56, 62 sonoriense, sonoriense, 56, 60 species, 48-49 steindachneri, 55 subrubrum, 1\-11, 31, 35, 41, 55 subrubrum group, 55 subrubrum hippocrepis, 11 subrubrum steindachneri, 31, 55 subrubrum subrubrum, 35 subrubrum triliratum, 23 Lepomis cyanellus, 11 gulosus, 11 macrochirus, 11 megalotis, 11 Leptodira hotamboeia, 163 Lichanura, 163 L ins to wiella szidati, 110-111, 1 1 9- 1 20 Loxocemus, 163 Lutra canadensis, 109-110 Lynxrufus, 109, 112 Malaclemys, 85-88, 90, 92-100 terrapin, 86, 91-93, 95-98, 100 terrapin centrata, 87 terrapin littoralis, 87 terrapin macrospilota, 87 terrapin pileata, 87 terrapin terrapin, 70 Maritreminoides nettae, 110-111, 1 1 9- 1 20 Microphallidae, 117 Microphallus opacus. Ill, 115, 1 17-120 ovatus, 117 Micropterus coosae, 11 punctulatus, 11 Moxostoma duquesnei, 11 Mustek vison, 109-110, 114, 117 Naja tripudians, 137 Natricinae, 124, 129, 137, 164 Matrix chrysarga, 129, 137 erythrogaster alta, 1 1 subminiata, 129, 137 trianguligera, 129, 137, 164 vibakari, 164 vittata, 129, 137 Nerodia, 126-127, 134-137, 139, 141, 144-147, 149-152, 155, 157 cyclopion, 125, 128, 130-132, 135-136, 138, 140-143, 145-156, 159 erythrogaster, 125, 128-132, 134-136, 138, 140-143, 145-156, 159 fasciata, 125, 128, 130-132, 135-136, 138, 140-143, 145-156, 159 rhombifera, 125, 128-132, 135-136, 138, 140-143, 145-156, 159 sipedon, 125, 128, 130-132, 135-136, 138, 140-143, 145-156, 159 valida, 125, 128, 130-132, 134, 138, 140-143, 145-156, 159 Notocotylidae, 118 Notocotylus quinqueserialis, 118 urbanensis, 118 Notropis asperifrons, 11 callistius, 11 lirus, 11 ornatus, 1 stilbius, 11 trichroistius, 11 venustus, 11 xaeocephalus, 11 Nyctanassa violacea, 114 Nudacotyle novicia, 119 Ondatra zibethica, 109, 118 Opisthorchidae, 113 Ozotheca hirtipes, 45 odorata, 21, 45, 48 Paragonimidae, 118 Paragonimus kellicotti. 111, 1 18-120 Paramonostomum pseudalveatum, 119 P ere in a caprodes, 11 nigrofasciata, 77, 81 Phagicola angrense, 1 1 9 nana, 119 Pharyngostomoides procyonis, 110-111, 119 Phenacobius ca tost am us, 11 Philothalmus semivariegalus, 1 63- 1 64 Phopalias macracanthus, 119 Platylhyra flavescens, 22 Poecilia reticulatua, 82 Pomatiopsis lapidaria, 1 1 8 Procambarus clarki, 118 Procyon lotor, 109, 112-113, 116-117 Psamnophis sibilans, 163 Pseudemys, 85-90, 92-96, 98-99 alabamensis, 101 concinna, 87-88, 94, 101 floridana, 88, 94, 101 nelsoni, 101 rubiventris, 87-88, 101 5cr//?/a. 7, 10-11, 14, 70, 87-88, 101 scrip ta elegans, 18 stejnegeri, 101 P/v'a5 korros, 137 mucosa, 137 Quinqueserialis hassali, 118 quinqueserialis, 111, 115, 1 1 8- 1 20 /?eg/«fl, 126-127, 135, 137, 139, 141, 144, 146, 147 a//e/j/, 125, 128-132, 135-136, 138, 140, 145, 146-151, 153-156, 159 grahamii, 125, 128, 130-132, 134-136, 138, 140-143, 145-156, 160 rigida, 125, 128, 130-132, 134-136, 140-143, 145-156, 160 septemvittata, 125, 128, 130-132, 135-136, 138, 140-141, 145-156, 160 Rhabdophis chrysarga, 129, 137 subminiata, 129, 137 Rhinoclemmys, 88-90, 98-99 areola ta, 101 pulcherrima, 101 sp., 101 Rhopalias macracanthus, 110-111 Sellacotyle vitellosa, 110-111, 120 Seminatrix, 126-127, 134-135, 139, 141-142, 144, 155, 157 pygaea, 126, 128-132, 134-136, 145-156, 160 Semotilus atromaculatus, 11 Sinonatrix trianguligera, 129, 137, 164 Staurotypus triporcatus, 23 Sternotherus odoratus, 9, 21, 46, 50 Storeria, 126, 129, 134, 137, 139, 141, 144, 155, 157, 159 dekayi, 125, 127-128, 130-132, 134-136, 138, 140-143, 145-156, 161 occipitomaculata, 126-128, 130-132, 134-138, 140-143, 145-156, 160 Swanka henricii, 41 Sytvilagus aquaticus, 119 Terrapene, 87-90, 92-95 Carolina, 101 ornata, 101 Testude pensilvanica, 21-22 Thalassophis anomalus, 137 Thamnophiini, 124, 137, 164 Thamnophis, 126-127, 129, 134, 137, 139, 151-152, 154-155 angustirostris, 126 brachystoma, 126, 128, 130-132, 145-156, 160 butleri. 126, 128, 130-132, 145-156, 160 chrysocephalus, 130-132, 135-136, 145-156, 160 couchii, 125 couchii A, 128, 130-132, 135-136, 138, 140-143, 145-156, 160 couchii B, 128, 130-132, 135-136, 138, 140-143, 145-156, 160 couchii couchii, 126, 128, 138 couchii hydrophilus, 126, 128, 160 cyrtopsis, 126, 128, 130-132, 135-136, 138, 140-143, 145-156, 160 elegans, 125, 135 elegans gxouv, 126, 130, 132, 139, 141-144, 157-158 elegans A, 128, 130-132, 135-136, 138, 140-143, 145-156 elegans B, 128, 130-132, 135-136, 138, 140-143, 145-156 elegans terrestris, 126, 128, 160 elegans vagrans, 126, 128, 160 eques, 125-126, 128, 130-132, 134-136, 138, 140-143, 145-156 eques megalops, 125, 160 eques virgatenuis, 125, 160 godmani, 126, 128, 130-132, 135-136, 145-156, 160 marcianus. 126, 128, 130-132, 135-138, 140-143, 145-156, 160 megalops, 126 melanogaster, 126, 128-132, 134-138, 140-143, 145-156, 160 nigronuchalis, 126, 128-136, 138, 140-143, 145-156, 160 ordinoides, 126, 128, 130-132, 145-156, 160 proximus, 126, 128-132, 134-139, 140-143, 145-156, 160 radix, 126, 128, 130-132, 134-156, 161 radix gxonp, 126, 130, 132, 139, 141-144, 157-158 rufipunctatus, 126, 128-136, 145-156, 161 sauritus, 126, 128-132, 134-138, 141-143, 145-156, 161 sauritus gxoxx^p, 126, 130, 132, 139, 141-144, 153, 157-158 scalaris. 126, 128, 130-132, 135-137, 145-156, 161 sirtalis A, 128, 130-132, 135-136, 138, 140-143, 145-156 sirtalis B, 128, 130-132, 135-136, 138, 140-143, 145-156 sirtalis fitchi, 126, 128, 138, 161 sirtalis group, 126, 130, 132, 139, 141-144, 157-158 sirtalis sirtalis, 126, 128, 161 Tocotrema concava, 114 Thyrosternum henrici, 41 hirtipes, 45 sonoriense, 41 Trachemys, 88, 94-95 Trachyboa gularis, 163 Trimeresurus gramineus, 1 37 Trionyx spiniferus, 4, 10 Tropidoclonion, 126-129, 134, 137, 139, 141-142, 144, 155, 157 lineatum, 125-126, 134-136, 161 lineatum A, 130-131, 133, 145-151, 153-156 lineatum B, 130-131, 133, 135-136, 138, 140-143, 145-156 Tropidophiidae, 164 Tropidophis, 163 Uca pugilator, 103-106 Ulocentra, 76, 79 Viperidae, 163 Virginia, 126-127, 134, 137, 139, 141, 144, 155, 157 striatula, 126, 128, 130-131, 133-138, 140-143, 145-156, 161 valeriae, 126, 128, 130-131, 133-138, 140-143, 145-156, 161 valeriae elegans, 128 Vulpesfulva, 109 Xenochrophis vittata, 137 Xenopeltis, 163 unicolor, 137 Zamenis florulentus, 163 rhodorachis, 163 i(D®IL(D(g^ JAN 4 iQn9 Volume23 . Number 1 $5.50 K^ft}^,9n5?iCSo, 1981 UN/IVK.cjsiTV BIOSYSTEMATICS OF THE KINOSTERNON HIRTIPES SPECIES GROUP (TESTUDINES: KINOSTERNIDAE) JOHN B. IVERSON p. 1 LIFE HISTORY OF ETHEOSTOMA COOSAE (PISCES: PERCIDAE) IN BARBAREE CREEK, ALABAMA PATRICK E. O'NEIL p. 75 THE TAXONOMIC RELATIONSHIP BETWEEN MALACLEMYS GRAY, 1844 AND GRAPTEMYS AGASSIZ, 1857 (TESTUDINES: EMYDIDAE) JAMES L. DOBIE p. 85 TULANL UNIVERSITY NEW ORLEANS TULANE STUDIES IN ZOOLOGY AND BOTANY, a publication of the Biology Department of Tulane University, is devoted primarily to the biology of the waters and adjacent land areas of the Gulf of Mexico and the Caribbean Sea, but manuscripts on areas outside this geographic area will be considered. Each number contains an indivi- dual monographic study or several minor studies. Normally two numbers plus an index and a table of contents are issued annually. Preferred citation of the journal is Tulane Stud. Zool. and Bot. INFORMATION FOR AUTHORS: Manuscripts submitted for publications are eval- uated by the editors and by an editorial committee selected for each paper. Contrib- utors need not be members of the Tulane faculty. Manuscripts of 20 or more pages, double-spaced, are preferred. We recommend conformance with the principles stated in CBE Style Manual, 4th ed., 1978. Manuscripts should be typewritten and double spaced. Two additional copies should accompany the original to expedite editing and publication. Legends for figures should appear on a separate page and in sequence. Illustrations should be proportioned for one or two column width corresponding to our printed page size, and should allow for insertion of the legend if occupying a whole page. Guidelines for letter and other extraneous markings should be done with a non-photo blue pencil such as Eagle Prismacolor. Photographs should be on glossy paper. Many tables, if carefully prepared with a carbon ribbon and electric typewriter, can be photographically reproduced, thus helping to reduce publication costs. Lettering in any illustrative or tabular material should be of such a size that no letter will be less than 1 Vi mm high when reduced for publication. An abstract not exceeding three percent of the length of the article must accompany the manuscript. Separates of published articles are available to authors at a nominal cost. Page charges, calculated at $45/page, are solicited from authors who have funds for this purpose through their institutions or grants. Acceptance of papers is not dependent on ability to underwrite costs but excessive illustrations and tabular matter may be charged to the author. EXCHANGES, SUBSCRIPTIONS, ORDERS FOR INDIVIDUAL COPIES: Ex- changes are invited from institutions publishing comparable series. Subscriptions are billed in advance. A price list of back issues is available on request. Individuals should send their remittance, preferably money order, along with their orders. Remittances should be made payable to "Tulane University." Subscription rates: Volume 23. $8.50 domestic, $9.50 foreign. Copies of Tulane Studies in Zoology and Botany sent to regular recipients, if lost in the mails, will be replaced if the editorial offices are notified before the second subsequent issue is released. COMMUNICATIONS: Address all queries and orders to: Editor, TSZ&B, Depart- ment of Biology, Tulane University, New Orleans, Louisiana 701 18, U.S.A. Harold A. Dundee, Editor TULANE STUDIES IN ZOOLOGY AND BOTANY Volume 23, Number 1 December 30, 1981 BIOSYSTEMATICS OF THE KINOSTERNON HIRTIPES SPECIES GROUP (TESTUDINES: KINOSTERNIDAE) JOHN B. IVERSON' Dept. of Biology Earl ham College Richmond, Indiana 47374 Abstract Geographic variation in scute and shell measure- ments (via muhivariate statistical analysis), body size, head scale and chin barbel morphology, size of first neural bone, shell carination, and head size and pat- terns in populations of the Kinostemon hiritpes species group were analyzed. The results support the retention of allopatric K. sonoriense and K. hirtipes as full species in the group, and the recognition of two eillopatric sub- species (one new) of K. sonoriense and six subspecies (four new and all apparently allopatric) of K. hirtipes. The description of each taxon includes complete synon- omies and ecological and reproductive data. Also in- cluded are a key to adults and a discussion of all taxa. Introduction Prior to 1970, members of the Kinoster- non hiriipes species group were cited more than 233 times in the literature. At least half of those citations contained errors in identi- fication, locality, and/or orthography. Iver- son (1976, 1978), Conant and Berry (1978), Iverson and Berry (1979), and Berry and Legler (1980) have each addressed some of the problems dealing with members of this group in the American southwest, adjacent northwestern Mexico, and northeastern Mexico. Clearly the distribution, identifica- tion, systematics, and phylogeny of the tur- 'Adjunct Assistant Curator of Herpetoiog)', Florida State Museum, University of Florida, Gainesville, FL 32611 ties of the Kinosternon hirtipes species group are poorly understood. The purpose of this report, as part of a continuing analysis of relationships within the family Kinosternidae, is to rectify this situation. My objectives here are 1) to redefine the members of this group taxonomically, 2) to analyze patterns of geographical variation in external morphological characters, 3) to develop a phylogeny of these members, and 4) to correct and bring order to the confus- ing and erroneous literature. Identification Of The Kinosternon Hirtipes Species Group One of the primary obstacles to the study of Mexican kinosternids has been the diffi- culty in distinguishing members of the K. hirtipes species group (K. hirtipes and K. sonoriense) from those of the A', scorpioides group (fide Berry 1978; including K. scor- pioides, K. alamosae, K. oaxacaedSid^K. in- tegrum), especially where the groups occur sympatrically. Adult males of the hirtipes group are readily distinguished by the pres- ence of a patch of elevated scales on the posterior crus and thigh of each hindleg (vinculae: fide H. M. Smith and R. B. Smith 1980), absent in turtles of the scor- pioides group, but adult females of the hir- tipes group lack these structures and are thus often difficult to identify. An elabora- tion of the differences between K. integrum EDITORIAL COMMITTEE FOR THIS PAPER: Dr. James F. Berry, Assistant Professor of Biology, Elmhurst College, Elmhurst, Illinois 60126 Dr. Robert G. Webb, Professor of Biological Sciences, University of Texas at El Paso, El Paso, Texas 79999 Tulane Studies in Zoology and Botany Vol. 23 and members of the K. hirtipes group is therefore justified (see also Iverson and Berry 1979), especially since they coexist in several Mexican drainage basins (see MA- TERIALS AND METHODS). Kinosternon scorpioides is not sympatric with members of the K. hirtipes group. Identification Of The Kinosternon Hirtipes Species Group One of the primeuy obstacles to the study of Mexican kinostemids has been the diffi- culty in distinguishing members of the K. hirtipes species group (K. hirtipes and K. sonoriense) from those of the K. scorpioides group (fide Berry 1978; including K. scor- pioides, K. alamosae, K. oaxacae and K. in- tegrum), especially where the groups occur sympatrically. Adult males of the hirtipes group are readily distinguished by the pres- ence of a patch of elevated scales on the posterior crus and thigh of each hindleg (vinculae: fide H. M. Smith and R. B. Smith 1980), absent in turtles of the scor- pioides group, but adult females of the hir- tipes group lack these structures and are thus often difficult to identify. An elabora- tion of the differences between K. integrum and members of the K. hirtipes group is therefore justified (see also Iverson and Berry 1979), especially since they coexist in several Mexican drainage basins (see MA- TERIALS AND METHODS). Kinosternon scorpioides is not sympatric with members of the K. hirtipes group. The primary criteria for distinguishing adults of K. integrum and the K. hirtipes group appear in Table 1. Juveniles are much more difficult to distinguish and remain poorly studied. In general, small specimens of the K hirtipes group have smaller plastra, narrower bridges, and more axillary-inguinal scute contact than K. inte- grum (Fig. 1). More precise discrimination of small turtles must await additional data. Members of the K. hirtipes group also differ from K. integrum ecologically. The former are virtually restricted to permanent water habitats, rarely leaving the water except to nest; migrating behavior is unre- ported. Kinosternon integrum is an excel- lent colonizing species (fide MacArthur and Wilson, 1967). It is extremely vagile, mi- grates considerable distances during the rainy season, and may aestivate under ground as K. fiavescens and K. alamosae do. The number of specimens and locality records for K. integrum in museum collec- tions (see lists in H. M. Smith and R. B. Smith, 1980) reflects the more frequent oc- currence of A", integrum than K. hirtipes on roads. Thus, K. integrum may be found in almost any temporary pond or roadside pool, habitats where K. hirtipes would al- most never occur. In addition, although their thermoregula- tory behavior has not been studied in detail, I suspect thermal preference and tolerance levels are higher in K. integrum than in the K. hirtipes group. This is reflected in the very different basking habits of the two forms. I have observed K. integrum basking at many Mexican localities in Michoacdn, Jalisco, Sinaloa, and Oaxaca, but K. hir- tipes basking only once at 2400 m elevation in Durango, and once (adult females only) at 1800 m in Jalisco. This perhaps reflects their coastal lowland (integrum) versus high plateau (hirtipes) origins. The absence of A'. integrum in the highest (i.e. coldest) basins of the southern Mexico Plateau (Pa'tzcuaro, San Juanico, and Villa Victoria; see later) may be related to thermal requirements rather than to historical zoogeography. Likewise, despite its vagility, K. integrum ranges no farther northward in Sonora than 29 °N latitude in the Rio Yaqui basin. Per- haps its range there is also limited by tem- perature regime. Further study may show other behavioral and ecological differences between these two groups on the Mexican Plateau. Materials And Methods Specimens and field work I have examined nearly all specimens of the species of the Kinosternon hirtipes group in United States museums. In addi- tion, most of the world's museums were canvassed for locality data of other speci- mens. All available type specimens were examined. Each locahty was pinpointed (and its elevation determined) on the No. 1 Kinosternon Biosystematics Figure 1. Plastral comparison of juvenile Kinoster- non hirtipes (left; UMBM 2403) and A:. integrum (UMBM 2411), both from 3.2 km SE Ocotlan, Jalisco, Mexico. 1:500,000 sheets of "La carta general de la Repiiblica Mexicana" (published by the Ex-Comision Intersecretarial de Mexico, D. F., 1958), with the help of the "Offi- cial Standard (Geographic) Names of Mexico", published by the Office of Ge- ography of the U.S. Dept. of Interior (1956). These localities were then mapped on Miller's (1968) drainage map of Mex- ico (Figs. 3 and 4). Field trips to sample critical areas for K. hirtipes were taken in May 1977 (12 days; 10 localities in Durango and Chihuahua), June 1978 (11 days; 12 localities in San Luis Potosi, Guanajuato, Michoacan, and JaUsco), June 1979 (part of 14 day trip; 5 localities in Mexico State), July 1980 (12 days; 7 localities in Coahuila, Chihuahua, and Durango), and May 1981 (8 days; 10 localities in Jalisco, Mdxico State, Michoacan, and Puebla). Field work with K. sonoriense in Arizona was also undertaken in January 1971 (2 days), May 1974 (4 days), January 1976 (4 days), and July 1980 (1 day), and in Chi- huahua in August 1980 (1 day). All known specimens and localities for members of the K. hirtipes species group are in the SPECIMEN LIST and plotted in Figures 3 and 4; museum acronyms fol- low Duellman, Fritts, and Leviton (1978) except for the following: CAS-SU DMNH EAL ENMU FB FWMNH JBI JFB LTU MES MSU MU NMSU RSF SENCK SM SRSU TAI UAZ UF UG UMKC UNSM UOK USA USL California Academy of Sciences - Stanford University Collections Dallas Museum of Natural History Ernest A. Liner, Houma, Louisiana Eastern New Mexico University Thomas R. VanDevender, Tucson, Arizona Fort Worth Museum of Natural History John B. Iverson, Richmond, Indiana James F. Berry, Elmhurst, Illinois Louisiana Tech University Michael E. Seidel, Huntington, West Virginia Michigan State University Midwestern University, Wichita Falls, Texas New Mexico State University R. S. Funk, Normal, Illinois Senckenberg Museum Strecker Museum, Baylor University Sul Ross State University Texas A&I University University of Arizona University of Florida, Florida State Museum University of Georgia University of Missouri, Kansas City University of Nebraska State Museum University of Oklahoma University of South Alabama University of Southwestern Louisiana Tulane Studies in Zoology and Botany Vol. 23 This study is based on the examination of at least 1298 museum specimens of the Kinosternon hirtipes species group, as well as other specimens collected and re- leased in the field. Population samples of turtles correspond to the inhabited drain- age basins, which are listed and described below from approximately north to south. The reader is referred to Blasques L. (1959) and Tamayo (1962, 1964) for more general descriptions of the geogra- phy and hydrography of the drainage basins in Mexico. Bill Williams River basin, Arizona (WILL). — The Bill Williams River and its major tributaries, the Big Sandy, Bur- ro, and Santa Maria rivers drain a small area in west central Arizona and empty into the Colorado River at Parker Dam, about 90 km below the Nevada border. Kinosternon sonoriense is the only fresh- water turtle known from this basin (four localities; 800-1200 m). Stebbins' (1966) K. flavescens records from this basin were based on K. sonoriense (Iverson, 1978: 477). Gila and Lower Colorado River basins, Arizona, California, New Mexico, and Sonora (GILA). — Most of Arizona south of the Mogollon Rim and a portion of west central New Mexico are drained by this system. Miller (1961), Ohmart, et al. (1975), and McNatt (1978) described changes in the aquatic habitats along the Colorado and Gila rivers and their tribu- taries over the past 100 years. Kinoster- non sonoriense occurs throughout the basin (Iverson, 1976, 1978) and reaches its maximum known elevation (2042 m) in the Gila River in western New Mexico (Niles, 1962; Degenhardt and Christian- sen, 1974), and its lowest known elevation (ca. 43 m) near Yuma, Arizona. The range of K. sonoriense in the Colo- rado River is poorly known. It apparently once occurred there at least upstream to southeastern Nevada (LaRivers, 1942, as K. flavescens; see Iverson 1978:476). More field work is needed along the Colo- rado River between Needles, California, and Yuma, Arizona, to establish the pre- sent range of K. sonoriense. The only other freshwater turtle which may occur naturally in this basin is K. flavescens, but I have elsewhere (Iverson, 1978) ques- tioned its recent occurrence in the Gila. The introduced Trionyx spinferus does, however, also occur in the Gila and Colo- rado rivers from southwestern Utah and western New Mexico to the mouth of the Colorado River (Webb, 1973). Southwestern New Mexico interior drainages (SWNM). — K. sonoriense occurs in the permanent water basins of the eastern and western slopes of the Pe- loncillo Mountains of Hidalgo Co. in southwestern New Mexico and adjacent Arizona from 1150 m to 1700 m (Niles, 1962, and Degenhardt and Christiansen, 1974, briefly discussed turtle habitats in f^^\ Figure 2. Comparison of papillae on the tail of female Kinosternon hirtipes (above; FMNH 71029; 137 mm CL; Guanajuato, Taboado) and A', integrum (FMNH 71031; 126 mm CL; same locality). No. 1 Kinosternon Biosystematics Table 1 . Character Comparison of adults of the Kinosternon hirtipes species group and Kinosternon integrum (K. scorpioides species group). Kinosternon hirtipes group Kinosternon integrum Elevated scale patches on hind legs of males Head shield in adults geographically variable, from reduced crescent-shape to large, V-shape, or to even larger tri- angle or bell shape (latter characteris- tic only of A", sonoriense and Valley of Mexico K. hirtipes); posterior margin of shield often concave. Axillary and inguinal scutes nearly al- ways in broad contact. Plastron usually yellow or greenish yel- low, sometimes darkly stained. If carination present on posterior cara- pace, then only one medial keel usual- ly evident. Skin very papillose; tail with numerous rows of large papillae (Fig. 2). Maximum carapace length, 185 mm; maximum plastron length, 160 mm. First vertebral scute width averages 24.5 (range 20 to 3207o) and 25.1% (20 to 30%) of carapace length in male A". hirtipes and K. sonoriense, respective- ly; 24.7 (18 to 31%) and 26.1% (20 to 32%) in females, respectively. Bridge length averages 20.1 (range 16 to 24%) and 21.4% (18 to 25%) of cara- pace length in male K. hirtipes and K. sonoriense, respectively; 23.6 (18 to 29%) and 24.8% (22 to 28%) in fe- males, respectively. Bridge length averages 82.0% (range 61 to 120%) and 85.3% (62 to 115%) of first vertebral scute width in male K. hirtipes and K. sonoriense, respective- ly; 95.5% (64 to 133%) and 95.0% (70 to 123%) in females, respectively. Maximum posterior width of plastral forelobe averages 43% (range 36 to 51%) and 47% (range 42 to 53%) of carapace length in male K. hirtipes and K. sonoriense, respectively; 48% (42 to 54%) and 49% (44 to 54%) in females, respectively. Nuchal and first neural bones occasion- ally (41% in K. sonoriense; 10% in K. hirtipes) in contact. No elevated scale patches on hind legs of males Adult head shield large, triangular or bell shaped, with posterior margin convex; shield not reduced or furcate behind Axillary and inguinal scutes usually not in contact; if touching, contact is nar- row. Plastron usually yellow-orange, almost never darkly stained. If carination present on posterior cara- pace, then three keels usually evident. Skin hardly papillose; tail with few, very reduced papillae (Fig. 2). Maximum carapace length, 210 -(- mm; maximum plastron length, 195-1- mm. First vertebral scute width averages 21.3% (range 17 to 26%) of carapace length in males; 22.5% (19 to 28% in females). Bridge length averages 26.1% (range 20 to 28%) of carapace length in males; 26.8% (20 to 30%) in females (ex- cludes coastal Jalisco specimens). Bridge length averages 114% (range 88 to 151%) of first vertebral scute width in males; 115% (91 to 158%) in fe- males (excludes coastal Jalisco speci- mens). Maximum posterior width of plastral forelobe averages 47% (range 42 to 54%) of carapace length in males; 53% (45 to 57%) in females (excludes coastal Jalisco specimens). Nuchal and first neural bones not in contact. Tulane Studies in Zoology and Botany Vol. 23 this area). Huntington (1914:70) reviewed the historical isolation of the Animas Valley, which lies to the east of Peloncil- los and receives the drainages of the east- ern slopes of those mountains. Hubbs and Miller (1948) examined the geography of this and other independent drainage bas- ins in southwestern New Mexico. Rio Sonoyta (- Sonoita) basin, Arizo- na and Sonora (SNTA). — The Rfo Sonoyta lies along the northwestern boundary of the state of Sonora, Mexico. The river disappears in the desert sands near the eastern border of the Pinacate lava flows. The physical geography of the basin is reviewed by Ives (1936). K. sonoriense is found in the more perma- nent portions of the basin between about 350 and 450 m, near the U.S. border. It is abundant at Quitobaquito Pond in Organ Figure 3. Distribution of the subspecies of Kino- sternon sonoriense. Dots indicate actual records; hatching, suggested total ranges. The range of K. sonoriense longifemorale is marked (A); the re- maining hatched area marks the range of the nominate subspecies. Question mark in Nevada is discussed in Iverson (1978); that in southeastern Chihuahua, in the present text. Stippled area il- lustrates portion of the allopatric range of Kinosternon hirtipes (see Fig. 4). Pipe Cactus National Monument, Pima Co., Arizona (Hulse, 1974; Iverson, field notes). The aquatic habitat at Quitoba- quito was described by Cole and Whiteside (1965). Hubbs and Miller (1948:113) discussed the historical geography of the basin. Kinosternon flavescens is the only other native aquatic or semiaquatic turtle known from the Rio Sonoyta system (H. M. Smith and Hensley, 1957; Iverson, 1979), but Hulse (1974:94) reported that Chrysemys picta dorsalis has been intro- duced into Quitobaquito Spring. Figure 4. Distribution of the subspecies of Kinoster- non hirtipes. Dots indicate actual records; hatch- ings, suggested total ranges. Subspecies ranges are marked: K. hirtipes murrayi (A-E); K. h. chapalaense (F); K. h. chapalaense x A", h. murrayi (G); A", h. magdalense (H); K. h. tarascense (I); K. h. megacephalurn (J), and A", h. hirtipes (K). Problematical localities (San Luis Potosi, C; Balsas, D; and Puebla, E) are discussed in text. Stippled area indicates a portion of the allopatric range of K. sonoriense (see Fig. 3). No. 1 Kinosternon Biosystematics Rio Magdalena, Sonora (MAGD). — The Rio Concepcidn (the major tributary of the Rio Magdalena) arises in the hills near Nogales, Arizona, flows as a perma- nent stream through the Magdalena Val- ley, and disappears into the coastal sands of northwestern Sonora (Tamayo 1964: 102). Kinosternon sonoriense is known from numerous permanent water habitats between about 300 and 1200 m elevation. Kinosternon flavescens is also known from this basin (Iverson, 1979). Rio Sonora, Arizona and Sonora (SNRA). — Like the Rfo Magdalena, the Rio Sonora rises near the Arizona-Sonora border and disappears (below Hermosillo) into coastal sands (Tamayo 1964:102). Kinosternon sonoriense is locally very abundant in this basin in permanent water habitants between at least 200 and 1200 m elevation. Kinosternon flavescens is the only other freshwater turtle known from this basin (Iverson, 1979). Rio Yaqui basin (excluding the Papigo- chic drainage), Arizona, Chihuahua, and Sonora (YAQ). — Because of the zoo- geographic dissimilarity of the Yaqui basin west of the Continental Divide (Rios Yaqui, Moctuzuma, Bavispe, and Aros) and those east of the Divide (Rios Papigo- chic and Tomochic) (Meek, 1904; Miller, 1958), and because obvious differences were observed early in the study between the turtles of the K. hirtipes group on either side of that Divide in those rivers {Kinosternon sonoriense in streams to the west, K. hirtipes in those to the east), the Yaqui sample was divided into a Plateau portion (hereafter called the Rfo Papi- gochic sample) and a non-Plateau portion (hereafter restrictively called the Rfo Yaqui sample). Kinosternon sonoriense is known only from headwater populations in permanent water situations between 1200 and 2000 m elevation in southeastern Arizona and adjacent northeastern Son- ora, and western Sonora near the Chihua- hua border. A specimen from El Novillo, Sonora, catalogued as A', sonoriense (UAZ 36505) but unseen by me (not mapped on Fig. 3, but mapped in Iverson, 1976) is probably misidentified since it is the only record of K. sonoriense in the lower Ri'o Yaqui. Additional field work in this basin is badly needed. The freshwater turtles, Kinosternon integrum, K. flavescens, K. alamosae, and Pseudemys scripta also occur in the Yaqui basin (Legler and Webb, 1970; Iverson, 1978, 1979; Berry and Legler, 1980), but the microsympatry of any pair of species in the Yaqui basin has not been established Rio Fuerte basin. Chihuahua and Sinaloa. — Three specimens of Kinoster- non sonoriense (identification verified) collected by Wilmer Tanner bear the lo- cality data "Cerocahui, Chihuahua" (question mark in Fig. 3). As mapped by Tanner and Robison (1960), and Conant (1978:466), the locality Hes along a tribu- tary of the Rio Fuerte (Pacific drainage). No other members of the K. hirtipes group are known from this basin, whereas K. integrum is abundant at lower eleva- tions (Berry, 1978; Iverson, unpublished). Contreras-Balderas (1975) suggested that the fish Notropis (Codoma) ornatus (pri- marily an inhabitant of the Mexican Pla- teau) may also inhabit the headwaters of the Rfo Fuerte. This would indicate hist- orical faunal interchange (perhaps stream capture) between the Plateau and the upper Rfo Fuerte, and might have permit- ted K. hirtipes, but not K. sonoriense, to reach the Fuerte. The Cerocahui locality must therefore remain problematical until this rarely visited area in southwestern Chihuahua is better studied. Rio Casas Grandes Interior Basin, Chi- huahua (CSGR). — The headwaters of the Rio Casas Grandes are in the Sierra Madre Occidental, very close (25 km radius) to the headwaters of the Rfo Bavispe (Yaqui) and Ri'o Papigochic basins. In fact, headwater streams of the Bavispe and Casas Grandes reach within 6 km of one another at about 2000 m, southwest of Pacheco, Chihuahua. The basins are there separated by a divide less than 200 meters high. Tulane Studies in Zoology and Botany Vol. 23 Tamayo (1962:475) provided a photo- graph of the Rio Casas Grandes, presum- ably south of the town of that name, and Goldman ( 1 95 1 : 1 1 9- 1 22) describes several habitats in this basin. Below (north) the town of Nueva Casas Grandes, the river is diverted for agricultural purposes and can at best be called intermittent. The river terminates in Laguna Guzman (1180 m; photographs in Henrickson, 1977) but seldom (only during floods) does that Laguna receive water via the Rfo Casas Grandes. K. sonoriense apparently occurs only above (south oO Nueva Casas Grandes (1475 m), up to an elevation of between 1500 and 1600 m in the Rio Piedras Verdes near Colonia Juarez. My trapping operations were unsuccess- ful on 13 May 1977 in the main channel of the Rfo Casas Grandes at a bridge on Highway 2 between Janos (ca. 10 km N) and Ascension; only catfish (Ictalurus sp.) were trapped in apparently permanent pools even though locals told me that "rock" turtles lived in the stream. Inde- pendently, Conant (1978:488) took only catfish and bullfrogs, and collected no turtles in his traps during two days of field work at the same site. On 1 August 1980, the Rio Piedras Verdes above Colonia Juarez was very shallow (average depth, 10-20 cm; maxi- mum depth, 0.5 m) and slowly moving. In one hour, two K. sonoriense were collect- ed by hand in shallow water and two more were taken in traps set in the deepest areas. Roger Conant (pers. comm.) trap- ped eight K. sonoriense near this same lo- cality on 19 August 1974 when the river was in flood. No other aquatic turtles are known from the Casas Grandes basin; Van De- vender and Van Devender's (1975) Chrys- emys picta record was based on specimens actually from the Rio Santa Maria basin. Rio Santa Maria interior drainage, Chi- huahua (STMR). — The Rio Santa Maria rises in the Sierra Madre Occidental very close to the Rfo Papigochic basin, and flows northward across the desert floor in northwestern Chihuahua. The river bed terminates in Laguna de Santa Maria (1172 m), and like Laguna Guzman sel- dom receives water from its confluent stream. The Laguna de Santa Marfa (photographed in Henrickson, 1978) is separated from the Laguna de Guzman by a divide of not more than 61 m elevation (Goldman, 1951:123). Geological evi- dence suggests the lakes were continuous during Wisconsin time when the water level reached at least 1225 m (Axtell 1978: 509). Kinosternon hirtipes reaches its north- ernmost distribution in the Rio Santa Maria, and is common in permanent wa- ter situations throughout the basin be- tween at least 1400 and 1600 m elevation. On 12 May 1977, in a tributary of the Rio Santa Maria, southeast of Galeana, two assistants and I captured nearly 100 individuals of K. hirtipes by hand (most of which were subsequently released) in less than two hours, primarily by feeling under stream banks. Seven man hours of hand collecting on 2 August 1980 at the same locality produced 140 turtles, which were measured, marked and released as part of an ecology study. Habitats near that location were described by Van Devender and Van Devender (1975). Semmler et al. (1977) reported similar suc- cess in the Rio Santa Maria, also near Galeana. Both K. flavescens (Iverson, 1979) and Chrysemys picta (H. M. Smith and Taylor, 1950a; Iverson, field notes; Roger Conant, pers. comm.) also occur in the Santa Maria basin. Rio Carmen (^ Rio Santa Clara) interi- or drainage, Chihuahua (CRMN). — The headwaters of the Rfo Carmen lie in the Sierra Madre west of the Sierra del Nido. The river once flowed (at least during floods) to the Lago de Patos (1175 m) near Villa Ahumada, but since the con- struction of a dam (Presa de Las Lajas) just south of Ricardo Flores Magon ( = El Carmen) the river no longer flows south past Ricardo Flores except in cement irrigation flumes. When visited on 11-12 May 1977, the remnants of the riparian No. Kinosternon Biosystematics woodland below the dam were still in evi- dene, but rapidly disappearing (see also Conant 1977:488). Sixty trap hours along the shores of the Presa yielded no turtles. The rocky shoreline lacks aquatic vegeta- tion and the continually changing shore- line (evident from water marks on the rocks) presented habitats which were un- doubtedly generally unsuitable for Kino- sternon turtles. However, Conant (1978: 473) indicated that he recently obtained K. hirtipes in the impoundment. Records for K. hirtipes are available from the Rfo Carmen below (north of) the dam at Ricardo Flores (15(X)-1600 m), up (South) to the region near Santa Clara (1800 m). Kinosternon flavescens is the only other aquatic turtle known from this basin (Iverson, 1979). Rio El Sauze (= Encinillas) interior drainage. Chihuahua (SAUZ). — The Rfo El Sauz rises in the eastern slopes of the Sierra del Nido, and flows intermit- tently to the desert floor in the vicinity of the town of Sauz. It then flows intermit- tently northward, paralleling the Sierra del Nido, until it disappears into the ground about 100 km north-northwest of Ciudad Chihuahua. According to Hubbs (in Hubbs and Springer 1957:299; and in Miller 1961: 393) the entire Sauz Valley went dry in 1947; however, Minckley and Koehn (1965) recorded an apparently permanent, though artificial pond (with fishes) in the Sauz Valley in 1964 and Contreras-Bal- deras (1974:182) reported fish collections made in 1964 and 1968. Kinosternon hirtipes has been collected from at least four localities in the Sauz Valley between 1500 and 1700 m; how- ever, we trapped none at either of two sites of apparently permanent water (con- taining fishes) near Sauz on 1 1 May 1977. Local children at that time confirmed the occurrence of turtles in the stream, but said they were uncommon; they were also unsuccessful at securing any for us. The only other aquatic turtle supposed- ly recorded from this basin is Sterno- therus odoratus, but the single record re- mains problematical (Moll and Williams, 1963; Conant and Berry, 1978). Although unknown, K. flavescens may occur in this basin (Iverson, 1979). Alamito Creek drainage, Presidio Co., Texas (TEX). — Alamito Creek is an ephemeral tributary of the Rio Grande, east and north of Presidio, Texas. K. hir- tipes is known from only two permanent ponds in this drainage in Texas (about 1050 m). Conant and Berry (1978:11-15) elaborated on the specific localities and field work in the area. This species has not been found in the Rio Grande itself or its American tributaries in southwest Texas, but populations may reside in permanent Mexican tributaries between Presidio (or Ojinaga, Chihuahua) and the Big Bend of the Rio Grande. Kinosternon hirtipes does not likely occur today above Presidio in the Rio Grande since even by 1919 that river was sometimes completely dry between the New Mexico border and the mouth of the Rfo Conchos (Udden, Baker, and Bose 1919:23). Kinosternon flavescens is the only other aquatic turtle occurring in the Alamito drainage with K. hirtipes; based on museum records, the two species occur microsympatrically in at least one of the spring-fed ponds in this basin (Iverson, 1979). Rio Conchos drainage. Chihuahua and Durango (CNCH). — The Rio Conchos is the major tributary of the Rio Grande (= Rio Bravo), accounting for 18% of the latter's total flow (Tamayo 1964:89). With its headwaters in the Sierra Madre in southwestern Chihuahua and extreme northern Durango, it drains more of the state of Chihuahua than any other single river. More specimens of K. hirtipes have been collected in the Rfo Conchos basin than in any other basin. The species is known from the mouth of the Conchos (ca. 800 m; the lowest altitudinal record for the species) near Ojinaga (Legler, 1960) to its more accessible headwaters (in the Rfo Florido) near Las Nieves in north- 10 Tulane Studies in Zoology and Botany Vol. 23 ern Durango (ca. 1800 m). Bushnell (1971:332) provided a photograph of Lago Boquillo on the Rfo Conchos, a lo- cality from which K. hirtipes is known. Kinosternon flavescens, Pseudemys scripta, and Trionyx spiniferus are also known from the Conchos basin (Legler, 1960; H. M. Smith, et al., 1963; Webb, 1973; and Iverson, 1979). Laguna Bustillos interior drainage, Chihuahua (BVST). — The small Laguna Bustillos basin (2720 square km) is in the foothills of the Sierra Madre west of Ciu- dad Chihuahua. The Laguna itself lies at approximately 19(X) m. A single collection (UMMZ; 4 specimens) of K. hirtipes is available from a tributary 27.4 km north of Ciudad Cuauhtemoc. I trapped along a clear, shallow, permanent, though inter- mittently flowing stream with an adjacent cattle tank (pond) 24.3 km north of Cuauhte'moc on Hwy. 28 on 10 May 1977. These two localities probably are the same, since I could find no other tribu- taries on the road north from Cuauhte'- moc. No turtles were captured, although the microhabitat seemed adequate in some of the deeper areas (maximum, 0.5 m). Populations of K. hirtipes may exist in the Laguna itself or in this tributary where it leaves the mountains to the west (and thus presumably has more flow). Further field work is warranted to deter- mine if K. hirtipes is still extant in the Bustillos basin. No other aquatic turtle is known from the basin. Rio Papigochic, Chihuahua (PAP). — As mentioned above in regard to the Rfo Yaqui basin, the plateau portion of that basin, the Rio Papigochic, is here consid- ered a separate sample area. The Papigo- chic arises in the Sierra Madre west of Ciudad Chihuahua and flows northwest- ward to the vicinity of Yepdmera, where it turns sharply southward for about 25 air- line km to the confluence of the Rfo Tom6chic and another sharp turn to the northwest. It continues in that direction for almost 150 airHne km before heading southwestward to its confluence with the Rio Aros (tributary to the Yaqui) just in- side the Sonora border. Because of this anomalous drainage pattern and because the faunal affinities of the Papigochic are with the Rfo Conchos and not the Rfo Ya- qui (sensu stricto), zoogeographers believe the Rfo Papigochic was until re- cently (prehistorically) a tributary of the former stream (Meek, 1904; Miller, 1958; among others). Kinosternon hirtipes is known from both the Tomochic and Papigochic from elevations of 1200 to at least 2000 m. Like Van Devender and Lowe (1977), I found K. hirtipes very common near Yepomera. On 10 May 1977, 5 traps set in a broad, ponded stretch of a Papigochic tributary yielded 25 turtles in two hours. No other turtle species were collected (or are known from the basin), but Tom Van Devender (pers. comm.) reports that natives near Yepdmera told him of a "tortuga pinta" that lives in the Papigochic basin. The va- lidity of the report and the identity of the turtle (perhaps Chrysemys picta) are un- confirmed. Rio Nazas interior drainage, Durango and Coahuila (NAZ). — More of the state of Durango is drained by the Rio Nazas than any other single drainage sys- tem. The Nazas rises in the Sierra Madre in western Durango as two major tribu- taries, the north-flowing Rio Santiago and the south-flowing Rfo Tepehuanes. The confluence of these two streams, about 20 airline km northwest of Santiago Papasquiaro begins the Rio de Ramos. The major northern tributary, the Rfo del Oro rises in the Sierra in northwestern Durango and joins the Rfo Ramos near El Palmito below which it is called the Rfo Nazas. The Nazas then flows eastward across the Chihuahuan desert (photo- graphed in Spieth, 1950:34), formerly as far as the Bdlson de Mayran in southwest- ern Coahuila. Diversion of the waters for agriculture near Torredn has, however, severed the Nazas-Mayran connection (Conant, 1963, provides an excellent de- scription of the topography of this basin). Trapping on the Rio Nazas west of Tor- redn (78 trap hours, 5 - 6 May 1977, 3 No. I Kinosternon Biosystematics 11 Kinosternon; 156 trap hours, 23 - 24 July 1980, 6 Kinosternon) and near El Palmito (98 trap hours, 7-8 May, 1977, 5 Kinoster- non) by me, and near El Palmito (30 trap hours, 20-21 July 1976, 5 Kinosternon). Roger Conant (pers. comm.) indicated that K. hirtipes is uncommon in the river itself. The absence of a large series from any one locality in the Nazas basin sup- ports this statement. The species is known from between 1100 (Lerdo) and 1400 m (El Palmito). Pseudemys scripta was collected at all three locations I visited and that species probably is sympatric with K. hirtipes in most of the Nazas drainage. K. flavescens is the only other aquatic turtle known from this basin (Iverson, 1979). Viesca interior basin, Coahuila (VSCA). — Only a single collection of A". hirtipes (and Pseudemys scripta) is known from the small area south of the city of Viesca, Coahuila (symbol J in Fig. 4). Bryce Brown seined 8 Kinosternon and 2 Pseudemys scripta on 4 June 1961 from 2 drying ponds south of Viesca at about 1 100 m. Natives informed Brown that the ponds were remnants of a once active hot spring (pers. comm.). At my suggestion, ichthyologist Robert Rush Miller visited the Viesca area in the spring of 1978 and verified (pers. comm.) the fact that a spring did once exist along the mountains southwest of the city, but he could find no permanent aquatic habitats suitable for turtles or native fishes. We visited the area on 23 July 1980 and an elderly Viesca resident showed us the locations of the extinct springs (8 total; 1 hot) and confirmed the lack of permanent surface water today. He told us that the springs had gone dry "about 20 years ago" but that prior to that time there had been much water, with many turtles, snakes, and fish. The only permanent water near Viesca of which anyone there knew was a spring near the small town of La Pena, about 20 miles to the East. We visited that spring on 23 July 1980 (as Miller did in 1978), found only introduced fish and no Kinosternon, and failed to locate the Pseudemys script a-WV.^ turtles locals told us "used to occur" in the two imy presas that remain. The Viesca turtle populations must therefore be considered extinct. The specimens of Pseudemys scripta from near Viesca are very similar to those in the Rio Nazas immediately to the west. The Kinosternon, however, show little af- finity with Nazas specimens and in fact have the most unique morphology of any member of the hirtipes group (see later). Rio Aguanaval interior drainage, Coa- huila, Durango and Zacatecas (AGUN). — The Rio Aguanaval rises on the Mex- ican Plateau in the mountains northwest of the city of Zacatecas and flows inter- mittently northward across the Chihua- huan Desert. It once emptied into the La- guna de Viesca in southwestern Coahuila, before its diversion for agricultural pur- poses (Conant 1963, 1969). K. hirtipes is the only turtle known to occur in the Aguanaval basin and has been collected only in the headwaters im- mediately northwest of Fresnillo between 2000 and 2200 m. Natrix erythrogaster alta, endemic to the Aguanaval system, is likewise known only from the headwaters (Conant, 1969:46). Laguna de Santiaguillo interior drain- age, Durango (STGO). — The Laguna Santiaguillo is isolated at just under 2000 m in the Sierra Madre northwest of Canatlan, Durango. Only two collections (four specimens) of K. hirtipes have been made in the Santiaguillo basin, both ap- parently from the same tributary to the Laguna near the village of Guatimape. On 7 May 1977, this tributary was temporari- ly reduced to isolated pools (maximum depth 1 m) in the stream bed. Trapping and seining produced three K. hirtipes. A presa with permanent abundant water was subsequently located about 0.7 km up- stream from the highway bridge, but was not sampled; it probably supports a good population of A', hirtipes. No other turtle is known from the basin. Rio Mezquital drainage, Durango (MEZ). — The Rk) Mezquital is the ma- 12 Tulane Studies in Zoology and Botany Vol. 23 jor tributary of the Rio San Pedro (Pacif- ic drainage), and drains that portion of the Mexican Plateau near Ciudad Du- rango, Durango (Albritton 1958: Conant 1963). Both K. hirtipes and K. integrum occur in the Plateau portion of the Rio Mezquital, but only AT. integrum is known from the Pacific coastal plain portion of the Mezquital-San Pedro system (Iverson, unpublished). Because of the inaccessibili- ty of the area south and southeast of Ciu- dad Durango, the lower Hmit of the range of K. hirtipes in the Mezquital is uncer- tain. The locality farthest downstream is at Mezquital on the Rio Mezquital, south- east of Ciudad Durango at about 1100 m. The species reaches its highest known ele- vation in this basin (2600 m) at Otinapa. Conant (1978:467, 473) correctly ques- tioned the record (H. M. Smith and Tay- lor, 1950a:26) of K. sonoriense from Durango, Durango (this basin); it was ap- parently based on a specimen of K. hir- tipes. The Rfo Mezquital and its Plateau trib- utaries near Ciudad Durango contain much permanent water even at the end of the dry season. On 6 May 1977 at the Ri'o La Sauceda (tributary to the Mezquital) bridge on Hwy 40 (Figure 5), 60 K. hir- tipes (pre-dominately juveniles and sub- adults) entered 11 traps in just three hours. The most productive traps had been set along steep, undercut dirt banks; traps in areas of gently sloping shoreHnes were unproductive. Six traps set for 45 minutes at the same locality on 25 July 1980 produced 14 K hirtipes. Based on museum records, K. hirtipes is very common in the Mezquital system (148 specimens), whereas K. integrum is un- common (1 am aware of but 14 specimens); a single collection of turtles from 0.8 km N Graceros contains 12 K. hirtipes (KU 68733-36, 68738-45) and a single K. integrum (KU 68737). El Salto area, Rio Acaponeta basin, Durango (SALT). - K. hirtipes appar- ently occurs in the Rfo Acaponeta only in its headwaters northeast of El Saho (Symbol B in Fig. 4). On 25 July 1980, at a shallow (maximum depth 0.75 m) ap- parently permanent stream 9.7 road km ENE El Salto (about 2400 m) in a moun- tain meadow surrounded by pine-oak woodland, six K. hirtipes were collected in 12 trap hours and three more were taken by hand. K. integrum is very common in the lower Rio Acaponeta basin (Berry 1978; Iverson, unpublished), but is unknown from the headwaters region. Southwestern San Luis Potosi (interior?) basin (SLP). - In an isolated portion of the Rio Santa Maria drainage basin (Panuco, i.e., Atlantic drainage southwest of Villa de Reyes (symbol C in Fig. 4), K. hirtipes and K. integrum co- occur abundantly. Iverson and Berry (1979) argued that this population of K. hirtipes is the resuh of an introduction. I continue to support that view, especially since K. integrum has been collected at numerous localities in the Rfo Santa Maria system (see list in Iverson and Berry, 1979), yet K. hirtipes is known only from the Laguna de las Rusias ( = Presa El Refugio; ca. 1900 m) area. As de- scribed by Iverson and Berry (1979:320) the only remaining aquatic habitat found on 11 June 1978 was a small permanent stream that was diverted entirely for agriculture within 2 km below the broken dam. The Arroyo below the dam was lined with seepage springs and quaking ground. The stream varied from one to four m wide (x ^ 1) and averaged only 0.25 m deep (over a soft mud bottom at least 1 m deep). The water was quite clear along most of its length, but odor and Figure 5. Rio La Sauceda (Rio Mezquital basin) at Highway 40 near city of Durango, 6 May 1977. No. 1 Kinosternon Biosystematics 13 refuse in the water indicated its use as a human sewage effluent. Kinosternon was abundant in the stream on 11 June; 13 K. hirtipes and 3 K. integrum were dipnetted or trapped in one hour. Rio Aguascalientes drainage, Aguas- calientes (AGUAS). - Because the distance between the two clusters of localities for K. hirtipes in the Rfo Aguas- caHentes- Verde system is so great, I have arbitrarily divided tTie system into two parts: the Rio Aguascalientes basin in the state of Aguascalientes and the remainder of the Rio Verde, primarily in Jalisco. Collection data for museum specimens indicate that Kinosternon hirtipes and K. integrum co-occur in the Rio Aguas- cahentes between at least about 1900 and 2000 m elevation. Oswaldo Mooser (pers. comm.) indicated that K. hirtipes is much less common than K. integrum in Aguascalientes, The fact that 21 museum specimens of K. integrum are available from ten localities in Aguascalientes whereas 15 specimens of K. hirtipes are known from only five localities (Iverson, unpubl.) support his contention. Mooser's field observations also indicate that the former occurs only in permanent water situations, wereas the latter is common in those situations as well as temporary aquatic habitats. I have not visited the Rfo Aguascalientes basin. Rio Verde drainage, Jalisco (VERD). - Draining most of northeastern Jalisco, the Rio Verde empties near Guadalajara into the Rio Grande de Santiago, which flows through the Sierra madre Occidental to the Pacific Ocean. K. integrum occurs throughout the entire Verde-Santiago system (Berry 1978; Iver- son, unpublished), but within the Rio Verde system (excluding Aguascalientes) K. hirtipes is known from only three localities. At least two are permanent water situations, their permanence enhanced by the construction of dams. On 16 June 1978, I collected four K. integrum and three K. hirtipes in one hour at the most southerly known locality in this basin Gust over 1800 m), a tributary of the Rfo Verde north of the city of Valle de Guadalupe. At that time the stream was reduced to a series of isolated, well- vegetated pools (one to two m across and P 1 m deep) in the channel below a large stone dam. The impounded reservoir (Presa Canada Grande) was unvegetated and reduced to a small (50 m X 50 m), deep (2 m *), muddy pond immediately behind the dam. The heads of literally hundreds of Kinosternon were visible at the pool's surface on that day. The isolat- ed pools in the stream channel, however, contained few turtles, usually only one per pool. Turtles were alsoablmdant here on 10 May 1981, when, gravid female K. hirtipes were seen basking. The other locality (ca. 2000 n\) below the Presa el Cuarenta on the Rio del Cuarenta (= Rfo de Lagos = Rio San Juan de los Lagos) near the village of Paso de Cuarenta, was also visited on 16 June 1978. The large cement dam im- pounds a huge, muddy, unvegetated res- ervoir. Because of the apparent lack of turtle habitat along the shoreline, the presa was not trapped. Below the dam, however, were numerous rocky-shored (in the main channel and well vegetated, mud-shored overflow and seepage pools. The more vegetated ponds were most pro- ductive, and eight trap settings produced 15 K. hirtipes and two K. integrum in one hour. K. hirtipes is probably much more common in the Rio Verde than is indi- cated by the paucity of locality records. Additional field work should verify this prediction. Of particular interest is the downstream limit of A", hirtipes in the Rfo Verde. Rio Grande de Santiago drainage. - This river links the Lago de Chapala with the Pacific Ocean and passes through the Sierra Madre Occidental. Tanner and Robison (1960) reported the collection of three unidentified Kinosternon from 7.5 mi. north of Magdalena, Jalisco (1000 m elevation) in this basin. One of the included specimens (BYU 14630) is un- questionably K. sonoriense and the 14 Tulane Studies in Zoology and Botany Vol. 23 locality data for that specimen must be considered in error. The other two speci- mens supposedly collected at the same locality (BYU 14631-32) are presently unlocatable (Tanner, pers. comm.). I doubt the natural occurrence of a K. sonoriense-WkQ member of the K. hirtipes species group in the Rio Grande de Santiago below Guadalajara. Of those major systems joining the Rio Santiago below Lake Chapala, only the Verde harbors K. hirtipes (see above). K. integrum however, ranges throughout the Lerma and Santiago basins and all their tributaries, both on the Plateau and off (Iverson, unpublished). Conant's (1978:465; Map 10) two most southerly records for K. hirtipes in Durango were erroneously plotted in the Rio Atengo drainage, a tributary to the Rio Santiago (Roger Conant, pers. comm.); the records belong in the Rio Mezquital basin near Ciudad Durango. Rio Lerma drainage, Jalisco, Guana- juato, Michaocan, and Mexico. — Most of the southern portion of the Mexican Plateau is drained by the Rio Lerma and its tributaties. The Lerma originates in the springs and lakes in the southern end of the Toluca Valley at over 2400 m (Gold- man 1951:185, 305, plates 59 and 60; Tamayo, 1964:104; Romero, 1965), and flows northward and then westward across the southern Plateau to Lake Chapala. The basic physiography of the river along most of its course is discussed by Barbour (1973:541) and Tamayo (1962). Like the Rio Grande de Santiago, the Rio Lerma' s south bank tributaries are not extensive (the river flows parallel to and immediately north of the Sierra Volvanica Transversal), whereas several of those on the north bank are very large (notably the Rios Turbio and de la Laja). Because the Lerma basin is well over 400 airUne km long, I chose to subdivide it for analysis of its resident turtle popu- lations. I have followed Barbour (1973: 540) in his division of the Rio Lerma basin into four physiographic regions: the Valley of Toluca (TOL; above the canyon below Temascalcingo in the state of Mex- ico), the Maravati'o basin (MAR; from near Temascalcingo, Mexico through Michoac^n and Guanajuato to the rapids near Salvatierra, Guanajuato), the Baji'o (BAJ; from near Salvatierra, Guanajuato to the region between Piedad and Yure'- cuaro, Michoac^n), and the Lake Chapala basin (CHAP; in Jalisco and Michoacan. [The reader is referred to Barbour (1973) for discussion of these physiographic provinces.] I have also considered the turtles in the lower Rio Lerma tributary, the Rio Duero (DUER) (historically a tributary of Lake Chapala; see Tamayo, 1962:404), as a separate pop- ulation for purposes of analysis. Each of these subdivisions is discussed separately. K. hirtipes and/or K. integrum are the only aquatic turtles presently known to occur in these basins (But Pleistocene fossils of Pseudemys cf. scripta are known from near Lake Chapala; Tom Van Devender, pers. comm.). Valley of Toluca basin, Mexico (TOL). - No turtles are known from the Toluca basin, but K. hirtipes probably occurs in the springs and lakes near the Rio Lerma headwaters in southeast Mex- ico state. Maravatio basin, Mexico, Michoacan and Guanajuato (MAR). - Only a single broken K. hirtipes shell (KU 43637) is available from this basin, and this region is thus unrepresented in subsequent analy- sis. No K. integrum are known from the Maravatio (Iverson, unpublished), but both species probably occur throughout the basin. Bajio basin, Guanajuato and Micho- acan (BAJ). - This basin includes the drainages of the Lerma tributaries, the Rio de la Laja and Rio Turbio. Both K. hirtipes and K. integrum occur in the Bajio up to at least 1900 m (north of San Miguel de Allende; Iverson, unpublished). On 12 June 1978 I sampled two marshy areas near the Rio de la Laja between San Miguel de Allende and Dolores Hidalgo, Guanajuato. An hour at each locality pro- duced three K. hirtipes (two by hand; one No. 1 Kinosternon Biosystematics 15 trapped) and two K. hirtipes (one trapped, one seined) respectively. Rio Duero drainage, Michoacdn (DUER). - K. hirtipes is known from only one locality in this drainage. At spring- fed, cypress-lined Lago Came'cuaro (1700 m), east of Zamora (Symbol G in Fig. 4), I found K. hirtipes very abundant on 14 June 1978; 17 trap hours produced 20 K. hirtipes and one K. integrum. The latter species is known from several other locali- ties in the Rio Duero and likely occurs throughout the basin. Villa Victoria basin, Mexico (state) (VILLA). - K. hirtipes has been collected at only three localities within a 3 km radius in this basin (part of the Rio Balsas basin; see later) at about 2500 m; K. inte- grum is therein unknown. K. hirtipes is apparently not common in the basin as evidenced by our collection of only two specimens in 134 trap hours at four locaH- ties below the Presa Villa Victoria on 21-22 June 1979. Lago de Chapala basin, Jalisco and Michoacdn (CHAP). - Lake Chapala (el- evation 1525 m) is 80 km long, east to west and covers about 1685 km^ (Debuen 1945; Deevey 1957; see photographs in Tamayo, 1962). Average depth is only 8 m (Tamayo 1964:105), and maximum depth is probably 9.8 m (Cole 1963:413). The Rio Lerma flows into the extreme eastern end of the lake and the Rk) Grande de Santiago exits the lake about 15 km north of the Lerma's mouth. There are no other large confluent streams. All records but one (Jiquilpan; Duellman, 1961) of K. hirtipes from the Chapala basin are from along its shores. K. integrum has also been commonly collected along the lake shores (Berry 1978; Iverson, unpublished). Trapping along the south shore near Tuxcueca on 15 June 1978 produced no turtles whatsoever, although J.F. Berry (pers. comm.) obtained a series oi K. inte- grum at the same locality in June 1975. The once extensive marshes at the east- ern end of Lake Chapala probably sup- ported dense populations of Kinosternon turtles, but drainage operations have un- fortunately nearly eliminated this habitat (Goldman, 1951:173-174). Laguna de Zapotldn interior drainage, Jalisco (ZAPO). - The Zapotlan basin lies north of Ciudad Guzman, Jalisco in the Sierra Volcanica Transversal. Only K. hirtipes is known from the basin and all specimens apparently originated at the southern end of the lake (ca. 1 500 m) near Ciudad Guzman. Gadow's (1908) record of Cinosternum integrum from this basin must therefore be based on K. hirtipes. San Juanico Valley interior drainage, Michoacdn (SNJ). - The Valley of San Juanico (north of Cotija, Michoacan) was until recently an isolated, interiorly drained basin, formed prehistorically by the damming of a northward-flowing trib- utary of the Chapala basin by a lava flow (Alvarez 1963, 1972; Barbour 1973). The construction of the Presa San Juanico across the valley's southern end has en- larged Lake San Juanico, and directed its effluents southward to the Balsas system (Alvarez 1972:158; Barbour 1973; pers. observ.). Turtles of the Kinosternon hirtipes group are the only turtles known from the valley behind and above the Presa San Juanico (ca. 1500 m; Symbol H in Fig. 4). Field work by Clyde Barbour (pers, comm.) and my own field crew (75 trap hours, 14-15 June 1978; 180 trap hours, 6-7 May 1981) in i\iQ presa (Figure 6) have produced only 7 specimens, three Hving (one seined by Barbour; two trapped by me) and four articulated shells (by my crew). Ichthyological field work in the presa on three dates in 1962 and 1963 by Alvarez (1963) apparently produced no turtles. The diversion of effluents from the San Juanico Valley to the Balsas appears to be permitting K. integrum (known through- out the Balsas system; Berry, 1978; Iver- son, unpublished) to expand its range toward the presa. Although no K. inte- grum are known from above the presa, and although we obtained no turtles 3 km below the dam in 1978 in one of the two effluent irrigation ditches (during one 16 Tulane Studies in Zoology and Botany Vol. 23 hour of seining and 12.5 trap hours), a single K. integrum observed sunning along the other ditch (ca. 100 m below the dam) was seined. An additional juvenile K. integrum (TUL 19504) is also known from one of the effluent ditches where it crosses the Cotija-Tocumbo road (dis- tance below dam uncertain). K. integrum likely will soon invade Lake San Juanico. In June 1978, the lake itself was very low; much of the muddy bottom was ex- posed due to evaporation and diversion for agriculture (Fig. 6). A small water hyacinth population represented the only obvious vegetation in the muddy lake. In 1981 the lake was even lower and new ditches were draining the lake even more. No turtles were collected and their exist- ence seems tenuous. Lago de Cuitzeo interior basin, Micho- acan CUIT). - Lago de Cuitzeo (just over 1800 m) is the largest interiorly drained natural lake in Mexico. It is fed primarily by the Rio Grande de Morelia, which heads in the mountains east of P^tzcuaro and flows southeastward to its confluence with the southeastern shore of the lake (Camacho, 1925). The lake is very shal- low, and has been known to be nearly dry (Debuen, 1943). Aquatic vegetation is accordingly uncommon. When visited on 12 June 1977, the lake level was very high and therefore not trapped for turtles. Based solely on museum specimens, K. hirtipes (one specimen) is much less com- mon than K. integrum (35 specimens, 7 localities; Iverson, unpublished). The single available specimen is a poorly pre- served male. Lago de Pdtzcuaro, interior basin, Michoacan (PATZ). - Lago de Patzcuaro (2,035 m; Symbol I in Fig. 4; see phot- graphs in Tamayo, 1962:493 and Solor- zano Preciado 1961:55) has been well- studied limnologically (summary in Cole, 1963), but its turtles have been only in- frequently mentioned (Duellman, 1961; Altini, 1942) or completely ignored (Martin del Campo, 1940). It has a sur- face area of only about 1 1 1 km^ and has a maximum depth of 15 m (DeBuen, 1944). Figure 6. Presa San Juanico, Michoacdn, looking northwestward from dam, on 15 June 1978. Reservoir was much reduced due to irrigation de- mands and the dry season. Kinosternon hirtipes magdalense was collected along the dredged canal in foreground. Emergent vegetation (primarily Scirpus) is common along the shoreline, especially the southern margin (Goldman, 1951:195, plate 58; Barbour, 1973:543), where the mats often extend out 20 m or more from the shoreline (pers. observ.). On 12-13 June 1978, the lake was quite clear; however, we saw no turtles during the day or night in shallow water (< 1 m) in narrow strips of shoreline on the south- eastern shore where emergent vegetation has been removed for docks (Fig. 7). Kinosternon hirtipes (the only turtle known from the lake) were, nevertheless, for sale the next day in the Patzcuaro (city) market, an apparently frequent oc- currence. The smaller (8 km^) higher (2120 m), younger, and deeper (maximum depth 45 m) Lago de Zirahuen (DeBuen, 1943, 1944) immediately to the southwest, is believed to have been historically drained by a tributary of the Rio Lerma flowing through the Lake Patzcuaro and Lake Cuitzeo basins (DeBuen, 1943). No turtles are known from Lake Zirahuen. Rio Balsas drainages, Michoacan and Puebla (BALS). - Turtles of the hirtipes group have been recorded from only three localities in the Rio Balsas basin, the river sytem which drains most of southern Mexico south of the Sierra Volcanica No. 1 Kinosternon Biosystematics 17 Transversal. The Villa Victoria (VILLA) localities have already been discussed. Duellman (1961) recorded K. hirtipes from the Balsas on the basis of UIMNH 24707 from 8 km W Ciudad Hidalgo, Michoacan (ca. 2200 m; Rio Tuxpan Valley; Symbol D in Fig. 4). I have exam- ined the specimen and believe it to be a female K. integrum. However, another specimen from the same locality (AMNH 62257) is unquestionably a female K. hirtipes. In addition, I collected a single male K. hirtipes in 128 trap hours in a stream of approximately the same locality on 4-5 May 1981. Kinosternon integrum occurs throughout the Tuxpan (and Balsas) system (Berry, 1978; Iverson, un- published); but K. hirtipes is now definite- ly known in the Tuxpan only near Ciudad Hidalgo. The only other supposed Balsas speci- men of K. hirtipes is an adult male (UU 12096) from a tributary of the Rfo Atoyac, 4.5 km S Molcaxac (just below 2000 m; Symbol E in Fig. 4). The identifi- cation is correct, but I question the valid- ity of the data for three reasons. First, because of the numerous highway access- es to the RTo Atoyac drainage and the fact that I know of at least 88 specimens of K. integrum (73 of which I have seen) from 13 localities in the Atoyac-Balsas system- in the state of Puebla, additional speci- Figure 7. Southeastern shore (foreground) of Lago de Pitzcuaro, Michoacan, 13 June 1978. mens o[ K. hirtipes would likely have been collected if the species did occur in that system. Second, nine trap hours at the Molcaxac locality on 3 May 1981 pro- duced 23 K. integrum and no K. hirtipes. Third, based on field numbers and collec- tion dates, Clyde Barbour collected K. hirtipes along the Rfo Lerma in Jalisco (UU 12120) on 7 May 1969 and K. inte- grum in the Rio Turbio in Guanajuato (UU 12083-84) on 8 May and in the Rfo Villeto in San Luis Potosf (UU 12085) on 12 May] immediately before he collected near Molcaxac (18 May 1969). Of the eleven turtles recorded as collected near Molcaxac, ten are definitely K. integrum. I submit that through a mixup, the Mol- caxac locality datum was mistakenly applied to the eleventh specimen, and that the specimen possibly originated some- where in the Lerma basin (Rfos Lerma or Turbio?) where Barbour also collected. Valley of Mexico interior basin (VALLE). - The physiography and histor- ical geology of the Valley of Mexico in which Mexico City lies has been well- studied (Bryan 1946, 1948; De Terra et al., 1949; Arellano, 1953; Sokoloff and Lorenzo, 1953; Zeevaert, 1953; Foreman, 1955; Hibbard, 1955; Maldonado-Koer- dell, 1955; Sears and Clisby, 1955; Mooseretal., 1956; Mooser, 1957, 1963; Deevey, 1957; Lorenzo, 1958; Bernal, 1959; Bribiesca Castrejon, 1960; White, 1962; Golomb, 1965; Bradbury, 1971; and an excellent summary in Barbour 1973:537). The entire basin is about 24 km wide and 113 km long (Foreman, 1955) and covers about 8000 km- (Maldonado- Koerdell, 1955:15). At the time of the Spanish conquest (ca. 1520), the Valley of Mexico was one of the largest interior drainage basins in the Transverse Vol- canic Arc, supporting five large spring- fed lakes (Tamayo, 1964; De Terra et al., 1949; among others). So extensive were the lakes at that time, that the early city of Mexico had been built on an island and the Spaniards were forced to build ships in order to besiege the city (Huntington 1914:96). Tremendous fluctuations in 18 Tulane Studies in Zoology and Botany Vol. 23 water level in the Valley prompted drain- age operations in the late 16th century, and by 1608 some of the Valley's water was diverted northward to the Rfo Tula (Atlantic drainage; Tamayo, 1964; Barbour, 1973:540). This artificial drain- age system was finally completed in 1900 (Huntington, 1914:97; Bribiesca Castrejon, 1960) and they only sizeable lakes in the Valley today are Zumpango (2243 m elevation) and Texcoco (2236 m) (Bar- bour, 1973). Only about lOiVo of the Valley floor is covered with water (Foreman, 1955). For discussions of the changing conditions of the lakes since about 1500 AD, see Bribiesca Castrejon (1960). Aquatic habitats in the Valley were discussed and photographed by Gadow (1908:6) and Goldman (1951: 138-39, plate 56). , , Apparently the first record of a turtle from the Valley of Mexico is Wagler's (1830) description of Cinosternon hirtipes (see later justification). Numerous liter- ature records (See synonymies) and muse- um specimens confirm the presence of K. hirtipes in the Valley of Mexico. Kino- sternon integrum has possibly been col- lected in the Valley only three times. FMNH 116521 bears only the data "Dis- trito Federal". Data associated with SM 9722-23 indicate they were purchased in the Xochimilco market on 11 June 1962. Fourteen turtles purchased for me by Gustavo Casas-Andreu in the Xochimilco market in August 1977 are all K. inte- grum. The merchants told him the turtles were from the "Valle de Mexico". Because there is no verifed record of the occurrence of K. integrum in the Valley before 1962, and because all three subse- quent records are apparently from mark- ets, I strongly doubt the natural occur- rence of K. integrum in the Valley of Mexico. K. integrum is, however, very abundant southeast of the Valley in Puebla and Oaxaca (Berry, 1978; personal observation; see discusssion under Rfo Balsas), and may have been imported to the Valley for sale in the markets. Support for such an hypothesis comes from Berry's (1978:83, Fig. 17) discriminant analysis of data from the turtles sent to me by Casas Andreu (UF 41651-64), which clearly showed their affinities to be with turtles from the Upper Rfo Balsas (Rio Mexcala) and the Upper Rfo Papalo- apan (Rfo Santa Domingo basin, Puebla and Oaxaca). Whatever their true origin, K. integrum will likely soon establish itself in the Valley. A study of the inter- action of that species with the native K. hirtipes would be significant. Mittermeier (1971:16) found Pseud- emys scripta elegans, obviously intro- duced, in the markets of Mexico City, where he was told that the species had been introduced into ponds near Mexico City. No other turtle is known from the Valley of Mexico. Unrepresented or unsampled basins. - Several other isolated and/or interior drainage basins within or adjacent to the range of A', hirtipes should be investigated for that species. These include the Laguna de Babicora (2100 m), northwest of Gomez Farias, Chihuahua; the Laguna de Los Mexicanos (2100 m), south of Cuauhtemoc, Chihuahua; the Laguna de Zapacu and Presa de Copandaro, near Zacapu in northern Michoacan; the crater lakes of the Llanos of Pueblo (see Alva- rez, 1949); the Rio Mezquitic (= Rfo Balanos), tributary to the Rio Grande de Santiago and accessible near Valparaiso, Zacatecas; the Laguna de Sayula (about 1300 m), north of Sayula, Jalisco; and the Laguna de San Marcos (ca. 1300 m), near Zacoalco, Jalisco. K. integrum occurs in the latter three basins (Iverson, un- published), where it is the only turtle species known. CHARACTERS Nineteen shell measurements were made with dial calipers on museum speci- mens of the hirtipes species group from drainage basins discussed above. Only No. I Kinoslernon Biosyslematics 19 data from specimens over 80 mm cara- pace length (except three females) with the full complement of measurements were used in the morphometric analyses; vari- ous ratios of characters were also em- ployed to minimize ontogenetic variation (see STATISTICAL TECHNIQUES). Character means and ranges by popula- tion (Appendices 1 and 2) and taxon (Ap- pendices 3 and 4) for each sex are avail- able from NAPS'. Mensural characters recorded and their abbreviations follow [Methods of measurement were given by Iverson (1977a); midline plastral scute measure- ments were always made on the animal's right side.]: carapace length (CL), cara- pace width (CW), maximum plastral length (PL), plastral widths measured at the lateral edges of the seams between the humeral, pectoral, abdominal, femoral, and anal laminae (WA, WB, WC, and WD respectively), bridge length (BL), gular length (GL), gular width (GW), interhumeral seam length (IH), inter- pectoral seam (IP), interabdominal (lAB), interfemoral (IF), interanal (IAN), first vertebral width (VW) and length (VL), maximum length of plastral fore- lobe (FL), and maximum length of plas- tral hindlobe (HL). The ratios of each character to CL as well as the ratios of IH, IP, lAB, IF, and IAN to PL (total, 23 ratios) were employed in the analysis. In some analyses, the number of variables was reduced to those thirteen (excluding the ratios CW, HL, PWD, IH, IP, lAB, IF, IAN, VW and VL to CL) with the greatest variation in the species group. ,See NAPS document 03915 for 20 pages of supple- mentary material. Order form NAPS, c/o Micro- fiche Publications, P.O. Box 3513, Grand Central Station, New York, NY 10017, USA. Remit in advance for each NAPS accession number. Insti- tutions and organizations may use purchase orders when ordering; however, there is a billing charge of $5.00 for this service. Mack checks payable to Microfiche Publications. Photocopies are $5.00. Microfiche are $3.00 each. Outside the United States and Canada, postage is $3.00 for a photocopy and $1.00 for a fiche. Those analyses are noted in the text. Sexes (males have long tails and scale patches on the hind legs) were always analyzed sep- arately. Relative shell height has been used to distinguish Kinosternon hirtipes from K. sonoriense (e.g., Ernst and Barbour, 1972; Wermuth and Mertens, 1961); how- ever, the character is difficult to measure consistently and preliminary analysis revealed it would not reliably separate the two taxa. It has therefore not been used in this analysis. Qualitative characters also recorded in- cluded relative head size, plastral color, and shell carination as well as the follow- ing. Nasal scale. - As described by Conant and Berry (1978:3), adult kinosternids have a patch of cornified epithelium which extends from the dorsal margin of the rostrum for a variable distance poster- iorly on the dorsum of the head. A drawing of the shape and extent of the nasal scale on each individual turtle was made. Chin barbels. - The number, relative sizes, and locations of chin and neck barbels were recorded. Head pattern. - Although often quite variable and difficult to describe, an at- tempt was made to qualify head patterns. The procedure involved photographing the heads of as many specimens as possi- ble (over 500 total head photographs available) for later simultaneous exam- ination and comparison. STATISTICAL TECHNIQUES Character ratios were employed in the statistical analyses despite recent criticism of their use by Atchley et al. (1975, 1976). This decision is based on arguments in favor of their use by Corruccini (1977), Nussbaum (1976), Dodson (1978), Heyer (1978), and Iverson (1979), as well as the articulate demonstration by Berry (1978) that, for at least one other Kinosternon species group, the use of ratios as input variables in both multiple discriminant 20 Tulane Studies in Zoology and Botany Vol. 23 analysis and distance (D-) analysis pro- duced results almost identical to those obtained by using residual values from re- gression analysis as input variables (the standardization technique recommended by Atchley et al., 1976). My own unpub- lished data on other kinosternid species also support the effectiveness and reliabil- ity of multivariate analyses using char- acter ratios for at least this family of turtles. Furthermore, I attach no statist- ical significance to multivariate output generated from ratios. The output is only used as a tool to pinpoint distinctive samples, and to suggest the characters most important to those distinctions. Simple statistics were performed using the Statistical Analysis System (Service, 1972). Standard deviations accompany mean character ratios only as a relative measure of dispersion; no statistical sig- nificance is implied. Step-wise discimi- nant analyses (see Gould and Johnston, 1972, and Sneath and Sokal, 1973, for review of the procedure) utilized the Bio- medical Programs BMD07M (W.J. Dixon, 1973), and BMDP7M (W.J. Dixon, 1977). Cluster analyses employed the NT-SYS (Rohlf and Kispaugh, 1972) and BMDP2M programs. Specific appli- cations of these analyses are outlined un- der MORPHOMETRIC ANALYSIS. Because of the pronounced sexual di- morphism in members of the K. hirtipes species group, sexes are considered sep- arately in all cases. Turtles from basin samples represented by only one or two individuals of either sex were included in the analysis as unknowns, and assigned to the most phenetically similar sample by discriminant analysis. BIOSYSTEMATIC TACT Like many evolutionary biologists be- fore me, the problem of interpreting the genetic (and taxonomic) relationships of closely-related, allopatric vertebrate pop- ulations is a perplexing one (see Inger, 1961; Amadon, 1966, 1968; Mayr, 1970:210-211; Amadon and Short, 1976). The propensity (perhaps restriction) of members of the K. hirtipes group for permanent water habitats, coupled with the geographical isolation of inhabited river basins due to historical geology and desertification have produced at least thirty allopatric populations of members of this group. Many of these populations differ notably from geographically adjac- ent populations, but are quite similar to other populations far removed (see RESULTS). Interpreting such complex variational patterns is difficult. In this paper I have taken a conserva- tive approach to the species-subspecies dilemma. Within a species morphologi- cally distinct, isolated (i.e. allopatric) pop- ulations are afforded only subspecific status even though additional data (especially breeding information) may show that some are full biological species. The genetic relationships of the popula- tions so named are unfortunately clouded by this taxonomic designation [Amadon and Short, 1976, define "megasubspecies" and "allospecies" in an attempt to counteract this confusion]; however, complementary studies of protein varia- tion now in progress should perhaps further clarify the specific-subspecific (i.e. genetic) relationships of these turtle populations. LITERATURE Because far more than half the litera- ture records of kinosternid turtles in Mex- ico are in error I had to assume that every literature record was incorrect until per- sonally verified by examination of the respective specimens or by analysis of dis- tributional information (for example, in cases where only one kinosternid occurs in a particular basin). I have therefore attempted to substantiate every literature record for any kinosternid from through- out the range of the members of the K. hirtipes group (or stated as being from that range), and any member of the Kino- sternon hirtipes group (i.e. K. sonoriense or K. hirtipes). Complete chronological synonymies were then compiled for K. sonoriense and K. hirtipes, and each ref- No. 1 Kinosternon Biosystematics 21 erence was annotated to indicate the reas- ons for its inclusion. Copies of this annotated synonymy have been deposited in the Florida State Museum Herpetology Library (University of Florida), and are available from the author as well. Most of this information appears in this paper in the SYSTEMATICS Section, with some elaboration in the next section. RESULTS AND DISCUSSION Literature Corrections Because of the past difficulty in the discrimination of the K. scorpioides group (including K. integrum and K. alamosae) from the K. hirtipes group (including K. hirtipes and K. sonoriense), the literature on Mexican Kinosternon has accumulated so many errors of identifica- tion that it is almost unusable. The fol- lowing literature corrections (ordered by the valid taxon with which the K. hirtipes group member was confused) are an attempt to bring some order and accuracy to the error-plagued literature. Sternotherus odoratus. Apart from Meek's El Sauz, Chihua- hua specimen of Sternotherus odoratus (see discussions in Moll and Williams, 1963 and Conant and Berry, 1978:15), that species (or its nomenclatural equiva- lent) has been frequently, though errone- ously, recorded from Mexico. Duges (1869:143) was apparently the first to record "Ozotheca (odorata?)" from "Guanajuato y Mexico", but in a subse- quent list (1888:106) he apparently changed his identification to Cinosternon pensylvanicum. Because K. hirtipes occurs in both Guanajuato and Mexico states, because it is more similar to S. odoratus than is K. integrum (the only other turtle recorded from those two states), and because Duges was apparently not aware of Wagler's (1830) description of A", hirtipes from "Mexico" (the species is not on his 1869 list), Duges' Ozotheca record was almost certainly based on Kinosternon hirtipes. The following orthographic variations of Duges' record were apparently based on his 1869 list and are thus considered synonymous (in part) with K. hirtipes: Ozotheca odorata, Velasco (1890b:54, 1891:52, 1892b:40, 1893b:81, 1894:40, 1896a:30, 1898:62); Ozothea odorata, Velasco (1892a:76, 1892c:79, 1895:38, 1896b:37); and Ozho- teca odorata, Garcias-Cubas (1884:179) and Velasco (1890a:35, 1893a:64, 1897:41). In addition, Conant and Berry (1978:15) have clarified Brown's (1950:230) record of Sternotherus odoratus from Presidio Co., Texas; the specimen on which the record was based is TCWC 650, the holotype of Kinosternon murrayi Glass and Hartweg 195 1(= K. hirtipes murrayi). The last erroneous record is that of Altini (1942:159) for Kinosternon odoratum in Veracruz; based on his specimen description, it is appar- ently referable to K. leucostomum. Kinosternon subrubrum. Testudo pensilvanica Gmelin (1788: 1042) has been recognized as a synonym of Kinosternon subrubrum (Lacepede, 1788:132) at least since 1917 (see Iverson, 1977b). Prior to that time, however, the former specific name had an active his- tory in the Mexican herpetological litera- ture, despite the fact that the species does not range in Mexico. Lichtenstein (1856:2) was the first to apply the name (as Cinosternon pensylvanicum) to speci- mens from Mexico in the Berlin Museum. Over the next 50 years, no fewer than 22 papers recorded that species name (or orthographic variations thereoO for speci- mens from Mexico. Based on the greater similarity of T. pensilvanica (i.e. K. subrubrum) to K. hirtipes than to K. integrum, and the fact that most of these references are based on specimens from Guanajuato (where integrum and hirtipes co-occur) and/or the Valley of Mexico (where only hirtipes naturally occurs; see MATERIALS AND METHODS), the following binomials and references most probably refer to K. hirtipes: Cinosternon pennsylvanicum, Duges (1888:106; 1890, in Velasco 1890b:291; 1895:5; 1896a:lv; sylvanicum, Bocourt (1876:5), Herrera 22 Tulane Studies in Zoology and Botany Vol. 23 (1890:330; 1891:46; 1893:339; 1904:5), (1890:330; 1891:46; 1893:339; 1904:5), Herrera and Lope (1899:281), Westphal- Castelnau (1872:278), and Strauch (1890:88); Cinosternonus pensylvanicum, Herrera (1899:28; for discussion see H.M. Smith and R.B. Smith, 1975:86); Cino- sternum pennsilvanicum, Cope (1900:1229); Cinosternum pennsylvan- icum, Gadow (1905:209); Cynosternon pensylvanicum, Herrera and Lope (1899:131); Cynosternon pennsylvani- cum, Herrera (1893:342); and Kinostern- um pennsilvanicum, Cope (1896:1021). The failure of these person to recognize their specimens as K. hirtipes Wagler is probably a consequence of the lack of a nuchal scute by Wagler's only type spec- imen (see later). Unaware that a missing nuchal scute (actually worn away) is an uncommon, though natural anomaly, A.M.C. Dumeril and G. Bibron (1834:370), A.H.A. Dumeril (1870:25), Bocourt (1876:50) and Duges (1888:106) used the absence of that scute as the key character in identifying hirtipes. Several additional orthographic vari- ations were not, however, based on K. hirtipes. Gadow's (1905:194) record of Cinosternum pennsylvanicum from Guer- rero must be based on K. integrum if the datum is correct, because it is the only Kinosternon found there. Lampe's (1901:185) description of Cinosternum pensylvanicum from north Mexico makes it clear he is referring to Kinosternon subrubrum hippocrepis (probably from Texas). Siebenrock's (1905:465) erroneous record of Testudo pensylvanica from Veracruz is possibly based on a specimen of A', herrera i. Cinosternon hippocrepis (another synonym of K. subrubrum; see Iverson, 1977b) was erroneously recorded from Sonora by Strauch (1865:100, 184) pre- sumably based on a specimen of A^. sonor- iense. Kinosternon flavescens. Several K. flavescens records are in part based on members of the K. hirtipes species group. Most of these have been previously discussed (Iverson, 1978). In addition. Cooper (1870:66) recorded Platythyra flavescens from the Colorado River Valley along the California border (precise locality unknown). 1 have else- where (Iverson, 1978:477) questioned the existence of A', flavescens in the Colorado River basin and here suggest that Cooper's record was almost certainly based on K. sonoriense. Kinosternon scorpioides group. The true identity of the species of Kino- sternon occuring on Marfa Madre Island in the Tres Marfas Islands off the coast of Nayarit has plagued herpetologists. Gunther (1885:15) first recorded and fig- ured K. hirtipes from the island, but the same specimens were called K. integrum by Boulenger (1889:42). Both Strauch (1890:91) and Stejneger (1899:64) sup- ported Boulenger's view, yet Gadow (1905:209) advocated Gunther's original designation. Siebenrock (1906:96) was the next to support Boulenger's position. H.M. Smith and Taylor (1950a:25) avoided the problem by recording both species from the islands. Zweifel (1960:94) next addressed the problem in his study of the herpetofauna of the islands. In collaboration with Norman Hartweg, he finally corrected the record; K. integrum is the only species of the genus occurring in the Tres Marfas. Wermuth and Mertens (1961 :Fig. 13, p. 20) reproduced Gunther's (1885) figures and recommitted the latter's error. Casas Andreu (1967:44) likewise repeated the error, apparently following Smith and Taylor (1950a). Hardy and McDiarmid (1969:218) were next to discuss the problem and they supported Hartweg, Zweifel, and Boulenger's position. In what I hope is the final chapter in this prolonged story, I can only repeat and emphatically support Hartweg's opinion (in Zweifel 1960:95) that K. integrum is "the only species of the genus that gets to the Tres Marias." Kinosternon hirtipes group. Garman (1887:16) erroneously record- No. 1 Kinosternon Biosystematics 23 ed Cinosternum hirtipes from San Luis Potosi, Mexico. Taylor (1952:793) re- peated that record, listing it as "possibly doubttui". Ihe specimen on which Carman's record was based (MCZ 4545), from the mountains of Alvarez, is un- questionably AT. integrum. The occurrence of K. hirtipes in the state of San Luis PotosC has thus been verified at only one other locality (see MATERIALS AND METHODS). J.R. Dixon et al. (1972:228) recorded K. hirtipes from Cadereyta, Queretero on the basis of AMNH 71570. That speci- men, an articulated shell, is referrable to K. integrum; K. hirtipes does not occur in that part of Quere'taro, Liner (1964:221) recorded the deposi- tion in the Tulane collections of K. hirtipes he collected in Guanajuato (precise locality not published). TU 17563 (adult male) from that collection, from 2 mi. N. Ojo de Agua, is not K. hirtipes, but K. integrum. Four papers (Martin del Campo, 1937:265; Caballero y Caballero, 1938a: 103, 1938b:448; Casas Andreu, 1967:45) erroneously recorded K. hirtipes^ from Tasquillo, Hidalgo, lying in the Rio Tula basin. Because only K. integrum occurs in that basin, those records must pertain to that species. Similarly Caballero y Cabal- lero's (1940a:225) record of K.^ hirtipes from Uruapan, Michoacan (Rio Balsas basin) is based on K. integrum, since only the latter species occurs in that area. Altini (1942:154) recorded Kinosternon hirtipes from Lake Patzcuaro and Lake Chapala, Mexico, and Lake Pete'n, Cuat- emala; the species occurs in both of the Mexican lakes, but clearly does not occur in Guatemala. He also recorded K. leuco- stomum, K. cobanum (= K. acutum), and K. triliratum (= K. scorpioides) from Lake Pet en. All three of those species are known from the Pete'n region: Claudius angustatus and Staurotypus triporcatus. Which of the latter two species was mis- identified by Altinas K. hirtipes cannot be determined by the data available to me. In the same paper, Ahini also erroneously recorded K. leucostomum from Mexico's Lake Chapala; the species is not found there. Because only K. integrum and K. hirtipes occur in Lake Chapala and because Altini also recorded K. hirtipes from the lake (presumably correctly iden- tified), his K. leucostomum record is probably based on K. integrum. Clearly, an examination of Altini's specimens (presumably at the University of Bologna) will be necessary to rectify these misiden- tifications. Based on my discussions with the author, it is clear that Wiewandt's (1971:34; and Wiewandt et al., 1972:162) records of A^. sonoriense from Sonora, 3.5 miles W. Alamos, were based on speci- mens of the recently described K. alamosae (K. sonoriense does not occur there). Similarly, as explained by Berry and Legler (1980), Herenghi's (1969) Sonora K. hirtipes are also referable to K. alamosae. Morphometric Analysis An NTSYS cluster analysis was per- formed early in the study (1977) employ- ing population means for all 23 variables as OTU's (males and females separately). Two major phenetic groups were evident in both the male (Fig. 8) and female dis- tance phenograms. The first group in each analysis included the Casas Grandes, southwest New Mexico, Magdalena, Sonora, Yaqui, and Sonoyta samples (i.e., the populations of K. sonoriense as previously recognized; Iverson, 1976 and 1978), and the Aguascalientes sample in the male analysis and the Nazas sample in the female analysis. Of later significance is the fact that the Sonoyta sample was the most distinct of the sonoriense population in both the male and female analysed. The second group in each case included all populations from the Rio Santa Maria in Chihuahua south and eastward to sout- ceral Mexico (representing populations of K. hirtipes). In both analyses the main hirtipes cluster was divided phenetically into two subgroups; however, the included samples 24 Tulane Studies in Zoology and Botany Vol. 23 were different in each analysis. For both sexes, one subgroup included all of the northern-most K. hirtipes samples (Santa Marfa, Carmen, Sauz, and Texas); but for the males the subgroup also included sev- eral of the southern-most populations (Valley of Mexico, Villa Victoria, and Bajio) and for the females it included another northern population (Conchos), a central population (Mezquital), and two southern populations (Patzcuaro and Chapala). No other obvious morpho- geographic correlations or discontinuities were noted in these preliminary clusters, but many of the sample sizes on which the means were based were quite small. The level of differences between clusters and/or samples were generally higher in females than males, substantiating my subjective observation that there is less variation among females. The final male (Fig. 9) cluster analysis (BMDP2M) of population means for all 23 variables for all samples with N > 2 (Appendices 1 and 2) suggests that six groups were evident. In declining order of distinctiveness they are the 1) Viesca, 2) Sonoyta, 3) Villa Victoria, 4) K. Figure 8. Preliminary NT-SYS cluster (based on the distance matrix, with complete averaging and low values considered similar) of population means of all 23 variables for males of the K. hirtipes species group (including K. sonoriense). Abbreviations as in text. N > 3 for all samples but SLP and PATZ (N = 2 each). Figure 9. BMDP2M cluster of population means of all 23 character ratios (Appendices 1 and 2) for males of the K. hirtipes species group (including K. sonoriense). Abbreviations as in text. N > 4 for all samples but VSCA (N = 2) and TEX (N = 3). Numbers are amalgamation distances (i.e., distance between the clusters joined). sonoriense, except Sonoyta, plus Aguas- calientes, 5) Duero,and6) the remaining K. hirtipes samples. In the final female analysis, the nine most distinct groups are the 1) San Juanico (but N = only 2), 2) Sonoyta, 3) Viesca (no female Villa Victoria sample was included), 4) Balsas (N = only 2), 5) Zapotlan (N = 2), 6) Santiaguillo (N = 2), 7) Duero, 8) K. sonoriense except Sonoyta, plus Nazas and Verde, and 9) the remaining K. hirtipes samples. Again, the female anal- ysis differences were not at as low levels as the males'. Stepwise discriminant analysis of popu- lations with N > 2, based on all 23 character ratios, produced plots of popu- lation means on the first two (most im- portant) canonical axes for males and females (Figure 10). Two groups separate along the first canonical axis in both anal- yses: 1) the seven K. sonoriense samples and 2) the K. hirtipes samples. Within the cluster of K. hirtipes sample means, two patterns are apparent (especially along the No. 1 Kinosternon Biosystematics 25 second canonical axis). First, the popula- tions are arranged from northern-most (Santa Man'a; Carmen) to central (Mex- quital; Aguascalientes) to southern-most (Duero, Vajio, Patzcuaro, and Chapala); except the Valley of Mexico sample, which falls with the northern populations. Second, there appears to be a weak phenetic break in this clinal arrangement 1) in males, between populations north of and including the Nazas (plus the Valley of Mexico) and those south and east of that basin and 2) in females, (less distinctly), between populations north of and including the Conchos (plus the Valley of Mexico) and those south and east. Other morpho-geographical dis- continuities include the relative isolation of Duero, Patzcuaro, and Aguascalientes samples (and the proximity of the latter to the K. sonoriense samples) in the male plot, and the relative isolation of the Zapotlan (but N == 2), Valley of Mexico, and Sonoyta samples in the female plot. The character ratios most important for discrimination of the samples were deter- mined in the stepwise discriminant anal- ysis in two ways: 1) by the highest F values calculated for each variable before any were entered into the discriminant function and 2) by the order in which the variables were entered into the function. The first (most important) variable in each is always the same, but the remaining frequently are not, especially if character information is redundant in two or more variables. For males the five most im- portant variables were PWB/CL, PWA/CL, GW/CL, BL/CL, and IP/PL by F value, and PWB/CL, IAN/PL, IP/CL, and BL/CL by order of inclusion. For females, they were IP/PL, IP/CL, IF/CL, AN/PL, and IF/PL, and IP/PL, IH/CL, IF/CL, BL/CL, and FL/CL, respectively. The analysis reveals that K. sonoriense in general has smaller inter- pectoral seam lengths, larger interfemoral seam and gular scute lengths, and a wider plastron, gular scute, and bridge (see also Appendices 3 and 4). Bivariate plots (Figures 11-14) of the most important characters (by F value) illustrate that K. sonoriense is both geographically and morphologically disjunct. Because they are the most morphometrically distinct groups within the species group, because they are nowhere known to hybridize, because there is no evidence of introgres- sion (based on the morphometric calcula- tions) in the two most geographically proximate populations (Casas Grandes and Santa Maria), and because several other morphological characters (see later) also show a sharp phenetic break between OUfB **'• • •CHAT • 2»fO »&UN • VSCA* , »GUAS • »AO CNCH • '*^ • v»ilf SauZ It! • StM» CSG« CIMN $TmI SAU7 ca**N v«.. • •' C^«:M fcAJ c^ HA2% aOUAS • <^'»«,M*OC • M(Z • AGUN» ,;»»o Figure 10. Plots on the first (k,) and second (k,) canonical axes of the population means of males (above) and females (below) of the Kinosternon hiriipes species group (including K. sonoriense). Abbreviations as in text. Analysis based on all 23 character ratios for populations with N > 2. First two axes account for 30.4 and 20.6% of the total variation, respectively, in males; and 23.8 and 18.0%, respectively, in females. 26 Tulane Studies in Zoology and Botany Vol. 23 the two morphometric groups, K. sonoriense and K. hirtipes are considered valid species as previously defined (Iver- son, 1976, 1978; Wermuth and Mertens, 1977). Therefore, populations of each species were analyzed separately. Variation within K. sonoriense. The above analyses (see especially Figures 9, 10 and 12-14) suggest that the Sonoyta sample is the most distinct of the populations of K. sonoriense. Stepwise discriminant analyses of the seven sonoriense populations with N > 4 (Fig- ure 15), employing 13 variables (those indicated as the most important in the overall analyses) clearly support this suggestion. Those plots also suggest additional variational patterns. Although most of the non-Sonoyta samples are morpholog- ically very homogeneous (note cluster overlap in Figure 15), both the male and female Yaqui sample contain some apparently anomalous individuals. In the male plot, all but one Yaqui specimen lie within the main cluster. The outlier (BYU 39 J V,U \ NAZ * *SAU J < 36 < ^ ui SIMR *" 'Sez '"' .^ Figure 11. Graph of percent posterior width of plastral forelobe/carapace length (PWB/CL) versus percent plastral width at humero-pectoral seam/carapace length (PWA/CL) for males of all populations of the Kinosternon hirtipes species group (including Kinosternon sonoriense). Only population means are plotted. Abbreviations as in text. Figure 12. Graph of percent posterior width of plastral forelobe/carapace length (PWB/CL) versus percent gular width/carapace length (GW/CL) for males of all populations of the Kinosternon hirtipes species group (including Kinosternon sonoriense). Only population means are plotted. Abbreviations as in text. 14629) is the only male available from the entire Ri'o Bavispe portion of the Yaqui basin and the only specimen available from that locality. The significance of its apparent distinctiveness must await the collection of additional material. The most important characters in the discrim- ination are IAN/PL, GW/CL, IF/PL, PL/CL and GL/CL by F-values, and IAN/PL, GW/CL, GL/CL, IP/PL, and PWC/CL by order of inclusion. In the female plot all but one of the Yaqui specimens lie outside the main cluster, and separate fairly well along the second canonical axis. Only a specimen from the Rio Gavilan (MVZ 46646), a tributary of the Rfo Bavispe, falls within the main cluster. Although the recorded locality is clearly in the Gavilan-Bavispe drainage (see map in Marshall, 1957), the possibility exists that the specimen actually originated in the Rio Casas Grandes basin; the two drainages inter- digitate near the locality. The outlying cluster of female Yaqui specimens is problematical, especially since it includes specimens from the same localities that clustered in the main group in the male No. 1 Kinosternon Biosystematics 27 VMLI CliNN •CMW iim IWHM Til ^ITOO • SAUl • AOUU MI «CIUCi • •CNCM lAJ CMM ,,yyj,f »l» • N» ,, Vu SNBA AOUH ,«. no> • ^«T. Figure 13. Graph of percent interanal seam length/carapace length (lAN/CL) versus percent interpectoral seam length/carapace length (IP/CL) for females of all populations of the Kinosternon hirtipes species group (including Kinosternon sonoriense). Only population means are plotted. Abbreviations as in text. analysis. Only additional field work in the Yaqui basin can clarify these variational anomalies. The most important charac- ters in the discrimination are IF/PL, lAN/CL, PL/CL, IH/PL, and PWB/CL by F-values and IF/PL, PL/CL, IH/PL, PW/CL, and IP/PL by order of inclu- sion. Two morphotypes are thus recogniz- able within K. sonoriense, represented by Sonoyta basin turtles and the other samples, respectively. The Sonoyta basin turtles have shorter plastra, longer inter- femoral seams, shorter interanal seams, wider first vertebrals, and narrower gulars, than the other populations. Variation within K. hirtipes Cluster analysis (BMDP2M) of popu- lation means for all 23 variables, suggest that three main groups exist for males and females (Fig. 16). For males, the most dis- tinct is the Viesca sample (but N = 2), followed by the Villa Victoria sample and all remaining hirtipes populations. Within the last group, the only relationship be- tween phenetics and geography is a small SNIA* JN«»^ »Ao' MAGO, •AGON .NAZ «MT2 'STGO SWNM* *CMCM uuz SNJ. STMI •stf CHAT .CtMN •miii Figure 14. Graph of percent interpectoral seam length/carapace length (IP/CL) versus percent interfemoral seam length/carapace length (IF/CL) for females of all populations of the Kinosternon hirtipes species group (including Kinosternon sonoriense). Only population means are plotted. Abbreviations as in text. subgroup including the Patzcuaro, San Juanico, Chapala, and Duero samples. Contrary to the results of the discriminant analysis which included K. sonoriense (Fig. 10), no clinal relationships or phenetic breaks between northern and southern populations are evident. For females, the most distinct group is the San Juanico sample (but N = 2), followed by the Viesca sample and the remaining populations. No obvious phenetic- geographic relationships are suggested within the last group. However, Villa Victoria was not represented in the female analysis since only one female specimen is known from that basin. Raw data for that specimen (Appendix 2) suggest it might be as distinct as the male sample (see later). Stepwise discriminant analysis of K. hirtipes populations with N i 2, based on all 23 character ratios, produced the canonical plots in Figure 17. The male analysis reveals the same north-south clinal pattern along the first axis, the same phenetic break between the Nazas and Aguanaval basins, and the same similarity between the Valley of Mexico population and northern populations that the earlier 28 Tulane Studies in Zoology and Botany Vol. 23 overall discriminant analysis illustrated (Fig. 10). It further suggests the distinct- iveness of the Villa Victoria population and possibly also the Patzcuaro sample. The female analysis again reveals the general north-south clinal pattern along the first axis (ahhough no Nazas- Aguanaval break is evident), and also suggests the distinctiveness of the Viesca, Zapotlan, San Juanico and possibly Duero populations. Figure 15. Plots on the first (k,) and sec ond (k2) canonical axes of population means (solid dots) of males (above) and females (below) of K. sonoriense. Lines connect the most dispersed values about the population mean. Population mean symbols are 1, SNTA; 2, SWNM; 3, GILA and WILL; 4, MAGD; 5, SNRA; 6, CSGR; and 7, YAQ. Individual Yaqui turtles are marked with Y (Yaqui River proper) or B (Bavispe River). Analysis based on 13 character ratios. First two axes account for 71 .0 and 10.2% of the variation, respectively, in males; and 47.9 and 23.1%, re- pectively, in females. The most important characters in the male discriminant analysis are BL/CL. GL/CL, PWC/CL, IP/CL, and IAN/PL, based on F-values and BL/CL, GL/CL, PWC/CL, IAN/PL, and PWD/CL based on order of inclusion. For the females, the important characters are IP/CL, BL/CL, IP/PL, PWB/CL, and FL/CL, and IP/CL, BL/CL, IH/CL, FL/CL, and PWB/CL, respectively. Bivariate plots of the most important characters (Figures 18-21; see also Figures 11-14) do not suggest that a significant phenetic break exists between northern and remaining populations. However, as in the previous analyses, they again indicate the distinct- iveness of several samples, including the Valley of Mexico, Viesca, San Juanico, Patzcuaro samples, and possibly also a group of three geographically adjacent samples occupying the ancestral Lake Chapala basin (Tamayo, 1964:108): Chapala, Zapotlan and Duero (see Figs. 18 and 21). MORPHOMETRIC CONCLUSIONS The numerous analyses have strongly suggested that 1) K. hirtipes and K. sono- riense are distinct morphometrically, 2) the Sonoyta sample within K. sonoriense is morphometrically distinct, 3) there is tremendous variation within K. hirtipes, and 4) the most morphometrically distinct populations of hirtipes are the San Juan- ico, Viesca, Patzcuaro, Villa Victoria, and Valley of Mexico samples and possibly also the combined Chapala-Za- potlan-Duero samples. As detailed above, the basic patterns of morphometric variation in the K. hirtipes species group involve several character complexes, the most important of which are 1) body size (see later), 2) relative plas- tron size (measured primarily as PL, PWA, PWB, PWC, and/or PWD), 3) relative bridge length, 4) the relative lengths of the gular, interhumeral seam, and interpectoral seam to the forelobe length (the forelobe length itself is not as important), and 5) (of much less impor- tance) the relative lengths of the inter- No. 1 Kinoslcrnon Biosyslcmatics 29 femoral and interanal seams to the hind- lobe length (the hindlobe length is also not as important). Other characters clearly are important in individual population comparisons, but these complexes are ap- parently the most important when consid- ering variation in the group as a whole. Variation in relative plastron size is much greater in males than females. Females appear to be much more conser- vative regarding plastral reduction or modification. For relative male plastron size there is somewhat of a continuum from the relatively large plastron of K. sonoriense (Fig. 22b) to the small plastra of San Juanico and Viesca populations (Fig. 22, 1 and m) with the remaining populations somewhat intermediate. For females, the range is from the medium- size plastra of the San Juanico and Viesca populations (Fig. 22n) to the relatively ex- tensive plastra of the remaining popula- tions. Plastral reduction, typically cor- related with an increase in aquatic habits among turtles (Zangerl, 1939:386; Berry, 1977; Iverson, MSI) and presumably an adaption thereto (Iverson, MSI), is con- sidered derived. Relative bridge length is extremely vari- able in the genus Kinosternon. Males vir- tually always have shorter bridges than fe- males. For males, bridge length ranges from short in San Juanico, Pa'tzcuaro, Valley of Mexico, and Viesca turtles to medium length in the other populations. For females, it ranges from medium length in San Juanico, Patzcuaro, and Valley of Mexico turtles to long (but not as long as some members of the K. scor- pioides group) in the remaining popula- tions. Its reduction is not necessarily cor- related with plastral reduction [for exam- ple, Patzcuaro turtles have medium (male) or large (female) plastra and rela- tively short bridges]. I consider relatively reduced bridge length in males or females a derived character, both because of its rarity in this species group and because many of the most specialized members of the genus have short bridges. The components of the plastral fore- lobe are quite variable in this species group, but because the interhumeral seam length is essentially of medium length in all samples (except perhaps in the Villa Victoria basin), variational patterns are ^ •-c -c STMR CRMN VALLE PAP CHAP DUER PATZ AGUN CNCH AGUAS SLP MEZ VERD NAZ ZAPO SAUZ VSCA SNJ I I I I I »— I 1 I I ) O O >OCJ CJ NJ ~- Figure 16. BMDP2M cluster of population means of all 23 character ratios (Appendices 1 and 2) for male (top) and female K. hirtipes. Abbreviations as in text. For males, N > 4 for all samples except VSCA (N = 2); for females. N > 5, except SNJ (N = 2) and ZAPO (N = 2). Numbers are amalgamation distances (i.e., distance between the clusters joined). 30 Tulane Studies in Zoology and Botany Vol. 23 dominated by the relative lengths of the gular and the interpectoral seam. Patzcuaro, Viesca, and San Juanico tur- tles have very short gulars (Fig. 22, i-n), whereas the remaining populations have medium to long gulars {K. sonoriense having the longest). Patzcuaro and San Juanico turtles (and possibly Villa Vic- toria) have the longest interpectoral seams as well (Fig. 22, i-1); K. sonoriense, the shortest; and the remaining populations have intermediate lengths. Since most Figure 17. Plots on the first (k,) and second (kj) canonical axes of the population means (solid dots) of males (above) and females (below) of Kinosternon hirtipes (excluding K. sonoriense). Abbreviations -as in text for females; but further shortened for males. Polygons in male plot enclose total dispersion of each population. Analysis based on all 23 character ratios for populations with N > 2. First two axes account for 33.7 and 16.0% of the total variation, re- spectively, in males; and 27.4 and 14.5%, respectively, in females. Kinosternon have interpectoral seams of medium length (frequently used to diag- nose the genus; e.g., Conant, 1975), deviations from that condition are consid- ered derived. For similar reasons, the con- ditions of reduced and extensive gular scutes are considered derived. The plastral hindlobe components do not show as much variation as the other complexes, but a few patterns are evident. Hindlobe length is greater in K. sono- riense and Villa Victoria turtles than in re- maining populations. The interfemoral seam is relatively shorter in the Valley of Mexico sample (Fig. 22, e-f) than in all other samples; and the interanal seam is relatively longer in that sample and the Chapala-Zapotlan combined sample, and shorter in the Viesca sample. Variation in these characters within K. sonoriense (i.e., shorter interanal and longer inter- femoral seams in Sonoyta turtles) has already been discussed. Each of these deviations from the modal hindlobe con- dition found in the group are considered derived. Other Morphological Characters Nasal scale. — The cornified epidermal shield (here called the nasal scale) found Figure 18. Graph of percent gular length/carapace length (GL/CL) versus percent bridge length/carapace length (BL/CL) for males of populations of Kinosternon hirtipes. Only popu- lation means are plotted. Abbreviations as in text. No. Kinosternon Biosystematics 31 • • DUEH ■ CH«P «UN« ilu; 4 Mf; •s,. ^ '•CNCM ZAPO •^ ., •p.i/ VSCA* Figure 19. Graph of percent interpectoral seam length/carapace length (IP/CL) versus percent bridge length/carapace length (BL/CL) for males of populations of Kinosternon hirtipes. Only population means are plotted. Abbreviations as in text. on the anterior dorsum of the head of all subadult and adult kinosternid turtles has received little attention (but see Sieben- rock, 1907) until recently (Conant and Berry, 1978). Cornification of the scale begins near the rostrum in the juveniles of these turtles and spreads posteriorly with age. Development is usually more rapid laterally than medially, such that older juveniles or subadults may have V-shaped scales even though adults might have tri- angular to rhomboidal scales. Due to its distinctive shape in the adults of kino- sternid turtle populations, it is an impor- tant taxonomic character. Intraspecific variation in nasal scale shape has already been noted in Kinosternon subrubrum. Ernst et al. (1974) have shown that the subspecies Kinosternon subrubrum stein- dachneri usually (81*7o) has a head scale that is furcate posteriorly (their "divided nasal") whereas turtles of both other sub- species (hippocrepis and subrubrum) usu- ally (90 to 97%) have a non-furcate, bell- shaped scale (posterior margin convex). Variation within the K. hirtipes species group is considerable but intrapopulation variation is insignificant. Three distinctive nasal scale shapes are evident in inter- population comparisons of adults of this group. Populations from the Rios Casas Grandes and Yaqui (excluding the Papi- gochic) and those westward and north- ward (i.e., populations of K. sonoriense as previously recognized; Iverson, 1976), and the population from the Valley of DUER CHAP* ■ •aGun ■ VSCA^ Zapo J ^^^ .C.MN • --• *"" I -'. Figure 20. Graph of percent interpectoral seam length/carapace length (IP/CL) versus percent bridge length/carapace length (BL/CL) for females of populations of Kinosternon hirlip s. Only populations means are plotted. Abbrevi- ations as in text. Figure 21. Graph of percent gular length/carapace length (GL/CL) versus percent bridge length/carapace length (BL/CL) for females of populations of Kinosternon hirtipes. Only population means are plotted. Abbreviations as in text. 32 Tulane Studies in Zoology and Botany Vol. 23 No. Kinosternon Biosystematics 33 Figure 22. Plastral variation in members of the Kinosternon hirtipes species group: Gila River basin Kinosternon sonoriense, JBI 563 female (a) and JBI 386 male (b); Rio Papigochic basin K. hirtipes murrayi, UF 40391 female (c) and UF 40396 male (d); Valley of Mexico A', h. hirtipes, UMMZ 99458 female (e) and UMMZ 80357 male (0; Lake Chapala basin K. h. chapalaense, UMMZ 97123 female (g) and UMMZ 97128 male holotype (h); Lake Patzcuaro K. h. tarascense, UF 43505 female (i and j; illustrating plastral scute staining and loss of stain with scute shedding) and UF 43506 male (k); Presa San Juanico A', h. magdalense. UF 45035 male holotype (1); Viesca A', h. megacephahim SM 1 1464 female (m) and SM 9823 male (n). 34 Tulane Studies in Zoology and Botany Vol.23 Mexico exhibit a triangular, rhomboidal or bell-shaped nasal scale as adults (Fig. 23: a and d). Turtles from the Zapotlan, Lake Chapala, and Rfo Duero basins pos- sess a nasal scale that typically is crescent- shaped and extremely reduced in size. It nearly always lies completely anterior to the orbits in turtles from the former two basins (Fig. 23: c), but may reach to mid- orbit in Ri'o Duero turtles (Fig. 23: 0- All remaining populations of this group have a nasal scale deeply furcate behind (Fig. 23: b and e; but see Synthesis). Because most of the members of the Figure 23. Nasal scale variation in members of the Kinosternon hiriipes species group: A", sonoriense, JBI 697 (a); A', hiriipes murrayi, UF 40396 (b); A', h. chapalaense, UMMZ 97130 paratype (c); A', h. hiriipes, UMMZ 99449 (d); A. h. larascense, UF 43505 paratype (e); A'. //. chapalaense x A', h. murravi, UMMZ 97135 (f). No. 1 Kinosicrnon Biosystemalics 35 genus Kinosternon have triangular or bell- shaped nasal scales as adults (A', dunni, K. angustipons, K. acutum, K. baurii, most A', subrubrum, and all members of the A. scorpioides and A. leucosiomum species groups), and because 1 believe that the furcate condition in A. subrubrum slein- dachneri is derived from the bell-shaped condition found in a A. subrubrum subru- brum-WkQ ancestor (see also Ernst et al, 1974), 1 consider the large nonfurcate shape to be the primitive adult condition in the genus Knujsternon. Therefore, the condition in A. sonuriense and Valley of Mexico A. luriipes is considered primi- tive, whereas the nasal scale reduction in remaining populations of A. hirtipes is considered derived. Chin Barbels. — Variable numbers of barbels are present on the chin and/or gular region of all kinosternid turtles; however, two basic patterns exist in the K. hirtipes species group. In the first, char- acteristic of all populations of K. sono- riense as previously defined (Iverson, 1976, 1978), 3 or 4 pairs of barbels are present and the largest 2 pairs are sub- equal and relatively long ( < one half orbit diameter) with one pair mentally lo- cated and the other at the level of mid- tympanum. Populations of K. hirtipes are charac- terized by the presence of at least two pairs of chin barbels, the largest two pairs both located on the chin with the anterior pair decidedly the largest (yet P half orbit diameter). Because the typical Kinoster- non condition is one with two mental pairs of barbels, the condition in K. sono- riense is considered derived and that of AT. hirtipes ancestral. Head color. — Head patterns in this group are extremely variable, even within populations. Patterns vary nearly con- tinuously from broadly mottled (common in K. sonoriense; Fig. 24b; Conant and Berry, 1978, Fig. 7) to medium or finely reticulated (as in most populations of K. hirtipes from Chihuahua to Mexico City; Fig. 24: e, f, g, and n; Conant and Berry, 1978, Fig. 7) to finely spotted (common in female Patzcuaro A. hirtipes; Fig. 24: h). Whatever the general pattern, however, the lateral markings are more or less lon- gitudinally oriented, such that two yellow, cream, or white lateral stripes (one ex- tending across the temporal region, poste- riorly from the posterodorsal margin of the orbit; and the other extending poste- riorly from the posteroventral margin of the orbit, along the posterodorsal margin of the maxillary sheath to the angle of the jaw) are vaguely to very well developed. The more ventral of those stripes is almost always apparent, no matter how finely reticulated or spotted the pattern, or me- lanistic the head coloration. Most of this general range of pattern variation may occur in a single population; however, females usually have less dark pigment on the head, have finer mottling or reticula- tions, and are more likely to be spotted (compare Fig. 24: a versus b or e versus f; see also Conant and Berry, 1978, Fig. 7). The jaw sheaths are also variably pigmented, but in general the more dark pigment on the head, the more darkly pig- mented (streaked) are the jaw sheaths. The only two significant deviations (considered derived conditions) from this general (considered primitive) color scheme are in K. hirtipes from the Lake Chapala and Zapotlan basins and the Val- ley of Mexico basin. Turtles from the lat- ter basin have typical amounts of dark pigment but most specimens have both light lateral head stripes very well-defined (Fig. 24: c and d). In the former two basins, the dark markings are generally broad, but the overall amount of dark pigment is significantly reduced (compare Figure 24b versus i-1); in other popula- tions broadness of marking is correlated with abundance of dark pigment. In addi- tion, in Chapala and Zapotlan turtles, the lateral temporal head stripe is typically bordered ventrally by a broad dark stripe and the ventral stripe is bordered dorsally by a similar dark stripe. The general appearance is one of two dark stripes rather than two light ones (Fig. 24: i-1). Although their nasal scales are similar 36 Tulane Studies in Zoology and Botany Vol. 23 No. 1 Kinosternon Biosystematics 37 K. - * '' ^^ Y*^ ' ''*"; W Figure 24. Head pattern variation in members of the Kinosternon hirtipes species group: Gila River Basin Kinosternon sononense, JBl 563 female (a) and JBI 387 male (b); Valley of Mexico, A', hirtipes hirtipes, UMMZ 99458 female (c) and UMMZ 80357 male (d); Rio Papigochic A', h. murrayi. UF 40391 female (e) and UF 40395 male (0; Patzcuaro basin A. h. tarastense, UF 43596 female paratype (g) and AMNH 82628 female (h); Lake Chapala basin, A. h. chapalaense, UMMZ 97128 male holotype (i), UMMZ 97127 male paratype (j), UMMZ 97123 female paratype (k), and UMMZ 97130 male paratype (1); Viesca, Coahuila K. h. megacephatum, SM 11462 female paratype (m,n). 38 Tulane Studies in Zoology and Botany Vol. 23 to Chapala-Zapotlan turtles, specimens from the Rfo Duero have much darker head pigment as in more northerly and easterly populations. Head size. — Only one population de- viates from the typical (clearly primitive) condition. Turtles from the Viesca basin have distinctly enlarged heads (especially females) with greatly expanded alveolar surfaces (Fig. 24: m-n). Plastral staining. — Although the plas- tron of members of the K. hirtipes group is typically cream, yellow, or yellow orange, it may be variably stained dark brown to black. The turtles from Lake Patzcuaro have plastra consistently (and uniquely?) stained red-brown. At least in that population the color is due to envi- ronmental staining since the color is shed with the scute (Fig. 22: i-j). Shell carination. — Due to sexual dif- ferences and considerable ontogenetic change, quantification of variation in the development of keels in members of this group is difficult. In general, adult K. sonoriense are much more obviously tri- carinate than K. hirtipes. The latter species often appears almost unicarinate, the former, very rarely. The dorsum of the shell thus has a flatter appearance in K. sonoriense than in K. hirtipes. Body size. — Average carapace lengths of males and females in Appendix 1 and 2 reveal that females average larger than males in populations of K. sonoriense, whereas males average larger in most pop- ulations of K. hirtipes. There is also con- siderable variation in body size among populations of K. hirtipes. The most ob- vious deviations from the mode are in tur- tles from the Viesca and San Juanico basins. Although these basins both have a small sample size, I believe the data truly reflect the small size of the inhabitant tur- tles. Patzcuaro turtles also tend to be smaller than the mode, although not so distinctly. A more complete analysis of population and sexual variation in body size in K. hirtipes is in preparation. Nuchal-neural bone contact. — Berry and Legler (1980:11) report that the nuchal bone contacts the first neural bone in 73% of the K. sonoriense and only 4% of the K. hirtipes they examined (sample sizes not reported). However, only 38.1<^o of the K. sonoriense (N = 42) and 10.2% of the K. hirtipes (N = 98) I examined have nuchal-neural contact. The discrepancy between our results for K. sonoriense is possibly due to their smaller sample size, but the character is obviously of only limited value in distin- guishing the two species. Synthesis Of Character Variation Analysis of geographic variation in morphological characters in the Kinoster- non hirtipes species complex supports the recognition of two allopatric species, both previously recognized (Wermuth and Mertens, 1977; among many others): K. sonoriense and K. hirtipes. Analysis of populations of K. sonoriense suggests the existence of two distinct morphotypes, represented by 1) the population inhabit- ing the Rio Sonoyta basin and 2) the remaining populations previously recog- nized as K. sonoriense (Fig. 3). Stepwise discriminant analysis of those two sam- ples using only 13 morphometric charac- ters is capable of distinguishing 100% of the males and 98.6% of the females. Be- cause the holotype oi K. sonoriense was collected in the Gila River basin (Iverson, 1976) the Rfo Sonoyta population is here described as a new subspecies. Patterns of geographic variation in morphology within Kinosternon hirtipes suggest the existence of several undes- cribed taxa (Fig. 4). The most distinct morphological samples in this highly vari- able species are the 1) Viesca, 2) San Juan- ico, 3) Pa'tzcuaro, 4) Valley of Mexico, 5) Chapala-Zapotlan, 6) possibly the Duero sample, 7) possibly the Villa Victoria pop- ulation (see below) and 8) the remaining populations of K. hirtipes. Stepwise dis- criminant analysis of the seven samples, excluding the Villa Victoria population (see below), using all 23 morphometric variables, was able to classify turtles into the correct morphotype at least 75% of No. 1 Kinosternon Biosystematics 39 the time for any given morphotype of either sex. San Juanico and Viesca turtles were always classified correctly, and only one Valley of Mexico turtle was misclassi- fied (a female, into sample 8, above). A single male and one female from Patz- cuaro were misclassified (into San Juan- ico, in both cases). Two males and one female Duero turtle were misclassified into the Chapala-Zapotlan sample; and two other male Duero turtles were mis- classified in the Patzcuaro sample. Chapala-Zapotlan turtle misclassification included three males and one female into the Patzcuaro, and one female into the Viesca sample. The large and highly vari- able sample of the remaining K. hirtipes populations included the following mis- classifications: 12 males and 13 females into the Chapala-Zapotlan sample; 7 males and 6 females into the Valley of Mexico sample; 9 males and two females into the Duero sample; seven females into the Patzcuaro sample and three females into the Viesca sample. Based on the various morphological analyses, I conclude that at least the fol- lowing samples should be recognized tax- onomically: 1) Viesca, 2) San Juanico, 3) Patzcuaro, 4) Chapala-Zapotlan, 5) Val- ley of Mexico, and 6) the remaining popu- lations of K. hirtipes (perhaps excluding the Villa Victoria sample). I tentatively consider the Duero population as inter- grading between samples 4 and 6. The first four samples have not been named and are described here. The holotype of Kinosternon hirtipes murrayi clearly be- longs in the last group and hence that group should bear that trinomen. The sta- tus of the Valley of Mexico and Villa Vic- toria samples are not as clear. Several of the early analyses (see Fig- ures 9, 11, 16, and 17) suggested that the male Villa Victoria sample was morpho- metrically distinct. Unfortunately, only one female is known from that basin, and although not as distinct (see Appendix 2), it does exhibit some of the characters which seem to distinguish the males (long- er hindlobe, shorter interhumeral seam, longer interpectoral seam, and longer first vertebral scute). However, the complete lack of geographically proximate com- parative material from the entire upper Lerma system and the near lack of mate- rial from the Balsas drainage system (one female from 45 miles to the west) make a decision regarding the distinctiveness of this population difficult. I have therefore tentatively included the population with those of K. h. murrayi until field work in the upper Balsas and Lerma basins can clarify distribution and morphological variation in those regions. Even less clear is the correct allocation of the holotype o{ Kinosternon hirtipes, a very old, worn male specimen, with no associated data except "Mexico" (see SYSTEMATIC ACCOUNTS). Plastral erosion makes clear morphometric alloca- tion impossible (Fig. 25). In addition, the shape of the nasal scale (Fig. 25) is some- what intermediate between a V-shaped condition of A^. h. murrayi and the rhom- boidal condition of Valley of Mexico tur- tles. Schmidt (1953) restricted the type- locality to "lakes near Mexico City" (in the Valley of Mexico) but without varia- tional analyses. Because the correct al- location can only be solved by field work in the Valley of Mexico and adjacent ba- sins, I tentatively follow Schmidt (1953) in the application of the name K. h. hirtipes to the populations in that Valley. Systematic Accounts A chronological list of all synonyms and orthographic variations thereof is given for each valid taxon. Each ortho- graphic combination appears with refer- ence to its first use, including author, date and pagination (referenced in Literature Cited). Selected subsequent usages, espe- cially those incorrect or of taxonomic sig- nificance, and including all pre- 1930 ref- erences, are included in species and sub- species synonymies. Most citations are al- so parenthetically annotated. Localities and location of all available specimens are also included. 40 Tulane Studies in Zoology and Botany Vol. 23 Figure 25. Nasal scale shape (top) and plastron shape (bottom) in holotype of Kinosternon hinipes (ZSM 1374/0). Kinosternon sonoriense LeConte Sonora Mud Turtle Kinosternum sonoriense LeConte, 1854: 184 [type-locality, "Tucson, in Sonora", Arizona. Type, a male, col- lected by Dr. J. L. LeConte (author's son) and placed in Philadelphia Acad- emy of Sciences; presently unlocat- able]; Troschel, 1855:415. Kinosternon sonoriense Gray, 1855:79 (first use of this combination; Tucson); Stejneger, 1902:149 (Fort Huachuca and Babacomari creek, Arizona); Ruth- ven, 1907:594 (Sabino Canyon, Santa Catalina Mountains, Arizona); Mearns, 1907:117; Van Denburgh and Slevin, 1913:396 (Gila River and its tributaries; 8 specific localities); Grin- nell and Camp, 1917:200 (in part; result of incorrect synonymy of K. flavescens with A^. sonoriense; lower Colorado River, California); Stejneger and Bar- bour, 1917:112 (in part; southern New Mexico and Arizona into southeastern California; northern Mexico); Schmidt, 1922:618; Van Denburgh, 1922:967 (Arizona, 18 localities; California, 2 localities; and Sonora, 5 localities); Pratt, 1923:238 (in part; western Texas [= K. hirtipes] into southern Califor- nia); Van Denburgh, 1924:229 (New Mexico; "Fort Union" locality in er- ror, see Iverson, 1978); Strecker and Williams, 1927:15 (in part; Bexar Co., Texa§ locality based on K. flavescens); Storer, 1930:430; Ditmars, 1936:397 (in part; southwestern Texas records based on K. flavescens); Dunn, 1936:472 (in part; Chihuahua locality based on K. hirtipes); Pickwell, 1947:60 (in part; southwestern Texas record based on K. flavescens); Brown, 1950:228 (in part; Texas localities based on K. flavescens); H. M. Smith and Taylor, 1950a:26 (in part; western Texas localities based on K. flavescens; Chihuahua and Durango localities based on K. hirtipes); Carr, 1952:90 (in part; Texas records based on K. flavescens); Schmidt, 1953:91 (in part; Texas records based on K. flavescens; erroneously restricted type- locality of the synonym Kinosternum henrici to Las Cruces, New Mexico); Mertens and Wermuth, 1955:336 (in part; Texas records based on K. flaves- cens; Chihuahua and Durango records, on K. hirtipes); Cagle, in Blair et al., 1957:281 (in part; Texas records based on K. flavescens); Gijzen and Wermuth, 1958:44 (in part; photograph apparently K. integrum); Wermuth and Mertens, 1961:27 (in part; Texas rec- ords based on K. flavescens ; Chihua- hua and Durango records based on K. hirtipes); Casas Andreu, 1965:382 (in part; Chihuahua and Durango records based on A', hirtipes); Stebbins, 1966:82 (in part; Texas records based on K. flavescens; Durango records based on K. hirtipes); Casas Andreu 1967:51 (in No. 1 Kinosternon Biosystematics 41 part; Chihuahua and Durango records based on K. hirtipes); Pritchard, 1967: 37 (in part; Coahuila records incorrect; Texas records based on K. flavescens); Cochran and Coin, 1970:136 (in part; Texas records based on K. flavescens); Legler and Webb, 1970:163 (in part; Chihuahua records based on K. hirtipes); Wiewandt, 1971:34 (in part; southern Sonora records based on K. alamosae); Wiewandt, Lowe and Larson, 1972:162 (in part; southern Sonora records based on K. alamosae); Ernst and Barbour, 1972:64 (in part; Chihuahua and Durango records based on A', hirtipes); Hambrick, 1976:291 (in part; Texas records invalid); Iverson, 1976:1 (in part; upper Ri'o Yaqui rec- ords in Chihuahua based on K. hirtipes); Wermuth and Mertens, 1977: 10; Conant and Berry, 1978:1; Iverson, 1978:476; H. M. Smith and R. B. Smith, 1980:156; Berry and Legler, 1980:1. Thyrosternum sonoriense Agassiz, 1857: 428; Blair, 1859:3, Troschel, 1860:270; Carman, 1885:8. Cinosternum sonoriense Agassiz, 1857: Plate V, fig. 8-11; Cope, 1875:52, Coues, 1875:589; Yarrow, 1883:31; Gunther, 1885:13; Boulenger, 1889:40; Siebenrock, 1907:551; Siebenrock, 1909:444. Kinosternum henrici LeConte, 1859:4 (type-locality, "New Mexico"). Type, a male, collected by Dr. T.C. Henry and placed in Philadelphia Academy of Sciences (ANSP 83). Locality data with type is "Gila River, New Mexico." Type-locality erroneously restricted by Schmidt (1953:91) to vicinity of Las Cruces; Cope, 1880:13 (in part; Texas record based on K. flavescens). Thyrosternum henrici Troschel, 1860: 270; Carman, 1884:8. Cinosternon henrici Strauch, 1862:41; Strauch, 1865:101; Strauch, 1890:89 (in part; Dallas, Texas record based on K. subrubrum). Cinosternon sonoriense Strauch, 1862:41; Strauch, 1865:100. Thylosternum sonoriense Muller,1865: 598. Kinosternon punctatum Gray, 1870:67 (in part; Sonora; eastern United States rec- ords based on K. subrubrum) . Swanka henricii Gray, 1870:69. Platythyra flavescens Cooper, 1870:66 (possibly in part; see Iverson, 1978; Colorado River Valley). Cinosternum henrici Cope, 1875:52; Yarrow, 1875:583; Coues, 1875:590; Yarrow, 1883:31; Boulenger, 1889:40; Ditmars 1907:26; Strecker, 1915:10 (in part; Texas records based on K. flaves- cens); Malnate, 1971:353. Aromochelys carinatus Yarrow, 1875:582 (in part; Arizona); Coues, 1875:589 (in part; Arizona). Cinosternum flavescens Yarrow, 1883:31 (in part; "Utah" and "Fort Mora", specimens actually K. sonoriense, see Iverson, 1978). Cinosternum hirtipes Gunther, 1885:15 (in part; result of his synonymy of K. henrici LeConte with K. hirtipes Wag- ler); Cope, 1887:23 (in part; result of his synonymy of C. henrici with C. hir- tipes); Gadow, 1905:209 (in part; Ari- zona and New Mexico). Cinosternon integrum Strauch, 1890:91 (in part; result of his synonymy of C. hirtipes Gunther with C. integrum LeConte). Kinosternon flavescens Van Den burgh, 1922:972 (in part; Ft. Verde and Graham Co. records actually K. sono- riense; see Iverson, 1978); LaRivers, 1942:66 (in part; Nevada; see Iverson, 1978); Stebbins, 1966:82 (in part; northwest Arizona; see Iverson, 1978). Kinosternon sp. Little, 1940:264 (Roose- velt Reservoir and Sallymae Creek, Gila Co., Arizona); Tanner and Robison, 1960:59 (in part; specimens are K. sonoriense but locality doubtful). Kinosternon sonoriensis Bogert and Ohver, 1945:396; Smith and Buechner, 1947:10; H. M. Smith, Wiiliams and Moll, 1963:207. Kinosternon hirtipes H. M. Smith and E. H. Taylor, 1950a:25 (in part; Arizona). 42 Tulane Studies in Zoology and Botany Vol. 23 Kinosternon sonorensis Weise, 1962: 165. Kinosternon seonoriense Berry and Shine, 1980:189. Type. Lost; see synonymy. Content. Two subspecies, one new, are described: K. s. sonoriense and K. s. long- ifemorale. Diagnosis. A Kinosternon of the hirtipes species group with: 1.) the adult nasal scale large and triangular, rhomboi- dal or bell shaped (not furcate behind); 2.) usually three or four pairs of relatively long chin or neck barbels (at least one pair more than half orbit diameter); 3.) male plastron relatively wide (PWB 42-53% of CL; X = 47.2%); 4.) first neural often (38.1%) in contact with nuchal bone; 5.) the female generally larger than the male; and, 6.) populations confined to Arizona, California, New Mexico, Sonora, western Chihuahua, and possibly Baja California. Description. The adult carapace gen- erally is tricarinate with the medial keel most apparent; some turtles possess well- defined keels, others have only the medi- an keel present with mere hints of the two lateral keels, and still others have a virtu- ally smooth shell. The algae covered shells of some individuals are extremely rugose and densely pock-marked (Fig. 26; found in 15 of 164 turtles by Hulse, 1976:347), a condition perhaps induced by the algae (the condition is known for no other kino- Figure 26. Articulated shell (without scutes) of adult K. sonoriense (JBI 800) from Sonora, near Cucurpe (Rio Sonora basin). Note rugosity. sternid). The average female is larger than the average male. Maximum female size is 175 mm carapace length; males 155 mm. The nasal scale is not furcate behind in adults. The first vertebral scute usually touches the second marginal. The axillary is nearly always in broad contact with the inguinal, and the inguinal contacts the eighth marginal. The ninth marginal scute is not elevated above the preceding marg- inals. The tenth marginal is higher than the ninth marginal and the eleventh mar- ginal may or may not be elevated to the height of the posterior portion of the tenth marginal. Interpectoral seam length is less than one-half of gular length. The nuchal bone often contacts the first neural bone. The carapace is brown to olive in color, the seams darker. Well-developed transverse plastral hinges border the ab- dominal scutes. The male plastron is rel- atively extensive (PWB = 42 to 53% CL). The plastron is yellow to brownish with darker brown seams. The bridge area is dark brown. The skin is dark gray and the head and neck bear cream colored mot- tlings that tend to form at least one pair of stripes extending back from the orbit, one above and the other below the typanum after intersecting the angle of the jaw. A yellow or cream stripe also extends from the palmar surface of each foot to the base of the hmb along its posterior sur- face in some adults. Three to four pairs of relatively long chin or neck barbels usual- ly are present. Mature females possess short, stubby tails, with a small terminal spine, whereas males possess long, thick- ened tails with a large terminal spine and a patch of elevated (tubercular), acute, nonimbricated scales on the posterior sur- face of the crus and thigh of each hind leg. Remarks. Iverson (1976) has synthe- sized most of the hterature. Additional important references include Hulse (1976); Morafka (1977); Bowler (1977); Conant and Berry (1978); Iverson (1978); H.M. Smith and R.B Smith (1980); Berry and Legler (1980); and Iverson and Wey- man (MS). No. 1 Kinosternon Biosystematics 43 Kinosternon sonoriense is the largest Kinosternon in which the females are gen- erally larger than the males. Perhaps con- comitantly it produces the largest number of eggs per clutch of any kinosternine — up to at least eight (Hulse, 1974; Iverson, unpubhshed data). I have observed copu- lation in the field near Fort Huachuca, Arizona (Gila River basin) on 4 May 1974, much later than the March-April records of Hulse (1974). The smallest tur- tles I measured were 22.3 mm CL (20.0 mm PL), 23.9 mm CL (18.3 mm PL), and 25.7 mm CL. In the southern part of its range, this turtle is apparently active year round; I have observed activity at Quito- baquito Pond, Arizona on several occa- sions in January. Range. Kinosternon sonoriense occurs in the United States from the Lower Colo- rado and Gila rivers in Arizona and New Mexico, southward to and including the Rio Yaqui basin west of the continental divide, and eastward through the Rfo Casas Grandes basin of northwestern Chi- huahua. It is known from between at least 43 and 2042 m elevation. The species may also occur in the Rio Fuerte (see MATER- IALS AND METHODS). Specimens examined and Additional Records. See Locality list. Etymology. The specific name sonor- iense refers to the Sonoran Biotic Prov- ince, wherein the turtle is found. Kinosternon sonoriense sonoriense (LeConte) Sonora Mud Turtle Synonymy. See species synonymy, ex- cept those references in synonymy of K. sonoriense longifemorale. Holotype. Lost; see species account. Diagnosis. A subspecies of K. sonori- iense with 1) a relatively long interanal seam (male x lAN/CL, 19.5'^o; female x , 23.0%); 2) a relatively short interfemoral seam (male X, IF/CL, 10.1%; female x , 10.1%); 3) a first vertebral scute of medi- um width (male x , VW/CL, 24.4%; fe- male X , 25.5%); and 4) a relatively wide gular scute (male x , GW/CL, 20.0%; fe- male X , 19.4%). Range. Kinosternon s. sonoriense is definitely known from the Bill Williams, lower Colorado, Gila, Sonora, Magda- lena, Yaqui, southwest New Mexico, and Casas Grandes basins of Arizona and New Mexico, and Sonora and western Chihuahua, Mexico. Specimens examined and Additional Records. See locality list. Etymology. See species account. Kinosternon sonoriense longifemorale ssp. nov. Sonoyta Mud Turtle Kinosternon sonoriense M earns, 1907: 1 17 (Sonoyta); Van Denburgh, 1922:969. (Sonoyta River three miles from Sonoy- ta); Stebbins, 1966:83 (Quitobaquito Spring); Hulse, 1974:15, 94 (Quitoba- quito Spring); H.M. Smith and R.B. Smith, 1980:160 (3 localities in Sonoyta basin). Holotype. USNM 21710, young male, preserved whole, from "artificial pond fed by springs", Sonoyta, Sonora, Mex- ico (31°5rN, 112°50'W); collected 15 January 1894, app^arently by E.A. Mearns. Paratypes. USNM 21709 and 21711, topotypic male and female, preserved whole, and USNM 21708, aduh female, preserved whole, from Sonoyta River, 3 mi. from Sonoyta, collected on 22 Janu- ary 1894 by B.A. Wood; UAZ 27987 and 27996, adult male and female, respective- ly, preserved whole, Quitobaquito Springs, Organ Pipe Cactus National Monument, Pima County, Arizona, col- lected on 14 May 1967 and 10 April 1965, respectively, by R.D. Krizman and T.J. Cox, respectively; and UF 47719 and 47720 (Fig. 27), skeletal aduh male and female, respectively from Quitobaquito Pond, Pima County, Arizona, collected on 19 January 1976 by John B. Iverson. Diagnosis. A subspecies of K. sonori- ense with 1) a relatively short interanal seam (male x , lAN/CL, 14.4%; female 44 Tulune Studies in Zoology and Boiany Vol.23 Figure 27. Plastron of female Kinoslernon sonoriense longifemorale (UF 47720) from Quitobaquito Pond, Pima County, Arizona. Note short interanal and long interfemoral seams. X, 18.5%); 2) a relatively long interfem- oral seam (male x , IF/CL, 12.8%; female X, 13.5%); 3) a wide first vertebral scute (male x, VW/CL, 28.9%; female x, 28.8%); and 4) a relatively narrow gular scute (male x, GW/CL, 17.7%; female X, 17.8%). Range. Kinoslernon sonoriense longi- femorale is known only from the Ri'o Sonoyta basin in Arizona and Sonora, Mexico. Specimens examined and Additional Records. See locality list. Etymology. The subspecific name long- ifemorale is from the Latin longiis (long) and femoralis (of the femur; here the femoral scute) and refers to the long inter- femoral seam which, along with the short interanal seam, diagnoses the taxon. Kinoslernon hiriipes (Wagler) Rough-footed Mud Turtle Cinosternon hiriipes Wagler, 1830:137, plate 5, fig. 29-30 (Type-locality, "Mex- ico", restricted to "lakes near Mexico City" by Schmidt 1953:89, but see RE- MARKS under A'. /?. hiriipes). Holo- type, Miinchen Museum (Germany) 1374/0, a male, collected by Baron Kar- winski, collecting date unknown. Type- locality incorrectly restricted to "Maz- atlan, Sinaloa" by H.M. Smith and E.H. Taylor 1950b:25; see discussion in Hardy and McDiarmid, 1969:70, 218); Wagler, 1833:plate 30; Fitzinger, 1835:125; A.M.C. Dumeril and Bibron, 1834:370; A.M.C. Dumeril and Dum- eril, 1851:17; Gray, 1855:46 (in part; "Brazils" record in error); Strauch, 1862:41; Strauch, 1865:101; A.H.A. Dumeril, 1870:25; Westphal-Castelnau, 1872:278; Gray, 1873:113; Bocourt, 1876:8; Duges, 1888:106. Kinoslernon oblongum Gray, 1844:33 (in part). Cinoslermon hiriipes Gray, 1844:33 (in synonymy). Kinosternum hiriipes LeConte, 1854:186; LeConte, 1859:5; MuUer, 1885:716. Kinoslernon hiriipes Grdiy, 1855:47, 1869: 183, 1870:67; Stejneger, 1899:64; Rust, 1934:59; Taylor, 1936:529 (in part; Sin- aloa records based on K. integrum); Martin del Campo, 1937:265 (in part; Hidalgo record based on A', integrum); Rust, 1938:22; Caballero y Caballero, 1938a: 103 (in part; Hidalgo record based on A', integrum); Caballero y Caballero, 1940a:225 (in part; Urua- pan, Michoacan locality based on K. in- tegrum ); H.M. Smith and E.H. Taylor, 1950a:25 (in part; Chihuahua, Michoacan, Guanajuato, Mexico, and Distrito Federal; other locality based on K. sonoriense, K. flavescens, or K. inte- grum); H.M. Smith and E.H. Taylor, 1950b:342 (in part; type-locality re- striction to Mazatlan, Sinaloa invalid); Glass and Hartweg, 1951:50; Taylor, 1952:793; Schmidt, 1953:89; Mertens and Wermuth, 1955:336; Cable in Blair et al. 1957:281 (in part; Arizona records based on A. sonoriense); Malkin, 1958:75 (in part; Nayarit records based on A. integrum); Zweifel, 1960:94 re- jects Tres Marias records; Wermuth and Mertens, 1961:19; Croulet, 1963:4 (in part; Nayarit record based on K. in- tegrum); Liner, 1964:221 (in part; Guanajuato records on A', integrum); No. 1 Kinosternon Biosystematics 45 Casas Andreu, 1965:285 (in part; Sina- loa, Colima, and Hidalgo records based on K. integrum); Pritchard, 1967:37; Casas Andreu, 1967:44 (in part; Sina- loa, Colima, Hidalgo and Nayarit rec- ords based on K. integrum); Hardy and McDiarmid, 1969:104 (rejects Sinaloa records, including H.M. Smith and Taylor's 1950b:343 type-locality re- striction); Cochran and Coin, 1970:135 (in part; Arizona records based on K. sonoriense); Ernst and Barbour, 1972: 66 (in part; Arizona records based on K. sonoriense); Dixon, Ketchersid, and Leib, 1972:228 (in part; Queretaro rec- ord based on K. integrum); Greene, 1972: 2 (in part; Puebla record based on K. integrum); Bravo-HoUis and Cabal- lero Deloya, 1973:109; Conant and Berry, 1978:1; Iverson, 1978:1, Iverson and Berry, 1979:318; Pritchard, 1979:537; H.M. Smith and R.B. Smith, 1980:137; Berry and Legler, 1980:1. Cinosternon pensylvanicum Lichtenstein, 1856:2 (in part; Mexico); Westphal- Castelnau, 1872:278 (in part; Guana- juato); Bocourt, 1876:5 (in part; Mex- ico); Herrera, 1890:330, 1891:46 (in part; Valley of Mexico); Strauch, 1890:88 (in part; Mexico); Herrera, 1893:339 (in part; Mexico); Duges, 1898:40 (in part; Mexico); Herrera and Lope, 1899:281 (in part; Mexico); Herrera, 1904:5 (in part; Mexico). Thryrosternum hirt ipes Agassiz, 1857:429. Ozotheca hirtipes LeConte, 1859:6; Tros- chel, 1860:270. Ozotheca odorata Duges, 1869:143 (in part; states of Guanajuato and Mexico); Velasco, 1890b:54 (in part, Guanajuato); Velasco, 1891:52 (in part; Queretaro record unsubstantiated); Velasco, 1892b:40 (in part; Tlaxcala record in- correct); Velasco, 1893b:81 (in part; Sonora record incorrect); Velasco, 1894:40 (in part; Zacatecas record un- substantiated); Velasco, 1896a:30 (in part; Aguascalientes record unsubstan- tiated); Velasco, 1898:62 (in part; Chiapas record incorrect). Ozhoteca odorata Garcia Cubas, 1884: 179 (in part; Mexico); Velasco, 1890a: 35 (in part; Nuevo Leon record incor- rect); Velasco, 1893a:64 (in part; Dur- ango record unsubstantiated); Velasco 1897:41 (in part; Coahuila record incor- rect). Cinosternum hirtipes Gunther, 1885:13 (in part; Sinaloa records based on K. in- tegrum; Arizona and New Mexico rec- ords based on K. sonoriense); Cope, 1885:390; Cope, 1887:23 (in part; Col- ima and Sinaloa records based on K. in- tegrum); Garman, 1887:16 (in part; San Luis Potosf record based on K. inte- grum); Boulenger, 1889:38; Gadow, 1905:209 (in part; Arizona and New Mexico records based on K. sonori- ense); Siebenrock, 1906:94, 1907: 551; Gadow, 1908:5; Siebenrock, 1909: 444; Gadow, 1930:50. Cinosternon pennsylvanicum Duges, 1888:10 (in part; Valley of Mexico and Guanajuato state); Velasco, 1890b:291 (in part; Guanajuato); Duges, 1895:5 (in part; Guanajuato); Duges, 1896a: Iv (in part; Mexico); Duges, 1896b:329 (in part; Mexico); Duges, 1896c:479 (in part, Guanajuato). Ozothea odorata Velasco, 1892a:76 (in part; Guerrero record incorrect); Velas- co, 1892c:79 (in part; Tamaulipas rec- ord incorrect); Velasco, 1895:38 (in part; Campeche record incorrect); Vel- asco, 1896b:37 (in part; Colima record incorrect). Cynosternon pennsylvanicum Herrera, 1893:342 (in part; Valley of Mexico). Cinosternum pennsylvanicum Gadow, 1905:209 (in part; Valley of Mexico). Kinosternum pennsilvanicum Cope, 1896: 1021 (in part; Austrocentral district of Mexico). Cinosternonus pensylvanicum Herrera, 1899:28 (in part; Mexico). Cynosternon pensylvanicum Herrera and Lope, 1899:131 (in part; Valley of Mex- ico). Cinosternum pennsilvanicum Cope, 19(X): 1229 (in part; Valleys of Mexico and Toluca northward through Guana- 46 Tulane Studies in Zoology and Botany Vol. 23 juato). Cinosternum pensylvanicum Lampe, 1901:184-85 (North Mexico). Cinosternon sp. Herrera, 1904:6 (Valley of Mexico). Cinosternum integrum Gadow, 1908:518 (in part; Laguna de Zapotlan, Jalisco). Kinosternon sonoriense Dunn, 1936:472 (in part; Chihuahua); H.M. Smith and E.H. Taylor, 1950a:26 (in part; Chi- huahua and Durango); Mertens and Wermuth, 1955:338 (in part; Chihua- hua to Durango); Casas Andreu, 1965: 386, 1967:52 (in part; Chihuahua and Durango); Legler and Webb, 1970:163 (in part; western Chihuahua); Iverson, 1976:1 (in part; upper Rfo Yaqui, Chi- huahua; see Iverson, 1978). Chinosternum hirtipes Caballero y Cabal- lero, 1938b:449 (in part; Hidalgo local- ity based on K. integrum). Sternotherus odoratus Brown, 1950:230 (in part; Presidio Co., Texas; see Conant and Berry, 1978). Kinosternon murrayi Glass and Hartweg, 1951:50 (type-locality, "Harper Ranch, 37 miles south of Marfa, Presidio County, Texas." Holotype, Texas Co- operative Wildlife Collection 650, a young male, collected by S.H. Wheeler on 12 August 1941); Peters, 1952:54; Legler, 1960:139 (Lajitas, Texas record in error); Cochran, 1961:232. Kinosternon flavescens Stebbins, 1966:82 (in part; Durango; see Iverson, 1978); Morafka, 1977:70, Map 25 (in part; some northern Mexico records based on K. hirtipes.). Kynosternon hirtipes Lopez 1975:2 (Val- ley of Mexico). Kinosternon hertipes Semmler et al., 1977: 18 (Chihuahua). Types. Only the holotype (Fig. 25), an adult male, preserved whole, is available, contrary to Bocourt's (1876:8) suggestion that Wagler's (1830, 1833) figures (Plate 5:fig. 29-30 and Plate 30:figs. 1-3, respec- tively) of Cinosternon (- Kinosternon) hirtipes were based on two different speci- mens. Content. Six subspecies, four new, are described: K. h. hirtipes, K. h. chapal- aense, K. h. murrayi, K. h. magdalense, K. h. tarascense, and K. h. megacephal- um. Diagnosis. A Kinosternon of the hir- tipes species group with 1) the adult nasal scale reduced and crescent-shaped, or larger and furcate behind, or still larger and triangular or bell shaped (the latter combination characteristic only of Valley of Mexico turtles); 2) usually three pairs of relatively short chin barbels ( < half orbit diameter); 3) male plastron relative- ly narrow (PWB 36 to 50% of CL; k = 43'Vo); 4) first neural rarely (10.2%) con- tacting nuchal bone; 5) the male generally larger than the female; and 6) populations confined to Central Mexico from Chihua- hua (and adjacent Texas) to Jalisco, Michoac^n, and Mexico (state). Despite its anomalous absence on the holotype (Fig. 25), a nuchal scute is typically pres- ent. Description. As for K. sonoriense ex- cept as stated above, and 1) the carapacial keels are almost never absent (i.e., the median keel is virtually always evident at least posteriorly); 2) maximum female size is 157 mm carapace length, male 182 mm; 3) carapace light to dark brown to nearly black in color; 4) plastron usually yellow to brown with darker brown seams but sometimes (stained ?) nearly black; 5) head markings extremely variable (coarse- ly mottled, reticulated or spotted to almost unmarked; see subspecific ac- counts). Remarks. Most of the literature is synthesized in Iverson (in press). A discus- sion of the evolutionary significance of the geographically variant sexual size di- morphism of this species appears in Iver- son (MS 2). Reproductive parameters are summarized here (Table 2) and in Iverson (MS 2). Clutch size data (4-5 and 4-7 eggs) in Moll and Legler (1971) are all referable to the subspecies murrayi. Scaling of skel- etal components is discussed in Iverson (MS 1) and Iverson and Weyman (MS). No. 1 Kinosternon Biosystematics 47 ^ ."s* "•*3 3 g c o TS a> CO cd X> CO (U J3 CJ •<-> ^ "o ^.^ cd 3 C C cd • Cm c o cd (L> a ^ o^ 3 C T3 CO ^ kH 6 a CO aj w (U 00 ^ c .1 •^ -^ *S s; J3 o o 5 3 5J to U O • s: CO ^ ^ Lh O O Cm C4-C c 73 o !U 00 5 3 "c a a> o C4-1 0 >» CO x> ■t-> 0) cS CO -!-> c« IZ) T3 _D (L» a > CTJ %-> a> O -<-> 3 ^ -a o c« u> l-l O. 0 u 0, oi 0 0 CO cd H CO « 3 O 01 01 ■g O _ CO ^ Q> 3 ^ E ^ n ii tS £ E 3 w 2 ''" « -3 o « 3 on ^ exi O ^ u ba 3 R ^^ O) ,^ >> CO _aj w^ "« V E C/3 3 00 3 < 03 ON 00 fs O OS r- 00 3 03 cd OS >% 3 c^. c^- 00 1> (>• 3 3 < c^- C '—1 _>. _>^ 3 (U >.c"- (U 3 3 "-5 c 03 c^- C ■— ) '— > >, 3 c 3 I—) 3 c3 03 03 — ) fS '^ vO ro ^ ^ r-" r-" 00 r~-' ^— < 1-^ ^^ ^^ ^^ ' >< s X •r^ 1 ^ w X vo X OS 1 rsi ^^ On QO ^^ 0 ^^ 00 OS d d Os" _^ Q m 4 4 N H < CU 48 Tulune Sludles in Zoology und Botany Vol. 23 Range. Primarily distributed on the Mexican Plateau, Kinosternon hirdpes ranges from Alamito Creek in Texas in the United States and the Rfos Santa Maria, Carmen, and Conchos in northern Mexico south and eastward to the Chap- ala, Zapotlan, San Juanico, Pa'tzcuaro, and Valle de Me'xico basins of the Sierra Volcanica Transversal of southern Mex- ico. It is known from between at least 800 and 2600 m in elevation. Specimens examined and Additional Records. See locality list. Etymology. The specific name hirtlpes is from the Latin, hirtus, meaning rough, and pes meaning foot, and refers to the rough scales on the feet of the species. Kinosternon hirtlpes hirtlpes Wagler Valley of Mexico Mud Turtle CInosternon hirtlpes Wagler, 1830:187 (see species synonymy). Ozotheca odorata Duges, 1869:143 (in part; State of Mexico). CInosternon pennsylvanlcum Duges, 1888:107 (in part; Valley of Mexico). CInosternum /?//7//7e5 Boulenger, 1889:38; Siebenrock, 1906:94, 1907:551 (State of Mexico); Gadow, 1908:5 (Chalco lakes, Valley of Mexico). CInosternon pensvlvanlcum Herrera, 1890:330, 1891:46 (in part; Valley of Mexico). Cynostenum pennsylvanlcum Herrera, 1893:342 (in part; Valley of Mexico). Cynosternon pensylvanlciim Herrera and Lope, 1899:131 (in part; Valley of Mexico). CInosternom pennsllvanlcum Cope, 1900: 1229 (in part; Valley of Mexico). CInosternon sp. Herrera, 1904:6 (Mexi- calzingo. Valley of Mexico). CInosternum pennsylvanlcum Gadow, 1905:209 (in part; Valley of Mexico). Kinosternon hirtlpes Martin del Campo, 1938:391 (Valley of Mexico); Caballero y C, 1939:279 (Xochimilco, Mexico, Distrito Federal); H.M. Smith and Taylor, 1950a:25 (Distrito Federal); Glass and Hartweg, 1951:50 (Valley of Mexico); Schmidt, 1953:89; Beltz, 1954:124 (Mexico City, Mexico); Martin del Campo, 1955:66 (Valley of Mexico); Deevey, 1957:240 (Valley of Mexico); Casas Andreu, 1965:385 (Dis- trito Federal); Kranz, Smith, and Smith, 1971:23 (near City of Mexico); Greene, 1972:2 (in part; Mexico City, Puebla locality based on A'. Integrum); Perez Villegas and Reyna Trujillo, 1978:215 (southern region of Valley of Mexico). Kinosternon hirtlpes hirtlpes Mertens and Wermuth, 1955:336 (first use of this combination; in part; State of Mexico); Wermuth and Mertens, 1961:20 (in part; State of Mexico); Duellman, 1961:57, 1965:653 (in part; Michoacan localities not this subspecies); H.M. Smith, Williams and Moll, 1963:209; Liner, 1964:221 (in part; Guanajuato records not this subspecies); Pritchard, 1967:37 (in part; State of Mexico); Casas Andreu, 1967:44 (in part; State of Mexico); Parsons, 1968:1238; Legler and Webb, 1970:163 (in part; Chihua- hua records based on A", h. murrayl); Mittermeier, 1971:16 (Mexico City); Moll and Legler, 1971:92 (in part; Chi- huahua records based on K. h. murrayl); Wermuth and Mertens, 1977: 7; Pritchard, 1979:537 (in part; Mexico City). Kynosternon hirtlpes Lopez 1975:2 (Val- ley of Mexico. Kinosternon sp. Niederberger, 1979:134 (Valley of Mexico archeological re- mains: 5500 BC). Types. Only the holotype (Fig. 25) is available (see Remarks below). Diagnosis. A subspecies oi Kinosternon hirtlpes with 1) the adult nasal scale tri- angular, rhomboidal, or bell-shaped (fur- cate behind in subadults, but not in large adults); 2) a mottled head pattern, typical- ly organized into a light streak extending posteriorly from the angle of the jaw, with a similar light postorbital streak vari- ably evident; 3) one or (typically) two pairs of mental chin barbels, the anterior pair largest; 4) medium body size (maxi- No. 1 Kinosternon Biosystematics 49 mum male size 140 mm CL; female, 140 mm); 5) relatively short bridge length (male BL/CL, 17.6<7o; female 21.7'^o); 6) relatively short interfemoral seam length (male IF/CL, 6.9%; female , 7.1%); 7) relatively long inter- anal seam length (male lAN/CL, 20.6%; female ,25.8%) and 8) popula- tions confined to the Valley of Mexico. Remarks. As mentioned under SYN- THESIS (above), the allocation of the holotype of Kinosternon hirtipes to the Valley of Mexico must remain uncertain until additional material is available from the southern and southwestern margins of the Mexican Plateau. Specific natural history data are un- available for Kinosternon hirtipes hir- tipes. A photograph of the plastron of UMMZ 99449, an adult female, appears in H.M. Smith and R. B. Smith (1980; plate 19, bottom). Range. Kinosternon hirtipes hirtipes is known only from the drainages of the Valley of Mexico. Specimens Examined and Additional Records. See locality lists. Etymology. See species account. Moll and Legler, 1971:92 (in part; Chi- huahua records based on k. h. murrayi); Wermuth and Mertens, 1977: 7; Pritchard, 1979:537 (in part; Mexico City). Kynosternon hirtipes Lopez 1975:2 (Val ley of Mexico. Kinosternon sp. Niederberger, 1979:134 (Valley of Mexico archeological re- mains: 5500 BC). Types. Only the holotypes (Fig. 27) is available (see Remarks below). Diagnosis. A subspecies of Kinosternon hirtipes with 1) the adult nasal scale tri- angular, rhomboidal, or bell-shaped (fur- cate behind in subadults, but not in large adults); 2) a mottled head pattern, typical- ly organized into a light streak extending posteriorly from the angle of the jaw, with a similar light postorbital streak vari- ably evident; 3) one or (typically) two pairs of mental chin barbels, the anterior pair largest; 4) medium body size (maxi- mum male size 140 mm CL; female, 140 mm); 5) relatively short bridge length (male x BL/CL, 17.6%; female x , 21.7%); 7) relatively short interfemoral seam length (male x IF/CL, 6.9%; female x , 7.1%); 8) relatively long inter- anal seam length (male x lAN/CL, 20.6%; female x, 25.8%) and 9) popula- tions confined to the Valley of Mexico. Remarks. As mentioned under SYN- THESIS (above), the allocation of the holotype of Kinosternon hirtipes to the Valley of Mexico must remain uncertain until additional material is available from the southern and southwestern margins of the Mexican Plateau. Specific natural history data are un- available for Kinosternon hirtipes hir- tipes. A photograph of the plastron of UMMZ 99449, an adult female, appears in H.M. Smith and R. B. Smith (1980; plate 19, bottom). Range. Kinosternon hirtipes hirtipes is known only from the drainages of the Valley of Mexico. Specimens examined and Additional Records. See locality lists. Etymology. See species account. Kinosternon hirtipes murrayi Glass and Hartweg Murray's Mud Turtle Ozotheca odorata Duges, 1869:143 (in part; Guanajuato); Velasco 1890b: 54 (in part; Guanajuato). Cinosternon hirtipes Westphal-Castelnau, 1872:278 (Guanajuato). Cinosternum hirtipes Cope, 1887:23 (in part; city of Chihuahua, Guanajuato). Cinosternum pennsylvanicum Duges, 1896c:479 (Guanajuato). Cinosternon pennsilvanicum Cope, 1900: 1229 (Toluca Valley northward through Guanajuato. Kinosternon sonoriense Dunn, 1936:472 (in part; Rio Conchos, Julimes, Chi- huahua); H.M. Smith and Taylor 1950a:26 (in part; Chihuahua and Dur- ango); Mertens and Wermuth, 1955: 338 (in part; Chihuahua and Durango); 50 Tulane Studies in Zoology and Botany Vol. 23 Casas Andreu, 1965:386 (in part; Chi- huahua and Durango); Legler and Webb, 1970:163 (in part; Rfos Papi- gochic and Tomuchic in western Chi- huahua); Iverson, 1976:1 (in part; Upper Rfo Yaqui, Chihuahua; see Iver- son, 1978). Kinosternon hirtipes Caballero y C, 1940b: 562 (Rio Lerdo, Guanajuato); Caballero y C. y Cerecero, 1943:534 (Rio Lerdo del Valle de Santiago, Guanajuato); H.M. Smith and Taylor, 1950b:25 (in part; Chihuahua, Guana- juato); Williams, Smith, and Chrapliwy, 1960:36 (Chihuahua, 1 mi. E La Cruz); Casas Andreu, 1965:385 (in part; Chi- huahua, Guanajuato); Conant, 1978: 465 (Texas, Chihuahua, Durango and Zacatecas). Sternothenis odoratus Brown, 1950:230 (in part; Presidio Co., Texas; based on holotype oiK. murrayi; see Conant and Berry, 1978:15). Kinosternon murrayi Glass and Hartweg, 1951:50 (Type-locality, "Harper Ranch, 37 miles south of Marfa, Pres- idio County, Texas." Holotype, TCWC 650, a young male, collected 12 August 1941 by S.H. Wheeler.); Peters, 1952:54 (Texas); Legler, 1960:139 (Jet. Rio San Pedro and Conchos, and Ojin- aga. Chihuahua). Kinosternon hirtipes murrayi Schmidt, 1953:89 (first use of combination; Texas); Mertens and Wermuth, 1955: 336 (Texas); H.M. Smith, Williams and Moll, 1963:207 (Chihuahua); Casas Andreu, 1967:45 (Texas, Chihuahua, and Durango); Parsons, 1968:1238; Cochran and Goin, 1970:135 (Texas); Moll and Legler, 1971:92 (Durango and Chihuahua); Ernst and Barbour, 1972: 66 (Texas); Hambrick, 1976:292 (Texas); Wermuth and Mertens, 1977:7 (Texas); Conant and Berry, 1978:1 (Texas and Chihuahua); Iverson, 1978: 476 (Chihuahua). Kinosternon hirtipes hirtipes Mertens and Wermuth, 1955:336 (in part; Chihua- hua); Duellman, 1961:57 (in part ?; Michoacan, 8 km W Ciudad Hidalgo and Lago'de Cuitzeo); Casas Andreu, 1967:44 (in part; Chihuahua, Michoa- can, and Guanajuato); Legler and Webb, 1970:163 (in part; Rios Papi- gochic and Tomuchic, western Chihua- hua); Moll and Legler, 1971:92 (in part; Chihuahua); Wermuth and Mertens, 1977:7 (middle and western Mexico); Pritchard 1979:537 (in part). Kinosternon flavescens Stebbins, 1966:82 (in part; Durango; see Iverson, 1978). Kinosternon hirtipes murryi Ashton et al., 1976:51 (lapsus pro murrayi). Kinosternon hertipes Semmler, et al., 1977:18 (near Galeana, Chihuahua). Types. Holotype: see subspecies synon- ymy. Paratypes: USNM 15860, adult male, preserved whole, from "Marfa, Presidio County, Texas", collected by Vernon Bailey: UMMZ 101294, adult male, preserved whole, and UMMZ S- 1083, shell of adult male, both topotypic and collected 12 June 1950 by Herndon G. Dowling. Diagnosis. A subspecies oi Kinosternon hirtipes with: 1) a large posteriorly furcate nasal scale (typically exending posterior to the orbits); 2) an extremely variable mot- tled to reticulated head pattern; 3) typical- ly two pair of mental chin barbels, the anterior pair largest; 4) medium to large body size (maximum known male size, 182 mm CL; female, 157 mm);^) relative- ly long bridge length (male x BL/CL, 20.0<^o; female x, 23. 7%); 6) relatively long gular length (male x GL/CL, 14.7<^o; female x, 15.8%); and 7) popu- lations confined to the Big Bend region of Texas and adjacent Chihuahua southward across the Mexican Plateau to northern Jalisco, northern Michoacan, and eastern Mexico (state). Remarks. As discussed in the results, there appears to be a slight morphometric distinction between populations of K. h. murrayi in the Ri'o Nazas northward, and populations in the Rio Aguanaval south- ward. This difference is not considered significant enough to warrant subspecific designation, but has some interesting zoo- geographic impUcations. No. I Kinosternon Biosystematics 51 Range. Kinosternon hirtipes murrayi is known from the following basins in Aguascalientes , Chihuahua, Coahuila, Durango, Guanajuato, Jalisco, Mexico, Michoacan, San Luis Potosi, Texas, and Zacatecas: Santa Maria (Chihuahua), Carmen, El Sauz, Conchos, Bustillos, Papigochic, Nazas, Viesca, Aguanaval, Santiaguillo, Mezquital, El Salto, Santa Maria (San Luis Potosi; presumably in- troduced), Aguascalientes, Verde, Lerma (except Chapala), Cuitzeo, Balsas, and Villa Victoria (with reservation). Specimens examined and Additional Records. See locality lists. Etymology. The subspecific name mur- rayi is a patronym, honoring Dr. Leo T. Murray of Texas A & M College. Kinosternon hirtipes chapalaense ssp. nov. Lake Chapala Mud turtle Cinosternum integrum Gadow, 1908:518 (in part; Laguna de Zapotlan, Jalisco). Kinosternon hirtipes Altini, 1942:153 (in part; Lake Chapala, Jalisco). Kinosternon hirtipes hirtipes Duellman, 1961:57, 1965:653 (in part; Jiquilpan, La Palma, Lago de Came'cuaro = 14 km E Zamora, Michoacan). Kinosternon hirtipes chapalaense Pritch- ard, 1979:557 (nomen nudum; Lake Chapala). Holotype. UMMZ 97128, adult male, preserved whole, from Lake Chapala, 0.25 mile off Chapala, Jalisco, Mexico [20°18'N, 103°12'W]; collected 15 July 1947, by Norman Hartweg. Paratypes. All preserved whole: UMMZ 97122-23, topotypic adult females; UMMZ 97124, topotypic subadult male; 97125-27 and 97129-30, topotypic adult males; and UU 12126-12128, adult male, subadult fe- male, and juvenile, Lago de Chapala, 3.2 km W Chapala; and UU 12125, adult fe- male, Lago de Chapala, 6.1 km W Ajijic, all collected on 21-22 June 1969 by Clyde Barbour. Diagnosis. A subspecies of Kinosternon hirtipes with 1) a reduced crescent-shaped nasal shield, which nearly always lies an- terior to the orbits (Figure 23); 2) a reduc- tion of dark pigment on the head and neck, dark markings confined to isolated spots or reticulations dorsally (Figures 23 and 24), but laterally sometimes organ- ized as two dark, nearly parallel post- orbital stripes; 3) the neck and chin virtu- ally unmarked and the mandibular and maxillary sheaths bearing only a few dark streaks, if any; 4) one, two, or three pairs of mental barbels present, the anterior- most pair (near the mandibular symphysis) usually the largest; 5) medium body size (maximum known size for males is 152 mm CL; females, 149 mm); 6) relatively long bridge length (male x BL/CL, 20.3<^o; female x, 25.3%); 7) relatively long interanal _seam (male x lAN/CL, 19.1%; female x, 25.2%); and 8) popula- tions confined to the Chapala and Zapot- lan (and possibly Duero) basins of Jalisco and Michoacan. Remarks. Field notes accompanying the topotypes provide no additional geo- graphical and ecological information. However, Clyde Barbour (pers. comm.) obtained the non-topotypic paratypes (during the night 21-22 June 1969) along the shore of Lake Chapala, on trot-lines baited with liver. These lines were neces- sarily buoyed off the lake bottom with floats to avoid bait removal by crabs. Peter Meylan found a single rotten car- cass of this species on the south shore of Lake Chapala just east of Tuxcueca dur- ing my field trip to the area on 15 June 1978; trapping at that locality produced no turtles. Trapping in isolated spring-fed pools just northeast of the town of Chap- ala on 9 May 1981 produced only K. inte- grum. Range. Kinosternon hirtipes chapalense is known only from the Lago de Chapala and Laguna de Zapotlan drainage basins in Jalisco and Michoacan, Mexico. Speci- mens from the Rio Duero basin are tenta- tively considered intergrades with K. h. murrayi. Specimens Examined and Additional Records. See locality list. Etymology. The subspecific name chap- 52 Tulane Studies in Zoology and Botany Vol. 23 alaense refers to Lake Chapala wherein the type series was collected. tKinosternon hirtipes megacephalum ssp. nov. Viesca Mud Turtle Holotype. SM(BCB) 1 1466, adult male, preserved whole, from 3.2 km SE Viesca [25 °2rN, 102 °48'W], Coahuila; collected 4 June 1961 by Bryce C. Brown and John Wottring by seining a drying pond. Paratypes. SM(BCB) 11460-65, adult females, preserved whole, all topotypic (11461 photographed in H.M. Smith and R.B. Smith, 1980); and SM(BCB) 9823, adult male, preserved whole, from 9.7 km SW Viesca, also collected on 4 June 1961 by Brown and Wottring (see Figs. 22 and 24). Diagnosis. A subspecies of Kinosternon hirtipes with: 1) enlarged head, hyper- trophied head musculature, and broad al- veolar jaw surfaces (Fig. 24); 2) the nasal scale furcate posteriorly; 3) the head pat- tern mottled or reticulated as in K. h. murrayi; 4) three to four pairs of chin barbels are present, two to three mental pairs (anterior usually the largest) and one small pair at level of anterior edge of tympanum; 5) small body size (maximum known size for males 99 mm CL; females, 1 17 mm); 6) plastron extremely reduced in size (Fig. 22); 7) relatively short bridge length (male x BL/CL, 17.3^70; female x , 23.90/0); 8) relatively short gular length (male x GL/CL, ll.Oo/o; female x, 12.8°7o); 9) relatively short interanal seam length (male x lAN/CL, 15.9"7o; female X , 20.90/0); and 10) populations confined to southwestern Coahuila. Remarks. This subspecies is known only from the type series. Field work in the area of the type locality (see MATER- IALS AND METHODS) suggests that K. h. megacephalum is now extinct; natural permanent water habitats apparently no longer exist near Viesca. Future field work should be concentrated in the mountains south of the city of Viesca in hope of discovering permanent water situ- ations where turtles (and fishes?) might still exist. The distinctive trophic apparatus of this subspecies is likely an adaptation to stenophagous molluscivory. Range. Known only from the two local- ities in Coahuila at which the type series was collected. Etymology. The subspecific name megacephalum is from the Greek mega, meaning large, and kephale, meaning head, and refers to the enlarged head, di- agnostic of the subspecies. Kinosternon hirtipes tarascense ssp. nov. Patzcuaro Mud Turtle Kinosternon hirtipes Altini, 1942:153 (in part; Lake Patzcuaro, Michoacan). Kinosternon hirtipes hirtipes Duellman, 1961:57, 1965:653 (in part; Lago de Patzcuaro, Michoacan); Casas Andreu, 1967:45 (in part; Patzcuaro, Canal de la Tzipecua, Michoacan). Holotype. UF 43506, adult male, pre- served whole, from Lago de Patzcuaro, adjacent to city of Patzcuaro [19°32'N, 101°36'W]; purchased in Patzcuaro market 13 June 1978 by John B. Iverson. Paratypes. All topotypic and preserved whole: UF 43505 and 43596, adult females; and UF 43507 and 43595, adult males. Diagnosis. A subspecies of K. hirtipes with: 1) a typically finely mottled to spotted head (Fig. 24); 2) variable red- brown to brown staining on the otherwise light yellow plastral scutes [The dark plas- tral scutes are apparently a result of natural staining; the character is exhibited to variable degrees by individuals and the dark color is lost when plastral scutes are shed (Fig. 22)]; 3) the large nasal scale posteriorly furcate; 4) two pairs of mental chin barbels typically present; 5) small to medium body size (maximum known size for males 136 mm CL; females, 132 mm)^: 6) relatively short bridge length (male x BL/CL, 18.0^^0; female x, 21. 40/0); 7) rel- No. 1 Kinosternon Biosystematics 53 atively short gular length (male x GL/CL, 10.6%; female x , 12.6%); 8) rel- atively long interpectoral seam length (malex IP/CL, 10. lo/o; female x, 8.5%); and, 9) populations confined to the Lago de Patzcuaro drainage basin. Remarks. Despite considerable study of other components of the biota of the Lago de Patzcuaro (see review in Cole, 1963 and Barbour, 1973), the mud turtles have been ignored. Reproductive infor- mation resulting from my studies appears in Table 2. Range. Known only from the basin of the Lago de Patzcuaro, Michoacan. Specimens examined and Additional Records. See locality list. Etymology. The subspecies name tara- scense honors the native tribe of Indians, the Tarascas, inhabiting the Patzcuaro area. Kinosternon hirtipes magdalense ssp. nov. San Juanico Mud Turtle Holotype. UF 45035, an adult male, preserved whole, from along the face of the dam at Presa San Juanico, Michoacan [ca. 19°50'N, 102°40'W] (Fig. 28). Holo- type collected 15 June 1978 by John B. Iverson, Peter A. Meylan, and Ron Magill. Paratypes. UF 45036, a subadult fe- male, UF 45038, female shell; UF 45039- 40, male shells, all topotypic; and TUL 18677, aduU male, collected atop Presa San Juanico 9 August 1963 by Clyde D. Barbour and Salvador Contreras-Balderas. Diagnosis. A subspecies o{ Kinosternon hirtipes with: 1) a finely mottled to spot- ted head pattern with jaw streaking mini- mal or absent; 2) a large nasal scale, fur- cate behind; 3) two pairs of mental chin barbels present; 4) small body size (max- imum known male size 94 mm CL; fe- male, 91 mm); 5) a relatively small plas- tron (male x PWB/CL, 41.9%; female X, 43.5%); 6) relatively short bridge length (male x BL/CL, 18.5%; female x , 19.7%); 7) relatively short gular length (male x GL/CL, 9.9%; female x, 11.0%); 8) relatively long interpectoral seam length (male x IP/CL, 8.7%; fe- male X, 11.0%); and, 9) populations re- stricted to the Magdalena Valley, Micho- acan. Remarks. As mentioned earlier (MAT- ERIALS AND METHODS), the turtles inhabiting the Presa San Juanico are poorly known. Future field work in the area should help delimit the subspecific range within the Magdalena Valley and also provide basic natural history infor- mation. Range. Kinosternon hirtipes magdal- ense is known only from the type series, all from the reservoir above Presa San Juanico in the Magdalena Valley of Michoacan, Mexico. Etymology. The subspecific name mag- dalense refers to the Magdalena Valley of Michoacan to which the subspecies is apparently endemic. Figure 28. Holotype (UF 45035) of Kinosternon hirtipes magdalense. 54 Tulane Studies in Zoology and Botany Vol. 23 Evolution Based on the derived characters of the turtles of the Kinosternon hirtipes species group (Table 3), I have constructed a phy- logeny of the included taxa (Fig. 29). For reasons discussed by Farris (1966) and Kluge and Farris (1969) (e.g., high intra- familial variation and indiscrete character shifts), I have not always assumed that taxa sharing derived morphometric char- acters are closely related. In fact, the dis- tribution of some character states among taxa clearly indicates that those characters are not a result of single origin, but rather of convergence. For example, the length- ening of the interanal seam in K. sonor- iense (longifemorale) and K. hirtipes (hir- tipes and chapalaense) certainly illustrates multiple origin of a derived character state. In addition, Viesca (megacephal- um), Patzcuaro (tarascense), and San Juanico turtles (magdalense) all share a relatively short bridge (with Valley of Mexico turtles), a short gular, and small body size (the latter two also share a long interpectoral seam), yet geographically and zoogeographically (Iverson, in prep- aration) the three populations likely do not represent a monophyletic divergence from a pre-murrayi stock. Rather, the evolution of these character states is more likely a response to selection in similar, very narrow adaptive zones (i.e., isolated, very small basins). Unfortunately, the functional significance of those characters is unknown, as is that of most of the other characters herein examined (but see Iver- son, MS 2) Table 3. Tally of subspecific taxa exhibiting derived character states in the Kinosternon hirtipes species group. Primitive states are discussed in the text. Derived Character 1 small plastron 2 short bridge 3 short gular 4 long gular 5 long interpectoral 6 short interpectoral 7 short interfemoral 8 long interanal 9 short interanal 10 V-shaped nasal 11 reduced nasal 12 multiple, long chin barbels 13 reduced head pigment 14 well-developed head stripes 15 large head 16 female > male 17 small body size 18 tendency toward unicarination Taxa exhibiting derived character magdalense, megacephalum magdalense, tarascense, hirtipes, megacephalum magdalense, tarascense, megacephalum sonoriense and longifemorale magdalense, tarascense sonoriense, longifemorale hirtipes chapalaense, hirtipes, longifemorale megacephalum murrayi, magdalense, tarascense, megacephalum, chapalaense chapalaense sonoriense, longifemorale chapalaense hirtipes megacephalum sonoriense, longifemorale (possibly tarascense, megacephalum) magdalense, tarascense, megacephalum murrayi, hirtipes, magdalense, tarascense, megacephalum, chapalaense No. I Kinosternon Biosystematics 55 The Kinosternon hirtipes species group apparently evolved on the Mexican Pla- teau from an ancestor as yet unknown. Despite the fact that several coastal streams have come to drain the Plateau due to headwater stream erosion (e.g., Rios Yaqui, Me^quital, Santiago, Balsas; see Fig. 4), K. hirtipes has nowhere left the Plateau. This is surprising since K. integrum has apparently moved both up and down several of these basins (Balsas, Santiago-Lerma, and Mezquital; Iverson, unpublished). K. sonoriense apparently evolved from a K. hirtipes-\\ke ancestor isolated in the Sonoran Desert, possibly following migration across the well- documented Sonora Desert-Chihuahua Desert filter barrier in southeastern Ari- zona, southwestern New Mexico and ad- jacent Mexico (see review in Morafka, 1977). Because so much geological infor- mation concerning the Mexican Plateau is now available (see reviews in Barbour, 1973 and Wauer and Riskind, 1978), a discussion of the historical zoogeography of the K. hirtipes species group will appear elsewhere (Iverson, in prepara- tion). The relationship between the Kino- sternon hirtipes species group and other Kinosternon is unclear. Siebenrock p* „.''^/^ ..^' y ./- j^"- y Figure 29. A theory of relationships among the subspecific taxa of the Kinosternon hirtipes species group. Numbers refer to derived character states listed in Table 3. Solid lines cutting line- ages mark identical shifts (convergence) in character states. (1907:551) included K. hirtipes and K. sonoriense, K. baurii, K. subrubrum, K. flavescens, and K. steindachneri (= K. subrubrum) in the K. subrubrum species group. However, I believe that K. baurii and K. subrubrum (including steindach- neri) represent a species group distinct from the K. hirtipes group, and that K. flavescens is similarly distinct. Perhaps the closest relative of the hirtipes group is K. herrerai (found in the Tampico Em- bayment of eastern Mexico; i.e., non- Plateau), which shares with most K. hir- tipes the elevated scale patches on the hindlegs of males, the tendency toward unicarination in adults, the furcate nasal scale, the reduced plastron, the broad inguinal-axillary contact, and several morphometric plastral characters. Un- fortunately, the determination of the phylogenetic relationship of the K. hir- tipes group to the other species groups in the genus must await further analysis. Key To Adult Turtles Of The Kinosternon hirtipes SPECIES GROUP lA. Nasal shield triangular, rhomboidal, or bell shaped; largest 2 pairs of chin barbels relatively long (at least one pair > half orbit diameter, with one pairmental and other at mid-tympan- um level); interpectoral length averages 5.0% of plastron length in males (less than 8% in 95% of cases) and 4.0% in females (less than 7% in 96% of cases); posterior width of plastral forelobe (PWB) averages 47.2% of carapace length in males (more than 44% in 95% of cases) and 49.0% in females (more than 45% in 96% of cases); maximum gular width averages 19.7% of carapace length in males (more than 18% in 94% of cases) and 19.1% in females (more than 17% in 94% of cases); first neural bone often (38.1%) in contact with nuchal bone; northwestern Chihuahua and Sonora, Mexico and adjacent New Mexico, Arizona and California Kino- sternon sonoriense 2 56 Tulane Studies in Zoology and Botany Vol. 23 IB. Nasal shield large and deeply notched posteriorly (V-shaped), or reduced to crescent-shaped scale lying fully anterior to level of orbits, or triangular, rhomboidal, or bell shaped if from Valley of Mexico; largest 2 pairs of chin barbels relatively short (< half orbit diameter), mentally located, with anterior pair larger; interpectoral length averages 8.2*^0 of plastron length in males (more than 4.5% in 97<^o of cases); and 6.6% in females (more than 3.5% in 94% of cases); posterior width of plastral forelobe (PWB) averages 42.8% of carapace length in males (less than 48% in 98% of cases) and 47.6% in females (less than 51% in 95% of cases); maximum gular width averages 17.3% of carapace length in males (less than 20% in 98% of cases) and 17.0% in females (less than 20% in 98% of cases); first neural rarely (10.2%) in contact with nuchal; Chihuahua, Mexico and adjacent Texas south- ward to Jalisco, Michoacan, and Me'xico, Mexico (state) Kinosternon hirtipes 3 2A. Interanal seam length averages 19.5% of carapace length in males (more than 16.5% in 97% of cases) and 23.0% in females (more than 21% in 90% of cases); interfemoral seam length averages 10.1% of carapace length in males (less than 13% in 93% of cases) and 10.1% in females (less than 12.5% in 95% of cases); maximum first vertebral width averages 24.4% of carapace length in males Oess than 28% in 97% of cases) and 25.5% in females (less than 28% in 90% of cases); and gular width averages 20.0% in males (more than 18.5% in 93% of cases) and 19.4% in females (more than 17.5% in 90% of cases); Bill Williams, lower Colorado, Gila, Sonora, Magdalena, Yaqui, southwest New Mexico, and Casas Grandes basins K.s. sonoriense 2B. Interanal seam length averages 14.4% of carapace length in males (less than 16% in 90% of cases), and 18.5% in females (less than 22% in 100% of cases); interfemoral seam length averages 12.8% of carapace length in males (more than 10% in 100% of cases) and 13.5% in females (more than 11.5% in 91% of cases); maximum first vertebral width averages 28.9% of carapace length in males (more than 28% in 90% of cases) and 28.8% in females (more than 26% in 100% of cases); and gular width averages 17.7% of carapace length in males (less than 19% in 100% of cases) and 17.8% in females (less than 20% in 100% of cases); Rio Sonoyta basin, Arizona, and Sonora, Mexico K. s. longifemorale 3A. Nasal shield reduced to crescent- shaped scale lying anterior to level of orbits; dark reticulate head markings reduced or nearly absent; plastral width at humero-pectoral seam (PWA) averages 33.3% of carapace length in males (less than 35.5% in 100% of cases) and 37.0% in females (less than 40% in 93% of cases); bridge length averages 20.3% of carapace length in males (over 18% in 100% of cases) and 25.3% in females (more than 22% in 100% of cases); gular length averages 11.8% of carapace length in males (less than 13% in 93% of cases) and 14.5% in females (less than 18.5% in 100% of cases); forelobe length averages 30.5% of carapace length in males (less than 33.5% in 100% of cases) and 31.8% in females (less than 34% in 100% of cases); interhumeral seam length averages 14.0% of maximum plastron length in males (more than 12% in 93% of cases) and 12.7% in females (more than 10% in 88% of cases); interabdominal seam length averages 28.6% of maximum plastron length in males (more than 26% in No. I Kinosternon Biosystematics 57 93*^0 of cases) and ZQ.S'Vo in females (more than 25. 5 "^o in 10024 months. Reproductive condition and length of the spawning season were determined by field observations and by determination of female gonosomatic indices (GSI). The GSI is the ratio of gonad weight to corrected body weight. Corrected body weight is total weight minus the viscera and gonad weight (Mathur and Ramsey, 1974). Fecundity is defined as the number of ova equal to or exceeding 0.2 mm diameter. A large number of ova less than 0.2 mm diameter were present in each ovary, but past studies of darters with retracted spawning seasons (Fahy, 1954; Winn, 1958a; Scalet, 1973) have suggested that these minute oocytes never differen- tiate into fully yolked, enlarged ova and were, therefore, not spawned that year. The smallest differentiating ova of the larger egg group was 0.2 mm diameter, so this size was used as the lower Hmit. For food and feeding studies, whole stomachs were removed, and the contents were identified to family and enumerated. RESULTS AND DISCUSSION Habitat Etheostoma coosae adults and juveniles were consistently collected over rubble in raceways and around boulders, near sand bars and occasionally in the foot of rif- fles. This habitat preference was main- tained seasonally with no indication of age or size specific habitat utilization for foraging or reproduction. The basis of habitat selection by darters is influenced, if not determined, by phys- iological and/or ecological requirements of the species. Ultsch et al. (1978) con- ducted a series of critical O2 experiments with six species of Etheostoma and ob- served that four ecological groups exist with respect to oxygen requirements ver- sus habitat selection. They suggested that one such group, typified by E. (Ulocentra) duryi and E. (Catonotus) flabellare, preferred relatively fast water but main- tained its ability to tolerate periods of hypoxia. This group was the most diverse physiologically in terms of oxygen use strategies. As a result of this, these darters maintained a diverse array of habitat types. The applicability of this expla- nation to habit selection by E. coosae Ues in the close phylogenetic and ecological relationships between it and E. duryi. Demography Etheostoma coosae was the dominant percid species in Barbaree Creek. It com- prised 5.9 percent of the total number of fish specimens collected (Table 1). The overall age class distribution of E. coosae for the year studied is seen in Table 2. Approximately 64 percent of the popula- tion occupied the - 1 age class, 29 percent the 1 + age class, and 7 percent the 2 + age class. Seasonal changes in age class compo- sition (Figure 1) indicate that maximum contribution to population size occurred during winter in the - 1 age class as it approached 12 months of age. From this point, percent contribution to population size declined throughout the older age classes. Of the 750 specimens examined, 32.9 percent and 55.6 percent of the males and females, respectively, survived from the - 1 to the 1-1- age class, whereas 10.8 percent and 10.3 percent of the males and females respectively, survived from the 1 + to the 2-1- age class (Table 2). The overall sex ratio, 1:1.3, was signifi- cantly different (X ' = 9.86; p < .01) from the expected 1:1. This skewed sex ratio was most evident in the 1 + age class, 1:1.8 (X' = 18.96; p <.01), whereas the - 1 and 2 -^ age class sex ratios were not statistically different from 1:1. Age and Growth The oldest individuals collected, two females and one male (Figure 2), were 36 No. 1 Etheostoma Life History 77 Table 1 . Percent relative abundance and frequency of occurrence of fishes collected in Barbaree Creek from April 1977 through April 1978. Species Abundance Occurrence Family Cyprinidae Campostoma anomalum Notropis asperifrons Notropis callistius Notropis lirus Notropis stilbius Notropis trichroistius Notropis venustus Notropis xaenocephalus Phenacobius catostomus Semotilus atromaculatus Family Catostomidae Hypentelium etowanum Moxostoma duquesnei Family Ictaluridae Ictalurus natal is 0.04 8.3 Family Centrarchidae Ambloplites rupestris Lepomis cyanellus Lepomis gulosus Lepomis macrochirus Lepomis megalotis Micropterus coosae Micropterus punctulatus Family Cyprinodontidae Fundulus stellifer 0.02 25.0 Family Percidae Etheostoma coosae Etheostoma jordani Etheostoma stigmaeum Percina caprodes Percina nigrofasciata Family Cottidae Cottus carolinae Total 1.69 91.7 5.87 100.0 8.81 100.0 0.19 25.0 2.97 75.0 45.62 100.0 0.34 41.7 16.90 100.0 0.01 8.3 0.09 25.0 2.00 100.0 0.39 66.7 0.07 41.7 0.01 8.3 0.07 33.3 0.04 8.3 0.76 66.7 1.11 91.7 0.06 25.0 5.93 100.0 1.18 100.0 0.97 100.0 0.31 66.7 0.96 91.7 2.94 100.0 99.35 78 Tulane Studies in Zoology and Botany Vol. 23 Table 2. Age-class distributions and survival of Etheostoma coosae collected in Barbaree Creek from April 1977 through April 1978. S, and S2 equal survival calculated from the - 1 and 1 + age classes, respectively. Sex Year Number of specimens Surviva 1 class S, S, -1 231 1.000 _ 1 + 76 0.329 1.000 2 + 25 0.108 0.329 -1 252 1.000 — 1 + 140 0.556 1.000 2 + 26 0.103 0.185 -1 483 1.000 — 1 + 216 0.447 1.000 2 + 51 0.106 0.236 Males Females Combined Sexes months of age assuming a May hatching. Each specimen had two annuli and the third was in the process of being estab- lished. Scale studies of E. coosae in Barbaree Creek indicated that annuli are established in early to middle spring. Males grew faster than females, and were on the average significantly longer (p <.05) by the first spawning season (Figure 3). Females attained 70.3 percent and 90.0 percent of their average maxi- mum standard length (40.5 mm) in 12 and 24 months, respectively, whereas males Eq _^a_^B_ M SUMMER Figure 1 . Seasonal changes in age-class composition of Eiheosloma coosae collected in Barbaree Creek. Age-class designations are: - 1 , 1 to 1 2 months; 1 -i- , 13 to 24 months; and 2 + , >24 months. attained 69.5 percent and 94.5 percent of their average maximum standard length (43.1 mm) in 12 and 24 months. The longest male and female were 47.0 mm and 44.1 mm, respectively (Figure 2). There were no significant differences (p >.05) between male and female length- weight equations as tested by analysis of covariance (Fg = 0.867, df = 1,136). The relationship for combined sexes was log wt. (grams) = 3.1657 log SL -5.1200, N = 140, r = .978. Page and Schemske (1978) stated that one possible factor affecting size in darters of the subgenus Catonotus was interspecific competition. They concluded that maximum size in sympatric popula- tions of Catonotus was determined by the presence or absence of Etheostoma squamiceps. In the presence of E. squaini- ceps, other Catonotus species were re- duced in size, whereas in the absence of £. sqamiceps their maximum size was greater. They suggested that the major function of the size differences was for size-specific utilization of potential repro- ductive habitats. The overall growth trend of E. coosae was for rapid growth, both sexes reach- No. 1 Etheostoma Life History 79 ing approximately 70 percent of the aver- age maximum standard length, the first year of life, with a subsequent reduction of this rate in later years (Figure 3). This pattern is quite common in darters (Page, 1974 and 1975; Page and Burr, 1976; Starnes, 1977) and in fishes generally. As Ricker (1971) has pointed out, this phe- nomenon is usually attributed to physio- logical size limitations primarily influ- enced by the heavier reproductive effort by older individuals. Reproduction Female gonosomatic indices (Figure 4) and field observations indicate that E. coosae spawned from mid March through early to mid May with peak spawning in April. The spawning periods for species of the subgenus Vlocentra are similar. Winn (1958a, 1958b) reported that males of Etheostoma sp. (Barren River form) in Tennessee established territories near the beginning of April, and spawning began in about one week. Stiles (1975) reported that E. simoterum spawned from early E E. I »- O z lU -* O et < O z < ^ A-<^ A ^ d^ A A A o A o OV A A A A A O OA A O A o o Oft A A A OA o o o o o *-? o O CA c* <> o o o o o A <) (TA () o o o o o O O OA (> •o o o A-? A G* O O O o •o » o o o o O O o • •o K) o o o O* O o •o •o o o o O o •o • o QA O •u • •o o O • • •u • o O «> • • o m> • • o O O 1 •o • • • • •o • • • • • • • o • • • • • ? MALE FEMALE m'j'j'a's'o'n'd'j'f'm'a 1 • — 1+ o 2+ A — 19 20 21 31 32 33 MONTHS OF AGE 22 23 34 35 Figure 2. Length-age relationship of Etheostoma coosae collected in Barbaree Creek. n1.0 Age 40.1 194 156 72 422 2 38.5 300 116 80 496 2 36.0 278 56 4 330 2 35.9 320 118 54 492 2 34.5 284 92 34 410 33,7 220 92 42 354 33.5 172 78 38 288 31.5 280 46 22 348 30.5 260 39 0 299 Mean 256.4 88.0 38.4 382.8 Percent 67.0 23.0 10.0 100.0 No. 1 Etheostoma I ife History 81 the anterior dorsal and ventral interradial membranes of the caudal fin. The inter- radial membranes throughout the length of the spinous and soft dorsal fins had rusty-red quadrate spots. Females main- tained a ground color of light tan overlaid by brown to black lateral blotches and mottling above the lateral line. The total egg complement increased proportionally with length, r = .682 (Table 3). Fecundity studies of other darters have substantiated positive size-fecundity rela- tionships: E. squamiceps, r = .692 (Page, 1974); E. barbouri, r = 530 (Flynn and Hoyt, 1979); E. kennicotti, r = .631 (Page, 1975); and Percina nigrofasciata, r = .721 (Mathur, 1973). An exception to this general relationship was reported for E. proeliare (r <.l) by Burr and Page (1978). This vagary was attributed to the short life span (one year) of the popula- tion studied, which yielded females of a similar size. Feeding The overall diet of E. coosae, consisted of 78 percent Diptera (Chironomidae and Simuliidae), 12 percent Crustacea (Cope- poda and Cladocera), 3 percent Ephemer- optera (Baetidae and Siphlonuridae) and 5 percent miscellaneous items (Acarina, Mollusca, Nematoda, Trichoptera, and sand). The diet of various size classes as well as the seasonal diet of combined age and size classes is seen in Figure 5. Etheostoma coosae consumed midge larvae as juveniles and expanded their diet as adults to include mayflies and caddis- flies. Midge larvae decreased whereas crusaceans increased in importance from spring to winter. Mayflies, caddisflies, and molluscs were important items during summer months. 11-20mm N = 6 21-30nnm N = 78 31-40mi N = 143 41-50mi N = 25 jf, Misc E|)h SPRING N=25 SUMMER N=97 WINTER N=34 Figure 5. Stomach contents of Etheostoma coosae collected in Barbaree Creek by size class of darter and season collected. Seasonal analyses include all size and age classes. Food items are abbreviated as follows: (Crus)tacea, (Dip)tera, (Eph)emeroptera, (Mis)ellaneous, (Mol)lusca, (Plec)optera, and (Tri)choptera. N equals sample size. 82 Tulane Studies in Zoology and Botany Vol. 23 The feeding mode of darters has been reported to be largely selective in some species: E. fonticola (Schenck and Whiteside, 1977), E. nigrum (Roberts and Winn, 1962), and E. radiosum cyanorum (Scalel, 1972); and largely opportunistic in others: E. acuticeps (Bryant, 1979), E. blennioides (Fahy, 1954), and E. gracile (Braasch and Smith, 1967). These papers have illustrated that within the genus Etheostoma feeding behaviors are quite variable and complex. Based on the liter- ature and my own studies, I believe that feeding behavior is not so restrictive but rather lies along a dynamic continuum between selectivity and opportunism. Species will adapt to prey abundance and type assuming the most energetically re- warding feeding response. Prey switching as a possible behavioral mechanism involved in feeding is supported by Ihe studies of Murdoch et al. (1975) on Poecilia reticulatua and Roberts and Winn (1962) on the role of visual cues in the feeding of £. nigrum. ACKNOWLEDGEMENTS 1 wish to thank Dr. Herbert Boschung, Dr. Maurice F. Mettee, and John Williams who read and discussed various parts of this paper; the USDA Forest Service for a grant to Boschung from which this study was funded; and finally Irene Thompson who performed all typing services on the manuscript. Literature Cited BOSCHUNG, H. and P. O'NEIL. 1980. The effects of forest clearcutting on fishes and macroin- vertebrates. U.S.D.A. Forest Service, Atlanta, Georgia. 32 pp. BRAASCH, M. and P. SMITH. 1967. The life history of the slough darter, Elheostoma gracile (Pisces: Percidae). 111. Nat. Hist. Surv. Biol. Notes No. 58 12 pp. BRYANT, R. 1979. The life history and compar- ative ecology of the sharphead darter, Etheo- sioma acuticeps. Tennessee Wildlife Resources Agency Technical Report No. 79-50. 60 pp. BURR, B. and L. PAGE. 1978. The life history of the cypress darter, Elheostoma proeliare. in Max Creek. 111. Nat. Hist. Surv. Biol. Notes No. 106. 15 pp. FAHY, W. 1954. The life history of the northern greenside darter, Etheostoma blennioides. Elisha Mitchell Scientific Society 70:139-205. FLYNN, R. and R. HOYT. 1979. The life history of the teardrop darter, Etheostoma barbouri Kuehne and Small. Amer. Midi. Nat. 101:127- 141. MATHUR, D. 1973. Some aspects of life history of the blackbanded darter, Percina nigrofasciata, in Halawakee Creek, Alabama. Amer. Midi. Natur. 89:381-393. . and J. RAMSEY. 1974. Reproductive bio- logy of the rough shiner, Notropis baileyi, in Halawakee Creek, Alabama. Trans. Amer. Fish. Soc. 103:88-93. MURDOCH, W., S. AVERY, and M. SMYTH. 1975. Switching in predatory fish. Ecology 56:1095-1105. PAGE, L. 1974. The life history of the spottail darter, Etheostoma squamiceps, in Big Creek, Illinois and Ferguson Creek, Kentucky. 111. Nat. Hist. Surv. Biol. Notes No. 89. 20 pp. . 1975. The life history of the stripetail darter, Etheostoma kennicoiti, in Big Creek, Illinois. 111. Nat. Hist. Surv. Biol. Notes No. 93. 16 pp. . and B. BURR. 1976. The life history of the slabrock darter, Etheostoma smithi, in Ferguson Creek, Kentucky. 111. Nat. Hist. Surv. Biol. Notes No. 99. 12 pp. PAGE, L. and D. SCHEMSKE. 1978. The effects of interspecific competition on the distribution and size of darters of the subgenus Catonotus (Percidae: Etheostoma). Copeia 1978:406-412. RICKER, W. 1971. Methods for assessment of fish production in freshwater. Blackwell Scientific Publications, Oxford, England. 348 pp. ROBERTS, N. and H. WINN. 1962. Utilization of the senses in feeding behavior of the johnny darter, Etheostoma nigrum. Copeia 1962:567- 570. SCALET, C.G. 1972. Food habits of the orange- belly darter, Etheostoma radiosum cyanorum (Osteichthyes: Percidae). Amer. Midi. Natur. 87:515-524. . 1973. Reproduction of the orangebelly darter, Etheostoma radiosum cyanorutn. Amer. Midi. Natur. 89:156-165. SCHENCK, J. and B.C. WHITESIDE. 1977. Food habits and feeding behavior of the fountain darter, Etheostoma fonticola (Osteich- thyes: Percidae). Southwest Nat. 21(4):487-492. STARNES, W.C. 1977. The ecology and life history of the endangered snail darter, Percina (Imo- stoma) tansi Etnier. Tenn. Wildlife Resources Agency Tech. Report No. 77-52. 143 pp. STILES, R.A. 1975. The reproductive behavior of Etheostoma simoterum (Cope)(Perci formes: Percidae). Program Abstracts, American Soci- ety of Ichthyologists and Herpetologists 55th Annual Meeting, June 8-14, Williamsburg, Vir- gia. p. 121. (Abstr.) No. 1 Etheostoma Life History 83 ULTSCH, C, H. BOSCHUNG, and M. ROSS. 1978. Metabolism, critical oxygen tension, and habitat selection in darters (Etheostoma). Ecology 59:99-107. WINN, H. 1958a. Comparative reproductive be- havior and ecology of fourteen species of darters (Pisces: Percidae). Ecol. Mongr. 28:155-191. 1958b. Observations on the reproductive habits of darters (Pisces: Percidae). Amer. Midi. Natur. 59:190-212. December 30, 1981 84 Tulane Studies in Zoology and Botany Vol. 23 THE TAXONOMIC RELATIONSHIP BETWEEN MALACLEMYSGKM , 1844 AND GRAPTEMYS AGASSIZ, 1857 (TESTUDINES: EMYDIDAE) JAMES L. DOBIE Department of Zoology-Entomology Auburn University, Alabama 36849 Abstract The turtle genus Graptemys is a distinctive group clearly separable from Malaclemys on the basis of external and osteological features. The difference between the groups indicate that the degree of genetic relationship is no closer than that resulting from their both having presumably arisen from a Pseudemys - like stock or Malaclemys from a Graptemys stock. INTRODUCTION Investigators of Malaclemys and Grapt- emys have based their taxonomic alloca- tions on penial, skull, shell, hind limb and pelvic girdle morphology and on head patterns. Osteological comparisons, when indicated, were usually limited to the skull, and in most cases, head patterns were used to distinguish taxa. The degree of evolutionary conservatism and paral- leUsm exhibited by turtles argues against the use of external characters (e.g., head striping), alone in determining taxonomic and phylogenetic relationships. Thus, both osteological and surficial features have been examined in this study. HISTORICAL REVIEW The controversy about the relationship between Malaclemys and Graptemys be- gan as a resuh of the lumping of Grapt- emys with Malaclemys by Boulenger (1889) and the re-establishment of the genus Graptemys by Baur in 1890. Since that time, W.P. Hay (1904) and O.P. Hay (1908) followed Baur in recognizing the two genera, as did Carr in 1949. Later, however, Carr (1952) questioned the validity of separating the two genera and McDowell (1964), without presenting sup- porting data, lumped Graptemys with Malaclemys. Zug (1966, 1971), on the basis of similiar penial, pelvic girdle, and hind limb morphology for the two genera considered them congeneric, and Parsons (1960, 1968) found the choanal structures of both genera to be so variable that the evidence did not particularly support or refute the congeneric idea. Several other authors (Ernst and Barbour, 1972; McKown, 1972; Dundee, 1974; Killebrew, 1979; Dobie and Jackson, 1979; Pritch- ard, 1979; Vogt, 1978, 1980) have not supported the synonymy of Graptemys with Malaclemys; they evidently must believe that sufficient evidence has not been presented to lump the two genera together. The purpose of this study is to clarify the generic status of Malaclemys and Graptemys. MATERIALS AND METHODS Representatives of each of the ten ex- tant Graptemys species (Vogt, 1980) and their subspecies and individuals of several subspecies of the monotypic Malaclemys were examined. External features, includ- Editorial Committee for this Paper: Dr. Eugene S. GAFFNEY, Associate Curator, Department of Vertebrate Paleontology, American Museum of Natural History, New York, New York 10024 Dr. John J. IVERSON, Assistant Professor of Biology, Earlham College, Richmond, Indiana 47374 85 86 Tulane Studies in Zoology and Botany Vol. 23 ng scute contracts, plastral patterns, and striping on the head and leg were analyzed in juvenile and adult turtles of both sexes. Skull and shell characters were analyzed on large sub-adult and adult females. Skull terminology is that of Gaffney (1972 a); scute and bone terminology is that used by Zangerl (1969). The method used to elucidate the rela- tionship between Malaclemys and Grapt- emys and to other North American emy- did turtles is the search for taxa that have shared derived characters. This method was described by Hennig (1966), and has been used by others (Gaffney, 1972 b, 1975; W.E. Clark, 1978) and is called phylogenetic systematics or cladism. DIAGNOSTIC CHARACTERISTICS The diagnostic characteristics of Grapt- emys, Malaclemys and an outgroup com- parison of those genera with the other North American emydid genera are listed in Table 1 . Each feature is also designated as either ancestral (primitive) or advanced (derived). SIGNIFICANCE OF DIAGNOSTIC CHARACTERISTICS The number (s) in a bracket refers to the number of the diagnostic features in Table 1. SKULL FEA TURES (1) Quadratojugal - maxilla contact. If the absence of contact between these two bones represents the primitive state, then the possession of the derived condition in three Graptemys species (in one pseudo- geographica and in all pulchra and bar- bourij, in M. terrapin, and in some Pseudemys species suggests that M. terrapin could have been derived from one of these Graptemys or Pseudemys species. Graptemys could have come from any group lacking contact between the two bones. (2) Spoon-shaped symphysis of lower jaw (Fig. 1). The flattened spoon-shaped nature of the symphyseal part of the lower jaw apparently is a derived feature in Graptemys. The absence of such a struc- ture in Malaclemys suggests that Grapt- emys was not ancestral to Malaclemys and that Malaclemys may have arisen from some Pseudemys species. Figure 1. Shape of the symphyseal area of the lower jaw in mature females of (A) Malaclemys ter- rapin. (B) Graptemys pseudogeographica, (C) G. geographica, (D) G. pulchra, (E) G. barbouri, (?) G. caglei, (G) G. versa, (H) G. ouachitensis sabinensis, (I) G. o. ouachitensis, and (J) G. flavimaculata (the shape of the symphysis is the same for flavimaculata, oculifera, and nigrinoda). No. 1 Malaclemys- Graptemys Relationship 87 (3) Bones surrounding the foramen pala- tinum posterius (Fig. 2). The bones sur- rounding that foramen in Terrapene and in the species of the Pseudemys rubriven- tris complex are the same as Graptemys; the other species of Pseudemys and the other N.A. emydid genera are like Mala- clemys. Therefore, Graptemys and Mala- clemys were possibly derived from differ- ent species of Pseudemys. (4) The absence of contact between the ophisthotic and pterygoid due to the in- volvement of the exoccipital. If the condi- tion in Malaclemys and Deirochelys repre- sents a derived feature, this would strong- ly suggest that Malaclemys was not the ancestral stock from which Graptemys evolved. It could also indicate that a Graptemys, Deirochelys, or any other species of North American emydid turtle could have been ancestral to Malaclemys. (5) The lack of a notch in the premaxil- lary bones. The lack of a notch in those bones in Graptemys and the presence of a notch in Malaclemys and the rest of the N.A. emydids, precludes determination of the possible ancestor for Graptemys and Malaclemys based on this feature. SHELL FEA TURES (6) Flaring of carapace. The presence of such in Graptemys and to varying degrees in all other N.A. emydids except Mala- clemys and some Terrapene, may indicate that flaring is an ancestral feature. If so, the upturning of the carapace in the last two genera would be a derived feature. This implies that Graptemys did not come from a Malaclemys stock. (7) Double notching of some peripherals. The double notching of some of the per- ipherals is found only in Graptemys and in some individuals of Pseudemys scripta and P. concinna. This could indicate that Graptemys was not ancestral to Mala- clemys and that a Pseudemys species was ancestral to Graptemys. (8 and 9) The keel and its associated bosses (Fig. 3). A number of reports have dealt with the extent and development of the keel in Malaclemys. The last vertebral scute is variable with respect to keel devel- opment. Say (1825) reported that the last vertebral in M. terrapin centrata was un- keeled; Wied (1865) noted that all of the vertebrals of M. t. pileata have a well developed keel. The keel in Malaclemys t. centrata was stated by W.P. Hay (1904) "to be rather low and rounded," whereas it was "always well developed," in M. t. macrospilota. A keel is thus not always present on the last vertebral, and I have not observed the end of the keel (the fifth boss area) to extend more than four-fifths the length of the last vertebral scute. W.P. Hay's (1904) statement about the keel and bosses of M. t. littoralis was: "the first vertebral plate is raised on the middle line to form a broad, low carina; on the second plate the elevation is greater, and stands out as a smooth boss . . . ; the elevation on the third plate has the form of a hemispherical button with a well- marked constriction around the posterior half of the base . . . ; on the fourth plate the elevation is raised into a knob-like protuberance from a base which is con- stricted all around . . . ; the fifth vertebral plate is flat or with only a trace of an elevation." Thus Hay's statement suggests that four or five bosses are present on the keel in Malaclemys. This is not always the case. Auburn University Museum of Paleontology (AUMP) speci- men 2179 has only three bosses, and its shell structures are normal. Concerning the total number of bosses on the keel in Graptemys pulchra, Carr and Coin (1955) said, "the dorsal keel . . . comprises a boss on each of the first four centrals, . . . weak to nearly lacking on the first and completely lacking on the fifth." A boss on the fifth central (verte- bral) is not lacking \n pulchra. Although it is not prominent in G. pulchra or in any other species of Graptemys, a terminal boss can be detected in all species. Cagle (1954), p. 182, Fig. 11) illustrated a juvenile G. flavimaculata that had five bosses on the carapace. I have never examined any specimen of Graptemys, including G. flavimaculata, in which the 88 Tulane Studies in Zoology and Botany Vol. 23 §i1 S -a E > c . — Q tS; E I >< ?i "^ cxo. .S E - o o ^ E -o "£^ £li O '^ r\ to ^ E |g ^ 5 o ? S Cu ^^ 5^S ;3 ■53 O •& 3 s: % 1 most !ome emys •^ 3 13 C ^ 1 doide and i u ^»«K g b '^* c X O ^ b S ^*^ M Uj 5 "^ i c .E|-S 1 i ^"21 O g o 2> C JJ ^ c lU b 'u •- E _aj o X3 ^ c« HJ p 3 O CU E ££ ^^ D, S o 18 >J c« i^ -b lr> S-Si ^ (0 o ^ U J.. c Q ><• ^-^ c^- E CL ^— o 00 C .^ o c w o ■ u (fl eO 00 3 3 O 3 C/5 C8 co tn^ 0 w XJ ~ u O E-§ 3 2 C lU ^ S 5 "■= 15.2iS _ 1> D. S E >. 1 ^t 2 o S o ^^ - S o c X o ti -a — " w c ^ — — o ° s 2 = * ci -c "5 2 "5 nJ 00 C ^ « >^ o — .2 UJ D..E 7^ (Ll — ^^ C k. L. U E I IJ c ^ T3 s: ^ iE U Sb T3 3 ^ • C o ? 00 w •^ c _>. 0:e "w 15 _ o E C w 1) — o O Q. ^ => 1> 13 00 C 3 T3 <-» C » X c E 3 B3 O 1/5 T3 0^ 00 u .-.^ Ji n. "" "O E — ?; « II ll «> U O JO 9 c E - c — o o -c _, — (U I) u- W > " c3 _[- E£ X o — ; 1> .2 t;; x c " ■a o D Q. w O u. IJ rt 3 ul 2 -= I" >, ii .^ ^ — C "r 2 E - -c J£ 3 J2 ,::; ~ •- "^ ^ i! iJ g w 1> o Ei: 2-S _ (U ^ CO -- -- o « E •- _ i_ t w. T3 nJ O w. C U C O IJ O No. 1 Malaclemys- Graptemys Relationship 89 5 g S E .E 5 ? -. '^ 1^ « -Q o 5 Q, s ^' ^ ej - ■Q « W (U -u _ T- • E O i^.E^^ •- ^ ^ ^ 5 ^ £ ■- E ^ °"c.E2 o 0 T3 pj — m ^S o S . 2 -S « ^ O -^ g Ci. -Si "t« ~ -s; s - I .5 5 § ,. ^ o^^ g o E-a Z Q; E 8 a S rt •i S 2 o O — "^ TS ci c« a 2 o cs — T3 =*- O "O D 1> ™ — ^ E c^Q ui X) X) 3 c o c 1) £ ^ X) -o 15 x: XI IS E c _>v o XI 3 "rt _ C E o (0 O CO K o c C o x: D. o u HJ ;a O o a « _>. (/^ "5 D. ■« k- U E aj 1) ca W x: 3. o ■a c "o c u u > O o ;o ■§:5 1^ 2 S3 ^.E ^ C 2, «j 5 2.- gQ S E ^ S o -Si E E ^ •a c s: c — := g| i 2 c g u- -^ r* H Z -r> Cu u ^^ o s? Z ■a ^^ ZO § o Z Si 3 3 "5 1> 03 ii - 5 ^:! uo C O Du X) ;^ r: Q o 3 i_ o o -o XI c aj nJ c • .1 15 :i 5 ts = •- Z D. Cu O > C r ° I S i_ CIS a ^G «^ ^ 2! o Z ? • £ S?5 E E iS 2 a — CO c« o t. Z c .4/ •- > 3 O JJ 3 2 ° 3 XI J 15 <^ X3 cjo 3 "O O 15 -o a> D. 31§ •a y ~'CU C r~ "^ X) age ^ JC CO -eS^ ^Q x: ^ Ai 3 .E 3 . T3 CO — ?s 3 !>, ^ [^ C C X> _ o o _- .E o CO ; - x: i> ° P 3 O Z 5 c X) c ■2 § O '-> X) TO ^J «lo Q U (U u ^ r c < > Xi 1> CO 3 "Z <■ I =^ s C -C a| . CO ? ^ ~ — ? 1» CO ^ c .^ ? 3 Q fc ^ iT '^ id- 3 =^ __ C CO CO Cb c/5 o OS S O I' I o II w- 'o aj CO 00 T3 J= c c cj aj 30^. 't* j= — Ci Si S U I ■a 2 § ^S. o •- -a cu o 5; 1^ p — 1^ 5 0^ L^ I* js K a> O 3 00 ^ •- c5 S O. o -s: y o d 3 ^-CIO 3 ^ ° ^ .5 -c o ^ ECO -a o c TO ^ E c 3 J E < 55 3 c < b; 3 a.) Ci, in f:: s c e8 O 00 D. 3 ^1 a> CO ts '-> _ o^ S 2 fc 15 ^'^ — x: E 13 :2 Q j: x: -a ■53 ° u .t; > 15 Q 0 :Ei ap_ c 3 ■53 CO 0 E CO 0 g 0 = „^ 2 -= CO c .£ E 0 60 c/^ U 'C s a •g x:=ii-scj='-i ^S6 O JJ E o § •^15 E ^B >:^ 3 I- CO XI ^ C a> w <— x: ^ 3 ^ -5 g E 5:5 c c ^ g -g 5 § T3 « .2 O 1) 1) -15 " tJ x: g .b .1-' 2 0 - - t: ^. b o o CO o ^ 2 .i 3 O 1- aj a> ^^ — •£ - j= ■£ £ 00 c cu _ 3 ^ 00— •- ^ ■Sl5_^ y .SP'Clo 3 ^ u aj 3 ■ - x: j= — OJ w- tU c 90 Tulane Studies in Zoology and Botany Vol. 23 5 S, c a c o X> .2 > ^ S o I'll '3 '^ ^ o O t3 J| = '- C ^ s -Si o -S ce; in o S ^•5 1 s' o X X) ^ >> c 5 ^-^ E'Z CO i> c? (U OD X ^ o « — ■5'ZS d *^ C CJ 3 O id S o s 5 "? - 1o 5 A — CO w-i 2 w A 1/1 m CO ■^^ /^ = En u CO 1^ ifl ^ U 1) > — 3 X T3 A ' A (U — •5 A CO CO =: CO c 2 o Q ^ o o i: 1/, c 5 ^ (> o CO Q. CO C CTJ CO ,, U o ti t3 S - CO E I I s x> -a CO E 2i b a; ii.E .5 a^ 11 -s: .E S- 5)Q <-s E 5 u CO "T 3 /^ J2E§ 3 O '-' ^■^ O £ CO c- - V b S lo ^ CO — o .y CJ Q.10 5. V >^ — H CO V PD.5? cog !=;-^ -O (U o CO CO c '5b E « 2 E- C ^ CO on CJ U c S 1> CO > D c .ti < T3 V A eg "w •^-^ iS Xi c^- 3 CO CU E V o o — ■ w 00 CO CJ CO o il^ .e| E O in i> CO r> u- •5 J> 6- (U Tj _ 3 O -2 o «^ •E -JQ E c lu ;^• ■Op -^ .5 — 3 3 CO CJ t; c ^^ •-so >> ^ F. H •= 75 S.2? ° J o b S^^ § J E E .5 CJ c^. 5 o r^ CO < Ji 00 CJ .5 o >. c u CO w 2 y ?> "cO „, J D CO T3 '^ CO « -S a "listed II; -. o - C OT) .S CO o« D. "^ -^ o .^ ^ 1> T3 nj •o 3 U5 I + + + I + + + + + +-H + + ++ M + + + + a T3 o % c o 9> es c« 1) u ■o C ^F tS o *_, p.-o o a j: =S-S J2 u T3 ■a 2 S o 2 3 _^ _ >" T3 "5 £? .£ =5 3 "" 00 >> c 13 o j= * a. X! ■o tj F o ni o c QU o >. n Lh O c >< o W X3 O 3 c t5 C "3 '3) iS CO 3 UU0Z.-O/0Z. T&- Volume23 . Number 2 $3.50 ^ ,, .D^eember 15, 1982 OtC2 0ii^82 MARvArxD I lMI\/irR~lTY CHANGES IN MELANIN MIGRATION INDUCED BY NORADRENERGIC AND HISTAMINERGIC AGENTS IN THE FIDDLER CRAB, UCA PUGILA TOR MUKUND M. HANUMANTE AND MILTON FINGERMAN p. 103 ADDITIONAL TREMATODES OF MAMMALS IN LOUISIANA WITH A COMPILATION OF ALL TREMATODES REPORTED FROM WILD AND DOMESTIC MAMMALS IN THE STATE WESLEY L. SHOOP AND KENNETH C. CORKUM p. 109 COMPARATIVE VISCERAL TOPOGRAPHY OF THE NEW WORLD SNAKE TRIBE THAMNOPHIINI (COLUBRIDAE, NATRICINAE) NITA J. ROSSMAN , DOUGLAS A. ROSSMAN and NANCY K. KEITH P- 123 TULANE UNIVERSITY NEW ORLEANS TULANE STUDIES IN ZOOLOGY AND BOTANY, a publication of the Biology Department of Tulane University, is devoted primarily to the biology of the waters and adjacent land areas of the Gulf of Mexico and the Caribbean Sea, but manuscripts on areas outside this geographic area will be considered. Each number contains an indivi- dual monographic study or several minor studies. Normally two numbers plus an index and a table of contents are issued annually. Preferred citation of the journal is Tulane Stud. Zool. and Bot. INFORMATION FOR AUTHORS: Manuscripts submitted for publications are eval- uated by the editors and by an editorial committee selected for each paper. Contrib- utors need not be members of the Tulane faculty. Manuscripts of 20 or more pages, double-spaced, are preferred. We recommend conformance with the principles stated in CBE Style Manual, 4th ed., 1978. Manuscripts should be typewritten and double spaced. Two additional copies should accompany the original to expedite editing and publication. Legends for figures should appear on a separate page and in sequence. Illustrations should be proportioned for one or two column width corresponding to our printed page size, and should allow for insertion of the legend if occupying a whole page. Guidelines for letter and other extraneous markings should be done with a non-photo blue pencil such as Eagle Prismacolor. Photographs should be on glossy paper. Many tables, if carefully prepared with a carbon ribbon and electric typewriter, can be photographically reproduced, thus helping to reduce publication costs. Lettering in any illustrative or tabular material should be of such a size that no letter will be less than 1 V2 mm high when reduced for publication. An abstract not exceeding three percent of the length of the article must accompany the manuscript. Separates of published articles are available to authors at a nominal cost. Page charges, calculated at $45/page, are solicited from authors who have funds for this purpose through their institutions or grants. Acceptance of papers is not dependent on ability to underwrite costs but excessive illustrations and tabular matter may be charged to the author. EXCHANGES, SUBSCRIPTIONS, ORDERS FOR INDIVIDUAL COPIES: Ex- changes are invited from institutions publishing comparable series. Subscriptions are billed in advance. A price list of back issues is available on request. Individuals should send their remittance, preferably money order, along with their orders. Remittances should be made payable to "Tulane University." Subscription rates: Volume 23. $8.50 domestic, $9.50 foreign. Copies of Tulane Studies in Zoology and Botany sent to regular recipients, if lost in the mails, will be replaced if the editorial offices are notified before the second subsequent issue is released. COMMUNICATIONS: Address all queries and orders to: Editor, TSZ&B, Depart- ment of Biology, Tulane University, New Orleans, Louisiana 701 18, U.S.A. Harold A. Dundee, Editor TULANE STUDIES IN ZOOLOGY AND BOTANY Volume 23, Number 2 December 15, 1982 CHANGES IN MELANIN MIGRATION INDUCED BY NORADRENERGIC AND HISTAMINERGIC AGENTS IN THE FIDDLER CRAB, UCA PUGILA TOR* MUKUND M. HANUMANTE AND MILTON FINGERMAN Department of Biology, Tulane University New Orleans, Louisiana 70118 U.S.A. Abstract The effects of the H, receptor blocker SA-97, the Hi receptor blocker cimetidine, the tyrosine hydroxy- lase inhibitor a -methyl-para-tyrosine and the H, receptor and norepinephrine uptakci blocker diphen- hydramine on histamine- or 4-methyl histamine-in- duced inhibition of melanin dispersion in the fiddler crab, Uca pugilator undergoing a background trans- fer from white to black were determined. Only cimeti- dine significantly antagonized the 4-methyl histamine- evoked decrease in melanin dispersion. a-Methyl- para-tyroslne by itself significantly diminished where- as diphenhydramine by itself significantly potentiated the amount of this centrifugal melanin migration in the fiddler crabs. None of these drugs affected melanin migration in vitro. The results are consistent with the hypotheses that norepinephrine triggers re- lease of a melanin-dispersing hormone and that H, re- ceptor activation decreases impulse-mediated nore- pinephrine release in this crab. INTRODUCTION Translocation of the melanin in the melanophores of the fiddler crab, Uca pugilator, is regulated by antagonistic neurohormones, a melanin-dispersing hormone (MDH) and a melanin-concen- trating hormone (Carlson, 1935; Sandeen, 1950; Fingerman, 1956). Norepinephrine (NE) triggers release of MDH in this crab (Fingerman et al., 1981; Hanumante and Fingerman, 1981a,b; 1982a,b,c; Hanu- mante et al., 1981). Recently histamine (HA) has been shown to inhibit melanin ♦Supported by Grant PCM-8 1-08864 from the National Science Foundation. dispersion in a dose-dependent manner (Hanumante and Fingerman, 1981b). Use of a variety of histaminergic agonists and antagonists led to the hypothesis that two types of HA receptors, called H, and H2, are present on NE neurons that trigger MDH release and that HA exerts its inhibi- tory action by stimulating the H2 recep- tors. The present investigation was devised to obtain further support for this hypothesis. This objective was carried out by observing the effects of specific mam- malian histaminergic and noradrenergic agents not used previously on the inhibi- tory action of HA and 4-methyl histamine (4-MeHA; a selective H2 receptor agonist, Owen et al., 1979; Douglas, 1980; Polanin et al., 1981) on melanin dispersion in Uca pugilator transferred from a white to a black background. Materials and Methods Adult male fiddler crabs, Uca pugilator, from the vicinity of Panacea, Florida, (Gulf Specimen Company) were used. Their melanophores were staged according to the system of Hogben and Slome (1931) whereby stage 1 .0 represents maximal pig- ment concentration, stage 5.0 maximal pigment dispersion and stages 2.0, 3.0, and 4.0 the intermediate conditions. When intact crabs were used, the melanophores seen through the cuticle on the anteroven- tral surface of the second walking leg on the right side were staged at the time a sub- EDITORIAL COMMITTEE FOR THIS PAPER: DR. RAY W. FULLER, Research Advisor, Eli Lilly and Company, Indianapolis, Indiana 46206 DR. WILLIAM S. HERMAN, Professor and Head, Department of Genetics and Cell Biology, University of Minnesota, MinneapoHs, Minnesota 55108 103 104 Tulane Studies in Zoology and Botany Vol. 23 stance was injected and 15, 30, 60, 90, and 120 minutes thereafter. To facilitate com- parison of the responses of the experimen- tal and control crabs, mean differences between the 15 through 120 minute melanophore stages for the control and ex- perimental groups were calculated for use in Table 1 . The depicted data are based on the mean melanophore stages of 20 intact crabs (10 experimental and 10 control) or 20 isolated legs (10 experimental and 10 control). When assays were performed on isolated legs, the melanophores were staged only at the time the legs were removed from the crab (at which time the legs were perfused with the test or control solution) and 15, 30, 45, and 60 minutes thereafter. The second and third walking legs from both sides of the crab were removed; the legs from the right served as experimentals and the legs from the left side received control solution; the melano- phores on the anteroventral surface of these isolated legs were observed for staging. The assays were performed using isolated legs having initially either maxi- mally concentrated melanin (stage 1.0) or maximally dispersed melanin (stage 5.0). Melanophores in isolated legs of this crab remain responsive for at least 120 minutes (Herman and Dallmann, 1975). The statis- tical significance of the data was deter- mined using Standard Errors of the Means (SEM) the Student's t test with significance set at the 95% confidence interval. None of the data for isolated legs were statis- tically significant. The volume of the solution injected into each crab or isolated leg was always 0.05 ml. The experiments with intact crabs and isolated legs were performed at 24 °C under an illumination of 1190 Ix. 4-MeHA dihy- drochloride (Smith, Khne and French), cimetidine (N"-Cyano-N-methyl-N'-{2- (5-methylimidazol-4-yl) methylthioethyl} guanidine) (Smith, Kline and French) and SA-97 (homochlorcyclizine) (Eisai) were generous gifts. In addition, HA, amethyl- para-tyrosine (a-MPT) and diphenhydra- mine HCl (all from Sigma) were used. The concentration used for each drug, whether injected alone or in combination, was 20 ug/dose of the free compound. All drugs except cimetidine were dissolved in Pantin's physiological saline (Pantin, 1934). Cimetidine was dissolved in acidi- fied (a drop of 1,2 M HCl) saline. Conse- quently, a drop of HCl (1.2 M) was added to control saline for the cimetidine exper- iments. The rest of the controls received pure saline. Results and Discussion 4-MeHA, an H2 receptor agonist, slowed the rate of melanin dispersion, as observed earlier by Hanumante and Fin- german (1981b), in intact crabs transferred from a white to a black background (Table 1). Cimetidine, which selectively blocks mammalian H2 receptors (Douglas, 1980; Polanin and McNeill, 1981) significantly antagonized the 4-MeHA. On the other hand, the H, receptor blocker SA-97 not only did not antagonize the 4-MeHA but the combination of 4-MeHA plus SA-97 resulted in significantly further inhibition. None of these drugs affect melanin migra- tion in vitro nor do SA-97 and cimetidine by themselves have an effect on the rate of melanin dispersion in crabs undergoing a background change from white to black (Hanumante and Fingerman, 1981b), a black background fostering melanin dis- persion (Brown and Hines, 1952) which will be effected by MDH. a-MPT selectively inhibits tyrosine hydroxylase. This enzyme catalyzes the synthesis of dihydroxyphenylalanine from tyrosine. At least in mammals this is the rate-limiting step in the biosynthesis of NE (Terrasawa et al., 1975; Lofstrom and Backstrom, 1978). a MPT by itself signi- ficantly decreased melanin dispersion. HA by itself, as reported earlier (Hanumante and Fingerman, 1981b), sig- nificantly reduced centrifugal melanin migration in intact crabs transferred from a white to a black background. However, in the crabs that were co-administered either 4-MeHA and a-MPT or HA and a-MPT (Table 1), 4-MeHA and HA were not able to produce further, significant No. 2 Melanin Migration in Crabs 105 reduction of the melanin dispersion. Diphenhydramine, a blocker of H, recep- tors and NE uptake, in mammals (Isaac and Goth, 1965; Fantozzi et al., 1975; Marco et al., 1980), by itself significantly enhanced melanin dispersion. However, when HA was co-administered with diphenhydramine, the HA-induced inhi- bition in melanin dispersion was still evident (Fig. 1). The present data, in light of our earlier report (Hanumante and Fingerman, 1981b) and the pharmacological actions of noradrenergic and histaminergic agents in mammals, further strengthen the hypo- thesis that (a) NE serves as a neurotrans- mitter triggering release of MDH and that (b) activation of H2 receptors located on NE neurons which control MDH release results in a decrement of melanin disper- sion in Uca pugilator transferred from a white to a black background. The observations that cimetidine, a selective H2 receptor blocker, antagonized the 4-MeHA-induced inhibition in melanin dispersion, whereas the Hi blocker SA-97 did not, reveal that this effect is mediated specifically by activation of HA H2 recep- tors. The marked increase in inhibitory effect of 4-MeHA when co-administered with the Hi antagonist SA-97 was probably due to the fact that excitation of H, receptors evokes enhanced melanin dispersion (Hanumante and Fingerman, 1981b), blocking them would prevent any endogenous Hi stimulation of the crabs. This would enable 4-MeHA, an agonist of H2 receptors, to produce an even greater inhibition of the melanin dispersion. On the contrary, in the crabs whose H2 recep- tors were blocked by cimetidine, 4-MeHA was unable to significantly decrease the action potential-mediated release of NE, which in turn resulted in a near normal quantity of MDH being released into the hemolymph of these crabs transferred to the black background. The fact that metiamide, another H2 receptor blocker, significantly antagonized the 4-MeHA- stimulated decrease in centrifugal melanin migration (Hanumante and Fingerman, 1981b) in vivo further strengthens this conclusion. NE has been found (0.51 pg/g) in the supraesophageal ganglia of male fiddler crabs (Hanumante and Fingerman, 1982b). Also, we have provided evidence that Hi and H2 receptors occur on NE neurons because in fiddler crabs pretreated with 6-hydroxydopamine (which presumably destroys NE neuroterminals in Uca as it does in vertebrates) (Hanumante and Fingerman, 1982b,c) HA is unable to significantly reduce further the melanin dispersion (Hanumante and Fingerman 1981b). We have not determined (i) the levels of NE in a-MPT injected crabs or (ii) the exact mechanism of action of a-MPT in Uca puilator. However, data that we obtained using noradrenergic and histadre- nergic agents (Hanuamante and Finger- man, 1981b) reveal that 20 MPT clearly interferes with NE neurotransmission. This probably was either by way of its well- established (at least in mammals) pharma- cological NE synthesis-inhibiting effect (Terraswawa et al., 1975; Lofstrom and HOURS Figure 1. Relationships between melanophore stage and time. Circles with bottom-half darkened, crabs that received diphenhydramine; circles with top-half darkened, crabs that received histamine; solid circles, crabs that received histamine plus diphenhydramine; open circles, salme-injected controls. Vertical bars indicate SEM. 106 Tulane Studies in Zoology and Botany Vol. 23 Table i . The means ( ± SEM) of the differences between the melanophore stages de- termined at 15, 30, 60, 90, and 120 minutes of the intact crabs that received a drug versus the saline-injected controls. The minus sign indicates decreased melanin dispersion relative to the controls. *Statistically significant p^ .05 relative to respective controls. 4-Methyl histamine (4-MeHA) Cimetidine 4-MeHA plus cimetidine 4-MeHA plus SA-97 a Methyl-p-Tyrosine (a-MPT) 4-MeHA plus a-MPT Histamine (HA) HA plus a-MPT -0.67*(± 0.08) -0.17 (± 0.01) -0.39 (± 0.07) -1.43*(± 0.12) -1.15*(± 0.15) -1.44*(± 0.21) -1.18*(± 0.18) -1.01*(± 0.12) Backstrom, 1978; Douglas, 1980) or by stimulating H2 receptors, thereby leading to the observed decrement in MDH release (Table 1). Hence, the melanin of these a -MPT-treated crabs did not disperse to the extent it did in the control animals. As stated above, in the crabs co-injected with 4-MeHA and a-MPT or HA and a-MPT, neither 4-MeHA nor HA signifi- cantly affected the melanin dispersion compared with that which occurred in re- sponse to a-MPT alone (Table 1). This pre- sumably was due to the interference with NE neurons by a-MPT in such a way that the impulse-mediated decrement in NE secretion by the H2 stimulators 4-MeHA and HA was not large enough to affect significantly the NE-mediated MDH release. The diphenhydramine-evoked increment in melanin dispersion (Fig. 1) was pre- sumably due to its blocking action on NE uptake, (Marco et al., 1980). NE uptake, inhibitors like nisoxetine (Koe, 1976) have already been shown to potentiate MDH release (Hanumante and Fingerman, 1981a). Diphenhydramine antagonizes H, receptors (Isaac and Goth, 1965; Fantozzi et al., 1975; Marco et al., 1980) also. How- ever, because H, receptor blockers do not significantly abolish HA- or 4-MeHA- (an H: receptor agonist) mediated inhibition of melanin dispersion, we suggest that the NE uptake, blocking action of diphenhydra- mme is responsible for the potentiation of melanin dispersion. The observation that even when HA is co-administered with di- phenhydramine there is still a decrease in melanin dispersion (Fig. 1) indicates that HA does not evoke its effect by stimulating NE uptake,; uptake, being the major mechanism of inactivating the postsyn- aptic actions of monoamines including NE (Fuller and Wong, 1977). That none of these drugs affect significantly melanin migration in isolated legs (Hanumante and Fingerman, 1981b) is consistent with the hypothesis that these drugs elicit changes in melanin dispersion indirectly by inter- acting with the neuroendocrine system of Uca pugilator. Literature Cited Brown, F.A., JR., and M.N. HiNfES. 1952. Modifications in diurnal pigmentary rhythm of Uca affected by continuous illumination. Physiol. Zool. 25: 56-70. Carlson, S.P. 1935. The color changes in Uca pugilator. Proc. Nat. Acad. Sci. 28: 549-551. Douglas, W.W198O. Histamine and 5-hydroxy- tryptamine (serotonin) and their antagonists, pp. 609-646 In: A.G. Oilman et al., Eds. Goodman and Oilman's The pharmacological basis of thera- peutics, 6th ed., Macmillan Publishing Co., New York. Fantozzi, R., F. Franconi, P.E. Man- NAioNi, E. Masini, F. Moroni. 1975. Interaction of H,- and Hj- receptor antagonists with histamine uptake and metabolism by guinea- pig isolated atrium and mouse neoplastic mast cells No. 2 Melanin Migration in Crabs 107 in vitro. Br. J. Pharmacol. 53: 569-574. FiNGERMAN, M. 1956. Black pigment concentrat- ing factor in the tiddler crab. Science 123: 585-586. M.M. Hanumante, S.W. FiNGERMAN, and D.C. REINSCHMIDT. 1981. Effects of norepinephrine and norepine- phrine agonists and antagonists on the melano- phores of the fiddler crab, Uca pugilator. J. Crust. Biol. 1: 16-26. Fuller, R.W., and D.T. Wong. 1977. inhibi- tion of serotonin reuptake. Fed. Proc. 36: 2154-2158. Hanumante, M.M., and M. Fingerman. 1981a. Responses of the melanophores of the fiddler crab, Uca pugilator, to drugs affecting noradrenergic neurotransmission: Further evidence for norepinephrine as a neurotransmitter triggering release of melanin-dispersing hormone. Comp. Biochem. Physiol. 70C: 27-34. . 1981b. Inhibitory effect of histamine on the release of melanin-dispersing hormone in the fiddler crab, Uca pugilator. Amer. Zool. 21: 1011. . 1982a. Additional evidence for nore- pinephrine as a neurotransmitter triggering release of melanin-dispersing hormone in the fiddler crab, Uca pugilator: The effects of alpha, and beta adrenoceptor blocking drugs on melanin migration. Comp. Biochem. Physiol. 71C: 15-19. . 1982b. Further evidence for norepine- phrine as a neurotransmitter stimulating release of melanin-dispersing hormone in the fiddler crab, Uca pugilator: The changes in the melanophores of the crabs following reserpine, 6-hydroxydopamine and bretylium administration. Gen. Pharmacol. 13: 99-103. 1982c. Pharmacological involvement of presynaptic alpha: adrenoceptors in norepinephri- nergic neurotransmission triggering the release of melanin-dispersing hormone in the fiddler crab, Uca pugilator. J. Crust. Biol. 2: 22-30. , S.W. FiNGERMAN, and M. FiNGERMAN. 1981. Antagonism of the inhibitory effect of the polychlorinated biphenyl preparation, Aroclor 1242, on color changes of the fiddler crab, Uca pugilator, by norepinephrine and drugs affecting noradrenergic neurotransmission. Bull. Environ. Contamin. Toxicol. 26: 479-484. HERMAN, W.S., and S.H. DALLMANN. 1975. Linnilus chromatophorotropin: action on isolated Uca legs and in various crustaceans. Experientia 31: 918-919. HOGBEN, L., and D. SLOME. 1931. The pig- mentary effector system - VI. The dual character of endocrine co-ordination in amphibian colour change. Proc. R. Soc. Lond. B. !08: 10-53. Isaac, L., and A. Goth. 1965. interaction of antihistaminics with norepinephrine uptake: cocaine-like effect. Life Sci. 4: 1899-1904. KOE, K. 1976. Molecular geometry of the inhibitors of the uptake of catecholamines and serotonin in synaptosome preparations of rat brain. J. Pharmac. Exp. Ther. 199: 649-661. LOFSTROM, A., and T. BACKSTROM. 1978. Relationship between plasma estradiol and brain catecholamine content in the diestrus female cat. Psychoneuroendocrinology 3: 103-107. Marco, E.J., G. Balfagon, J. Marin, B. Gomez, and S. LLUCH. 1980. indirect adrenergic effect of histamine in cat cerebral arteries. Naunyn-Schmiedeberg's Arch. Pharmacol. 312: 239-243. Owen, D.A.A., C.A. Harvey, and R.W. GRESTWOOD. 1979. Cardiovascular studies with impromidine (S. K. and F. 92676), a new very potent and specific histamine Hj-receptor agonist. J. Pharm. Pharmacol. 31: 577-582. PANTIN, C.P.A. 1934. The excitation of crusta- cean muscle. J. Exp. Biol. 11: 11-27. POLANIN, A., and J.H. McNEILL. 1981. Char- acterization of the histamine receptors in rabbit left atria. Can. J. Physiol. Pharmacol. 59: 19-24. T.E. Tenner, jr., and J.H. McNEILL. 1981. The characterization of cardiac histaminergic chronotropic receptors in the rabbit. Can. J. Physiol. Pharmacol. 59: 14-18. SANDEEN, M.I. 1950. Chromatophorotropins in the central nervous system of Uca pugilator, with special reference to their origins and actions. Physiol. Zool. 23: 337-352. Terrasawa, E., W.E. Bridson, J.W. Davenport, and R.W. Gay. 1975. Roie of brain monoamines in release of gonadotropin before proestrus in the cyclic rat. Neuroendocrin- ology 18: 345-359. 108 Tulane Studies in Zoology and Botany Vol . 23 ADDITIONAL TREMATODES OF MAMMALS IN LOUISIANA WITH A COMPILATION OF ALL TREMATODES REPORTED FROM WILD AND DOMESTIC MAMMALS IN THE STATE WESLEY L. SHOOP AND KENNETH C. CORKUM Department of Zoology and Physiology, Louisiana State University Baton Rouge, Louisiana 70803 Abstract The following trematodes were collected from hunter-trapped mammals in the Atchafalaya basin of Louisiana during the winters of 1981 and 1982: Alaria alarioides (Dubois, 1937) Dubois, 1970 from mink, Mustela vison Schreber, and river otter, Lutra cana- densis (Schreber); Alaria marcianae (La Rue, 1917) Walton, 1949 from raccoon, Procyon lotor (Linn.) and bobcat, Lynx rufus (Schreber); Alaria mustelae Bosma, 1931 from raccoon and mink; Amphimerus speciosus (Stiles and Hassal, 1896) Barker, 1911 from raccoon and the domestic cat. Fells domesticus Linn.; Baschklrovitrema incrassatum (Dies., 1850) Skrjabin, 1944 from mink and river otter; Brachylaima virgin- iana Dickerson, 1930 from opossum, Dldelphis vir- giniana Kerr; Carneophallus basodactylophallus Bridgman, 1969 from raccoon; Cryptocotyle concava (Creplin, 1825) Lube, 1899 from mink; Fibricola cratera (Barker and Noll, 1915) Dubois, 1932 from mink, opossum, and raccoon; F. lucida (La Rue and Bosma, 1927) Dubois and Rausch, 1950 from mink and opossum; Gyrosoma stngulare Byrd, Bogitsh, and Maples, 1%1 from raccoon and mink; Hasstllesia texensis Chandler, 1929 from muskrat. Ondatra zibe- thica (Linn.); Heterobllharzia americana Price, 1929 from mink, raccoon, and bobcat; Isthmiophora mells (Schrank, 1788) Luhe, 1909 from raccoon and mink; Linstowiella szldati (Anderson, 1944) Anderson and Cable, 1950 from opossum and raccoon; Marltremtn- oides nettae (Gower, 1938) Rankin, 1939 from rac- coon and mink; Microphallus opacus (Ward, 1894) Ward, 1901 from raccoon and mink; Paragonlmus kellkottl Ward, 1908 from opossum; Pharyngosto- moldes procyonis Harkema, 1942 from raccoon; Quinqueserialis qulnqueserialis (Barker and Laughlin, 1911) Harwood, 1939 from muskrat; Rhopalias ma- cracanthus Chandler, 1932 from opossum; and Sella- cotyle vitellosa Sogandares-Bernal, 1961 from mink. Alaria alarioides, A. marcianae, Amphimerus spe- ciosus, Cryptocotyle concava, Isthmiophora mells. Microphallus opacus, Paragonlmus kellkottl, and Qulnqueserialis qulnqueserialis have not been pre- viously reported from Louisiana mammals. Diag- noses are presented for the species representing state records along with pertinent notes on the biology of each. New host records include Heterobllharzia americana, Cryptocotyle concava, and Maritremi- noides nettae from mink; Alaria marcianae, Amphi- merus speciosus, and Linstowiella szldati from rac- coon; and Hasstllesia texensis from muskrat. A com- pilation of trematodes previously reported from Loui- siana mammals is presented. INTRODUCTION Recently, we reported some trematodes collected from mammals in south Louisi- ana (Shoop and Corkum, 1981a). Since that time we have continued our examina- tion of hunter-trapped mammals from the Atchafalaya basin of Louisiana during the winters of 1981 and 1982. The following mammals were examined for trematodes: 42 minks, Mustela vison Schreber; 37 rac- coons, Procyon lotor (Linn.); seven river otters, Lutra canadensis (Schreber); five muskrats, Ondatra zibethica (Linn.); three bobcats, Lynx rufus (Schreber); four domestic cats, Felis domesticus Linn.; two opossums, Dldelphis virginiana Kerr; and three red foxes, Vulpes fulva (Desmarest). The red foxes were found uninfected with trematodes. Trematodes were fixed in steaming 10% EDITORIAL COMMITTEE FOR THIS PAPER: DR. BERT B. BABERO, Professor of Biological Sciences, University of Nevada, Las Vegas, Las Vegas, Nevada 89154 DR. WALTER E. WILHELM, Associate Professor of Biology, Memphis State University, Memphis, Tennessee 38152 109 no Tulane Studies in Zoology and Botany Vol. 23 formalin and stained in Semichon's aceto- carmine. All measurements are in micro- meters unless otherwise stated; means are followed by the ranges in parentheses. Line drawings were prepared with the aid of a microprojector. Representative specimens of the species for which diagnoses are given were deposited in the Manter Lab- oratory, University of Nebraska State Museum, Lincoln, Nebraska. Table I lists the trematodes recovered from the eight species of mammals. Lumsden and Zischke (1961) reported and diagnosed Fibricola cratera, F. lucida, Hasstilesia texensis, Brachylaima virgin- iana, and Rhopalias macracanthus from Louisiana mammals. Our specimens agree in all respects with Lumsden and Zischke's (1961) diagnoses. Our specimens of Hassti- lesia texensis from the muskrat represent a new host record. Shoop and Corkum (1981a) reported and diagnosed Alaria mustelae, Baschkirovitrema incrassatum, Gyrosoma singulare, Maritreminoides nettae, and Pharyngostomoides procyonis from Louisiana mammals. In that report we noted M. nettae in raccoons; it is herein reported from the mink as well (new host record). In a more recent note, we (Shoop and Corkum, 1982) commented further on the status of G. singulare in this state. He- terobilharzia americana has been reported from Louisiana mammals by Malek et al. (1961) and Kaplan (1964). Our collections of H. americana from mink represent a new host record. Carneophallus basodac- tylophallus was originally described by Bridgman (1969) from raccoon in Louisi- ana as was Sellacotyle vitellosa from mink by Sogandares-Bernal (1961). Lumsden and Winkler (1962) reported Linstowiella szidati from opossum. We have found it in opossum as well as in raccoon. In addition to these trematodes, we identified eight other species that have not been previously reported from Louisiana mammals and that are of importance from epidemiolo- gical or zoogeographical standpoints. Table II compiles all trematodes reported heretofore from mammals in the state of Louisiana. Family DIPLOSTOMIDAE Poirier, 1886 Alaria alarioides (Dubois, 1937) Dubois, 1970 (Figure 1) Synonyms: Diplostomum alarioides Dubois, 1937; Enhydrodiplostomum alar- ioides (Dubois, 1937) Dubois, 1944. Hosts: Mustela vison Schreber and Lutra canadensis (Schreber). Location: Small intestine. Locality: Belle River, Assumption Parish, Louisiana. Deposition: Univ. Nebraska State Mus., Manter Lab. Coll. No. 21367. Diagnosis (based on ten mature specimens): Body elongate, distinctly bisegmented, 1650 (1400-1800) long by 540 (450-650) at the widest point. Forebody spathulate, 777 (640-940) long by 540 (450-650) wide; pseudosuckers present as depressions on either side of the oral sucker, never observed evaginated. Hind- body claviform, 907 (760-1050) long by 430 (400-480) wide, containing reproduc- tive organs. Forebody tegument covered with small spines; hindbody smooth. Oral sucker terminal, 92 (80-100) long by 106 (90-120) wide; acetabulum weak, spher- ical, 75 (60-80) long by 76 (60-90) wide, often covered by the tribocytic organ; tribocytic organ broadly elliptical when evaginated, 348 (240-400) long by 280 (240-330) wide, with a longitudinal cleft. Prepharynx and esophagus extremely short or absent; pharynx usually in contact with oral sucker, 77 (70-90) long by 65 (50-80) wide; paired ceca extend to the posterior end of body. Testes tandem, not equal; anterior testis asymmetrical, laterally dis- posed on either side of midline, 215 (200-250) long by 317 (290-350) wide; pos- terior testis symmetrical, dumbbell- shaped, much wider than anterior testis, 218 (190-250) long by 394 (350-410) wide, with a ventro-median groove to allow pas- sage of ceca, uterus, and vitellaria; ejacula- tory duct opens into the genital atrium; genital atrium opens posterior, subterm- inally on the dorsal surface. Ovary spher- ical, located in hindbody just in front of No. 2 Trematodes of Mammals 111 Table I. Trematodes recovered from hunter-trapped mammals in Louisiana during the winters of 1981 and 1982. Trematode Hosts No. Examined No. Infected % Location Alaria alarioides (Dubois, 1937) Dubois, 1970 A. marcianae (La Rue, 1917) Walton, 1949 A. mustelae Bosma, 1931 Amphimerus speciosus (Stiles and Hassal, 1896) Barker, 1911 Baschkirovitrema incrassatum (Dies., 1850) Skrjabin, 1944 Brachylaima virginiana Dickerson, 1930 Carneophallus basodactylophallus Bridgman, 1969 Cryptocotyle concava (Creplin, 1825) Luhe, 1899 otter 7 mink 42 raccoon 37 bobcat 3 raccoon 37 mink 42 raccoon 37 domestic cat 4 otter 7 mink 42 opossum 2 raccoon 37 mink Fibricola cratera (Barker and Noll, 1915) Dubois, 1932 F. lucida (La Rue, and Bosma, 1927) Dubois and Rausch, 1950 mink raccoon opossum mink opossum Gyrosoma singulare Byrd, Bogitsh, and Maples, 1961 raccoon mink Hasstilesia texensis Chandler, 1929 Heterobilharzia americana Price, 1929 muskrat raccoon mink bobcat Isthmiophora melis (Schrank, 1788) Luhe, 1909 raccoon mink Linstowiella szidati (Anderson, 1944) Anderson and Cable, 1950 raccoon opossum Maritreminoides nettae (Gower, 1938) Rankin, 1939 mink raccoon Microphallus opacus (Ward, 1894) Ward, 1901 raccoon mink Paragonimus kellicotti Ward, 1908 opossum Pharyngostomoides procyonis Harkema, 1942 raccoon Quinqueserialis quinqueserialis (Barker and Laughlin, 1911) Harwood, 1939 muskrat Rhopalias macracanthus Chandler, 1932 opossum Sellacotyle vitellosa Sogandares- Bernal, 1961 mink 42 42 37 2 42 2 37 42 5 37 42 3 37 42 37 2 42 37 37 42 2 37 2 42 2 24 2 2 1 1 1 1 2 21 1 22 4 12 2 26 2 7 2 1 20 2 1 6 2 1 1 3 6 5 4 1 31 29 57 5 67 3 2 3 25 29 50 50 52 Sm. Int. Liver Sm Int. 10 32 100 62 100 19 5 20 Cecum 54 Mes. Ven 5 33 16 Sm. Int. 5 3 50 7 16 14 10 50 Lungs 84 Sm. Int. 40 Cecum 50 Sm. Int. 5 » 112 Tulane Studies in Zoology and Botany Vol. 23 the anterior testis, 103 (90-120) long by 1 14 (110-120) wide; uterus courses anteriad into the forebody and turns immediately posteriad where it opens in the genital atrium; vitellaria penetrate the forebody and extend in two bands through the ven- tro-medial grooves of the testes to the level of the genital atrium; vitelline reservoir median, intertesticular. Eggs large, operculate, 101 (90-1 10) long by 55 (50-60) wide. Excretory system not observed. Discussion: Dubois (1937) originally des- cribed Diplostomum alarioides from a Brazilian otter. He (Dubois, 1944) subse- quently purged the genus Diplostomum of all mammalian parasites, retaining it for avian parasites, and erected the new genus Enhydrodiplostomum for D. alarioides and a second otter parasite, D. fosteri. Chandler and Rausch (1946) assigned two additional species, Alaria clathrata and A. pseudoclathrata, both also parasites of the otter, to the genus Enhydrodiplostomum. In a later revision, Dubois (1970) agreed that these four species are closely related, but reassigned them to the genus Alaria where additional mustelid parasites are found. Sawyer's (1961) collection of A. alari- oides from river otter in Georgia was the first report from North America. Since then. Miller and Harkema (1964, 1968) reported y4. alarioides from both mink and river otter in North Carolina, and Fleming et al. (1977) reported it from river otter in Alabama. A. alarioides is also a common parasite of mink and river otter in Louisi- ana. Measurements oi A. alarioides from the two hosts compare favorably with the descriptions of Dubois (1937, 1970). Alaria marcianae (La Rue, 1917) Walton, 1949 (Figure 2) Synonyms: Cercaria marcianae La Rue, 1917; Agamodistomum marcianae (La Rue, 1917) Cort, 1918; Alaria americana Hall and Wigdor, 1918; Alaria canis La Rue and Fallis, 1934; Alaria minnesotae Chandler, 1954. Hosts: Lynx rufus (Schreber) and Procyon lotor (Linn.). Location: Small intestine. Locality: Pierre Part, Assumption Parish, Louisiana. Deposition: Univ. Nebraska State Mus., Manter Lab. CoU. No. 21368. Diagnosis (based on ten mature specimens): Body elongate, distinctly bi- segmented, 1375 (1000-1600) long by 478 (350-600) at the widest point. Forebody spathulate with lateral margins folded ven- trally where they meet at the midline, the entire forebody serving as an organ of attachment, 883 (650-1050) long by 478 (350-600) wide; ear-like appendages pre- sent on either side of the oral sucker, rarely observed invaginated to form pseudo- suckers. Hindbody conical, 535 (400-650) long by 363 (280-500) wide, containing re- productive organs. Forebody tegument covered with small spines, hindbody tegu- ment smooth. Oral sucker terminal 90 (60-105) long by 73 (60-81) wide; aceta- bulum weak, spherical, 74 (60-95) long by 75 (60-95) wide, rarely covered by the tribocytic organ; tribocytic organ elongate when evaginated, 453 (310-550) long by 200 (155-225) wide, with a longitudinal cleft. Prepharynx present, 5 (4-6) long; pharynx pyriform, 102 (75-215) long by 64 (55-85) wide; esophagus 6 (4-10) long; paired ceca extend to the posterior end of the body. Testes tandem, not equal; ante- rior testis asymmetrical, typically wedge- shaped, laterally disposed on either side of the midUne, 160 (128-215) long by 225 (175-300) wide; posterior testis symmetri- cal, dumbbell-shaped much wider than anterior testis, 210 (165-276) long by 340 (275-425) wide, with a ventro-medial groove to allow passage of ceca and uterus; muscular ejaculatory pouch lies posterior to the testes and empties into the genital atrium; genital atrium located in the poste- rior end of the body, opening on the dorso-subterminal side. Ovary reniform, located in front of the anterior testis on either side of midline, 72 (60-99) long by 167 (100-180) wide; Mehlis' gland opposite No. 2 Trematodes of Mammals 113 the ovary; uterus courses briefly into the forebody and turns immediately posteriad where it empties into the genital atrium; vitellaria located only in the forebody, from just in front of the acetabulum to the forebody-hindbody juncture; vitelline re- servoir prominent, located in the hindbody at the level of the anterior testis. Eggs few, large, operculate, 122 (110-128) long by 65 (60-75) wide. Excretory pore terminal, remainder of excretory system not observed. Discussion: Apparently, adult Alaria marcianae have not previously been reported from Louisiana. A single speci- men of A. americana (= A. marcianae) from a dog from Baton Rouge was de- posited by G. Dikmans (USNM Helm. Coll. No, 25159). We examined that speci- men and identify it as /I . marcianae, being similar to our material from the bobcat. In a previous report, the epidemiology of A. marcianae mesocercariae was studied in Louisiana and evidence was presented that this species was responsible for an authochtonous human infection (Shoop and Corkum, 1981b). In experimental in- fections only juvenile raccoons served as definitive hosts for A. marcianae. Adult raccoons proved to be refractory to the development of the mesocercarial stage, which remained undifferentiated in the subcutaneous fat. These findings were cor- roborated in the present study because no adult raccoons were found infected. Two yearlings, however, harbored several adult A. marcianae in their duodena. This is the first report of raccoon naturally infected with this species. Though these worms from the yearlings exhibited no morpholo- gical anomalies, they were smaller than specimens from the bobcat. The known definitive hosts for A . mar- cianae in Louisiana now include the domestic dog, bobcat, and juvenile rac- coons. In experimental laboratory infec- tions we have found that the domestic cat is a suitable definitive host and that it, as well as feral cats, may play a significant role in the maintenance of A. marcianae in Louisiana. Family OPISTHORCHIIDAE Braun, 1901 Amphimerus speciosus (Stiles and Hassal, 1896) Barker, 1911 (Figure 3) Synonyms: Amphimerus caudalitestis Caballero, Grocott, and Zerecero, 1953; A. guayaquilensis {Rodriguez, Gomez, and Montalvan, 1948) Caballero, Grocott, and Zerecero. 1953; A. interruptus (Braun, 1901) Barker, 1911; A. minimus Thatcher, 1970; A. neotropicalis Caballero, Mon- tero-Gei, and Caballero, 1963; A. parcio- vatus Franco, 1967; A. pricei (Foster, 1939) Yamaguti, 1958; A. pseudofelmeus (Ward, 1901) Barker, 1911. Hosts: Felis domes ticus Linn, and Procyon lotor (Linn.). Location: Liver and bile ducts. Locality: Ramah, Iberville Parish, Louisi- ana. Deposition: Univ. Nebraska State Mus., Manter Lab. Coll. No. 21369. Diagnosis (based on ten mature specimens): Body elongate, sharply ta- pered anterior to the acetabulum, 10.25 (8.0-12.25) mm long by 2010 (1150-2400) at the widest point. Tegument beset with small, stout spines. Oral sucker 268 (240-300) long by 313 (270-340) wide; acetabulum 200 (150-240) long by 218 (170-250) wide. Prepharynx absent; pharynx 183 (160-200) long by 173 (150-190) wide; esophagus 170 (120-200) long; paired ceca extend to the posterior end of body. Testes tandem, in posterior Vi of body, transversely elongate, sHghtly lobed; anterior testis 498 (410-600) long by 925 (550-1150) wide; posterior testis 573 (450-720) long by 925 (550-1150) wide; seminal vesicle elongate, coiled, opens into the genital atrium which is immediately preacetabular. Ovary oval to reniform, may be slightly lobed, 325 (240-450) long by 470 (370-610) wide; seminal receptacle large, lying immediately postovarian, 525 (200-700) long by 473 (320-600) wide; Laurer's canal present, opening on dorsal surface; Mehlis' gland preovarian, sinistral to midline; uterus forming transverse, intercecal coils between the ovary and ace- 114 Tulane Studies in Zoology and Botany Vol. 23 tabulum; vitellaria lateral, extracecal, con- sisting of two pairs of disjunct bundles on each side, each pair separate at level of the ovary; four vitelline ducts fuse mesially at the level of the ovary to form a vitelline reservoir. Eggs small, 28 (25-32) long by 12 (11-14) wide. Excretory pore terminal or slightly subterminal; excretory vesicle sigmoid, coursing anteriorly between the testes and bifurcating immediately poste- rior to the seminal receptacle. Discussion: Reports of species of Amphi- merus from North American mammals have almost exclusively been A. pseudofe- lineus and this name has become well en- trenched in veterinary literature. However, Nasir and Diaz (1972) synonymized the following species with A. speciosus: A. caudalitestis; A. guayaquilensis; A. inter- ruptus; A. minimus; A. neotropicalis; A. parciovatus; A. pricei; and A. pseudofe- lineus. Lumsden and Zischke (1963) reported A: <;;himerus interruptus from a yellow- crowned night heron, Nyctanassa violacea. Their measurements fall within the ranges we recorded and the specimen figured is remarkably similar to ours, indicating that they are the same species. Lumsden and Zischke also noted similarities between their specimens and the description of A. speciosus. These observations corroborate, in part, Nasir and Diaz's (1972) synony- mies and further indicate the ability of these organisms to live in both avian and mammaUan hosts. A. speciosus has been reported in cats and dogs from several states in the United States (Rothenbacher and Lindquist, 1963). Chronic morbidity associated with infection includes liver and biliary cirrhosis and pancreatitis. Also, Thatcher (1970) commented on the unassessed possibility of human infection with this species. A. speciosus was collected from the liver and bile ducts of one of four domestic cats and two of 37 raccoons in Louisiana. The rac- coon apparently is a new host record for this species. Family HETEROPHYIDAE (Leiper, 1909) Odhner, 1914 Cryptocotyle concava (Creplin, 1825) Luhe, 1899 (Figure 4) Synonyms: Distoma concava Creplin, 1825; Tocotrema concava Looss, 1899; Cryptocotyle echinata Linstow, 1878. Hosts: Mustela vison Schreber. Location: Small intestine. Locality: Belle River, Assumption Parish, Louisiana. Deposition: Univ. Nebraska State Mus., Manter Lab. Coll. No. 21370. Diagnosis (based on ten mature specimens): Body foliate, 904 (780-1050) long by 612 (560-680) wide. Tegument be- set with small spines. Oral sucker terminal, 47 (35-55) long by 54 (40-65) wide; aceta- bulum 41 (35-50) in diameter, found within the genital atrium and comprising a part of the acetabulogenital apparatus; acetabulo- genital apparatus 67 (60-75) long by 91 (70-125) wide, located medially and equa- torially. Prepharynx 10 (5-15) long; pharynx 49 (40-55) long by 48 (45-60) wide; esophagus 76 (65-100) long; paired ceca extend to the posterior end of body where they turn medially just posterior to the testes. Testes opposite, distinctly lobate, 152 (125-175) long by 233 (210-250) wide, located in posterior end of body; seminal vesicle courses from testes to the acetabu- logenital apparatus; cirrus pouch absent. Ovary wedge-shaped, lobate, 93 (70-115) long by 138 (100-175) wide, located dextral to the midline, between the ovary and right testis; uterus makes 3-4 intercecal loops before opening into the acetabulogenital complex; vitellaria mostly lateral, com- mence behind the level of the cecal bifurca- tion and extend to the posterior end of body where they meet at the midline; vitel- line reservoir is located medially, at the level of the seminal vesicle. Eggs small, operculate, 36 (33-40) long by 15 (13-20) wide. Discussion: Wootton (1957) first reported Cryptocotyle concava from North Amer- ica and elucidated the Ufe cycle. It included an operculate snail, Amnicola longiqua, in which rediae gave rise to pleurolophocer- cous cercariae; these penetrated and No. 2 Trematodes of Mammals 115 Figures 1-7. 1. Alarm alarioides from mink and river otter. 2. Alaria marcianae from bobcat and raccoon 3 Amphimerus speciosus from raccoon and the domestic cat. 4. Cryptocotyle concava from mink 5 Isthmio- phora melis from raccoon and mink. 6. Microphallus opacus from raccoon and mink. 7. Quinquesenalis quinqueserialis from muskrat. Scales in micrometers. 116 Tulane Studies in Zoology and Botany Vol. 23 encysted in three-spined sticklebacks, Cas- ter osteus aculeatus. When infected fish were fed to both chicks and ducklings adult worms were recovered. Hoffman (1957) found metacercariae of C. concava in suckers, Catostomus commersoni, and also obtained adults from experimentally infected chicks. The only other report of C. concava from North America was that of Burrows and Lillis (1965) who collected specimens from a dog in New Jersey. We compared our specimens with theirs (USNM Helm. Coll. No. 60902) and find no differences between them. Our report is the first record of C concava from mink. Its occurrence in them is not surprising due to the prevalence of fish in their diet and the lack of definitive host specificity common in heterophyids. Quite possibly, Louisiana veterinarians may encounter eggs of this trematode in routine stool examination of pets. In addi- tion, the possibility of human infection can not be overlooked because Cryptocotyle eggs have already been reported from humans elsewhere (Babbot et al., 1961). Family ECHINOSTOMATIDAE (Looss, 1902) Poche, 1926 Isthmiophora melis (Schrank, 1788) Luhe, 1909 (Figure 5) Synonyms: Fasciola putori Gmelin, 1790; Fasciola trigonocephala Rud., 1802; Euparyphium melis (Schrank, 1788) Railliet, 1919; Echinocirrus melis (Schrank, 1788) Mendheim, 1943. Mendheim, 1943. Hosts: Procyon lotor (Linn.) and Mustela vison Schreber. Location: Small intestine. Locality: Belle River, Assumption Parish, Louisiana. Deposition: Univ. Nebraska State Mus., Manter Lab. Coll. No. 21371. Diagnosis (based on ten mature specimens): Body lanceolate, 2450 (2000-3500) long by 650 (520-700) wide. Anterior tegument densely covered with spines until the posterior level of the aceta- bulum, where they diminish in number towards the posterior end of the body. Head collar reniform, bearing 27 spines; each side with 4 corner spines, 59 (57-61) long by 13 (12-14) wide; six marginals on each side, 46 (43-48) long by 11 (9-13) wide; and a double, uninterrupted row of dorsal spines composed of four oral and three aboral spines, 40 (36-44) long by 1 1 (8-12) wide. Acetabulum large relative to the oral sucker, 380 (350-410) long by 385 (350-430) wide. Prepharynx not discern- ible; pharynx 130 (110-160) long by 115 (110-140) wide; esophagus 173 (110-210) long; ceca bifurcate immediately anterior to the cirrus sac and extend to the posterior end of the body. Testes tandem, irregular in shape, from strongly indented to com- pletely lobed, posterior testis always more indented or lobate than the anterior testis, both testes wider than long; anterior testis 242 (200-310) long by 348 (310-370) wide; posterior testis 285 (220-410) long by 341 (320-360) wide; cirrus sac ovate, extending from middle of the acetabulum to just pos- terior to the cecal bifurcation, 265 (220-300) long by 168 (130-200) wide; seminal vesicle distinct; cirrus long, coiled when withdrawn, beset with minute spines. Ovary spherical, dextral to midline. 111 (90-130) long by 1 14 (90-130) wide, located between the acetabulum and anterior tes- tis; MehHs' gland broadly oval to reniform, lying immediately in front of the anterior testis; seminal receptacle absent; uterus short, with 3-5 intercecal coils; vitel- laria extend from the level of the ovary to the posterior end of body; vitelline reser- voir well developed, at the anterior half of the anterior testis. Eggs large, operculate, 97 (95-100) long by 53 (50-60) wide. Excre- tory pore dorsal and subterminal. Discussion: Dawes (1946) and Skrjabin and Bashkirova (1956) transferred all the species of Isthmiophora to the genus Euparyphium, however, Yamaguti (1971) retained the former based on: (1) body shape (lanceolate in Isthmiophora whereas Euparyphium is subcylindrical); and (2) shape of testes (irregular with lateral in- dentations in Isthmiophora whereas in No. 2 Trematodes of Mammals 117 Euparyphium they are longitudinally elongated). Based upon a comparative study of several hundred specimens from Louisiana mink and raccoons, our speci- mens agree with the generic diagnosis of Isthmiophora as presented by Yamaguti, This is the first report of Isthmiophora melis from the raccoon and, to our know- ledge, the only report of this species from North America. We have found this spe- cies in the small intestine of six of 37 rac- coons and two of 42 minks. The only other echinostomes found in raccoon are Euparyphium beaveri reported by Harkema and Miller (1964) and Bufundo et al. (1980) and Echinostoma revotutum which was regarded as an aberrant condi- tion (Larson and Scharf, 1975). Because Euparyphium beaveri is also found in minks we compared the type material de- posited by Beaver (1941) to our specimens. We find they are very similar in head collar spination and body anatomy, but that they differ strikingly in two respects: (1) the range in size of our specimens (2000-3500) is not concordant with the ranges provided by Beaver (3860 -10500) and the averages are markedly dissimilar (2450 for our material to 6100 for that of Beaver's); and (2) the testes in our specimens are broader than long with either deep marginal inden- tations or completely lobate, whereas that oi Euparyphium is longitudinally oval with only slight evidence of indentations in the larger specimens. We conclude that our material is distinct from Euparyphium beaveri. Lumsden and Zischke (1961) rediag- nosed Euparyphium beaveri from Louisi- ana minks. A close inspection of their diagnosis indicates they probably were not dealing with E. beaveri but with the closely related Baschkirovitrema incrassatum. At the time of their diagnosis B. incrassatum had not been reported from North Amer- ica. It is now known to be a common inha- bitant of mustelids from the Gulf and Atlantic coasts (Sawyer, 1961; Miller and Harkema, 1964; Fleming et al., 1977; Shoop and Corkum, 1981a). At the time we diagnosed B. incrassatum from a river otter in Louisiana we had only specimens from a single otter. We now, however, have a large series of B. incrassatum from both river otter and mink and they include the ranges of both our previous material and that given by Lumsden and Zischke (1961). We, therefore, regard Eupary- phium beaveri of Lumsden and Zischke, 1961 conspecific with Baschkirovitrema incrassatum. Family MICROPHALLIDAE Travassos, 1920 Microphallus opacus (Ward, 1894) Ward, 1901 (Figure 6) Synonyms: Microphallus ovatus Osborn, 1919. Hosts: Procyon lotor (Linn.) and Mustela vison Schreber. Location: Small intestine. Locality: Belle River, Assumption Parish, Louisiana. Deposition: Univ. Nebraska State Mus., Manter Lab. CoU. No. 21372. Diagnosis (based on ten mature specimens): Body oval to pyriform, 1233 (1160-1300) long by 664 (620-700) wide. Tegument spined throughout. Oral sucker subterminal, 67 (60-70) long by 80 (75-90) wide; acetabulum 86 (80-90) long by 91 (90-100) wide. Prepharynx 60 (35-85) long; pharynx weak, 36 (35-40) long by 29 (25-30) wide; esophagus 340 (275-400) long; ceca short, rarely extending beyond the seminal vesicle, occasionally with a single sac. Testes two, opposite, 190 (150-230) long by 135 (75-190) wide, very often the testes are not discernible in gravid specimens; seminal vesicle saccular, preacetabular, opening into the genital atrium; genital atrium lies sinistral to the acetabulum, 62 (55-75) long by 74 (65-80) wide. Ovary spherical to oval in shape, dextral to midline, 150 (110-175) long by 160 (130-205) wide; oviduct sinistral to ovary, courses posteriad to the Mehlis' gland; Mehlis' gland prominent, on the midline of the body between the two bundles of vitellaria; uterus makes several loops in posterior half of body and opens 118 Tulane Studies in Zoology and Botany Vol. 23 into the genital atrium; vitellaria in two symmetrical clusters of spherical follicles, located in the posterior Vi of body; vitel- line ducts fuse in the middle of the body at the level of the Mehlis' gland to form a viteUine reservoir. Eggs small, numerous, 25 (25-26) long by 13 (12-14) wide. Excretory vesicle V-shaped, extending to the anterior level of the vitellaria; a single collecting duct arises from each side of the vesicle and courses anteriad to the level of the pharynx. Discussion: Though Microphallus opacus is generally regarded as a fish parasite (Yamaguti, 1971) it has been experimental- ly established in various reptilian species as well as opossum and raccoon by Rausch (1947) and in white mice by Sogandares- Bernal (1965a). Rausch (1946) also reported it from a naturally infected rac- coon from Ohio and provided a brief diag- nosis. Our material from raccoon and mink agrees well with that description. Sogandares-Bernal (1965a) surveyed the crayfish parasites in Louisiana and found Cambarellus puer and Procambarus clarkii naturally infected with the metacercariae of Microphallus opacus. He noted that snails of the genus Amnicola, "probably Integra", released several different types of microphallid cercariae at his study site (Rosedale, Louisiana), one of which he be- lieved to be M. opacus. The definitive host at that time was unknown. The life-cycle of M. opacus in Louisiana can be postu- lated using Sogandares-Bernal's report and that of the present work to include the fol- lowing; an amnicolid snail as first interme- diate host; several crayfish species as second intermediate hosts; and the raccoon and mink as definitive hosts. At present, the extent to which M. opacus uses fishes as definitive hosts in Louisiana is unas- sessed as it has yet to be reported from fishes in this state. Family PARAGONIMIDAE DoUfus, 1939 Paragonimus kellicotti Ward, 1908 Hosts: Didelphis virginiana Kerr. Location: Lungs. Locality: Baton Rouge, East Baton Rouge Parish, Louisiana. Deposition: Univ. Nebraska State Mus., Manter Lab. Coll. No. 21394. Discussion: We have recovered three ma- ture Paragonimus kellicotti from the lungs of a single opossum. We have not figured or diagnosed P. kellicotti owing to the paucity of specimens in our possession and to the fact that our specimens are similar to those described by Byrd et al. (1942) which came from the lungs of a Tennessee opos- sum. Paragonimus kellicotti metacercariae were reported from crayfish in Louisiana by Ameel (1934) and La Rue and Ameel (1937). Sogandares-Bernal (1965b) re- ported natural infections of the snail, Pomatiopsis lapidaria, with Paragonimus kellicotti. Since those accounts, P. kelli- cotti is commonly acknowledged to be pre- sent in Louisiana although neither the adult nor the definitive host have been reported from this state. That the infected opossum was trapped in residential Baton Rouge is epidemiolog- ically significant. The location was an upper middle class neighborhood which borders on the flood plain of the Mississip- pi River. The area of the flood plain in heavily treed, with numerous bayous, and low lying grounds which are nearly always water laden. This scenario is a classical nidus capable of maintaining all of the hosts essential to the life-cycle of P. kelli- cotti and has the potential of including man into the life-cycle owing to his close proximity and crustacean cuisine. Family NOTOCOTYLIDAE Luhe, 1909 Quinqueserialis quinqueserialis (Barker and Laughhn, 1911) Harwood, 1939 (Figure 7) Synonyms: Notocotylus quinqueserialis Barker and LaughUn, 1911; Quinqueser- ialis hassali (Mcintosh and Mcintosh, 1934) Harwood, 1939; Notocotylus urban- ensis of Harrah, 1922. Hosts: Ondatra zibethica (Linn.). Location: Cecum. Locality: Belle River, Assumption Parish, 1 = E c E c = 3 c iSSoi^^uBS -iSmSS £ 2 g I II S i- 8 S o S S 5 : 11 |J lli £ ;j|l|S|sil|u|S|i|(;|a No. 2 Trematodes of Mammals 119 Louisiana. Deposition: Univ. Nebraska State Mus., Manter Lab. Coll. No. 21373. Diagnosis (based on ten mature specimens): Body elongate, oval, slightly attenuated anteriorly, 3850 (3420-4150) long by 1050 (960-1300) at the greatest width. Tegument aspinous. Ventral sur- face with five longitudinal rows of spher- ical glands. Oral sucker subterminal, 335 (320-350) in diameter; acetabulum absent. Pharynx absent; esophagus short, paired ceca extend to posterior end of body. Testes opposite, highly branched, in pos- terior end of body, 513 (405-610) long by 305 (260-390) wide; external seminal vesicle tubular, coursing anteriad to the base of the cirrus sac; cirrus sac elongate, claviform, 1277 (1050-1500) long by 145 (125-170) at the greatest width; cirrus often extruded and much coiled, densely beset with spines; genital pore median, near intestinal bifurcation. Ovary deeply lobed, intertesticular, 334 (300-390) long by 210 (150-250) wide; Mehlis' gland immediately anterior to ovary; uterus comprised of transverse loops which may extend beyond the ceca; metraterm distinct, 775 (700-900) long; vitellaria pretesticular, in two, extra- cecal bands. Eggs oval, 17 (16-18) long by 8 (7-9) wide, without polar filaments. Excretory system not observed. Discussion: Penn (1942) examined 1,780 muskrats from coastal Louisiana and re- covered the trematodes Nudacotyle novicia, Echinochasmus schwartzi, and Paramonostomum pseudalveatum. Byrd and Reiber (1942) examined three musk- rats from the New Orleans area and reported E. schwartzi and Phagicola nana ( = P. angrense). Because of their declining numbers, we were unable to obtain a large series of muskrats from trappers, but we were successful in obtaining five carcasses. Two of the muskrat harbored hundreds of Quinquesehalis quinqueserialis in their ceca. Although this species is considered a ubiquitous parasite of muskrats in North America, this is the first report of it from Louisiana. Our measurements agree well with those provided by Kinsella (1971) in his study of intraspecific variation of Q. quinserialis. The life-cycle has been eluci- dated by Herber (1942) and includes the freshwater snail, Gyraulis parvus, from which monostome cercariae are released and encyst on vegetation. The muskrat becomes infected while grazing on vegeta- tion containing the cysts. Incidentally, one muskrat was infected with thousands of Hasstilesia texensis in the cecum (new host record). We have found H. texensis in all of the swamp rab- bits, Sylvilagus aquaticus, that we have examined in Louisiana. As all of the speci- mens from the muskrat were gravid and showed neither stunting nor any anoma- lies, we presume that the muskrat may serve occasionally as a normal, definitive host for this species. SUMMARY The following trematodes were collected from hunter-trapped mammals in the Atchafalya basin of Louisiana during the winters of 1981 and 1982: Alaria alarioides (Dubois, 1937) Dubois, 1970; Alaria mar- cianae (La Rue, 1917) Walton, 1949; Alaria mustelae Bosma, 1931; Amphi- merus speciosus (Stiles and Hassal, 1896) Barker, 1911; Baschkirovitrema incras- satum (Dies. 1850) Skrjabin, 1944; Brachy- laima virginiana Dickerson, 1930; Carneophallus basodactylophallus Bridg- man, 1969; Cryptocotyle concava (Creplin, 1825) Luhe, 1899; Fibricola cra- tera (Barker and Noll, 1915) Dubois, 1932; Fibricola lucida (La Rue and Bosma, 1927) Dubois and Rausch, 1950; Gyrosoma sin- gulare Byrd, Bogitsh, and Maples, 1961; Hasstilesia texensis Chandler, 1929; Heterobilharzia americana Price, 1929; Isthmiophora metis (Schrank, 1788) Luhe, 1909; Linstowiella szidati (Anderson, 1944) Anderson and Cable, 1950; Mari- treminoides nettae (Gower, 1938) Rankin, 1939; Microphallus opacus (Ward, 1894) Ward, 1901; Paragonimus kellicotti Ward, 1908; Pharyngostomoides procyonis Harkema, 1942; Quinqueserialis quinque- serialis (Barker and Laughlin, 1911) Har- wood, 1939; Phopalias macracanthus 120 Tulane Studies in Zoology and Botany Vol. 23 (Chandler, 1932; and Sellacotyle vitellosa Sogandares-Bernal, 1961. Adult trematodes reported from Louisi- ana for the first time are: Alaria alarioides, A. marcianae, Amphimerus speciosus, Cryptocotyle concava, Isthmiophora melis, Mircophallus opacus, Paragonimus kellicotti, and Quinqueserialis quin- queserialis. New host records include Heterobilhar- zia americana, Cryptocotyle concava, and Maritreminoides nettae from mink; Alaria marcianae, Amphimerus speciosus, and Linstowiella szidati from raccoon; and Hasstilesia texensis from muskrat. Natural infections of adult Alaria mar- cianae were found only in juvenile rac- coons. This substantiates previous experi- mental work which demonstrated that adult raccoon are unsuitable for the maturation of this trematode. The larvae, however, are able to employ the adult rac- coon as a paratenic host where they remain undifferentiated in the subcutaneous fat. Amphimerus speciosus is a well known pathogen of dogs and cats in North America, being herein reported from a domestic cat and a raccoon. Synonymiza- tion of the better known A . pseudofelineus with A. speciosus is corroborated by our observations. Whether Isthmiophora is distinct from Euparyphium has been debated by several authors. We place our specimens in the genus Isthmiophora on the basis of body shape and testicular morphology. We com- pared our specimens to those of Eupary- phium beaveri and conclude they are dis- tinct. This is the first report of /. melis from a raccoon and, to our knowledge, the only report of this species from North America. We consider Euparyphium bea- veri of Lumsden and Zischke, 1961 to be a synonym of Baschkirovitrema incras- satum. Microphallus opacus is a common para- site in the mink and raccoon in Louisiana. Sogandares-Bernal (1965a) stated that the aquatic snail, Amnicola, probably served as first intermediate host and that several species of crayfish served as second inter- mediate hosts. Therefore, a hypothetical life-cycle of M. opacus from Louisiana can be proposed: the first intermediate host is the aquatic snail, Amnicola; several cray- fishes serve as second intermediate; and the raccoon and mink are definitive hosts. Fishes have yet to be reported with M. opacus from Louisiana. The first and second intermediate hosts, as well as the larval stages, of Paragonimus kellicotti have been previously reported from Louisiana. However, this is the first report from this state of the adult fluke in a naturally infected definitive host, the opossum. The locality of the infection is noteworthy in that it was found in an upper middle class Baton Rouge residential area. Literature Cited AMEEL, D.J. 1934. Paragonimus, its life history and distribution in North America and its taxo- nomy (Trematoda: Troglotrematidae). Am. J. Hyg. 19: 279-317. Babbot, F.L., W.W. Frye, and J.E. Gordon. 1961. intestinal parasites of man in Arctic Greenland. Am. J. Trop. Med. Hyg. 10: 185-190. BABERO, B.B. 1960. A survey of parasitism in skunks {Mephitis mephitis) in Louisiana, with ob- servations on pathology due to helminthiasis. J. Parasitol. 46: 26-27. 1972. A record of progenesis in Trema- toda. Proc. Helm. Soc. Wash. 39: 128-131. , and J. Lee. 1961. Studies on the hel- minths of nutria, Myocastor coy pus (Molina), in Louisiana with checklist of other worm parasites from this host. J. Parasitol. 47: 378-390. BAFUNDO, K.W., W.E. WiLHELM, and M.L. Kennedy. 1980. Geographic variation in helminth parasites from the digestive tract of Tennessee raccoons, Procyon lotor. J. Parasitol. 66: 134-139. Beaver, P.C. 1941. studies on the life history of Euparyphium melis 'Trematoda: Echinostomi- dae). J. Parasitol. 27: 35-44. Bennett, H.J. 1938. a partial check list of the trematodes of Louisiana vertebrates. Proc. La. Acad. Sci. 4: 178-181. , and A.G. Humes. 1939. studies on the pre-cercarial development of Stichorchis subtriquetrus (JTemaioda.: Paramphistomidae). J. Parasitol. 25: 223-231. , and L.L. Jenkins. 1936. The longe- No. 2 Trematodes of Mammals 121 vity of the miracidium of Cotylophoron cotylo- phorum. Proc. La. Acad. Sci. 8: 5-13. BRIDGMAN, J.F. 1969. Life cycles of Carneo- phallus choanophallus n. sp. and C. basodactylo- phallus n. sp. (Trematoda: Microphailidae). Tulane Stud. Zooi. & Bot. 15: 81-105. Burrows, R.B., and W.G. Lillis. i965. Trematodes of New Jersey dogs and cats. J. Para- sitol. 51: 570-574. BYRD, E.E., and R.W. MACY. 1942. Mam- malian trematodes. III. Certain species from bats. J. Tenn. Acad. Sci. 17: 149-156. , and R.J. REIBER. 1942. Mammalian trematodes. II. Three flukes from small mam- mals. J. Tenn. Acad. Sci. 17: 143-156. , and M.V. Parker. 1942. The anatomy of a lung fluke from the opos- sum (Didelphis virginiana Kerr). J. Tenn. Acad. Sci. 17:116-129. Chandler, A.C, and R. Rausch. i946. a study of strigeids from Michigan mammals, with comments on the classification of mammalian strigeids. Trans. Am. Microsc. Soc. 65: 328-337. Custer, J.W., and D.B. Pence. 1981. Ecolo- gical analysis of helminth populations of wild canids from the Gulf Coastal prairies of Texas and Louisiana. J. Parasitol. 67: 289-307. DAWES, B. 1946. The Trematoda with special refer- ence to British and other European forms. Cam- bridge Univ. Press, 664 p. DiKMANS, G. 1931. A new nematode worm, Vian- naia bursobscura, from the opossum with a note on the other parasites of the opossum. Proc. U.S. Natl. Mus. 79: 1-4. 1945. Check-list of the internal and ex- ternal animal parasites of domestic animals in North America. Am. J. Vet. Res. 6: 211-241. Dubois, G. 1937. Sur quelques Strigeide. Rev. Suisse Zool. 44: 391-396. 1944. A propos de la specificite' parasi- taire des Strigeida. Bull. Soc. Neuch. Sci. Nat. 69: 5-103. 1970. Statut des Alariinae Hall et Wigdor, 1918 (Trematoda: Diplostomatidae) et revision de quelques alariens. Bull. Soc. Neuch. Nat. 10: 259-727. Fleming, W.J., C.F. Dixon, and J.W. LOVETT. 1977. Helminth parasites of river otters {Lutra canadensis) from southeastern Alabama. Proc. Helminthol. Soc. Wash. 44: 131-135. HARKEMA, R. and G.C. MiLLER. 1964. Hel- minth parasites of the raccoon, Procyon lotor in the southeastern United States. J. Parasitol. 50: 60-66. HERBER, E.C. 1942. The life history studies on two trematodes of the subfamily Notocotylinae. J. Parasitol. 28: 179-196. Hoffman, G.L. 1957. studies on the life cycle of Cryptocotyle concavum from the common sucker and experimentally in the chick. Proc. North Dakota Acad. Sci. 11: 55-56. Kaplan, E.H. 1964. HeterobUharzia americana Price, 1929, in the opossum from Louisiana. J. Parasitol. 50: 797. KINSELLA, J.M. 1971. Growth, development, and intraspecific variation of Quinqueserialis quinque- serialis (Trematoda: Notocotylidae) in rodent hosts. J. Parasitol. 57: 62-70. Larson, O.R., and W.C. SCHARF. 1975. New helminth records from Minnesota mammals. Proc. Helminthol. Soc. Wash. 42: 174-175. La Rue, G.R., and D.J. AMEEL. 1937. The distribution of Paragonimus. J. Parasitol. 23: 382-388. LUMSDEN, R.D., and C.A. WiNKLER. 1962. The opossum, Didelphis virginiana (Kerr), a host for the cyathocotylid trematode Linstowiella szidati (Anderson, 1944) in Louisiana. J. Parasitol. 48: 503. , and J. A. ZlSCHKE. 1961. Seven trematodes from small mammals in Louisiana. Tulane Stud. Zool. & Bot. 9: 87-98. , and 1963. Studies on the trematodes of Louisiana birds. Z. Parasitenk. 22: 316-366. Malek, E.A., L.R. Ash, H.F. Lee, and M.D. Little. 1961. HeterobUharzia infection in the dog and other mammals in Louisiana. J. Parasitol. 47: 619-623. Martin, D.R. 1976. New host and distribution records of helminth parasites of the Mexican free- tail bat, Tadarida brasiliensis, from Texas and Louisiana. Proc. Helminthol. Soc. Wash. 43: 85-86. Miller, G.C, and R. Harkema. i964. Studies on helminths of North Carolina vertebrates. V. Parasites of the mink, Mustela vison Schreber. J. Parasitol. 50: 717-720. , and 1968. Helminths of some wild mammals in the southeastern United States. Proc. Helminthol. Soc. Wash. 35: 118-125. NASIR, p., and M.T. DiAZ. 1972. Avian flukes from Venezuela. Rivista di Parassitologia 33: 245-276. PENN, G.H. 1942. Parasitological survey of Loui- siana muskrats. J. Parasitol. 28: 348-349. Rausch, R. 1946. The raccoon, a new host for Microphallus sp., with additional notes on M. ovatus from turtles. J. Parasitol. 32: 208-209. 1947. Some observations on the host re- lationships of Microphallus opacus (Ward, 1894) (Trematoda: Microphailidae). Trans. Am. Microsc. Soc. 66: 59-63. 122 Tulane Studies in Zoology and Botany Vol.23 ROTHENBACHER, H., and W.D. LlND- QUIST. 1963. Liver cirrhosis and pancreatitis in a cat infected with Amphimerus pseudofelineus. J. Am. Vet. Med. Assoc. 143: 1099-1102. SAYWER, T.K. 1961. The american otter, Lutra canadensis vaga, as a host for two species of trema- todes previously unreported from North America. Proc. Helminthol. Soc. Wash. 28: 175-176. SHOOP W.L., and K.C. CORKUM. 1981a. Some trematodes of mammals in Louisiana. Tulane Stud. Zool. & Bot. 22: 109-121. , and . 1981b. Epidemiology of A laria marcianae mesocercariae in Louisiana. J. Parasitol. 67: 928-931. _, and . 1982. Progenesis re- considered in Ribeiroia ondatrae (Price, 1931) nee Babero, 1972, a junior synonym of Gyrosoma singulare. J. Parasitol. 68:424. SKRJABIN, K.L., and E.I. BASHKIROVA. 1956. Echinostomatidae Dietz, 1909. In SKRJABIN, K.I., Trematodes of Animals and Man. 12: 53-932. SOGANDARES-BERNAL, F. 1961. Sellacotyle vi- lellosa, a new troglotrematid trematode from the mink in Louisiana. J. Parasitol. 47: 911-912. . 1965a. Parasites from Louisiana cray- fishes. Tulane Stud. Zool. & Bot. 12: 79-85. . 1965b. Studies on American paragoni- miasis. 1. Age immunity of the snail host. J. Para- sitol. 51: 958-960. , and J.F. BRIDGMAN. 1960. Three Ascocoiyle complex trematodes (Heterophyidae) encysted in fishes from Louisiana, including the description of a new genus. Tulane Stud. Zool. & Bot. 8: 31-39. Thatcher, V.E. 1970. The genus Amphimerus Barker, 1911 (Trematoda: Opisthorchiidae) in Co- lombia with the description of a new species. Proc. Helminthol. Soc. Wash. 37: 207-211. WOOTTON, D.M. 1957. The life history of Cryp- tocotyle concavum (Creplin, 1825) Fischoeder, 1903 (Trematoda: Heterophyidae). J. Parasitol. 43: 271-279. YAMAGUTI, S. 1971. Synopsis of digenetic trema- todes of vertebrates. Vol. I. Keigaku Publ. Co., Ltd., Tokyo, 1,100 pp. December 15, 1982 COMPARATIVE VISCERAL TOPOGRAPHY OF THE NEW WORLD SNAKE TRIBE THAMNOPHIINI (COLUBRIDAE, NATRICINAE) NITA J. ROSSMAN and DOUGLAS A. ROSSMAN Museum of Zoology, Louisiana State University Baton Rouge, Louisiana 70893 NANCY K. KEITH Dept. of Experimental Statistics, Louisiana State University Baton Rouge, Louisiana 70893 Abstract The positions and lengths of a variety of visceral organs in 631 preserved adult thamnophiine snakes were determined in terms of ventral scute number and converted into a per cent of total ventral number; a mean was calculated for each taxon to allow compar- ison with other taxa. Dice-Leraas diagrams were then constructed for the following organ positions and lengths: posterior end of heart, anterior and posterior ends of Uver, posterior end of pancreas, anterior and posterior ends of right and left kidney, Uver length, right and left kidney lengths, heart-liver interspace, and kidney overlap. Sexual dimorphism is apparent in many of the characters examined. Apparently corre- lated with their need for space to accommodate developing young, females tend to have their anterior and midbody organs placed more anteriorly and their kidneys more posteriorly than those in males. Stepwise discriminant analysis was performed on the following four variables in male thamnophiine snakes: posterior end of heart, anterior end of right kidney, posterior end of left kidney, and kidney overlap. The 294 specimens represented 11 groups — 7 genera plus Ruthven's four species groups of Thamnophis. Two of four linear discriminant functions were retained as they explain 83.2 l<^o of the relative variation. Function 1 is generally an anterior end of right kidney dimension, and function 2 is a kidney overlap and posterior end of heart dimension. More than 66% of the specimens were correctly classi- fied by use of the model. All groups except Clonophis could be classified with greater success than the 21% prior probability obtained by placing them all in the Elegans group of Thamnophis, the numerically largest sample. The discriminant analysis was able to distinguish among the seven genera (as well as among Ruthven's four species groups of Thamnophis) at the 0.05 level except that Clonophis and Tropidoclonion could not be distinguished from each other. Although visceral topographic data alone do not clearly delimit thamnophiine genera nor establish inter- or intrageneric relationships, some trends are apparent that serve to support taxonomic conclusions based on other kinds of characters. Clonophis and Regina can be distinguished from Nerodia, in which genus they were formerly included. Thamnophis (less proximus and sauritus) can also be distinguished from Nerodia (less erythrogaster and valida). The Sauritus group of Thamnophis differs markedly from the other three species groups established by Ruthven in most visceral topographic features. The ribbon snakes (Sauritus group) frequently tend to have a posterior displacement of organs, a condition often occurring also in the short, semifossorial genera (Clonophis, Seminatrix, Storeria, Tropidoclonion, Virginia). One unique feature shared by all of the semifossorial genera is the possession of a relatively long liver. Introduction The technique of determining snake vis- ceral topography using ventral scutes as re- ference points has received little attention since its introduction by Thompson seventy years ago. Although a moderate amount amount of descriptive anatomical work has appeared in print, very little has EDITORIAL COMMITTEE FOR THIS PAPER: DR. SAMUEL B. McDOWELL, Professor of Zoology, Rutgers University, Newark, New Jersey 07102 DR. JAMES S. ROGERS, Associate Professor of Biology, University of New Orleans, New Orleans, Louisiana 70122 DR. ROBERT A. THOMAS, Director, Louisiana Nature Center, New Orleans, Louisiana 70127 123 124 Tulane Studies in Zoology and Botany Vol. 23 been done of a comparative nature that might be of taxonomic value, and none using discriminant analysis. The present study was undertaken to investigate the possible taxonomic significance of visceral topography in the tribe Thamnophiini of the colubrid subfamily Natricinae. Beddard (1908, 1909) characterized the position of visceral organs in three genera of boid snakes in terms of the distance from the snout to the organ. He also mea- sured organ length and the distance be- tween organs. Beddard was convinced that the position of viscera within the body of snakes generally had systematic impor- tance. Subsequent authors who also used distance measurements were Atwood (1916, 1918), Bergman (1941 et seq.), and Brongersma (1951, 1957 a & b). Bergman expressed the organ positions and lengths as a per cent of snout-vent length, and both he and Brongersma also presented their data diagrammatically. Thompson (1913a & b, 1914) was the first to relate the position of the various visceral organs to the ventral scutes in an attempt to provide a simple, yet objective, technique for stating the location of the organs. The position of an organ was expressed as a percentage of the total num- ber of ventrals in order to compensate for individual, sexual, and geographic varia- tion in ventral number. This technique has been utilized subsequently only by Thorpe (1975), Underwood (1976), and Rasmussen (1979). Thorpe determined the midpoint of an organ rather than the anterior and posterior ends, so his data are not comparable to ours or to those of other authors. Inasmuch as one has to ascertain the anterior and posterior ends in order to determine the midpoint, the latter would appear to be an unnecessary complication and if used alone it also results in a loss of information. Garrigues (1962), Bogert (1968), Collins and Carpenter (1970), and Frenkel and Kochva (1970) also gave organ positions and lengths in terms of ventral number, but they did not express their data as a per cent of total ventrals. Also, by lumping his samples for each species, Garrigues failed to take sexual dimorphism into account. Valle (1944-45), Bragdon (1953), and Camazine et al. (1981) used ventral number to pinpoint the location of various posterior organs so that surgical proce- dures could be carried out using the smallest incisions possible. In each case, the investigator counted ventral scutes from the vent forward. Materials and Methods We examined 63 1 preserved adult speci- mens, representing 8 thamnophiine genera (only Adelophis was omitted because of its rarity) and 35 species (4 being represented by two subspecies or populations). Large subadults were used only if their data fitted into the range of variation for the taxon under consideration. Juveniles were rejected because their values tend to lie outside the normal range of variation in adults (see Bergman, 1958a, 1961b). Only nongravid females or those with undeveloped eggs were used because of the distortion caused by developing embryos (also noted by Bergman, 1961a; CoUins and Carpenter, 1970; Thorpe, 1975). Be- cause females tend to have their anterior organs situated more anteriorly and their kidneys more posteriorly than those of males, each sex was considered separately (see the Sexual Dimorphism section for further discussion). Using the Dowling method for counting ventral scutes, we inserted insect pins in the 20th scute and in every 15th scute there- after. Several midventral slits were made to expose the organs being studied. The ven- tral scute numbers at the anterior and posterior ends of each organ were re- corded; to faciUtate inter- and intraspecific comparisons, a percentage was calculated by dividing the scute number by the total number of ventrals. The following organs were considered where possible: heart, liver, gall bladder, pancreas, right and left kidneys. Lungs, thyroid, spleen, and adrenals were not considered because they were difficult to locate in many specimens. Testes and ovaries were not considered No. 2 Visceral Topography of Snakes 125 because of the varying size depending on whether the specimens were in a breeding or non-breeding state (see Matthews and Marshall, 1956; Manna and Sircar, 1978). Organ lengths, expressed as the total number of ventral scales covered, were also recorded and treated as a percentage of total number of ventrals. The following distances were measured and expressed in the same manner: posterior end of heart to anterior end of liver, posterior end of liver to anterior end of gall bladder, distance between or overlap of the right and left kidneys. On museum material other than that in the Louisiana State University Museum of Zoology (LSUMZ), only the heart, anterior end of liver, and kidneys were examined in order to minimize the number of incisions. Preliminary data on LSUMZ specimens had indicated that these organs were the most relevant to the study. The statistics used in the Inter- and Intrageneric Comparisons section con- sisted of calculating the mean, standard deviation, and standard error of the mean for each sex of each taxon, then construct- ing graphs by the Dice-Leraas method as discussed in Simpson et al. (1960). This method presents a graphic representation of differences between populations, and the results appear in Figs. 1-19. The 95% confidence interval of the mean was deter- mined by dividing the standard deviation by the square root of the sample size and multiplying this figure by a value from the Student's t-test table using n-1 degrees of freedom (Runyon and Harber, 1968). Be- cause of the very large confidence interval generated by a sample of two specimens, we constructed a Dice-Leraas diagram only in those cases where we had a minimum sample of three specimens of the same sex. The confidence interval results in a plus or minus figure relative to the mean. Where a determination of the statistical significance of the differences between means could not be obtained from this graphic representa- tion (using the three general rules on p. 353 in Simpson et al., 1960), then a Student's t-test was used. When data are stated as being significantly different in this paper, it refers to the fact that the differences are significant at the p< .05 level. To minimize the possible effects of geo- graphic variation, we attempted to sample populations from as restricted an area as possible. In four instances (Thamnophis couchii, T. elegans, T. sirtalis, Tropido- clonion lineatum) we treated different sub- species or geographically distant popula- tions as separate taxon samples. Because enough male and female Thamnophis eques could not be obtained from one geo- graphic area, we used females of T. e. megalops and males of T. e. virgatenuis. Due to the existence of sexual di- morphism, data for males and females could not be combined for discriminant analysis. We chose to restrict the discri- minant analysis to the data for males; only a relatively few confidence intervals could be shown for females on the Dice-Leraas diagrams because many of the confidence intervals exceeded the ranges of variation. Only those specimens that had data avail- able for all characters were used. Six variables (posterior end of heart, anterior and posterior ends of right kidney, ante- rior and posterior ends of left kidney, and kidney overlap) were first run after the values were standardized at the mean to allow for comparisons. Because the poste- rior end of the right kidney and the ante- rior end of the left kidney were signifi- cantly correlated, those characters were eliminated to obtain a four-variable explanatory and predictive model. The posterior end of the right kidney and anterior end of the left kidney values are reflected in the kidney overlap figures. Because of the relatively small number of specimens in each sample, the 294 speci- mens were placed in the following eleven groups to achieve greater statistical signi- ficance of the discriminant values: 1 . Clonophis kirtlandii — 6 specimens 2. Nerodia (cyclopion, erythrogaster, fasciata, rhombifera, sipedon, valida) — 51 3. Regina {alleni, grahamii, rigida, sep- temvittata) — 24 126 Tulane Studies in Zoology and Botany Vol. 23 4. Seminatrix pygaea — 8 5. Storeria {dekayi, occipitomaculatd) — 15 Thamnophis (groups from Ruthven, 1908) 6. Sauritus group (proximus, sauritus) — 14 7. Radix group [brachystoma,^ butleri, eques { = megalops in Ruthven), mar- cianus, radix] — 44 8. Elegans group [couchii couchii,^ c. hydrophilus,' elegans terrestris,^ e. vagrans,^ melanogaster, nigronucha- lis,^ ordinoides, rufipunctatus { = an- gustirostris in Ruthven), scalaris] — 63 9. Sirtalis group [chrysocephalus,^ cyr- topsis ('eques in Ruthven), godmani,^ sirtalis fitchi, ' s. sirtalis] — 47 10. Tropidoclonion lineatum (Nebraska, New Mexico, Texas) — 5 11. Virginia {yaleriae, striatula) — 17 Prior probabilities of group membership were calculated by dividing the number in any group by the total number in the study. These prior probabilities are used in classifying the specimens with the discri- minant model. Sexual Dimorphism Details on sexual dimorphism in this study appear in Tables I and II and in Figs. 1-19. A comparison of sexual dimorphism data from this study with other studies appear in Table III. Anterior organ positions The posterior end of the heart and the anterior end of the liver in males are located posteriorly to those positions in fe- males in 11^0 and 81 % of the taxa, respec- tively. Male Clonophis, Seminatrix, Stor- eria, Tropidoclonion (for heart only), and Virginia have the posterior end of the heart and the anterior end of the liver located posteriorly to those positions in females in all species. In Nerodia, Regina, and Thamnophis there is interspecific variabi- Uty in both features. Male Thamnophis 'taxon described since Ruthven (1908) Haxon not recognized by Ruthven (1908) have the posterior end of the heart situated posteriorly to that of females in 74% of the taxa; male Nerodia in 67<^o; male Regina in 67% . The anterior end of the liver in males lies posteriorly to that of females in 8O070 of the species of Nerodia, 73% of the taxa of Thamnophis, and in the only species of Regina for which data are available. Midbody organ positions Sexual dimorphism of the midbody organ positions is not pronounced. Males have the posterior end of the liver located posteriorly to that of females in 50% of the taxa, the posterior end of the gall bladder posteriorly to that of females in 69%. The posterior end of the liver is more poste- riorly placed in males in 60% of the species of Nerodia, both species of Storeria, and in the one species of Virginia examined. Males have the posterior end of the gall bladder located more posteriorly than do females in 60% of the species of Nerodia, 70% of the taxa of Thamnophis, and in the one species of Storeria examined. Posterior organ positions In contrast to most of the preceding characters, the kidneys exhibit marked sexual dimorphism in many of their fea- tures. The anterior ends of the right and left kidneys in males are anterior to those of females in all taxa, as are the posterior ends of the right and left kidneys in 86% and 73% of the taxa, respectively. The posterior end of the right kidney in males is situated anteriorly to that of females in all species of Regina, Seminatrix, Tropido- clonion, and Virginia, and 95% of the taxa of Thamnophis. In Nerodia the posterior end of the right kidney of males is situated posteriorly to that of females in 67% of the species. In both species of Storeria and in half the species of Nerodia, the posterior end of the left kidney of males is situated posteriorly to that of females. The poste- rior end of the left kidney of males is anterior to that of females in 95% of the taxa of Thamnophis, 67% of the species of Regina, one population of Tropidoclo- nion, and in both species of Virginia. No. 2 Visceral Topography of Snakes 127 Table 1. Sexual dimorphism in certain thamnophiine snakes. Character post ant . post post ant . post ant . post , heart^ liver . liver i gall bladder r. kidney • r. kidney 1 . kidney 1. kidney 35 27 18 16 37 37 37 37 Position in eft? Position in 99 posterior to posterior to that in 99 (or that in cfcC (or o" organ longer') 9 organ longer") 7 77. 17% 81% 15% 50% 44% 69% 31% 0% 100% 14% 86% 0% 100% 24% 73% liver length 17 gall bladder length 16 r. kidney length 3 7 1. kidney length 37 post, heart-ant. liver interspace 26 kidney overlap 3 7 76% 69% 5% 0% 2 7% 13% -a' significantly cfd=99 different (pS05) from 9? 6% 4% 6% 0% 0% 0% 0% 3% 6% 0% 3% 0% 0% 3% :>7% 41% 22% 25% 81% 49% 86% 32% 18% 6% 46% 73% 12% 22% means of the taxa were used in computing the figures in this table Organ lengths and interspaces The liver and gall bladder of females are longer than those of males in 76% and 69% of the taxa, respectively. However, the right and left kidneys of males are longer than those of females in 92% and 100% of the taxa, respectively, probably due to the presence of a hypertrophied sexual segment in males (Matthews and Marshall, 1956; Prasad and Reddy, 1972). The male heart-liver interspace is longer than that of females in 73% of the taxa, as is the male kidney overlap in 84% of the taxa. In 80% of the taxa of both Nerodia and Thamnophis, males have a shorter liver than do females; the hver is also shorter in male Storeria dekayi (in S. occi- pitomaculata the Hver shows no sexual dimorphism). Data were available for both sexes in only one species of Regina and one of Virginia. Males have a shorter gall blad- der than do females in 80% of the species of Nerodia, in 70% of the taxa of Tham- nophis, and in Storeria dekayi. Males of Clonophis, Nerodia, Regina, Seminatrix, Storeria, Tropidoclonion, and Virginia have longer right and left kidneys than do females. In all taxa of Thamnophis, males have a longer left kidney than do females; in 86% of those taxa, males also have a longer right kidney. In all species of Clonophis, Regina, Seminatrix, Storeria, and Virginia, males have a longer heart- Hver interspace than do females, as is the case for 60% of the species of Nerodia and 64% of the taxa of Thamnophis. In all species of Clonophis, Nerodia, Regina, Seminatrix, Tropidoclonion, and Virginia, males have a more extensive kidney overlap than do females, as is the case for 76% of the taxa of Thamnophis. Asymmetry of kidney lengths In 76% of the taxa, females have the right kidney longer than the left (24% differ significantly). On the other hand, males have the left kidney longer than the right in 55% of the taxa (5% differ signi- ficantly). In all species of Nerodia, the right kidney is the longer one in both sexes. Summary and conclusions In general, the anterior and midbody organs are placed more posteriorly in males than in females, whereas the kidneys of males are positioned more anteriorly than those of females. This more anterior positioning of the anterior organs and more posterior positioning of the posterior organs in females would allow greater space for the developing young. Inter- and Intrageneric Comparisons To facilitate comparisons, each set of 128 Tulane Studies in Zoology and Botany Vol. 23 dE'xaaAO Xsupiii ^OOC/iOO C/50C/) ■b "b 'b^ ■b -D ^ tj ^ z ■b /I CO z z bo yjtotototototoocococototncotocotoo "3t2i3 zzzzzzz -zzzzzzzzz^zzz bo'b'bO'b'b'b'b'D'b'b'D'b'b'bo*" ^ii°* "b z z ■b -b aOEdSJ33UT J8ATT -IJEBl) b ■b O 1 jn c/1 c/1 t/) o 1 Z Z Z Z • 1 ^■bf>--b ^ Z 1 1 ■b z ■b II 1 IZZZZZZIZZZI IZZZI 'g^ 'b^i'boo'b ooo 'o'b'b '^ "o 1 to o o ■b ~^l o o -1 in -I o o o o o o ■b 'b -b -b ■b -b *-l n o o o to ■bbl z •b o to O oo^^ .-H^o o inocNo^ o ^ ooootoooocootooooootoo ^o^ ■b-b'o'bl'b'b^^'b^'b'b'b'b'b^'b ■b'b'b o o b o CO o H3du3x A3Upi51 -J o b -M -1 O -H m O O O O CO o •b "b •£ 'b^ "b to o to CO z ■b o CO o .-1 O O .H .H O totototoototootoototootoococoo MCOCO zzzz-zz.*z.*zz.*z.-zz.* zzz ■o'b'b'b^'b'b^'bT^'b'b^'b^'b^^ iioo o o •b o o ■b-b II53U3X •Xq TIES 1 z z z z ri 1 0.0.r,.,> -n 1 1 1 1 Z 1 ■b 1 1 ! itototootoitocoi icocol 1 'H;' 1 1 1 IZZZ -ZIZZI IZZl 1 IZl ■bo'b'^'fa ^'> c/*o* o ] 1 1 J3AIX 1 c/1 c« to O C/1 1 Z Z Z • Z 1 o. a. "D o. c» 1 1 1 1 to CO z z 1 1 1 itotototooitotoi ItOOl 1 ii"] 1 1 1 IZZZZ -IZZl IZ-I 1 IZl o ■□ ooO "bo o*o o* i - ASUpi^l "X •3S0d ■b z z z z z z •b -0 o ■bOo z z z z o z z ■b -b inorgino-— 1 in CN m t-HOm tococooooooococoocoocoococo ooo ^ooooooooooo'-b'^o^o'b <><>o* z o z z o o* Aaupi^ -X •3UB 1 O ^ O CNi ,-( CW o o o o o ^ o O O C« o o to O o o 0'-l*H^OOOOOO^'-l O^OtM^*-^OC^ OOOOOOOOOOOOCOOOOOOOOO o*oo*oooooooooo*o*oooo ooo o o o o o (> o* CH Cw o o ^ ^ O* O* o* Aaupt^i -J ■ 3Sod 1 en c/l en to en c« z z z z z z •d o -D -o 'b o to to CO z z z 0*0 0* z o to o z • oo ,_4,_i(/^,_)0'Hinooo CM i-i cNjin --lo ooooooooootoocootoootoooto 2 -2 'Z • 'Z • 'Z OOOOOOOOOOo.O*oOo*OOi^3 OOo* z o to O o °* Xaupt^i -a •3UB 1 C -^ C tn o o tn to o z • • z z • 0+0*0*0 0*0* in o o o to O O o* o o o to o ^r-< .H^^ ^r-I.H ^ -H.-I •-* ^ OO^OOO-HOOO'-lO.-I^OOrHO ino oooooooooooooooooo oo^ o o o tN O o o o o o o 000*OOOC>000*0*IOOOOOOO OOo •xq -xTeS •qsod 1 1 1 to to to to to I Z Z Z Z Z 1 o-b-oo*^ 1 1 1 1 in 1 O 1 -6 ' llll cNinl 11^ III 1 1 1 1 itntocoooitotoi lotoi 1 igl 1 1 1 IZZZ ■ -IZZl 1-ZI 1 IZl ■b ■doO*'^ -b -o ^ -b -b 1 1 1 jaAXX • 5Sod 1 to to to CO to 1 z z z z z 1 0+ "b "d o* "o 1 1 1 1 1 1 1 c o to b? 1 1 1 Itotototooitotoi lOOl 1 1^' 1 1 1 IZZZZ -IZZl l.-^-l 1 IZl o*.boo^ •bo bo II 1 1 1 z J3ATX ■3UE O ■d O eg 1 to o to to O 1 z • z z • 1 ■b b o* -o •□ O 1 1 ■o o o o to • z ■d 'b 111 r-H 1 1 in 1 --^ 1 itoitototoototototoi itootoiotoco 1 IZIZZZ-ZZZZI IZ.-ZI'ZZ •d fe b II ^ "d o -d o -d d o ^ ■o\) i ^ o o o aJEaq •jsod z u-1 .-1 O o o to to o to • ■ z z • z ■b ■b 'b o •b o* to to to z z z •b o "b o ■b o o to • z ■b -b 1 .H o in .-1 1 in tococoto 1 totootototocoototootoo oto ZZZZIZZ -ZZZZ -ZZ -Z •! -z •b'b'bb ^■b>Doii'bo'b'b'b'bo'b "bii o o O '-I o o ■b -b CO / CO / •C / CJ / / ^ / c3 'a c c: 1. a: ."■ C c K O .Sf -til ^ tJ 't! < <3 s; <32 to « Cm 1 1^ o o •tJ 'pi 2i SX 13 O .a !, CO o to g^ en's P ^ V a 5^ .^ -^ ^.QU5jy^coco CO 3 « c o £X o C3 S* •vi IS (3 •t-1 i No. 2 Visceral Topography of Snakes 129 Table III. Data on sexual dimorphism reported In the literature. X indicates that the organ is longer or located more caudally in sex indicated; ND that there is no appreciable dimorphism. ^ ^ ^"■^"--^^^ Character ^ u ^ c c Xi ^ > >. r. •o £ TJ X 4J ■M U Taxon ^^■^'^^ M OO J^ 00 Ul > tfl T3 *j -a Ul T3 CU C > c c O tu C -H c ■-< 0 ^ C -H O'-H C -H CO M • u ■U QJ m > O -H OJ • >-i •U U tfl c o « c T3 ■H • J^ U C • c • -H o • a. u > C ■a ■H c • c •H Q> OJ c •H U c • QJ U rH QJ c T3 x: •H U J^ 00 c • OJ 1— 1 tH a, >. n) QJ r-\ C u t3 QJ ■H > ^ O QJ OJ > ^ H O. 1 tn ■u U U QJ 03 AJ QJ C a -H o" 9 cf 9 d 9 (^ 9 a- 9 d 9 a- 9 d 9 Cf 9 o" 9 Cf 9 Cf 9 d 9 Clonophis kirtlandii L M M L - - - - H H H H H H M M M M M M M M M L Nerodia cyolopion M M M M M M M M M M H M M M M M M L M M M M H H M M erythrogaster L M M L M M M M M L M M M L M M M L M H H H H M M M fasoiata M M M M M M M M M M H H M M M M M L H H H M H H M M rhombifera H H H M M M M M M M H M M M M M L L M H M M H H M M sipedon M M M M M M M M M M H M M M M M M L M H M H H M M M valida M M - - - M - L M M M M M M M M - - M M M M H M Regina alleni H H H H _ _ _ _ M M M M M M M M _ _ M M M M M M H H grahamii M M - M - M - M M M H H M H M H - L M M M M M M - M rigida M M - M - M - H M M M M M M M M - L M M M M M M - M septemvittata M - M - M - - - M - M - M - M - M - M - M - M - M - Seminatvix pygaea H H H H H H H H H H M H - - L M L L M M H hI Storeria dekayi M M M L H M H M M M M M H H H H H M M M M M M M M M OGcipitomaoulata M M M L M M H - M H M M H H H M H M M L M L L L M M Thamnophis proximus M M M M M M H H H H H H H H H H M L M M M M M M M M Sauritus sauritus H H - M - M - M H H H H H H H H - L M M L L M M - M group 1 Thamnophis brachystoma L M _ _ _ _ _ _ M M M M M M M M _ _ M M M M M M _ _ butleri M M - - - - - - M M M M M M M M - - M M M M M M - - Radix eques L L L L L L L L L M L M L M L M L L M H M H M M M - group marotanus L L L L L L M M M M M M M M M M M L M H M M M M M M radix L L L L L L M M M M M H M M M H M M M H M H M H M L Thamnophis aouchii A M M L - L - M - L M L M L L L M L - M M M M M M M L couahii B M M M L M M M M L M L M M M M M M L M M M H M M M M elegans A L L L L L L M - L L L M L L L M L L M H H M M M M M elegans B L L L L L L M M L L L M L L L M M L M M H H M M M L Elegans melanogaster M H M M M M M L M M M M M M M M M L M H M H M M M M group nigronuchalis M M - L - M - L L L L L L L L L - L M M M M M M - M ordinoides L M - - - - - - M M M M M M M M - - M M M M M M - - rufipunctatus M M M M - - - - L M L M L L L L - - M M M M M M M M soalaris - L M L - - M M M M M M M M - - M H M H H H - L Thamnophis chrysoaephalus M M M M - _ _ _ M M M M M M M M _ _ M M M M M M M M ayrtopsis L L L L L L M L L L M M L L M M M L M H M H H H L M Sirtalis godmani L M M L L M L M L M M M - - M M H M M M M M group sirtalis A L L L L L L M M M M M M M M M M M L M H M M M M M M' sirtalis B L L L L - M - M M M M M M M M M - L M H M M M M M M Tropidoalonion lineatum A L L - - - - - - M H M M H H M M _ _ L L L L M M _ _ _ linp.ntum B - L - L - M - H - H - M - H - M - M - T, - T, - T, - M Vvrg%nia striatula M M M L - M - M H H M H H H M M - M L L L L L L M L Valerias M M M L H H - M H H M M H H M M H H L L L L M M M L Ruthven's species groups No. 2 Visceral Topography of Snakes 131 "' Males 14 16 18 20 22 24 26 C. kirtlandii N. eye lop ion erythrogaster fasciata rhombifera sipedon valida R. alleni grahamii rigida septemvittata Se, pygaea St. dekayi occipitomaculata Th. brachystoma butleri chrysocephalus couchii A couchii B cyrtopsis elegans A elegans B eques godmani marcianus melanogaster nigronuchal is ordinoides proximus radix rufipunctatus sauritus scalaris sirtalis A sirtalis B Tr. 1 ineatum A lineatum B V. striatula valeriae I I - ^• r -Hi' f^ -1-5 Figure 1. Location of the posterior end of the heart in thamnophiine snakes (expressed as a % of total ventrals). Construction of this and subsequent graphs is explained on pp. 127-129 132 Tulane Studies in Zoology and Botany Vol. 23 aoBdsja^ui: Xaupi:^ q^Suai; Xaup"p^ '2 jaATi •3Sod Xaupx^ 'X -* vO ro CT\ lO r~ cs >— I I— t u-i 0-) iH vo rH r^ I r-| t-t CN ~3- r^ I r^ r^ -J- ~* iH ro U-| iTl lO lO lA -J" ON I I r^ n in vo I 1 I I l~- O rS; -v^ rCl "tS ^ t-^ +^ 1^ E o in CO "5^ s 35 ^ « , rs; Q) CO •r^ w O '--i ^"ll t) -t^ a +^ ?; ?; 'tl S^ 3 (■V (3 « ^ rCl *i) fc- !^ (3 U No. 2 Visceral Topography of Snakes 133 ^ vO^Om^o^oo^ 1 O^O vO V.O ^D 0^ 0^ ■H 1 \0 CM 3DBdsa9:}ut (NCS)CNH (N CM iH iH r-s 1 CN CO a3Ai][-5ap3q m a> r^ in rH 1 i ~* i in sj- o o CM 1 1 CM m rH rH fO rH ,H 1 1 O" 1 rH sj- rn rO CN 1 1 CN iH ^,HcNjrgvooo >H rH c^ m o CO -3- 00 CJs dB^aSAO 1— l-a'iHirivJ-vOCMi— IvO in m -a- -* r-i r^ 00 r^ in AaupT^ •— lrO<3-CTN^r^lrlCT^<^J .H ro iH CJN 00 CM VO m CN qaSua^ ro r-s rH tH 1 in in 1 in 00 vO 1 m CM m cvi ro 1 1 1 1 r^ 1 CN CM 1 rs 00 OS aaATj CM rH ,H M ^ 1 1 1 1 1 ^O 1 .H 1 1 1 1 1 -* CNI CN CM CN 1 1 1 1 1 CN 1 CN 1 1 1 1 1 XaupT^i '\ ro m 00 r^ r^ \o 00 1 r-s r^ iC9up-i2i -a vCvO^fvOONr^LOvOOO in CNi ^o in CT> r-s 00 so in oor-~oooo 1 lO ro n 1 1 1 1 m 1 ro ^ vo 1 1 1 1 m 1 in cN 1 vO 1 rO CM 1 .H 1 OS 00 00 St sl- • 3 sod m u~i ro vo r-^ 1 1 1 1 vo rn in r^ ^ 1 1 1 1 1 CN 1 cys 1 1 00 1 in 1 1 1 1 1 1 1 1 1 1 O 00 ^o o vr> 1 1 1 1 00 1 o 00 1 vo 00 so a9AT^ 1 vD ^D r-^ vC c^ i 1 i 1 00 1 00 sl- 1 r-s so CJs •3 sod CTN to CO (3 rs; < m -w +i 4J / s -t-i -t-i s s (3 i-i / •r^-r^csoao S !^ o ex, s f-i t-i •S O •rJ cu / y y (33 CO CO •fJ SL, ^-i / SS ■t-^ <3 / CO CO O CO ;i / •ri •r^ t-^i / r«; rs; CO ti a / fX CO EX -^-i rS •t^ / o p O t-^ ^ s / S s 5 CL S « 5- •V^ •t-i E (33 3 fc 4-i 3 £X <35 / X <3 H o en cJ^ r^ O en -3- en \D osvocnincMp^cMvo ooo 0 ^ .— t CM ^ CNl tH en CM rH cMCM>3-cn VO 0 -a- in-l rHrHiHCMrHrHiHCM r-\ ^ r-A 00 CN CM -3- O 00 in CM CM CJsos . .vXJOsOOSinsOOsOsOO 00 00 • in fi . o • • • • u ^ T— 1 ^D ^ r-^ r^ o o .-H -H »--i s CO <3 -w -u C3 <: oa 0 pa S « Sh +^ ^ <:pqco<:ffl toco.* •vi S«Oeo cococo ■tj <3 O C35 <3 ^ ■^ -t-i c^ !S g t-i ca •ri O 4^ 't-. C •-^a,Ks; «osE +jtoTO r~-i +i « « S a ») ix ,«.c.s;oooto-^s;ov^"^'3a OS3?HC»(3>0^3t.SH poo Sj'--' t-^ <3- (3 -T-i (i .s; -^ -u trS'?^ .S ^ •t^ o Si CO o a.'-i t3 S c cs Q> y •t^ -tJ ■tJ -M « O 0 t) C» H-, Ss CO a •- talis group are distinguishable from each other. In the classification matrix (Table VIII), members of the Sirtalis group are misclassified as members of the Elegans group much more frequently (42. 6*^0) than they are correctly classified (29.8%); they also are often misclassified (21.3%) as members of the Radix group. One of the most interesting results of the discriminant analysis is the wide separation of the Sauritus group from the other three groups (see Fig. 20). Conclusions Visceral topographic data alone do not clearly delimit thamnophiine genera nor establish inter- or intrageneric relation- ships. Nevertheless, some trends are % cyclop ion erythrogaster fasciata rhombifera sipedon valida R, grahamii rigida St. dekayi occipitomaculata Th. couchii A couchii B cyrtopsis elegans A elegans B eques marcianus melanogaster nigronuchal is proximus radix sauritus sirtalis A sirtalis B Tr. lineatum B V. striatula valeriae 46 48 50 52 54 56 58 60 62 64 66 Figure 7. Location of the posterior end of the pancreas in female thamnophiine snakes (expressed as a % of total ventrals). 144 Tulane Studies in Zoology and Botany Vol. 23 o o o o ^r~- OO OO OO OO OO OO OO OO 0-10 oo .HO OO OO OO OO tr^o- OOCsl OvO OO OO cnon cNro OO c^jn OO oOvT) cn(>j on OO OO OO H04 OO Or^ OO x>m n-) 00 rv. 00 0» « GO o^ ^^ 03 00 ^ 00 OO 00 gg Oe r^. CD a ^» 5J) oc t/5 II II m 0\ ti .5" t3 P (11 !>: ^ S II a * CM h. •a 0 -sr Q a O 3 sr 0 o O lo ■5 s c a B 3 0 0 tuj ■T3 C 0 a C 0 ^, w •a II w n, (Tl 3 x> hi «5 ■a ou -c II 8 ^ * c c •«: cw S i^ 0 0 II ■3 a 3 < 0 0 on 0 fi !0 s- PS ri u CHI \^ n) >t 0 tM ■x:i ^ Ci. C II 0 0 VO ft; 0 ^ II ^ v, s: ? V <3 K, a sr tM ?= 0 51 iX CO 3 II 0 ■* t3U 158 Tulane Studies in Zoology and Botany Vol. 23 groups designated by Ruthven (1908) do not appear to be distinguishable from one another solely on the basis of visceral topo- graphy. As was implied above, in many cases the small, semifossorial thamnophiines tend to have a posterior displacement of organs, a condition they share frequently with the ribbon snakes (Thamnophis proximus, T. sauritus) and occasionally with some species of Nerodia and Regina. Posterior displacement is a general trend, not an in- variable phenomenon, and both inter- and intrageneric variation occur from one character to the next. The semifossorial genera also show a definite trend toward Table IX. Significance of Ruthven' s Thamnophis groups compared as four se irate populations. NS indicates difference not significant at p>.05. Character Sex Sauritus- Radix Sauritus - Elegans Sauritus - Sirtalis Radix- Elegans Radix- Sirtalis Elegans - Sirtalis post, heart 9 .001^ .01 .001 .001 .001 .001 NS NS NS NS NS NS ant. liver d 9 .001 .001 .001 .001 .001 .001 .05 .05 .05 .001 NS NS post, liver d 9 .001 .001 .001 .001 .001 .001 NS NS NS NS NS .05 post, gall bladder d 9 .001 .001 .001 .001 .001 .001 NS NS NS NS NS NS post, pancreas cf 9 .001 .001 .001 .001 .001 .001 NS NS NS NS NS NS ant. r. kidney d 9 .001 .01 .001 .001 .001 .001 .001 .001 .001 .01 .01 NS post . 1. kidney d 9 .001 .001 .001 .001 .001 .001 .001 .001 .PI .dOl NS NS ant. 1 kidney d 9 .001 .001 .001 .001 .001 .001 .001 .001 .00- .001 .01 NS post 1. kidney d 9 .001 .001 .001 .001 .001 .001 .001 .001 .001 .001 .001 NS liver length d 9 NS NS .02 NS .01 NS NS NS NS .02 NS NS r. kidney length d 9 .01 .001 .01 .001 .01 .001 NS .01 NS NS NS NS 1. kidney length d 9 .001 .001 .001 .001 .001 .001 NS NS NS NS NS NS heart-liver interspace cC 9 .02 NS .01 NS NS NS NS NS VS NS NS NS kidney overlap d 9 NS .001 NS .01 NS .001 NS NS NS K: NS NS Significance levels determined using 2-tailed Student's t-test. No. 2 Visceral Topography of Snakes 159 having relatively short kidneys, but the data for Storeria are equivocal and the characteristic is not unique to those genera. One unique feature that is shared by all of the semifossorial genera is the possession of a relatively long liver. We do not know why small snakes would possess a proportionally longer liver than large snakes, but perhaps there are physiological constraints that prevent the mutual reduc- tion of body and of liver from being directly proportional — perhaps a mini- mum quantity of liver tissue is required for the proper functioning of that organ. Acknowledgments For the loan of specimens and for other courtesies, we are indebted to the follow- ing curators: Charles W. Myers and Richard G. Zweifel (American Museum of Natural History); Harry A. Shankman (Arizona State University); Douglas C. Cox (Brigham Young University); C.J. McCoy, Jr. (Carnegie Museum); Hymen Marx (Field Museum of Natural History); Walter Auffenberg and Peter Meylan (Florida State Museum); Kenneth CHffer and Philip J. Regal (James Ford Bell Museum of Natural History); Joseph T. Collins and William E. Duellman (Uni- versity of Kansas Museum of Natural History); Gloria Z. Wurst and David B. Wake (Museum of Vertebrate Zoology, University of California at Berkeley); Harold A. Dundee (Tulane University); T. Paul Maslin (University of Colorado Museum of Natural History); Gary Brei- tenbach and Arnold G. Kluge (University of Michigan Museum of Zoology); James F. Jackson (University of Southwestern Louisiana); Jonathan A. Campbell (Uni- versity of Texas at Arlington); and Robert G. Webb (University of Texas at El Paso). We are also grateful to Darrel R. Frost, Mark S. Hafner, Dominique G. Homber- ger, and Randy H. Vaeth for helpful sug- gestions at various stages in the develop- ment of this manuscript. Specimens Examined' Clonophis kirtlandii. ILLINOIS, Christian Co.: LSUMZ 40065; Cook Co.: FMNH 23166, 25437; /Gross Pt.y: FMNH 2989; Will Co.: FMNH 55562, 65902. INDIANA, Delaware Co.: FMNH 64670; Porter Co.: FMNH 42069; /Orange Co.?y FMNH 3060. KENTUCKY, Jefferson Co.: FMNH 25535. OHIO, Hamilton Co.: LSUMZ 7445, 13539. Nerodia cyclopion. LOUISIANA, Ascension Par.: LSUMZ 13703; Calcasieu Par.: LSUMZ 12150; Cameron Par.: LSUMZ 18671-2; Iberville Par.: LSUMZ 18286, 20703, 24669; Jefferson Par.: LSUMZ 8670, 13704; Lafourche Par.: LSUMZ 13557, 19183; St. Bernard Par.: LSUMZ 9280; St. Charles Par.: LSUMZ 18757, 29355; St. James Par.: LSUMZ 18293, 19174; St. Tammany Par.: LSUMZ 34308; Vermilion Par.: LSUMZ 24025, 33939. Nerodia erythrogaster. LOUISIANA, Acadia Par.: LSUMZ 20310; Cameron Par.: LSUMZ 20344; East Baton Rouge Par.: LSUMZ 17321, 17702, 19175, 20312, 20723, 22909, 24028; Iberville Par.: LSUMZ 18287, 22558-9; Jefferson Par.: LSUMZ 18716; Livingston Par.: LSUMZ 28812; St. Ber- nard Par.: LSUMZ 8992; St. John the Baptist Par.: LSUMZ 23864; St. Tammany Par.: LSUMZ 12983, 20279; Vermilion Par.: LSUMZ 34295; Washington Par.: LSUMZ 12540; West Baton Rouge Par.: LSUMZ 11887; West FeHciana Par.: LSUMZ 18758. Nerodia fasciata. LOUISIANA, Ascension Par.: LSUMZ 17698; Cameron Par.: LSUMZ 12731, 17315, 20281, 28666; Jefferson Par.: LSUMZ 8947, 8953; Natchitoches Par.: LSUMZ 30410; Plaque- mines Par.: LSUMZ 8653; Pointe Coupee Par.: LSUMZ 20274; St. Charles Par.: LSUMZ 7142, 7527; St. Landry Par.: LSUMZ 18113, 18122; St. Martin Par.: LSUMZ 19171, 19173. Nerodia rhombifera. LOUISIANA, East Baton Rouge Par.: LSUMZ 17687, 17794, 17945, 20799, 23662, 28008-10; Iberville Par.: LSUMZ 13756; St. Charles Par.: LSUMZ 9216. Nerodia sipedon. ALABAMA, Jackson Co.: LSUMZ 36375; Pickens Co.: LSUMZ 36399, 36400. ILLINOIS, Jackson Co.: LSUMZ 27610; Pope Co.: LSUMZ 27599. MISSISSIPPI, Greene Co.: LSUMZ 36379, 36381-3, 36385, 36387, 36390-3, 36396-7; Lauderdale Co.: LSUMZ 36403-4; Wilkinson Co.: LSUMZ 28712. MISSOURI, Lawrence Co.: LSUMZ 9107. Nerodia valida. MEXICO, Colima: LSUMZ 7876; Nayarit: LSUMZ 33099, 36266, 36268; Sinaloa: AMNH 36269, 84077, 84080-2, 87575, 87577, 88889-90, 88892; Sonora: AMNH 84074-6. Regina alleni. FLORIDA, Alachua Co.: FSM 2476, 2498, 6634, 6637, 7171, 9096, LSUMZ 13618-9; Collier Co.: LSUMZ 28992; Dade Co.: FSM 42527; Dixie Co.: LSUMZ 7473; Hillsborough Co.: FSM 42529; Indian River Co.: FSM 42524-6, 160 Tulane Studies in Zoology and Botany Vol. 23 42530; Polk Co.: FSM 1868; Sumter Co.: FSM 11157. Regina grahamii. LOUISIANA, East Baton Rouge Par.: LSUMZ 17947, 33460, USL 7623; Iberville Par.: LSUMZ 20271; Lafayette Par.: USL 20945; St. Landry Par.: LSUMZ 28665, USL 15936, 23236, 23414, 23427; St. Martin Par.: USL 22953, 24432; Terrebonne Par.: LSUMZ 36484-7; Vermi- lion Par.: USL 10687, 17353. TEXAS, Chambers Co.: LSUMZ 33462. Regina rigida. NO DATA: USL 6067, 8820. LOUISI- ANA, Iberville Par.: LSUMZ 22556; Lafayette Par.: USL 24245; Natchitoches Par.: LSUMZ 12988; Orleans Par.: LSUMZ 8982-3; Sabine Par.: USL 24453; St. Charles Par.: LSUMZ 8680; St. Landry Par.: USL 15930, 17620; St. Martin Par.: USL 14365, 19471, 22425, 24433; Terrebonne Par.: LSUMZ 36483. Regina septemvittata. ALABAMA, Baldwin Co.: LSUMZ 15783. NORTH CAROLINA, Orange Co.: LSUMZ 14353^. OHIO, Montgomery Co.: LSUMZ 24476, 30184-5. TENNESSEE, Clay Co.: LSUMZ 34795; Jackson Co.: LSUMZ 34798. Seminatrix pygaea. FLORIDA, Alachua Co.: FSM 9813 (-6,-12), 14146 (-4), 14147 (-1,-7), 14215 (-4), 14216 (-2,-4,-9), 14217 (-3,-5,-7), 14218 (-4,-6); Dade Co.: LSUMZ 6530, 7405, 24582. Storeria dekayi. LOUISIANA, Ascension Par.: LSUMZ 18776; Cameron Par.: LSUMZ 2764, 12196, 18168-70, 24038, 28819-20, 28822, 29977, 32649; Iberia Par.: LSUMZ 2771; Iberville Par.: LSUMZ 12229, 23877; St. Landry Par.: LSUMZ 18665, 20074; Vermilion Par.: LSUMZ 24733. Storeria occipitomaculata. LOUISIANA, Claiborne Par.: LSUMZ 24658; East Feliciana Par.: LSUMZ 16686; Natchitoches Par.: LSUMZ 24745, 33077-8; West Feliciana Par.: LSUMZ 12602, 17898. Thamnophis brachystoma. PENNSYLVANIA, Clar- ion Co.: CM 28292-3, 28295, 28297-9, 28302-3, 28306-9, 28311, 28313, 28317-8, 28320-1. Thamnophis butleri. CANADA, Ontario: UMMZ 90193. INDIANA, Noble Co.: UMMZ 132822. OHIO, Lucas Co.: UMMZ 68864, 99627(3). MICHIGAN, Sanilac Co.: UMMZ 98774; Wash- tenaw Co.: UMMZ 465234; Wayne Co.: UMMZ 89519. WISCONSIN, Waukesha Co.: AMNH 76178-80. Thamnophis chrysocephalus. MEXICO, Guerrero: AMNH 72500-1, 72503; Oaxaca, AMNH 91094-5, 93235, 97855-6, 97865-6, 97868-9, 97871. Thamnophis couchii couchii. CALIFORNIA, Ama- dor Co.: LSUMZ 16530, 16544; Kern Co.: LSUMZ 16549; Shasta Co.: LSUMZ 22938, 34587-8, 34590, MVZ 18824-5; Tehama Co.: LSUMZ 16550; Tulare Co.: LSUMZ 16547; Tuolumne Co.: LSUMZ 34585. Thamnophis couchii hydrophilus. CALIFORNIA, Humboldt Co.: LSUMZ 34578; Shasta Co.: LSUMZ 1655M; Trinity Co.: LSUMZ 34594-5. OREGON, Jackson Co.: LSUMZ 16560-4, 16567. Thamnophis cyrtopsis. ARIZONA, Coconino Co.: LSUMZ 29940, 30062, 30083, 30088; Gila Co.: LSUMZ 30061; Maricopa Co.: LSUMZ 30063, 30081; Pima Co.: LSUMZ 30066, 30090; Santa Cruz Co.: LSUMZ 10035, 30072, 30076-7; Yavapai Co.: LSUMZ 29943, 29945-6, 29948, 30064-5, 30067-8. Thamnophis eiegans terrestris. CALIFORNIA, Mendocino Co.: LSUMZ 34378, 34380; San Mateo Co.: LSUMZ 7922, 16502-3, 16507, 34371, 34373; Sonoma Co.: LSUMZ 34368-9, 34374-5; Sonoma- Mendocino Co.: LSUMZ 34367. Thamnophis eiegans vagrans. NO DATA: LSUMZ 20747-50. ARIZONA, Coconino Co.: LSUMZ 29957, 29959-62. COLORADO, Conejos Co.: LSUMZ 11571, 11609, 11611, 11615, 30051, 30055; Costilla Co.: LSUMZ 7985, 11603-5, 11607, 11614, 11618, 13929, 13931-2, 30050; Rio Grande Co.: LSUMZ 30056. Thamnophis eques megalops. MEXICO, Chihuahua: AMNH 104471, 104772, BYU 22701; San Luis PotosK LSUMZ 4374-5, 4879. Thamnophis eques virgatenuis. MEXICO, Durango: AMNH 102521, LSUMZ 16424-6, 16429-30. Thamnophis godmani. MEXICO, Oaxaca: AMNH 89604, 91101-2, 91105, 97853, 97873-4, 97884, 97888, 103090, 103092-5, 103101, 103103, 103105, 103113, 104394, 106993, 106995-8, 107002-5, 718170. Thamnophis marcianus. TEXAS, Bexar Co. LSUMZ 10315; Duval Co.: LSUMZ 23239, 23243 Hartley Co.: LSUMZ 10407; Jeff Davis Co. LSUMZ 29608; McMuUen Co.: LSUMZ 23248 Moore Co.: LSUMZ 10365; Presidio Co.: LSUMZ 23255; San Patricio Co.: LSUMZ 23249, 23252; Webb Co.: LSUMZ 30929; Zavala Co.: LSUMZ 23254. Thamnophis melanogaster. MEXICO, Jalisco: LSUMZ 16434; Michoaca'n: LSUMZ 14489-90, 14492-3, 16435, 34346, 36277, 36279-80, 36282-6. Thamnophis nigronuchalis. MEXICO, Durango: LSUMZ 11637, 16448, 16450-5, 16459-60, UTEP 3386-7. Thamnophis ordinoides. CALIFORNIA, Del Norte Co.: MVZ 30276-7, 30279. OREGON, Clatsop Co.: MVZ 34265-8, 36848; Polk Co.: MVZ 24808; TUlamook Co.: MVZ 47856. WASHINGTON, Clark Co.: MVZ 34259; King Co.: MVZ 38653, 38655, 38657, 38670, 38674; Lewis Co.: MVZ 70366; Pacific Co.: MVZ 34262. Thamnophis proximus. LOUISIANA, Acadia Par.: LSUMZ 17899; Cameron Par.: LSUMZ 33964; No. 2 Visceral Topography of Snakes 161 Claiborne Par.: LSUMZ 33966; East Baton Rouge Par.: LSUMZ 16912, 18714, 20254; Iberia Par.: LSUMZ 18077; Iberville Par.: LSUMZ 20255, 20316, 22548; Livingston Par.: LSUMZ 7960, 18974; Pointe Coupee Par.: LSUMZ 20220; St. Tammany Par.; LSUMZ 7934; Vermilion Fax.: LSUMZ 24052. TEXAS, Hidalgo Co.: LSUMZ 18621-3. Thamnophis radix. NO DATA: LSUMZ 20735^0, 20742-5. COLORADO, Denver Co.: LSUMZ 7465; Larimer Co.: UC 31837^0, 31842-3, 31847, 31851, 31873, 31888. ILLINOIS, Iroquois Co.: LSUMZ 8126. NEW MEXICO, San Miguel Co.: LSUMZ 7942, 7944, 7972. Thamnophis rufipunctatus. ARIZONA, Coconino Co.: LSUMZ uncatalogued, LSUMZ 36815. MEXICO, Chihuahua: AMNH 4342, 68286, ASU 17042, 5304-5, 5335, UTEP 2043, 2262, 3657. Thamnophis sauritus. FLORIDA, Alachua Co.: FSM 14183, 14550 (-1), 14550 (-2), 14550 (-3), 14550 (-4), 14550 (-7), 14550 (-8), 14550 (-9), 14551 (-2), 14551 {,-4), 39197; Collier Co.: FSM 39198, 39200-2; Dade Co.: FSM 22874, 39204-5; Franklin Co.: LSUMZ 21805-6, 21810; Pasco Co.: LSUMZ 22003. LOUI- SIANA, St. Tammany Par.: LSUMZ 8302, 23770. Thamnophis scalaris. MEXICO, Distrito Federal: AMNH 75934; JaUsco: UTA R-4932, R^949, 5991, 5993; Mexico: AMNH 71315 (2), 94714; Michoacan: AMNH 88724. Thamnophis sirtalis fitchi. CALIFORNIA, Amador Co.: LSUMZ 16486-8, 16489-92; Mendocino Co.: LSUMZ 16493; Modoc Co.: LSUMZ 8215; Plumas Co.: LSUMZ 16477-8, 16481-2; Shasta Co.: LSUMZ 16496-8. Thamnophis sirtalis sirtalis. INDIANA, Allen Co.: LSUMZ 7988. MINNESOTA, Carlton Co.: JFBM 1115, Cass Co.: LSUMZ 7991, 7996; Clearwater Co.: JFBM 2644-5, 2651-2, 2657, 2659; Isanti Co.: LSUMZ 23229, 23232, 23234, 24461-2; Pine Co.: LSUMZ 23230. Tropidoclonion lineatum. NEBRASKA, Jefferson Co.: KU 45252-65, 45267-8; Richardson Co.: KU 52228. NEW MEXICO, San Miguel Co.: LSUMZ 29998-9, 30096-7. TEXAS, Travis Co.: LSUMZ 20078-9. Virginia striatula. NO DATA: USL 5395, 15841. LOUISIANA, Acadia Par.: LSUMZ 12091; Ascen- sion Par.: LSUMZ 12087, 18777; Caddo Par.: LSUMZ 20210; East Baton Rouge Par.: LSUMZ 1598, 1604-5, 2786, 17348, 18712, 23536, 23745; East Feliciana Par.: LSUMZ 2779; Lafayette Par.: USL 11179, 22890; Livingston Par.: LSUMZ 12126; Sabine Par.: LSUMZ 20193; St. Helena Par.: LSUMZ 18360; St. Landry Par.: USL 18277; St. Tammany Par.: LSUMZ 2773. Virginia valeriae. FLORIDA, Alachua Co.: FSM 42545; Leon Co.: FSM 1942, 34858; Liberty Co.: FSM 42531-2, 42534-5; Wakulla Co.: FSM 32991. GEORGIA, Chattahoochee Co.: FSM 42546. LOUISIANA, Bossier Par.: LSUMZ 24656; Caddo Par.: LSUMZ 12094; East Baton Rouge Par.: LSUMZ 12147, 17671; East Feliciana Par.: LSUMZ 15536; Livingston Par.: LSUMZ 20256; St. Helena Par.: TU 5957; St. Tammany Par.: TU 1 1844, 14238, 18395; Webster Par.: LSUMZ 12142; West Feliciana Par.: LSUMZ 17901. MISSISSIP- PI, Hancock Co.: TU 14304, 15056, 17681. Literature Cited ATWOOD, W.H. 1916. The visceral anatomy of the black snake {Zamenis constrictor). Washington Univ. Stud. Ser. Sci. 4(13): 3-38. 1918. Visceral anatomy of the garter snake. Trans. Wisconsin Acad. Sci. Arts and Let- ters. 19: 531-552. BEDDARD, F.E. 1908. A comparison of the neo- tropical species of Corallus, C. cookii, with C. madagascariensis; and on some points in the ana- tomy of Corallus caninus. Proc. Zool. Soc. London. 1908: 135-158. 1909. Some notes upon Boa occiden talis and Boa {Pelophilus) madagascariensis. Proc. Zool. Soc. London. 1909: 918-927. Bergman, R.A.M. 1941. Acrochordus Javanl- cus Hornst. Treubia, Buitenzorg. 18: 207-211. 1950. The anatomy of Natrix vittata (L.). Zool. Meded. 31 : 13-24. 1951. The anatomy of Homalopsis buc- cata. Proc. Kon. Nederlandse Akad. van Weten., Ser. C. 54: 511-524. 1952. L'anatomie du genre Ptyas a Java. Rivista di Biologia Coloniale. 12: 5-42. 1953. The anatomy of Cylindrophis rufus (Laur.). II. Proc. Kon. Nederlandse Akad. van Weten., Ser. C. 56: 657-666. . 1954. Thalassophis anomalus Schmidt. Universidad Nacional de Cordoba, Rep. Argen- tina. 1954: 1-16. 1955a. The anatomy of Xenopeltis uni- color. Zool. Meded. 33: 209-225. . 1955b. Dendrophis pictus. Proc. Kon. Nederlandse Akad. van Weten., Ser. C. 58: 206-218. . 1955c. L'anatomie de Cerberus rhyn- chops. Archives neerlandaises de Zool. 11: 113-126. 1955d. L'anatomie de Enhydrino schis- tosa D. Archives neerlandaises de Zool. 11: 127-142. 'Standard museum acronyms (Duellman, Fritts, and Leviton, 1978) are used throughout this paper. 162 Tulane Studies in Zoology and Botany Vol. 23 . 1955e. L'anatomie de Enhydris enhy- dris. Revista di Biologia Coloniale. 15: 5-28. 1956a. The anatomy of Dryophis pra- sinus. II. Proc. Kon. Nederlandse Akad. van Weten., Ser. C. 59: 272-279. 1956b. The anatomy of Matrix submi- niata. Biol. Jaarb. 23: 306-326. . 1956-58. L'anatomie de £'/ff/7o/rfe5 /w5- cus B. Rivista di Biologia Coloniale. 16: 9-32. 1958a. The anatomy of the Acrochor- dinae. I, II, IV. Proc. Kon. Nederlandse Akad. van Weten., Ser. C. 61: 145-166, 173-184. . 1958b. The anatomy of Matrix piscator. Biol. Jaarb. 26: 77-99. . 1959a. Matrix chrysarga. Biol. Jaarb. 27: 73-86. 27: 87-97. .. 1959b. Matrix trianguligera. Biol. Jaarb. . 1960. The anatomy of some Homalop- sinae. Biol. Jaarb. 28: 119-139. . 1961a. The anatomy of Coluber radiatus and Coluber melanurus. Pacific Sci. 15: 144-154. . 1961b. The anatomy of some Viperidae. (I, II, and III). Acta Morph. Neer. Scan. 4: 195-230. . 1962a. The anatomy of Hydrophis fas- ciatus atriceps. Biol. Jaarb. 30: 389-416. 1962b. Die anatomic der Ela- pinae. Zeitschrift fiir Wissenschaftliche Zoologie. 167(3/4): 291-337. 1963. The anatomy of Ablabes balio- deira, a colubrid snake from Java. J. Ohio Herp. Soc, 4: 1-14. .. 1965. The anatomy of Calamaria multi- punclata (Boie). Bull. Nat. Mus. Singapore. No. 33, Pt. 7: 35-56. BOGERT, CM. 1968. A new genus and species of dwarf boa from southern Mexico. Amer. Mus. Novit. (2354): 1-38. BRAGDON, D.E. 1953. a contribution to the sur- gical anatomy of the water snake. Matrix sipedon sipedon; the location of the visceral endocrine organs with reference to ventral scutellation. Anat. Rec. 117: 145-161. BRONGERSMA. L.D. 1951. Some notes upon the anatomy of Tropidophis and Trachyboa (Ser- pentes). Zool. Meded. 31: 107-124. 1957a. Notes on the trachea, the lungs and the pulmonary artery in snakes. I-II. Proc. Kon. Nederlandse Akad. van Weten., Ser. C. 60: 299-313. 1957b. Notes upon the trachea, the lungs and the pulmonary artery in snakes. III. Proc. Kon. Nederlandse Akad van Weten., Ser. C. 60: 451-457. CAMAZINE, B., W. GARSTKA, and D. Crews. 1981. Techniques for gonadectomizing snakes (Thamnophis). Copeia 1981: 884-886. Collins, R.F., and C. Carpenter. 1970. Organ position-ventral scute relationship in the water moccasin (Agkistrodon piscivorus leucos- toma), with notes on food habits and distribution. Proc. Oklahoma Acad. Sci. 49: 15-18. Con ANT, R. 1961. A new water snake from Mex- ico, with notes on anal plates and apical pits in Matrix and Thamnophis. Amer. Mus. Novit. (2060): 1-22. DUELLMAN, W.E., T. FRITTS, and A.E. LEVITON. 1978. Museum acronyms. Herpetolo- gical Review. 9: 5-9. FRENKEL, G., and E. KOCHVA. 1970. Visceral Ematomy of Vipera palaestinae: an illustrated pre- sentation. Israel J. Zool. 19: 145-163. GARRIGUES, N.W. 1962. Placement of internal organs in snakes in relation to ventral scalation. Trans. Kansas Acad. Sci. 65: 297-300. Hull, C.H., and N.H. NIE, eds. 1979. SPSS update: new procedures and facilities for releases 7 and 8. McGraw-Hill Book Company, New York, 238 p. KOFRON, C.P. 1978. Foods and habitats of aqua- tic snakes (Reptilia, Serpentes) in a Louisiana swamp. J. Herpetol. 12: 543-554. Manna, D.C, and A.K. Sircar. 1978. Mor- phology of the hemipenis and cloacal gland and seasonal changes in the testes of the snake Ahaetulla nasuta. British J. Herpetol. 5: 711-717. Matthews, L.H., and F.H.A. Mar- shall. 1956. Cyclical changes in the reproduc- tive organs of the lower vertebrates. In A.S. Parkes, ed., Marshall's physiology of reproduc- tion, Ch. 3, 1: 156-225 Longmans, Green and Co., London. McDowell, S.B. 1979. a catalogue of the snakes of New Guinea and the Solomons with special reference to those in the Bernice P. Bishop Museum. Part III. Boinae and Acrochordoidea (ReptiHa, Serpentes). J. Herpetol. 13: 1-92. MUSHINSKY, H.R., and J.J. HEBRARD. 1977. Food partitioning by five species of water snakes in Louisiana. Herpetologica. 33: 162-166. NiE, N.H., C.H. Hull, J.G. Jenkins, K. STEINBRENNER, and D.H. BENT. 1975. SPSS: Statistical package for the social sciences, 2nd ed. McGraw-Hill Book Company, New York, 675 p. Prasad, M.R.N. , and P.R.K. Reddy. 1972. Physiology of the sexual segment of the kidney in reptiles. General and Comparative Endocrinology, Supplement, (3): 649-662. No. 2 Visceral Topography of Snakes 163 RASMUSSEN, J.B. 1979. An intergeneric analysis of some Boigine snakes — Bogert's (1940) Groups XIII and XIV (Boiginae, Serpentes). Vidensk. Meddr dansk naturth. Foren. 141: 97-155. ROSSMAN, D. A. 1963a. The colubrid snake genus Thamnophis: a revision of the Sauritus group. Bull. Florida State Mus. 7: 99-178. . 1963b. Relationships and taxonomic sta- tus of the North American natricine snake genera Liodytes, Regina, and Clonophis. Occ. Papers Mus. Zool., Louisiana St. Univ., (29): 1-29. RUNYON, R.P., and A. HABER. 1968. Funda- mentals of behavioral statistics. Addison-Wesley Publ. Co., Reading, Massachusetts, xi + 304 p. RUTHVEN, A.G. 1908. Variations and genetic re- lationships of the garter-snakes. Bull. U.S. Natl. Mus. 61: 1-201. Simpson, G.G., A. Roe, and R.C. Lewon- TIN. 1960. Quantitative zoology. Harcourt, Brace and Company, New York, BurUngame, 440 p. Smith, H., and J.E. HUHEEY. i960. The watersnake genus Regina. Trans. Kansas Acad. Sci. 63: 156-164. Thompson, J.C. 1913a. contributions to the anatomy of the Ophidia. Proc. Zool. Soc. London. 1913: 414-425. . 1913b. The variation exhibited by main- land and island specimens of the Hibakari snake. Matrix vibakari (Boie). Proc. U.S. Natl. Mus. 46: 157-160. . 1914. Further contributions to the ana- tomy of the Ophidia. Proc. Zool. Soc. London. 1914: 379-402. Thorpe, R.S. 1975. Quantitative handling of characters useful in snake systematics with particu- lar reference to intra-specific variation in the Ringed Snake, Natrix natrix (L.). Biol. J. Linn. Soc. 7: 27-43. Underwood, G. 1976. a systematic analysis of bold snakes. In A. d'A. Bellairs and C. Barry Cox, eds.. Morphology and biology of reptiles. Linnean Soc. Symposium Ser. No. 3: 151-175. VALLE, J.R. 1944-45. Sobrevida da parelheira {Philodryas sp.) depois da adrenalectomia. Mem. Inst. Butantan, Sao Paulo. 18:237-240. VARKEY, a. 1979. Comparative cranial myology of North American natricine snakes. Milwaukee Public Mus. Publ. Biol, and Geol. (4): 1-76. Appendix A Comparative data on non-thamnophiine snakes obtained from the literature fell within the outer parameters of the thamno- phiine data sets generated by our study except for the following taxa whose organ positions lie more posteriorly or which have longer organs or interspaces. Posterior end of heart — non-natricine and non-homalopsine Colubridae: Boiga,^ Chamaetortus,^ Coluber ( = Gonyosoma) oxycephalus, ^ Dipsadoboa, ' and male Zamenis rhodorhacis;^ Acrochordidae: Acro- chordus arafurae,^ A. granulatus,^ A. javanicus;^ Boidae: male Bolyeria/ male Corallus,* rnale Eunectes,* mdXt Licha- nura,* male Loxocemus,* male Xenc- peltis;* Viperidae: C<2W5W5 rhombeatus.^ Anterior end of right kidney — non-natri- cine and non-homalopsine Colubridae: Coluber ( = Gonyosoma) oxycephalus,^ female Philothamnus semivariegatus,^ male Psammophis sibilans,^ male Zamenis florulentus,^ Z. rhodorhacis? Posterior end of right kidney — non-natri- cine and non-homalopsine Colubridae: Coluber ( = Gonyosoma) oxycephalus, ^ male Leptodira ( = Crotaphopeltis) hotamboeia,^ female Philothamnus semivariegatus,^ male Psammophis sibi- lans;^ Viperidae: male Causus rhombeatus.^ Anterior end of left kidney — non-natri- cine and non-homalopsine Colubridae: male Coluber { = Gonyosoma) oxyce- phalus,^ female Philothamnus semi- variegatus,^ male Psammophis sibilans,^ male Zamenis florulentus.^ Posterior end of left kidney — non-natri- cine and non-homalopsine Colubridae: male Coluber { = Gonyosoma) oxyce- phalus,^ female Philothamnus semiva- riegatus,^ male Psammophis sibilans.^ Right kidney length — Viperidae: Causus rhombeatus.^ Heart-liver interspace — The following taxa had an overlap — Tropidophiidae: Trachyboa gularis,^ Tropidophis;^ Viperidae: Causus rhombeatus.^ Kidney overlap — all taxa reported in the literature have an overlap, but Causus 164 Tulane Studies in Zoology and Botany Vol. 23 rhombeatus^ (Viperidae) is the only one to have a greater overlap than any of the Thamnophiini. The following taxa have an organ posi- tion lying more anteriorly or have shorter organs than any of the Thamnophiini. Posterior end of left kidney — Tropido- phiidae: female Exiliboa placata.' Liver length — non-natricine and non- homalopsine Colubridae: female Philothamnus semivariegatus,^ Heart-liver interspace — Colubridae, Na- tricinae: male Natrix ( = Amphiesma) vibakari^ from Japan. Kidney asymmetry — In the present study males in 64°7o of the taxa have the left kidney longer than the right but the dif- ference is significant in only 5"7o. How- ever, the literature reveals that in the 'Rasmussen (1979) Thompson (1914) 'McDowell (1979) 'Underwood (1976) 'Thompson (1913b) 'Brongersma (1951) 'Bogert (1968) 'Bergman (1959b) 'Bergman (1955e) '"Bergman (1956-58) "Bergman (1960) males of most taxa the right kidney is longer than the left. The following are the taxa in which the left kidney is longer: Colubridae, Natricinae — Natrix ( = Sinonatrix) trianguligera;* Colu- bridae, Homalopsinae: Enhydris enhy- dris;" other Colubridae: Coluber ( = Gonyosoma) oxycephalus,^ Elapoides fuscus.'" Females in 76% of the thamnophiines have the right kidney longer than the left (28% significantly different) as do the females of all taxa reported in the literature except: Colubridae, Natricinae — Natrix (=Amphiesma) vibakari;^ Colubridae, Homalopsinae: Hypsirhina { = Enhy- dris) plumbea;'' other Colubridae: Elapoides fuscus. ' " December 15, 1982 087