I flilflljfifff I lllUfililHW * Uii linl? j Jill ' liliiiiflfiivH UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN * BIOLOGY APR 9^92 BIX HELDIANA Zoology NEW SERIES, NO. 52 I9RA 61988 Phylogeny of the Viperine Snakes (Viperinae): Part II. Cladistic Analysis and Major Lineages James S. Ashe Hymen Marx U/L in?* °Gy UK A Contribution in Celebration of the Distinguished Scholarship of Robert F. Inger on the Occasion of His Sixty-Fifth Birthday November 30, 1988 Publication 1396 PUBLISHED BY FIELD MUSEUM OF NATURAL HISTORY Information for Contributors to Fieldiana General: Fieldiana is primarily a journal for Field Museum staff members and research associates, although manuscripts from nonaffiliated authors may be considered as space permits. The Journal carries a page charge of $65 per printed page or fraction thereof. Contributions from staff, research associates, and invited authors will be considered for publication regardless of ability to pay page charges, but the full charge is mandatory for nonaffiliated authors of unsolicited manuscripts. Payment of at least 50% of page charges qualifies a paper for expedited processing, which reduces the publication time. Manuscripts should be submitted to Dr. James S. Ashe, Scientific Editor, Fieldiana, Field Museum of Natural History, Chicago, Illinois 60605-2496, USA. Three complete copies of the text (including title page and abstract) and of the illustrations should be submitted (one original copy plus two review copies which may be machine copies). No manuscripts will be considered for publication or submitted to reviewers before all materials are complete and in the hands of the Scientific Editor. Text: Manuscripts must be typewritten double-spaced on standard-weight, 8V2- by 1 1-inch paper with wide margins on all four sides. For papers longer than 100 manuscript pages, authors are requested to submit a "Table of Contents," a "List of Illustrations," and a "List of Tables." In most cases, the text should be preceded by an "Abstract" and should conclude with "Acknowledgments" (if any) and "Literature Cited." All measurements should be in the metric system. The format and style of headings should follow those of recent issues of Fieldiana. For more detailed style information, see The Chicago Manual of Style (13th ed.), published by The University of Chicago Press, and also recent issues of Fieldiana. In "Literature Cited," authors are encouraged to give journal and book titles in full. Where abbreviations are desirable (e.g., in citation of synonymies), authors consistently should follow Botanico-Periodicum-Huntianum and TL-2 Taxonomic Literature by F. A. Stafleu & R. S. Cowan (1976 et seq.) (botanical papers) or Serial Sources for the Biosis Data Base (1983) published by the BioSciences Information Service. References should be typed in the following form: Croat, T. B. 1978. Flora of Barro Colorado Island. Stanford University Press, Stanford, Calif., 943 pp. Grubb, P. J., J. R. Lloyd, and T. D. Pennington. 1963. A comparison of montane and lowland rain forest in Ecuador. I. The forest structure, physiognomy, and floristics. Journal of Ecology, 51: 567-601. Langdon, E. J. M. 1979. Yage among the Siona: Cultural patterns in visions, pp. 63-80. In Browman, D. L., and R. A. Schwarz, eds., Spirits, Shamans, and Stars. Mouton Publishers, The Hague, Netherlands. Murra, J. 1946. The historic tribes of Ecuador, pp. 785-821. In Steward, J. H., ed., Handbook of South American Indians. Vol. 2, The Andean Civilizations. Bulletin 143, Bureau of American Ethnology, Smithsonian Institution, Washington, D.C. Stolze, R. G. 1981. Ferns and fern allies of Guatemala. Part II. Polypodiaceae. Fieldiana: Botany, n.s., 6: 1- 522. Illustrations: Illustrations are referred to in the text as "figures" (not as "plates"). Figures must be accompanied by some indication of scale, normally a reference bar. Statements in figure captions alone, such as "x 0.8," are not acceptable. Captions should be typed double-spaced and consecutively. See recent issues of Fieldiana for details of style. Figures as submitted should, whenever practicable, be 8V2 by 1 1 inches (22 x 28 cm) and may not exceed 1 1 Vi by I6V2 inches (30 x 42 cm). Illustrations should be mounted on boards in the arrangement you wish to obtain in the printed work. This original set should be suitable for transmission to the printer as follows: Pen and ink drawings may be originals (preferred) or photostats; shaded drawings should be originals, but within the size limitation; and photostats should be high-quality, glossy, black and white prints. All illustrations should be marked on the reverse with author's name, figure number(s), and "top." Original illustrations will be returned to the author upon publication unless otherwise specified. Authors who wish to publish figures that require costly special paper or color reproduction must make prior arrangements with the Scientific Editor. Page Proofs: Fieldiana employs a two-step correction system. Each author will normally receive a copy of the edited manuscript on which deletions, additions, and changes can be made and queries answered. Only one set of page proofs will be sent. All desired corrections of type must be made on the single set of page proofs. Changes in page proofs (as opposed to corrections) are very expensive. Author-generated changes in page proofs can only be made if the author agrees in advance to pay for them. THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER. FIELDIANA Zoology NEW SERIES, NO. 52 Phylogeny of the Viperine Snakes (Viperinae): Part II. Cladistic Analysis and Major Lineages James S. Ashe Department of Zoology Field Museum of Natural History Chicago, Illinois 60605-2496 Hymen Marx Department of Zoology Field Museum of Natural History Chicago, Illinois 60605-2496 A Contribution in Celebration of the Distinguished Scholarship of Robert F. Inger on the Occasion of His Sixty-Fifth Birthday Accepted for publication March 17, 1987 November 30, 1988 Publication 1396 PUBLISHED BY HELD MUSEUM OF NATURAL HISTORY © 1988 Field Museum of Natural History ISSN 0015-0754 PRINTED IN THE UNITED STATES OF AMERICA Table of Contents List of Illustrations Abstract 1 Introduction 1 Methods 1 monophyly of vlperinae 8 JMajor Lineages of Viperinae 9 ; Causus (Basal Lineage) 9 i Adenorhinos 13 Athens and Related Lineages 13 ! Bitis, Echis, Vipera russelii 14 I Cerastes-Eristicophis Lineage 15 Comparison of Trees Based on Qualita- tive Characters 16 Piscussion 17 Lineage Causus 20 1 Lineage Adenorhinos 20 Lineage Athens 20 I Lineage Bitis 20 Echis 21 I Cerastes Group 21 " Vipera" 21 Acknowledgments 22 Literature Cited 22 Appendix 23 1 . Cladistic branching patterns and charac- ter-state distributions of major lineages of the Viperinae based on all characters . . 10-11 2. Simplified summary cladogram based on all characters showing branching patterns among major monophyletic lineages and demonstrating lack of resolution of rela- tionships among Eristicophis, Pseudoce- rastes, and Cerastes and uncertainty in placement of Echis 12 3. Trees generated from qualitative charac- ters 18-19 List of Tables 1 . Binary coding of character-states used in cladistic analysis of viperine taxa 2-7 2. Number of steps in each character step . . 8 3. Areas of congruence between trees based on all characters and three trees based on qualitative characters 16 Phylogeny of the Viperine Snakes (Viperinae): Part II. Cladistic Analysis and Major Lineages Abstract Cladistic analysis of 33 species of the nine genera of viperine snakes was performed, based on a transformation series in 66 characters of which 1 1 do not vary with the Viperinae. In most instances, species-level taxa segregate into lineages which represent traditional genera. However, taxa in- cluded in the genus Vipera are distributed widely in the cladogram. The traditional lineages Causus, four of five Athens, Bitis, Echis, and Cerastes are clearly indicated to be monophyletic units. Among taxa generally included in the Viperinae, the genus Causus represents the most basally derived unit. The genera Cerastes, Pseudocerastes, and Eristi- cophis form a monophyletic lineage, but the re- lationships among them cannot be resolved at present. The position of the genera Echis and Ad- enorhinos is unclear. Adenorhinos is indicated to be either a very basally derived lineage among viperines or a close relative to Athens. This con- flict cannot be resolved with available data. The monophyletic nature of the genus Vipera is not supported in this study. Introduction This text represents the results of cladistic anal- ysis based on the character analysis of 66 char- acters developed in a previous publication (Marx et al., 1988). Eleven of the characters have only a single state each within the Viperinae. Cladistic analysis was performed using species-level taxa as the operational taxonomic units. However, this paper only discusses the major lineages within the Viperinae. This is possible because most species- level taxa unite into well-supported monophyletic groups which are, in many instances, equivalent to traditionally recognized genera. Relationships among taxa included in these major lineages will be discussed in subsequent papers. Methods The quality of any phylogenetic analysis is pri- marily sensitive to two factors, analysis of the characters and the algorithm used for reconstruc- tion of cladistic branching patterns. Criteria for analysis of the character set used in this study and the transformation series of 66 characters have been discussed elsewhere (Marx et al., 1 988). When these 66 characters are binary-coded for all avail- able states, the total number of binary options becomes 238. States of these characters among study taxa are presented in Table 1 . Characters chosen for analysis are those for which we have sufficient comparative information across all taxa included in this study. While we expect that other character systems will provide phylo- genetically useful information when compared across diverse taxa, the characters included rep- resent a broad array of primarily morphological features which we believe provide an adequate base for initial phylogenetic analysis of viperine taxa. Analysis of a variety of other characters available, of morphology, myology, physiology, behavior, and biochemical characteristics, will provide an opportunity for reevaluation and test- ing the conclusions presented here. Cladistic analysis of taxa of the Viperinae was ASHE & MARX: VIPERINE SNAKES. PART II. OOOOO OOOOOOOO O ©O — — < — OOOOO O OOOOO— ' O O OOOOO OOOOOOOO O — — — « — • O OO — OO O OOOOOOOO OOOOO OOOOOOOO O OO — — — o O — — — O OOOOOOOO OOOOO OOOOOOOO — o— — — — — oooo o OOOOO OOOOO OOOOOOOO OOOOOOOO O o o o o o — — OOOOO — OOOOO o OOOOOOOO OOOOO — OO OOOOO — — OO — — OO — — — — — — OO — OO o OOOOO — — — O — — — O — — — — — — OO — — o o OOOOO — — — — — — — — — — — — — — _^-^Hrt« o OOOOO OOOOOOOO — OO OO O OOOOO o oooooo — o OOOOO OOOOOOOO — OO o— o OOOOO o oooo — o — o OOOOO OOOOOOOO O OO OO O OOOOO — OOOOOOOO OOOOO — — — — — — — — — — — — — — rt_^-,^-,_ _ __^-,_rtrt^__ OOOOO OOOOOOOO O OO OO O OOOOO O OO — oooo- oo o o o — OOOO — OO OOOOOOOO o o o o OO- OOO OOOOO o OOOOO — OOOOOOOO OOOOOOOO OOOOO OOOOOOOO O OO OO O OOOOO — OOOOOOOO OOOOO OOOOO OOOOOOOO — — — — — — o — — — — OO— — — — — — — OOOOO o — o — — OOOOO OOOOOOOO — OOOOO O — — o — ooo — _ _ rt _, _ OOOOOOOO o OO OO O OOOOO o OO OO O OOOOO o OO OO O OOOOO o o — o — OOOOOOOO OOOOOOOO OOOOOOOO 2.3 > OOOOO OOOOO 3 -5 a a :~ 3 3 3 3 3 3 3 3 3 3 a a a a « OOOOO — OO — — — — — — o — — — — — — — — — — — — — 5 <" il ? .5$ c >^ ^. ^ 2 ill's 2 2 2 S 8 3 3 1 2 •-, ,£p a -> -Si s « •si •S1 -S3 e^-s & "3 §■•§ a s ,i( g.,i3 s; S&3 •=: 1 1 <0 'N a 'a 3 3 -?; ■<: o, -5; c ^ -s: 8 -!3 -S3 -S3 -S3 -!3 ■~; c c c c c &s a s k t o ji S 2 R> ~ ~ rf J* ^ .S ^ .T^ a 3 53 <3 <^ 60-C >*J "^ "^ T ^ T QqQqcqQqcqoqQqQq FIELDIANA: ZOOLOGY 3 2.3 ooooo oooooooo _____ OOOOOOOO ooooo oooooooo — ooooo o oooooooo oooooooo oooooooo o o o o o o ooooo — o o o o o o ooooo o o o o o o o ooooo o ooooo oooooooo oooooooo o o o o o o o O O — ooooo oooooooo oooooooo ooooo oooooooo o oo oo o ooooo oooooooo o oo — oooo oooooooo o — o _ o o — — > oooooooo o oo oo o ooooo o oooooooo _____ OOOOOOOO o oo oo o ooooo o oooooooo ooooo ooooo oooooooo o oo oo o o — — — — o oooooooo oooooooo o o — oo o o — — — — o oooooooo ooooo oooooooo o oo oo o ooooo — oooooooo OO — OO OOOOOOOO O — — OO O OO — — — — oooooo — o ooooo ooooo oooooooo oooooooo o o — — o o — — ooooo ooooo oooo — — oo o — oo — — oo sis ill! 3 3 3 3 3 3 3 3 3 3 ra a a a « s 3 •£ 4= 2 ? c <3 s 1-. ho c *s .** y 4l ? 3 •2 -2 -2 -2 2 -2 -2 -2 ASHE & MARX: VIPERINE SNAKES. PART II. ooooo oooooooo o oo oo o ooooo o ooooooo OOOOO OOOOOOOO O OO OO O OOOOO O O O — O O O O OOOOO OOOOOOOO O OO OO O OOOOO O O — ' — OOOO OOOOO OOOOOOOO o o — OO O OOOOO O O — ' — OOOO OOOOO -"OOOOOOO O — ' — < — O O OOOOO O — — — OOOO ooooo — ooooooo o — — — o o — " — • — O O O ——. — OOO- OOOOO — OOOOOOO O — — — O O — — — O o o — — — o — © — OOOOO — OOOO — — O O — — — — — — — — OO o — — — o — — — OOOOO — O — OO — — O — — — — — — — — — OO o — — — — — — — ooo — o — — — — — — — — — — — — — — — — — — o — — — — — — — — — OO — O — — — — — — — — — — — — — — — — — — o — — — — — — — — ooooooo o o — o o o o o — — o — o o OOOOO OOOOOOOO o oo oo o ooooo o OOOOO OOOOOOOO o oo oo — ooooo o OOOOO OOOOOOOO o o— — o — ooooo o ooooo — ooooooo — — — — — — ooooo o ooooo — o — o — — oo — — — — — — oo — oo o OOOOO — — — O — — — o — — — — — — — — — — o o ooooo — — — — — — — — — — — — — — — — — — — o OOOOO OOOOOOOO o oo oo o ooooo o OOOOO OOOOOOOO — — — — — o ooooo o OOOOO — — — — — — — — — — — — — — O — — — — — OOOOO — O — O — — — O — — — — — — — — — — — o OOOO— OOOOOOOO o oo — — ooooo o ooooo — oo — — ooo — — o — — OOOOOOOO OOOOOOOO ooo o — — o o- o o o ooooo O —OOOO o — o o — — ooooo — — — — — — — — — — — — — — — — — — — — — — — — — — — — OOOOO — — — — — OO— o — o oo o OOOOO — — — — — — — — o — o o— o ooooo ooooo OOOOOOOO OOOOOOOO o o o o o — o — — — — o ooooo o o — — — OOOOOOOO OO — oo — o — OOOOO OOOOOOOO — oo oo — ooooo — •i s III si's 3 3 3 a a a ooo 3 I 11 3 3 3 3 o o 3-1 C -3 « § ~ 5 3 & y -2 3 3 £•£> ^ ^ k_ ^ '-~ ^ V. k_ bo s a 'a 5>" y tn -a >] ^ O) &] (/] a 3 .<-> -a1 a *• a 53 .g, .§, .5, .5, .§ .5, kv ^ ^ ^ ^ ^ v) <*, ■& a -a g g o u S •« -a c; k. k. i» ik. Lq ^; t* ^5 ^ ^ sl « .a 5 &o 5 ^>ia p a iS .y § -S3 -2 -2 -2 -52 -2 -2 -S3 ■Si -5 S C a d aqQqoqQqcqQqcqaq FIELDIANA: ZOOLOGY s •- 91 > o o o — o ooooooo— ooo — o ooooooo— ooooo oooo — ooo o o o o o o O — O O — — o o o o o o OOOO — — o o o o o o ooooo o ooo — o — oo ooo — o — oo oooooooo oo — oo oooo — o — oo oooo — o — oo oooo — o — oo oooo — o — oo o — — o — o — oo o — — o — o — oo o — — o — O — OO — — ' — — oo o oo oo o ooooo o oooooooo oo o oo o— o ooooo o oooooooo o o o — o OOO— — ooooo ooooo ooooo ooooo — — _ , — — — i o — o o oooooooo oooooooo oooooooo oooo — ooo — — — 0 — ooo o — • — — o — oo- ooooo oooooooo o oo oo o ooooo o o — oooooo ooooo oooooooo o oo oo o ooooo o o — — ooooo — oooo oooooooo o oo oo o ooooo o o — — ooooo — OOOO OOOOOOOO — OO O— O OOOOO O O — — — — oo — — OOOO OOOOOOOO — OO O— O OO — — o o — — — — — — o — — OOO— OOOOOOOO — OO — — O OO — — o o — — — — — — o — — ooo— oooo — — oo — o— — — o — o — — o o — — — — — — — — — OOO— OOOO — — OO — — — — — O — o — — o o — — — — — — — — — o — — — — — — O — — — o — — — — — — — — — — — — — — — — — — — — — O — — — — — — — — — — o — — — — — — — — — — — — — — — — — — — — — oooo — o — o — oo— o — o oo o o — ooo o ooooo — oo ooooo oo — oooo— o oo oo o o — ooo o ooooo — oo ooooo oo — ooooo o oo oo o ooooo o oooooooo ooooo oooooooo o oo oo o o o © — — oooo- oooo ooooo oooooooo o — — o o o o — ooooo ooooo ooooo o o — o — o o — o — oooooooo o oooooooo o oooooooo — ooo — oo — o — ooo — oo — o — oo oo o oo — oo o o OO OO O OO — — — o o oo — — — OO — — — o — — — o — — o ooo- ooo— — o o — — o o — — oo- ooooo oooooooo o oo oo o ooooo — oooooooo — oo — o oooooooo o oo oo o ooooo o oooooooo — o — — o oooooooo o oo oo o ooooo o oooooooo — o — — o oooooooo o oo oo o ooooo o o — oooooo — — — — o ooooo — — o o oo oo o oo — oo o o — oooooo — — — — — OOOOO — — o o oo oo o oo — oo o o — — oooo — — __ — — — —OOOO — — o o OO OO O OO — o— o o — — ooo — — — — — — — — OOOO — — O O OO — O O OO — o— o o — — ooo — — — — — — — — OOOO — — — — O— — O — OO — — — O — — — — — O — — — — — — — — — O — — — — — — — — — — — — — — O — — — — — — — — OOOOO — OOOOOOO o ooooo oooooooo o oo — — — oooooooo o — o — — — oooooooo o — — — — — oooo — o — — o o o o o o o o o o o o o o — o — o o — — — ooo — — o oooooooo ooo — o o oooooooo ooooo — oooooooo ooooo — — — — oo — © — o — ooo — — — — — — — — — — — — — — o — — — — — — — O — O — — — — — OOO — OOOO— OOOOOOOO o oo o— o ooooo o oooooooo oooo— oooooooo o oo o— o ooooo — oooooooo — — — o — o — o — ooo— o oo o— — o — ooo — — — — — — — — — — — — — — o — o — — oo— — oo — — — o — ooo — — — — — — — — — — — — — — o — ' — o o — — oo — — — — — — — — — — — — — — — — — — — — 1 — — — — — — — — — — o — — — — — — — — — — — — — — — — — ooooo oooooooo o o— oo o ooooo o oooooooo a Si *» 3 3 5 a Si III 3 "5 3 3 3 3 ,° rQ r° It 3 3 -S 3 a c •SIS •Si H k.V.h.V.h.k.V.h. 1 1 -I ^ £ ^ -5-5 S 5 * B 0 ■g C Q §j 5 5 g 5 •2 ! ris supercili ris hindii ris nitschei ris squamig ris hispidus 8 c s: 8 | arietans gabonica heraldica 3 o peringueyi worthingto nasicornis ^ ^ ^ ^ w yi y> l»J y> ."> y> -c -s: -c -c -s: ■« i, UjUj 00 kj rq-^-^-^'T; ^ Q5Q5Q5OQQ5QQQQQ5 ASHE & MARX: VIPERINE SNAKES. PART II. OOOOO — ' O — ' — — ' O — o o oo oo o ooooo o ooooo — o — o o o o o o oo oo o ooooo o oo — oo oooooooo o oo oo o ooooo o oooooooo oooooooo oooooooo o — . — .OO OOOOOOOO O OO OO o ooooo o oooooooo oooooooo oooooooo oooooooo oooooooo 0000000— OOOOOOO— ooooo — 0000000 ooooo oooooooo ooooo oooooooo ooooo oooooooo — o — — — oooooooo — — — — — OOOOOOOO — — — — _ , OOOOOOO — 00 00 00 00 00 00 OO — o o ooooo — O —OOOO — O —OO — O — O —OO — O — O — 1 o — — — — oooooooo oooooooo oooooooo oooooooo OOOOOOO — 0 — — O O O O — OOO 0 0 O — O O O OOOOO 0 0 O — O O O ooooo 0 0 O — O O O ooooo 0 0 O O O O O ooooo 0 0 O O O O O ooooo 0 0 O O O O O 0 0 0 — — 0 0 — 0 — 0 ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo ooooo 00 — — ooo — oooooooo oooooooo 000 000 000 000 000 — 00 — 00 — 00 — 00 — 00 o o o o o o o o o o o o o o o o o — o — 000 000 000 000 000 000 000 000 — — o — . — 1 o ooooo ooooo ooooo ooooo ooooo ooooo ooooo 00—00 o — — o o — 00 — 0000 — 000000—0 oooooooo oooooooo oooooooo oooooooo ooooo — oooooooo ooooo — oooooooo ooooo o 00 — 00000 o ooooo o o — o ooooo o o — o ooooo o o — o 00 — 00 o o — O OO — — o o o — O OO — — o o o — O OO — — o o o — O OO — — o o o — o — O — — o o — — — — o — — — o — — —ooooo — o o o o — — — o o o — — — — © — — — — — o — — — — — o — — — — — o — — — — — O O — O — — OOO— O — O OO o — — OO OO — — OOO— O OO OO OO — OO OOO — OOOO O OO OO OOOOO OOOOOOOO O OO OO OOOOO OOOOOOOO O OO — o OOOOO OOOOOOOO O OO — — OOOOO OOOOOOOO — OO — — OOOOO OOOOOOOO — OO — — OOO — O OOOOO — OO — 00 — — OOO — O OOOOO — OO — OO — — O — OOO— — — OOOOOOO O — OOO— — — OOOOOOO O — OOO— — oooooooo o ooooo — oooooooo o ooooo o oooooooo O OOO — O o 00 — 0 — 000 O OO — O OO — o o — — — o — — o o o — o o o — o o o — OOO — — O — OOO —0—00— ———00— — — — o — — O — OOO ooooo oooooooo oooooooo o o o o o o o o ooooo ooooo oooooooo 0—000000 OOOOO OOOOOOOO O OO OO OOOOO OOOOOOOO O O— OO OOOOO — OOOOOOO O — — OO OO — OO — OOO — — OO — — — o — OO — OO — OOO — — OO — — — — — — O — OO — O — o — — — o — — — — — — O — O— — — — — — — — O — — — — — — — o — — — — — 0 — 0— — — — — — — — — — — — — — — — — — — — — ooooo o o — — ooooo o ooooo o o — o ooooo o o — O OO — — o o — — O OO — — o o — — o — o — — — o — — —ooooo — — o o o — — — OOO — — — o o o — — — — — o — > ■P £ £ £ 4? & — '« a c*. fa SI* 5-8 3 3 3 3 3 33 3-6 a-^i 3 3 3 3 5- Si II 1 as 3 2 8 8 2 5 •& -3 a <3 a q a uuoau d. .c*. .cj. ,§, .5, .§, .§, .d. to Cg •« Sis § i 8 -2 -2 -2 -2 & 5^ 5 5 .a ^. X S i» &, U3 fcj Uj "^ "^ ^ ^ x 3 Q '5 S.i "^ <- Q 3 C 60? *> *>*?•«■ ;£«?»£ 9 s m c ^3 >! Sols © 3 a a &o-s: o cj, s 6 ■§ :« :S :| :3 :S :c :S :| ^ QqQqaqaqQqQqcqQq FIELDIANA: ZOOLOGY $ 8. 3 > _ _ _ _ _ oo — oo — — o — oo _ _ _ _ _ OO— 'OOO — ' O _ _ _ _ _ oooooo — © _ _ _ _ _ OOOOOOOO _ _ _ , _ _ OOOOOOOO _ _ _ _ _ OOOOOOOO — , O — • — — OOOOOOOO o O — — — OOOOOOOO OO — OO OOOOOOOO OOOOO OOOOOOOO OOOOO OOOOOOOO OOOOO OOOO — OOO o o o — _ o o o — « _ o o o o — o o o o — o o o o o o o o o o o o o o o o o o o o o o o o o o o o o o O O © — — — o o o o — o O OOOOO O OOOOO O OOOOO O OOOOO O OOOOO o — — i — — — — o ______ o ______ o — ■ o o — o — O — o O — © — O OOOOOO o — o — o — o — o o o o OOOOO — OOO— "OOO O OO OO O OOOOO O OOOOOOOO OOOOO — ' — OO — OO— O O— OO O OOOOO O OOOOOOOO OOOOO OOOOOOOO O OO OO O OOOOO O — OO — o OOOOO OOOOOOOO O OO OO O OOOOO O — OO — o OOOOO OOOOOOOO O OO OO O OOOOO O — — — — ' o OOOOO OOOOOOOO O OO OO O OOOOO o — — _ _ _ — OOOO OOOOOOOO O OO OO O OOOOO o — — — — — _ O — — — OOOOOOOO O OO OO O OOOOO o — — — _ — _ o — — — OOOOOOOO O OO OO O OOO — o o — — — — — _ o — — — OOOOOOOO O OO O— O OOO — o o _ — — — — _ o — — — OOOOOOOO O OO O— O OOO — — O — — — — — i _ O — — — OOOOOOOO O OO — — O OO — — — o — — • — — ■ — _ O — — — OOOOOO — O — OO — — O OO — — — o — — — — — _ O — — — OOO — O — — O — OO — — o OO — — — o — — — — — _ _ _ _ _ OOO — — — — o — — o — — — oo — — — o — — — — — _____ OOO — — — — o — — o — — — — o — — — o — — — — — _____ ooo — — — — O — — O — — — _____ o _ _ _ — — _____ _oo — — — — O — — O — — — _____ o — — — — — _____ _______0 — — — — — — _____ o — — — — — _____ _______o — — — — — — _____ _ _____ o o o — o — o — o — o — o — o — o — o — o — OOO — — OOOOOOOO O OO o OOOO— OOOOOOOO O OO o OOOOO OOOOOOOO O OO o OOOOO OOOOO — OO o o— o OOOOO OOOOO — OO — o— o OOOOO OOOOO — OO OOOOO OOOOO — OO OOOOO OOOO — — OO OOOOO — OOO — — OO OOOOO — OOO — — — o OOOOO — OOO — — — o OOOOO — — — — — — — o OOOOO _ — — __ — — — OOOOO — — O — — OOO OOOOO — — OOOOOO O — — OO OOOOOOOO _____ OOOOOOOO _____ OOOOOOOO _____ OOOOOOOO _____ OOOOOOOO 3 I § £ :~ s I aj s^ mil 3 3 333 3 3 333 a a a 13 a _ _ o _ _ o o o o o o — o — o — o — o — o — OOOOO O OOOOOOOO OOOOO O OOOOOOOO OOOOO O OOOOOOOO OOOOO O OOOOOO — — OOOOO O OOOOOO — — OOOOO O OO — OOO — — OO — OO o — o — o — o — — OO — OO O — O — — — o — — OO — OO O — o — — — o — — o o o o o o o o o o o o o o o o o o o o o o o o o — o o o OOOOO OOOOO OOOOO O O — o 2 5 5 3 O —OOOO O —OOOO O — o o — o "3 Q g •S- -a £ 3 3 s 3 §•■5 -5 *5 ^ j .52 .53 .53 .53 .53 tl c c c c O — OOOOOO OOOOOOOO OOOOOOOO OOOOOOOO OOOOOOOO OO — OOOOO OO — OOOO — a g ^ Sb'5 G = 3 >« ■?- & ~ cs 5 ■5 C •3-3 3 3 -S i: C a <^ «3 a «>o-s: "> ."» y» y> y> .*> .*> y> UU U} — — ^ T T t C k s o Se s 2 •« ■« ■& s s s s s s :« := QQQQQQQ^QQQQOQQq ASHE & MARX: VIPERINE SNAKES. PART II. Table 2. Number of steps in each character step (in- cluding branches within genera shown in fig. 1 ). Character Digits no. 0 1 2 3 4 5 6 7 8 9 0 1 6 1 3 1 7 4 1 1 10 4 1 1 1 4 2 1 2 3 2 20 1 4 3 3 3 4 3 6 1 4 30 3 2 1 3 1 3 1 2 6 4 40 3 1 3 1 1 1 4 4 5 1 50 4 3 1 5 4 4 2 2 3 1 60 1 2 4 2 4 2 1 4 3 5 70 6 7 8 4 2 1 1 1 1 2 80 4 6 8 3 2 5 7 7 5 1 90 3 3 6 7 9 8 5 6 3 2 100 2 1 10 10 4 2 2 1 1 1 110 5 9 3 4 2 8 9 5 5 5 120 3 2 1 7 9 5 4 4 5 4 130 3 1 6 7 2 4 6 8 9 4 140 3 2 1 1 7 7 3 4 4 1 150 1 5 7 9 3 5 4 5 3 4 160 3 3 1 2 1 3 3 4 1 1 170 1 2 2 2 6 6 5 4 3 2 180 2 1 2 5 5 3 2 1 2 2 190 6 1 2 2 3 5 7 8 6 5 200 5 1 1 2 2 2 3 4 5 6 210 6 6 6 5 4 5 4 3 2 2 220 1 2 2 1 5 2 1 1 2 2 230 3 3 4 4 4 6 8 6 facilitated by use of a phylogenetic tree reconstruc- tion program entitled QTREE, written by Larry E. Watrous and run on a PDP 1 1 -40 computer at Field Museum of Natural History. This program is based on a Wagner tree algorithm described by Farris (1970), and mimics the results of commer- cially available phylogenetic tree reconstruction programs (L. E. Watrous, pers. comra.). Options included in the QTREE program and used in the analysis of viperine taxa include sorting and ad- dition of taxa to the tree, both by random and by increasing or decreasing difference from the start- ing taxon, and optimization of derived character- states of higher taxonomic units after each oper- ational taxonomic unit is added. The QTREE program does not provide options which allow searching for all possible trees within N steps of the shortest tree produced. However, it is designed to print out all trees discovered which contain the shortest number of steps. Therefore, we cannot be certain that we have discovered all of the shortest possible trees. Nonetheless, after considerable run time (at least 1 hour), the com- puter produced only a single tree based on all the characters at 836 steps (designated "all-tree"). Therefore, our analysis is based on this length of tree, with a clear understanding that other trees of this length are likely to exist. Because of the pos- sible effect of quantitative characters with numer- ous states on the branching patterns of the clado- gram, we removed all meristic and mensurate characters (see Marx et al., 1988, for discussion and list of these characters) from the data matrix and repeated the analysis using only qualitative characters (designated "qualitative trees"). Initial analysis of the data was done using all characters available and using the species-level taxa as the operational taxonomic unit (otu). The use of species-level taxa for initial analysis was chosen because of uncertainty about the monophyletic na- ture of many higher-level taxa within the Viperi- nae. Under this analysis the minimum number of steps required for reconstruction of the most par- simonious tree with all data (all-tree) is 836 for an overall consistency of 28.47%. This relatively high number of steps required to generate the clado- gram results from extensive parallelism of many characters. This is especially evident in parallel- isms in structural features among species-level taxa in different monophyletic lineages. The number of steps in each binary-coded option is included in Table 2. In order to recognize major lineages in the vi- perines more clearly, we have simplified the initial tree by combining the species-level taxa of those lineages which are robustly demonstrated to be monophyletic. In most instances this results in terminal taxa which represent traditional genera (i.e., Bitis, Causus). In one instance, taxa included in a traditional genus did not segregate into a monophyletic unit. Taxa traditionally included in the genus Vipera are distributed widely within the cladogram (fig. 1). Except for those taxa of Vipera which show clear affinities to other groups, the genus Vipera will not be considered in detail in the following analysis. The subsequent simplified cladogram is exclusive of Vipera and is illustrated in Figure 2. Monophyly of the Viperinae The taxa included in this cladistic analysis are those traditionally included in the subfamily Vi- perinae. The subfamily Viperinae in this sense plus the subfamilies Crotalinae and Azemiopinae can be demonstrated to form a monophyletic group based on three clearly derived character-states (lat- FIELDIANA: ZOOLOGY eral process of the palatine absent, a single max- illary tooth, and the fangs posterior with anterior groove). In addition, Lombard et al. (1986, fig. 11) show that the form of the ectopterygoid bone is unique to the Viperidae, which is made up of these three recognized subfamilies. If the Colubridae are considered as an out-group, then structure of the ectopterygoid can be interpreted as a derived char- acter-state of the Viperidae as a whole. Within the Viperidae, however, we have not been able to demonstrate that the subfamily Vi- perinae is monophyletic in relation to the Crotal- inae. Clearly, the Crotalinae are monophyletic based on the highly derived and unique nature of the loreal pits and associated structures. Unfor- tunately, we have been unable to find any derived character-states which are shared by all members of the Viperinae but absent in the Crotalinae. This raises the possibility that the subfamily of the Vi- perinae in the traditional sense may be paraphy- letic in relation to the Crotalinae. However, we have not been able to demonstrate the sister group relationship of the pit vipers (Crotalinae) to any monophyletic lineage presently included within the Viperinae. Therefore, the paraphyletic nature of the Viperinae as presently considered is a distinct possibility that has not been conclusively dem- onstrated. After considering these circumstances, we feel that an analysis of the relationships within the Viperinae is not seriously compromised by the possibility of paraphyly. This conclusion primar- ily results from the fact that exclusion of a highly derived subgroup will not affect the branching pat- tern or sister group relationships of the remaining taxa, and, in addition, paraphyly of the Viperinae has not been undeniably shown. Major Lineages of Viperinae We have relied on the cladogram based on all characteristics (qualitative, mensurate, and meris- tic) for initial analysis of relationships and will contrast these with cladistic relationships based only on qualitative characteristics. As noted above, we have considerably simplified the all-tree by combining species into traditional genera if they were indicated to form a monophyletic group (fig. 1 ). The resulting all-tree may be broken down into five clearly defined lineages: Causus lineage; Ad- enorhinos lineage; Athens lineage; Bitis lineage; and Cerastes- Pseudocerastes lineage which in- cludes Cerastes, Pseudocerastes (from Vipera per- sica), and Eristicophis. The position of Echis is uncertain as described further below. Causus (Basal Lineage) The Viperinae as considered here can be divided into two clearly monophyletic groups: Causus lin- eage (lineage AA) and a lineage (lineage A) which leads to all other Viperinae. Causus represents the most basally derived stock within the Viperinae. It is shown to be monophyletic in our analysis based on the presence of 12 derived character- states (fig. 1). Of these 12, seven are derived in parallel elsewhere in the Viperinae. However, the distribution of these character-states is such that they do not suggest that Causus should be consid- ered for possible placement in other parts of the tree. The five remaining character-states are unique to Causus (2-2, 40-2, 48-7, 50-7, 53-3). Character- states 2-2 (single, large, modified rostral shield) and 48-7 (increased number of palatine teeth) are especially strong and suggest that Causus is mono- phyletic. Both character-states represent the end points of a transformation series in a direction of modification not shown by other Viperinae. In addition, character-state 53-3 (narrow lateral flange and spine on ectopterygoid) differs significantly from conditions of the ectopterygoid found among other viperine snakes. The other unique character- states (40-2, 2 compound processes with medial larger; and 50-7, large number of dentary teeth) represent states of these characters that are not found among other viperines but which were in- terpreted by Marx et al. (1988) to be intermediate between states which are presently found among other Viperinae. Two possibilities for the origin of this latter character-state distribution are suggested. If the first instance of the state found in Causus was indeed intermediate between states found among other vipers, then the condition found in Causus must have occurred within the lineage but has been lost in all extant species. The second possibility is that Marx et al. (1988) were incorrect in their char- acter analysis, and that the state found in Causus is not truly intermediate but represents a unique derivation of those characters. We are unable to decide between these possibilities; however, nei- ther of these alternatives significantly weakens the hypothesis that Causus is a monophyletic unit. If the latter possibility is true, it lends considerable additional support. ASHE & MARX: VIPERINE SNAKES. PART II. «0— v o c C 3 °i5 5s C i»8 "to" c a> CO 5& •2 to 2 o E 2 E a3 a* ii™ &* s> s^ 5 5> s> si

e. si c o .eo-C Q.E o o O (0 1 6 65-8t MM M 63-14 56-3 -66-10 52-2 -65-11 45-4r -63-13 42-6f -62-3 41-4r -45-7 39-7r -42-9 38-4r -38-9 20-31 -8-2 11-7 -5-2r Fig. 1. Cladistic branching patterns and character-state distributions of major lineages of the Viperinae based on all characters. Simplified from most parsimonious tree produced, using species-level taxa as OTUs and all characters for analysis (= "all-tree"). See text, page 8, for discussion, r = Reversal; i = independently evolved elsewhere on the cladogram. The monophyly of all other viperines in relation to Causus is strongly supported by 10 derived character-states (lineage A). Seven character-states (1-2, 7-2, 16-3, 41-4, 57-3, 62-2, 63-4) are unique- ly derived for all viperines less Causus. The 10 character-states can be divided into four recog- nizable complexes of modification: (1) shingling of head scales (1-2, 3-2, 7-2, 62-2, 63-4); (2) elon- gation of bony elements in the feeding arch (39-1, 41-4); (3) increase in body scale rows (16-3); and 10 FIELDIANA: ZOOLOGY Figure 1. (4) "live-bearing" reproduction (57-3). Derived character-states in several different functional sys- tems provide additional support that this lineage of viperines is indeed monophyletic. It is especially interesting that the all-tree in- dicates that the Viperinae cannot be shown to be monophyletic if Causus is included; however, it is clearly monophyletic if Causus is excluded. This provides some support for the possibility that Causus should be removed from the Viperinae and ASHE & MARX: VIPERINE SNAKES. PART II. 11 Fig. 2. Simplified summary cladogram (modified from fig. 1) based on all characters showing branching patterns among major monophyletic lineages, and demonstrating lack of resolution of relationships among Eristicophis, Psendocerastes, and Cerastes and uncertainty in placement of Echis. See text, pages 14-15, for discussion. Number in parentheses = number of species used in analysis. 12 FIELDIANA: ZOOLOGY elevated to a hierarchal category of equivalent sta- tus as done by Cope (1 859, p. 334, Causinae). This possibility will be discussed below (see Discus- sion). Adenorhinos Within the monophyletic lineages of viperines (less Causus), the most basally derived lineage in the all-tree is that leading to Adenorhinos (lineage BB). The monotypic genus Adenorhinos is shown by our analysis to be highly autapomorphic among the Viperinae. Of 1 7 derived character-states found in Adenorhinos, eight are unique to this lineage. Among these eight character-states, 4-2 (presence of nasal shield depression), 5-2 (anterior position of nostril in nasal shield), 25-1 (length of the an- terior portion of the skull), and 53-4 (ectopterygoid bone centrally broad without spine) represent transformation of structural features in directions not found among other viperines. In addition, character-states 27-2 (posterior process of lateral arms of premaxilla present) and 1 3-3 (large eye size) represent transformation series unique to the viperines as a whole. Finally, character-states 7-3 (absence of loreal shield) and 45-1 (relative length of longest maxillary tooth 1 3.3%— 16.9%) are unique end points of transformation series found in less highly derived forms among the other vi- perines. This large number of highly derived character- states found uniquely in Adenorhinos reflects the unique evolutionary path taken by this lineage in comparison to other viperines, including the entire family Viperidae. Despite its highly autapomorph- ic condition, its basal position in the cladogram is clearly contradicted in trees based on qualitative characters (see below). In our analysis, the sister group to Adenorhinos is represented by the monophyletic group here des- ignated as lineage B. Within lineage B, fragmen- tation of the species hitherto included in the genus Vipera presents problems in discussion of mono- phyletic groups. In order to discuss cladistic re- lationships within this lineage, we have chosen to discuss the species of Vipera in a relatively cursory fashion at this point while concentrating on char- acteristics which define major lineages. Lineage B is found to be monophyletic based on seven de- rived character-states. Of these, four are derived in parallel elsewhere on the tree and three are uniquely derived in this lineage. Among those which are uniquely derived, character-states 9-2 (more than 3 anterior temporal shields) and 1 1-2 (number of supralabials intermediate beween 5-7 and 8-10) represent the first indication of frag- mentation of the head shields in this lineage. Char- acter-state 37-2 (a moderately bulbous parietal bone) appears to be a good, uniquely derived state for this lineage. The most basally derived member within this lineage is that leading to Vipera ursini (lineage CC). Lineage C is monophyletic in relation to V. ursini based on six derived character-states. None of these seven states are uniquely derived for lineage C; they therefore provide no additional evidence that the lineage bounded by D through H represents a monophyletic group. Cladistic relationships among taxa included in lineage C show that Vipera berus (lineage DD) is derived basally and forms the sister group to all other vipers included in this lineage (lineage D). Of the nine character-states which de- fine lineage D, four are unique to this lineage: 1 1 -3 (number of supralabials 8-10), 38-4 (relative length of quadrate 44.3%-49.7%), 41-5 (relative length of the compound 76.3%-81.7%), and 61-2 (tem- poral scales keeled). Therefore, if the species of Vipera which are derived independently are not included, then the lineage bounded by lineages B, C, and D forms the monophyletic basis for con- sideration of the remainder of the viperines. This lineage (B, C, D) is indicated to be monophyletic based on seven uniquely derived character-states. Of these character-states, 37-2 and 61-2 are not directly related to fragmentation of head shield and therefore, provide unique transformation in independent character systems and considerably strengthen the proposition that the viperines bounded by lineages B, C, and D form a mono- phyletic assemblage. Atheris and Related Lineages In the all-tree, the five species of Atheris con- sidered formed a single monophyletic group united by the shared presence of nine derived character- states (lineage EE). Of these nine, only two states were uniquely derived: 28-3 (vomer ring laminae deeply emarginate or partially absent) and 52-3 (base of posterodorsal projection of premaxilla ex- panded). An additional character-state 54- 1 (round pupil shape) is treated as a reversal here and is shared with Adenorhinos. Those character-states which are independently derived elsewhere on the cladogram reflect a re- petitive trend in the viperines for fragmentation ASHE & MARX: VIPERINE SNAKES. PART II. 13 of the head scales (1-3), strengthening of the feed- ing arch (43-3), and shortening of the skull (66-9, 51-2). Among the species examined in the group and included in our analysis were representatives of the lineages of At her is as recognized by Marx and Rabb (1965), including both terrestrial and ar- boreal forms. These disparate forms, which are here grouped into a monophyletic unit (but see discussion of qualitative trees below), are striking evidence of the adaptive divergence of this stock. The sister group to Atheris includes all of those vipers bounded by lineage E. Lineage E is shown to be monophyletic based on five derived char- acter-states, of which only two are unique to this lineage— 10-2 (interoculabials present and absent [intraspecific variation]) and 1 1 -4 (number of su- pralabials intermediate between 8-10 and 1 1-1 3)— and represent initial transformational stages in fragmentation and increase in cephalic scalation. One of these (10-2) is the first occurrence of this scale series among the viperines. Though the monophyly of this lineage cannot be presumed to be strongly supported, alternative arrangements of character-states do not provide better resolution of the available data. The lineages E, F, G, and H represent a series of derived character-states which primarily reflect the progressive fragmentation of the head shields and similar modifications of various skull char- acteristics. These are most highly derived in the character-states which indicate that lineage H is monophyletic. Most of the character-states at this point (lineage H) are more derived states of fea- tures which are found to be derived on one or more of the lineages E, F, and G. The character-states found on lineages E, F, and G are bounded by the independent derivation of three species of the tra- ditional genus Vipera (respectively, V. aspis, V. latastei, and V. ammodytes), each representing a different progression of these transformation se- ries. A major division in the viperines takes place at the sister lineages I and II. These two lineages are indicated to form a monophyletic group based on those character-states at the level of lineage H. However, the monophyly of lineage H cannot be considered to be irrevocably established, for two reasons. First, six of the 10 character-states which define lineage H (2-3, 20-5, 25-6, 50-2, 63-11, 66-9) are independently derived in other parts of the cladogram. Second, and more important, all character-states except 1-3 (all small dorsal head scales) are intermediate states of transformation series which are found in more highly derived con- ditions elsewhere on the tree. Furthermore, all rep- resent quantitative characteristics. This suggests that the taxa united by this lineage may represent a grade of evolution rather than a monophyletic clade. Alternatively, the possibility that these qualitatively intermediate character-states actual- ly reflect the condition in the hypothesized ances- tor of taxa included on lineages I and II cannot be refuted with the characters available. We, there- fore, are accepting the working hypothesis that lineage H is a monophyletic unit. Bitis, Echis, Vipera russelii The taxa Bitis, Echis, and Vipera russelii are shown to form a monophyletic group in the all- tree based on derived character-states indicated at lineage II. These character-states are primarily as- sociated with (1) shingling of the head shields (character-states 3-3, nasal shield separated from rostral and supralabials; 11-5, number of supra- labials; 59-2, number of scales in nasal shield; and 62-3, fragmentation of scales between rostral and nasal shields); (2) enlargement of the gape (char- acter-states 39-7, length of squamosal; and 41-8, length of compound); and (3) other characters as- sociated with prey capture (character-states 45-8, lengthening of longest maxillary tooth; and 48-4, reduction in number of palatine teeth). In addi- tion, reduction of subcaudals (character-state 24-2) provides another unrelated synapomorphy sup- porting the monophyly of this lineage. The monophyly of the taxa traditionally placed in the genus Bitis is one of the better supported in the Viperinae. Of the 1 2 apomorphic character- states distinguishing this lineage, five are uniquely derived. Two of these five character-states (53-2, broadly flanged anterodorsal shape of the ecto- pterygoid; and 64-2, broadly shaped postorbital bone) are especially striking because of their unique structure among all snakes and are unusual among viperine characters because they are not elements of a gradient. Character-state 51-2 (position of the parietal relative to the postorbital with medial contact) is also unique to Bitis and may form part of a functional complex with character-state 64-2. The other two character-states (59-3, number of scales in nasal shield; and 45-10, relative length of longest maxillary tooth) are extreme values of gradients found among other vipers. The sister lineage to Bitis is made up of Vipera russelii and Echis. The position of these two groups 14 FIELDIANA: ZOOLOGY relative to Bitis is problematic. Character-states indicating monophyly of Echis plus V. russelii are primarily either reversals or are independently de- rived in other parts of the phylogeny. The rela- tionship of Echis to Bitis is particularly disturbing. Previously, Echis had been associated with Ce- rastes based on the presence of serrate scales (Marx & Rabb, 1965). It is important to note that among viperines, serrate scales are found elsewhere in only two (of 3) arboreal species of Atheris. How- ever, Echis shares only two derived character-states with Cerastes ( 1 8-2, serrated keels on lateral scales; and 56-2, lateral body scale rows angular to body axis), as opposed to Cerastes sharing six derived character-states with Eristicophis (see below). Based on these characters, Cerastes must be hypothe- sized to be more closely related to Eristicophis. If true, serrate scales must have been derived inde- pendently at least three times, once each in Atheris, Cerastes, and Echis. An additional derived char- acter-state that Echis shares with Cerastes is 43-3, the presence of a single, fused angular and splenial bone. However, this derived characteristic is also shared with Bitis and other genera (Marx & Rabb, 1965) and is not indicated to be derived at this level in our analysis. In contrast, Echis shares with Bitis a number of characteristics associated with fragmentation of head shields and characteristics associated with prey capture (see above) indicated on lineage II. These characteristics, however, re- flect evolutionary trends which have occurred in a variety of lineages of vipers and do not provide strong support for the monophyletic relationship of Echis and Bitis. It is particularly interesting that V. russelii con- sistently couples with the Bitis lineage in the all- tree. This represents considerable phylogenetic distance from all other members of the genus Vi- pera and serves to underscore the gross polyphy- letic nature of the genus Vipera. Cerastes- Eristicophis Lineage In comparison to lineage II, lineage I unites Eris- ticophis, Pseudocerastes, Cerastes, and two species of Vipera into a monophyletic lineage. This lineage is basally defined by four derived character-states, two of which are unique to this lineage (25-9, rel- ative length of anterior portion of skull; and 26-6, relative width of skull). Vipera xanthina and V. lebetina, respectively, are derived independently from the lineage which leads to the monophyletic grouping Eristicophis, Cerastes, and Pseudoce- rastes. The monophyly of these three genera is very strongly supported by derived character-states in- dicated on lineage K. Lineage K is indicated to be monophyletic based on 1 2 derived character-states, 1 1 of which are uniquely derived in the context of the viperines, including those found on the head (3-3, nasal shield separated from rostral and supralabials; 11-5, number of supralabials 11-13; 14-2, midthroat scalation with numerous gular scales; and 55-2, supranasal sac present), skull (26-7, relative width of skull; 49-4, number of pterygoid teeth), body (20-4, number of ventrals; and 60-2, shape of dor- sal scales squarish), and modifications of scales ( 1 5-2, keeled gular scales; and 2 1 -2, keeled ventral scales). The weight of these 1 1 character-states provides substantial evidence that Eristicophis, Pseudocerastes, and Cerastes are a monophyletic unit; however, at least one of these, 60-2 (squarish dorsal scales), must be assumed to show reversal in Pseudocerastes. The genera Pseudocerastes and Cerastes are hy- pothesized to form a monophyletic group based on nine derived character-states indicated on lin- eage L. Six of these are uniquely derived in the viperines, including presence of supraocular horns in some species (8-2); multiple small scales be- tween nasal and rostral shields (62-3); relative length of the quadrate (38-9); relative length of the dentary (42-9); relative length of longest maxillary tooth (45-7); and relative length of frontal suture (66- 1 0). However, if lineage L is taken to be mono- phyletic, it requires reversion back to the primitive condition in at least five characters in the lineage leading to Pseudocerastes. These include reversion from keeled to smooth gular scales (15-lr), keeled to smooth ventral shields (21-1 r), and squarish to normal-shaped dorsal scales (60- lr). This large number of reversals, and especially in the previ- ously mentioned characters, leads us to believe that the monophyly of Pseudocerastes plus Ce- rastes should be seriously questioned. Based on this assumption, it seems most reasonable at this time to consider Eristicophis, Pseudocerastes, and Cerastes to represent an unresolved trichotomy (fig. 2). This implies that those derived characters which Pseudocerastes shares with Cerastes are not reliable indicators of phylogeny at this position on the tree. Treating these groups as an unresolved trichotomy also helps resolve the relatively large number of reversals required for Eristicophis (also see discussion of relationships of these genera in cladograms based on qualitative characteristics, below). ASHE & MARX: VIPERINE SNAKES. PART II. 15 Table 3. Areas of congruence (x) between trees based on all characters (A) and three trees based on qualitative characters (Q1,Q2,Q3). Taxon No. of species in study Congruence Qi Q2 Q3 Causus, monophyletic 5 Causus, basal lineage Athens, arboreal, monophyletic 3 Athens, arboreal, and A. hindii 4 Athens, monophyletic 5 Bitis, monophyletic 8 Echis, monophyletic 2 Cerastes, monophyletic 2 Cerastes, Eristicophis, Pseudocerastes, monophyletic 4 Cerastes, Eristicophis, monophyletic 3 Vipera russelii & V. lebetina, monophyletic 2 Vipera, monophyletic 8 Adenorhinos, basal to all vipers less Causus 1 Despite the problems of resolving the genera Eristicophis, Cerastes, and Pseudocerastes into a dichotomous branching array, it appears clear that the three genera taken together form a well-sup- ported monophyletic group. Comparison of Trees Based on Qualitative Characters In order to test the above cladistic hypothesis and resulting conclusions, as well as to examine the effect of characters with numerous states on the branching patterns of the cladogram based on all characteristics, a data matrix was generated us- ing only the 38 qualitative characters (matrix equivalent to the first 38 characters in table 1). The minimum length cladogram generated based on these data required 1 36 steps. The total number of binary characters included were 60, resulting in a consistency index of 43.5%. This is significantly better than the consistency of the tree generated using all data. Three trees were generated with this minimum length. Summaries of these trees are shown in Figure 3. While differences among these three qualitative trees and that based on all data are apparent, the striking similarities of branching patterns of most major lineages is encouraging (ta- ble 3). With the exception of differences discussed below, the similarities suggest that the addition of characters with numerous states did not have a gross effect on generated branching patterns. The basal position and monophyletic nature of the genus Causus is consistent among all clado- grams. This consistency provides strong support for the early derivation of the Causus lineage in the phylogenetic history of the Viperinae. How- ever, in contrast to the cladogram based on all characteristics which do not provide any derived states linking Causus to the rest of the Viperinae as a monophyletic group, trees based solely on qualitative characteristics are uniform in provid- ing two uniquely derived character-states, indi- cating that Causus plus all other Viperinae are monophyletic. These derived states are 3 1 -2 (pal- atine-pterygoid articulation with overlapping joint) and 32-1 (medial wing of prefrontal present but small). The trees based on qualitative character- istics also are uniform in treating character state 54-2 (pupil shape round) as a derived feature for the entire Viperinae, with a reversal to the prim- itive condition (pupil elliptical) in the ancestors of Athens, Adenorhinos, and Bitis. While this is most parsimonious in the overall cladogram, we have some reservations about treating presence of el- liptical pupils as a reversal rather than retention of a primitive characteristic in these lineages. If elliptical eyes are treated as primitive, then the derived condition of round pupils must have aris- en independently in several lineages. In contrast to the strongly supported monophyly of all species of Athens in the all-tree, trees based on qualitative characters provide consistent evi- dence for the monophyly of only the three arboreal species of Athens plus A. hindii. In all qualitative 16 FIELDIANA: ZOOLOGY trees A. superciliaris is separated from the re- mainder of Atheris either by Adenorhinos (tree 2 + 3) or by Adenorhinos plus Bitis (tree 1). The association of Adenorhinos with species of Atheris in all of the qualitative cladograms is in marked contrast to the basal position of Adenorhinos when all characteristics are considered (fig. 1). The sug- gestion that Atheris plus Adenorhinos is a mono- phyletic group and that Adenorhinos is actually an aberrant member of the genus Atheris must be given serious consideration (see below). As in the cladogram based on all characteristics, the species of Bitis are consistently indicated to form a monophyletic group in the qualitative trees. The eight species of Bitis group together in a single node in all trees, providing strong evidence that Bitis is a monophyletic group. The species of both Echis and Cerastes, respec- tively, unite into traditional genera, further pro- viding evidence that they are monophyletic. All cladograms also show the lineage including Ce- rastes, Eristicophis, and Pseudocerastes as a mono- phyletic group. However, Cerastes and Eristico- phis are shown to be sister groups in cladograms based on qualitative characters (fig. 3), while Cerastes and Pseudocerastes are hypothesized to be sister groups when all characters are considered (fig. 1). However, we do not have suitable confi- dence in the branching sequence hypothesized among these three taxa in the cladograms based on all characteristics, and we have indicated our uncertainty by showing relationships among them as a trichotomy (fig. 2). The position of Echis in the qualitative trees is instructive. In these three trees Echis always forms a part of a monophyletic group which includes Cerastes, Pseudocerastes, and Eristicophis (plus Bitis in trees 2 and 3). The relationships of Echis to these three genera in contrast to its close as- sociation with Bitis (trees 2, 3; fig. 1) remains equivocal as indicated in Figure 2. Two groups of importance show considerable instability in their position in relation to other taxa in the cladograms. The position of Bitis is espe- cially problematic. This genus is indicated either to be part of a lineage which includes Echis, Eris- ticophis, Cerastes, and Pseudocerastes (trees 2, 3, and all), or it is more closely related to Atheris (tree 1 ). The position of Bitis indicated in Figure 2, based on a cladogram generated by using all characteristics, cannot be given support under these circumstances. The most dramatic shift in posi- tion between the cladogram based on all charac- teristics and the qualitative cladograms is in the position of Adenorhinos. The all-tree shows Ad- enorhinos to be derived very early in the evolution of the Viperinae. In contrast, all three qualitative trees show Adenorhinos to be very closely related to Atheris. Indeed, qualitative trees 2 and 3 indi- cate that Adenorhinos could effectively be consid- ered to be an unusual member of a monophyletic genus Atheris as Loveridge (1930) so indicated in the original description. Based on the fundamental differences in the positions of Adenorhinos in the qualitative trees and the all-tree, the position of Adenorhinos indicated in Figure 2 must be strongly suspect. Except for consistent pairing between Vipera russelii and V. lebetina in qualitative trees, trees produced in our analysis fail to show congruence or to resolve structure in the relationships among the species of Vipera. This provides additional strong evidence that the fundamental polyphyly of the genus Vipera indicated in the all-tree is cor- rect. In conclusion, while a more limited set of qual- itative characters give strikingly different cladistic positions for Bitis and Adenorhinos, they provide considerable support for a number of other lin- eages. The basal position and monophyly of Cau- sus as well as the relationships between Cerastes, Pseudocerastes, and Eristicophis are well support- ed. The monophyly of Bitis is well established, as is that of four of the five species of Atheris studied. The differences suggest areas where further study is badly needed. Especially striking is the associ- ation of Adenorhinos with Atheris as well as the uncertainty in relationships of Bitis. Also, the con- tinued indication of the polyphyletic nature of Vi- pera cannot be ignored. Discussion Based on the characters used in this analysis, the phylogenetic study of viperine taxa is remark- able, both for the complexity of character-state distributions and, in contrast, for the confirmation of the monophyly of a number of traditionally recognized taxa. The character-state complexity is primarily reflected in a great deal of parallelism in all lineages of the trees. This parallelism reflects changes in the viperines presumably associated primarily with modifications of the prey capture mechanisms. It is particularly interesting that the level of homoplasy demonstrated in this analysis of the viperines did not obscure the monophyly ASHE & MARX: VIPERINE SNAKES. PART II. 17 CO tO (0 O "0 4m c to <0 to Q. to o (0 to fl> Q. (0 ^ (0 o c Q) to to 3 4m 4m c X ^% to (0 k. a> o to S Q. CM ■ — * (0 a •Q <0 ce •c ^■N V. CM O 0) to to v. 00 ^s o o 4m (0 k. (0 v. Q> (D Q. (0 Q> Q. (0 Q. (0 v. Q> Q. 4m v. a> 4m o c "0 to 4m (0 Q) Q. ID V. Q> Q. CD v. a> Q. to o "0 (0 (0 (0 v. 0> ^ ^ ^ ^ ^ «* <* ■^ OQ £ ^ ^ UJ a. Uj O Fig. 3. Trees generated from qualitative characters (first 38 characters in table 1). A, Qualitative tree 1; B, qualitative trees 2 and 3. (The tree shown is tree 2. Tree 3 differs from tree 2 in that Vipera ursini and V. berus reverse positions.) of major lineages and relationships among taxa which have been traditionally grouped as genera, except for one notable exception. This exceptional case is that the monophyly of the genus Vipera cannot be supported, and indeed, species tradi- tionally placed in Vipera are widely separated in the cladograms. The extreme level of polyphyly indicated for the genus Vipera is remarkable (see below). In comparison to Vipera, the species usually in- cluded in Causus, Bitis, Echis, Cerastes, and four of the five species of Atheris clearly unite in mono- phyletic groupings in all cladograms generated (ta- ble 3). 18 FIELDIANA: ZOOLOGY CD (0 Q. Q. Q. o V. CD Q. CO CO CD CO 1 ? O c CD T3 fit Q. CO CO CO CD 4m o E CO CD ^ C 4m __ **» CO CO •"■» CO f{j CO CO CD CD Q. CD Q. CD Q. CD Q. CD CM •Q CD CO "■» CD 4m (0 CO CD CO Q. v. CD fit Q. O o 4m CO u^^^^^^^^^-^om CO CD 4m CO CO V. CD o o •D 3 CD CO Q. CM ^ *-» CO co w •-. (0 o £ Uj cq B Figure 3. ASHE & MARX: VIPERINE SNAKES. PART II. 19 Lineage Causus Historically, Causus has been a recognized primitive unit of the Viperinae based on having the nine head plates distinct— similar to the Co- lubridae, the out-group of the entire Viperidae. This is confirmed in this study by the basal po- sition of Causus in the cladograms. However, in spite of the primitive character-states of the head plates, a number of derived character-states of oth- er features unite the species into a cohesive mono- phyletic unit. Our cladistic analyses provide limited support for the opinion that Causus should be of an equiv- alent hierarchal rank to the Viperinae. This study documents that the genus Causus is derived early in the history of the Viperinae and forms the sister group to all other viperines. Additionally, cladistic analysis based on all characters shows that the Viperinae cannot presently be demonstrated to be monophyletic if Causus is included; however, it is clearly so if Causus is excluded. This adds con- siderable support to Groombridge's ( 1 984) finding that all Viperinae except Causus share the derived condition of presence of a ventral course in the facial carotid artery (which he interpreted as a de- rived reversal). In contrast, cladograms based on qualitative characteristics indicate that Causus is a part of the monophyletic lineage of the Viperinae. There- fore, though it seems possible that Causus should be elevated to subfamilial status, at present we feel that this decision should be deferred pending ad- ditional investigation. Certainly, the qualitative differences in uniquely derived character-states separating Causus from other viperines are insuf- ficient to warrant subfamilial status in themselves. orhinos. These cladograms clearly indicate a re- lationship between Adenorhinos and the species of Athens. It is interesting that the species barbouri was originally described in the genus Athens (Loveridge, 1930). However, Marx and Rabb (1965) suggested that this taxon be placed in a distinct genus Adenorhinos. Our analysis does not resolve this conflict. The unusual combination of uniquely derived features of Adenorhinos suggests a unique life-style regardless of its relationships. Marx and Rabb ( 1 965) suggested that this life-style was similar to some colubroid slug-eating snakes. Lineage Atheris The all-tree cladogram is fully consistent with the monophyly of the genus Atheris as defined by Marx and Rabb (1965). They suggested that both terrestrial and arboreal species should be included in the genus Atheris. Previously, terrestrial and arboreal species had been placed in separate gen- era, with the two terrestrial species incorrectly placed in Bitis and Vipera. It has been generally accepted that these snakes of two fundamentally different adaptive types should be placed together in one genus (Ashe, 1968; Rasmussen & Howell, 1982; Broadley, 1983; Groombridge, 1984). Our all-tree cladogram provided considerable evidence for the monophyly of these adaptively divergent snakes. In trees based on qualitative characteris- tics, only A. superciliaris does not unite as part of a monophyletic Atheris. However, it is clearly closely related to other Atheris and is separated from other members of the genus by Adenorhinos (see above). Lineage Adenorhinos Lineage Bitis The position in the cladogram of a recently de- scribed species, A. barbouri (Loveridge, 1930), is especially instructive. It shares a number of de- rived character-states with other viperines thus clearly establishing its position as a true viper. However, the all-tree cladogram indicates that it occupies a very basal position in the diversifica- tion of the vipers. If correct, it is interesting that, though derived from an early stock of vipers, Ad- enorhinos is very highly specialized in the presence of a wide variety of apomorphic features. How- ever, cladograms based on qualitative characters suggest a different set of relationships for Aden- The species included in the genus Bitis form one of the most well-supported monophyletic lineages in our cladograms. The stability of this lineage is especially gratifying when the diversity of adaptive types and habitats found among members of this lineage is considered. Included in Bitis are taxa adapted to a wide variety of habitats (deserts, for- ests, and mountains) which is reflected in the species-level structural diversity. This diversity includes snakes with large, heavy bodies in addi- tion to a small sand-burrowing form and a variety of others. In addition, a number of species of Bitis show considerable parallelism to other viperines. 20 FIELDIANA: ZOOLOGY The fact that such diverse taxa unite together into a strongly supported monophyletic group graphi- cally illustrates the variety of derived character- states that show evolutionary stability in spite of considerable species-level adaptive radiation. However, the relationships of this genus within the Viperinae remain equivocal. Echis As noted above (see analysis of Bitis-Echis- Vi- pera russelii), the position of Echis as a sister group (with V. russelii) to Bitis is equivocal. If the position as indicated in Figure 1 is correct, then serrate body scales must have been derived independently at least three times, once each in Athens, Cerastes, and Echis. Trees based on qualitative character- istics also fail to resolve this conflict (see above). It is interesting that Echis has previously only been considered as a close relative of Cerastes (Marx & Rabb, 1965). Our conclusions do not provide any support for this position, and the character-state distributions available do not provide a clear al- ternative. At present we must conclude that the relationships of Echis cannot be resolved with characters used in this study. The possible rela- tionships of Echis are indicated in Figure 2. Cerastes Group The three genera Eristicophis, Pseudocerastes, and Cerastes are noted above to represent an un- resolved trichotomy (fig. 2). Two of the taxa in- cluded in this lineage, Pseudocerastes and Eristi- cophis, represent monotypic genera, while the third taxon, Cerastes, is comprised of two species. The two species traditionally included in the genus Cerastes clearly unite as a monophyletic group in our analysis, thus providing considerable support for the traditional concept of this genus. Within this monophyletic lineage, two genera, Cerastes and Eristicophis, have always been rec- ognized as genus-level taxa (Boulenger, 1896). In contrast, the third genus has been considered either as a distinct genus (Smith, 1943; Guibe, 1957; Anderson, 1963; Harding & Welch, 1980;Leviton & Aldrich, 1 984) or as a member of the genus Vipera under the name Vipera persica (Marx & Rabb, 1965). It is evident from our analysis that the species persica should occupy a hierarchal rank equivalent to Cerastes and Eristicophis. The array of character-states which Marx and Rabb noted to be similar between Pseudocerastes persica and Vipera lebetina and which caused them to suggest that both should be included in Vipera, exhibit widespread parallelism and are not pertinent to a cladistic analysis. In addition, in consideration of the widespread polyphyly of Vipera in our cladistic analysis and the array of apomorphic features shared between Pseudocerastes and other mem- bers of the Cerastes group, we feel that recognition of the genus Pseudocerastes is fully warranted. "Vipera" One of the most remarkable characteristics of our cladistic analyses is the total polyphyly of eight examined species traditionally placed in the genus Vipera. The dispersion of these taxa throughout the cladogram is so extreme (figs. 1, 3) that no reanalysis of characters will allow grouping of any significant subset of these into a monophyletic lin- eage (except for linkage of V. russelii and V. le- betina in all qualitative trees). Additionally, no derived character-states are available which would suggest that Vipera, in a traditional sense, could be a monophyletic lineage. It seems clear from our cladistic analysis that the genus Vipera has been composed of taxa which are united by the shared presence of primitive character-states. We can find no other explanation for the level of polyphyly indicated by our analysis. An additional surprise is that no two species of Vipera were shown to be sister groups with con- sistency across trees. Instead, species of Vipera were scattered as independent lineages throughout the cladograms in a distribution that primarily reflects the viperine trend of fragmentation of head plates. Our data provide absolutely no sup- port for the traditional concept of the genus Vi- pera; however, neither do the data provide a strong basis for understanding the relationships of the species included within Vipera to the remainder of the Viperinae. We must conclude that the structural charac- teristics examined in this study are inadequate for resolving the phylogenetic relationships of species presently included in the genus Vipera. We there- fore suggest that these species remain assigned to the genus Vipera for convenience until phyloge- netic studies based on other character systems either confirm or refute our analysis, or otherwise provide clearer evidence of the relationships of these taxa. ASHE & MARX: VIPERINE SNAKES. PART II. 21 Acknowledgments We are grateful to the National Science Foun- dation for support by grant GB 5814 to Hymen Marx and George B. Rabb. We appreciate Molly Ozaki's attention to this manuscript. We thank Claire Richardson for constructing the illustra- tions. Literature Cited Anderson, S. C. 1963. Amphibians and reptiles from Iran. Proceedings of the California Academy of Sci- ence, 31: 417^98. Ashe, J. 1968. A new bush viper. Journal of the East Africa Natural History Society and National Museum, 27(1): 53-59. Boulenger, E. G. 1896. Catalogue of the Snakes of the British Museum (Natural History). Vol. 3, Colu- bridae (Opisthoglyphae and Proteroglyphae), Ambly- cephalidae, and Viperidae. Quaritch, London, 727 pp. Broadley, D. G. 1983. FitzSimons' Snakes of South- ern Africa. Delta Books, Johannesburg, South Africa, 376 pp. Cope, E. D. 1859. Catalogue of the venomous serpents in the Museum of the Academy of Natural Sciences of Philadelphia, with notes on the families, genera and species. Proceedings of the Academy of Natural Sci- ence of Philadelphia, 11: 332-348. Farris, J. S. 1970. Methods for computing Wagner trees. Systematic Zoology, 19: 83-92. Groombridge, B. C. 1984. The facial carotid artery in snakes (Reptilia, Serpentes): Variations and possible cladistic significance. Amphibia-Reptilia, 5: 145-155. Guibe, J. 1957. Reptiles d'Iran recoltes par M. Francis Petter. Description d'un Viperide nouveau: Pseudo- cerastes latirostris, n. sp. Bulletin du Museum National d'Histoire Naturelle, ser. 2, 29: 136-142. Harding, K. A., and R. G. Welch. 1980. Venomous Snakes of the World. Pergamon Press, London, 188 pp. Leviton, A. E., and M. L. Aldrich. 1984. Checklist of the amphibians and reptiles of the Arabian Pen- insula, pp. xv-xxv. In Anderson, J., Herpetology of Arabia [facsimile]. Society for the Study of Amphib- ians and Reptiles, Athens, Ohio. Lombard, R. E., H. Marx, and G. B. Rabb. 1986. Morphometries of the ectopterygoid in advanced snakes (Colubroidea): A concordance of shape and phylogeny. Biological Journal of the Linnean Society, 26: 133— 164. Loveridge, A. 1930. Preliminary description of anew tree viper of the genus Athens from Tanganyika Ter- ritory. Proceedings of the New England Zoological Club, 11: 107-108. Marx, H., J. S. Ashe, and L. E. Watrous. 1988. Phy- logeny of the viperine snakes (Viperinae): Part I. Char- acter analysis. Fieldiana: Zoology, n.s., 51: 1-16. Marx, H., and G. B. Rabb. 1965. Relationships and zoogeography of the viperine snakes (family Viperi- dae). Fieldiana: Zoology, 44: 161-206. Rasmussen, J. B., and K. M. Howell. 1982. The cur- rent status of the rare Usambara Mountain forest- viper, Athens certatophoms Werner, 1895, including a probable new record of A. nitschei mngweensis Bo- gert, 1 940, and a discussion of its validity (Reptilia, Serpentes, Viperidae). Amphibia-Reptilia, 3: 269-277. Smith, M. A. 1943. Serpentes. Fauna of British India, Including Ceylon and Burma. Vol. 3, Reptilia and Amphibia. Taylor and Francis, London, 583 pp. 22 FIELDIANA: ZOOLOGY Appendix Viperinae Taxa and Number of Examined Skulls Used in This Study Taxa No. of skulls Taxa No. of skulls Adenorhinos barbouri Athens hindii Athens hispidus Athens nitschei Athens squamiger Athens superciliaris Bitis arietans Bitis atropos Bitis cornuta Bitis gabonica Bitis heraldica Bitis nasicornis Bitis peringueyi Bitis worthingtoni Causus difilippii Causus lichlensteini Causus lineatus Causus resimus Causus rhombeatus Cerastes cerastes Cerastes vipera Echis carinatus Echis coloratus Eristicophis macmahoni Pseudocerastes persica Vipera ammodytes Vipera aspis Vipera bents Vipera latastei Vipera lebetina Vipera russelii Vipera ursini Vipera xanthina Sample Size Summary Subfamilies Genera Species Skulls Azemiopinae Viperinae Crotalinae 1 9 6 1 33 46 3 109 87 ASHE & MARX: VIPERINE SNAKES. PART II. 23 Field Museum of Natural History Roosevelt Road at Lake Shore Drive Chicago, Illinois 60605-2496 Telephone: (312) 922-9410 UNIVERSITY OF ILLINOIS-URBANA 590.5FIN.S. C001 FIELDIANA : ZOOLOGY $ NEW SERIES SCHGO 40-54 1988-89 009378735 01