Novitatesamerican MUSEUM PUBLISHED by the AMERICAN MUSEUM of NATURAL HISTORY CENTRAL PARK WEST at 79TH STREET NEW YORK, N.Y
Total Page:16
File Type:pdf, Size:1020Kb
NovitatesAMERICAN MUSEUM PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET NEW YORK, N.Y. 10024 U.S.A. NUMBER 2607 NOVEMBER 10, 1976 NORMAN I. PLATNICK AND WILLIS J. GERTSCH The Suborders of Spiders: A Cladistic Analysis (Arachnida, Araneae) k g - si 0,.00<t 0i 000:0::0; ,0;f\:Nv: ::l zA :::}-0%0, :; ;, :f 41$ 4 AMERICAN MUSEUM Novttates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 2607, pp. 1-15, figs. 1-18 November 10, 1976 The Suborders of Spiders: A Cladistic Analysis (Arachnida, Araneae) NORMAN I. PLATNICK1 AND WILLIS J. GERTSCH2 "What, for instance, shall we do if we find fossils that are typical of the Mygalomorph and Arachnomorph forms save for the presence of segmentation? It is well within the bounds of possibility and we shall then have to decide whether a 'grandfather' is to be grouped with his descendants or 'his cousins'." -W. S. Bristowe, 1933, p. 1033 "The synthetic or evolutionary method of classification . agrees with cladistics in the postulate that as complete as possible a reconstruc- tion of phylogeny must precede the construction of a classification . ." -E. Mayr, 1974, p. 95 ABSTRACT The methods of phylogenetic systematics are group relationships (between the Liphistiidae applied to the problem of the subordinal classifi- and, in the first case all of, and in the second cation of spiders. Synapomorphies in external case some of, the mygalomorph spiders) docu- morphology, internal morphology, embryology, mented only by symplesiomorphic characters. and karyology indicate that the Liphistiidae A return to the earlier classification of Pocock, represent the sister group of all other Recent recognizing two suborders (Mesothelae and spiders. The two currently prevailing subordinal Opisthothelae) and two infraorders of Opistho- classifications of spiders (those of Bristowe and thelae (Mygalomorphae and Araneomorphae), Gertsch) are rejected because they imply sister is advocated. INTRODUCTION The present paper represents an attempt to be strictly monophyletic (i.e., must contain all examine the higher classification of spiders from species assumed to be descendants of a hypo- the viewpoint of phylogenetic systematics. The thetical ancestral species; Nelson, 1971, 1973) principles involved are that all named taxa must and that such groups can be recognized only by 'Assistant Curator, Department of Entomology, the American Museum of Natural History. 2Curator Emeritus, Department of Entomology, the American Museum of Natural History. Copyright i) The American Museum of Natural History 1976 ISSN 0003-0082 / Price $1.00 2 AMERICAN MUSEUM NOVITATES NO. 2607 the presence of shared, derived (synapomorphic) it can be shown that amblypygids and spiders characters and not by the presence of shared, are not each other's closest relatives, much of our primitive (symplesiomorphic) characters (Hennig, analysis may be falsified; we would point out, 1966). By this means, the reconstruction of however, that several authors who have recently phylogeny is placed on an objective basis, as it addressed the problem of the interrelationships involves character analysis only, and not (as seen of the various arachnid orders (Petrunkevitch, in one recent classification of araneomorph spi- 1955; Savory, 197 1; Firstman, 1973; Yoshikura, ders) the subjective recognition and evaluation of 1975) have supported this hypothesis. "basic patterns and evolutionary trends" (Leh- We thank Dr. Mohammad U. Shadab for pro- tinen, 1967, p. 204). The cladograms produced viding the illustration ofLiphistius, Mr. Robert J. differ from phylogenetic trees in that they do Koestler for assistance with the scanning electron not attempt to specify unknowable ancestor- microscope, and Dr. Herbert W. Levi of the descendant relationships or unmeasurable differ- Museum of Comparative Zoology, Harvard Uni- ences in genetic similarity; they are, however, versity, for the loan of specimens. Drs. W. S. Bris- predictions of general synapomorphy and are towe, R. R. Forster, H. W. Levi, and R. T. Schuh thus testable and potentially falsifiable (and read and commented on the manuscript. therefore scientific) hypotheses (Wiley, 1975). As Schaeffer, Hecht, and Eldredge (1972) HISTORY have pointed out, character analysis involves two separate processes, the recognition of transforma- Since the time of Pocock (1892) it has been tion series of homologous character states and recognized that the basic problem in the higher the determination of the polarity (primitive to classification of spiders centers around the derived sequence) of those transformation series; family Liphistiidae and its relationships; as he the first process presents no real difficulties for summarized it (pp. 307-308), "In the presence the characters described below. To determine of chitinous plates on the upper surface of the which of two or more homologous states is prim- abdomen and of two sternal plates on the an- itive and which derived, we have used two sources terior extremity of its under surface, in the of evidence, immediate out-group comparison extreme narrowness of the sternum, but above and ontogeny; where such evidence is unavailable all in the position and structure of its spinning- we have refrained from establishing polarities mammillae, Liphistius differs from all known and merely indicate that the character distribu- spiders; and no gradational forms are known tion supports the monophyly of one or the other which would lessen the value of these peculiari- of two taxa even though we do not know which. ties.... The isolated position that Liphistius The use of out-group comparison requires knowl- occupies with respect to other spiders can per- edge of the closest relative of the entire group haps be best expressed by setting it apart by under consideration, which in this case we sug- itself in a group equal in value to a group con- gest is the arachnid order Amblypygi. The taining all the others. For these I propose the hypothesis that spiders and amblypygids are names Mesothelae and Opisthothelae, the terms sister groups is supported by at least two appar- being derived from the position of the spinning- ently autapomorphic characters: they are the organs." A cladogram can be easily derived from only arachnids with subchelate chelicerae and Pocock's strictly subordinated classification with both a pumping pharynx and a pumping (fig. 1). stomach (Kaestner, 1968). We thus hypothesize Petrunkevitch (1923, pp. 150-152) reacted that any character state found in some but not to this classification: "Pocock's idea in separating all spiders and also in amblypygids is plesio- the Liphistiidae in a special sub-order Mesothelae morphic, and its homologs apomorphic; this hy- is undoubtedly sound, although the name chosen pothesis can be falsified in any particular case by by him is misleading and therefore objectionable. incongruence with more numerous synapomorphy . To correct this error and to make the names patterns, but only one such incongruence has more uniform, I shall call this first sub-order been detected in the characters used below. If Liphistiomorphae. ... One might regard all spi- CD °=- /Liphistiidae Mygalomorphae C0~ CD 0 qoQ CD O XArachnomorphae CD o CD~' CD * Liphistiomorphae 1CD 5_ . Mygalomorphae ca. xCD tr Arachnomorphae ;s 0oC, Liphistiomorphae oCso 0 Mygalomorphae CDCD*_ CA) ,,oe _j CD * Araneomorphae CD CD (including Liphistiidae) CD Ctenizoidea 4 AMERICAN MUSEUM NOVITATES NO. 2607 ders with an unsegmented abdomen as belonging Finally, Gertsch (1949) presented a classifica- to a sub-order equivalent in value and opposed tion accepting only two suborders (Mygalo- to the Liphistiomorphae. Without calling such a morphae and Araneomorphae) in which the group a sub-order, Pocock has applied to it the Liphistiidae are associatedwiththemygalomorph name Ophisthothelae. This, however, seems to families Mecicobothriidae, Antrodiaetidae, and me objectionable for two reasons: First, it would Atypidae in the superfamily Atypoidea and the necessitate the creation of divisions under the other mygalomorph families are associated in the sub-order, and, second, it would be reasonable superfamily Ctenizoidea (fig. 4). This scheme has only if the composition of the abdomen in all since won widespread acceptance by authors non-segmented spiders were the same. This, such as Archer (1948; although published prior however, is not only far from being certain, but to Gertsch's classification Archer credits Gertsch may altogether not be true. It is, for example, for the changes relevant here), Kaston (1972), known that the heart in Arachnomorph spiders Lehtinen (1975), and recently even Bristowe has three pairs of ostia, while in Mygalomorph (1976). Gertsch's classification appeared in a spiders it has four pairs." From these comments popular book, and the justification of it was and the classification he presented, it is evident intended for publication as part of a revision of that in 1923 Petrunkevitch accepted the mono- the North American Atypoidea that, for various phyly of Liphistiomorphae, Mygalomorphae, and reasons, has not yet been published. As the Arachnomorphae, but not of Opisthothelae; adoption of this classification has never been thus his scheme must be presented as a trichotomy adequately defended in print, we have thought (fig. 2). Among recent authors, Lehtinen has it best to begin the present analysis with a evidently accepted this scheme, as his dendro- summary of the arguments