Portal de Periódicos da CAPES

Trumpeting nose‐horn diversity

2004; Wiley; Volume: 20; Issue: 1 Linguagem: Inglês

10.1111/j.1096-0031.2003.00007.x

1096-0031

Lawrence J. Flynn,

Diverse Musicological Studies

Phylogénie et Évolution des Elasmotheriina (Mammalia, Rhinocerotidae) (Phylogeny and Evolution of the Elasmotheriina [Mammalia, Rhinocerotidae]). By Pierre-Olivier Antoine. Publications Scientifiques du Muséum, Paris, 359pp. €69 plus shipping. http://www.mnhn.fr/publication/. In French. The refrain of small-mammal paleontologists, notably Oldrich Fejfar of Prague, is “More and Better Fossils”. It is with more and better fossils that greater paleodiversity within clades emerges, and more complete data sets for known taxa are built. “To More and Better Fossils” could be the subtitle to Pierre-Olivier Antoine's monograph on the extinct group of rhinocerotids known as elasmotheres. One phenomenon in the evolution of small mammals, particularly rodents, is high within-group diversity, for example numerous coexisting genera (and sometimes species) within one family. Such diversity in the fossil record is perhaps not surprising on the global scale, but it is striking when single localities, arguably snapshots of past biotas, yield large numbers of closely related taxa. One level in the Chinji Formation (middle Miocene) of Pakistan produced over 20 muroids (Flynn et al., 1998). Rodents have rhinos beat, but for a while during the Miocene Epoch large mammals (elephantoids and rhinos) also showed considerable diversity; clearly constraints governing community structure are not constant through time. For the late Miocene Dhok Pathan Formation of Pakistan, Tassy (1983) indicated over a half dozen contemporary elephantoids on the Siwalik paleolandscape. The late early Miocene paleolandscape of the Bugti Hills, Baluchistan, Pakistan, was populated by seven contemporary rhinocerotids (Welcomme et al., 2001). One of these rhinos was Bugtirhinus praecursor, an early member of the fantastic Subfamily Elasmotheriinae. The name “rhinoceros” derives from two Greek roots, meaning nose-horn. There is, of course, diversity in the number and location of horns. The namesake of the elasmotheres is the derived end member Elasmotherium, in which the horn is nowhere near the nose. The name Elasmotherium is based on the Greek term “elasmos”, for “plate”, in reference to the plate-like outer walls of enamel of its cheek teeth. Nearly 200 years ago Fischer de Waldheim (1808) named the extinct Elasmotherium—before most of the living rhinos were named. The taxon was based on a lower jaw, whose molars show a straight, high ectolophid (complement to the flat ectoloph of the upper molars). Elasmotherium shows many derived features, including: enlarged molars reduced premolar series incisor loss hypsodonty highly infolded enamel cementum covering molar crowns arc-shaped lower border of the mandible nasal horns absent huge frontal horn present slender limbs How these features are distributed across rhinocerotids and to what extent they define a higher taxon is the subject of the work by Antoine. In systematic studies, elasmotheres are usually submerged within a clade containing living rhinocerotines. Antoine's study establishes them as a major taxon including a dozen genera in parallel to a separate clade for living rhinos plus the extinct hippo-like teleoceratines. His work reveals apparent homoplasy, which alters the perceived manner in which the derived features are distributed, and thus had led to the view that elasmotheres nest with living rhinocerotines. Antoine's work is un grand travail. It is explicit and falsifiable. It acknowledges and tests previous work by extending research to include more taxa than previously assembled and more characters than others have cited. The computer-assisted parsimony analysis is based on 282 characters surveyed for 36 terminal taxa. This is a lot of work. The title of the volume purports to treat Subtribe Elasmotheriina, but surveys quite well, in fact, the Subfamily Elasmotheriinae, as well as key elements (10 taxa) of the Rhinocerotinae. Also included are three rhinocerotids not assigned to a subfamily, and two other outgroups, Hyrachyus eximius and Tapirus terrestris. This approach balances possible effects by distant outgroups on the perceived polarity of characters: in addition to the tapir, other rhinos closer to the elasmotherine-rhinocerotine group are employed. The resultant tree is well resolved without a lot of pesky polytomies. (One polytomy involves similar populations not earlier recognized as separate at the species level.) The rhinocerotines and elasmotheres turn out to be sister taxa in this analysis. The volume is attractive, a large-format contribution, number 188 in the series of scientific publications of the Muséum National d'Histoire Naturelle. It is nicely presented on semi-glossy paper, with few errors that got by the proofing process. The cover is a simplified cladogram superimposed on a cool image of Bugti fossil country in Baluchistan. Figures and tables are numerous. Cladograms are fine; most of the line drawings, crucial for an explicit work on sometimes-controversial morphology, are okay (although simple). A couple of the drawings fall short of their goal of clearly showing what morphology is being examined. It is important to present the anatomy clearly to non-specialists—you can never tell when someone who works on rodents will be watching. This is a minor problem in the totality—after all, there are 312 figures, 15 tables, and 7 appendices (plus an index). If this is not enough for you, then check out the accompanying CD, which is a facsimile of the volume (identical as far as I can tell). The characters analyzed were carefully chosen. Given broad familiarity with available living and fossil rhinos, Antoine was able to define features that were likely to be meaningful in a phylogenetic sense. He controlled for sexual dimorphism, and based especially on a fine series of Hispanotherium (Aegyrcitherium) beonense, he was able to control for ontogenetic effects (allometry) due to individual age. Large museum collections that approximate biological populations are crucial to advancing understanding of dimorphism and growth. A new museum was established in September, 2003, in Hezheng, Gansu Province, China, for fossils derived from local “dragon bone” sands. These Miocene deposits produce scores of skulls representing Paraelasmotherium, Chilotherium, Shansirhinus, and Dicerorhinus—all await study and integration into extended future phylogenetic analysis of Rhinocerotidae. The author lays out his methodology clearly and rigorously. He prefers a character definition that permits binary analysis. Features that for other researchers might form a series with more than two states, are redefined (but not always!) as couplets (for example, feature present or absent; if present, feature of state a or b). On the skull, the postorbital process is considered present or absent; if present, it incorporates the jugal or the squamosal. However, the external auditory pseudomeatus is considered to display three states: open, partially closed, closed. Cementum on the cheek teeth is present/absent; if present, there is a little or a lot. This could be an additive series. The reasons for differing approaches that result in splitting or not splitting characters are sometimes apparent, but not always. It is not apparent why P2 is considered separately (character 94) from P3-4 (102). In a case like this, it would seem that separating P2 from P3-4 increases the weight of the structures in question (unless there are different developmental trajectories controlling these parts of the dental series, thus justifying splitting). Still, the effects on the analysis of these few traits likely are minor. Pierre-Olivier takes pains to avoid the bias forced by presumed morphoclines. In ultimate analyses, he optimizes features in poorly represented taxa. The parsimony analysis under Hennig86 generated three shortest trees differing only in the position of closely related species of Hispanotherium and in the position of the rhinocerotid Protaceratherium minutum (in any case, the latter taxon has autapomorphies excluding it from rhinocerotine or elasmotheriine membership). The strict consensus tree is chosen as the reference tree. It has 36 terminal taxa, including four outgroups. Succeeding the outgroups in the tree, Protaceratherium falls external to a monophyletic Rhinocerotinae and Elasmotheriinae. Among the Rhinocerotinae are sister taxa Rhinocerotina and Teleoceratina. Menoceras arikarense and Diceratherium armatum take up basal positions within Elasmotheriinae. This is interesting, of course, insofar as these are North American genera at the base of an otherwise Old World group. The Elasmotheriina, then, include 19 taxa, beginning with Kenyatherium bishopi and Bugtirhinus praecursor. Kenyatherium was named 30 years ago and recognized as an elasmothere, although known only for a couple of teeth from the undated site of Nakali in Africa (but see Flynn and Sabatier, 1983). It attests to expansion of the group in the early late Miocene. Bugtirhinus is recently named for material from the early Miocene deposits of Dera Bugti, Pakistan. In Antoine's phylogenetic tree (fig. 309), it is plotted at c. 21 Ma, creating significant ghost lineages for other taxa; a position at about 19 Ma, coeval with Menoceras arikarense, seems more consistent. This idea is testable with study of associated fauna—and more field work. Evolution of the elasmotheres produced species of Caementodon, Hispanotherium, Procoelodonta, and derived taxa like Sinotherium and Elasmotherium. Their history documents increasing crown height and distribution of cementum on the cheek teeth, increasing dental complexity, commitment to browsing adaptations in anterior dentition and unreduced limbs with proximal muscle attachments, increasing body mass, diminution of nasal horns, and ultimate development of the prominent central frontal horn. One of the great contributions of this volume is an investigation of the effects on trees due to sample size and breadth of data sets. As in many cladistic studies, the relatively high number of terminal taxa (and characters surveyed) results in heightened observed homoplasy. When homoplasic characters are excluded from the analysis, resolution decreases. Partial analyses restricted by character set show interesting effects on resolution. Cranial characters are highly homoplasic, yielding unstable trees; there is considerable parallelism in skull shape and horn morphology. Mandibular features are almost as bad. Dental features yield better resolution, but with low coherence and retention indices. Of the dentition, permanent teeth give more stable results than the deciduous teeth, but results are not highly congruent with the reference tree! Postcranial data are most robust, and yield a tree corresponding well to the reference tree. The conclusion is that homoplasy effects each area, dental features are less congruent, but more discriminant (better representation), and that the skull and jaw shape and the milk dentition are most subject to strong functional convergence. Analysis of the data set excluding a monophyletic Rhinocerotinae increases tree stability because the group is convergent with elasmotheres in cranial features, for example lateral projection of the orbit. Antoine proceeds to compare his phylogeny with previous studies. It is not strictly comparable with that of either Groves (1983) or Prothero et al. (1986) due to non-overlap of taxa surveyed. The latter phylogeny included only one elasmothere (Elasmotherium), and it plotted among the rhinocerotines due to convergence. Fortelius and Heissig (1989) agree with Antoine's placement of terminal taxa on the reference tree, where accumulated apomorphies establish relationship; lower on the tree there are differences, due to the relatively greater effects of the fewer characters. There are also interesting divergent opinions on character recognition and polarity. The study of Cerdeño (1995) is more fully comparable in its computer-assisted phylogenetic analysis. Still there are significant differences in the operational taxonomic units (OTU's) due to species synonymies or omission of taxa, in character recognition and interpretation, and in coding. This leads to disagreements in the resultant phylogeny Is there art in phylogenetic analysis? This work says yes. Similar approaches with equally valid methodologies can lead to quite different results that hinge on starting points: character recognition, outgroups chosen, recognition of OTUs, notions of polarity. For example, Antoine reassigns polarity to number of roots on the deciduous first premolar, with concomitant difference in apomorphy distribution than in other studies. What is the key to initial decisions about anatomy and included taxa in any study? Individual authority and choice—art, if you will. There is art in perceiving which features carry diagnostic significance. There is also art in transforming the phylogenetic analysis into a classification. Interestingly, Antoine does not mind applying paraphyletic genera for successive species on the cladogram; their relationship is explicit, the nomenclature is negotiable. One species pair is considered to pose candidacy for anagenesis. Such interpretations employ biostratigraphy and biogeography at the same time as cladistics; melding them is an art. So Pierre-Olivier Antoine's exploration of elasmothere evolution is both art and science. Like all science, it stands on the shoulders of those that have gone before, benefiting from previous work, and developing new and broader data sets. “To more and better fossils”: this renewed phylogenetic analysis is made fruitful because the data set could be expanded, thanks to increased intellectual control of fossil collections on a global scale. Future analyses of rhinos and other groups will also bear fruit thanks to ongoing field work, especially in Asia, and development of museum collections. The work is synergistic—we can answer more and different interesting questions than we could with a limited data set. Globalization of science leads to improved science; Phylogénie et évolution des Elasmotheriina sets an example.