J Mammal Evol (2013) 20:239–248 DOI 10.1007/s10914-012-9211-4 ORIGINAL PAPER Evolutionary Divergence and Convergence in Shape and Size Within African Antelope Proximal Phalanges Julien Louys & Shaena Montanari & Thomas Plummer & Fritz Hertel & Laura C. Bishop Published online: 27 July 2012 # Springer Science+Business Media, LLC 2012 Abstract Morphological convergence amongst species convergent evolution of the morphology of the proximal inhabiting similar environments but having different evolu- phalanges was observed, but only in the externally ordinated tionary histories is a concept central to evolutionary biology. morphospace. Size shows less correlation with phylogeny Cases of divergent evolution, where there is morphological than does shape. Therefore, we suggest that divergence in divergence between closely related species exploiting dif- size will occur more readily when a species encounters new ferent environments, are less well studied. Here we show environmental conditions than divergence in shape. These divergent evolution in the morphology of the proximal findings are compatible with observations of rapid dwarfing phalanges of several closely related African antelope species on islands (Foster’s rule). inhabiting different environments. This morphological di- vergence was consistently observed in both a neutral mor- Keywords Ecomorphology . Phalanges . Convergent phospace and an externally ordinated morphospace. evolution . Divergent evolution Divergence, but not convergence, was also observed when size and shape were considered independently. Finally, Introduction Electronic supplementary material The online version of this article Morphological convergence amongst species sharing a sim- (doi:10.1007/s10914-012-9211-4) contains supplementary material, which is available to authorized users. ilar environment but different evolutionary histories is a : well-reported phenomenon that has been observed in many J. Louys L. C. Bishop lineages across time. It remains a key concept in evolution- Research Centre in Evolutionary Anthropology and Palaeoecology, School of Natural Sciences and Psychology, ary biology (e.g., Hertel 1994; Schluter 2000; Stayton 2005, Liverpool John Moores University, 2006). However, an understanding of divergent evolution is Liverpool, UK critical for many theories examining the interaction between the environment and the evolution of phenotype. Adaptive J. Louys (*) School of Earth Sciences, The University of Queensland, radiation, for example, relies on the divergence of many St. Lucia, QLD, Australia species from a common ancestor once new environments e-mail: [email protected] are encountered (Schluter 2000). Documentation and, more importantly, empirical tests for morphological divergence S. Montanari Richard Gilder Graduate School, are critical for modern evolutionary theory and provide a American Museum of Natural History, useful comparator for modern studies of genetic divergence New York, NY, USA (Gatesy et al. 1997; Hassanin and Ropiquet 2004). Antelopes (Mammalia: Bovidae) are found throughout T. Plummer Department of Anthropology, Queens College, the Old World but are most common in African environ- CUNY and NYCEP, ments where they show the highest level of continent-wide Flushing, NY, USA diversity. In Africa, they inhabit a range of habitats from open grasslands to dense forest to wetland. They are abun- F. Hertel Department of Biology, California State University, dant in fossil assemblages and have long been used as Northridge, CA, USA paleoenvironmental indicators (Gentry 1970; Vrba 1980). 240 J Mammal Evol (2013) 20:239–248 Subsequently, postcranial elements commonly preserved in remove size from shape (e.g., Kappelman 1988), this ap- the fossil record have been the subject of ecomorphological proach has been subsequently questioned and abandoned analyses (Kappelman 1988, 1991; Plummer and Bishop because of the intimate relationship between size and ecol- 1994; Kappelman et al. 1997; DeGusta and Vrba 2003, ogy (DeGusta and Vrba 2005; Kovarovic and Andrews 2005; Plummer et al. 2008; Bishop et al. 2011). These 2007; Plummer et al. 2008; Bishop et al. 2011). studies have revealed a correlation between habitat use and In order for any ecomorphological analysis to be valid, it morphology of the antelope postcranial skeleton, as success- must be demonstrated that the organisms under study ex- ful locomotor adaptation is intimately associated with or- hibit both convergent and divergent evolution in response to ganismal success (Plummer et al. 2008). Morphology will new environmental conditions, and changes in size as well also be a consequence of the mechanical interaction between as shape are important aspects to consider. Although mor- the postcranial skeleton and the substrate encountered (Scott phological convergence (and divergence) is assumed and 1985; Hamrick 2001; DeGusta and Vrba 2003). The mor- qualitatively validated by ecomorphological studies (starting phology of the phalanges, in particular, should reflect these with Gentry (1970)), this has never been tested quantitative- interactions as these elements are often in direct contact with ly, and particularly within a phylogenetic context. Here, the substrate being moved across (cursorial locomotion) or therefore, we test the following two hypotheses: 1) that through (aquatic, arboreal, scansorial, and fossorial locomo- two closely related species in different environments are tion) (Kent and Miller 1997). Previous morphological more different morphologically than two closely related analyses of mammalian phalanges have focused more on species in the same environment (divergence); and 2) that the distal phalanges (e.g., Macleod and Rose 1993; Hamrick two distantly related species in the same environment are 2001) as these are the most extreme elements of the limb more similar morphologically than two distantly related and the most likely to be in contact with external environ- species in different environments (convergence). This is mental stimuli. There is also a rich body of literature the first study where the influence of phylogeny (as opposed examining primate and human phalanges in the context of to taxonomy) is examined explicitly for analyses seeking to hominin evolution (Deane and Begun 2008; Rolian et al. reconstruct habitats on the basis of bovid ecomorphological 2009; Almécija et al. 2009, 2010, 2012;Griffinand discriminant models. We focus on proximal phalanges be- Richmond 2010;Kivelletal.2011;Nelsonetal.2011; cause the morphology of these elements should reflect both Congdon 2012). a species’ ambulation and the substrate because of their Recently, the use of ecomorphological analyses of ante- articulation with both the metapodials, and hence the limb, lopes for habitat reconstructions has been questioned (Klein and the more distal phalanges, and hence the substrate. et al. 2010) on the basis that size, rather than shape, explains most of the morphological variation in antelope metapodials and astragali. Therefore, these authors argue, any discrimi- Materials and Methods nant analysis attempting to reconstruct habitats on the basis of elements that have not been independently linked to We examined 343 proximal phalanges from 36 antelope habitat or foraging strategy will make classification errors species from: the American Museum of Natural History, among like-sized genera. Body size is probably the most New York, NY (AMNH), the National Museum of Natural fundamental ecological trait, correlating with all aspects of History, Washington, DC (NMNH), and The Natural Histo- an animal’s biology including life history, physiology, be- ry Museum, London, UK (NHML). Adult, wild-shot speci- havior, and evolution (Peters 1983; Schmidt-Nielson 1984; mens were measured in preference to zoo specimens, when Damuth and MacFadden 1990), and understanding how available (zoo specimens, n010). Each species was assigned body size changes in response to different environments to one of four broadly defined habitat preference categories has important implications for understanding macroevolu- used in previous ecomorphic studies (Scott 1985;Kappelman tionary processes (Evans et al. 2012). In bovids, the physical 1988, 1991; Kappelman et al. 1997) and the ethological structure of the environment, as well as diet, has been shown literature (Dorst and Dandelot 1986;Kingdon1997). These to correlate with size (Brashares et al. 2000; Bro-Jørgensen categories partition the environmental continuum from hab- 2008). Closed habitats tend to favor smaller-bodied species itats generally lacking trees and bush to those with a contin- because this trait facilitates movement through dense vege- uous tree canopy. The habitat preference categories are: open tation and allows for effective camouflage. Open habitats (grassland, arid country, ecotones bordering open country), tend to favor larger-bodied species in part because of in- light cover (light bush, tall grass), heavy cover (heavy bush, creased seasonality and the body-size restrictions related to woodland, densely vegetated swamp), and forest (Table 1). a grazing lifestyle (Brashares et al. 2000; Bro-Jørgensen We measured proximal phalanges following the measure- 2008). Although early ecomorphological analyses attempted ment scheme detailed by Bishop et al. (2011)(Fig.1, Table 2). to adjust morphological data for size, effectively trying to We calculated a species mean for each measurement, and their J Mammal Evol (2013) 20:239–248 241 Table 1 Antelope species ex- amined, common names,