Conclusions

Our data, analyses, and interpretations support the thesis that the ancestral primates were small animals that moved and foraged "in the dense tangle of small twigs and vines, which characterizes the canopy and forest margins" (Cartmill, 1970: 328)—what Charles-Dominique and Martin (1970) independently identified as "the fine branch and creeper niche." This idea has been accepted by some authors (Allman, 2000; Hamrick, 1998; Larson, 1998; Lemelin, 1999; Rasmussen, 1990; Rollinson and Martin, 1981; Schmitt, 1999; Schmitt and Lemelin, 2002), but it continues to provoke controversy and misunderstandings. Some of those misunderstandings deserve to be laid to rest. For example, the proposition that ancestral primates were fine-branch foragers implies nothing about their diets. That proposition is equally compatible with the thesis that the primitive primates were predominantly herbivorous (Sussman, 1991; Szalay, 1972) or with the "visual-predation theory" proposed by Cartmill (1970, 1972, 1974a,b, 1992). The "fine-branch theory" and the "visual-predation theory" are not competing alternatives (contra Fleagle, 1999: 346). Though the second was originally proposed in conjunction with the first (Cartmill, 1970), the two are logically independent. The first is a theory about primate locomotor adaptations, whereas the second offers to explain the primate synapomorphies of the visual system. It is possible that the ancestral (eu)primates originally adapted to the fine-branch milieu to exploit insect resources. But it is also possible that they developed their distinctive locomotor synapomorphies as adaptations for feeding on plant tissues, and later developed specializations of the visual system to allow them to more effectively stalk the insects they encountered while foraging among slender terminal branches (Cartmill, 1975; Rasmussen, 1990).

A similar mosaic sequence of evolutionary changes may account for some of the apparent conflict between our conclusions and those of Szalay and his collaborators. They argue that "... rapid, successive, leaping, and landing with a habitual grasp (i.e., graspleaping)" (Szalay and Dagosto, 1988: 27), represents the ancestral primate locomotor habit, and that the morphotypic peculiarities of the primate locomotor and neural apparatus originated as adaptations to this sort of fast, jumpy arboreal locomotion (Szalay and Dagosto, 1988; Szalay and Delson, 1979; Szalay and Sargis, 2001; Szalay et al., 1987). Szalay and Sargis (2001: 299) accordingly dismiss the relevance of marsupial analogies in understanding the adaptive significance of primate traits. "Didelphid arborealists," they aver, "are drastically unlike a variety of explosive arboreal graspleapers encountered among, and probably in the very ancestry of, the Euprimates."

These conclusions are hard to reconcile with the manifest and unique resemblances that we and others have documented between primitive primates and small, fine-branch-haunting marsupials such as Caluromys (Lemelin and Schmitt, this volume). If marsupial arboreality is drastically unlike that of the ancestral primates, and if the ancestral primate locomotor repertoire consisted largely of jumping and other fast, irregular, asymmetric forms of locomotion, then it is difficult to understand why small arboreal marsupials and primates should share a complex of behaviors— diagonal-sequence walking gaits, increased hindfoot duty-factors, and peak vertical forces on the hindlimb that are higher than those on the forelimb— that have not been found as a complex in any other mammals and appear to be functionally related to cautious or deliberate arboreal locomotor habits. If ancestral primates were habitual "explosive grasp-leapers," then why should primates have a distinctive symmetrical slow gait?

It might be argued that increased hindlimb loading in primates appeared originally in connection with an ancestral habit of vertical clinging and leaping, and that DS gaits evolved in various descendants that reverted to quadrupedalism. But the manifest morphological and behavioral convergences between primates and less saltatory arboreal marsupials (Lemelin and Schmitt, this volume; Lemelin et al., 2003; Schmitt and Lemelin, 2002) seem to argue for the opposite conclusion—that ancestral primates evolved these marsupial convergences as adaptations to foraging habits resembling those of Caluromys, and that specializations for leaping evolved secondarily in later phases of primate evolution.

It is of course possible that the Caluromys-like phase in primate phylogeny predated the last common ancestor of the Euprimates. Bloch and Boyer (2002; Chapter 16, this volume) have suggested that grasping feet are a synapomor-phy of a clade that includes Carpolestes and Euprimates, but excludes Plesiadapis and other "archaic primates" that lacked grasping feet. If this were so, then DS gaits might have been acquired along with grasping feet during a pre-euprimate phase of primate evolution. This account could in principle reconcile our ideas about DS gaits with the reconstruction of the ancestral primates as "graspleapers" (Szalay and Dagosto, 1988; Szalay and Sargis, 2001; Szalay et al., 1987). Unfortunately, the phylogeny of Bloch and Boyer implies that the highly derived dental features that carpolestids share uniquely with other so-called "archaic primates" are convergences (Kirk et al., 2003). We think that it is more likely that grasping extremities arose independently in car-polestids and euprimates. This conclusion is still compatible with the proposition that leaping became an important locomotor mode for many lineages in later phases of primate evolution, or even in the last common ancestor of the living primates. However, the marsupial parallels for the peculiar features of primate gaits favor the conclusion that nonsaltatory, Caluromys-like locomotor adaptations were established earlier in the euprimate ancestry.

In some respects, this conclusion is reminiscent of the ideas of Boker (1926, 1932), who thought that the ancestral mammals had been cautious arboreal "clamp-climbers" (Klammerkletterer) with lemur-like grasping hands and feet, adapted to a slow "branch walking" locomotor habit (Schreiten aufdenÂsten), from which their descendants had departed in various ways to increase speed. This is almost certainly not true of mammals as a whole, but something similar may be true of primates. Although the last common ancestors of the living primates were surely not restricted to exaggeratedly slow-motion movements like those of a loris and could no doubt leap or bound when they needed to, they must have been equally well adapted to more deliberate, cautious, or stealthy locomotion when circumstances called for it—for example, in searching for food items in the terminal branches of tropical forests.

0 0

Post a comment