There is debate as to the precise ecological significance of the relatively high levels of orbital convergence seen in primates. Cartmill attributes it to selection for nocturnal visual predation (Cartmill, 1992), whereas others associate it with manual manipulation or visual detection of small objects, including fruits, insects, and small branches in a nocturnal rainforest environment (Crompton, 1995; Rasmussen, 1990; Sussman, 1991, 1995). Common to all these models is the relationship between orbital convergence and nocturnality. Although relatively high degrees of orbital convergence have not been extrapolated down to basal primates using rigorous character optimization methods, all primates, living and fossil, with only one exception (Megaladapis) have more highly convergent orbits (Heesy, 2003; Ni et al., 2004; Ravosa and Savokova, 2004) than seen in nonprimate mammals, including plesiadapiforms. It, therefore, seems probable that basal primates also had relatively high degrees of orbital convergence, and hence were nocturnal.
This conclusion is congruent with character optimization studies of the evolution of activity pattern and chromacy in primates, and their relatives (Heesy and Ross, 2001, 2004), but runs counter to recent claims that basal primates were diurnal (Li, 2000; Ni et al., 2004; Tan and Li, 1999). We have discussed our objections to Tan and Li's arguments elsewhere (Heesy and Ross, 2001, 2004).
The eye shape data presented here suggest that the eyes of these basal primates were probably not distinguished from those of their ancestors on the basis of shape, as anthropoids are the only mammals with a distinctive eye shape (Figure 7). The reason for the lack of correlation between eye shape and diel activity pattern in nonprimate mammals is not obvious. One possibility is that the nocturnality generally assumed for the mammalian stem lineage resulted in a nocturnal-shaped eye (i.e., with a large cornea relative to axial length), and that nocturnality and its characteristic eye shape persisted in the lineage leading from basal mammals to basal primates. Of course, this does not explain why all nonanthropoid diurnal mammals possess a "nocturnal eye shape," including many visually dependent diurnal mammals. Another possibility is that the measures of eye shape used here are poor indicators of image brightness; in particular, axial diameter of the eye may not accurately reflect focal length in mammals. Future work should evaluate this possibility.
In contrast with these conclusions regarding eye shape, it can be hypothesized that basal primates, if they were nocturnal, were distinguished from their ancestors by larger eye size (Figure 5): extant nocturnal primate eyes have larger axial diameters than similarly sized nonprimate mammals. As noted earlier, when accompanied by photoreceptor pooling, increase in eye size increases image brightness (Land and Nilsson, 2002). Increase in axial length of the eye in basal primates will also increase visual acuity in a nocturnal environment, the same way as it increases acuity for diurnal animals (i.e., by enlarging the image and spreading it over a greater number of photore-ceptors). Of course, this would make the image dimmer if the cornea and pupil did not also increase in size to maintain image brightness, but image brightness is maintained in primates (Figure 7) regardless of differences in eye size (Figure 5).
Increased visual acuity is also expected in the context of the increased orbital convergence that also characterized basal primates. One effect of increased orbital convergence is to improve image quality along the visual axis (by aligning optic and visual axes), so it seems reasonable to expect that the eye would be altered to take advantage of the improved image quality. Increasing axial length and spreading the image over a greater number of photoreceptors is one way to do this.
If the basal primate eye was characterized by features functioning to increase visual acuity in a nocturnal environment, this acuity could have been put to a number of uses, including visual predation on insects (Cartmill, 1992), detection of small fruits, and locomotion in the terminal branches (Crompton, 1995; Sussman, 1995). Several workers have criticized the "nocturnal visual predation" model of primate origins by pointing out that many nocturnal primates use their auditory sense to detect prey, suggesting that this weakens the link between orbital convergence and visual predation (Crompton, 1995; Rasmussen, 1991; Sussman, 1991, 1995;). Clearly, however, basal primates could well have been using both senses to find their prey. R.S. Heffner and H.E. Heffner provide evidence that in extant mammals, increased acuity in sound localization is positively correlated with both increased width of the binocular visual field (1985) and a narrowing of the field of highest visual acuity, estimated by the width of the area centralis (in degrees) (1992). The sound localization threshold is a measure of acuity, such that the lower the threshold, the smaller the difference in the angular position of a sound source that can be detected. Hence, animals with large binocular visual fields and narrow fields of high-acuity vision tend to have the highest auditory acuity. Heffner and Heffner argue that auditory and visual acuity are correlated because hearing is used to guide the eyes toward the target more precisely. Indeed, they go so far as to suggest that "it is the function of sound localizing, i.e., directing the attention of other senses toward the sound-producing object . . . which underlies the variation in mammalian sound-localizing acuity" (R. S. Heffner and H. E. Heffner, 1992: 711). This suggests that if basal primates exhibited adaptations for prey localization, these adaptations probably were found in both the hearing and visual systems.
Heffner and Heffner's data do not include many primates, but support for a link between visual and auditory acuity, and degree of insectivory among primates is found in Tetreault et al.'s (2004) study of retinal ganglion cell densities in Cheirogaleus and Microcebus. Microcebus has a higher retinal ganglion cell density than Cheirogaleus, is more insectivorous, and has larger more mobile pinnae, an important determinant of sound localizing ability (Brown, 1994; Coleman and Ross, 2004; Heffner and Heffner, 1992). Their data also suggest that Microcebus has a narrower field of high-acuity vision than Cheirogaleus. Clearly more research is needed into the sound localizing and visual acuity of strepsirrhines, as well as the interactions between the two systems.
The increase in orbital convergence in anthropoid primates over and above that of most prosimians (Ross, 1995), combined with the decreased pupil diameter associated with diurnality, probably further improved image quality along the visual axis of anthropoids. In this context it would have been worthwhile to both further increase image size by increasing axial length of the eye (producing the unusually long eyes of extant anthropoids), and add a retinal fovea to the visual axis. It is noteworthy in this regard that diurnal anthropoids fall with diurnal birds on the plot of cornea diameter and axial diameter (Figure 9), and most diurnal birds have retinal foveae as well (Ross, 2004).
The tarsier eye exhibits adaptations for increased acuity in a nocturnal environment over and above those predicted for basal primates. The orbits of tarsiers are highly convergent for their size, suggesting that tarsiers are maximizing convergence as much as possible to improve image quality on the retina. The eyes of tarsiers are longer in axial length than any other mammals, plotting with strigiform and caprimulgiform birds. This may reflect increases in overall eye size, to increase either image brightness or the range of light levels over which their eyes are sensitive. It may also be an attempt to increase visual acuity. Tarsiers possess a retinal fovea characterized by a high density of photoreceptors and ganglion cells (Hendrickson et al., 2000), and the exclusion of blood vessels from the center of the fovea, or foveola (Hendrickson et al., 2000; Polyak, 1957; Ross, 2004). Tarsiers lack a tapetum (Hendrickson et al., 2000; Martin, 1973), also probably an adaptation for increased acuity (Cartmill, 1980; Ross, 2004) and possess a postorbital septum to insulate their fine-grained retina against movements in the temporal fossa during mastication (Cartmill, 1980, Heesy et al., this volume; Ross, 1996). In most of these features, tarsiers resemble owls, animals with similar relative axial diameters of the eye (Figure 6), supporting Niemitz' (1985) suggestion of ecological convergence between the two.
Tarsiers and anthropoids share several features of the visual system that are divergent from the basal primate condition. They both possess retinal foveae and lack tapeta, even when nocturnal, and their eyes exhibit large axial diameters. Their orbits are highly convergent for their size and are characterized by a postorbital septum. One explanation is that these shared features are adaptations to diurnality that have been retained by the tarsier lineage when it adopted nocturnal habits (Cartmill, 1980; Ross, 2000). These features of the tarsier eye may also be adaptations for high acuity in a nocturnal environment. Parsimony suggests that the last common ancestor of extant hap-lorhines was nocturnal (Heesy and Ross, 2001, 2004; Ross, 2004), but definitive resolution of this question must await discovery of fossils closer to the ancestral haplorhine node.
Was this article helpful?