Howler Groups Research Articles

The Relationship of Size of Feeding Aggregate to Size of Food Patch: Howler Monkeys (Alouatta palliata) Feeding in Trichilia cipo Fruit Trees on Barro Colorado Island

The influence of one socioecological variable, food patch size, on grouping patterns of animals is examined. Food patch size is measured by the number of feeding spaces, S, a patch holds for a given animal species. The values of S should be determined by structural or spatial criteria if the size of the feeding aggregate (the animals feeding together within a patch) is limited by the availability of space or by interference behavior. Otherwise, S is limited by the abundance and density of food items within the patch, and may vary, depending on the relative profitability of alternative available to the animal, a result deduced from simple optimal foraging theory. If the feeding aggregate sizes favored by selection for efficient foraging differ from those favored by other selective pressures, aggregate sizes will be adjusted to reflect the relative strengths of these counteracting selective pressures. Data from Barro Colorado Island on the sizes of howler monkey feeding aggregates seen in different-sized Trichilia cipo fruit trees are applied to test the hypothesis that: food patch size limits the sizes of feeding aggregates. This hypothesis is supported by demonstrating a positive correlation between howler feeding aggregate size and Trichilia DBH (diameter at breast height), since DBH is shown, in turn, to predict accurately both the spatial sizes and crop sizes (numbers of ripe fruits) of Trichilia patches. Some observations support the explanation that howlers were sensitive to the densities and abundances of ripe fruits within and not to the spatial size of patches, per se. A X2 analysis is used to establish that small feeding aggregates were underrepresented in large patches, indicating that, in this instance, feeding aggregate size is adjusted to and not merely limited by patch size. To account for this unexpected result, a hypothesis is proposed: that howler feeding aggregates are often part of larger foraging units (the group of animals maintaining contact while ranging) which travel between clumps of food patches, and uponi reaching a clump, the monkeys apportion themselves among the according to the sizes of the patches. The result that howlers were seen feeding more frequently in Trichilia that are near to Quararibea asterolepsis trees (the other major food source at this time) than in those located farther from Quararibea supports this explanation. This manner of ranging exemplifies one potential way that groups of animals can maintain cohesion while using too small for the entire foraging unit. THERE IS WIDESPREAD AGREEMENT that the size of resource patches influences patterns of grouping in animals by limiting the number of animals able to use a patch (Altmann 1974, Bradbury and Vehrencamp 1976, Crook 1972, Horn 1968, Jarman 1974, Klein and Klein 1973, Kummer 1971, Ward and Zahavi 1973). Despite its frequent use as a premise by several of these authors in explanations of grouping patterns, the relationship between patch size and group size has been developed only casually. Since there is little evidence yet available for examination, it is premature to consider the patch size-grouping pattern relationship as a sociobiological axiom. It is more appropriate, at this stage, to treat the relationship as an hypothesis and to develop this hypothesis into a testable form. In this paper we define the role of patch size in grouping patterns and test the hypothesized relationship between them with data on howler monkeys. Fieldwork was carried out in old (greater than 140 years) secondary forest on the central plateau (Foster 1973, Knight 1975) of Barro Colorado Island (BCI) during August and September 1976 (fig. 1). Howler monkeys (Alouatta palliata) on BCI normally forage in cohesive groiups of 10-30 animals (Altmann 1959, Bernstein 1964, Carpenter 1934, Chivers 1969, Mittermeier 1]973, Smith 1977), but, in contrast, the howlers we met during the initial weeks of our study were usually in small, scattered subgroups. They appeared to rely heavily on two tree species, Trichilia cipo (A. Juss.) C.D.C. and Quararibea asterolepsis Pittier, as fruit sources. Fig trees, the principal food source of howlers (Smith 1977), were not discovered fruiting in this area, although large howler groups were encountered feeding in fig trees located elsewhere. Since these fig trees (principally Ficus yoponensis and F. insipida) generally produce very large crops (Morrison 1975) and have mature crowns which appeared to be two or more times larger than those of the average Trichilia or Quararibea, we hypothesized that the sizes of food available to howlers in the plateau area at this time were influencing their grouping patterns; i.e., the temporary reliance of howlers on relatively small encouraged small, scattered feeding aggregates. We therefore designed a test of this hypothesis which we will present in detail together with our lPresent address: Department of Anthropology, Peabody Museum, Harvard University, Cambridge, Massachusetts 02138, U.S.A. BIOTROPICA 14(2): 81-9