Foraging ecology of howler monkeys in a cacao (Theobroma cacao) plantation in Comalcalco, Mexico.

Recent evidence indicates that primate populations may persist in neotropical fragmented landscapes by using arboreal agroecosystems, which may provide temporary habitats, increased areas of vegetation, and connectivity, among other benefits. However, limited data are available on how primates are able to sustain themselves in such manmade habitats. We report the results of a 9-month-long investigation of the feeding ecology of a troop of howler monkeys (n = 24) that have lived for the past 25 years in a 12-ha cacao plantation in the lowlands of Tabasco, Mexico. A vegetation census indicated the presence of 630 trees (> or =20 cm diameter at breast height (DBH)) of 32 shade species in the plantation. The howlers used 16 plant species (13 of which were trees) as sources of leaves, fruits, and flowers. Five shade tree species (Ficus cotinifolia, Pithecellobium saman, Gliricidia sepium, F. obtusifolia, and Ficus sp.) accounted for slightly over 80% of the total feeding time and 78% of the total number trees (n = 139) used by the howlers, and were consistently used by the howlers from month to month. The howlers spent an average of 51% of their monthly feeding time exploiting young leaves, 29% exploiting mature fruit, and 20% exploiting flowers and other plant items. Monthly consumption of young leaves varied from 23% to 67%, and monthly consumption of ripe fruit varied from 12% to 64%. Differences in the protein-to-fiber ratio of young vs. mature leaves influenced diet selection by the monkeys. The howlers used 8.3 ha of the plantation area, and on average traveled 388 m per day in each month. The howlers preferred tree species whose contribution to the total tree biomass and density was above average for the shade-tree population in the plantation. Given the right conditions of management and protection, shaded arboreal plantations in fragmented landscapes can sustain segments of howler monkey populations for many decades.