In many areas, forest remains only in small patches such as on hilltops, along rivers, or in areas where the topography makes exploitation difficult. Populations of game animals are usually reduced in such areas, and since they are isolated, it is difficult or impossible for the game species to return. Wildlife populations could be maintained (or established) in these isolated patches, and in the absence of their natural predators (which the patches are too small to support), they could be cropped as a sustainable yield protein.
Most forest game animals are frugivorous, and as a result, they experience high juvenile mortality during seasonal scarcity of fruit (Smythe, 1986). Pre-Hispanic forest natives may have left waste crops in the fields, inadvertently or intentionally. Game animals that came to feed were exploited as a supplementary source of protein. The natives thus practiced provisioning of the population and may, in so doing, have reduced seasonal juvenile mortality of the game animals. A principal aim in this project is to develop methods to enhance juvenile survivorship in the field by provisioning in times of fruit scarcity.
In any wild animal management scheme, it is necessary to census the populations as accurately as possible. Since no single method of population censusing appears to be adequate in tropical forests, two different methods are used where possible. Populations of potential meat animals are being estimated with strip-census techniques in a 62-hectare peninsula adjacent to Barro Colorado Island, which is isolated from the contiguous forests by a 1-kilometer-long, five-strand electric fence. The animals being censused in this way are the agoutis (Dasyprocta punctata), pacas, spiny rats (Proechimys semispinosus), collared peccaries (Tayassu tajacu), and white-tailed deer (Odocoileus virginianus). The three species of rodents are also being censused in a trap-mark-release program. This program will be used to make more accurate estimates of the rodent populations but is confined to the period when food is scarce in the forest (between September and March) and animals are willing to enter traps.
The study area is divided into quadrants, one of which is artificially provisioned during the season of scarcity. The effect of the provisioning on the survival of juvenile animals is being determined by the censuses.
Progress in this project has been slower than in those discussed earlier, chiefly due to the unpredictable elements inherent in working with wild animals and the necessary restrictions on manipulation of predators or competing species imposed by working within a protected area. Furthermore, the electric fence is less effective than expected: predators have moved in and taken advantage of trapped animals, and unwanted animals have settled in the provisioned areas and used provisions intended for the targeted species. Nonetheless, much valuable knowledge is being gained.
STRI personnel undertook these projects several years ago with the full understanding that instant solutions to the problems that they were trying to solve were unlikely. But significant progress has been made in these high-risk ventures. Further progress, no matter how slow it is in coming, and the dedication necessary to achieve that progress are essential if conservation goals are to be achieved at the same time as the livelihood of the rural poor is improved.