RELATING PATTERNS OF LAND-USE CHANGE TO FAUNAL BIODIVERSITY IN THE CENTRAL AMAZON

Large-scale deforestation in threatening the diversity of tropical for ests. Given the paucity of data on life-history characteristics of tro pical species, methods are needed to identify those species susceptibl e to extirpation following forest fragmentation. The approach develope d in this paper provides a method to examine potential effects of fore st fragmentation on biodiversity considering both land-use changes and traits of susceptible species. Based on certain behavioral characteri stics, the effects on forest fragmentation in the Brazilian Amazon wer e projected for nine groups on animals. The taxonomically diverse spec ies were characterized by gap-crossing ability and area requirements. The probability of local extinction due to destruction of habitat over a 40-year period was estimated for these animal groups under three sc enarios of land-use practice, projected with a computer model. These s cenarios include the typical land use of central Rondonia, Brazil, and two extremes of land-use practice that bracket the range of possible agricultural land-use changes. Animals with gap-crossing ability propo rtional to area requirements respond similarly to fragmentation, regar dless of their taxonomic affiliation. The available habitat for those species in proportional to the amount of remaining forest under all th ree agricultural management scenarios. In contrast, species that have large area requirements but that will cross only small gaps are more a dversely affects by forest fragmentation. The available habitat for su ch species is reduced at a rate disproportionately greater than the ra te of forest clearing. For a tropical frog, the effects of forest frag mentation were evaluated considering not only gap-crossing ability and area requirement but also specialized habitat requirements and edge e ffects. For the worst-case scenario of land management, gap-crossing a bility and area requirements of the tropical frog imply that only 60% of the forest remaining after seven years of management is suitable ha bitat. Considering breeding habitat requirements and possible edge eff ects further reduces the suitable habitat to 39% of the remaining fore st. These reductions in the proportion of suitable forest area are lik ely to change with rate and spatial pattern of forest loss.