122 

 presented to show that shifting cultivation affected the biomass of terrestrial mammals. In addition, 

 new evidence now suggests that the reference points used to compare animal abundances may not be 

 appropriate. Linares (1976) compared the relative abundance of mammals at Cerro Brujo, based on the 

 identification of some 6,000 bones classified into 1,437 identifiable specimens belonging to 14 species. 

 These identifications were converted to "minimum number of individuals" values for each species for a 

 total number of 466 individuals taken by Cerro Brujo hunters over a 20 y period. Based on these 466 

 individuals, Linares (1976) compiled a hunting inventory based on number of individuals taken and 

 their biomass. The relative proportions of the Cerro Brujo mammals, based on biomass, were 

 compared with similar data for Surinam (biomass recalculated by Eisenberg and Thorington [1973] 

 from Walsh and Gannon [1967]) and Barro Colorado Island (Panama; Eisenberg and Thorington 

 [1973]). The comparison of these three sites showed that the percent biomass by species for the major 

 taxa of terrestrial mammals at Surinam and Barro Colorado Island data were quite comparable. The 

 percent biomass of these mammals, however, was quite different from that at Cerro Brujo. Linares 

 (1976) concluded that the differences in faunal assemblages among sites were due to human hunters and 

 shifting cultivation; Cerro Brujo was defined as having a "cultural faunal assemblage," whereas 

 Surinam and Barro Colorado Island, due to a lack of hunters and gardens, were defined as having 

 "natural" faunal assemblages. Based on recent research by Glanz (1990, 1991), it now appears, 

 however, that mammal densities of many species at Barro Colorado Island are not natural, compared 

 with other Neotropical sites. On one hand, 13 species of terrestrial mammals, including the jaguar 

 (Panthera onca) and the white-lipped peccary (Tayassu pecan), are extinct or possibly extinct. In 

 addition, densities for agoutis (Dasyprocta punctata), pacas {Agouti paca), sloths (Bradypus and 

 Choloepus), and squirrels (Sciurus) are substantially higher than at other Neotropical sites. Glanz 

 (1990, 1991) concluded that these relatively high population densities for some mammals were due 

 primarily to the protected status of Barro Colorado Island and its protection from unregulated hunting. 



The relationships between gardens and game abundance were further elucidated by Irvine 

 (1987) in a two-phase study of resource management by the Runa Indians in the lowland tropical forests 



