220 

 outing) and the thicket tinamou {Crypturellus cinnamomeus; not known to use gardens or eat crops) 

 were not taken in gardens. 



Indigenous people throughout South America exphcitly recognize that a function of gardens is to 

 attract game for hunting (Dufour, 1990). Crops and the surrounding vegetation are especially attractive 

 to deer, peccary, and rodents. Some indigenous groups, such as the Ka'apor and the Kuikuru of 

 Brazil, respond to crop predation by planting more crops dian necessary for their personal consumption 

 (Balee, 1984; Balee and Gely, 1989; Cameiro, 1983). Other groups, such as the Kay^o (Parker et 

 al., 1983; Posey, 1982, 1983, 1984), disperse their gardens throughout the area. 



Indigenous groups in Mesoamerica also recognize that gardens can attract game. In southern 

 Mexico, for example, Lacandon Maya farmers leave a portion of their com harvest standing in the field 

 (Nations and Nigh, 1980). This enables hunters to harvest wild animals such as the deer, squirrels, 

 pacas, coatis, and peccaries attracted to the com. In Nicaragua, wild animals such as pacas, agoutis, 

 white-tailed deer, collared peccaries, and Baird's tapir {Tapirus bairdii) are major crop pests in Miskito 

 Indian gardens (Nietschmann, 1973). Paca and white-tailed deer are frequently harvested in gardens 

 and plantations by Miskito hunters. 



Wildlife densities in three forest successional stages . Due to small sample sizes, wildlife 

 population density estimates could not be determined for all species and vegetation types. Instead, 

 sighting frequencies were calculated for three forest successional stages: Late Secondary Forest without 

 Gardens, Late Secondary Forest with Gardens, and Early Secondary Forest. Statistical comparisons 

 could be made for four species. Significant differences were detected for plain chachalaca {Ortalis 

 vetula) sighting frequencies between the three stages (x^ ^proximation = 8.4900, d.f. = 2, P = 

 0.0143; Table 4-1). No significant differences in sighting frequencies were detected between forest 

 successional stages for squirrels (Sciurus spp.; P = 0.1596), coatis (P = 0.0877), or kinkajous (Potos 

 flavus; P = 0.0665). When summarized by game versus nongame taxa, significant differences were 

 detected for game bird sighting frequencies between the three forest stages (x^ approximation = 

 7.2947, d.f. = 2, P = 0.0261; Table 4-1). The greatest average sighting frequency was 5.8 sightings 



