66 

 as Other and the similarity between Early Secondary Forest and Plots & Gardens, it was decided to 

 combine these categories. For the species analyses, the expected number of kills per vegetation type 

 was determined by multiplying the total number of kills by the proportion of each vegetation type. 



Minimum Harvest and Catchment Areas 



The minimum harvest area for each taxa was defined as the region encompassing the kill site 

 locations for that taxa. The minimum catchment area for each hunter was defined as the region 

 encompassing the kill site locations of all game taken by that hunter. The minimum catchment area 

 was determined only for the seven main hunters. Since hunters did not always report the entire area 

 searched, the minimum catchment area reported here is conservative and includes only the region 

 encompassing the actual kill site locations. The harvest and catchment areas were determined using the 

 minimum convex polygon method (Hayne, 1949; Mohr, 1947) and measured using a compensating 

 polar planimeter. 



Use of Dogs 



Dogs frequently accompanied hunters to the forest and gardens, and often located or killed 

 game. Usually this game was collected and brought home to be eaten by the hunters. The presence of 

 dogs on hunting outing, the game taken, and the total number of dogs with the hunting party were 

 recorded. 



Statistical Methods 



Parametric and nonparametric tests were used to analyze hunting data. Sex of game taken was 

 tested against an expected 1:1 (male:female) ratio using the x^ test. Homogeneity for the distribution of 

 take by month was tested using the x" test. For mammals, the expected monthly take was equal to 1/17 

 of the total take, while for birds it was 1/7 of the total take because hunters apparently did not report 

 their bird kills during 1989. The distribution of take by hour was tested using the x^ test and assumed 



