Estimating primate densities using home range and line transect methods:: A comparative test with the black colobus monkey Colobus satanas

The line transect method is one of the main methods used to estimate primate densities. Several protocols have been proposed to analyze the data recorded under this method but none of them have been widely accepted since there is a considerable controversy about their respective accuracy. In this study, densities of the black colobus monkey Colobus satanas calculated using eight different protocols were compared with the actual density given by the home range method. Rates of polyspecific associations were also compared. The two most accurate estimates of group density (under- or overestimation < 10%) were yielded by the protocol that used the maximum transect-to-animal distance and by that of using a fixed distance of 100 m. These protocols, however, underestimated individual density because counts performed from transects underestimated by 23% the average group size. The six other protocols overestimated group density by 20 - 195%. Factors that could explain these overestimation were discussed. Because histogram of sighting frequency showed several classes of distances with no records and because groups have been detected as far as 160 m, we suggested that the uneven topography of the study site increased the variability of the sighting distances. Combined with a relatively low number of sightings (n =23). this did not allow to identify a sharp detection distance. Rates of polyspecific association found with the two methods were similar. We recommend to investigate the influence of the topography for bias in density estimates when using the line transect method.