Census techniques for southern reedbuck, Redunca arundinum, on forestry lands in the Natal Midlands, South Africa.

摘要

Census techniques for southern reedbuck, Redunca arundinum, were tested on a forestry estate in the Natal Midlands, South Africa. Both transect and systematic sampling strategies were evaluated. Data were also collected on sex ratio and age structure. Systematic dawn flush counts and night habitat counts (before and after first significant frosts) were used to estimate minimum population size. Repeated measures ANOVA, post hoc Scheffe tests and coefficients of variation were used to assess the comparative repeatability of the census techniques. Dawn flush counts yielded very low estimates of abundance. Night habitat counts (before frosts) performed best of the systematic methods, obtaining highest repeatability and minimum population estimate. For transect counts, three enumeration and two estimation (computation) methods, including both modified and fixed strip width transects, were tested. Transect counts consistently over-estimated population size, but were relatively accurate and repeatable when fixed strip width (determined using the mean visibility index) sampling was used. This method was recommended as the transect method most suitable for monitoring reedbuck abundance on forestry lands in the Natal Midlands. The sex ratios obtained showed a bias towards females; and quantification of population structure showed poor repeatability and limited potential for future application..