Comparing the precision and cost-effectiveness of faecal pellet group count methods.
Abstract
Faecal pellet group (FPG) count data are widely used to estimate animal abundance, with two alternative methods normally employed. Faecal accumulation rate (FAR) techniques measure the daily accumulation rate of pellet groups, while faecal standing crop (FSC) techniques measure overall density. To estimate abundance, both methods require estimates of the animal defaecation rate. FSC techniques also require an estimate of pellet group decomposition rate. In general, FAR techniques are considered less prone to bias while FSC methods are considered more precise and cost-effective. On balance, the majority of authors and practitioners prefer FSC methods, although little empirical evidence supports this decision. FPG count data were obtained from 26 study areas to compare the precision of FAR and FSC count techniques when applied to wild deer populations in the UK uplands. The time needed to collect count data was quantified in 10 study areas. The coefficients of variation (CV) of FSC pellet group count data ranged from 9% to 23% and were approximately 0.7-0.9 times those of equivalent FAR data. The precision of both methods was related to the density of pellet groups. On average, FSC count data took 80 min per plot to obtain, with FAR taking 1.6-1.9 times longer. For the precision of FSC and FAR abundance estimates to be comparable in the range of conditions studied, decomposition rate trials would require a CV of 5-20%. While a number of studies report this to be possible, estimates of the time needed to obtain this level of precision generally exceed the net available time that results from the deployment of FSC rather than FSC pellet group counts. Using the levels of finance available to most deer managers in the UK uplands, deer abundance estimates obtained using FSC techniques on individual study sites up to 20 000 ha appear generally less cost-effective than FAR when compared in terms of their overall precision. As FAR methods are also thought to have less potential for bias when applied in the appropriate environmental conditions, they should be preferred over FSC when estimating deer abundance in concealing habitats.