Estimating acridid densities in grassland habitats: A comparison between presence–absence and abundance sampling designs
Résumé
Sampling methods to estimate acridid density per surface area unit in grassland habitats were compared using presenceÐabsence data and count data. Sampling plans based on 6 yr of surveys were devised to estimate the density of Chorthippus spp., Euchorthippus spp., and Calliptamus italicus L. These acridids represented >90% of species in the study area. Sampling plans based on count data provided a reasonable tool when densities were >1/m2 and when the level of precision was 0.20-0.30. A binomial sampling plan can be used to estimate C. italicus density with a level of precision >0.28. Sampling characteristics, i.e., estimated mean, actual precision, and sample size, were established on validation data sets with bootstrapping analysis. Sampling costs were also calculated according to density-dependent functions. Comparison between binomial sampling and enumerative sampling of C. italicus showed that binomial sampling required less time than enumerative sampling when densities were 2/m2 and when fixed precision was 0.35. Plot area had no signiÞcant effect on sample variances of counts