engleski

The comparison of accuracy and efficiency of different spatial sampling design in earthworm research

Object. Earthworms (Lumbricidae) are important members of terrestrial biocoenoses, due to the relatively large proportion in biomass and numerous interactions with the environment. However, the complexity and opacity of the media in which they live requires a costly sampling. A well-planned sampling is essential to obtain representative data and their subsequent quantitative analysis and interpretation. When planning the sampling it is important to know the characteristics of species or group of organisms to be sampled, and accordingly determine the sampling parameters. In earthworm research, except for the specificity of ecological groups that certain species belong to, their particularities with respect to seasonal dynamics, spatial distribution, and limited mobility in space are also important. In previous earthworm studies various spatial sampling designs were implemented, with both variable sample size and the distance between samples. Method. Earthworm were sampled at 7 locations in two consecutive seasons. At each location a grid of 49 regularly separated sampling points (7x7) was used for sampling. Earthworms were determined to species level. A computer simulation was ran with the obtained results in order to compare them with random, transect and regular grid sampling design with different number of sampling points and different distance between them. Parameters (number of earthworms, number of species, Chao, Simpson and Shannon index, density) were calculated in order to compare their accuracy and efficiency. All simulations were made in R software. Result. The best results were obtained with a regular (raster) grid sampling design with at least 36 sampling points. The linear (transect) design had the highest variability when compared with raster or random design the same number of sampling points. Conclusion. The results of this study could helped in finding the balance between the sampling effort and accurate estimates of local species richness and the tendency to optimize the spatial arrangement of sampling points and their required number when investigating earthworms.

spatial sampling, simulation, earthworm, random, transect

nije evidentirano