Point and interval statistical performances of Morisita patchiness index estimators in different sampling schemes: simulated data
Description
This dataset comprises 1,440,000 lines according to different simulated scenarios of varying fragment shape, sampling unit shape and size, aggregation effect, edge effect and sampling effort. The variables (columns) are: rep (replication number of a scenario), sampling (A: hexagonal lattice sampling, B: quadrat lattice sampling, C: circlet lattice sampling, D: east-to-west transect sampling, E: south-to-north transect sampling), spptype (type of point process: either 'poisson' for complete spatial randomness or 'thomas' for modified Thomas process), agg (gaussian radius of the modified Thomas process; 0: no aggregation, i.e., spptype=='poisson', 1: 0.04 radius, 2: 0.02 radius, 3: 0.01 radius), ftype (either convex or square shape for the fragment), quad.side (area of the sampling units; S: small- each sampling unit with 1x10-4 area; M: medium- each sampling unit with 4x10-4 area, B: big- each sampling unit with 1.6x10-3 area), sample.area (sampling effort; ss: small- 1% of the total area of the fragment, ms: medium- 5% of the total area, bs: big- 10% of the total area), dens (population density- constant in Hausdorff distance), affty (affinity for core area; 0: equal preference for edge and core area; 1-4: increasing edge avoidance), frag.area (total area of the fragment), n.points (number of points generated in the point process), in.points (number of points within all sampling units of a given shape), Id (populational Morisita patchiness index), Id.hat (MLE of Id), Id.jackk (jackknife estimator of Id), V.jackk (jackknife variance estimate of Id-- this was calulated wrong!), Id.boot (boostrap estimate of Id), V.boot (boostrap variance estimate of Id), P025.boot (2.5 bootstrap percentile), P975.boot (97.5 bootstrap percentile).
Files
Steps to reproduce
The dataset was simulated in R software. We will happily provide the source codes for anyone interested in replicating our results. Please send an e-mail to davibg@ufsj.edu.br with your request.