Comparison of Sampling Methods Used to Estimate Spider (Araneae) Species Abundance and Composition in Grape Vineyards

The effectiveness of 3 methods for sampling spiders in grape vineyards was tested. The sampling methods were as follows: (1) a drop-cloth method in which spiders were dislodged from a 5.1-m2 area of grapevine onto a drop cloth, (2) a funnel method in which spiders were dislodged into a 0.74-m2 funnel, and (3) a D-vac method in which 50 sections of grapevine (3D-em sections, total area of 5.4 m2) were suctioned with a gasoline-powered vacuum. Data from all sampling methods were adjusted for area sampled and compared with an absolute control, in which foliage from an entire vine was removed and searched for spiders. Analyses were made on the follOwing 8 most abundant spider species: Allyphaella pacifica Banks, Metaphidippus vitis (Cockerell), Theridion dilutum Levi and Theridion melanumm Hahn (grouped as Theridion spp.l, Cheiracanthium inclusum (Hentz), Hololena nedra Cham­ berlin & Ivie, Trachelas pacificus (Chamberlin & Ivie), and Erigone dentosa (O.P.-Cambridge). Mean spider abundance from each sampling method differed Significantly from the absolute control in (::=1) speCies. The funnel method often overestimated spider abundance·while the drop-cloth and D-vac methods often underestimated spider abundance. Estimates of spider species composition were most accurately measured by the drop-cloth method and were sig­ nificantly altered by the D-vac method. We tested 3 data transformations (In[x + 1], V [x + 0.5], and x l •bl2 ) to stabilize the variance and satisfY the requirements of analysis of variance (ANOVA). No single transformation uncoupled the mean and variance for all spider species over the 3 sampling methods. Using mean and variance of mean spider abundance, sample size and disperSion parameters were created by way of the Taylor power law. Sample sizes were determined for each sampling method; estimates of sample size and sampling costs varied considerably among spider species and sampling methods. All but 1 spider species collected with the drop-cloth method had a clumped distribution, the funnel method produced variable dispersion indexes, and all spider species collected with the D-vac method had a random distribution pattern. Whereas no Single sampling method was best for all spider species, the funnel method is favored because of its efficiency in estimating spider densities and its reduced cost.