Effectiveness of sampling methods and further sampling for accessing spider diversity: a case study in a Brazilian Atlantic rainforest fragment

GUILHERME HENRIQUE FERNANDES DE AZEVEDO; BÁRBARA TEIXEIRA FALEIRO; IVAN LUIZ FIORINI DE MAGALHÃES; ALÍPIO RESENDE BENEDETTI; UBIRAJARA OLIVEIRA; JOÃO PAULO PEIXOTO PENA-BARBOSA; MARCUS THADEU TEIXEIRA SANTOS; PAULA FIGUEIREDO VILELA; MÁRIO DE MARIA; ADALBERTO J. SANTOS

INSECT CONSERVATION AND DIVERSITY

WILEY-LISS, DIV JOHN WILEY & SONS INC

Lugar: New York; Año: 2014 vol. 7 p. 381 - 391

1752-458X

The current biodiversity crisis makes the quantification of the diversity and the description of organism distribution particularly pressing. Biological inventories are among the most effective ways to improve the knowledge about local biota, but they can be very time and money-consuming. The determination of adequate sampling effort and the selection of cost-effective collecting methods are critical issues. In this article, a spider diversity inventory in an Atlantic semi-deciduous forest fragment in Brazil was used to compare the efficiency of three collecting methods in two different seasons in order to propose an optimised sampling protocol. The worthiness of increasing sampling effort in the target area and similar tropical ecosystems was estimated and evaluated in terms of its cost-effectiveness. For a better sampling of the spider community, it is suggested that a proportion of 55, 29 and 16% of total sampling hours should be dedicated to nocturnal hand collecting (NHC), pitfall traps and beating trays, respectively, in the rainy season. If only one method can be applied, the most efficient in terms of species per sampling is the NHC. A completeness of 70% of the estimated spider species richness (as predicted by the Chao1 estimator) was observed in the complete inventory and increasing sampling effort in the studied area may be highly ineffective when the costs involved are considered. Other studies in similar tropical rainforest areas also presented completeness around 70%, which might be a threshold from which the sampling effort necessary to raise the observed species richness substantially starts to be ineffective.