A inventory of the spider fauna was carried out at Serra do Cachimbo, inside the "Campo de Provas Brigadeiro Velloso", Novo Progresso municipality, Pará state, Brasil. Two expeditions were made, during both dry (August and September, 2003) and wet (March and April, 2004) seasons. Three collectors participated in each expedition. The sampling effort, represented by 240 samples, was divided through the following methods: beating tray and sweeping net (96 samples); nocturnal hand search (96 samples) and concentrated litter sorted by a combination of hand search and Winkler extractors (48 samples). The spiders assemblages from four vegetation types (open rainforest, Formiga river's riparian forest, arboreal savanna and white sand vegetation) were compared. A total of 4964 spiders were obtained, from which 2724 adults, resulting in 397 recognized morpho-species in 37 families. The most abundant families were Theridiidae, Salticidae and Araneidae and the ones with more species were Salticidae, Araneidae e Theridiidae. The singletons represented 40% of the total species richness and only two species presented more than 100 individuals. The estimated richness species curves reached from 473 (bootstrap) to 674 (jackknife2) species. The higher alpha-diversity (Shannon-Wiener index) was observed in open rainforest, followed by Formiga river's riparian forest, white sand vegetation and arboreal savanna. the higher beta-diversity (Jaccard and Morisita-Horn indexes) was observed between Open rainforest and white sand vegetation and the lowest one was between open rainforest, arboreal savanna and Formiga river's riparian forest. The dry season presented more species than the wet season. However, no such difference was detected in white sand vegetation. A principal component analysis revealed that some species are linked to arboreal vegetations and others are linked to the open vegetation in white sand. Those differences in diversity and taxonomic composition between the various vegetations can be explained by drifts in availability of food items (preys), space resources (refugees and substratum to web attachment), as well as micro-climatic factors (temperature and moisture) of a given vegetation type. The most experienced conectar sampled more species than any other collector and all less experienced collectors sampled similar number of species. The differences in abundance among samples obtained by each conectar were not significant. The night hand sampling was the most efficient technique regarding the detection of differences in species richness among vegetation types. The methods beating tray/sweep net and litter sampling do not showed any significant difference in relation to species richness.
|
