As forest destruction and fragmentation advance throughout the Brazilian Amazon, it has become important to determine how these processes affect the fauna in various vegetation types in order to predict impacts and the conservation value of habitat fragments for different animal taxa. Invertebrates are useful bio-indicators because of their adaptive and dispersal potential and intimate connections with the environment. The use of spiders to evaluate forest fragmentation is recent and only slightly employed, although the group is mega-diverse and biologically tied to environmental composition and structure. This study used spiders to evaluate the effects of forest fragment size, degree of isolation, and distance to roads in 15 forest islands in a savanna matrix and 6 areas of continuous forest, in Alter do Chão district, Santarém municipality, Pará state, Brazil. Capture of spiders involved 252 man-hours of effort with an entomological umbrella and manual nocturnal collections. Transects of 250 m were collected three times, and the summed results of each transect comprised a sample unit. In total, 7751 spiders were captured, including 5477 immature and 2274 adult specimens in 306 species belonging to 32 families. Spider community characteristics, analyzed by MDS (Multidimensional Scaling) with Bray-Curtis distance, showed separation between continuous forest and forest fragment habitats. For species with more than 10 individuals in the collections, an analysis was made of the response to the first dimension of ordination, and a direct ordination was made using characteristics of each collection area (distance from the forest island to continuous forest, area of forest fragment, and form index of the forest island). GLM analysis, used to evaluate the effects of environmental degradation, showed significant differences for number of trees in each forest fragment and for distance to roads: forest fragmentation was significant to spider communities only in terms of size of forest fragment (on axis 1 of the MDS). An ANOVA used to compare species richness showed greater richness in continuous forest, differing from the result of rarefaction curves that predict slightly higher species richness in forest islands. The hierarchical standard of the spider community was obtained with the Nestedness Temperature Calculator Program–Nestcalc.
|
