Variação na detectabilidade e padrões de ocupação de anuros na Amazônia Oriental

Abstract

Difficulty in finding strong distribution patterns in studies with frogs is frequent. However, we do not know how much of the difficulty in detecting patterns are related to the low influence of environmental variation in the distribution of the animals or to the sampling process itself. To minimize this problem, the sampling was based on the proposal form in occupation of studies, including repeated survey to each of the sampling occasions. This sample design allows to evaluate the influence of environmental variables on the distribution patterns of frogs taking into account the influence of the detectability of species on these patterns. Our hypothesis is that there is variation in the probability of detection among frogs species, and at different times for the same species, and that this variation may be important in models of occupancy of these species. We tested eight models for the 10 most abundant species and assess their fit to the data using the Akaike Information Criterion (AIC) and the weight of the Akaike (AICwgt). For all species, among the models that best fit the data its is included one that is likely to occupation influenced by the distance of the stream (variable “Dist”) and whose probability of detection varies between surveys (variable "Survey"). Of the 10 species analyzed eight included the variable "Survey" between the three models of best fit. In addition, the probability of detection also varied among species, indicating that it is important that these variations in detection are considered when interpreting the occupancy patterns of species. This is the first work in Brazil, which analyzes the frogs distribution patterns taking into account the probability of detection of the species.