Methodology
The WPI targets communities of terrestrial forest and savanna birds and mammals that occupy top trophic levels in their ecosystems. To develop a WPI, the chapter begins with a camera trap survey designed to be (i) spatially representative of the area of interest, (ii) of sufficient sampling intensity to detect a representative sample of the species in the target community, and (iii) of sufficient duration to develop unbiased occupancy estimates for the sampled species. Then, it presents a preliminary bootstrapping analysis of the WPI to examine the precision and robustness of the index.
Data description
It is based on the state variable of occupancy, rather than abundance (or density) because of the difficulty of developing unbiased estimates of abundance for a set of species when animals are not individually recognizable. Typically, a community of terrestrial birds and mammals will contain a few species that can be recognized as individuals and tracked though space and time. For these species, it is possible to develop unbiased estimates of abundance or density using capture-recapture methods
and spatially explicit capture-recapture models. Many other species, however, are difficult to identify reliably as individuals, and some relative abundance index must be used in place of density or abundance estimates. The result is a set of abundances composed of unbiased estimates and many relative abundance indices with unknown, species-specific biases. A composite index based on such a mix of biased and unbiased estimates will be biased to an unknown degree, and trends developed from such indices
will be open to several interpretations, depending on the strength of the underlying assumptions (O’Brien, 2011).
and spatially explicit capture-recapture models. Many other species, however, are difficult to identify reliably as individuals, and some relative abundance index must be used in place of density or abundance estimates. The result is a set of abundances composed of unbiased estimates and many relative abundance indices with unknown, species-specific biases. A composite index based on such a mix of biased and unbiased estimates will be biased to an unknown degree, and trends developed from such indices
will be open to several interpretations, depending on the strength of the underlying assumptions (O’Brien, 2011).