Conservation of rare and endangered species often
faces urgent challenge to gather sufficient data for
policy making in a short time period. In this study we
demonstrated a simple and quick approach to obtain
adequate distribution data for a globally vulnerable
species, the Fairy Pitta. By integrating sighting data
from all possible sources, we developed a species
predictive model and designed a standardized
sampling protocol to evaluate the model effectiveness.
The rule-based model applied niche concept and
assumed all the potential variables had equal
Fig. 2. Predictive distribution of the Fairy Pitta (Pitta
nympha) and field sampling design. Three categories of
sampling cells were included: i.e., random cells within:
random cells within the predictive areas, active selection
cells within: active selection cells within the predictive
areas, and active selection cells outside: active selection
cells outside the predictive areas. All the selections were
based on the predictive areas and were used to conduct
field surveys.
contributions. The resulting 40% overall predictive
accuracy indicated that our model and the overall
approach was straightforward and valuable. We also
conducted similar field work in 2002 and 2005, the
results suggested that the model accuracy could reach
over 50% (R. S. Lin, unpublished data). In addition, the
gathered information provides sufficient pitta presence
data for decision makers and further insight on the model
refinements (Lee et al., 2006).