Linking the Data
We did not include in the final analyses some of the
66 survey sites because data on fishes were incomplete,
social and economic data were lacking, and association
between villages and ecological data was unclear (due
largely to complex resource-use rights determined by genealogies)
and because we reduced ecological variability
by omitting fish-survey locations classified as sheltered
from prevailing weather (Karlson et al. 2004). Consequently,
for 25 sites we had data on 3 potential distal