Immediate effect
To test the immediate effect of the disturbance on aquatic insect
communities, we estimated the difference between the probabilities of
occurrence of each genus before (B ) and after (A ) the
water flow change. For this analysis we used the artificial substrates
sampled on day 21. The model used was:
\begin{equation}
{a_{g}=\frac{b_{g}}{e_{g}}\backslash n}{B_{i}\sim\text{Bin}\left(b_{g},n\right)\backslash n}{A_{i}\sim\text{Bin}\left(a_{g},n\right)\backslash n}{b_{g}\sim\text{Beta}\left(1,1\right)\backslash n}{a_{g}\sim\text{Beta}\left(1,1\right)\backslash n}{e_{g}\sim\text{Gamma}\left(0,0\right)\backslash n}{\text{Effect}_{g}=\exp\left(e_{g}\right)}\nonumber \\
\end{equation}(model 1)
In model 1, Bi and Ai are
the number of occurrences of genera i in the streams before and
after the disturbance, n was the number of possible events for
the occurrence (total observations), and Effectgwas the estimated value for the effect of the disturbance on groupg .