Explanatory variables used to model AIV prevalence in Spanish wetlands.

<p>Variables marked with * were excluded as highly correlated (Spearman's coefficient ≥ |0.6|) with other variables within their factor.</p>1<p>Flamingos and White storks were considered separately due to previous epidemiological data obtained in the area that identified these species as important AIV carriers <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046418#pone.0046418-Busquets2" target="_blank">[17]</a>, <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046418#pone.0046418-PrezRamrez1" target="_blank">[18]</a>.</p>2<p>In this index we grouped dabbling ducks and flamingos due to similar feeding habits, as both avian groups feed on surface water, which has been identified as a risk factor in AIV epidemiology <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0046418#pone.0046418-Munster1" target="_blank">[24]</a>.