Limnological characteristics
Limnological characteristics varied among the sampled reservoirs (see Appendix S1 in supporting information). Principal component analysis (PCA) identified three major axes of limnological variation (axes that explain more than 10 % of the variation). PCA reduced the data set to three components that explained 64.27% of the variance in the original variables. The first and second axis of the PCA explained 51.07% of the total variation observed. PC1 explained 28.69% of the variation and can be interpreted as a measure of altitude, surface area, total nitrogen (TN) and total phosphorus (TP) (Fig. 2). PC2 explained 22.38% of the limnological variation and was loaded heavily by electrical conductivity, pH and depth. PC3, which explained 13.2% of the variation, was loaded mostly by Temperature and transparency. As can be seen in Fig. 2 three major clusters of reservoirs is created based on the PCA components. Cluster 1 is composed of 6 lowland reservoirs and cluster 2 is composed of 12 highland reservoirs while cluster three is composed on 17 reservoirs of mixed elevation but mostly with high nutrient concentration (Fig. 2; see Appendix S2 in supporting information).