Total Species Turnover, Total Rewiring, and Traits
Using linear regression models, we explored how different species traits and abundances influenced the total number of rewirings (bottom panel of Figure 3). While we did not find any meaningful models regarding key life history traits or species’ abundances, we found that species traits related to maximum length and feeding helped explain around 35% of the total number of rewirings. Specifically, total number of rewirings had a significant and negative relationship with a species piscivore status and a significant positive relationship with the interaction between maximum length of a species and its piscivore status, Figure 4 (βMax length =0.39, p >0.05; βPiscivore=-9.49, p <0.01; βMax length*Piscivore=0.05, p <0.01; adjusted R2 =0.344; Table S2). The negative relationship between piscivore status and the total number of rewirings suggested that non-piscivores have a greater number of rewirings than piscivores when controlling for the maximum length of a species. Interestingly, despite there being no overall effect of max length on the total number of rewirings, the significant interaction indicates that for piscivores, there is a strong positive relationship between the total size and the number of rewirings that does not exist for the non-piscivores.
We also note that while no specific traits or species abundances explained the total number of species turnover connections, all fish that contributed to species turnover (n =7) had preferences for large rivers (compared to only 60% of the non-turnover fish). Furthermore, Micropterus salmoides , one of the seven turnover species, explained about 24% of the total number of species turnover connections at a threshold of 0.5 (βMicropterus salmoides=7.136, p <0.001;R2 =0.24).