Materials and Methods
Using fish abundance data from NEON, we constructed seasonal interaction
networks using a two step-process: (i) developed Fall and Spring Poisson
log-normal (PLN) models to explain species abundances in each season,
and (ii) adopted the EMtree approach that used the fitted PLN models to
create nondirectional seasonal species interaction networks. We then
quantified species turnover and rewiring using a measure of
beta-diversity. Finally, we used linear regression models to explore how
species-specific traits explained the total number of interactions
classified as “rewiring”, and “species turnover”.