Stream fish Bayesian size spectrum model

Biomass size spectra are useful tools for ecologists to investigate macroecological processes such as trophic energy transfer and productivity. However, little is known about how different methods of aggregating data across spatial scales of river networks may affect community size spectra results. We used size-binned data (0 – 2048 g) of fish assemblages from three Lake Ontario watersheds to compare fish size spectra slopes across multiple stream classification systems and the effects of sampling design on size spectra at broader spatial scales. The slope of individual site-based size spectra ranged from -2.901 to -1.382 (median -1.718) while watershed-level size spectra had an average slope of -1.77. Size spectrum slopes did not differ across stream classes, though sites with salmonid species exhibited less negative slopes. Aggregated size spectra showed better model fit than individual site models regardless of stream order. Precision improved with stratified random sampling and larger sample sizes (>15 sites) at the watershed scale. Aggregating sites using different strategies offers effective approaches for modeling size spectra, supporting investigations into macroecological processes in river ecosystems.

Methods
This repository includes model parameters for the size selectivity correction (b0.summary.csv; beta.summary.csv) and simulated stream fish data (samplefishdata.csv) that can be used in the described Bayesian size spectrum model. The fish data includes body size measurements, both length and weight, of all fish captured at the site.