Data Package Metadata   View Summary

Nine mountain lakes with brook trout (Salvelinus fontinalis) and four reference fishless lakes in the Canadian Rockies were included in a replicated whole lake ecosystem experiment that involved intensive harvesting and detailed population demography estimation of brook trout from 2017 to 2020 to assess harvesting effects (Brookes et al. 2022; Matte et al. 2023), as well as and associated ecological effects in the lakes (this paper). The experimental harvesting treatment resulted in lake-specific differences in the effective density of brook trout among the mountain lakes at different elevations. With three exceptions, all study lakes that were historically stocked contained brook trout as the sole fish species in the lakes. The exceptions are Mud Lake with a small population of longnose dace (Rhinichthys cataractae), Margaret Lake with small numbers of cutthroat trout (Oncorhynchus clarkii), and Dog Lake which also contains rainbow trout (Oncorhynchus mykiss), longnose dace, longnose sucker (Catostomus catostomus), and bull trout (Salvelinus confluentus) (Messner et al. 2013; Parks Canada, unpublished data). We also included four fishless lakes (n=2 high elevation, above treeline; n=2 low elevation, below treeline) in our study as an ecological reference for aquatic communities that had not been or had been minimally impacted by historical exotic brook trout stocking.

For each of the nine lakes containing brook trout, Mark-recapture studies were conducted in the summer of 2018, between May 27 and June 30, except for Cobb Lake, where isolated marking events occurred until September 12. Detailed methods for estimating brook trout population demography are provided in the Supplementary Information. To estimate the population-level consumptive impact of brook trout on zooplankton and macroinvertebrate communities, we used effective density as our metric of abundance (Post et al. 1999). This metric assumes that individual-level consumptive effects scale allometrically (Peters 1983; Schmidt-Nielsen 1984; Wilson 1975), according to the following formula: Consumption = Individual Mass^2/3

The aggregate population-level consumptive effect of brook trout can thus be modelled using a detailed formula, which can be found in the paper

We note that this metric assumes that our size-structure estimates were representative of the size structure of the entire population.

Water chemistry, zooplankton, macroinvertebrate, and eDNA samples were collected near the end of the open-water season (August to September) in 2018 from all 13 lakes. All study lakes were sampled for pH, dissolved oxygen (DO, mg/L), total phosphorus (TP, µg/L), total nitrogen (TN, mg/L), dissolved organic carbon (DOC, mg/L) and algal biomass (Chla, µg/L) to characterize lake environments.

We analyzed crustacean zooplankton and macroinvertebrates by manual, morphological methods between August and September 2018 for all 13 lakes to cover taxa potentially not captured by COI eDNA metabarcoding (methods in Supplementary Information). We applied eDNA metabarcoding with cytochrome oxidase I (COI) mitochondrial DNA using universal primers designed by Leray et al. (2013) with the goal of capturing sequence variation for taxonomic groups represented by the crustacean zooplankton and by macroinvertebrates.