The distribution of native brook trout (Salvelinus fontinalis) in eastern North America is often limited by temperature and introduced brown trout (Salmo trutta), the relative importance of which is poorly understood but critical for conservation and restoration planning. We evaluated effects of brown trout on brook trout behavior and habitat use in experimental streams across increasing temperatures (14–23 °C) with simulated groundwater upwelling zones providing thermal refugia (6–9 °C below ambient temperatures). Allopatric and sympatric trout populations increased their use of upwelling zones as ambient temperatures increased, demonstrating the importance of groundwater as thermal refugia in warming streams. Allopatric brook trout showed greater movement rates and more even spatial distributions within streams than sympatric brook trout, suggesting interference competition by brown trout for access to forage habitats located outside thermal refugia. Our results indicate that removal of introduced brown trout may facilitate native brook trout expansion and population viability in downstream reaches depending in part on the spatial configuration of groundwater upwelling zones.

Quantifying spatial variability in fish growth and identifying large‐scale drivers of growth are fundamental to many conservation and management decisions. Although fish growth studies often focus on a single population, it is becoming increasingly clear that large‐scale studies are likely needed for addressing transboundary management needs. This is particularly true for species with high recreational value and for those with negative ecological consequences when introduced outside of their native range, such as the Flathead Catfish Pylodictis olivaris. This study quantified growth variability of the Flathead Catfish across a large portion of its contemporary range to determine whether growth differences existed between habitat types (i.e., reservoirs and rivers) and between native and introduced populations. Additionally, we investigated whether growth parameters varied as a function of latitude and time since introduction (for introduced populations). Length‐at‐age data from 26 populations across 11 states in the USA were modeled using a Bayesian hierarchical von Bertalanffy growth model. Population‐specific growth trajectories revealed large variation in Flathead Catfish growth and relatively high uncertainty in growth parameters for some populations. Relatively high uncertainty was also evident when comparing populations and when quantifying large‐scale patterns. Growth parameters (Brody growth coefficient [K] and theoretical maximum average length [L∞]) were not different (based on overlapping 90% credible intervals) between habitat types or between native and introduced populations. For populations within the introduced range of Flathead Catfish, latitude was negatively correlated with K. For native populations, we estimated an 85% probability that L∞ estimates were negatively correlated with latitude. Contrary to predictions, time since introduction was not correlated with growth parameters in introduced populations of Flathead Catfish. Results of this study suggest that Flathead Catfish growth patterns are likely shaped more strongly by finer‐scale processes (e.g., exploitation or prey abundances) as opposed to macro‐scale drivers.

Flathead Catfish Pylodictis olivaris have been either intentionally or accidentally introduced into Atlantic Slope drainages extending from Florida to Pennsylvania and have quickly become established. In Pennsylvania, Flathead Catfish were first detected in the Schuylkill River at the Fairmont Dam in 1999 and in the Susquehanna River at Safe Harbor Dam in 2002. The species has since moved throughout the respective basins, with subsequent detections during 244 riverine surveys in these drainages. Fishway and electrofishing surveys in the tidal Schuylkill River, a Delaware River tributary, have documented an increase in abundances since 2004, when the surveys were first implemented. Hoop‐net surveys in nontidal large‐river reaches found mean (±SD) catch rates varying from 0.00 to 4.51 ± 4.38 fish/series. A Bayesian hierarchical Poisson regression model indicated that Flathead Catfish abundance decreased as the distance from the initial point of detection increased, demonstrating a general pattern of fish expansion upstream from the point of detection. The distance downstream of the nearest dam, although not significant, had a relatively high posterior probability of being negatively correlated with Flathead Catfish abundance. Ongoing and future targeted surveys should help to better understand changes in the distribution and abundance of Flathead Catfish in these systems.

Walleye (Sander vitreus) population declines have been linked to climate change, but it is unclear how the growth of this cool-water species may be affected by warming water temperatures. Because warming rates vary among lakes, it is uncertain whether lake characteristics may mediate the temperature effects on walleye growth may vary as a result of differences in lake habitat or productivity. In this study, we (1) quantified walleye annual growth from 1983-2015 in 61 lakes in midwestern United States; (2) estimated the relationship between annual early life growth (\omega; mm·year-1) and water growing degree days (GDD); and (3) identified lake characteristics affecting log(\omega)-GDD relationships. On average, \omega estimates significantly increased with increasing GDD; however, this relationship varied in direction and magnitude among lakes. We estimated an 84% posterior probability of a negative effect of water clarity on the log(\omega)-GDD relationship, suggesting that water clarity may mediate the effect of warming water temperatures by affecting the magnitude and direction of the log(\omega)-GDD relationship. Our results provide insights into the conservation of cool-water species in a changing environment and identifies lakes characteristics in which walleye growth may be more resilient to climate change.

Various abiotic and biotic factors affect fish and their habitats at macroscales. For example, changes in global temperatures will likely alter demographic rates, including growth. However, to date, there is no statistical framework for assessing the ability to detect macroscale effects on fish growth under different sampling scenarios. We provide a generalized framework for calculating the frequentist and Bayesian power of detecting macroscale effects on fish growth. We illustrate this framework for a range of sampling scenarios which varied in the number of fish sampled per lake, the number of lakes sampled, and the magnitude of the temperature effect on growth for two case study species. However, the framework can be adapted to investigate other species, sampling scenarios, and environmental drivers. The ability to detect macroscale effects was more affected by the number of lakes sampled, rather than the number of fish sampled from each lake. Confidently detecting macroscale effects likely requires sampling hundreds of lakes. This was true for both case study species, despite different life histories and extents of spatial variability in growth.