Resources and temperature play major roles in determining biological production in lake ecosystems. Lakes have been warming and ‘browning’ over recent decades due to climate change and increased loading of terrestrial organic matter. Conflicting hypotheses and evidence have been presented about whether these changes will increase or decrease fish growth within lakes. Most studies have been conducted in low-elevation lakes where terrestrially derived carbon tends to dominate over carbon produced within lakes. Understanding how fish in high-elevation mountain lakes will respond to warming and browning is particularly needed as warming effects are magnified for mountain lakes and treeline is advancing to higher elevations. We sampled 21 trout populations in the Sierra Nevada Mountains of California to examine how body condition and individual growth rates, measured by otolith analysis, varied across independent elevational gradients in temperature and dissolved organic carbon (DOC). We found that fish grew faster at warmer temperatures and higher nitrogen (TN), but slower in high DOC lakes. Additionally, fish showed better body condition in lakes with higher TN, higher elevation and when they exhibited a more terrestrial δ13C isotopic signature. The future warming and browning of lakes will likely have antagonistic impacts on fish growth, reducing the predicted independent impact of warming and browning alone.

Lake sampling

At the deepest point in each lake, in situ measurements of temperature was taken using a YSI probe (YSI Incorporated, Yellow Springs, Ohio, USA). Surface water samples were filtered through 63-μm mesh to remove zooplankton and processed for chlorophyll-a (chl-a), particulate organic matter (POM), total nitrogen (TN), total phosphorus (TP) and dissolved organic carbon (DOC). For chl-a quantification, a known volume of water was filtered through 0.7 μm glass fiber filters (GF/F Fisher Scientific) and frozen. Chl-a, a proxy for phytoplankton biomass, was measured using a fluorometer after a 24 h cold methanol extraction. For POM isotope analysis, a known volume of water was filtered through pre-weighed pre-combusted (7 h, 500 °C) 0.7 μm glass fiber filters. Upon returning to the laboratory, filters were dried for 24 h at 60 °C, weighed and packaged in tin capsules for 13C and 15N isotope analysis. Total nitrogen and total phosphorus samples were collected in HDPE vials and preserved with H2SO4 to a pH < 2 and stored at ~ 4 °C until analysis. TN and TP were measured using an auto analyzer (LaChat QuikChem 8500, persulfate digestions). Leaves of several common plant species were collected from shoreline and frozen until processing for isotopic analysis. Leaves were sorted into broad functional groups (grasses, shrubs, pine) and dried at 60 °C for 2 days. A mortar and pestle was used to grind the leaf samples before packaging in tin capsules for isotope analysis. Based on a subset (10 lakes) of the plant data, we chose to process a grass and pine sample to capture the maximum variation in isotopes within the terrestrial organic matter entering lakes.

To quantify DOC, water samples were filtered through pre-combusted glass fiber filters (Whatman GF/F, pore size 0.7 μm) into triple-rinsed 20 mL glass vials and preserved with HCl to a pH < 2. DOC was measured using a total organic carbon analyzer (TOC-V CSN, Shimadzu Scientific Instruments, Japan). To characterize DOC quality, we used UV–Vis absorbance, spectrofluorometry and spectrophotometry, which reflect several aspects of the molecules comprising the light-absorbing and fluorescing DOM pool, respectively. We used excitation-emission matrices (EEMs) as a three-dimensional representation of fluorescence intensities scanned over a range of excitation/emission wavelengths (Coble 1996; Chen et al. 2003). EEMs were collected with a JY-Horiba Spex Fluoromax-3 spectrofluorometer (HORIBA, Japan) at room temperature using 5 nm excitation and emission slit widths and an integration time of 1.0 s. The Aqualog spectrophotometer simultaneously collects both fluorescence and absorbance spectra on a sample. All fluorescence spectra were collected in signal-to-reference (S:R) mode with instrumental bias correction. Instrument-specific corrections, Raman area normalization and Milli-Q blank subtraction were conducted with Matlab (2009). From the UV–Vis absorbance and EEMs data, we calculated two indices of DOC quality: the freshness index (FI) and specific UV absorption (SUVA). FI (β:α) is a ratio of emission intensity at 380 nm to that of the region between 420 and 435 nm at an excitation of 310 nm and is reflective of recently produced algal organic matter (Parlanti et al. 2000). SUVA is a DOC-normalized index of aromaticity calculated as UV absorbance at 254 nm/[(DOC (mg L−1) × path length (0.01 m)] (Weishaar et al. 2003). FI increases with autochthonous carbon production, whereas SUVA increases with allochthonous carbon production.

All isotope samples were analyzed by the University of California, Davis Stable Isotope Facility for 13C and 15N, using an elemental analyzer interfaced to a continuous flow isotope ratio mass spectrometer.

Fish sampling

At each of the 21 lakes, we caught fish by angling. Each fish was identified to species, weighed, photographed and measured (TL; maximum length). We collected a dorsal muscle sample from each individual which was frozen until processing for stable isotope analysis. Upon returning to the laboratory, muscle samples were freeze dried for 24 h, ground with a mortar and pestle and packaged for 13C and 15N analysis. Otoliths were removed, cleaned, dried and stored in vials for age determination and growth rate analysis. We calculated catch per unit effort (CPUE) as the total catch divided by the number of person-hours spent angling at each lake. All applicable institutional guidelines for the use of animals were followed and approved by the Institutional Animal Care and Use Committee at the University of California, San Diego (Protocol #S14140).
Fish sample processing

Fish in temperate regions can be aged by examining calcified structures called otoliths, which form annuli—rings that correspond to low winter growth. The width of the annuli is an indicator of annual growth (Casselman 1990). To determine age and annual growth, the sagittal otoliths were mounted on the edge of a microscope slide with the core positioned just within the microscope slide’s edge and polished to section the otolith. The otolith was then flipped onto the transverse cross-section and polished again until the core was exposed in the transverse section similar to Taylor and McIlwain (2010). Annuli were counted by two independent readers in the absence of information about fish size or lake. Ages were in agreement for 84% of the otoliths, and never differed by more than 1 year. For otoliths where the age determinations disagreed, the two readers examined the otoliths together and were able to reach a consensus. The width of each annuli was measured using imaging software (Image J).

Please see ReadMe files for units.