The temperature dependence of predation rates is a key issue for understanding and predicting the responses of ecosystems to climate change. Using a simple mechanistic model, we demonstrate that differences in the relative performances of predator and prey can cause strong threshold effects in the temperature dependence of attack rates. Empirical data on the attack rate of northern pike (Esox lucius) feeding on brown trout (Salmo trutta) confirm this result. Attack rates fell sharply below a threshold temperature of þ118C, which corresponded to a shift in relative performance of pike and brown trout with respect to maximum attack and escape swimming speeds. The average attack speed of pike was an order of magnitude lower than the escape speed of brown trout at 58C, but approximately equal at temperatures above 118C. Thresholds in the temperature dependence of ecological rates can create tipping points in the responses of ecosystems to increasing temperatures. Thus, identifying thresholds is crucial when predicting future effects of climate warming.

Classical methods for estimating the abundance of fish populations are often both expensive, time-consuming and destructive. Analyses of the environmental DNA (eDNA) present in water samples could alleviate such constraints. Here, we developed protocols to detect and quantify brown trout (Salmo trutta) and Arctic char (Salvelinus alpinus) populations by applying the droplet digital PCR (ddPCR) method to eDNA molecules extracted from water samples collected in 28 Swedish mountain lakes. Overall, contemporary fish CPUE (catch per unit effort) estimates from standardized survey gill nettings were not correlated to eDNA concentrations for either of the species. In addition, the measured environmental variables (e.g. dissolved organic carbon concentrations, temperature, and pH) appear to not influence water eDNA concentrations of the studied fish species. Detection probabilities via eDNA analysis showed moderate success (less than 70% for both species) while the presence of eDNA from Arctic char (in six lakes) and brown trout (in one lake) was also indicated in lakes where the species were not detected with the gillnetting method. Such findings highlight the limits of one or both methods to reliably detect fish species presence in natural systems. Additional analysis showed that the filtration of water samples through 1.2 μm glass fiber filters and 0.45 μm mixed cellulose ester filters was more efficient in recovering DNA than using 0.22 μm enclosed polyethersulfone filters, probably due to differential efficiencies of DNA extraction. Altogether, this work showed the potentials and limits of the approach for the detection and the quantification of fish abundance in natural systems while providing new insights in the application of the ddPCR method applied to environmental DNA.

The characteristics and dynamics of lake ecosystems strongly reflect the relative sizes of the littoral (the nearshore habitat where photosynthetically active radiation penetrates to the lake bottom in sufficient quantities to support photosynthesis) and pelagic zones (the rest of the lake; Lodge et al., 1988;Wetzel, 1990;Vander Zanden & Vadeboncoeur, 2020). However, predictive relationships for patterns of the relative size of these habitats across landscapes are not available, probably due to the paucity of bathymetric data relative to the global abundance of lakes (Hollister et al., 2011;Seekell, 2018;Wetzel, 1990). In particular, there is need to develop scaling relationships that relate habitat size to commonly measured lake characteristics (Cael et al., 2017;Seekell et al., 2013). Such relationships provide the simple rules used to generalize understanding of aquatic ecosystem patterns and processes at regional to global scales (Downing, 2009).Among lakes, the littoral zone varies in size from nearly 0% to 100% of the total lake area. This variation is captured in a conceptual model created by Wetzel (1990), and popularized in his widely read text book, that depicts the relationship between the logarithm of lake abundance and the logarithm of the pelagic:littoral area ratio. However, this log-ratio formulation can neither accommodate lakes completely comprised by littoral area, which is relatively common for shallow lakes, nor does it explain why some similarly sized lakes are completely comprised by littoral area while others have almost none. An empirical analysis by Henson (1993) identified an inverse scaling relationship between relative littoral habitat size (littoral area/ total surface area) and mean depth for lakes with mean depths greater than 4 m. However, about 85% of Earth's lakes have a mean depth less than 4 m, and this scaling relationship predicts that these lakes are comprised of >100% littoral area (Cael et al., 2017;Henson, 1993

Benthic gross primary production (GPP) is often the most important part of aquatic food webs in northern lakes, which are gradually warming and receiving increased terrestrial colored dissolved organic carbon loadings due to global change. Yet, measurements of benthic GPP are fairly uncommon, and methods and unit dimensions of benthic GPP are unstandardized and rarely compared. In this study, we measured benthic GPP in 27 headwater lakes from three regions in northern Sweden and analyzed potential constraining drivers of benthic GPP z rates at discrete depths and estimates of benthic GPP averages across the whole lake, as well as across the littoral zone. We also compared in situ measurements of benthic GPP averages across the whole lake with modeled values using the "autotrophic structuring model." We found that benthic GPP z rates were best explained by, and positively related to, available light (i.e., a function of depth and water color) and temperature. Benthic GPP averages across the whole lake, on the contrary, were best explained by the relative size of the littoral area, which is a measure that combines lake bathymetry and water color. The comparison between in situ measured and modeled estimates of benthic GPP averages across the whole lake revealed that (1) the autotrophic structuring model underestimates GPP at low values and overestimates GPP at high values compared with measured data, and that (2) measured values were related to temperature, which is not included as a variable in the autotrophic structuring model. Considering future predicted changes impacting northern latitude lakes, our results suggest that increased lake water temperatures can to some extent mitigate the negative impacts of reduced light availability from lake browning on benthic GPP z rates. The combined impact of these changes on benthic GPP averages across the whole lake will depend on, and be moderated by, lake bathymetry determining the relative size of the littoral area.