Increase in benthic trophic reliance on methane in 14 French lakes during the Anthropocene

Lakes play a key role in the regulation of the global carbon cycle. However, their functioning can be strongly impacted by anthropogenic pressures and climate variability. Understanding the response of the carbon cycle to environmental changes remains a crucial, elusive goal for both ecosystem managers and aquatic ecologists. In particular, the relations among lake physical and chemical properties, landscape structure and lake carbon cycling must be studied to predict future trends in lake functioning. Sediment cores were collected from the deepest part of 14 small French lakes that differed in lake properties (elevation, conductivity, area, area of the watershed) and land‐use class (forest, wetland, agricultural land and urban area). The sampling strategy employed the top‐bottom approach (a comparison between present‐day conditions and ‘reference’ conditions at Medieval period, c. AD 1000). For each sample, the following variables were analysed: isotopic carbon composition of sedimentary organic carbon (δ¹³COM), δ¹³C of chironomid remains (δ¹³CHC), and sedimentary pigments (total carotenoids, TC). Stepwise multiple regression analysis showed that the size of the catchment area may affect δ¹³COM values for the Medieval samples (R² = 0.36, P < 0.05), such that the flux of terrestrial organic matter increases with the size of the watershed. However, this relation is not observed in the present‐day samples, and the influence apparently becomes largely anthropogenic. For these present‐day samples, the proportion of agricultural land in the watershed appears to be the primary driver of lake biogeochemical cycles through a direct effect on nutrient availability (R² = 0.24, P < 0.05) and through an indirect effect on the benthic carbon cycle. The results also confirm the widespread existence of a pathway for methane‐derived carbon contribution to chironomid biomass (up to 61% of chironomid biomass) and suggest that high‐conductivity lakes are highly sensitive to the presence of this pathway (more than 75% of the lakes in our dataset have benthic food webs apparently dependent on biogenic methane; CH₄). The results may indicate that the high‐conductivity lakes (from the Jura Mountains) are more vulnerable to anthropogenic activities than low‐conductivity lakes because high‐conductivity water provides an excellent nutritive medium for the development of photoautotrophic production. Studying within‐lake CH₄ dynamics and the response of the CH₄ cycle in high‐conductivity lakes appear to be crucial for understanding both regional carbon budgets and lake trophic functioning.