LAUSR

Biogeochemical processes are active year-round in ice-covered lakes, such that processes in one season can affect limnological conditions in subsequent seasons. However, the extent and nature of these legacy effects are poorly understood, particularly for the CO2 content of lakes and when considering gas exchange with the atmosphere. Here, we used a unique 36-yr dataset of weekly limnological measurements of Buffalo Pound Lake in the northern Great Plains to assess seasonal changes in CO2 concentration and flux and determine how dependent lake pCO2 is on limnological conditions of previous seasons. We found that the lake was a net source of CO2 to the atmosphere (mean 18.5 7.4 mol CO2 m yr), with spring potentially accounting for the majority (~ 64% 20%) of CO2 efflux, assuming ice in spring was permeable to gas exchange (32.9% 19.8% if not). Analysis with generalized additive models (GAMs) demonstrated that current and antecedent seasonal conditions combined to explain 72.6% of deviance in spring pCO2, but that the strength of model predictions and the importance of antecedent conditions diminished in GAMs of summer (43.6%) and fall (23.3%) CO2 levels. This research suggests that pCO2 is regulated by a combination of coeval and historical environmental conditions and shows that quantification of seasonal and annual fluxes requires a mechanistic understanding of the legacy effects of preceding time intervals. It is now well established that inland waters contribute significantly to the global carbon budget (Cole et al. 2007; Prairie 2008; Tranvik et al. 2009), although many questions remain about the factors regulating variability in water-column pCO2 at broad spatial and temporal scales. One such uncertainty relates to the legacy effects of antecedent water-column conditions on current ecosystem function. For example, biogeochemical cycling under ice can substantially alter the abundance and chemical form of macronutrients in spring (Kratz et al. 1987; Hampton et al. 2017) and, in the case of carbon (C), substantially increase CO2 concentrations under ice (Kratz et al. 1987; Finlay et al. 2015). Additionally, although spring CO2 flux has been shown to contribute significantly to total annual CO2 flux in many lakes (Maberly 1996; Striegl et al. 2001; Ducharme-Riel et al. 2015), relatively few measurements of pCO2 are available for shoulder seasons of summer, owing to logistical issues related to sampling during ice melt and formation. Given that lake pCO2 is frequently elevated in spring and fall seasons relative to summer (Baehr and DeGrandpre 2002; Denfeld et al. 2015), it is important to better understand the magnitude and drivers of seasonal contributions to annual CO2 fluxes to improve estimates of the role lakes in the global C cycle. Seasonal variation in water-column pCO2 in boreal lakes frequently follows predictable annual patterns of change in metabolic processes, particularly in ice-covered dimictic systems. In these lakes, CO2 accumulates under ice in winter (Baehr and DeGrandpre 2002; Denfeld et al. 2015), causing a large efflux of CO2 in spring when the ice melts and the water column circulates (overturn). pCO2 levels are reduced in summer when the water column is stable and primary production *Correspondence: [email protected] This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. Additional Supporting Information may be found in the online version of this article. Special Issue: Long-term Perspectives in Aquatic Research. Edited by: Stephanie Hampton, Matthew Church, John Melack and Mark Scheuerell.