Using oxygen stable isotopes to quantify ecosystem metabolism in northern lakes

Abstract

In remote regions of the world, whole lake metabolic estimates are scarce, largely because long incubations, intensive sampling and deployment of monitoring equipment are impractical. The oxygen isotope (δ18O) mass balance approach represents a simple and efficient alternative to measure whole-lake gross primary production (GPP) and respiration (R) from a single point sample, yet this option has not been extensively explored in habitats such as remote northern lakes. Here, we explored the applicability of the method using a sensitivity analysis on simulated data, showing that in large, heterotrophic (i.e., R > GPP) lakes, model outputs are sensitive to input terms for isotopic fractionation and air–water gas exchange. Despite these sensitivities, field applications of the δ18O method generated promising results that were generally consistent with parallel, free-water diel DO metabolic modelling, but greater than in vitro incubation measurements. The isotopic approach captured both wide-ranging metabolic conditions in in situ experimental mesocosms, and the seasonal trends in GPP and R in a shallow, dystrophic lake. In a clearer, deeper heterotrophic lake, the isotope approach integrated a fraction of metalimnetic metabolism missed by diel DO metabolic estimates. Overall, metalimnetic contributions to surface δ18O–DO dynamics had the greatest impact on model outputs, but with accurate information on air–water gas exchange, mixing depth, and the vertical DO and light regime of a given system, these effects can be accounted for and the isotopic approach can yield well constrained, spatio-temporally integrated rates of GPP and R. The approach is clearly suitable for use in oligo- and mesotrophic lakes, especially in remote regions where sampling is logistically difficult.