The carbon budget induced by water-level fluctuation in a typical shallow lake

Shallow lakes are characterized by strong water-level fluctuations, which control shifts in carbon (C) sources and sinks. However, there remains uncertainty in elucidating the alteration in the carbon budget with water-level changes. Here, carbon budget variation with water level was estimated by between carbon sequestration (C_s) and carbon emission (C_e) (including carbon dioxide, methane) using an improved logistic model and the DeNitrification-DeComposition (DNDC) model in Baiyangdian Lake (BYD), a typical shallow lake. According to the results, the C_s and C_e were 5.33 Tg/a and 1.72 Tg/a, respectively, which indicated that BYD acted as a carbon source (3.60 Tg/a net carbon emission). The spatial pattern and high value areas (HVA) of the carbon budget showed significant change with water depth variation under water level fluctuation. The HVA of C_s, C_e and C_net in the BYD was distributed at water depths of -0.24±0.77 m, 2.41±0.50 m and 2.53±0.26 m, respectively. At the lake scale, increasing water level will increase the unevenness of the water depth contribution of C_s, with a higher variation coefficient of the contribution structure, while the impact on the C_e contribution is not significant. Furthermore, elevating the water level increases the contribution of C_s in HVA to the whole lake, while the effect of water level on C_e is characterized by segmentation. In combination with the goal of increasing sinks and reducing emissions and the feasibility of management operation, the water level threshold based on the net carbon emissions reduction can be more accurately identified with vegetation coverage demand.