Climate warming is affecting nutrient cycling in Arctic lake ecosystems, impacting global carbon emissions. Indeed, Arctic lakes are significant carbon sinks, where low temperatures and nutrient scarcity limit carbon release both from the soil and from waters. However, lakes are shrinking at approximately 1500-2000 Km² annually due to warming, while nutrient levels are rising due to reduced snow cover and growing herbivore populations. This study aims to elucidate the connections between lake drying, nutrient inputs, and greenhouse gas emissions, which is essential for accurate predictions and effective mitigation strategies.

We designed a space-for-time experiment using stable isotopes, in-situ gas-flow measurements and remote sensing to understand the mechanisms linking changes in snow cover with terrestrial productivity, N inputs by migratory geese, CH₄, CO₂, N₂O, and H₂O emissions across nine Arctic lake ecosystems (Svalbard Islands). At each lake, measurements in lake waters, inundated, and dry soils simulated lake drying effects.

In catchments where summer snowmelt occurred earlier, simulating a warmer climate scenario, grass productivity and geese abundance were higher. This led to increased N concentrations both in soil and in water, promoting carbon emissions and turning lakes from carbon sinks into sources. CH₄ emissions decreased while CO₂ emissions rose from lake waters to dry soil. On average, CO₂ equivalent emissions from dry soil were 19.2 g m^-2 s^-1 higher than from lake waters, with differences intensifying with N soil inputs by geese. Data indicate that the estimated economic impact of losing 1 km² of Arctic lake surface is nearly $32,000 annually due to rising carbon emissions. Our findings reveal positive feedback mechanisms linking climate warming, snow, herbivores, and greenhouse gas emissions, highlighting the ecological and economic implications of Arctic lake loss and enhancing Arctic-climate interaction predictions.