Lakes and ponds are important natural sources of the potent greenhouse gas methane (CH₄), with small shallow waters identified as particular hotspots^1,2. Ebullition (bubbles) of CH₄ makes up a large proportion of total CH₄ flux^3,4. However, difficulty measuring such episodic events⁵ makes prediction of how ebullition responds to nutrient enrichment and rising temperatures challenging. Here, the world’s longest running, mesocosm-based, shallow lake climate change experiment was used to investigate how the combination of warming and eutrophication (that is, nutrient enrichment) affects CH₄ ebullition. Eutrophication without heating increased the relative contribution of ebullition from 51% to 75%. More strikingly the combination of nutrient enrichment and experimental warming treatments of +2–3 °C and +4–5 °C had a synergistic effect, increasing mean annual ebullition by at least 1900 mg CH₄-C m⁻² yr⁻¹. In contrast, diffusive flux showed no response to eutrophication and only a small increase at higher temperatures (average 63 mg CH₄–C m⁻² yr⁻¹). As shallow lakes are the most common lake type globally, abundant in highly climate sensitive regions⁶ and most vulnerable to eutrophication, these results suggest their current and future contributions to atmospheric CH₄ concentrations may be significantly underestimated.

DOI: 10.1038/s41558-017-0063-z