Umeå universitets logga

Lake ecosystems receive, transmit and process terrestrial carbon and thereby link terrestrial, aquatic and global carbon cycles. Most lakes evade CO2 to the atmosphere, but the annual magnitude of CO2 evasion, as well as sources and mechanisms underpinning CO2 evasion from lakes are still largely unresolved. CO2 evasion from lakes can be sourced from direct external input from the catchment, but CO2 can also be produced in-lake from organic carbon breakdown. Both sources have been shown to be of importance to individual systems, but a landscape perspective is still missing. Globally, most lakes are in northern high latitudes, but due to infrequent seasonal sampling the magnitude of CO2 evasion on an annual scale is largely unknown, as are constraining variables of in-lake metabolism (i.e. production and consumption of CO2). As a consequence of these knowledge gaps, there is little possibility to predict future lake carbon cycling, for instance due to changing dissolved organic carbon (DOC) input or lake temperature resulting from global warming.

In this thesis I aim to resolve these knowledge gaps surrounding the magnitude, cycling and sources of CO2 evasion from high-latitude (mainly arctic) lakes. By combining the estimates of annual CO2 evasion and metabolism, I investigated the magnitude of CO2 evasion, as well as the contribution of the internal carbon processing to CO2 evasion. Inclusion of ice-melt evasion allows to assess the importance, and drivers, of ice-melt CO2 evasion on the annual scale. Furthermore, by pooling lakes from multiple different lake surveys I was able to analyse the lake and landscape variables associated with high-latitude lake metabolism. Finally, through use of an experimental pond facility I manipulated dissolved organic carbon input and temperature to explore the effects of future climate conditions on lake carbon cycling and CO2 evasion.

I found that both external input and internal CO2 production can contribute to CO2 evasion from lakes, but it is often dominated (>75%) by a single source and forest cover increased the amount to which the internal source contributed to annual CO2 evasion. I also found that the concentration of DOC in the lakes was inversely correlated to the proportion of CO2 lost at ice-melt. As a result, the ice-melt season is of significant importance to the annual CO2 evasion from low DOC high-latitude lakes, and omission can underestimate the magnitude of annual CO2-evasion by ~50%. Metabolism in these types of clear-water, low nutrient systems is dominated by benthic (on the sediment) production. Consequently, in-lake metabolism in these high-latitude clear-water lakes is largely constrained by lake depth and basin shape, and the potential for ice-scouring to disturb the benthic system in littoral areas. Convex lakes with predominantly shallow sediments were thus less productive compared to concave lakes where benthic production is less affected by ice-scouring. Finally, increasing DOC inputs (e.g. as a result of changes in climatic conditions) positively related to the amount of CO2 produced within and evaded from the lakes. However, warming was found to decrease in-lake CO2 production and evasion, potentially via increased nutrient limitation of carbon mineralization (i.e. more energy expanded for nutrient uptake in order to break down organic carbon in warmer water), and changes in community structure (e.g. different macrophytes). This thesis thus clearly outlines the annual magnitude (both open water and the specific importance of ice-melt), source contribution (quantified for many lakes rather than single systems) as well as the lake and landscape factors of note to source contribution (i.e. forest cover and DOC input increased internal cycling, especially in shallow and concave systems). Taken together, the results advance understanding the mechanisms behind cycling and evasion of CO2 in earth’s most common lake type.