Umeå universitets logga

The majority of lakes on Earth are shallow, unproductive and located at high latitudes. These lakes are experiencing big changes due to climate change, where two environmental drivers operate simultaneously, browning and warming. How they affect lake ecosystems is not well understood. Here, I addressed this issue by using a theoretical and an experimental approach. In particular, I generated model predictions and compared them with the results of a realistic large-scale experiment, where browning and warming were manipulated in a factorial design. In addition, model outcomes were compared with data from 12 unproductive lakes sampled along a gradient of browning. Another novelty of my thesis is that it integrates benthic and pelagic food web components in the model and experimental approaches. I found that browning affected the resources availability for benthic and pelagic producers in the model and in the experiment. With browning, benthic primary producers became increasingly light limited and declined, while pelagic producers became less nutrient limited and increased. Pelagic nutrient limitation was alleviated by two non-exclusive mechanisms. Browning directly enriched the water with nutrients, and browning indirectly increased the nutrient flowing from the sediment to the pelagic habitat via suppression of benthic producers. To tease apart these two mechanisms I applied structural equation modeling (SEM). The indirect evidence by SEM suggests that both mechanisms contributed equally to the pelagic nutrient concentration in the experiment. Interestingly, a model food web with only primary producers shows similar qualitative behavior as a food web with grazers and carnivores included. This happens because carnivorous fish exert strong top-down control in the more productive habitat, which relaxes grazing pressure on primary producers and increases resource limitation in the adjacent habitat. Biomass of benthic and pelagic consumers followed the same pattern as their resources. The lake data were largely congruent with model expectations and supported the findings of the experiment. Furthermore, the model also predicted a negative relationship between total phosphorus and both primary and fish production, which was observed across the 12 lakes. Warming effects were more complex. The model predicts that warming effects should depend on browning and are expected to be strongest in the more productive of the two (benthic and pelagic) habitats. For example, at low levels of browning the biomasses of benthic algae and fish are expected to decline with warming, which was observed in the experiment. In contrast, observed warming effects at high levels of browning deviated from model expectations. The mechanisms by which browning and warming interactively affect lake food webs are still poorly understood. This thesis offers a conceptual foundation for their further study through the integration of within- and between-habitat interactions.