Regional aquatic biogeochemistry of the Northern Highlands Lake District

The movement of water across the landscape links streams, lakes and wetlands together into cohesive networks or hydroscapes. While lake, stream, or wetland ecosystems have often been the focus of a wide range of research efforts, there is much to be learned about how hydroscapes influence biogeochemical patterns and processes. The overarching goal of this dissertation is to quantify biogeochemical characteristics of hydroscapes in the water-rich landscape of Wisconsin's Northern Highland Lake District (NHLD) and understand how hydroscapes characteristics influence regional aquatic biogeochemical patterns. I address these broad goals by exploring (1) the capacity of different types of information to predict lake characteristics, (2) what insights can be gained regarding processes influencing regional biogeochemistry by integrating lake and stream biogeochemical surveys, and (3) how characteristics of hydroscapes influence regional dissolved organic carbon (DOC) patterns in streams.;This research demonstrates that information from neighboring lakes is more effective at reducing prediction errors of a lake's characteristics than the lake's history (Chapter 1). Thus, having continuous records in a select set of systems might allow the extrapolation of values in adjacent systems without the need for continuous data in all systems. However, several years of data are needed to consistently predict annual characteristic and, as a result, long-term observation is unavoidable. Comparisons of regional lake and stream biogeochemical surveys emphasize clear differences between lake and stream biogeochemistry, indicating that biases can emerge in large-scale studies that use only one type of ecosystem (Chapter 2). This research further demonstrates that integrating lake and stream information can lead to insights about regional aquatic biogeochemistry that would not have been apparent from considering a single type of ecosystem. Finally, I demonstrate that wetland extent is the best predictor of DOC in streams, and I was unable to detect an effect of upstream lakes on streamwater DOC in the NHLD (Chapter 3). This result confirms the importance of wetlands as a primary driver of stream DOC, but I suggest that the relative importance of lakes within stream networks may vary among regions depending on wetland extent.