Lakes are classically viewed as discrete ecosystems bounded on all sides by land. However, a narrow focus on lakes as discrete units is incompatible with the scale of many management programs and ignores the placement of lakes relative to their larger ecological context. While it is clear that lakes are not isolated units but are instead embedded components of lake-river networks and have a broader landscape (i.e. regional) context, it is not always clear how this embedding is borne out quantitatively. For example, is the position of a lake in a multi-lake network a dominant predictor of nutrient retention? Or, how strongly does the arrangement of streams and near-stream land-use (i.e. aquatic-terrestrial linkages) affect lake nutrient concentrations? To date, we have been unable to quantitatively address these questions in a synthetic manner because the necessary data has not previously been available for many lakes over large geographic extents (i.e. the macroscale). As a result, prior research has mostly been conducted on single lakes or in some cases groups of nearby lakes in a single region. In each of the following chapters, I developed macroscale lake databases to examine hundreds to thousands of lakes across diverse local and regional settings and used these databases to investigate the roles of both local and regional processes and aquatic-terrestrial linkages in determining lake nutrient retention, nutrient concentrations, and basic lake morphometry.In my first chapter, I show that throughout Northeastern and Midwest US, lakes with higher connectivity have lower nutrient retention but this "connectivity effect" is apparent at the scale of entire lake networks rather than more localized lake subwatersheds. My findings suggest that a broader whole-network perspective is likely to be more effective than a narrow lake-specific perspective in regulatory frameworks focused on eutrophication. In my second chapter, I show that lake nutrient concentrations are related to a variety of agricultural activity measures beyond the percent of the watershed in agricultural land use. I show that when one measure in particular, the percentage of agricultural land use in near-stream areas, is elevated, this signals a high likelihood of elevated lake nutrient concentrations. I further show that lake total phosphorus concentrations have different relationships with measures of agricultural activity compared to lake total nitrogen concentrations. My findings suggest that differences in lake nutrient sensitivity to agricultural activity may affect the outcome of policies to enhance water quality depending on whether they focus on lake N or P. In my third and final chapter, I test the utility of geometric models for predicting lake depth for lakes with missing depth information. Using bathymetric data for 5,000 lakes, I show that one of the assumptions of such models, that slope proxies of the surrounding land are representative of true in-lake slope, is not supported by available evidence. This lack of relationship has implications for predictive accuracy and model bias in lakes of different types or shapes. My findings specifically suggest that geometric models are likely to overestimate the depth of lakes with concave cross-section shapes and those classified as reservoirs. Across all chapters, I use the frameworks of landscape limnology and macrosystems ecology to explore the relationships between watershed and lake characteristics. I show that such macroscale analyses, which explicitly consider hierarchy in the freshwater landscape, provide a nuanced understanding of controls on lake nutrients and morphometry across a broad array of lakes.
Read
