Examining lakes at multiple spatial scales: Predicting fish growth, macrophyte cover and lake physio-chemical variables.

摘要

My research examines lake ecosystems focusing on two main topics: the role of landscapes and spatial scales for predicting lake characteristics, and the role of macrophytes in lake ecosystems. I used a combination of approaches to answer the following questions: (1) Within lakes, how is macrophyte cover related to fish growth in lakes? (2) Can landscape features, in addition to within-lake characteristics (lake morphometry and physio-chemical variables) predict macrophyte cover in lakes? and (3) What is the relative ability of regional landscape features, such as geology and land use/cover, and local landscape features, such as lake morphometry and hydrology, for predicting lake physio-chemical variables? To answer the first question, I performed a field test of the relationship between macrophytes and fish growth for two common fish species using survey data from 45 thermally stratified north-temperate lakes. Although theory and experimental evidence support the hypothesis of an optimal intermediate macrophyte density for fish foraging and growth, I found little evidence to support this idea. However, I did find that growth for some ages of both species was linearly related to some of the macrophyte metrics. To answer the second question of what lake and landscape variables can predict macrophyte cover, I performed a field test using survey data from 54 stratified north-temperate lakes. I found that macrophyte cover metrics were best predicted by some combination of at least two physio-chemical, morphometric and landscape predictors. Finally, using a digital landscape and lake database of approximately 500 Michigan lakes, I examined the relative ability of regional and local landscape features to predict lake physio-chemical variables. Because these scales are hierarchical, such that broad-scale landscape features constrain the occurrence of local lake features, I organized lakes using a multi-scale hierarchical framework, and used hierarchical linear modeling, which is a multivariate approach that accounts for the non-independence of lakes within regions and partitions variance into variance components at each spatial scale. I found that lakes can best be grouped at regional scales using fine-scaled subecoregions or major river watersheds, that both regional and local spatial scales are important for understanding variability in lake physio-chemical variables, and that different regional and local features are correlated to different lake physio-chemical variables. (Abstract shortened by UMI.)