Ecologists try to understand how changing habitats alter the populations of organisms living within them, and how, in turn, these changing populations alter the environment. By linking individual or cellular (physiological) processes to system level responses, mechanistic models can help describe the feedback loops between organisms and the environment. Aquatic systems have long used mechanistic models, but increasing model complexity over the last 50 years has led to difficulty in parameterization. In fact, it is often unclear how researchers are choosing parameters at all, even though small changes in parameters can change qualitative predictions. I explore the challenges in parameter estimation present in even an ideal situation. Specifically, I conduct individual experiments for all of the needed parameters to describe a simple lab-based, aquatic system; estimate those parameters using the results from these experiments supplemented with literature data; and run a large experiment designed to test how well the lab-estimated parameters predict actual zooplankton populations and nutrient changes over time. I document best practices for finding and reporting parameter choices and show whole ecosystem level consequences of a variety of decisions. To get the best predictions I find that a mix of parameter estimation methods are necessary. Trait-based approaches are another method to understand species-environment interactions. Trait-based methods aggregate species into functional traits, perhaps making qualitative predictions easier. Theory suggests that more functionally diverse systems will be more resilient. I test this prediction in a simple aquatic system but am unable to find consistent support for this hypothesis, and instead finding that results are highly dependent on what measures of ecosystem recovery are used. Overall, more species-specific information is critical to building better models for both mechanistic and trait-based approaches. I expand species-specific data by providing new information, and collating information from literature on a small, tropical Cladocera. / Thesis / Doctor of Philosophy (PhD) / Predicting what will happen to a habitat after a disturbance is critical for conservation and management. Species specific information is useful for building a mechanistic understanding of ecology. Predictions that include underlying processes (mechanisms) may be more robust to a changing environment than predictions based on correlations. Eutrophication, the addition of excess nutrients, is a common problem in freshwater habitats. Being able to predict the effects of nutrient addition is critical for ensuring the health of freshwater ecosystems. By using species-specific life history and morphological information and a simple lab system, I test different methods of predicting and understanding the consequences of eutrophication. I find that the ramifications of eutrophication are not easily predicted by species' categorizations or with a more detailed mechanistic model.
| Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/25842 |
| Date | January 2020 |
| Creators | Werba, Jo |
| Contributors | Kolasa, Jurek, Biology |
| Source Sets | McMaster University |
| Language | English |
| Detected Language | English |
| Type | Thesis |
Page generated in 0.0023 seconds