Testing for Unimodal Relationships between Diversity, Disturbance, and Productivity
Date
2020-01-30
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Many unimodal ecological theories, though implemented into resource management, remain contentious in the literature and are often analyzed with inappropriate techniques, namely quadratic regressions. Two of these theories, the intermediate disturbance and productivity hypotheses, are assessed as case studies to explore the hypotheses themselves, characterize the conditions in which various relationship shapes take place, and underscore the disparities between the analytical techniques that can be used to detect them. The power of various traditional and novel techniques used to distinguish between unimodal versus monotonic relationships is simulated to conclusively determine which are the most appropriate. The two-lines test has been found to display the most powerful detection of unimodal shapes and breakpoints regressions to have the most power to detect monotonic, concave shapes. According to these tests, diversity-disturbance and diversity-productivity relationships display primarily peaked shapes, dependent on various characteristics. The quadratic regression, even when paired with the Mitchell-Olds & Shaw test, is an inefficient and inappropriate test for unimodality. While the intermediate disturbance and productivity hypotheses are too simplistic, misused, and misunderstood, the findings within this thesis warrant further exploration into the scenarios in which they occur and can be implemented. Future analyses of unimodal hypotheses should consider the recommendations of technique usage within this thesis.
Description
Keywords
intermediate disturbance hypothesis, intermediate productivity hypothesis, unimodal
Citation
Tessier, R. S. (2020). Testing for Unimodal Relationships between Diversity, Disturbance, and Productivity (Master's thesis, University of Calgary, Calgary, Canada). Retrieved from https://prism.ucalgary.ca.