Evaluation and Improvement of the Stoichiometric Homeostasis Model for Understanding and Predicting the Structure and Functioning of a Low Arctic Tundra Plant Community

Abstract

Arctic tundra vegetation structure profoundly affects ecosystem processes and climatic regulation. Therefore, there is a growing urgency for more accurate biogeochemical models to better understand how tundra vegetation will respond to ongoing major environmental changes, including rising temperatures, enhanced soil nutrient availability, and increased snowfall. The stoichiometric homeostasis model (the H model) - which quantifies the ability of an organism to maintain its internal elemental concentrations despite variation in the availabilities of these elements as resources - may provide valuable insights into vegetation-environment feedbacks, but its application in the Arctic has not previously been investigated. In this thesis, I used a two-step approach to first evaluate and then improve the H model for understanding the structure and functioning of a low Arctic plant community. First, I tested the applicability of the H model in the tundra context. Second, I improved the accuracy for determining the homeostatic H values by comparing two methods for estimating soil nutrient availabilities (ion exchange membrane (IEM) incubation versus soil sample extraction methods). Third, I examined the applicability of the H model in predicting key aspects of species’ ecological performances under various experimental manipulation conditions over two successive 6-year periods at both the individual species and community levels. The IEM method was superior to the extraction method in providing biologically meaningful soil nutrient data for modeling the H values. The H indicator based on nitrogen (N):phosphorus (P) ratio (HN:P) was a more robust homeostatic indicator than those based on either single element (HN or HP). Furthermore, variation in plant P rather than N stoichiometry drove much of the differences in HN:P. In terms of functioning, tundra species with relatively high HN:P values were more dominant, more temporally stable, and less responsive to the effects of environmental change on soil available P. Finally, communities with higher HN:P values also had more aboveground biomass. Together, these results highlight the value of using the combination of both N and P for a more complete understanding of homeostatic regulation, and validate the potential of using H values to predict tundra plant community structure and functioning in a rapidly changing environment.