Evaluating how spatial heterogeneity in forage chemistry and abundance influences diet and demographics in a declining moose (Alces alces) population in northeast Minnesota.

University of Minnesota Ph.D. dissertation.May 2019. Major: Conservation Biology. Advisor: James Forester. 1 computer file (PDF); xii, 180 pages.Since 2006, the moose population in northeastern Minnesota has declined by nearly 50%. While recent warming has been implicated as a primary cause of this decline, there is little evidence to support this relationship. More recent evidence suggests that the influences of warming and poor nutrition may predispose moose to increased risk of mortality, including mortality attributed to predation and disease. During summer, moose begin to experience the detrimental effects of high temperatures at around 14 to 17 ºC, and mean-maximum summer temperatures in this area range from 19.5 ºC along Lake Superior, to approximately 24.5 ºC in the more central part of the region. Thus, moose in northeastern Minnesota are likely dealing with the negative effects of high temperatures on a routine basis throughout summer. While it has been suggested that nutrition and warming may be acting in concert to influence moose demographics in Minnesota, potential synergisms between these factors have not been investigated. Thus, I evaluated how spatial variation in the thermal landscape influences forage chemistry, the abundance and distribution of forage, and the spatial variation in moose diets and overwinter survival. To determine how high temperatures might influence the chemistry of moose forage, I used untargeted metabolomics to evaluate how varying combinations of temperature, moisture, and light in both experimental and natural conditions influence the production of plant secondary metabolites in moose forage. To investigate how the abundance and chemistry of moose forage varies across NEMN, I used a mixed-effects regression kriging framework to estimate spatial variation of δ13C and δ15N values in plants commonly eaten by moose, and then refined these predictions using species-specific allometric equations to estimate above-ground biomass of moose forage. Finally, to investigate the interaction between spatial variation in high summer temperatures, moose diet, and over-winter survival, I used stable isotope values from forage and hair to estimate moose diet via Bayesian mixing models, and then evaluated if diet composition and quality vary as a function of mean-maximum summer temperature, season, or winter mortality. In general, warming and high-temperatures had variable effects on the defensive chemistry of moose forage, only a minor influence on forage abundance, and a strong effect on diet quality, composition, and overwinter survival. Specifically, when investigating the influences of warming on PSM production in moose forage, I found that the influences of temperature can be modulated by the presence or absence of other abiotic factors, such as precipitation and light. As an example, the relative abundance of compounds known to negatively influence moose herbivory increased by 250% or more when high temperatures occurred in an open canopy setting. When modeling spatial heterogeneity in the chemistry and abundance of moose forage across northeastern Minnesota, I found that while mean-maximum summer temperature played a strong role in the isotopic composition of moose forage across the region, it had only a minor effect on distribution and abundance. Finally, when investigating interactions between spatial variation in high summer temperatures, moose diet, and over-winter survival, I found that the warmest parts of the moose range in Minnesota were those where moose diets were poorest and where winter mortality rates were highest. Specifically, I found that moose in the warmest parts of the range have diets containing the highest proportion of aquatic forage and the lowest proportion of high-preference forage. Additionally, moose that did not survive winter had diets containing substantially greater proportions of aquatic forage throughout the entire growing season when compared to moose that survived, which consumed mostly high-preference forage during early summer but increased their consumption of aquatics during late summer. Finally, while I estimated overall mortality to be at approximately 30% throughout the entire study region, mortality in the warmest parts of the range (69%) was approximately 4.5 times higher than that in the coolest parts of the range (15%). Given the evidence I present here, habitat-improvement projects may want to focus on promoting the regeneration of forage species that can adapt to future warming scenarios, while still providing thermal refuge, and proper nutrition during late summer. Also, future studies should evaluate spatially explicit differences in habitat use as a function of the thermal landscape and how variation in habitat use-behavior (i.e., movement) within the thermal landscape may influence diet composition, quality, and nutritional restriction. Identifying mechanistic links between movement, diet, and nutritional condition within the thermal landscape would advance our basic knowledge of large mammal behavior and ecology, as well as help develop sound management strategies in how we plan for future warming