Forage-selection and dietary quality of moose in northeastern Minnesota

Background/Question/Methods

Since 2006, moose populations in Minnesota have declined by nearly 70%, and recently, the Minnesota Department of Natural Resources proposed that moose be listed as a “species of concern,” due in part to the ongoing threat of climate change. Moose subjected to prolonged heat stress either fast or seek wetland habitats as thermal refuges. These areas provide forage of low caloric value and cannot sustain moose feeding on them exclusively. The extent to which animals cannot meet nutritional demands will be reflected in late-winter body condition. However, to attribute body condition to landscape context, we must first be able to compare forage available in the landscape to that consumed by individual animals. To do this we are using stable isotope analysis and C:N ratios of common forage from northeastern Minnesota across a 5°C summer temperature gradient (separated into warm, moderate, and cold temperature regimes). Additionally, we are comparing forage data to stable isotopes in moose tissues that integrate the isotopic composition of forage over a variety of spatiotemporal scales. C:N ratios serve as an indicator of plant quality, while stable isotopes of C and N allow us to determine where moose are feeding and what they are feeding on. 

Results/Conclusions

Preliminary analyses reveal substantial variation across northeast Minnesota in both summer and winter forage. Corylus cornuta samples (n=47) collected during summer-2012 from 39 plots incorporating three disturbance types (fire, insect defoliation, and clear-cut) and a control are homogenous for δ¹⁵N, but exhibit considerable variation in δ¹³C (-29.1±1.27‰; ANOVA, F_3,44­=4.52, p=0.008) and C:N ratios (ANOVA; F_3,44­=2.31, p=0.008). Insect defoliated areas exhibit distinct isotopic signatures and C:N ratios. Corylus cornuta samples from different temperature regimes exhibit homogenous δ¹³C and δ¹⁵N; however, C:N ratios exhibited considerable variation (ANOVA; F_2,45­=5.18, p=0.009) with significant differences between warm and cold plots. Mean winter δ¹⁵N values from seven different species, including C. cornuta, exhibited no relationship with C:N, whereas mean δ¹³C values decreased with decreasing C:N. Sequential samples along hairs from 10 different moose exhibited low variability in δ¹³C (variance=0.09‰) and indicate an increase in the δ¹³C of diet (suggesting a decrease in dietary quality) from late spring to kill date (fall 2011). Sequential samples showed no clear temporal trend in δ¹⁵N, which varied among individual animals (variance=1.81‰), suggesting that δ¹⁵N might be home range specific.