TWIG PRODUCTION PATTERNS AMONG MOOSE FORAGE SPECIES AND IMPLICATIONS FOR FOREST MANAGEMENT

Abstract

A declining moose (Alces alces) population in northeastern Minnesota is a serious concern of the Gichi Onigaming Anishinaabeg (Grand Portage Band of Lake Superior Chippewa). A better understanding of the variation in moose forage production at the plant level among tree and shrub species and over time may be important to moose recovery in areas where nutrition may be limiting, while managers need practical methods to monitor forage density. Using data from an extensive 2019 moose browse survey at Gichi Onigaming (Grand Portage Indian Reservation, MN) and Minong (Isle Royale, MI), we fit models of twig production (number of twigs per plant) within moose reach as a function of species, canopy cover, and stem height. We validated the best fit model against an independent data set and used the model to simulate stand-level forage density over time for 3 tree species preferred by moose. Twig production varied non-linearly with stem height. An increasing then decreasing unimodal curve with height fit better than allometric models or species means. Peak twig production, height at peak production, and rate of production decline with height varied among species. Paper birch (Betula papyrifera) and balsam fir (Abies balsamea) generally had the greatest peak twig production, greatest heights at peak production, and lowest rates of decline with height. Among trees, quaking aspen (Populus tremuloides) generally had lower peak twig production, lower height at peak production, and greater rates of decline with height than most other trees and some tall shrubs. Peak twig production was greater under open canopy than under closed canopy for seven species. Model validation indicated that predictions were well correlated with observations and outperformed alternative models but had consistent over-prediction bias. Simulations of regenerating aspen, paper birch, and red maple (Acer rubrum) forests indicated that whereas aspen produced ~1.4-2.2 times more peak forage biomass than paper birch or red maple, paper birch and red maple produced usable forage densities for ~ 2-17 years longer than aspen. Finally, we provide a procedure to use the regression equations to estimate moose forage density from common forest regeneration survey data. Although the equations are suitable for monitoring, they should be used cautiously in high accuracy applications. Our findings suggest that diverse mixes of deciduous trees and shrubs resulting from post-harvest treatments likely provide abundant moose forage for longer durations than do nearly pure aspen stands often resulting from clearcutting alone in northeastern Minnesota.