Peromyscus responses to alternative forest management systems in the Missouri Ozarks, USA

Abstract

Operational-scale forest management experiments are long-term investments because harvest treatment effects may be dynamic throughout one or more rotation lengths. We examined deermouse (Peromyscus spp.) abundance over the first 20 years of the Missouri Ozark Forest Ecosystem Project (MOFEP), which assesses ecological responses to even-aged, uneven-aged, and no-harvest forest management systems applied at landscape scales in the Missouri Ozarks. In the spring of each of 11 sampling years, we sampled Peromyscus populations on two permanent trap grids on each of nine study sites (n = 3 sites per management system). Management entries occurred in 1996 and 2011, with small mammal sampling conducted during two pre-treatment years (1994–1995), and Years 2–5 (1998–2001), Years 13–14 (2009–2010), and Years 16–18 (2012–2014) after the first entry. We estimated abundance for each grid in each sampling year with Bayesian closed-population mark-recapture models, and modeled variation in abundance with negative binomial log-linear mixed effects models. Uneven- and even-aged management systems caused similar increases in Peromyscus abundance that were detectable shortly after the first management entry [proportional effect of even-aged management on Peromyscus abundance vs. no-harvest management: posterior median = 1.8, 95% credible interval 1.0–3.2; proportional effect of uneven-aged vs. no harvest management = 1.7 (1.0, 2.8)]. These effects were not surprising given positive effects of harvest treatments on understory cover and food resources. However, the consistency of this increase was less expected, as we observed no conclusive dissipation of harvest effects even in Years 13–14 after the first entry or amplification of harvest effects after the second entry. Observing extremely high system-wide yearly variation in Peromyscus abundance, we did not detect evidence of increasing divergence in effects of these three management systems or of any area-wide trends in abundance during 1994–2014. However, over subsequent decades, we expect higher potential for divergence in Peromyscus abundance as the three management systems differentially shape forest structure and tree species composition. Thus, the MOFEP study offers a unique framework for building and testing hypotheses about patterns and mechanisms of long-term changes in Ozark forests and effects on vertebrate communities. 3 1994–2014. However, over subsequent decades, we expect higher potential for divergence in 44 Peromyscus abundance as the three management systems differentially shape forest structure and 45 tree species composition. Thus, the MOFEP study offers a unique framework for building and 46 testing hypotheses about patterns and mechanisms of long-term changes in Ozark forests and 47 effects on vertebrate communities.