The natural 'exclosure effect' and tree regeneration following post-windstorm salvage logging

Understanding the influence of post-disturbance forest management on tree regeneration is critical for assessing ecosystem recovery and guiding future responses. In particular, the influx of elevated coarse woody material (CWM) following wind disturbance, if left in situ, may impede herbivore access, thereby protecting saplings from browsing damage through a natural 'exclosure effect'.In 2013, a tornado in northcentral Maine, United States and subsequent salvage logging operations created three clear 'treatments' for evaluation of the exclosure effect: blowdown, blowdown plus salvage logging and an undamaged control. Nine years post-tornado, we inventoried tree regeneration within these treatments to evaluate differences in sapling abundance, species composition, size structure and browsing intensity. We also inventoried CWM, including the height above forest floor.Results revealed significant differences in sapling composition and browsing intensity among treatments. The salvage treatment had the highest proportion of browsed saplings (56 +/- 28%; mean +/- standard error), followed by the control (9 +/- 10%) and blowdown (5 +/- 8%). Blowdown had by far the greatest mean (50 +/- 9 cm) and average maximum (169 +/- 43 cm) heights for CWM. Binomial generalized linear models revealed that browsing probability was a function of mean CWM height and an interaction between sapling density and proportion of sapling hardwoods. Thus, browsing damage was less likely in plots with greater CWM heights and more likely in plots with greater sapling density and more hardwood saplings.Synthesis and applications. This study furthers our understanding of ecosystem recovery following blowdown and salvage logging. Results suggest that salvage logging created important differences in coarse woody material (CWM) abundance and height distribution, when compared to un-salvaged areas, and that these differences in turn altered sapling size structure and browsing intensity. These findings highlight the potential long-term effects of successive disturbances, as the differences evident in these early stages may persist for decades or longer. Importantly, we provide evidence of the exclosure effect, suggesting that CWM retained in the un-salvaged area protected saplings from moose browsing. Thus, in post-disturbance areas where browsing threatens regeneration, we recommend that managers consider retaining CWM to serve as a natural exclosure.