Effects of amur honeysuckle (Lonicera maackii [rupr.] herder) invasion and removal on native vegetation and white-footed mice (Peromyscus leucopus) in mixed-hardwood forests of Indiana.

摘要

The threat of non-native invasive species continues to compromise the ecological and economic integrity of our natural resources. Numerous investigators have documented the negative effects of invasive species on native biota. However, much work is still needed with regard to how invasive species spread in space and time, factors contributing to impacts on native biota within invaded ecosystems, and resultant effects of removing invasive species. In terms of invasion patterns, few studies have documented local patterns and rates of woody plant invasions, and even less is known about changes in spatial patterning and factors influencing structural characteristics (diameter and height) of individuals as invasion phase progresses from establishment to saturation. Furthermore, little is known about how the duration of occupation by an invasive species, along with the amount of growing space it occupies, influences native biota within invaded areas (i.e., microsite-level impacts). Finally, few investigators have examined the effects of removing invasive species on native flora and fauna. We (my graduate committee and I) examined the spread and ecological effects of Amur honeysuckle (Lonicera maackii [Rupr.] Herder), a non-native, invasive shrub that has invaded many ecosystems throughout eastern North America, at 12 mixed-hardwood forests in central Indiana representing a range of invasion intensity and overstory composition. At three of the mixed-hardwood forests we examined age distributions and spatial patterns of Amur honeysuckle invasions, and identified factors influencing life-stage characteristics of individual shrubs. Predicted age distributions indicated that Amur honeysuckle reached the exponential phase of invasion at ~10-15 years. Inhomogeneous L and cross- L functions indicated that Amur honeysuckle exhibited a clustered spatial pattern; immature individuals (no berries) clustered around mature individuals (with berries). However, spatial relationships between honeysuckle and trees rarely exhibited a clustered pattern. Regression analyses with Amur honeysuckle diameter and height as response variables revealed that incorporating spatial autocorrelation provided a better model fit than models where spatial autocorrelation was not considered and caused otherwise significant predictor variables to become non-significant (p ≥ 0.05). Our results suggest that local-scale invasion by this species follows a predictable temporal sequence of population establishment and expansion via neighborhood diffusion and the forest-scale distribution of nascent foci. Furthermore, our results highlight the importance of considering spatial autocorrelation when evaluating life-stage characteristics of invasive populations. At all 12 mixed-hardwood forests we examined the influence of density, percent cover, and duration of Amur honeysuckle (i.e., time since establishment), as well as other environmental factors, on native plant taxa. Overall, study sites with the highest taxonomic diversity (H'), richness (S), percent covers, and densities of native vegetation also had the lowest percent covers of Amur honeysuckle in the upper vertical stratum (1.01-5 m). Based on linear mixed model analyses (random effect = study site), percent cover of Amur honeysuckle in the upper vertical stratum was consistently and negatively correlated with H', S, total percent cover, and woody seedling densities of native taxa at the microsite scale (mixed model p values < 0.05). While duration of Amur honeysuckle occupation at the microsite scale was not significant when percent cover of Amur honeysuckle in the upper vertical stratum was included in models, duration was significantly correlated with dependent variables and with upper-stratum honeysuckle cover, suggesting that greater Amur honeysuckle age at the microsite scale results in higher light competition from above for native ground flora species. At six of the mixed-hardwood forests, we examined the short-term effects of removing Amur honeysuckle and other non-native shrubs on native herbaceous plants and woody seedlings, as well as white-footed mice( Peromyscus leucopus). Each study site contained two 80 m x 80 m sample areas (removal area where all non-native shrubs were removed and a reference area where no treatment was implemented). Native and non-native vegetation was sampled in the spring and summer of 2010 (before removing non-native shrubs) and again in the spring and summer of 2011 (after removals). Percent cover and diversity of native species and seedling densities of native woody species increased after shrub removal (permutation p values ≤ 0.10). Conversely, changes in reference areas were typically much lower and often non-significant. However, we also observed significant increases in Amur honeysuckle seedling densities and the percent cover of garlic mustard (Alliaria petiolata [M. Bieb.] Cavara & Grande) in removal areas. Our results suggest that removing Amur honeysuckle and other non-native shrubs allows the short-term recovery of native plant taxa across a range of invasion intensities. However, long-term recovery of native flora will likely depend on renewed competition with invasive species that re-colonize treatment areas, the influence of herbivores, and subsequent control efforts implemented by forest managers. To examine effects on white-footed mice, 49 live traps were placed in each of the two, 80 m x 80 m areas (reference and removal areas). The number of white-footed mice was recorded using mark-release-recapture (MRR). Trapping was done for six nights in the summer of 2010 and four nights in the fall of 2010 (before exotic shrub removals), and again during the summer and fall of 2011 (after removals). For each 49-trap grid, we assumed population closure, and abundance was estimated using Bayesian parameter-expanded data augmentation, with time and individual heterogeneity (model Mth). For each season (summer or fall) and each grid type (removal or reference), we calculated differences in abundance by subtracting estimates in 2010 from estimates in 2011. Permutation tests (assuming a paired design) were then used to test whether mean differences were significantly different from zero. For both trapping seasons, mean abundance increased from 2010 to 2011 (i.e., positive differences) in both removal and reference areas, but the magnitude of increase within removal areas was substantially greater (permutation p < 0.05 for removal areas). For the feasible subset of mice, we calculated mean squared distance (MSD) as an index of space use. Linear regression was then used to determine how environmental variables influenced space use by individuals. For mice captured in the summer, percent cover of leaf litter (p = 0.004) was the only significant predictor of MSD, whereas canopy cover (p = 0.001) and abundance (p = 0.003) were negatively correlated with MSD for mice captured in the fall. Our results suggest that management efforts to control the spread of Amur honeysuckle and other exotic shrubs may lead to short-term increases in the abundance of generalist rodents such as white-footed mice. Furthermore, factors such as leaf litter cover, canopy cover, and population-level abundance may influence space use by individual mice within invaded habitats.