The relative importance of emerald ash borer-caused tree mortality and abundance of the invasive shrub Amur honeysuckle on tree seedling survival and recruitment in Midwestern deciduous forests

Emerald ash borer (EAB), Agrilus planipennis, may cause extensive changes to forest community composition in North America. Canopy ash (Fraxinus spp.) trees succumb to EAB in 1-4 years, with effects on understory light likely mediated by sub-canopy tree cover. We hypothesize that the effects of ash mortality on tree seedling recruitment and survival are influenced by sub-canopy cover and abundance of the invasive shrub Lonicera maackii.

We sampled 16 sites, each of which contained three 400m² circular plots, from 2012-2014 that were initially established by the U.S, Forest service and represent a time range since EAB infestation throughout central and western Ohio. Nested within each plot, were a sub-plot (200m²) and four (4m²) micro-plots. Within plots we assessed ash health (visual scale 1(healthy) to 5(dead)), measured all trees >10cm diameter at breast height (DBH), and calculated percentage of ash basal area that were rated ≥ 3. Within each sub-plot we conducted transects to measure percent L. maackki cover, measured all sub-canopy trees (3-9.9cm DBH) and the two largest L.maackii in each quadrant to calculate sub-canopy BA and L. maackii BA. At the micro-plot level, we censused tree seedlings 20-100cm to calculate all response variables.

We investigated effects of percentage ash basal area (BA) rated 5, percentage ash BA rated ≥ 3, sub-canopy BA, and L. maackii BA, and L. maackii percent cover on tree seedling recruitment and survival using linear mixed models and Akaike’s Information Criterion for model evaluation. Response variables were seedling species richness, total seedling abundance, proportion of L. maackii seedlings, proportion of invasive seedlings (excluding L. maackii), proportion of tree seedlings, proportion of ash seedlings, proportion of seedling survival, and annual seedling recruitment for 2013 and 2014.

Collectively, more poor quality ash and higher percentages of L. maackii cover were the best predictors for seedling species richness; more species occurred in plots with poorer quality ash and less L. maackii cover. Independently, L. maackii cover was the best predictor for total seedling abundance, proportion of tree seedlings, number of seedling recruits, and proportion of L. maackii seedlings. For each variable there were fewer seedlings present in plots with more L. maackii cover; except for L. maackii seedlings which increased in plots with more L. maackii cover. Separately, plots with more poor quality ash had a greater proportion of invasive seedlings when L. maackii seedlings were excluded. Both the proportion of ash seedlings and total seedlings that survived from 2012-2014 were similar across all sites and seem to be independent of these two variables. 

Further analyses will be conducted using path analysis. Data will also be incorporated into a forest dynamics model to make long-term forest composition predictions that can assist in guiding resource management in forests impacted by emerald ash borer and Amur honeysuckle.