Weeds, worms, and deer – Oh my!


Biotic global change agents, such as non-native plants (‘weeds’), non-native earthworms (‘worms’), and overabundant herbivores (‘deer’), can be major ecosystem drivers in the forest understory. A new study by Schoodic Institute and the National Park Service, published in the journal Biological Invasions, examines the status and relationships among these stressors at eight national parks across seven northeastern states. Lead author, Schoodic Institute Forest Ecology Director Nicholas Fisichelli, sums up the study, “this research highlights the tremendous multipronged management challenge for areas already experiencing the combined effects of weeds, worms, and deer and the future vulnerability of other areas as temperatures warm and conditions become more amenable to biotic global change stressors.”

The results show that weeds, worms, and, deer were common across all parks in the study – overall, 46% of plots had non-native plants, 42% of plots had evidence of earthworms, and all parks had plots with heavy deer browse damage. “All biotic global change stressors were strongly and positively correlated with one another,” says study co-author, ecologist Kate Miller of the National Park Service. Deer browse positively influenced earthworm presence and both deer and earthworms promoted non-native plants. Consequently, 29% of plots included earthworms, non-native plants, and moderate or more severe browse damage, and 28% of plots had no earthworms, no non-native plants and only minor deer browse. Warmer air temperatures and higher soil pH also facilitated non-natives.

Parks in southern New England, New York, and New Jersey, such as Morristown National Historical Park and Roosevelt-Vanderbilt National Historic Sites, had the highest levels of weeds, worms, and deer while parks in northern New England including Acadia National Park and Marsh-Billings-Rockefeller National Historical Park had relatively lower levels of biotic stressors. Understanding the status and relationships among understory global change stressors and associated environmental variables is critical to identifying vulnerable sites and developing successful management strategies tailored to individual site conditions and priority native species.

Deer, as selective herbivores, browse palatable plants (e.g., oak and maple seedlings) and avoid other species (e.g., spruces). The forest understory is like an all-you-can-eat salad bar and deer move along and eat all the tastiest salad fixings, perhaps cucumbers and chickpeas, and leave the things they don’t like, perhaps olives or capers. Non-native earthworms are ecosystem engineers that alter the forest floor and soil environment and reduce native seedling growth and survival. Non-native plant invasions can reduce the abundance and growth of native plants.

Heavy deer browse damage to a northern white cedar sapling in Acadia National Park. Photo by N. Fisichelli


Non-native earthworm found in one of Acadia National Park’s forest health plots. Photo by N. Rosenberg

Locations of national park units included in the study: 1. Acadia National Park, 2. Marsh-Billings-Rockefeller National Historical Park, 3. Saint-Gaudens National Historic Site, 4. Saratoga National Historical Park, 5. Minute Man National Historical Park, 6. Roosevelt-Vanderbilt National Historic Sites, 7. Weir Farm National Historic Site, and 8. Morristown National Historical Park. Study parks are found in three ecological provinces: Laurentian Mixed Forest (light gray), Adirondack-New England Mixed Forest (medium gray), and Eastern Broadleaf Forest (Oceanic) Province (dark gray)

Bivariate relationships among biotic global change stressors (plot-level averages, n=350). Spearman rank correlations are shown in the upper left corner of each subpanel (all correlation p-values <0.001). Deer browse damage increases from 1 (low) to 4 (very high). Non-native plant values are plot averages per m2. Boxplot boxes are the interquartile range, thick horizontal lines are the medians, dashed vertical lines are the outer tails (1.5 x interquartile range), and open circles are plot values beyond the outer tails.