Skip to main content

Data from: Woody plant secondary chemicals increase in response to abundant deer and arrival of invasive plants in suburban forests

Cite this dataset

Morrison, Janet; Roche, Bernadette; Veatch-Blohm, Maren (2022). Data from: Woody plant secondary chemicals increase in response to abundant deer and arrival of invasive plants in suburban forests [Dataset]. Dryad.


Plants in suburban forests of eastern North America face the dual stressors of high white-tailed deer density and invasion by nonindigenous plants. The combination of chronic deer herbivory and strong competition from invasive plants could alter a plant’s stress- and defense-related secondary chemistry, especially for long-lived juvenile trees in the understory, but this has not been studied. We measured foliar total antioxidants,  phenolics, and flavonoids in juveniles of two native trees, Fraxinus pennsylvanica (green ash) and Fagus grandifolia (American beech), growing in six forests in the suburban landscape of central New Jersey, USA. The trees grew in experimental plots that had been subject for 2.5 years to factorial treatments of deer access/exclosure X addition/no addition of the nonindigenous invasive grass Microstegium vimineum (Japanese stiltgrass). As other hypothesized drivers of plant secondary chemistry, we also measured non-stiltgrass herb layer cover, light levels, and water availability. Univariate mixed model analysis of the deer and stiltgrass effects and multivariate structural equation modeling (SEM) of all variables showed that both greater stiltgrass cover and greater deer pressure induced antioxidants, phenolics, and flavonoids, with some variation between species. Deer were generally the stronger factor, and stiltgrass effects were most apparent at high stiltgrass density. SEM also revealed that soil dryness directly increased the chemicals; deer had additional positive, but indirect, effects via influence on the soil; in beech PAR positively affected flavonoids; and herb layer cover had no effect. Juvenile trees’ chemical defense/stress responses to deer and invasive plants can be protective, but also could have a physiological cost, with negative consequences for recruitment to the canopy. Ecological implications for species and their communities will depend on costs and benefits of stress/defense chemistry in the specific environmental context, particularly with respect to invasive plant competitiveness, extent of invasion, local deer density, and deer browse preferences.


Data were collected in six suburban forests in central New Jersey, USA, within 16 m2 plots, with half fenced to exclude deer. Variables measured for this study were:

- Proportion cover of all plants in the herb layer, scored in 10% intervals and measured in 16 0.25 m2 subplots per plot

- Ground-level photosynthetically active radiation (PAR), measured as the percent of full sun PAR, using five measurements per plot made with an AccuPAR ceptometer

- Soil water potential measured on two samples per plot with a benchtop WP4 meter

- A deer browse index for each plot, set to zero for fenced plots and assigned to one common value for all unfenced plots within one forest, based on the proportion of deer-browsed individuals in unfenced plots of five native plant taxa: Carya spp., F. grandifolia, Fraxinus pennsylvanica, Acer rubrum, and Rubus allegheniensis. These were included because they were sufficiently common in the forests’ understories to allow for one index applicable to all of the forests and because they were, in our sites, neither the most browsed species nor completely avoided by deer-browsed

In these same plots, leaf tissue was collected from Fagus grandifolia and Fraxinus pennsylvanica, dried at 50 deg C and used in chemical analysis of total antioxidants, total phenolics, and total flavonoids. 

All data were entered and managed in Excel data files, and statistical analysis was done in SAS for mixed model analysis and in R (piecewiseSEM package) for structural equation modeling

Usage notes

A ReadMe file is included with detailed information.


National Science Foundation, Award: NSF-DEB 1257833