Figure Captions
Figure 1. Proportion of sampled unfenced plants with deer browse signs in the understory of a central New Jersey suburban forest, Herrontown Woods Preserve. LIBE, Lindera benzoin ; NYSY, Nyssa sylvatica ; EUAL, Euonymus alatus ; ROMU, Rosa multifloraNumber of sampled plants from right to left: LIBE, 30, 476; NYSY, 29, 308; EUAL, 80, 743; ROMU, 55, 492. G-tests for heterogeneity indicated a significant difference among species only for the pooled 2012-2019 data; different letters indicate species that were significantly different (P< 0.02) in pairwise tests.
Figure 2. Comparison of the metabolite profiles of the tree species in the fenced and unfenced plots. A) The number of common and unique metabolite features identified for the four woody tree species is depicted in the Venn diagram. B) Principal component analysis (PCA) depicts the clustering of all the samples based on species. C) Dendrogram displays the relationship of the samples based on genotype and presence or absence of fences. D) Common and unique metabolic pathways predicted based on the differentially accumulating metabolites.
Figure 3. Identification of metabolites that accumulate differentially in the four species following the treatment gradient. A) Partial least squares discriminant analysis (PLS-DA) separates the samples into eight distinct groups corresponding to species and containment in fence or unfenced plots. B) Important features that contributed to the PLS-DA based separation of the samples are depicted with their pattern of accumulation shown by the color code. C) Normalized concentrations (mean+ SE) of the top five metabolite features identified by PLS-DA. Different letters indicate statistically significant differences (ANOVA and Tukey’s HSD).
Figure 4. Untargeted metabolomic analysis of N. sylvatica samples from fenced and open plots. A) Principal component analysis (PCA) grouped the samples into two clusters that correspond with the fencing treatment. B) PLS-DA clustering identified two distinct groups based on treatment. C) The top 15 metabolite features that contributed to the separation of the samples into two clusters on the PLS-DA plot and their relative abundances is shown. D) Hierarchical cluster analysis (HCA) computed based on the top 30 significantly different metabolite features groups the samples into two groups.
Figure 5. Metabolomic comparison of fenced and unfenced L. benzoin samples. A) PCA plot demonstrates the metabolite-based relationship of the fenced and unfenced L. benzoin samples. B) PLS-DA plot clusters L. benzoin samples into two groups based on significant features. C) The top 15 metabolite features that are responsible for the PLS-DA grouping of the samples and their relative abundances in fenced and unfenced plants is displayed. D) Pearson Correlation and HCA identified the top 25 metabolites that varied significantly among the samples and clustered the sampled into groups based on these features.
Figure 6. Untargeted metabolomic analysis of R. multiflora under fenced and unfenced plots. A) PCA plot displays the similarities and differences of R. multiflora samples based on the identified metabolite features. B) PLS-DA identifies two clusters that correspond to treatments. C) Important features that differentiate the samples into fenced and unfenced groups. D) HCA displays the top 25 significantly different metabolites and their relative abundances among the R. multiflora samples.
Figure 7. Comparative metabolomic analysis of E. alatus samples from fenced and open plots. A) PCA indicates no treatment-based clustering of the samples. B) PLS-DA clusters the E. alatussamples into two treatment-related groups. C) Fifteen of the top most important features were identified by PLS-DA analysis and their relative abundances in fenced and unfenced plots are shown. D) HCA depicts the relative concentrations of the top 25 metabolite features that differed among the fenced and unfenced samples significantly.
Figure 8. Means + 95% CL of percent cover and height of four woody and semi-woody species in the understory of a central New Jersey suburban forest, Herrontown Woods Preserve, in fall 2019. Fenced plots had excluded deer since spring 2013. LIBE, Lindera benzoin ; NYSY,Nyssa sylvatica ; EUAL, Euonymus alatus ; ROMU, Rosa multiflora . Top, N fenced = 18, N not fenced = 20. Bottom, N from left to right: 23, 41, 14, 5, 65, 18, 11, 17. * P<0.05, **P<0.01, ****P<0.0001, from one-tailed t-tests between fenced and no-fenced values for each species.
Figure 9. Phylogenetic relatedness of the four species included in the study. Species Order classifications are from the Integrated Taxonomic Information System (http://www.itis.gov), and the Order-level tree is from the Angiosperm Phylogeny Website (http://www.mobot.org/MOBOT/research/APweb/).