The effects of local filtering processes on the structure and functioning of native plant communities in experimental urban habitats

Abstract

Despite a growing literature-base devoted to document biodiversity patterns in cities, little is known about the processes that influence these patterns, and whether they are consistent over time. In particular, numerous studies have identified the capacity of cities to host a rich diversity of plant species. This trend, however, is driven primarily by introduced species, which comprise a large proportion of the urban species pool relative to natives. Using an experimental common garden study, we assessed the relative influence of local assembly processes (i.e., soil environmental filtering and competition from spontaneous urban species) on the taxonomic and functional diversity of native plant communities sampled over four seasons in 2016-2018. Taxonomic and functional diversity exhibited different responses to local processes, supporting the general conclusion that species- and trait-based measures of biodiversity offer distinct insights into community assembly dynamics. Additionally, we found that neither soil nor competition from spontaneous urban species influenced taxonomic or functional composition of native species. Functional composition, however, did shift strongly over time and was driven by community-weighted mean differences in both measured traits (maximum height, Hmax; specific leaf area, SLA; leaf chlorophyll a fluorescence, Chl a) and the relative proportions of different functional groups (legumes, annual and biennial-perennial species, C4 grasses, and forbs). By contrast, taxonomic composition only diverged between early and late seasons. Overall, our results indicate that native species are not only capable of establishing and persisting in vacant urban habitats, they can functionally respond to local filtering pressures over time. This suggests that regional dispersal limitation may be a primary factor limiting native species in urban environments. Thus, future regreening and management plans should focus on enhancing the dispersal potential of native plant species in urban environments, in order to achieve set goals for increasing native species diversity and associated ecosystem services in cities.

Keywords: biodiversity; community assembly processes; functional traits; urban ecology; vacant land.

FIGURE 1

(a) Map of study site location relative to Baltimore City, Maryland (orange outline) and (b) location on the University of Maryland, Baltimore County campus (star). (c) Illustration of the experimental plot layout. Soil treatment groups are indicated by color (orange‐topsoil; gray‐fill) and competition treatment groups are indicated by a “W” for weeded and no letter for unweeded plots; solid lines around boxes indicate control groups that were not seeded, and numbers in each box indicate the seeded native “community type.”

FIGURE 2

Plotmeans of (a) taxonomic diversity (Shannon H), (b) functional diversity (FDis), (c) CWM Hmax, (d) CWM specific leaf area (SLA), (e) CWM chlorophyll a, (f) CWM seed weight, (g) CWM C3 species, (h) CWM C4 species, (i) CWM forbs, (j) CWM legumes, (k) CWM annuals, (l) CWM biennials/perennials with 95% CIs for soil type (fill vs. topsoil). Letters indicate significant differences between pairwise comparisons of seasons (p ≤ .05).

FIGURE 3

Plotmeans of (a) taxonomic diversity (Shannon H), (b) functional diversity (FDis), (c) CWM Hmax, (d) CWM specific leaf area (SLA), (e) CWM chlorophyll a, (f) CWM seed weight, (g) CWM C3 species, (h) CWM C4 species, (i) CWM forbs, (j) CWM legumes, (k) CWM annuals, (l) CWM biennials/perennials with 95% CIs for competition treatment groups (unweeded vs. weeded). Letters indicate significant differences between pairwise comparisons of seasons (p ≤ .05).

FIGURE 4

Plotmeans of (a) taxonomic diversity (Shannon H), (b) functional diversity (FDis), (c) CWM Hmax, (d) CWM specific leaf area (SLA), (e) CWM chlorophyll a, (f) CWM seed weight, (g) CWM C3 species, (h) CWM C4 species, (i) CWM forbs, (j) CWM legumes, (k) CWM annuals, (l) CWM biennials/perennials with 95% CIs for season treatment groups (years 1–4). Letters indicate significant differences between pairwise comparisons of seasons (p ≤ .05).

FIGURE 5

Principal coordinate analysis (PCoA) of (a) taxonomic composition and (b) functional composition data (n = 24 for each season) with convex hulls differentiating each season and 95% confidence ellipses.

FIGURE 6

PCA biplot of soil variables for urban fill and topsoil treatments measured in seasons 1 (S1) and 4 (S4). Soil variables include % moisture, pH, organic matter (OM), aluminum (Al), arsenic (as), calcium (ca), cadmium (cd), chromium (Cr), copper (cu), iron (Fe), potassium (K), magnesium (mg), manganese (Mn), sodium (Na), phosphorus (P), lead (pb), zinc (Zn). Only soil variables that were significantly correlated with the first two principal component axes (p ≤ .05) were retained as vectors for the biplot.