Vegetated (green) roofs have become common in many cities and are projected to continue to increase in coverage, but little is known about the ecological properties of these engineered ecosystems. The objective of this study was to test whether green roofs conform to commonly-observed patterns from natural ecosystems, in terms of the relationship between plant species richness and ecosystem function. We used commercially available green roof trays as replicated plots with varying levels of plant species richness (0, 1, 3, or 6 common green roof species per plot) to explore the influence of species richness on water-quality relevant ecosystem functions: specifically, water retention and inorganic nutrient (N, P) retention following rain events. We hypothesized that greater plant species richness would result in increased water and nutrient retention, i.e., decreased hydrologic leaching losses. This hypothesis was based on the expectation that richer mixtures of species will have higher productivity as well as increased capacity to fully exploit available water and nutrients due to complementary resource use. To test this hypothesis, accumulated plant biomass was estimated near the peak of the first full growing season, and runoff volume and nutrient fluxes were measured from a subset of runoff events for 1 year.
Results/Conclusions
We found that: (1) total plant biomass increased with increasing species richness; (2) green roof plots were effective at reducing storm runoff, with vegetation increasing water retention more than soil-like substrate alone, but there was no significant effect of plant species richness on runoff volume; (3) green roof substrate was a significant source of phosphate, regardless of presence/absence of plants; (4) dissolved inorganic nitrogen (DIN = nitrate + ammonium) runoff fluxes were different among plant species, and decreased significantly with increasing plant species richness. The variation in N retention was positively related to variation in plant biomass. Our results suggest that complementarity effects rather than selection effects are the primary contributor to the link between biodiversity and ecosystem function in this system. Notably, the increased biomass and N retention with species richness in this engineered ecosystem are similar to patterns observed in published studies from grasslands and other well-studied ecosystems. We suggest that more diverse plantings on vegetated roofs may enhance the retention capacity for reactive nitrogen. This is of importance for the sustained health of vegetated roof ecosystems, which over time often experience nitrogen limitation, and is also relevant for water quality in receiving waters downstream of green roofs.