Biomass thresholds for species dominance and diversity in sown urban meadows: Is it possible to determine what happens in the longer term through design?

Naturalistic meadows inspired by the appearance of grassland communities in nature have been increasingly fashionable as a design tool in urban landscapes. Forb species richness and diversity in a meadow community often determine public appreciation in urban areas and the ecological values to pollinating invertebrates. However, grasses are often more competitive leading to dominance in the longer term particularly on fertile soils. Where grasses are excluded in seed mix design to try to slow down this process, weeds tend to be more problematic, potentially leading to parallel declines in forb species richness and standing biomass. This study aims to test the possibility to design meadow mixes that contain grasses but that do not lead to competitive elimination of the forbs at least in a short term. The research involved setting up a field experiment at the Green Estate Ltd, Manor Top, Sheffield, UK. Twenty nine forbs (15 species geographically distributed in both Western Europe and Inner Mongolia, China and 14 species distributed in Inner Mongolia only) and one grass species (Deschampsia cespitosa) were established by sowing in situ to test design variables: initial sowing rates (including 2 sowing densities and 3 sowing ratios of forb to grass), 2 ecotypes of Deschampsia cespitopsa (short and tall to represent a gradient of light competition), and 2 contrasting depths of sand mulch layer (i.e. difference in levels of moisture stress). The study showed that a forb (predominantly Achillea millefolium and Echinops sphaerocephalus in this study) dominated community can be achieved thought design, but sensible management is essential to maximise performance. Light competition and especially the ability to project taller foliage above grass canopy before it closes the ground proved to be a critical factor. However, dominant forbs reduced the diversity and abundance of subordinate forbs through competition in the same way as grasses do. Thus, this study suggests that species characteristics can play a more important role than the life form of forb or grass in a community. \nIncreasing forb sowing density or decreasing grass sowing density increased forb seedling numbers but not forb biomass in the longer term. Also, having a higher forb sowing rate speeded up the dominance effect. Thus, the most powerful design tool for a forb rich and diverse meadow community is not sowing density or ratio of forb: grass across all species but changing the ratio between the species, and in particular reducing seeds of the potential dominant species seeds to very low levels. The higher moisture stress increased the numbers of forb seedling and biomass but decreased the subordinate forb biomass. This may benefit forb species diversity in the short term but reduce the biomass coexistence in the longer term. Forbs that had a distribution in both Western Europe and Inner Mongolia appeared to be more persistent than Inner Mongolian forb species, with ability to compete for light more important than geographical origins per se.