Green roofs have the potential to address several of the environmental problems associated with urbanisation, and can be used as mitigation for habitats lost at ground level. Brown roofs (a type of green roof) can be used to mitigate for the loss of brownfield habitat, but the best way of designing these habitats remains unclear. This paper reports an experiment to test the effects of different types of recycled aggregate on the development of vegetation assemblages on brown roof mesocosms. Five recycled aggregates were tested: (1) crushed brick, (2) crushed demolition aggregate, (3) solid municipal waste incinerator bottom ash aggregate, (4) a 1:1 mix of 1 and 2, and (5) a 1:1 mix of 3 and 2. Each was seeded with a wildflower mix that also included some Sedum acre and vegetation development was studied over a six-year period. Species richness, assemblage character, number of plants able to seed, and plant biomass were measured. Drought disturbance was the key factor controlling changes in plant assemblage, but effects varied with substrate treatment. All treatments supported a similar plant biomass, but treatments with a high proportion of crushed brick in the growth substrate supported richer assemblages, with more species able to seed, and a smaller amount of Sedum acre. Crushed brick, or recycled aggregates with a high proportion of crushed brick, are recommended as good growth substrate materials for encouraging brown roof plant diversity. This investigation demonstrates the importance of multi-year studies of green roof development for the generation of robust findings.
Carefully designed green roofs have the potential to be used as mitigation for habitats lost at ground level. The development of plant assemblages on two green roofs designed to emulate diverse brownfield habitats (brown roofs), by using recycled demolition aggregate as part of a low-fertility growth substrate, were studied over the first four years of their development. The cover-abundance of flowering plants and habitat structural components (e.g. bare ground, moss) were measured on the Domin-Krajina scale within all identified microhabitats. Drought disturbance was one of the main controlling factors on assemblage development. Annual plants were abundant and successful in the first growth season, and thereafter only re-appeared in any numbers following drought disturbances in subsequent years. Moss and Sedum acre L. increased through the study period until these plants dominated coverage. The cover-abundance of perennial wildflower species was strongly influenced by drought disturbance. The influence of drought disturbance varied between different brown roof microhabitats, with plant assemblages in coarser and less fertile microhabitats more resistant to these disturbances. Observed responses to drought were consistent with the following two hypotheses: (i) Areas of coarse substrate can act as disturbance refugia for plants during drought by helping preserve pockets of water under large clasts and within absorbent materials such as brick. (ii) The plant assemblages living in areas of more fertile substrates, which grow more luxuriantly when water availability is high, are more vulnerable to drought disturbance. Green roofs should be designed to include a range of substrate types to create several microhabitats that will collectively support more species than any single microhabitat.