The comparative effects of three Sedum species on green roof stormwater retention

Abstract

Green roofs are typically dominated by Sedum species because they can tolerate hot, xeric environments. However, due to their high water use efficiency, some have questioned the selection of these species for stormwater management. We investigated (1) how three common Sedum species contribute to overall stormwater retention efficiency by green roofs in the mid-Atlantic region, and (2) whether species-specific differences in water use could be explained by morphological and physiological characteristics. Water use and CO2 exchange were continuously monitored in growth chamber studies under increasing drought stress for S. album, and S. kamtschaticum, two species known to variably cycle between CAM and C3 metabolisms. Under fall temperature conditions, S. kamtschaticum had gas exchange rates akin to C3 photosynthesis and used 35% more water compared to S. album. Interestingly, S. album conserved water and had malic acid accumulation confirming CAM metabolism for the duration of the experiment, even under well-watered conditions. In field studies, sixteen replicate green roof platforms (n=4 per species) were planted with S. album, S. kamtschaticum, S. sexangulare, or left unplanted during summer 2010. The platforms were monitored intensively for canopy growth, leaf area, root biomass, substrate moisture and runoff for two years (2011 and 2012). Plant treatment effects on stormwater runoff were significant, but most discernible for small and intermediate-sized rainfall events less than 62.5mm. The two species with the greatest stormwater retention efficiencies, S. kamtschaticumand S. sexangulare, also had the highest rates of evapotranspiration (ET), and higher ET rates resulted in less total runoff. Because evapotranspiration was identified as important for predicting performance by plants in the field study, I investigated how ET data from this study, combined with environmental data collected from a weather station at the study site, could be used to improve the application of the FAO56 Penman-Monteith evapotranspiration equations to green roofs. The incorporation of specific seasonal crop coefficients were found to improve correlations between predicted and measured rates of ET and these coefficients were related to plant characteristics. The refinement of ET equations can lead to more accurate hydrologic models of green roofs and design and management tools.