An experimental design to assess soil-plant-water relations on a Kansas green roof

Abstract

Green roofs are increasingly common as cities seek environmentally sustainable approaches to mitigate climate change impacts while providing urban amenities. However, water availability is one of the most limiting factors in a green roof system and understanding the substrate-plant-water relationships within green roof ecosystems is key for understanding plant success and failures for individual green roofs. To provide insight on green roof plant and substrate selection for green roofs in the Flint Hills Ecoregion (Kansas, USA) and similar climates, three experimental research green roof beds were designed for the Kansas State University (KSU) College of Architecture, Planning and Design (APD) building and were initially planted in October 2017, with replacement plantings completed by June 2018. The three green roof beds provide three different depths (approximately 4, 6, and 8 inches [10.2, 15.3, and 20.4 cm]). The experimental design is a strip plot design within a randomized complete block design for each green roof depth. Each green roof depth contains two substrates (Rooflite® Extensive MC and Kansas BuildEx), and three different species mixes (all-Sedums, Sedums and native grasses, and all-natives). Each of the three specified mixes contain 6 different species planted in triplicate within each replication cell. The main objectives for this APD Experimental Green Roof research are twofold: (1) understand how a locally blended substrate (Kansas BuildEx) and a commercially provided, regionally mixed substrate (Rooflite® Extensive MC) vary in the way they store and dispense water and how different plant mixes may affect the hydrologic processes within these substrates; and (2) understand how vegetative coverage and above-ground biomass of three mixed-species plantings and selected native plant species change over time. Using lessons learned from this experimental study of the two different substrate types, three different substrate depths, and three species mixes, the desire was to improve our collective understanding of the selected plants and substrates and ultimately to improve the design, implementation, and management of green roofs in this part of Kansas and in locations with similar climates. To realize these objectives, our interdisciplinary team has been investigating the relationships between micro-meteorological and subsurface temperatures and soil moisture dynamics, two different substrates installed at depths ranging from 2.4-5.2 inches (6.1-13.2 cm) called the 4-inch bed, 4.5-7.5 inches (11.4-19.1 cm) called the 6-inch bed, and 6.5-10.1 inches (16.5-25.7 cm) called the 8-inch bed, and vegetative coverage of specified mixes and biomass associated with the three distinct species mixes and seven native species (sideoats grama, blue grama, little bluestem, shortbeak or prairie sedge, purple prairie clover, prairie junegrass, and prairie dropseed). This dissertation research examined water holding capacities of the roof while soil moisture release curves were estimated to provide insight on how water is stored and the energy status of this water within the two selected green roof substrates. Lab tests to understand water holding capacities were done at KSU while lab tests on substrate-water energy status were sent to the Turf and Soil Diagnostics Lab in Linwood, Kansas. Soil moisture dynamics in each of the substrate types were investigated by analyzing the recession curve slopes for Rooflite® Extensive MC and Kansas BuildEx to provide insight as to how green roof substrate properties can cause variations in soil moisture retention and recession. Soil moisture recession rates were analyzed for 1-hour and 24-hour periods following rainfall events in two configurations. Configuration 1 assessed soil moisture recession rates within the all-natives species mix planted in both substrates in all three green roof depths from March 2018 to early July 2019. Configuration 2 assessed soil moisture recession rates for the Sedums only and all-natives species mixes planted in both substrates for the 4- and 8-inch depths from late July 2019 to May 2020. For both configurations, soil moisture monitoring was done in situ. In terms of soil moisture, it was found that Kansas BuildEx (BuildEx) had a greater roof capacity than the Rooflite® Extensive MC substrate, and these substrates varied in the energy status of water within the soil. However, there was little to no difference in the rate at which these substrates dispense water (recession rates). For Configuration 1, there was only an effect of substrate in the shallowest bed (4-inch depth) when looking at a 1-hour recession period. For Configuration 2, there was only a slight effect of mix on recession in the 8-inch bed for the 1-hour recession period. Plant coverage and above-ground biomass measurements were taken at the end of the 2018 and 2019 growing seasons. Coverage measurements utilized overhead photography. When looking at species mix performance in these beds, by the end of the second growing season there was a significant effect of mix type on amount of cover (or vegetative coverage within each plot), with the all-natives and Sedum and natives mixes having the greatest cover in the 4-inch and 8-inch beds. In the 6-inch bed there was a significant interaction effect between mix and substrate types. When looking at cover for each mix, Rooflite® Extensive MC yielded greater cover in the Sedums only mix; and when looking at cover in each substrate, the Sedums and natives mix having the greatest cover in Rooflite® Extensive MC, and the Sedums and natives mix and the all-natives mix having the greatest cover in Kansas BuildEx. Regarding individual species performance, by the end of 2019, little bluestem had greater biomass in the Rooflite® Extensive MC substrate than in the BuildEx substrate for the 4-, 6-, and 8-inch beds. In the 6-inch bed, sideoats grama had greater biomass in the Rooflite® Extensive MC substrate than in BuildEx, while purple prairie clover had greater biomass in Kansas BuildEx substrate than in Rooflite® Extensive MC. Buffalograss was one of the species planted in the Sedums and native grasses mix. Based on personal observations this grass performed exceptionally well throughout this study with photographs and visual assessments clearly indicating buffalograss dominance in most plots where it was planted, corresponding to findings in Liu et al., 2019. The outcomes of this study show that there are important relationships occurring between substrate type and mixed-species performance in varying substrate depths for green roof systems associated with the APD Experimental Green Roof. Plant above-ground biomass can be affected by substrate type and particular species. Substrate types influence the percent cover of green roof species mixes, and how water is stored and taken up by plants. The results and work related to this dissertation have enhanced the knowledge of soil-water relations of green roof ecosystems in this part of Kansas, which can help improve design, implementation, and management practices and make green roofs more sustainable. Future research should focus on analyses of how substrate chemical and physical properties change over time (if possible, five and ten after the first full growing season) and how these changes affect water movement within the substrate and plant species and mix performance over the long-term. It is likely that plant patterns will change over time depending on how well each species does over the long-term. Cover and biomass should continue to be monitored to see how the selected mixes and individual species perform over time.