Biofiltration systems are highly valued in urban landscapes as they remove pollutants from stormwater runoff whilst contributing to a reduction in runoff volumes. Integrating trees in biofilters may improve their runoff retention performance, as trees have greater transpiration than commonly used sedge or herb species. High transpiration rates will rapidly deplete retained water, creating storage capacity prior to the next runoff event. However, a tree with high transpiration rates in a biofilter system will likely be frequently exposed to drought stress. Selecting appropriate tree species therefore requires an understanding of how different trees use water and how they respond to substrate drying. We selected 20 tree species and quantified evapotranspiration (ET) and drought stress (leaf water potential; Psi) in relation to substrate water content. To compare species, we developed metrics which describe: (i) maximum rates of ET under well-watered conditions, (ii) the sensitivity of ET and (iii) the response of Psi to declining substrate water content. Using these three metrics, we classified species into three groups: risky, balanced or conservative. Risky and balanced species showed high maximum ET, whereas conservative species always had low ET. As substrates dried, the balanced species down-regulated ET to delay the onset of drought stress; whereas risky species did not. Therefore, balanced species with high ET are more likely to improve the retention performance of biofiltration systems without introducing significant drought risk. This classification of tree water use strategies can be easily integrated into water balance models and improve tree species selection for biofiltration systems. (C) 2018 Elsevier Ltd. All rights reserved.
展开▼
