Evapotranspiration of Urban Landscape Trees and Turfgrass in an Arid Environment: Potential Tradeoffs in the Landscape

摘要

Irrigation in arid urban landscapes can use significant amounts of water. Water conservation must be based on plant species and the ability to meet plant water requirements while minimizing overirrigation. However, actual evapotranspiration (ET) estimates for landscape trees and turfgrass in arid environments are poorly documented, especially direct comparisons to assess potential trade-offs. We conducted research to quantify ET of 10 common landscape tree species grown in southern Nevada and compared these values with the ET of both a warm season and cool season turfgrass species. The trees were grown in a plot with a high-density planting (256 trees/ha). A complete morphological assessment was made on each tree, and monitoring of plant water status was conducted monthly. ET was quantified with a hydrologic balance approach, irrigating based on the previous week's ET to eliminate a drainage component. Transpiration was estimated with sap-flow sensors, and evaporation was estimated by difference. Although ET in liters revealed no statistical difference based on species, there were many significant differences in tree morphological parameters (P < 0.05), such as found with basal canopy area. When ET was converted to centimeters based on standardizing the ET on a basal canopy area basis, statistically higher ET values (P < 0.05) were generated for three of the trees (Lagerstroemia indica, Gleditsia tricanthos, and Fraxinus velutina 'Modesto'). A clear separation of all tree ET values (lower ET) with turfgrass ET occurred (P < 0.001), with the exception of L. indica. Backward regression analysis revealed that all morphological and physiological parameters were eliminated with the exception of percent cover in predicting ET (cm, R-2 = 0.88, P < 0.001). In addition, a highly curvilinear relationship existed between decreasing percent tree cover and ET on a basal canopy area basis (R-2 = 0.96, P < 0.001), revealing that smaller trees located within the plot had significantly higher ET (centimeters). Tree-to-grass water use ratios demonstrated that all species except L. indica had ratios significantly below 1.0, indicating that on the basis of this study, landscapes dominated by mature trees irrigated at ET would have lower water use rates than similar areas planted to turfgrass, with the exception of the smaller L. indica. The results suggest that the smaller trees within the higher planting density plot were partially released from a negative feedback on transpiration that occurred in the larger trees based on reduced canopy atmospheric coupling.