Tree transpiration in semi-arid cities: An inter-city comparison
The municipal water demand of highly populated cities like Los Angeles, California and Salt Lake City, Utah largely exceeds the local precipitation of the semi-arid regions in which they are located. Moreover, these regions are likely to experience more frequent and prolonged droughts intensified by climate change. Currently, irrigation of urban landscapes in semi-arid cities constitutes approximately 65% of residential water use. There is a great need to understand the patterns of water use by urban trees in these regions at the species level. However, data on urban tree transpiration is very limited. Previously, we used thermal dissipation sapflux probes to measure transpiration of 158 trees belonging to 19 species that are commonly planted in semi-arid cities. Here, we compare the results between two cities: Los Angeles, California and Salt Lake City, Utah. Both cities have low annual precipitation, high rates of urban irrigation, and high atmospheric vapor pressure deficit. However, there is significant variability in climatic and other environmental factors both within and across these cities. We asked the questions: is urban tree transpiration similar within and across these cities? Or is transpiration highly site-specific?
We found that whole-tree transpiration varied greatly among species, study sites, and regions. However, a comparison of similar species at different locations revealed identical relationships between transpiration, atmospheric vapor pressure deficit, and sapwood area. In particular, London planetree and honey locust exhibited identical responses of normalized transpiration to vapor pressure deficit, while differences in sapwood area explained 62% of the variability in tree transpiration. Both of these species are recommended for planting in Salt Lake City for water conservation; however, in situ data indicate that these species have very high transpiration rates. The results of this study are promising for developing a systematic approach to modeling urban tree transpiration in semi-arid cities. This work will help reduce the uncertainty of urban evapotranspiration in hydrologic models and assist landscape planners with selecting water efficient species.