文摘
Constructed treatment wetlands have been well established as effective and sustainable solutions to the problem of urban water treatment and reuse. However, treatment wetlands located in aridland cities may behave differently relative to their more mesic and humid counterparts, and this could potentially impact their ability to deliver the ecosystem services that are expected of them. Specifically, in hot, dry climates large water losses via evaporation and plant transpiration may comprise a major component of whole-system water budgets. Our primary goal was to develop a rigorous and informed model of how well these “working wetlands” function in hot, arid climates by developing and comparing robust water and nutrient budgets, as our process-based understanding of how mesic constructed wetlands function may not be readily transferred to arid climates where constructed wetlands are becoming increasingly widespread. At the Tres Rios constructed treatment wetland in Phoenix AZ USA, we quantified water losses via plant transpiration and open water evaporation as well as inorganic N loads into and from the whole wetland system and into the vegetated marsh. We found that water losses due to transpiration and evaporation were remarkably high when compared to most mesic constructed wetlands. Total water losses via evaporation and transpiration peaked at 300,000 m3 mo−1 (714 L H2O m−2 mo−1) in the hot, dry summer months and averaged more than 70% of the whole-system water losses over a 27 month time period. At the same time, the vegetated marsh removed nearly all of the inorganic N that was supplied to it. Large transpirative water losses moved large volumes of replacement water into the marsh via a “biological tide” that provided more opportunities for vegetation and soil microbes to process N and other target solutes. This enhanced the N treatment efficacy of the Tres Rios constructed treatment wetland relative to humid, mesic systems. To our knowledge, this is the first time that biotically-mediated surface hydrology has been demonstrated in any wetland.
