Effects of urbanization-induced environmental changes on ecosystem functioning in the Phoenix metropolitan region, USA

Urban ecosystems are profoundly modified by human activities and thereby provide a unique "natural laboratory" to study potential ecosystem responses to anthropogenic environmental changes. Because urban environments are now affected by urban heat islands, carbon dioxide domes, and high-level nitrogen deposition, to some extent they portend the future of the global ecosystem. Urbanization in the metropolitan region of Phoenix, Arizona (USA) has resulted in pronounced changes in air temperature (T _air), atmospheric CO₂ concentration, and nitrogen deposition (N _dep). In this study, we used a process-based ecosystem model to explore how the Larrea tridentata dominated Sonoran Desert ecosystem may respond to these urbanization-induced environmental changes. We found that water availability controls the magnitude and pattern of responses of the desert ecosystem to elevated CO₂, air temperature, N deposition and their combinations. Urbanization effects were much stronger in wet years than normal and dry years. At the ecosystem level, aboveground net primary productivity (ANPP) and soil organic matter (SOM) both increased with increasing CO ₂ and N_dep individually and in combinations with changes in T _air. Soil N (N_soil) responded positively to increased N deposition and air temperature, but negatively to elevated CO₂. Correspondingly, ANPP and SOM of the Larrea ecosystem decreased along the urban-suburban-wildland gradient, whereas N_soil peaked in the suburban area. At the plant functional type (FT) level, ANPP generally responded positively to elevated CO₂ and N_dep, but negatively to increased T _air. C₃ winter annuals showed a greater ANPP response to higher CO₂ levels (>420 ppm) than shrubs, which could lead over the long term to changes in species composition, because competition among functional groups is strong for resources such as soil water and nutrients. Overall, the combined effects of the three environmental factors depended on rainfall variability and nonlinear interactions within and between plant functional types and environmental factors. We intend to use these simulation results as working hypotheses to guide our field experiments and observations. Experimental testing of these hypotheses through this process should improve our understanding of urban ecosystems under increasing environmental stresses.