An understanding of the formation of spatial heterogeneity is important because spatial heterogeneity leads to functional consequences at the ecosystem scale; however, such an understanding is still limited. Particularly, research simultaneously considering both external variables and internal feedbacks (self-organization) is rare, partly because these two drivers are addressed under different methodological frameworks. In this dissertation, I show the prevalence of internal feedbacks and their interaction with heterogeneity in the preexisting template to form spatial pattern. I use a variety of techniques to account for both the top-down template effect and bottom-up self-organization. Spatial patterns of nutrients in stream surface water are influenced by the self-organized patch configuration originating from the internal feedbacks between nutrient concentration, biological patchiness, and the geomorphic template. Clumps of in-stream macrophyte are shaped by the spatial gradient of water permanence and local self-organization. Additionally, significant biological interactions among plant species also influence macrophyte distribution. The relative contributions of these drivers change in time, responding to the larger external environments or internal processes of ecosystem development. Hydrologic regime alters the effect of geomorphic template and self-organization on in-stream macrophyte distribution. The relative importance of niche vs. neutral processes in shaping biodiversity pattern is a function of hydrology: neutral processes are more important in either very high or very low discharge periods. For the spatial pattern of nutrients, as the ecosystem moves toward late succession and nitrogen becomes more limiting, the effect of self-organization intensifies. Changes in relative importance of different drivers directly affect ecosystem macroscopic properties, such as ecosystem resilience. Stronger internal feedbacks in average to wetter years are shown to increase ecosystem resistance to elevated external stress, and make the backward shifts (vegetation loss) much more gradual. But it causes increases in ecosystem hysteresis effect. Finally, I address the question whether functional consequences of spatial heterogeneity feed back to influence the processes from which spatial heterogeneity emerged through a conceptual review. Such feedbacks are not likely. Self-organized spatial patterning is a result of regular biological processes of organisms. Individual organisms do not benefit from such order. It is order for free, and for nothing.

Humans have dramatically increased phosphorus (P) availability in terrestrial and aquatic ecosystems. As P is often a limiting nutrient of primary production, changes in its availability can have dramatic effects on ecosystem processes. I examined the effects of calcium carbonate (CaCO3) deposition, which can lower P concentrations via coprecipitation of phosphate, on P availability in two systems: streams in the Huachuca Mountains, Arizona, and a stream, Río Mesquites, in Cuatro Ciénegas, México. Calcium carbonate forms as travertine in the former and within the microbialites of the latter. Despite these differences, CaCO3 deposition led to lowered P availability in both systems. By analyzing a three-year dataset of water chemistry from the Huachuca Mountain streams, I determined that P concentrations were negatively related to CaCO3 deposition rates. I also discovered that CaCO3 was positively correlated with nitrogen concentrations, suggesting that the stoichiometric effect of CaCO3 deposition on nutrient availability is due not only to coprecipitation of phosphate, but also to P-related constraints on biotic nitrogen uptake. Building from these observations, bioassays of nutrient limitation of periphyton growth suggest that P limitation is more prevalent in streams with active CaCO3 deposition than those without. Furthermore, when I experimentally reduced rates of CaCO3 deposition within one of the streams by partial light-exclusion, areal P uptake lengths decreased, periphyton P content and growth increased, and periphyton nutrient limitation by P decreased. In Río Mesquites, CaCO3 deposition was also associated with P limitation of microbial growth. There, I investigated the consequences of reductions in CaCO3 deposition with several methods. Calcium removal led to increased concentrations of P in the microbial biomass while light reductions decreased microbial biomass and chemical inhibition had no effect. These results suggest that CaCO3 deposition in microbialites does limit biological uptake of P, that photoautotrophs play an important role in nutrient acquisition, and, combined with other experimental observations, that sulfate reduction may support CaCO3 deposition in the microbialite communities of Río Mesquites. Overall, my results suggest that the effects of CaCO3 deposition on P availability are general and this process should be considered when managing nutrient flows across aquatic ecosystems.

Though cities occupy only a small percentage of Earth's terrestrial surface, humans concentrated in urban areas impact ecosystems at local, regional and global scales. I examined the direct and indirect ecological outcomes of human activities on both managed landscapes and protected native ecosystems in and around cities. First, I used highly managed residential yards, which compose nearly half of the heterogeneous urban land area, as a model system to examine the ecological effects of people's management choices and the social drivers of those decisions. I found that a complex set of individual and institutional social characteristics drives people's decisions, which in turn affect ecological structure and function across scales from yards to cities. This work demonstrates the link between individuals' decision-making and ecosystem service provisioning in highly managed urban ecosystems.

Second, I examined the distribution of urban-generated air pollutants and their complex ecological outcomes in protected native ecosystems. Atmospheric carbon dioxide (CO₂), reactive nitrogen (N), and ozone (O₃) are elevated near human activities and act as both resources and stressors to primary producers, but little is known about their co-occurring distribution or combined impacts on ecosystems. I investigated the urban "ecological airshed," including the spatial and temporal extent of N deposition, as well as CO₂ and O₃ concentrations in native preserves in Phoenix, Arizona and the outlying Sonoran Desert. I found elevated concentrations of ecologically relevant pollutants co-occur in both urban and remote native lands at levels that are likely to affect ecosystem structure and function. Finally, I tested the combined effects of CO₂, N, and O₃ on the dominant native and non-native herbaceous desert species in a multi-factor dose-response greenhouse experiment. Under current and predicted future air quality conditions, the non-native species (Schismus arabicus) had net positive growth despite physiological stress under high O₃ concentrations. In contrast, the native species (Pectocarya recurvata) was more sensitive to O₃ and, unlike the non-native species, did not benefit from the protective role of CO₂. These results highlight the vulnerability of native ecosystems to current and future air pollution over the long term. Together, my research provides empirical evidence for future policies addressing multiple stressors in urban managed and native landscapes.

Constructed treatment wetlands (CTW) have been a cost-efficient technological solution to treat different types of wastewater but may also be sources of emitters of methane (CH4) and nitrous oxide (N2O). Thus, my objective for this dissertation was to investigate CH4 and N2O fluxes via multiple pathways from the Tres Rios CTW located in Phoenix, AZ, USA. I measured gas fluxes from the CTW along a whole-system gradient (from inflow to outflow) and a within-marsh gradient (shoreline, middle, and open water sites). I found higher diffusive CH4 release in the summer compared to spring and winter seasons. Along the whole-system gradient, I found greater CH4 and N2O emission fluxes near the inflow compared to near the outflow. Within the vegetated marsh, I found greater CH4 emission fluxes at the vegetated marsh subsites compared to the open water. In contrast, N2O emissions were greater at the marsh-open water locations compared to interior marsh. To study the plant-mediated pathway, I constructed small gas chambers fitted to Typha spp. leaves. I found plant-mediated CH4 fluxes were greater near the outflow than near the inflow and that CH4 fluxes were higher from lower sections of plants compared to higher sections. Overall, Typha spp. emitted a mean annual daily flux rate of 358.23 mg CH4 m-2 d-1. Third, using a 30-day mesocosm experiment I studied the effects of three different drydown treatments (2, 7, 14 days) on the fluxes of CH4 and N2O from flooded CTW soils. I found that CH4 fluxes were not significantly affected by soil drydown events. Soils that were dry for 7 days shifted from being N2O sources to sinks upon inundation. As a result, the 7-day drydown soils were sinks while the 14-day drydown soils showed significant N2O release. My results emphasize the importance of studying ecological processes in CTWs to improve their design and management strategies so we can better mitigate their greenhouse gas emissions.