Effects of Urbanization on the Water Cycle and Aquatic Ecosystems I
Presiding: C Welty, University of Maryland Baltimore County; C M Swan, University of Maryland Baltimore County
NB21F-01 08:30h
Towards Sustaining Water Resources and Aquatic Ecosystems: Forecasting Watershed Risks to Current and Future Land Use Change
Sustaining aquatic resources requires managing existing threats and anticipating future impacts. Resource managers and planners often have limited understanding of the relative effects of human activities on stream conditions and how these effects will change over time. Here we assess and forecast the relative impacts of land use on sediment concentrations in Mediterranean-climate watersheds in California. We focus on the Russian River basin, which supports threatened salmonid populations vulnerable to high levels of fine sediment. We ask the following questions: (1) What are the relative impacts of three different land uses (urban, exurban and agriculture) on the patterns of fine sediment in streams? (2) What is the relative contribution of past and current changes in land use activities on these patterns? and (3) What are the effects of future development on these sediment levels? First, we characterized land use at the parcel scale to calibrate the relative impacts of exurban and urban land use on stream substrate quality, characterized by the concentration of fine sediment surrounding spawning gravels (`embeddedness') in 105 stream reaches. Second, we built multiple ordinal logistic regression models on a subset of watersheds (n=64) and then evaluated substrate quality predictions against observed data from another set of watersheds (n=41). Finally, we coupled these models with spatially explicit land use change models to project future stream conditions and associated uncertainties under different development scenarios for the year 2010. We found that the percent of urban housing and agriculture were significant predictors of in-stream embeddedness. Model results from parcel-level land use data indicated that changes in development were better predictors of fine sediment than total development in a single time period. In addition, our results indicate that exurban development is an important threat to stream systems; increases in the percent of total exurban development in a watershed significantly reduced the odds of observing low embeddedness. Our 2010 forecasts highlight the sensitivity of watersheds to small changes in exurban growth. In previously unimpaired watersheds, small increases in future exurban growth resulted in cumulative impacts on substrate quality not predicted by models lacking this land use type. Because most previous analyses have characterized land use at a resolution that cannot capture exurban development, these results suggest that many such models may be missing an important type of development that can adversely impacting aquatic ecosystems. We suggest that parcel level data may be the fundamental unit for land use change analysis because it represents the economic decision unit for land owners and resolves issues of geographical scale and boundary issues that have long hampered progress in ecological forecasting.
NB21F-02 08:45h
Effects of Concrete Channels on Stream Biogeochemistry, Maryland Coastal Plain
In the 1950's and 60's, extensive networks of cement-lined channels were built in suburban watersheds near Washington, D.C. to convey storm water to downstream locations. These cement-lined stream channels limit interactions between surface and groundwater and they provide sources of alkalinity in Maryland Coastal Plain watersheds that normally have low alkalinity. This project was designed to 1) compare base flow water chemistry in headwater reaches of urban and non-urban streams, and 2) to evaluate downstream changes in water chemistry in channelized urban streams in comparison with non-urban reference streams. During a drought year, headwater streams in both urban and non-urban sites had significant concentrations of Fe(II) that were discharged from groundwater sources and rapidly oxidized by iron-oxidizing bacteria. During a wet year, the concentrations of Fe(II) were higher in headwater urban streams than in the non-urban streams. This suggests that impervious surfaces in headwater urban watersheds prevent the recharge of oxygen-rich waters during storm events, which maintains iron-rich groundwater discharge to the stream. Downstream changes in water chemistry are prominent in cement-lined urban channels because they are associated with distinctive microbial communities. The headwater zones of channelized streams are dominated by iron-ozidizing bacteria, that are replaced downstream by manganese-oxidizing zones, and replaced further downstream by biofilms dominated by photosynthesizing cyanobacteria. The reaches dominated by cyanobacteria exhibit diurnal changes in pH due to uptake of CO2 for photosynthesis. Diurnal changes range from 7.5 to 8.8 in the summer months to 7.0 to 7.5 in the cooler months, indicating both the impact of photosynthesis and the additional source of alkalinity provided by concrete. The dissolved oxygen, pH, and other characteristics of tributaries dominated by cyanobacteria are similar to the water chemistry characteristics observed in much larger urban river channels further downstream. These downstream redox zonations, microbial habitats, and pH characteristics observed in channelized tributaries are very different from non-urban watersheds in the Maryland Coastal Plain, which have pH values less than 7 and do not have the prominent redox zonations and associated microbial habitats. These downstream changes in redox chemistry and pH in urban stream channels have implications for the transport and retention of heavy metals in urban streams.
NB21F-03 09:00h
Urbanization and Water Quantity: Impacts and Mitigation
Urban and suburban development adversely impacts both surface and ground water resources by profoundly altering the hydrologic cycle. Conventional storm water management practices rely on storage to reduce runoff peaks and trap pollutants, but do not address impacts due to changes in the partitioning of water. The introduction of impervious surfaces increases the amount of storm runoff and decreases the amount of ground water recharge. Ground water pumping reduces the amount of ground water. Increased storm runoff causes channel enlargement and increased flood peaks, even when mitigation storage is employed. Reduced ground water results in decreased ground water discharges to aquatic systems. Enhancing the infiltration of storm runoff appears to be a feasible way to address the hydrologic alterations that usually accompany urban and suburban development. Practices that focus infiltration, such as rain gardens and bioretention facilities, appear to be particularly effective. Modeling indicates that a properly designed bioretention facility can preserve natural runoff amounts and increase ground water recharge amounts well above natural levels, perhaps even compensating for ground water pumping. Remaining research issues involve subsoil characterization, vegetation selection, ground water contamination, regulatory strategies, and long-term performance.
NB21F-04 09:15h
Dry Weather Runoff Patterns in Arid Watersheds
Dry weather flow in arid, urban watersheds is often dominated by non-point source runoff that may emanate from hundreds of individual inputs. These non-storm flows are typically assumed to be relatively constant and modeling and monitoring efforts are designed based on this steady-state assumption. This study investigated temporal patterns in dry weather urban runoff at multiple spatial scales. Flow was monitored at 5-minute intervals at four southern California catchments of various sizes over a 5-month period between December 2003 and May 2004. Non-storm flow was analyzed for spatial and temporal patterns. Mean flow was linearly related to catchment size and the area-normalized flow was consistently between 176 and 185 m3 d-1 km-2. Flow in all continuously flowing catchments exhibited varying degrees of 24-hour periodicity. This study showed that non-storm flow in urban watersheds should be based on daily averages because of observed periodicities. Finally, the similarity in area-normalized flow suggested that there might be consistent scaling relationships for dry weather urban runoff based on watershed characteristics.
NB21F-05 09:30h
Effect of Increases in Peak Flows and Imperviousness on Stream Morphology of Ephemeral Streams in Southern California
The relationship between increased basin imperviousness, increased effective runoff, and stream channel degradation has been well documented. In 2003 the Center for Watershed Protection reviewed over 220 studies on this topic and developed an Impervious Cover Model (ICM) as a key paradigm to explain and sometimes predict how indicators of stream quality change in response to different levels of watershed development. The ICM predicts that stream quality indicators will decline when watershed impervious cover exceeds 10% and that severe degradation will occur above 25% impervious cover. However, one of the major gaps in the ICM is that it has not been tested in the southwestern U.S. and it is not certain how well the existing ICM predicts stream stability in arid and semi-arid climates. The goal of this study was to begin filling this knowledge gap by investigating the relationship between basin imperviousness and stability of ephemeral streams in southern California. A combination of historic and current field geomorphic analysis was conducted on ten sites to establish a relationship between increases in basin imperviousness and peak flow and changes in stream channel morphology, given the boundary resistance in each stream. This data was used to generate an enlargement curve for southern California streams and compare that to relationships observed in other portions of the country. The apparent threshold of response for southern California's ephemeral streams in less than 5%, substantially lower than observed in other regions of the country. Furthermore, ephemeral streams appear to be less resilient than similarly sized perennial streams. Although all channels appear to have an inherent rate of degradation, the rate increases substantially when basin imperviousness exceeds 2-3%. This study constitutes the first step toward building a regional ICM for arid and semi-arid areas
NB21F-06 09:45h
Using Passive Sampling Devices to Assess Chemistry and Toxicity in Streams from six U.S Metropolitan Areas
The U.S. population is growing by almost 3 million people a year with concomitant increase in urban development. Increased urbanization causes changes to watersheds which may affect aquatic biota by altering the physical and chemical environment. We deployed semi-membrane-permeable-devices (SPMDs) for 30 days to passively sample hydrophobic organic contaminants (HOCs) from 180 streams in six major metropolitan areas in the U.S.: Atlanta, Georgia, Dallas/Fort Worth, Texas, Denver, Colorado, Milwaukee, Wisconsin, Portland, Oregon, and Raleigh/Durham, North Carolina. SPMD extracts were tested with two assays: (1)Fluoroscan which estimates PAH concentration (pyrene index) by exposing samples to UV light and (2)P450RGS which measures induction of CYP1A a liver enzyme involved in detoxification of organic contaminants. There was a strong positive relation between urban intensity and both the pyrene index and CYP1A in streams from all six metropolitan areas indicating higher HOC concentrations and greater potential toxicity at higher urbanization levels. Invertebrate community responses as measured by EPT taxa richness and benthic index of biotic integrity were also significantly and negatively correlated with both the pyrene index and CYP1A. Our results suggest that toxicity may be a factor in degradation of invertebrate communities in urban environments.