Grantee Research Project Results
2002 Progress Report: Alternative Urbanization Scenarios for an Agricultural Watershed: Design Criteria, Social Constraints, and Effects on Groundwater and Surface Water Systems
EPA Grant Number: R828010Title: Alternative Urbanization Scenarios for an Agricultural Watershed: Design Criteria, Social Constraints, and Effects on Groundwater and Surface Water Systems
Investigators: Lathrop, Richard C. , LaGro Jr., James A. , Nelsoni, Edward B. , Zedler, Joy B. , Bahr, Jean M. , Bradbury, Kenneth R. , Greb, Steven R. , Potter, Kenneth W. , Nowak, Peter
Institution: University of Wisconsin - Madison , Wisconsin Department of Natural Resources
EPA Project Officer: Packard, Benjamin H
Project Period: January 15, 2000 through January 14, 2003
Project Period Covered by this Report: January 15, 2002 through January 14, 2003
Project Amount: $886,105
RFA: Water and Watersheds (1999) RFA Text | Recipients Lists
Research Category: Watersheds , Water
Objective:
The objectives of this research project are to elucidate the mechanisms and to further develop the models used to assess hydrological impacts in relation to various conditions and scenarios, as we evaluate and assess groundwater relationships.
Progress Summary:
Wetland Vegetation Research. The ecological component of our research project pertains to wetland systems and how altered hydrologic regimes affect plant biodiversity and function. In temperate latitudes of the United States and Canada, many wetlands have been invaded by reed canary grass, Phalaris arundinacea, often replacing the natural diverse flora with dense monotypic stands of P. arundinacea. Our research has been particularly effective in explaining the factors that cause high-quality wetlands to become lower quality systems dominated by P. arundinacea. The key factor is the synergism of hydrologic disturbances (such as increased surface-water runoff and decreased groundwater supplies) and the special invasive characteristics of P. arundinacea.
Understanding how urbanization affects wetlands requires the study of patterns in the field as well as experimental evaluation of cause-effect relationships. Twelve wet meadows in Dane County, Wisconsin, showed vegetation patterns that correlate with hydrologic disturbance (e.g., increased stormwater runoff from urban areas) and presence of P. arundinacea. Sites (0.45 ha) with indicators of hydrologic disturbance had lower species richness and lower floristic quality index (FQI) values than reference sites, and both species richness and FQI were inversely related to the abundance of P. arundinacea per site. At the plot scale (1m2), only 5.5 species coexisted with P. arundinacea. In contrast, about twice as many species coexisted with native graminoids: 11.5 with Carex stricta and 10.6 with Calamagrostis canadensis. Cover of P. arundinacea was significantly higher in plots on sites with disturbed hydrologic regimes than on reference sites. These field results suggest a strong negative effect of hydrologic disturbance or presence of P. arundinacea on the quantity and quality of species in a wetland. When the two factors co-occurred, the difference was more severe. Plots with P. arundinacea on disturbed sites had one-third of the species of reference plots and the lowest cover and FQI values.
To determine which of the disturbances associated with urban stormwater runoff are most responsible for P. arundinacea invasions, we conducted an experiment in 150 mesocosms, each 1.1m2 in area. We planted a diverse assemblage of wet meadow species and grew them for 2 years before introducing plugs of P. arundinacea. Simultaneously, we initiated disturbance treatments with nutrient additions, sediment additions, and flooding (each at 3 levels in all 27 combinations). Native species richness decreased with the application of sediments or flooding of 4 consecutive weeks or longer, and losses of several matrix species in these treatments increased the overall light transmission through the plant canopy, which increased the final biomass of P. arundinacea. P. arundinacea also expanded in direct relation to the level of nutrients added.
Study results indicated that increases in the gross supply of resources (via nutrient addition and nutrient-rich topsoil sediment addition) and decreases in the uptake of resources by resident vegetation (due to the death and damage of native plants resulting from sedimentation and floods of increasing duration) increased the invasibility of the experimental assemblages. We further showed that combining two or more disturbances tended to amplify invasion. This study points to the importance of halting nutrient inputs and maintaining the integrity of the native plant canopy to control the spread of a clonal invasive species. In sedge meadow wetlands characterized by plant tussocks that provide microhabitat for a diverse flora, infilling between the tussocks by runoff sediments eliminates this microhabitat and provides a nutrient-rich substrate with little light limitation for rapid colonization and dominance by reed canary grass. Thus, our key ecological study component is linked to the physical attributes of our area's hydrologic system. To maintain the ecological health and biodiversity of natural wetland systems throughout the region, natural infiltration rates without an increase in runoff and associated nutrients and sediments must be maintained as an area urbanizes.
Other results of our wetland research component have a more direct linkage to our social research component on the property owner acceptance of rain gardens to increase infiltration. In our outdoor plant mesocosm experiments, we identified four native plant species that likely will grow well in bioretention areas, including rain gardens, because of their rapid biomass production and their ability to tolerate periods of both flooding and dryness: C. stricta, C. canadensis, Spartina pectinata, and Eupatorium perfoliatium. We also identified other species that were least suitable for rain gardens due to slow growth and inability to tolerate flooding.
Hydrogeologic Research. The physical component of our research project pertains to an analysis of the hydrologic changes that occur in a local watershed as a result of urbanization. Our research has derived the final conceptual model for the region's natural hydrology related to groundwater discharge as springs at the edge of wetlands. An understanding of these controls is essential to the development of numerical models that we will use to evaluate the hydrologic impacts of various urbanization scenarios and to address two other objectives of this project: (1) optimal siting and operation of municipal and other high-capacity wells; and (2) evaluating groundwater quality and quantity tradeoffs between unsewered and sewered subdivisions.
Extensive testing of a deep well borehole drilled near the major springs in our case study watershed (Pheasant Branch) provided stratigraphic information that confirmed the presence of a regional aquitard (shale) separating the upper and lower bedrock aquifers. The effects of pumping by nearby municipal and other high-capacity wells on water levels in the upper and lower bedrock aquifers were evaluated by examining continuous water level records from two zones of the test well separated using an inflatable packer. Comparison of the water level data to records of precipitation and pumping rates for nearby wells revealed that pumping causes frequent cycling of water levels in the lower aquifer, while the upper aquifer has relatively steady water levels that respond primarily to precipitation. These results illustrate the effectiveness of the regional aquitard in isolating the lower bedrock aquifer, and also confirm the hypothesis that steady spring flow rates are maintained by water discharging from the shallow bedrock.
Thus, a decrease in infiltration rates over an urbanized landscape would directly reduce spring flow discharge rates into area wetlands and other water bodies. Because typical urban development designs with their increased areas of impervious surfaces cause a substantial decline in infiltration with a concomitant increase in runoff, the linkage between land use decisions and wetland ecology via altered hydrology is made. We currently are developing alternative urban design scenarios, including one that maintains, as best as possible, natural infiltration rates. The hydrologic impacts of each urban scenario will be tested using the models developed and/or refined for this research project.
Transitional Agricultural Lands Research. Postharvest soil sampling (5 cm depth) in the fall of 2000 provided the basis for mapping P levels across a 1,220 ha (3,028 ac) study region comprising 10 animal feeding operations (AFOs) ranging in size from 92 to 1,600 animal units. The sampling effort was accompanied by on-farm interviews, which helped to establish working relationships with producers for long-term research in the area. The field research on nutrient management and land use revealed trends related to phosphorus (P) management that have implications for agricultural watersheds undergoing urbanization. First, strong sources of P greater than 400 lb ac-1 (Bray P1) have developed at the farm, field, and subfield scales, occupying approximately 15 percent of the watershed study area. These "hot spots" accumulate P due to a legacy of manure additions, establishing the potential for P migration from agricultural fields. High P levels in the soil in urbanizing watersheds create additional risks for nonpoint source pollution at the time of construction site scarification and regrading. Reducing these levels through improved nutrient management will be necessary to prevent large deliveries of P to nearby Lake Mendota. The timeline of urbanization in agricultural watersheds, however, are often much shorter than the time it will take to lower soil P levels such as these down to environmentally sound levels, under current cropping practices and manure management. For instance, with 45 percent of this study area planted to corn as it is now, this area would consume nearly 26 lb P ac-1 of P at a yield of 160 bu ac-1. Dairy farmers, however, will add between 13 and 26 lb P ac-1 yr-1 to corn fields, and may exceed these rates if their animals-to-land ratio is relatively high (<2 AU ac-1).
The second trend observed involves the use of land under different farming structures for AFOs. There is a trend in Wisconsin, as in other areas of the country, for dairy AFOs to expand herd sizes using the confinement model of dairying. Development pressures challenge this prevailing model. In our study region, confinement-style dairying requires producers to own or operate land not only to feed animals, but more relevant to nonpoint pollution, to distribute manure. Some AFOs are able to acquire land in close proximity to the animal housing centers, while other producers are left to find land that is more distant and often external to the watershed. Larger AFOs often rent land, determining rental rates based on personal relationships.
Currently, the study area is not zoned for urban use, but as urbanization encroaches on the region, land that was once usable for manure distribution most likely will be developed, leaving large AFOs no alternative but to haul manure considerable distances.
Watershed Modeling Research. This research project required the development of new hydrologic models, as well as the application and/or modification of existing models. One major objective of the modeling was to provide tools that could be used to design infiltration practices as well as to evaluate the benefits of infiltration practices at the site, development, and watershed scale. In particular, we were interested in evaluating the degree to which rain gardens (bioretention cells) can reduce surface runoff, increase groundwater recharge, and increase flows in springs. We also used a groundwater model to improve our understanding of the flow system that supplies springs in southern Wisconsin.
We developed two new models, each implemented using Matlab, to simulate the continuous hydrologic functioning of rain gardens. Each model can handle a three-layered rain garden, where the top layer is the rooting zone, the middle layer is a high-conductivity storage zone, and the third layer is the native subsoil. We explored rain garden performance as a function of the primary design variables: (1) area, depth of ponding (hd); (2) thickness of the storage zone (ST); and (3) the primary site parameter, subsoil hydraulic conductivity (Kss). Continuous simulations were conducted using hourly rainfall data for Madison, WI, for the months of April through September of 1992-1997. Results indicate that as the rain garden area ratio increases, the runoff (overflow from the garden) decreases to zero. The results for recharge (deep percolation from the bottom soil layer) are more complex. As the area ratio increases from zero, the recharge amounts increase to a maximum, and then decrease. The reason for the maximum in recharge, which was observed in all cases modeled, is that at larger area ratios, evapotranspiration becomes a progressively larger term in the soil column water budget. For simulation results, where ST = 90 cm, hd = 15 cm, and Kss = 1 cm/h, the maximum recharge depth is 45 cm, or about 80 percent of the annual average rainfall for the months of April through September. (Note that the depths represent total volumes divided by the sum of the area of the impervious surface and the rain garden.) This is more than twice the typical recharge rate (19 cm) for the silt loam soils in this region and does not account for any recharge that might occur in the months of October through March, a period that accounts for most natural recharge in the region.
At the development scale, we have developed a spreadsheet model based on Soil Conservation Service (SCS) modeling principles that performs standard SCS runoff calculations and additionally accounts for infiltration practices. Calculations can be made for design events and in a quasi-continuous mode. Two kinds of infiltration practices are handled: (1) spreading of runoff from impervious surfaces onto pervious surfaces; and (2) rain gardens. The former is handled simply by increasing the rainfall on the pervious surface to account for the impervious runoff. The rain garden model is a tank model that accounts for the soil and subsoil hydraulic conductivity. We are distributing our spreadsheet model to local design engineers to receive feedback on its usefulness.
For many planning activities, it is useful to evaluate the benefits of infiltration practices at the watershed scale. We are writing modules for the United States Geological Survey Modular Modeling System (Leavesley, et al., 1996) that will allow for the evaluation of rain gardens and spreading impervious surface runoff onto pervious surfaces. This model will be applied to the North Fork of Pheasant Branch to assess the benefits of infiltration practices in terms of both runoff reduction and infiltration enhancement. Spatially distributed recharge values from this model will be used as input to our groundwater model to assess the benefits of infiltration practices to spring flow.
One groundwater model of the North Fork of Pheasant Branch watershed was constructed using MODFLOW(McDonald and Harbaugh, 1988). The model was created using a telescopic mesh refinement technique, extracting initial layers, parameter distributions, and boundary conditions from a MODFLOW model of Dane County(Krohelski, et al., 2000). Modifications included the addition of model layers, refinement of the grid, addition of nodes to represent previously unrepresented stream reaches and wetlands, modification of the recharge array to correspond to recharge values estimated by Steuer and Hunt (2001) using the Precipitation Runoff Modelling System surface water model, and calibration of hydraulic conductivity values. This model will be used to evaluate the groundwater benefits of infiltration practices.
Social Research. Two areas of our social research relate directly to the barriers or impediments that must be overcome if low-impact development is to be adopted as the norm in urban design. The first set of barriers can be classified as institutional, such as municipal ordinances that restrict any new development to environmentally unfriendly standard practices such as excessive minimum street widths and traditional curbs and gutters to quickly convey runoff water away from the development. Our research project has identified ordinances that are barriers to low-impact development and will make recommendations to eliminate this set of barriers.
The second set of barriers generally can be grouped in the realm of human biases and/or inadequate knowledge by key players. Many of the environmentally unfriendly ordinances discussed above embody the interests and views of various institutional players (e.g., fire department and snow removal personnel). Their reluctance to change reflects the deeper problem that municipal officials (e.g., planners, engineers) and citizen oversight committees in general have little experience with low-impact development designs pertaining to water management. Indepth personal interviews conducted with a wide range of individuals (e.g., municipal officials, regional planners, builders, developers, engineers, and environmental consultants) instrumental in the adoption of alternative storm water management practices elucidated this problem further. These interviews disclosed that the adoption of even a single, simple practice such as a rain garden is potentially complex. Many factors and considerations are important including cost, the physical, institutional and legal environments, and the understanding of various key actors (engineers, builders, developers). Interestingly, homeowners and renters are of scant importance in driving the decision to install these practices in new developments. Decisions about these matters are made early on in the development phase; owners/tenants have little input or impact.
Another finding of our sociological research is a key barrier to the adoption of infiltration practices: the lack of knowledge about the practices and their effectiveness under different conditions. Conventional storm water management practices are well understood; their alternatives are not. Builders, developers, planners, regulators, and municipal officials need to know how these practices can be installed, how well they function in different settings, and how much they cost to install and maintain. Additional research and documentation of the practices can overcome this barrier. In particular, a high priority to funding and implementing demonstration projects of the low-impact practices in a variety of settings is critical to their rapid acceptance.
Although the previous discussion addresses the issues related to implementing low-impact development practices in areas undergoing urbanization, different problems exist for retrofitting these practices in established urban areas. Other than larger scale infiltration systems installed on public lands, individual homeowners and commercial property owners are the key players for implementing rain gardens or other infiltration systems at the scale of individual properties. Through the use of focus group listening sessions conducted in May 2002, we were able to document how residents of Maplewood, MN, accepted the voluntary installation of rain gardens as part of larger municipal projects to alleviate chronic flooding problems in their community. In general, homeowner "gardeners" tended to be younger, liked gardening in general, and often had experienced water problems on their property, as compared to "nongardeners." Voluntary signup brought a greater sense of buyin and prevented unwilling homeowners from having to plant and maintain gardens that were perceived as requiring too much care. Participants found the ideas of "helping your neighbor" and eliminating standing water that was disproportionately affecting some people in their neighborhood as more persuasive reasons for participating than broader appeals for beautifying the community and saving taxpayer costs for managing storm water in the larger watershed. Because few homeowners had any prior knowledge of rain gardens, they had not developed specific attitudes towards rain gardens other than the word "garden" has the connotation of "work." The common mode of reasoning against having a rain garden-"I don't have a water problem, so I don't need a garden"-must be overcome by educational programs. The value of well-organized rain garden demonstration projects is paramount to the infiltration practice's successful implementation by enough homeowners to make a difference in the watershed management of a community.
Future Activities:
We will continue to elucidate the mechanisms and further develop the models used to assess hydrological impacts in relation to various conditions and scenarios, and evaluate and assess groundwater relationships.
Journal Articles:
No journal articles submitted with this report: View all 72 publications for this projectSupplemental Keywords:
watershed, wetlands, stormwater runoff, groundwater, wastewater treatment, thermal pollution, eutrophication, urban development, ecology., RFA, Scientific Discipline, Geographic Area, Water, Ecosystem Protection/Environmental Exposure & Risk, Nutrients, Water & Watershed, Environmental Chemistry, Ecosystem/Assessment/Indicators, State, Wet Weather Flows, Ecological Risk Assessment, Environmental Engineering, Watersheds, nutrient transport, ecological effects, ecological exposure, urbanization, environmental monitoring, aquatic ecosystem, fate and transport, eutrophication, thermal pollution, alternative urbanization scenarios, biodiversity, streams, agricultural watershed, runoff, surface water, aquatic degradation, ecological impacts, eutrophication of lakes, urban development, aquatic ecosystems, water quality, agriculture, nutrient cycling, channel erosion, impervious surface areas, irrigation, Wisconsin (WI), social constraints, well location, aquatic biota, land use, groundwater, stream ecosystem, storm water, agriculture , phosphorous, land managementProgress and Final Reports:
Original AbstractThe perspectives, information and conclusions conveyed in research project abstracts, progress reports, final reports, journal abstracts and journal publications convey the viewpoints of the principal investigator and may not represent the views and policies of ORD and EPA. Conclusions drawn by the principal investigators have not been reviewed by the Agency.