Exchanges across Land-Water-Scape Boundaries in Urban Systems Strategies for Reducing Nitrate Pollution M.L. Cadenasso,a S.T.A. Pickett,b P.M. Groffman,b L.E. Band,c G.S. Brush,d M.F. Galvin,e J.M. Grove,f G. Hagar,g V. Marshall,h B.P. McGrath,i J.P.M. O’Neil-Dunne,j W.P. Stack,k and A.R. Troyj a

b

Cary Institute of Ecosystem Studies, Millbrook, New York, USA c

d

Department of Plant Sciences, University of California, Davis, California, USA

Department of Geography, University of North Carolina, Chapel Hill, North Carolina, USA

Department of Geography and Environmental Engineering, Johns Hopkins University, Baltimore, Maryland, USA e

Maryland Department of Natural Resources, Annapolis, Maryland, USA f

g h

Parks and People Foundation, Baltimore, Maryland, USA

Urban Design Program, Columbia University, New York, New York, USA i

j

USDA Forest Service, Northern Research Station, South Burlington, Vermont, USA

Parsons, The New School for Design, New York, New York, USA

Rubenstein School of the Environment and Natural Resources, University of Vermont, Burlington, Vermont, USA k

Baltimore City Department of Public Works, Baltimore, Maryland, USA

Conservation in urban areas typically focuses on biodiversity and large green spaces. However, opportunities exist throughout urban areas to enhance ecological functions. An important function of urban landscapes is retaining nitrogen thereby reducing nitrate pollution to streams and coastal waters. Control of nonpoint nitrate pollution in urban areas was originally based on the documented importance of riparian zones in agricultural and forested ecosystems. The watershed and boundary frameworks have been used to guide stream research and a riparian conservation strategy to reduce nitrate pollution in urban streams. But is stream restoration and riparian-zone conservation enough? Data from the Baltimore Ecosystem Study and other urban stream research indicate that urban riparian zones do not necessarily prevent nitrate from entering, nor remove nitrate from, streams. Based on this insight, policy makers in Baltimore extended the conservation strategy throughout larger watersheds, attempting to restore functions that no longer took place in riparian boundaries. Two urban revitalization projects are presented as examples aimed at reducing nitrate pollution to stormwater, streams, and the Chesapeake Bay. An adaptive cycle of ecological urban design synthesizes the insights from the watershed and boundary frameworks,

Address for correspondence: M.L. Cadenasso, Department of Plant Sciences, Mail Stop 1, PES 1210, One Shields Ave, Davis. CA

Ann. N.Y. Acad. Sci. 1134: 213–232 (2008). doi: 10.1196/annals.1439.012

 C

95816. Voice: +1-530-754-6151; fax: +1-530-752-4361. mlcadenasso@ ucdavis.edu

2008 New York Academy of Sciences. 213

214

Annals of the New York Academy of Sciences from new data, and from the conservation concerns of agencies and local communities. This urban example of conservation based on ameliorating nitrate water pollution extends the initial watershed-boundary approach along three dimensions: 1) from riparian to urban land-water-scapes; 2) from discrete engineering solutions to ecological design approaches; and 3) from structural solutions to inclusion of individual, household, and institutional behavior Key words: urban; cities; watershed; boundaries; nitrate; water; pollution; land cover; urban design; Baltimore, Maryland; riparian; nonpoint source; restoration

Introduction Urbanization is one of the most conspicuous facets of global environmental change (Vitousek et al. 1997b). How do the familiar frameworks and expectations from the ecology of nonurban systems apply in urban areas? How can improved ecological understanding of urban areas be used to promote conservation, both within and beyond cities and suburbs? These two questions broadly motivate this paper. We focus on two frameworks—watershed and boundary—as important concepts from nonurban ecology that may be applied in cities and suburbs. We describe how the watershed and boundary frameworks have guided the understanding of nonpoint stream pollution by nitrate in Baltimore, Maryland. The experience of agricultural and forested ecosystems led us to expect that riparian zones would be key locations within urban watersheds for converting nitrate via denitrification, and hence reducing the nitrate export from these complex watersheds. However, we discovered that the urban riparian zones did not reduce the amount of nitrate present in the streams, but that they actually served as sources for nitrate. The altered functions of urban riparian zones reflect altered urban hydrology, lowered ground water tables, and lower amounts of soil carbon. While these alterations can be understood within the boundary and watershed frameworks, we think that a spatially extensive and dispersed conception of how boundaries function within urban watersheds is more appropriate than the traditional view of riparian boundaries as spatially discrete locations within watersheds. Not only are streamside riparian

zones altered in urban systems (Groffman et al. 2004), but the alteration of hydrological flow paths disassociates the “riparian” zone from the stream. The tight spatial link between the structure and function of riparian zones in undeveloped landscapes is not representative of highly altered urban watersheds. The alteration of land-water-scapes in urban areas, therefore, requires a new conceptualization of riparian zones as spatially dispersed throughout the watershed. When this insight is placed in a management context, spatially extensive interventions in urban watershed functioning are suggested. We use ongoing restoration projects in Baltimore to illustrate how the watershed and boundary frameworks are combined in spatially new ways in practice. We close with a generalized approach that incorporates such insights into ecological urban design that can support the goals of urban conservation and reducing the pressure of urban land conversion on surrounding regions. We begin with the role and practical significance of nitrate in urban areas and move through the remainder of the points outlined above.

Nitrate Pollution and Conservation in Urban Systems The relationship of land and water in urban areas is a complex physical, ecological, and social interaction. Most of the world’s cities are located near water—both rivers and coastal waters. Water is crucial for drinking and food production but also for transportation and industry. Much of the history and development

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments

of urban areas is tied to the availability of water to transport goods, generate energy, and remove wastes. Water pollution, therefore, has plagued urban areas since their beginning and, once it was recognized as a human health concern (e.g., Johnson 2006), many strategies to prevent or minimize pollution through regulation, diversion, and treatment have been employed (Melosi 2000). Pollution sources are defined as either point or nonpoint. Point sources include those with a discrete identifiable location. Nonpoint-source pollution is from many diffuse sources, which makes it difficult to prevent and to regulate. Urban runoff is a common type of nonpoint-source pollution and is our focus here. Though concern with water pollution is common to all cities, we use the case of the Baltimore, Maryland metropolitan region to examine the nature and control of the complex relationship of land and water. Of the many potential urban pollutants, we focus on nitrate because it is a particularly mobile form of nitrogen, which is a key limiting nutrient in ecosystems. Nitrate is thus a critical pollutant that enters receiving waters from both point and nonpoint sources, and it has been identified as the most common groundwater pollutant in the United States by the USEPA (1990). Nitrate is a natural by-product of the nitrogen cycle, however, human activities now produce more than twice the amount of nitrogen that is fixed through natural processes including lightening and bacterial fixation (Vitousek et al. 1997a). Nitrate pollution is added to systems through human use of water for waste removal, the addition of fertilizer, and the combustion of fossil fuels. Pathways of release include leaky sanitary sewers, treated water released from septic systems, industrial spills, leaching from contaminated lands such as landfills and gasworks, stormwater runoff from roads, and deicing at airports (see Bernhardt et al. 2008). Atmospheric deposition of nitrate derived from automobile exhaust is a major urban pathway and is also important downwind of cities. Coastal waters, which are disproportionately

215

the location of large and growing urban areas worldwide (UN Population Fund 2007), are limited by nitrogen. Thus, they are particularly susceptible to nitrate-induced eutrophication (Carpenter et al. 1998; Diaz 2001; Howarth & Marino 2006). Nitrate exports are elevated in urban and suburban ecosystems relative to the natural ecosystems that they replace because land-use change results in increases in “sources” and decreases in “sinks” for nitrate (Kaye et al. 2006). Increased sources of nitrate include local hotspots of atmospheric deposition associated with roads (Weathers et al. 2001; Lovett et al. 2002), fertilizer use on lawns (Law et al. 2004), and sewage disposal, either through onsite septic systems, wastewater treatment plant discharges, or leaks from sanitary sewer lines flowing to treatment plants (Wakida & Lerner 2005). Urban land-use change decreases sinks for nitrate in riparian zones (Groffman et al. 2002) and streams (Paul & Meyer 2001). The result is that urban streams consistently exhibit nitrate concentrations well above those of native ecosystems, but generally lower than those of more heavily fertilized agricultural ecosystems (Jordan et al. 1997; Miller et al. 1997; Weller et al. 2003; Groffman et al. 2004; Wollheim et al. 2005). Baltimore is ideal for examining the mechanisms of nitrate pollution and evaluating alternative strategies for minimizing the impact of nitrate pollution for two reasons. First, it is located on the banks of the Chesapeake Bay, which is the largest estuary in the United States and has the greatest land-to-water ratio of any estuary in the world (Sprague et al. 2006). Nitrate is of great concern in the Chesapeake (Boesch et al. 2001), which has been declared “impaired waters” by the U.S. Environmental Protection Agency (EPA). This designation is accompanied by a mandate to decrease nitrate loadings into the bay by 40% by the year 2011 or face regulatory action (Koroncai et al. 2003). Point sources such as sewage treatment plants are being addressed through planning and management practices with some

216

success (Chesapeake Bay Foundation 2003). Nonpoint sources of nitrate, which include runoff from the surrounding urban and agricultural landscape into the bay, are much more difficult to control. Addressing nonpoint sources of nitrate comprehensively in the system is an ongoing concern (NRCS 2006). Second, Baltimore hosts the Baltimore Ecosystem Study (BES) Long Term Ecological Research (LTER) program (www.beslter.org). This National Science Foundation–funded effort is multidisciplinary and multiscalar and involves more than 50 academic, governmental, and nonprofit scientists, managers, and educators. The theoretical, empirical, and management efforts of the BES program provide the basis for this review and synthesis. Nitrate export across land-water-scape boundaries is the result of many processes and activities on the landscape. We use the term “land-water-scape” to emphasize the ubiquitous spatial and functional integration of land and water in urban areas due to the alteration of land cover and the introduction of infrastructure. The exchange of nitrate across the complex terrestrial–aquatic boundaries of urban areas needs to be understood in order to develop conservation strategies for reducing nitrate pollution into the Chesapeake Bay. Urbanized and urbanizing landscapes are estimated to contribute a significant amount of nitrate loading to the bay from its catchment (Koroncai et al. 2003), which includes parts of six states and the District of Columbia. Frameworks for Analysis of Urban Nitrate Pollution We use the watershed and ecological boundary frameworks to analyze the problem and to evaluate potential interventions. The watershed framework is relevant because of the integrating capacity of hydrological catchments (Naiman & D´ecamps 1997). This feature of watersheds is widely recognized in ecology and has been the mainstay of such research programs as the Hubbard Brook Ecosystem Study

Annals of the New York Academy of Sciences

(Bormann & Likens 1994) and the Coweeta hydrological laboratory (Swank & Douglass 1977), where stream outputs are monitored to assess important processes and changes within forested ecosystems. Though the watershed framework has been more commonly used in forested and agricultural landscapes to evaluate water quality (Nikolaidis et al. 1998; Valiela et al. 2000; Jones et al. 2001; Wickham et al. 2002; Wayland et al. 2003), the watershed framework applies to urbanized areas as well (Pickett et al. 1997; Pickett et al. 2007). Applying the watershed concept to urban areas must recognize the imposition of infrastructure such as curbs and storm drains (Paul & Meyer 2001). Urban stream or storm drain flows reflect the upstream sediment, pathogen, or pollutant inputs in these complex catchments. In the case of nitrate, the scientific approach to the watershed is to measure and understand the controls on nitrate flow and retention in the catchment ecosystem, while application aims to reduce nitrate release from the watershed. Of course, in urbanized watersheds, changes in land cover or human behavior become paramount controls (Pickett et al. 1997; Groffman et al. 2004; Cadenasso et al. 2007). The ecological boundary framework (Cadenasso et al. 2003) is also relevant to the issue of nitrate pollution in urban regions. The boundary framework conceives of a landscape as being organized into patches, which are spatially delimited areas of contrasting structure or function. The boundary framework also recognizes the flows between patches, which may, in general terms, consist of organisms, materials, energy, or information. The final component of the boundary framework is the boundaries themselves. Boundaries are areas of contact between patches and appear as structural or process gradients. Therefore, boundaries can be abrupt or gradual depending on the steepness of those gradients. For example, a structural boundary exists at a forest–field edge. This same location can also represent a process gradient, as in the case of wind velocity, which

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments

drops precipitously across the boundary from field to forest interior (Chen et al. 1995), or dissolved inorganic nitrogen flux, which is enhanced on the field–forest boundary (Weathers et al. 2001). In order to assess the function of boundaries, both the degree of contrast between patches and how boundaries control flows between patches must be known (Cadenasso et al. 2003). In watersheds, the riparian zone is generally considered to be the boundary between terrestrial and aquatic systems. Riparian zones are distinct features of the landscape, defined as streamside habitats. They have been recognized as “hotspots” of biological and chemical activity (Naiman & D´ecamps 1997; Ewel et al. 2001; McClain et al. 2003). It has been suggested that riparian forests function as buffers that prevent nutrient export from the terrestrial system into receiving waters (Peterjohn & Correll 1984; Naiman et al. 2005; Mayer et al. 2007). This has resulted in a common management strategy of planting riparian buffer strips to prevent nutrient enrichment of adjacent waters (e.g. Naiman & Bilby 1998; Naiman et al. 2005). The removal of nitrate in riparian boundaries is the result of denitrification. In the next section, we address insights about the connections between landscapes and receiving waters in an urban system and explore how scientific knowledge about boundary function informs interventions to control nitrate pollution. Patches, Flows, and Boundaries in Metropolitan Baltimore How well the watershed framework and the boundary framework apply to the problem of nitrate pollution in urban systems is based on the degree to which cities and suburbs share structural characteristics with the forest and agricultural systems in which those frameworks have been most commonly used. To assess the congruence of the frameworks and the urban environments, we give a brief overview of

217

the nature of urbanized patches, the key features of urban hydrology as a case of betweenpatch flow, and the structure and function of urban riparian zones, which are the boundaries between the streams and the upland patches of watersheds. Each of the three framework components—patches, flows, and boundaries—varies in an urban landscape relative to the traditional view of forest or agricultural field and hydrological flows traversing riparian boundaries. Urban upland patches need to capture the fine-scale heterogeneity characteristic of the system, and the flows of water must account for the alterations to flow paths that occur in urban lands. Urban infrastructure introduces new types of land–water boundaries and must be included. Urban Pattern: Delineating and Characterizing Land Cover Patches Urban patches will act as either the sources or sinks of nitrate within and around cities. Patches in urban systems may be characterized by distinguishable areas of land cover. Of course, there are other features of urban landscapes, such as social and economic processes (Gottdiener & Hutchison 2000; Grove et al. 2006), that must also be seen as criteria for the characterization of system heterogeneity. However, land use/land cover is frequently used as a fundamental criterion for urban heterogeneity (Pauleit & Duhme 2000; Alberti et al. 2003; Band et al. 2005), and here we focus on land cover. Note that patches do not have to be homogeneous, but only must contain heterogeneity that is distinct from the patterns of adjoining patches at a specific scale. The delimitation of urban patches has been problematic because landcover models that are often applied to urban systems make assumptions that constrain their utility in that context. Many of these models are derived from a system that was designed to represent use of natural resources and commodities at the scale of the entire United States (Anderson et al. 1976) and to contrast those

218

uses with urbanized land uses. Anderson and colleagues (1976) and the systems derived from it (e.g., DiGregorio & Jansen 2001) assume that human and nonbuilt covers are distinct features of landscapes and, consequently, they are kept separate in the classifications. But this does not allow for an accurate depiction of integrated urban landscapes. For example, not all lands usually classified as residential may be ecologically similar, even when differentiated into low, medium, and high density of people or structures. Of course, density of housing units likely correlates with proportion of the land in impervious roof or pavement material. However, other features of the land that may affect such ecological processes as nitrate retention may differ greatly within a given density of residential buildings (Cadenasso et al. 2006, 2007). For example, the presence or density of trees and the presence and amount of turfgrass or other herbaceous vegetation may differ considerably, and may do so independently of the housing type or density. The predominant scale of application is one practical assumption of Anderson-based systems that becomes a shortcoming when applied to urban areas. Social scientists, geographers, and urban designers consistently note the existence of fine-scale heterogeneity as a conspicuous feature of urban areas (Jacobs 1961; Clay 1973; Shane 2005). Cities, suburbs, and towns are well known for blockto-block differences in architecture, landscape plantings, and gardens. Even new subdivisions differ conspicuously from one another in layout, density, and vegetation. Commercial, industrial, and transportation corridors of different style and configuration cut across the grain of the rest of the settlement, or define its predominant dimensions (Forman et al. 2003). Across all of these contrasts, the type, intensity, and frequency of landscape management may differ, and the level of investment in repair, renovation, or replacement of buildings and infrastructure may differ. Many of these physical and ecological differences are related to social structure, such as demography, ethnicity,

Annals of the New York Academy of Sciences

income, lifestyle, and household size, composition, and lifestage, among many other factors (Gottdiener & Hutchison 2000). In spite of the large number of potential and actual fine-scale differences in urban areas, such as those listed above, the standard classifications do not account for many of them. However, we hypothesize that such fine-scale heterogeneity may have functional ecological implications. The existence of clear, fine-scale heterogeneity in urban areas and the growing understanding of the role of spatial heterogeneity in ecology in general (e.g., Kolasa & Pickett 1991; Hutchings et al. 2000; Lovett et al. 2005) call for a new, ecologically relevant way of modeling spatial heterogeneity in urban areas and watersheds to improve understanding of such functions as nitrate retention. A system developed by Cadenasso et al. (2007) can be used to exemplify a fine-scaled, ecologically focused approach to urban heterogeneity. This new classification, called High Ecological Resolution Classification for Urban Landscapes and Environmental Systems (HERCULES), focuses on the cover of land in urban areas, ranging from old central business districts to the emerging exurban fringe. By focusing on land cover, and not including land use, HERCULES models can be used to test how cover is related to function. Of course, land use and other spatial data can be added as a data layer if a specific research question requires it. Another key feature of HERCULES is its integrated nature. HERCULES incorporates vegetation structure and cover, the cover of buildings, and the cover of surfaces including pavement and bare ground. These are the three kinds of coverages that can exist throughout urban areas (Ridd 1995), and HERCULES allows them to vary independently of each other. (See Cadenasso et al. 2007 for a full description of the HERCULES land-cover model.) This flexibility avoids some of the constraints of the standard urban classifications, which separate vegetation from the built component of the environment (e.g., Anderson et al. 1976). Because

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments

HERCULES helps clarify the structure of urban heterogeneity, and does so in a way that integrates built and natural components, it facilitates understanding the fine-scale structure of urban watersheds. This more accurate ecological depiction of patches facilitates the quantification of nitrate storage and export from the landscape and furthers the exploration of management and conservation strategies. Regardless of what precise land-cover model is used to describe urban patches, social-ecologicalphysical integration and fine-scale heterogeneity must be accounted for (Cadenasso et al. 2006). Urban Fluxes: Hydrology as the Medium for Nitrate The boundary framework in urban areas must also address the nature of the flows from patch to patch. Because our focus is on nitrate as both an indicator parameter and a variable of practical management concern in receiving waters, we focus on water as a primary medium of transport of nitrate in urban systems. Hence, we must examine the hydrology of urban systems. Fortunately, there is much literature on this phenomenon (e.g., Paul & Meyer 2001; Strayer et al. 2003; Walsh et al. 2005b; Bernhardt et al. 2008), and it is a core aspect of the work of the BES. The land cover alteration from the historic forest and agriculture cover to urbanized land use leads to changes in the timing, quantity, location, and quality of water across the urban land-water-scape (Band et al. 2005; Goonetilleke et al. 2005). Urban streams are extraordinarily “flashy,” exhibiting very rapid increase in flow following storms, a higher peak than nonurbanized streams in the same region, and a rapid decline to low flow (Walsh et al. 2005b). In addition, urban hydrographs often have lower levels of base or persistent flow than nonurbanized streams (e.g., Rose & Peters 2001; but see Meyer 2005). The typical urban hydrograph is generated in part, by the large proportion of ground covered by

219

impervious pavement and buildings (Paul & Meyer 2001). Urban infrastructure, engineered to drain stormwater rapidly, contributes to the flashiness of urban streams by delivering runoff from surfaces distant from the stream channels quickly and directly through pipes. The increase in imperviousness, changes to the hydrological flow path, and the potential increase of sources of nitrate in the urban landscape may all contribute to a decrease in water quality (Arnold & Gibbons 1996; Hatt et al. 2004; Walsh et al. 2005a, 2005b; Atasoy et al. 2006). The high erosive force of urban flash floods often results in highly incised stream channels, which reduce interaction between the stream and the riparian zone by limiting overbank flows and lowering riparian water tables (Groffman et al. 2003). The biogeochemical implications of these engineered changes, and the potential to reverse them with designed ecological features are discussed below. Urban Boundaries: Terrestrial—Aquatic Interfaces Watershed studies in forested and agricultural landscapes suggest that riparian zones, with their expected high levels of denitrification due to their vegetation composition, microbial populations, soil conditions, and hydrological regimes, are the most relevant boundaries to examine for the control of nitrate in urban streams (Peterjohn & Correll 1984; Naiman & D´ecamps 1997; Lowrance 1998). This assumption has proven to be faulty in urban systems (Groffman et al. 2004; Walsh et al. 2005a; Craig et al. 2008). The structure of urban riparian zones is usually highly altered, in part due to the hydrological characteristics reviewed in the previous section. Stream incision and erosion of the banks often strand former floodplains high above all but the most unusual depths of stream flow. Storm drain pipes, piping and burying of streams, the installation of steep curbs, and downspouts connected to storm drains all reduce the interaction of stormwater with the soil. The reduced infiltration into urban soils

220

and the rapid drainage of stormwater leads to lowered water tables, so that riparian soils are no longer wet for long periods of time (Groffman et al. 2004). Therefore, urban groundwater tables are expected to be deeper and to be recharged more slowly than nonurban subsurface waters. Hence, urban riparian zones are drier than nonurban ones (Sukopp 1998). Drier conditions lead to altered riparian forest composition, and urban riparian zones have been found to support twice the number of upland plant species as nonurban riparian zones in the Piedmont region of Maryland (Brush et al. 1980; Brush & Bain, in prep.). The replacement of wetland species, such as Acer negundo (box elder), by upland species, such as Robinia pseudoacacia (black locust), suggests that the floodplains are undergoing a “hydrological drought,” which would result in a change from anaerobic to aerobic conditions (Brush, in prep.). Aerobic conditions reduce the potential for denitrification, an anaerobic microbial process that converts nitrate into nitrogen gases, and increase the potential for nitrification, an aerobic microbial process that produces nitrate. Loss of organic matter in urban riparian boundaries to aerobic decomposition robs anaerobic denitrifying bacteria of their energy source. Indeed, under some conditions, urban riparian zones can become sources of nitrate to streams (Groffman et al. 2002, 2003; Groffman & Crawford 2003). Policy in the Chesapeake Bay region has employed riparian revegetation as one of its restoration strategies (Chesapeake Bay Program 2000). The conclusion that urban riparian zones are altered to such a degree that they no longer function as nitrate sinks has stimulated managers and policy makers in Baltimore City and in the State of Maryland, to pay increased attention to strategies beyond the traditionally recognized riparian boundary for controlling stormwater flows and their contamination by nitrate (Pickett et al. 2007). In addition, the drying of the riparian soil and consequent shift in the plant commu-

Annals of the New York Academy of Sciences

nity from riparian-adapted to upland-adapted species further suggests that solely focusing on the riparian zone to enhance nitrogen retention in the urban watersheds is not adequate. Strategies that focus on patches distant from the major streams and recognize that land– water interfaces are now more spatially dispersed and located throughout the watershed, not just along the streams, have gained importance (Walsh et al. 2005a; Bernhardt & Palmer 2007; Pickett et al. 2007). The scientific and management question then becomes: What structural and functional features of urban patches beyond the stream can be altered or exploited to reduce nitrate pollution. Strategies for Decreasing Urban Nitrate Loading into the Chesapeake Bay The mandate to reduce nitrate and other water pollution to the Chesapeake Bay remains a compelling goal (Chesapeake Bay Commission 2006). The unexpected failure of urban riparian boundaries to contribute to nitrate reduction in urban streams has highlighted the need to develop and apply new strategies. An option is to shift attention from controlling nitrate via the existing riparian zones of urban streams to seeking replacements for this boundary function elsewhere in urban watersheds. Both behavioral and structural alterations are possible. Altering human behavior is one broad approach, which we briefly mention for the sake of completeness. It is beyond our scope here to present details of social processes. Examples include individual or household behaviors, such as nitrogen fertilization, automobile usage, consumption and disposal patterns, and management of pet waste (Grimm et al. 2003; Law et al. 2004; Baker et al. 2007). Education to reduce dumping in storm drains can also play a role in influencing individual behavior (Dietz et al. 2004). The behavior of institutions is similarly difficult to

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments

adjust, although street sweeping, parkland mowing schedules, and vacant lot rehabilitation are examples of municipal actions that may influence nitrate loading to streams (W.P. Stackpers. observ.). In fact, a street sweeping study in Baltimore City found that an appreciable portion of the contaminant loading could be associated with particulates greater than 0.25mm, which includes sand grains and leaf litter (W.P. Stackpers. observ.). We will concentrate on the second strategy to decrease nitrate loading— alteration of the biophysical structure of urban watersheds. Stream and Riparian Restoration Stream networks and riparian zones in urban areas have been highly modified, as mentioned earlier. Although strict restoration to a preurban condition is not a realistic goal, restoration in the urban context implies mitigation of ecological, hydrological, and aesthetic functions. We use the familiar term “restoration” in this context as the aim is to restore the ecological function of riparian zones as sinks rather than sources of nitrate for the receiving waters. In some cases the originally damaging modification was unintentional, as with down cutting and the isolation of streams from floodplains. In other cases the modification was intentional, as when streams were buried, piped, straightened, or channelized. In all of these situations, the internal heterogeneity of the stream, which can facilitate denitrification or permit nutrient uptake by biological processes in stream waters, has been damaged or removed entirely (Paul & Meyer 2001). Stream restoration can alter the interface between terrestrial systems and the drainage network in ways favorable to nitrate retention or absorption (Bukaveckas 2007; Craig et al. 2008; Kaushal et al. 2008). Increasing stream sinuosity, especially if this encourages pockets of organic matter in saturated sediments, provides important anaerobic substrates in which denitrification can occur. (See Bernhardt et al. 2008, for a discussion on the

221

tradeoffs between denitrification and atmospheric NO 2 generation.) “Daylighting” buried streams, or removing pipes, permits biological processes that can act as nitrate sinks to again occur in the restored stream reaches. In Baltimore City, records indicate that 90% of the originally present first-order drainages have been replaced by stormwater drainage pipes. Restoring streams and riparian zones is a classic approach to enhancing the functionality of the terrestrial–aquatic boundary (Bernhardt et al. 2005; Palmer et al. 2007). This approach, however, assumes that the boundary is discrete and easily identified. The alterations to the hydrological flow paths in the urban land-water-scapes require that a more sophisticated conceptualization of terrestrial–aquatic boundaries be recognized which greatly enhances opportunities for nitrate retention strategies (Walsh et al. 2005a; Bernhardt & Palmer 2007). The goal of most urban stream restoration projects is geomorphic stabilization to protect property and infrastructure, for example, sanitary sewer lines in the near-stream zone (Bernhardt et al. 2005; Hassett et al. 2005). While there is great interest in the effects of stream and riparian restoration on nitrogen retention, there has been very little evaluation of nutrient dynamics in restored streams (Bukaveckas 2007; Craig et al. 2008). There are significant uncertainties about the ability of geomorphic restorations to create the conditions necessary for significant amounts of denitrification, especially at high flows, when the residence time of nitrate-laden water in anaerobic, denitrifying sites is likely to be short or nonexistent. A second, and complimentary, need, therefore, is to retrofit catchments to slow and retain the water, resulting in a less flashy hydrograph. In BES, site-specific evaluations of the effects of stream restoration on nitrogen removal are coupled to whole-watershed studies of urban hydrology and nutrient dynamics. Recent results (Shields et al. 2008) focus on quantifying the nature and extent of nutrient export

222

under different flow regimes and on identifying subwatersheds where significant amounts of nitrogen are exported under relatively low, baseflow conditions. These sub watersheds are the most promising sites for stream restoration to significantly reduce nitrogen export, as geomorphic restorations are most likely to create favorable conditions for denitrification under low flow. Particularly promising are areas of low-density residential development served by septic systems. These areas have very high nitrate concentrations in streams (Groffman et al. 2004) but lack the extensive hydrological alteration common in more dense developments. As a result, significant amounts of nitrate are exported from these areas at the low-flow conditions that are amenable to improvement by stream restoration. Site-specific stream restoration work in BES has been centered on Minebank Run, which was restored by Baltimore County using geomorphic approaches to stabilize banks and reconstruct highly incised streambeds (Rosgen 1994, 1996). Work at the site has included comparison of restored upstream reaches with degraded downstream reaches, as well as, pre- versus post-restoration analyses of stream and hyporheic carbon and nitrogen dynamics, including denitrification (Mayer et al. 2003; Groffman et al. 2005). Results suggest that restoration can create hotspots of denitrification where the stream–riparian interface has been modified to facilitate interaction of stream water with the riparian zone (Kaushal et al. 2008). These modifications are common in geomorphic restorations, as stream water is directed out of the channel into the riparian zone to reduce runoff energy to protect downstream reaches from physical degradation. These results raise questions about altering stream and riparian restoration designs to facilitate more stream–riparian interactions that create more effective denitrification hotspots (Gift et al. 2008). Answering these questions will require collaborations among restoration designers and engineers, stream ecologists, and social scientists (Palmer et al. 2007).

Annals of the New York Academy of Sciences

New Functional Interfaces in the Drainage Network New functional interfaces between stormwater sources and streams are one important aspect of the nitrate-management strategy. In essence, the strategy is to slow down stormwater and increase its interaction with ecologically active substrates that might remove nitrate and other pollutants. This contrasts with the traditional engineering approach to stormwater, which aims to remove the water from the urban system as quickly as possible (Melosi 2000). In general, however, the distribution of nitrogen export is heavily weighted towards the few, very high-flow events of the year (Shields et al. 2008). Because under these circumstances the ability to gain the residence times needed for denitrification may be very limited, additional strategies to retrofit the catchment to slow or retain more of the water will also need to be implemented. Several structures or practices can slow and ecologically engage stormwater. Detention basins reduce the rate of stormwater delivery to streams and can support denitrification to replace activity lost in riparian zones affected by urban hydrological changes (Zhu et al. 2004; Hogan and Walbridge 2007). Home gutter systems can be designed to store some roof runoff and to disperse water to lawns and gardens rather than shunting all such water to storm drains. Level spreaders are engineered structures designed to slow water and spread it over a larger area at the same elevation (Hathaway & Hunt 2006). Roads constructed higher than surrounding lawns and landscaping can use level spreaders to move water through curb cuts into adjacent permeable areas. Bioswales are attractively vegetated ditches that can absorb rainwater, increase infiltration, and reduce the amount of surface runoff (Li et al. 1998). In old urban neighborhoods where there is little yard space available for major ecological redesign, alleys and vacant lots can be redesigned to help manage stormwater. Although little used at this time in the Baltimore region, green roofs have

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments

proven their ability to contribute to the effective reduction of stormwater (Mentens et al. 2006; Oberndorfer et al. 2007). These should also help reduce nitrate input into stormwater. The key to the ability of new functional interfaces to serve as hot spots for denitrification in urban watersheds is for them to capture nitrate-laden water and then hold it long enough under the anaerobic, high-carbon conditions suitable for denitrification to occur. This is quite a challenge given the need to avoid flooding of property. Many stormwater treatment strategies, for example, detention basins and bioswales, focus on fostering infiltration of stormwater. While this has the benefit of recharging groundwater and reducing stormflow, it reduces the opportunities for denitrification. Furthermore, there are uncertainties about the effects of this infiltration on groundwater quality. Some stormwater controls, such as bioretention ponds, are also designed to enhance the removal of nitrogen. Design and testing of bioretention ponds for achieving that goal is an active area of research (e.g., Hsieh & Davis 2005; Hunt et al. 2006). Balancing these constraints, and evaluating the effects of multiple small functional interfaces at the whole-watershed scale, is an ongoing science and technology challenge (see discussion of the BES W263 project below). There is also active research in BES to evaluate the effects of stormwater chemistry, such as high salt content, on denitrification in functional interface features (Hale & Groffman 2006). Urban Tree Canopy Goals The fact that so much urban drainage has been highly degraded and does not move through riparian land areas before reaching the receiving waters has encouraged managers and policy makers in the Baltimore region to consider a variety of activities beyond the riparian zone to provide functional options to reduce nitrate and to obtain other ecological benefits. For example, Baltimore County has a long tra-

223

dition of encouraging and assisting forest management within and across parcels. Indeed, the county established an urban–rural demarcation line in 1967 (Baltimore County Office of Planning 2006). This has proven to be an important policy instrument for maintaining ecologically active open space and promoting forest sustainability as indicated by its continuing role in the Baltimore County master plan (Baltimore County 2000). The county recognizes both carbon sequestration and water quality as conservation goals (Baltimore County Department of Environmental Protection and Resource Management 2007). We use the Urban Tree Canopy (UTC) program as an example of the role of greening in enhancing watershed function. This program aims to increase tree canopy cover in Baltimore City and was adopted by the City in response to community initiated greening activities and a voluntary greening goal suggested by the Chesapeake Bay Program. Baltimore City has an average tree canopy cover of 20%. The mayor’s office has stipulated a goal to increase the tree canopy cover to 40% over 30 years (Baltimore City 2006). Formation of this goal was made possible by the UTC assessment conducted by BES scientists in collaboration with local partners using high-resolution remotely sensed imagery (Galvin et al. 2006; Raciti et al. 2006). The UTC assessment process identified not only the amount of tree canopy currently present (termed existing UTC), but also the amount that is biophysically feasible given the constraints of buildings and impervious surfaces (termed possible UTC). The precision of the UTC assessment allowed existing UTC and possible UTC estimates to be computed at the individual parcel level, and results indicate the importance of private lands for achieving the canopy goal (Troy et al. 2007). BES scientists are also examining neighborhood sensitivity to environmental issues, economic resources, social cohesion, and other factors that may affect the success or desirability of tree plantings on public lands in different neighborhoods. In addition, sorting out

224

Annals of the New York Academy of Sciences

the potential for tree planting on public versus private lands refines the prioritization for planting to private property (Troy et al. 2007). This result suggests that many more institutional partners are needed to achieve the city goal. Similar initiatives following the same methodology are being conducted in New York City, NY; Washington, DC; Boston, MA; Pittsburgh, PA; Annapolis, MD; Burlington, VT; Des Moines, IA; and Tampa, FL. Many of the UTC projects are coordinated by the Urban Ecology Collaborative (http://www. urbanecologycollaborative.org/uec/), a partnership operated through nonprofit organizations in each city. Stormwater management and reduction of nitrate pollution are shared goals among these cities. Neighborhood Restoration and Stormwater Improvement An additional example of joint research and application targets a storm drain watershed, labeled number 263 (of 355) in Baltimore City. This 345 ha area is ultimately drained by a 10ft diameter pipe that empties into the Middle Branch of the Patapsco River, which is a tidal arm of the Chesapeake Bay at Baltimore. The Watershed 263 Demonstration and Restoration Project (http://www.watershed263.org) aims to have all city departments work together with community nonprofits to improve environmental and social benefits in the 11 neighborhoods of the storm-sewer-shed. The School Board and departments of Recreation and Parks, Public Works, Transportation, and Housing are all engaged. Based on BES results reviewed above, the project recognizes that in these center-city neighborhoods addressing a traditional riparian zone for nitrate control is futile, due to the highly modified nature of the drainage network that long ago buried surface streams to accommodate street grid development. Instead, the city Department of Public Works, the Parks & People Foundation, which is a community-focused nongovernmental orga-

nization, the Community Watershed Council, and researchers from the BES and the U.S. Forest Service have initiated a multifaceted, multiyear neighborhood revitalization project. Mitigation and revitalization efforts in Watershed 263 include cleaning and planting grass and trees in vacant lots, removing unneeded pavement from the 30% of land in public ownership in the watershed, greening schoolyards, planting street trees, and installing raingardens and other vegetation-based biofiltration systems along alleys where the topography is appropriate (Richardson 2006). In addition, the project helps identify and work with high-nitrate-producing private properties to reduce their stormwater pollution loading. The water quality monitoring component of the study focuses on two subcatchments of the storm drainage network of equal size. One of these catchments is currently undergoing the mitigation activities mentioned above, and mitigation will begin on the other at a later date. This allows for comparison between the currently altered and a control catchment. Both of these catchments are instrumented with automated water samplers in the storm drain outlet at the lowest accessible spot in each, with weekly sampling for water chemistry (www.beslter.org), temperature, and pathogens. This project is unusual in having been proposed by the city Department of Public Works and the Parks & People Foundation as a result of conversations with the scientists in BES, who shared their discoveries about the shortcomings of the urban riparian zones as sinks for nitrate. The combination of scientific results, management significance, and increasing the scale of how boundary functions are conceived in urban watersheds presents an opportunity for a practical synthesis. The need to intervene beyond the traditional limits of riparian boundaries in urban watersheds suggests an ecological involvement in the design process that can replace or compensate for the lost functions that would normally ameliorate nitrate concentration in urban streams. We construct a new design framework that can both guide hypothesis

225

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments

testing about urban watershed function and integrate ecological knowledge into designs that can be widely deployed in urban watersheds. Feedback between Structure and Function: An Urban Ecological Design Cycle The complexity, yet tractability, of both understanding the biophysical controls on nitrate flux and managing it for local and Chesapeake Bay improvement is illustrated by a “cycle of ecological urban design” (Pickett & Cadenasso 2008). The cycle is a synthetic framework, drawing on the insights discussed above. The synthetic framework aims at testing the hypothesis that landscape structure and vegetation management affect the nitrate loading or retention in different watersheds in the Baltimore region. It advances understanding of how biophysical processes affecting nitrate flux and various social processes feed back to affect one another. Furthermore, how this feedback is affected by creative urban designs that incorporate best management practices for stormwater and nitrate mitigation can also be revealed. The cycle of ecological urban design is being used to guide current collaborative research. How it operates is as described below (Fig.1). The cycle begins with a focus on effectively representing current land cover and the associated practices of vegetation management in that landscape. The kind of patch mosaics and heterogeneity discussed earlier in the paper, represented by the highly resolved categorical land cover model HERCULES, represent this step of the feedback loop. Based on such categorically refined descriptions of watershed heterogeneity, spatially explicit, fine-scaled hydrological models are developed to quantify the controls on current nitrate flux from the existing urban landscapes. Applying this approach to contrasting neighborhoods, such as center-city residential areas, early 20th century suburbs, and the currently changing suburban fringe, will indicate how different existing land-

External Policy

Landscape Structure & Management

Neighborhood Preferences

Bay water quality. Neighborhood quality of life. Bay tree canopy requirements. Reduce impervious surface.

Nitrogen Flux

Physical & Social Factors of Adoptability

New Vegetation Management Options

Design Options

Figure 1. Urban ecological design cycle. The ecological–social feedback loop is motivated by a concern to decrease nitrogen loading into the Chesapeake Bay by 40% by the year 2011. This motivation is depicted by the oval as a press from outside the loop. The feedback loop illustrates the relationships between land-cover heterogeneity and management, hydrology and nitrate export, vegetation management, design, and social decisions and activities. The heterogeneity of the land cover in the system as well as individual management decisions influence the hydrology and capacity of the terrestrial system to retain nitrogen. This relationship can be altered by affecting land-cover heterogeneity and management options at the neighborhood scale. Incorporating management options into design scenarios is the next step in the loop. Assessing the adoptability of those design options includes physical constraints and opportunities and desires and choices of individuals in the neighborhood. The selected design modifications can then be incorporated into a new model of landscape heterogeneity and management influences. The predicted influence of the new land-cover model on nitrate retention can be simulated through spatially explicit hydrological modeling. This dialog between landscape structure, management, hydrology, and design options is iterative and alterations to the land-water-scape can be tested for their influence on nitrate export to the Chesapeake Bay.

scapes, with their characteristic urban design and stormwater infrastructure, operate hydrologically. These models will allow the prediction of nitrate export from the existing watersheds. In order to improve the nitrate retention or conversion in the watersheds of the contrasting neighborhoods, current best management practices and potential new management practices and potential architectural and landscaping strategies will be identified by expert focus groups. With these tools in hand, an urban

226

design team including an architect and a landscape architect will generate new urban designs that might be built in the focal neighborhoods. The role of “ecological engineering” and of architectural and landscape architectural designs that maximize the ecological work done on stormwater and nitrate will be explored in the designs. Designs suitable for rehabilitation of vacant lots, retrofitting existing structures, new infill designs, and utilizing green space and improving the ecological function of neighborhood transportation infrastructure will be developed. The suitability of the designs proposed must be evaluated not only in terms of how they may physically fit into specific neighborhoods, but also how they relate to the social structure and values of the different neighborhoods. Social analyses will address density, gender, age, ethnicity, income, education, and national origin in the different neighborhoods, which are available from the U.S. Census. However, additional data, obtained from neighborhood surveys or focus groups, will also be required to understand what designs are actually preferred by the different neighborhoods. In addition, focus groups will be convened to evaluate and discuss the various designs from the perspective of local environmental and other benefits, as well as the larger regional benefits to the Chesapeake Bay. Discussion will also extend to the management and maintenance options that can be applied in the neighborhood. Additional data on consumer behavior, lifestyle, recreation, neighborhood cohesion, and attitudes about the environment will be obtained to help understand what shapes neighborhood differences in design preference. Based on the discussions with residents of the different neighborhoods, designs will be chosen for inclusion in new landscape maps, generated using HERCULES. These new maps will be assumed to incorporate features that can improve nitrate retention, since the designs have been conceived with those goals and appropriate features in mind. Using these new maps of the study watersheds, we will apply the hydroe-

Annals of the New York Academy of Sciences

cological models developed earlier using the existing landscape elements. Running the hydroecological models with the new landscape structures as their base will test whether, in fact, improvement in stormwater retention, quality, and nitrate retention can be achieved. The cycle is potentially an iterative, adaptive process. As new management strategies and design options are developed, they can be applied to both new and existing urban watershed structures. The cycle may also promote learning about local ecological services at a neighborhood scale, as well as connecting the neighborhood to larger regional environmental concerns. By engaging residents, policy makers and municipal managers, researchers, and the urban design professions, a joint learning process can be established and continued into the future. Discussion and Conclusions In the minds of many, conservation is an activity and philosophy most often associated with wild areas. For several decades now, it has also become recognized that it is necessary to design conservation strategies for cultural landscapes, or landscapes in which production activities occur side by side with natural processes that have clear conservation value (Farina 1997). However, urban areas are rarely considered the target of conservation, except perhaps for large wilderness parks that some cities are fortunate to possess. Conservation in urban areas more broadly can be considered to preserve existing ecological processes (Sukopp & Weiler 1988; Sukopp et al. 1990), or promote designs that establish ecological processes in parts of the urban mosaic where they had been reduced or lost (Felson & Pickett 2005). Applying conservation principles and practices to more extensive patch mosaics across urban areas may have several benefits. Increasing ecological structures and processes throughout metropolitan areas can reduce air and water pollution, both locally and regionally.

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments

Greening in urban areas is recognized to have psychological benefits (Kuo et al. 1998), to enhance biodiversity (e.g., McKinney 2002), to ameliorate microclimate (e.g., Heisler 1974; McPherson et al. 1997), and to improve stormwater management. This last feature is one that requires further empirical evaluation, as we have indicated earlier. Such improvements in urban environmental quality may reduce some of the pressures for residents to migrate from old city cores and older suburbs, and hence may contribute to regional conservation by reducing suburban and exurban sprawl. We have shown how the fundamental ecological concept of spatial heterogeneity can be applied to conservation in urban systems using the watershed and boundary frameworks. The application has required new, high categorical resolution models of urban land cover. The expected sink function of riparian zones as boundaries in this system was discovered not to operate due to hydrological alterations of urban streams and floodplains. This scientific result stimulated managers and policy makers to extend their stormwater-mitigation strategy spatially beyond the near-stream environment. Greening neighborhoods in storm drain catchments, reducing the coverage of paved surface, and the installation of several best management practices for stormwater should provide conservation value and social value more broadly in urban habitats. In terms of the boundary framework, this represents alteration of the source patches as well as the originally conceived boundaries exemplified by urban riparian zones. A general scheme for discovering the function of urban landscapes, inserting new design options that mimic natural processes into those landscapes based on the knowledge of ecologically aware engineers, managers, and urban designers, and vetted by the needs and perspectives of neighborhoods, brings the key information, actors, and interventions together in a learning loop. The cycle of ecological urban design serves as a generalizable way to achieve

227

watershed conservation goals in urban areas (Fig.1). Strategies for enhancing nitrate storage and reducing nitrate yield to receiving waters can be categorized into five types, all of which have been exemplified above. The first, and most traditional, approach is to restore the riparian boundary to regain the expected ecological function as a nitrate sink. Second, infrastructure can be added, as in the case of detention basins, or removed as in the case of daylighting buried urban streams. Third, infrastructural modifications intended to better mimic natural structures and ecological processes can be made, such as bioretention ponds and level spreaders. Fourth, materials used in the urban landscape can also be modified to enhance ecological processes. For example, the introduction of green roofs or permeable surfaces to enhance water infiltration farther upslope in the hydrological flow paths, thereby reducing the volume of water the riparian zone must process. Finally, reconceptualizing urban land-waterscapes from the standard watershed model of a riparian zone restricted to the stream side toward a model that recognizes the spatially distributed nature of land–water interfaces within urban watersheds suggest additional conservation strategies can be used throughout the watershed. In this paper we have focused on biogeophysical strategies that take advantage of the reconceptualization of urban watersheds as land-water-scapes to reduce nitrate yield from urban watersheds. We have only mentioned briefly how biogeophysical strategies can be complimented by strategies that address the human behaviors that control the input of nitrogen into the watershed, as social details are beyond our scope here. The insights can be seen as expanding on the current state of understanding of how urban watersheds work, and can thus contribute to the health of coastal waters. This expansion takes place along three dimensions: 1) from a focus on the relatively narrow riparian boundaries to extensive, functional urban land-water-scapes; 2) from engineering to remove water quickly to design that

228

Annals of the New York Academy of Sciences

slows water flow and enhances the ecological processes that can retain or convert nitrate; and 3) from emphasis on structural mitigation to include behaviors of humans organized as households, institutions, and agencies. Acknowledgments

Work reported in this paper was supported by the NSF LTER program through grant number DEB 0423476 and the Biocomplexity of Coupled Natural Human Systems program BCS-BE 508054. All BES participants are thanked for their generosity of ideas and enthusiasm for collaborative exploration of Baltimore. The comments of three anonymous reviewers enriched the manuscript. Conflict of Interest

The authors declare no conflicts of interest.

References Alberti, M., J.M. Marzluff, E. Shulenberger, et al. 2003. Integrating humans into ecology: opportunities and challenges for studying urban ecosystems. BioScience 53: 1169–1179. Anderson, J.R., E.E. Hardy, J.T. Roach & R.E. Witmer. 1976. Land use and land cover classification systems for use with remote sensor data (USGS Professional Paper No. 964): US Geological Survey. Arnold, C.L. & J.C. Gibbons. 1996. Impervious surface coverage: the emergence of a key environmental indicator. J. Am. Plann. Assoc. 62: 243–258. Atasoy, M., R.B. Palmquist & D.J. Phaneuf. 2006. Estimating the effects of urban residential development on water quality using microdata. J. Environ. Manage. 79: 399–408. Baker, L.A., P.M. Hartzheim, S.E. Hobbie, et al. 2007. Effect of consumption choices on fluxes of carbon, nitrogen, and phosphorus through households. Urban Ecosyst. 10: 97–117. Baltimore City, Office of the Mayor. 2006. Mayor O’Malley announces international award recognizing Baltimore’s urban tree canopy goal. Press Release, October 19 (www.ci.baltimore.md.us/ news/press/061019.html).

Baltimore County. 2000. Master plan 2010. Baltimore County Council. Towson, MD. http://www. baltimorecountymd.gov/Agencies/planning/ masterplanning/. Baltimore County Department of Environmental Protection and Resource Management. 2007. Forests and trees. Towson, MD. http://www.balt imorecountymd.gov/Agencies/environment /forestsandtrees/index.html. Baltimore County Office of Planning. 2006. Smart growth. Baltimore County. Towson, MD. http:// www.baltimorecountymd.gov/Agencies/planning/ masterplanning/smartgrowth.html. Band, L.E., M.L. Cadenasso, C.S.B. Grimmond, et al. 2005. Heterogeneity in urban ecosystems: patterns and process. In Ecosystem function in heterogeneous landscapes. G. Lovett, C.G. Jones, M.G. Turner & K.C. Weathers, Eds.: 257–278. Springer-Verlag. New York. Bernhardt, E.S., L.E. Band, C.J. Walsh & P.E. Berke. 2008. Understanding, managing, and minimizing urban impacts on surface water nitrogen loading. Ann. N.Y. Acad. Sci.The Year in Ecology and Conservation Biology. In Press. Bernhardt, E.S. & M.A. Palmer. 2007. Restoring streams in an urbanizing world. Freshw. Biol. 52: 738– 751. Bernhardt, E.S., M.A. Palmer, J.D. Allan, et al. 2005. Ecology–synthesizing U.S. river restoration efforts. Science 308: 636–637. Boesch, D.F., R.B. Brinsfield & R.E. Magnien. 2001. Chesapeake Bay eutrophication: scientific understanding, ecosystem restoration, and challenges for agriculture. J. Environ. Qual. 30: 303–320. Bormann, F.H. & G.E. Likens 1994. Patterns and Processes in a Forested Ecosystem. Springer-Verlag. New York. Brush, G.S. in prep. Urban riparian vegetation—defining the hydrological drought. Brush, G.S. & D. Bain. in prep. Historical land use and forest succession. Brush, G.S., C. Lenk & J. Smith. 1980. The natural forests of Maryland: an explanation of the vegetation map of Maryland. Ecol. Monogr. 50: 77–92. Bukaveckas, P.A. 2007. Effects of channel restoration on water velocity, transient storage, and nutrient uptake in a channelized stream. Environ. Sci. Technol. 41: 1570–1576. Cadenasso, M.L., S.T.A. Pickett & J.M. Grove. 2006. Dimensions of ecosystem complexity: heterogeneity, connectivity, and history. Ecol. Complexity 3: 1–12. Cadenasso, M.L., S.T.A. Pickett & K. Schwarz. 2007. Spatial heterogeneity in urban ecosystems: reconceptualizing land cover and a framework for classification. Front. Ecol. Environ. 5: 80–88.

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments Cadenasso, M.L., S.T.A. Pickett, K.C. Weathers & C.G. Jones. 2003. A framework for a theory of ecological boundaries. BioScience 53: 750–758. Carpenter, S.R., N.F. Caraco, D.L. Correll, et al. 1998. Nonpoint pollution of surface waters with phosphorus and nitrogen. Ecol. Appl. 8: 559–568. Chen, J., J.F. Franklin & T.A. Spies. 1995. Growingseason microclimatic gradients from clearcut edges into old-growth douglas-fir forests. Ecol. Appl. 5: 74– 86. Chesapeake Bay Commission. 2006. Focal points: annual report 2006. Chesapeake Bay Commission. Annapolis, pp. 42. http://www.chesbay.state.va.us/ Publications/CBC-AR-2006.pdf. Chesapeake Bay Foundation. 2003. Sewage Treatment Plants: the Chesapeake Bay Watershed’s Second Largest Source of Water Pollution. Baltimore, MD. Chesapeake Bay Program. 2000. Chesapeake 2000. Chesapeake Bay Program. Annapolis, MD. http://www.chesapeakebay.net/pubs/chesapeake 2000agreement.pdf. Clay, G. 1973. Close Up: How to Read the American City. Praeger Publishers. New York. Craig, L.S., M.A. Palmer, D.C. Richardson, et al. 2008. Stream restoration strategies for reducing nitrogen loads. Front. Ecol. Environ. In Press. Diaz, R.J. 2001. Overview of hypoxia around the world. J. Environ. Qual. 30: 275–281. Dietz, M.E., J.C. Clausen & K.K. Filchak. 2004. Education and changes in residential nonpoint source pollution. Environ. Manage. 34: 690. DiGregorio, A. & L.J.M. Jansen. 2001. Land Cover Classification System (LCCS). Classification Concepts and User Manual. Food and Agriculture Organization. Rome, MD. Ewel, K.C., C. Cressa, R.T. Kneib, et al. 2001. Managing critical transition zones. Ecosystems 4: 452–460. Farina, A. 1997. Principles and Methods in Landscape Ecology. Kluwer. New York. Felson, A.J. & S.T.A. Pickett. 2005. Designed experiments: new approaches to studying urban ecosystems. Front. Ecol. Environ. 3: 549–556. Forman, R.T.T., D. Sperling, J.A. Bissonette, et al. 2003. Road Ecology: Science and Solutions. Island Press. Washington, DC. Galvin, M.F., J.M. Grove & J.P.M. O’Neil-Dunne. 2006. A Report on Baltimore City’s Present and Potential Urban Tree Canopy. Maryland Department of Natural Resources, Forest Service. Annapolis, MD. Gift, D., P.M. Groffman, S.S. Kaushal, et al. 2008. Root biomass, organic matter, and denitrification potential in degraded and restored urban riparian zones. Restoration Ecology In Press. Goonetilleke, A., E. Thomas, S. Ginn & D. Gilbert. 2005. Understanding the role of land use in urban

229

stormwater quality management. J. Environ. Manage. 74: 31–42. Gottdiener, M. & R. Hutchison. 2000. The New Urban Sociology, 2nd edn. McGraw Hill. New York. Grimm, N.B., L.J. Baker & D. Hope. 2003. An ecosystem approach to understanding cities: familiar foundations and uncharted frontiers. In Understanding Urban Ecosystems: A New Frontier for Science and Education. A.R. Berkowitz, C.H. Nilon & K.S. Holweg, Eds.: 95–114. Springer-Verlag. New York. Groffman, P.M., D.J. Bain, L.E. Band, et al. 2003. Down by the riverside: urban riparian ecology. Front. Ecol. Environ. 1: 315–321. Groffman, P.M., N.J. Boulware, W.C. Zipperer, et al. 2002. Soil nitrogen cycling processes in urban riparian zones. Environ. Sci. Technol. 36: 4547–4552. Groffman, P.M. & M.K. Crawford. 2003. Denitrification potential in urban riparian zones. J. Environ. Qual. 32: 1144–1149. Groffman, P.M., A.M. Dorsey & P.M. Mayer. 2005. Nitrogen processing within geomorphic features in urban streams. J. N. Am. Benthol. Soc. 24: 613–625. Groffman, P.M., N.L. Law, K.T. Belt, et al. 2004. Nitrogen fluxes and retention in urban watershed ecosystems. Ecosystems 7: 393–403. Grove, J.M., M.L. Cadenasso, W.R. Burch, Jr., et al. 2006. Data and methods comparing social structure and vegetation structure of urban neighborhoods in Baltimore, Maryland. Soc. Nat. Resour. 19: 117– 136. Hale, R. & P. Groffman. 2006. Chloride effects on nitrogen dynamics in forested and suburban stream debris dams. J. Environ. Qual. 35: 2425–2432. Hassett, B.A., M. Palmer, E. Bernhardt, et al. 2005. Restoring watersheds project by project: trends in Chesapeake Bay tributary restoration. Front. Ecol. Environ. 3: 259–267. Hathaway, J.M. & W.F. Hunt. 2006. Urban Waterways: Level Spreaders: Overview, Design, and Maintenance. North Carolina Cooperative Extension Service, AGW 588 09. Hatt, B.E., T.D. Fletcher, C.J. Walsh & S.L. Taylor. 2004. The influence of urban density and drainage infrastructure on the concentrations and loads of pollutants in small streams. Environ. Manage. 34: 112– 124. Heisler, G.M. 1974. Trees and human comfort in urban areas. J. Forest. 72: 466–469. Hogan, D.M. & M.R. Walbridge. 2007. Best management practices for nutrient and sediment retention in urban stormwater runoff. J. Environ. Qual. 36: 386–395. Howarth, R.W. & R. Marino. 2006. Nitrogen as the limiting nutrient for eutrophication in coastal marine ecosystems: evolving views over three decades. Limnol. Oceanogr. 51: 364–376.

230 Hsieh, C.H. & Davis, A.P. 2005. Multiple-event study of bioretention for treatment of urban stormwater runoff. Water Sci. Technol. 51: 177–181. Hunt, W.F., A.R. Jarrett, J.T. Smith & L.J. Sharkey. 2006. Evaluating bioretention hydrology and nutrient removal at three field sites in North Carolina. ASCE J. Irrig. Drain. Eng. 132: 600–608. Hutchings, M.J., E.A. John & A.J.A. Stewart (Eds). 2000. The Ecological Consequences of Environmental Heterogeneity. Blackwell Science. Malden, MA. Jacobs, J. 1961. The Death and Life of Great American Cities: The Failure of Town Planning. Random House. New York. Johnson, S. 2006. The Ghost Map: The Story of London’s Most Terrifying Epidemic—and How It Changed Science, Cities, and The Modern World. Riverhead Books. New York. Jones, K.B., A.C. Neale, M.S. Nash, et al. 2001. Predicting nutrient and sediment loadings to streams from landscape metrics: a multiple watershed study from the United States Mid-Atlantic region. Landscape Ecol. 16: 301–312. Jordan, T.E., D.L. Correll & D.E. Weller. 1997. Relating nutrient discharges from watersheds to land use and streamflow variability. Water Resour. Res. 33: 2579– 2590. Kaushal, S.S., P.M. Groffman, P.M. Mayer, et al. 2008. Effect of stream restoration on denitrification at the riparian-stream interface of an urbanizing watershed of the mid-Atlantic U.S. Ecological Applications In Press. Kaye, J.P., P.M. Groffman, N.B. Grimm, et al. 2006. A distinct urban biogeochemistry? Trends Ecol. Evol. 21: 192–199. Kolasa, J. & S.T.A. Pickett. 1991. Ecological Heterogeneity. Springer-Verlag. New York. Koroncai, R., L. Linker, J. Sweeney & R. Batuik. 2003. Setting and Allocating the Chesapeake Bay Basin Nutrient and Sediment Loads: The Collaborative Process, Technical Tools, and Innovative Approaches. US Environmental Protection Agency. Annapolis, MD. Kuo, F.E., M. Bacaicoa & W.C. Sullivan. 1998. Transforming inner-city landscapes: trees, sense of safety, and preferences. Environ. Behav. 30: 28–59. Law, N.L., L.E. Band & J.M. Grove. 2004. Nitrogen input from residential lawn care practices in suburban watersheds in Baltimore County, Md. J. Environ. Plann. Manage. 47: 737–755. Li, J., R. Orland & T. Hogenbirk. 1998. Environmental road and lot drainage designs: alternatives to the curb-gutter-sewer system. Can. J. Civil Eng. 25: 26– 39. Lovett, G.M., M.M. Traynor, R.V. Pouyat, et al. 2002. Atmospheric deposition to oak forests along an urban-rural gradient. Environ. Sci. Technol. 34: 4294– 4300.

Annals of the New York Academy of Sciences Lovett, G.M., C.G. Jones, M.G. Turner & K.C. Weathers (Eds.). 2005. Ecosystem Function in Heterogeneous Landscapes. Springer. New York. Lowrance, R. 1998. Riparian forest ecosystems as filters for nonpoint-source pollution. In Successes, Limitations, and Frontiers in Ecosystem Science. M.L. Pace & P.M. Groffman, Eds.: 113–141. Springer. New York. Mayer, P.M., S.K. Reynolds, M.D. McCutchen & T.J. Canfield. 2007. Meta-analysis of nitrogen removal in riparian buffers. J. Environ. Qual. 36: 1172– 1180. Mayer, P.M., E. Striz, R. Shedlock, et al. 2003. The Effects of Ecosystem Restoration on Nitrogen Processing in an Urban Mid-Atlantic Piedmont Stream. Paper presented at the First Interagency Conference on Research in the Watersheds. Benson, AZ. Meyer, S.C. 2005. Analysis of base flow trends in urban streams, northeastern Illinois, USA. Hydrol. J. 13: 871–885. McClain, M.E., E.W. Boyer, C.L. Dent, et al. 2003. Biogeochemical hot spots and hot moments at the interface of terrestrial and aquatic ecosystems. Ecosystems 6: 301–312. McKinney, M.L. 2002. Urbanization, biodiversity, and conservation. BioScience 52: 883–891. McPherson, E.G., D. Nowak, G. Heisler, et al. 1997. Quantifying urban forest sturcture, function, and value: the Chicago urban forest climate project. Urban Ecosyst. 1: 49–61. Melosi, M.V. 2000. The Sanitary City: Urban Infrastructure in America from Colonial Times to the Present. Johns Hopkins University Press. Baltimore, MD. Mentens, J., D. Raes & M. Hermy. 2006. Green roofs as a tool for solving the rainwater runoff problem in the urbanized 21st century? Landscape Urban Plann. 77: 217–226. Miller, C.V., J.M. Denis, S.W. Ator & J.W. Brakebill. 1997. Nutrients in streams during baseflow in selected environmental settings of the Potomac River basin. J. Am. Water Resour. Assoc. 33: 1155–1171. Naiman, R.J. & R.E. Bilby (Eds). 1998. River Ecology and Management: Lessons from the Pacific Coastal Ecoregion. Springer-Verlag. New York. Naiman, R.J. & H. D´ecamps 1997. The ecology of interfaces: riparian zones. Annu. Rev. Ecol. Syst. 28: 621– 658. Naiman, R.J., H. D´ecamps & M.E. McClain. 2005. Riparia: Ecology, Conservation, and Management of Streamside Communities. Elsevier, Academic Press. Amsterdam. Natural Resources Conservation Service. 2006. Chesapeake Bay and agriculture. United States Department of Agriculture. Washington, DC. Nikolaidis, N.P., H. Heng, R. Semagin & J.C. Clausen. 1998. Non-linear response of a mixed land use

Cadenasso et al.: Reducing Nitrate Yield from Urban Catchments watershed to nitrogen loading. Agric. Ecosyst. Environ. 67: 251–265. Oberndorfer, E., J. Lundholm, B. Bass, et al. 2007. Green roofs as urban ecosystems: ecological structures, functions, and services. BioScience 57: 823–833. Palmer, M., J.D. Allen, J. Meyer & E.S. Bernhardt. 2007. River restoration in the twenty-first century: data and experiential knowledge to inform future efforts. Restoration Ecol. 15: 472–481. Paul, M.J. & J.L. Meyer. 2001. Riverine ecosystems in an urban landscape. Annu. Rev. Ecol. Syst. 32: 333–365. Pauleit, S. & F. Duhme 2000. Assessing the environmental performance of land cover types for urban planning. Landscape Urban Plann. 52: 1–20. Peterjohn, W.T. & D.L. Correll. 1984. Nutrient dynamics in an agricultural watershed: observations on the role of a riparian forest. Ecology 65: 1466–1475. Pickett, S.T.A., K.T. Belt, M.F. Galvin, et al. 2007. Watersheds in Baltimore, Maryland: understanding and application of integrated ecological and social processes. J. Contemp. Water Res. Educ. 136: 44–55. Pickett, S.T.A., W. Burch, Jr., S. Dalton, et al. 1997. A conceptual framework for the study of human ecosystems in urban areas. Urban Ecosyst. 1: 185– 199. Pickett, S.T.A. & M.L. Cadenasso 2008. Linking ecological and built components of urban mosaics: an open cycle of ecological design. J. Ecol. 96: 8–12. Raciti, S., M.F. Galvin, J.M. Grove, et al. 2006. Urban Tree Canopy Goal Setting: A Guide for Chesapeake Bay Communities. United States Department of Agriculture, Forest Service, Northeastern State & Private Forestry, Chesapeake Bay Program Office. Annapolis, MD. Richardson, D.C. 2006, Sept. Watershed 263: a resource uncovered. Stormwater: The Journal for Surface Water Quality Professionals. Vol. 7. http://www.forester.net/ sw_0609_watershed.html. Ridd, M.K. 1995. Exploring a V-I-S (vegetationimpervious surface-soil) model for urban ecosystem analysis through remote sensing: comparative anatomy for cities. Int. J. Remote Sens. 16: 2165– 2185. Rose, S. & N.E. Peters. 2001. Effects of urbanization on streamflow in the Atlanta area (Georgia, USA): a comparative hydrological approach. Hydrol. Process. 15: 1441–1457. Rosgen, D.L. 1994. A classification of natural rivers. Catena 22: 169–199. Rosgen, D.L. 1996. Applied River Morphology. Wildland Hydrology. Pagosa Springs, CO. Shane, D.G. 2005. Recombinant Urbanism: Conceptual Modeling in Architecture, Urban Design, and City Theory. John Wiley & Sons. Hoboken, NJ. Shields, C.A., L.E. Band, P.M. Groffman, et al. 2008. Export timing of nitrogen from catchments along an

231

urban-rural gradient in the Chesapeake Bay watershed. Water Resources Research In Press. Sprague, E., D. Burke, S. Claggett & A. Todd 2006. The State of Chesapeake Forests. The Conservation Fund & USDA Forest Service, Northeastern Area, State and Private Forestry. Arlington, VA & Annapolis, MD. Strayer, D.L., R.E. Beighley, L.C. Thompson, et al. 2003. Effects of land cover on stream ecosystems: roles of empirical models and scaling issues. Ecosystems 6: 407–423. Sukopp H. 1998. Urban ecology—scientific and practical aspects. In Urban Ecology. J. Breuste, H. Feldmann & O. Uhlmann, Eds.: 3–16. Springer-Verlag. New York. Sukopp, H. & S. Weiler 1988. Biotope mapping and nature conservation strategies in urban areas of the Federal Republic of Germany. Landscape Urban Plann. 15: 39–58. Sukopp, H., S. Hejny & I. Kowarik (Eds). 1990. Urban Ecology: Plants and Plant Communities in Urban Environments. SPB Academic Publishing. The Hague. Swank, W.T. & J.E. Douglass. 1977. Nutrient budgets for undisturbed and manipulated hardwood forest ecosystems in the mountains of North Carolina. In Watershed research in eastern North America. Vol. 1. D.L. Correll, Ed.: 343–364. Chesapeake Bay Center for Environmental Studies, Smithsonian Institute. Edgewater, MD. Troy, A.R., J.M. Grove, J.P.M. O’Neil-Dunne, et al. 2007. Predicting opportunities for greening and patterns of vegetation on private urban lands. Environ. Manage. 40: 394–412. United Nations Population Fund. 2007. State of World Population 2007: Unleashing the Potential of Urban Growth. United Nations Population Fund. New York. USEPA (US Environmental Protection Agency). 1990. National pesticide survey: nitrate. Office of Water, Office of Pesticides and Toxic Substances. Washington, DC. Valiela, I., G. Tomasky, J. Hauxwell, et al. 2000. Operationalizing sustainability: management and risk assessment of land-derived nitrogen loads to estuaries. Ecol. Appl. 10: 1006–1023. Vitousek, P.M., J.D. Aber, R.W. Howarth, et al. 1997a. Human alteration of the global nitrogen cycle: sources and consequences. Ecol. Appl. 7: 737–750. Vitousek, P.M., H.A. Mooney, J. Lubchenco & J.M. Melillo. 1997b. Human domination of the Earth’s ecosystems. Science 277: 494–499. Wakida, F.T. & D.N. Lerner. 2005. Non-agricultural sources of groundwater nitrate: a review and case study. Water Res. 39: 3–16. Walsh, C.J., T.D. Fletcher & A.R. Ladson. 2005a. Stream restoration in urban catchments through redesigning stormwater systems: looking to the catchment to save the stream. J. North Am. Benthol. Soc. 24: 690–705.

232 Walsh, C.J., A.H. Roy, J.W. Feminella, et al. 2005b. The urban stream syndrome: current knowledge and the search for a cure. J. North Am. Benthol. Soc. 24: 706– 723. Wayland, K.G., D.T. Long, D.W. Hyndman, et al. 2003. Identifying relationships between baseflow geochemistry and land use with synoptic sampling and R-mode factor analysis. J. Environ. Qual. 32: 180–190. Weathers, K.C., M.L. Cadenasso & S.T.A. Pickett. 2001. Forest edges as nutrient and pollutant concentrators: potential synergisms between fragmentation, forest canopies, and the atmosphere. Conserv. Biol. 15: 1506–1514.

Annals of the New York Academy of Sciences Weller, D.E., T.E. Jordan, D.L. Correll & Z.J. Liu. 2003. Effects of land-use change on nutrient discharges from the Patuxent River watershed. Estuaries 26: 244–266. Wickham, J.D., R.V. O’Neill, K.H. Riitters, et al. 2002. Geographic targeting of increases in nutrient export due to future urbanization. Ecol. Appl. 12: 93–106. Wollheim, W.M., B.A. Pellerin, C.J. Vorosmarty & C.S. Hopkinson. 2005. N retention in urbanizing headwater catchments. Ecosystems 8: 871–884. Zhu, W.X., N.D. Dillard & N.B. Grimm. 2004. Urban nitrogen biogeochemistry: status and processes in green retention basins. Biogeochemistry 71: 177–196.