The Triangle region of North Carolina is framed by the cities of Raleigh, Durham, and Chapel Hill. The Triangle spans 9 counties and more than 40 municipalities. The area is experiencing explosive population growth, accelerating steadily since the 1970's. With this growth come the myriad symptoms of urbanization: impacts on watersheds, wildlife habitat and open space, and so on.

The Triangle Landscape Change Project is an on-going effort that was initiated under funding from NSF's Urban Reseach Initiative (1999-2002), and which continues to engage members of the lab. The overall goal of the project is to better understand the causes and consequences of landscape change in this region. This broad goal has provided a framework for studies of landscape context on real estate values, forest pattern and composition, forest bird communities, and watershed impacts stemming from urban development. Recently, the emerging issue of ecosystem services has provided a more integrated conceptual framework for this collection of projects. Here, we provide a tasting of the various elements of the Triangle Landscape Change Project.

The Triangle region of North Carolina. Inset locator is North Carolina; colored section is study area based on a single Thematic Mapper scene intersected over 25 years. Land cover is classified for 2005 (greens, forests; yellows, fields; reds, developed).

Forest composition and dynamics

Perhaps the most obvious effect of urban development is that it fragments intact habitats (here, forests) into smaller patches. Smaller patches, for purely geometric reasons, also tend to have more edge and are farther separated. Increased edge has several implications: forest edges are more productive (and hence, sequester more carbon)(McDonald and Urban 2004), but also seem more vulnerable to invasive species (MacDonald et al. 2006a). Forests in this region also exhibit legacies of prior land uses, in that current forest pattern is related to topography and soils in ways that reflect the agricultural history of the Piedmont (Taverna et al. 2005).

Forest edges in the Triangle, as distance to nonforest landcover (forests only shown, from 2005 imagery; red is close to edge, green is far, scaled on deciles). 10% of forests are < 30 m from edge; 30%, < 60 m; and nearly 50% < 100 m.

Forest bird communities

Forests in the region are quite fragmented but still rather highly connected due to a large proportion of second-growth forests and especially their concentration in naturally connected riparian zones. This region has been a proving ground for the lab's foray into analyses of habitat connectivity based on graph theory (Minor and Urban 2007, 2008). This work has also documented a strong gradient in forest bird community composition, with similar species richness but quite different species composition in urban as compared to rural forests (Minor and Urban 2009). These patterns do not map simply into the prevailing model of "area-sensitive" species being limited to large, remnant tracts.

Graph model of forest bird habitat patches for the Triangle (from Minor and Urban 2008).

Stream ecosystems and the urban stream syndrome

Urbanization has well recognized impacts on hydrology, often modeled in terms of impervious surface area; these impacts are readily observable in the Triangle (Carle et al. 2005).

A new collaboration with Emily Bernhardt's lab has freshened our interest in land use impacts on stream ecosystems. In this, we are focusing explicity on the urban stream syndrome, a constellation of symptoms including the effects of increasingly flashy hydrographs in urban streams, decreased baseflow, decreased connections between streams and their riparian zones, and increased connections between streams and their contributing watersheds. Our work focuses on contrasting development impacts in space and time: local (stream-scale) versus larger (watershed-scale) influences, and episodic (hydraulic pulses) versus chronic pressures (contaminant concentration under low-flow conditions) influences on urban stream ecosystems. Under funding from NC's Water Resources Research Institute and Bernhardt's NSF CAREER grant, we have recently conducted a synoptic survy of ~70 low-order streams stratified over a gradient of land use intensity. We are currently developing models to help diagnose and treat the urban stream syndrome.

Watersheds surveyed in the Triangle. Dot size is proportional to total nitrogen. Land cover data is from 2005 (reds, developed; greens, forested; yellows, fields) (Somers et al., in prep.).

The pace and pattern of landscape change

The standard approach for studying landscape change is to classify land cover types from aerial photographs or satellite images from 2 or perhaps a few dates, and to then compare the maps. This is essentially an ANOVA model for studying landscape change. The increased availability of satellite imagery invites a different approaches. McDonald et al. (2007) showed that quite nuanced ecological information could be extracted from images analyzed in terms of continuous variables (e.g., greenness indices). Recently, Sexton developed an approach to take advantage of the Landsat imagery archive, which is now freely acessible. Sexton use signature extension methods to apply the 2001 NLCD land cover classification to be applicable to the entire Landsat archive, 1985-present (Sexton et al., in review). This is a game-changing innovation, as it invites a regression-based analysis of landscape change instead of the conventional ANOVA model. Thus, we can now ask questions about the shape of the trajectory of landscape change, rather than merely asking whether 2 maps are different. We are only beginning to explore the ramifications of this shift in analytic approach.

Trends in land cover in the Triangle, 1985-2005. Top series: fields (ag + lawns); middle: evergreen forest (pines); bottom: low-density developed (from Sexton et al., in review).

Urbanization and the provision of ecosystem services

Ecosystem services are those products and benefits that ecosystems provide to humans. These include food and fiber, watershed protection, wildlife habitat and open space, and so on. Thus, it is easy to frame much of our work in the Triangle in this general framework. Recently, there has been a huge flush of interest in ecosystem services as natural resource managers, conservation agencies, and land use planners attempt to ensure the sustainable and equitable provision of these services in the face of increased development pressure on ecosystems (e.g., the Millenium Ecosystem Assessment). The lab is participating in a working group on ecosystem services with Duke's Nicholas Institute and several other faculty. Our aim is to develop methods to link the ecology, economics, and policy framework for managing ecosystem services. As part of this, we have recently entered into the planning stage for an Urban Long-term Research Area (ULTRA), with our planning grant focused explicitly on the sustainable and equitable provision of ecosystem services across the Triangle. This project is in collaboration with colleagues at Duke (contacts: Dean Urban and Emily Berhardt), North Carolina State University (contacts: George Hess and Melissa McHale), the University of North Carolina at Chapel Hill (contacts: Larry Band and Phil Berke), and the Triangle J Council of Governments.

Urban systems can be viewed as the interaction between ecosystems and human institutions, mediated by markets and policies. In each system, processes generate patterns (from bottom to top) and in turn are constrained by these patterns or structures. A major challenge in urban systems is to reconcile the patterns of supply of ecosystem services with the patterns of demand for these services (figure from our ULTRA proposal).

A particular focus of our work on ecosystem services is to quantify the joint production of services. For example, what is the overlap in watershed protection, biodiversity support, and carbon sequestration potential for a particular parcel of land? What opportunities for leverage and efficient allocation might be available by taking advantage of this overlap? These general questions will engage us in future projects.


Carle, M. Vernon, P.N. Halpin, and C.A. Stow. 2005. Patterns of watershed urbanization and impacts on water quality. J. Am. Water Resources Assoc. 41:693-708.

Mansfield, C., S. Pattanayak, W. McDow, R.McDonald, and P. Halpin. 2005. Shades of green: measuring the value of urban forests in the housing market. J. Forest Economics 11:177-199.

McDonald, R.I., and D.L.Urban. 2004. Forest edges and tree growth rates in the North Carolina Piedmont. Ecology 85:2258-2266.

McDonald, R.I., and D.L. Urban. 2006a. Edge effects on species composition and exotic species abundance in the North Carolina Piedmont. Biological Invasions 8:1049-1060.

McDonald, R.I., and D.L. Urban. 2006b. Spatially varying rules of landscape change: lessons from a case study. Landscape and Urban Planning B 74:7-20.

McDonald, R.I., P.N. Halpin, and D.L. Urban. 2007. Monitoring succession from space: a case study from the North Carolina Piedmont. Applied Vegetation Science 10:193-203.

Minor, E.S., and D.L. Urban. 2007. Graph theory as a proxy for spatially explicit population models in conservation planning. Ecological Applications 17:1771-1782.

Minor, E.S., and D.L. Urban 2009. Forest bird communities across a gradient of urban development. Urban Ecosystems (on-line).

Minor, E.S., and D.L. Urban. 2008. A graph-theory framework for evaluating landscape connectivity and conservation planning. Cons. Biol. 22:297-307.

Sexton, J.O., M.J. Donohue, and D.L. Urban. Landcover change and ecosystem services on the North Carolina Piedmont Plateau (1985-2005). (ms in review)

Taverna, K., D.L. Urban, and R.I. McDonald. 2005. Modeling landscape vegetation pattern in response to historic land use: a hypothesis-driven approach for the North Carolina Piedmont. Landscape Ecology 20:689-702.