CHANS-Net: The International Network of Research on Coupled Human and Natural Systems

Brian Allan, postdoctoral fellow, Washington University
Research Interest: Disease ecology

Title: Effects of landscape change on the transmission of tick-borne diseases in the southeastern United States

Human alteration of natural landscapes can create hotspots for emerging zoonotic diseases through a complex pathway of ecological interactions between hosts, vectors, and their environment. Here, we examine the impacts of human-mediated landscape change in the Saint Louis, Missouri region, on the emergence of several bacterial pathogens (e.g., Ehrlichia spp., Borrelia lonestari) transmitted by the lone star tick (Amblyomma americanum). Utilizing a natural gradient in human disturbance in the Saint Louis metropolitan region, we implemented a combination of field- and laboratory-based approaches to assess the effects of landscape change in this region on the abundance of ticks and their vertebrate hosts, tick infection rates with pathogens, and the distribution of tick bloodmeals among vertebrate host species. We found a positive relationship between the abundance of lone star ticks and white-tailed deer (Odocoileus virginianus), the preeminent host for all three life stages of this tick species. We also found a unimodal relationship between the density of white-tailed deer and Geographic Information System (GIS)-based estimates of the percent forest cover associated with our study sites. Further, the density of lone star ticks was similarly related to percent forest cover, indicating a potential indirect mechanism by which landscape change may alter the abundance of ticks in this region via the response of a key vertebrate host. Finally, for a subset of these sites, we determined the prevalence of lone star tick-associated pathogens and thereby calculated the density of infected ticks. We found a unimodal relationship between the density of infected ticks and percent forest cover, suggesting that intermediate levels of forest cover are correlated with high human risk of exposure to tick-borne diseases. Our results indicate that the potential impact of tick-borne diseases on human health should be incorporated in landscape planning decisions in this emerging hotspot for vector-borne pathogens.

Andre Dribin, doctoral candidate, University of Illinois-Chicago
Research Interest: Landscape Ecology, History of Science, Design Criticism

Title: Lakefront Competition

In 2003, the Graham Foundation for Advanced Studies in the Fine Arts sponsored a competition on the “21st Century Lakefront Park” to extend and complete Chicago’s Lincoln Park, encouraging proposals to consider the possibility of ‘both/and’ between the built and natural environment. Soliciting over 100 entries from professionals across the nation, the six competition winners highlight the many changes in the disciplinary practice of architecture addressing some of the most complex political, ecological, and aesthetic challenges facing public lands and waters in cities today. Tackling issues related to history, context, energy, biodiversity, branding, globalism, and infrastructure, Chicago’s lakefront offers a site rich with landscape and programmatic diversity to challenge park-city relations. The various attitudes evident in the competition entries questioned how one might participate in the nature-society relation today and re-appropriate the role of landscape as a constructive activity. Following a brief account of the landscape and planning activities that have shaped Chicago’s lakefront, an evaluation of the winning entries helps to cultivate more active participation, renew discussion and develop fresh ideas about the role of the constructed landscape to expand the possibilities of the twenty-first century city.

Rebecca Kennedy, research ecologist, USDA Forest Service
Research Interest: forest landscape ecology, climate change, biodiversity, landscape wildlife ecology, old-growth forest ecology and conservation, landscape structure and dynamics, influences of disturbances and environmental changes on forest landscape diversity and resilience

Title: Assessing potential tradeoffs for carbon sequestration, wildlife habitat, and fire risk mitigation with climate change and fire management on the Olympic Peninsula, Washington, USA

Forests of the maritime Pacific Northwestern USA may have high carbon sequestration potential and high potential to sustain older forest and other forest structural types for threatened and valued wildlife species, via their high productivity and moderate to infrequent fire regimes. With climate change, there may be shifts in incidence and severity of fire, especially in the drier areas of the region, via changes to forest productivity and hydrology, and consequent effects to C sequestration and forest structure. To explore this issue in an area with relatively limited alternatives for fire and fuels management, I assessed potential effects of varying approaches to fire management (no suppression/wildland fire management/highly effective fire suppression) under two climate change scenarios on future C sequestration and wildlife habitat in Olympic National Park, WA, over a 500-year simulation period. I used the simulation platform FireBGCv2, which contains a mechanistic, individual tree succession model, a spatially explicit climate-based biophysical model that uses daily weather data, and a spatially explicit fire model that incorporates ignition, spread, and effects on ecosystem components, with stochastic properties implemented in a spatial domain. C sequestration patterns varied over time and spatial and temporal patterns differed somewhat depending on the climate change scenario applied and the fire management methods employed. Under the more extreme climate change scenario with little fire suppression, fires were most frequent and severe and older forest habitat was reduced, but early successional forest important to some components of other wildlife habitat were promoted. General trends were similar under the more moderate climate change scenario but spatial patterns differed. Some areas of the landscape served as refugia for older forest under increasing frequency of high severity fire and may be promising as anchors for the maintenance of habitat in a landscape experiencing increasing frequency of disturbance with climate change.

Quingmin Meng, associate director of computing and technology, University of North Carolina-Charlotte
Research Interest: I am a broadly trained spatial analyst with interests in basic and applied research related to ecosystem disturbance and processes including urban ecosystems, forest health, fire ecology, and biomass and carbon. I am particularly interested in quantifying and modeling of human-environmental interactions especially environmental geography, landscape processes, landscape epidemiology, and the spatial-temporal dynamics of the coupled human and natural systems. At the same time, I am interested in the basic research of theoretical and methodological aspects of geographical information science, remote sensing, geo-spatial statistical modeling, the integration of geographical methodologies, and applying them to ecosystem analyses.

Title: Landscape modeling of multi-strata forest fire severity using alternative remote sensing of Landsat, AVIRIS, and MASTER

Landscape models of fire severity are needed to better understand the behavior and ecological impacts of wildfires, especially in complex and spatially heterogeneous environments. Remote sensing models of fire severity predominantly focus on the two band Landsat magederived differenced normalized burn ratio (dNBR) and/or fitting relationships between field measures of composite burn index (CBI) and dNBR to map fire severity at unsampled places. Immediately following control of the fire, we surveyed fire severity of sixty 500 m2 forested plots that were established prior to the wildfire in Big Sur ecoregion. We measured CBI across five forest strata, including substrate, herb, shrub, intermediate-sized tree, and dominant canopy tree layers. Here, we use all the available Landsat spectral information to model the largest wildfire in California’s documented history, the 2008 Big Sur Basin Complex fire. Besides AVIRIS (Airborne Visible/Infrared Imaging Spectrometer), a new hyperspectral technology MASTER (the MODIS and ASTER simulator sensor) is used to regionally model fire severity across multiple forest strata in Big Sur. Statistical models with predictors of Landsat bands, Landsat derived NBR or dNBR, image differencing, image ratioing, AVIRIS, and MASTER images indicate the combinations of Landsat bands predicted more landscape variability in fire severity than all the other remote sensing data. Statistical tests of fire behavior and its effects on the heterogeneous landscape had different effects on different forest strata. The empirical comparisons between fire severity and spectral characteristics show similar fire damages occurred within the same or different vegetation stands but in different slope and aspect positions can result in significant different hyperspectral characteristics, which cannot be suitable for fire severity modeling of heterogeneous landscapes. This study shows the combined measure CBI is not optimal to represent multi-strata forest fire severity and dNBR performed poorer than Landsat TM bands for fire severity modeling.


Jian Peng, assistant professor, Peking University, China
Research Interest: LUCC; Landscape Patterns and Ecological Sustainability

Title: Assessing ecosystem health of rural landscapes based on landscape patterns: A case study in Lijiang County of China

Ecosystem health assessment is always one of the key topics of ecosystem management. Compared with various studies on the health assessment of single kind of ecosystems at local or national scale, few are focused on assessing ecosystem health of landscapes, which are geo-spatial units composed of different kinds of ecosystem mosaics. Meanwhile, few assessments have focused on the effects of landscape patterns on the healthy status of ecosystems.

Taking Lijiang County of China as a case, this study aims to assess ecosystem health of rural landscapes based on landscape patterns. In the assessment, ecosystem health are distinguished between physical health and integrated health of ecosystem. Regarded as the ability to sustain ecosystem structure and functions, physical health of ecosystem is assessed through three aspects, i.e. ecosystem vigor, organization and resilience. And focusing on ecosystem service function, integrated health of ecosystem is considered as the ability of ecosystem to keep physical health itself, and to satisfy human needs for ecosystem services. In details, ecosystem vigor is weighed by mean NDVI; ecosystem organization is measured by selected landscape metrics; ecosystem resilience is evaluated by resilience coefficient of different land use types and associated area ratio; and ecosystem service function is assessed according to ecosystem services coefficient of different land use types and associated area ratio, considering spatial neighboring effects among different ecosystems.

The results show that, during 1986-2002, both the status of physical health and integrated health of ecosystem for the whole county are healthy, with a little increase of each index. Although there are only a few changes of the two indexes for all the 24 towns during the study period, there is distinct spatial difference among the towns, and the town stationed by the county government has the lowest value of both indexes.

Salman Qureshi, assistant professor, University of Karachi, Pakistan
Research Interest: Urban Ecology; Urban Nature & Ecosystem Services; Socio-ecological Modelling in Megacities; Remote Sensing & GIS

Title: Epitomizing the simplicity in complexity of socio-ecological field sampling using an integrated spatial model of complex mosaic of urbanization

Scientists studying megacities are always confronted with the challenge of stratifying their sampling sites for in-depth field investigations, particularly in developing countries where the urban landscapes do not expand with predetermined plans. The challenge gets more complicated especially for socio-ecological studies; it is primarily because of the complexity of coupled human and natural systems where multifunctional composite of land-uses interfere. This paper presents a brief conceptual framework for developing an urban-rural gradient model, adapted with a set of postulates to systematically theorize the selection of research sites in larger urban agglomerations - formally the megacities. The urban gradient model helped to understand urban development prognosis and growth corridors. It is based on the presumption that it would examine samples from a variety of urban structural forms and functional characteristics; rather all measures of urbanization that would necessarily decrease with distance from the centre. It also transgresses the classical method of sampling along urban-rural transect or sampling grids. An example of framework implementation is presented to probe the pertinence of the proposed methodology. The case study was conducted in the megacity Karachi, Pakistan which is a city of 18 million inhabitants and the greater Karachi covers an area of 3600 km2. Results corroborate the urban gradient as an important research method rather than merely a modeling approach. The framework construct makes it adaptable to varying types of cities, especially fast growing cities with complex land-uses.

Maria Santos, doctoral candidate, University of California-Davis
Research Interest: Mediterranean ecosystems, spatial planning, conservation management, mammalian carnivores, habitat, GIS and remote sensing

Title: Contrasting ecology and culture in Mediterranean ecosystems of Portugal and California

Oak woodland persistence in Portugal and California may only be possible through the integration of both human and natural components. In this study we assessed how socio-economic systems (SES) and ecological requirements of wildlife species are linked to oak woodlands of Portugal and California. More specifically, we (1) applied the SES framework to four resources provided by oak woodlands: forestry, rangeland, agriculture and natural areas; (2) spatialized SES information into predictive models of wildlife presence using remote sensing data; and (3) analyzed how sensitive these resources are to future changes in land use and climate. We found that in both regions the sustainability of extractable resources may be threatened by replacement rate, land-use history, and interdependence with other resources. The non-extractable resources (natural areas themselves) are more susceptible and sustainable management is dependent on the voluntary nature of collective-choice rules. Inclusion of the SES spatial context of all four resources into models of wildlife presence produced the most parsimonious models, as the wildlife species responded to the heterogeneity of resource availability rather than to one resource alone. In addition, inclusion of productivity and stress parameters further improved the predictions of wildlife presence. These results demonstrate that the persistence of oak woodland SES’s is tightly linked with the persistence of current land use and productivity patterns, and changes in any of these parameters is likely to affect natural and human communities depending on them.

Mao-Ning TaunMu, doctoral candidate, Michigan State University
Research Interest: Spatial ecology, biodiversity conservation, climate change, GIS and remote sensing

Title: Effects of Human-Environment Relationships on the Spatio-Temporal Dynamics of Giant Panda Habitat

Habitat loss and degradation due to human activities are among major threats to biodiversity in the world. At the same time, conservation efforts have been increasing to mitigate the negative impacts of human activities through reducing exploitation of natural resources and actively restoring damaged landscapes. Therefore, understanding the effects of human-environment relationships, i.e., how human activities affect environment and how they respond to environmental changes, on the spatio-temporal dynamics of wildlife habitat is essential for conservation. In this study, we used the interactions between giant panda habitat and human activities in Wolong Nature Reserve and the adjacent Sanjiang Township, China as a case study to address these issues. While Wolong Nature Reserve sustains ca. 10% of the entire wild panda population, it is also home to ca. 4,500 local residents, and ca. 3,600 people in Sanjiang live just outside the reserve. We mapped panda habitat from 2001 to 2007 using remote sensing images and then related habitat change to anthropogenic factors (e.g., distance to local households and roads) to characterize the human-environment relationships. Overall, the amount of habitat increased ca. 10% in the reserve, which reversed the trend of habitat degradation in the past several decades. However, conspicuous differences in the patterns of habitat change and human-environment relationships were found between the reserve and Sanjiang Township. While panda habitat improved near local households in the reserve, it continued to degrade in Sanjiang, especially near roads and trails. The results suggested that different human-environment relationships determine the processes and patterns of habitat change, and the socioeconomic conditions of local people and implementation of conservation policies may explain the different relationships. This study not only has direct implications for panda conservation, but also increases our understanding of the complexity of human-environment relationships in a coupled human and natural system.

Molly Van Appledorn, doctoral candidate, University of Maryland-Baltimore
Research Interest: Watershed Ecology

Title: An analytical tool for delineation, assessment, and prioritization of riparian buffers for watershed management

Riparian areas have been a conservation priority because of their potential to attenuate nonpoint-source pollutants from upslope sources. Although the importance of riparian areas has long been recognized, land managers lack user-friendly tools that would facilitate regional planning by effectively scaling up understanding from local observations to entire watersheds. Here we present a strategic planning tool that integrates hydrologic characterizations with land-cover patterns to rapidly assess filtering potential within and among watersheds. This tool uses publicly-available data to 1) identify potential biogeochemically active zones based on a suite of topographic definitions, 2) calculate the width of buffers along preferential flow pathways from pollutant source areas to a stream network, and 3) prioritize stream-side areas for conservation or restoration based on the amount of source loading. We demonstrate the ability of the tool to calculate summary statistics such as mean buffer width and proportion of unbuffered cropland for all sub-basins within the Chesapeake Bay Watershed (CBW). We illustrate the sensitivity of the tool to regional differences in physiography and land cover patterns by comparing local buffering patterns within two case-study sub-basins in the CBW.

Eric White, faculty research associate, Oregon State University
Research Interest: Natural resource economics, forest sector responses to climate change and climate change policy, land use change, recreation economics

Title: Regional land use conversion involving forestry and agriculture land in the context of using biomass for bioenergy

Biomass is expected to be an important source of renewable energy as part of comprehensive climate change policy. Currently, biomass represents about half of U.S. renewable energy consumption (mostly associated with on-site energy production from timber mills) and the primary biomass feedstocks are waste residues (e.g., timber milling residues). Increased reliance on feedstocks produced specifically for bioenergy may yield changes in traditional production from the forest and agriculture sectors, the rate of land conversion between the two sectors, and resource management practices. As demand for bioenergy increases, biomass production from short-rotation woody crops (SRWC), logging residues, and non-merchantable trees will likely increase. Currently, SRWCs are estimated to comprise less than 0.1% of the agriculture and forest landscape. Timber harvests residues are currently believed to amount to about 64 million dry tons of woody biomass. Using existing research results, we identify considerations and some potential implications of increased use of woody biomass feedstocks. To examine possible future conditions, we use the Forest and Agriculture Sector Optimization Model—greenhouse gases (FASOM-GHG)—an economic dynamic optimization model of the U.S. forest and agriculture sectors—to project future biomass feedstock consumption under both reference and climate policy scenarios and possible effects on forest age classes, forest types, and other forest resource conditions. We report regional-level projections of feedstock consumption for future decades from the forest and agriculture sectors. Additionally, we examine projected impacts to land conversion (e.g., afforestation or deforestation involving agriculture) and management intensity as result of increased demand for bioenergy.