Jennifer Pontius

Title: Research Ecologist
Unit: Center for Research on Ecosystem Change
Previous Unit: Ecology and Management of Northern Forests
Address: Northern Research Station
271 Mast Road
Durham, NH 03824
Phone: 603-868-7739
E-mail: Contact Jennifer Pontius

Jump to Publications

Education

• Ph.D. in Earth and Environmental Sciences, University of New Hampshire, Durham, New Hampshire, September 2004
  Dissertation: Assessing Hemlock Woolly Adelgid Infestation, Hemlock Health and Susceptibility Using Hyperspectral Technologies
• M.S. in Natural Resources, University of New Hampshire, Durham, New Hampshire, May 1998
  Thesis: The influence of micro-environmental factors on sugar maple demographics in the White Mountains of New Hampshire: A study of potential migration
• BA in Environmental Science with a concentration in Terrestrial Ecology, May 1993
• BA in Spanish, University of Virginia, Charlottesville, Virginia, May 1993

Current Research

• My goal is to help develop hyperspectral remote sensing techniques to directly assess forest decline and species distribution on a landscape scale. To date, this work has focused on the detection and mapping of pre-visual decline symptoms in hemlock resulting from hemlock woolly adelgid infestation, using hyperspectral remote sensing imagery. Because foliar chemistry can also be mapped with hyperspectral instruments, we are also developing  GIS-based decline susceptibility models that include foliar chemistry, a potentially influential factor in forest decline rates.
• More recently, my work has expanded to include hardwood decline and forest species mapping. These techniques provide a much-needed tool for the early detection of new and existing stressors and will allow forest management agencies to focus management efforts before stands are severely impacted. Most recently, we have begun an effort to transfer this technology to commercially available hyperspectral sensors for more widespread application.

Why is This Important

The potentially severe consequences and large scale of invasive insect infestations require that scientists and land managers identify new infestations early on and target management activities where they are most likely to succeed. Field-based methods are time consuming and only offer spot information. Having the ability to detect trees in the very early stages of decline, and predict anticipated reates of decline across entire landscapes is crucially important to the development of integrated pest management strategies for managing infestations. We have found that hyperspectral remote sensing can be used to identify forest stands in the very early stages of decline, often before visual symptoms are even apparent on the ground. These sensors are also capable of mapping foliar chemistry, a tool which we are currently using to map the potential rates of decline in combination with landscape variables typically available in a GIS model. Utilizing these methods, forest service scientists are working to ensure that these high tech, remote sensing instruments will soon be used by land managers across the northeast in forest health management efforts.

Future Research

• Our next steps are to expand our current hyperspectral methods to multiple species, including hardwoods affected by invasives such as the emerald ash borer and Asian longhorned beetle. The ultimate goal is to develop cookbook methods for commercially available sensors that are available to land managers for application.
• We would also like to use these technologies to provide informative data layers to GIS models for forest susceptibility to external stressors such as invasive species, atmospheric deposition, and global warming. Such models can identify regions of potential impact to surface water quality, wildlife habitats, or economic resources. Pulling the data layers we, and other scientists, create into landscape-scale ecosystem models is critical.

Featured Publications