Graduate Programs

 

Penn State Cross-Disciplinary Statistical Ecology and Environmental Statistics

The pressing problems in environmental and ecological resource management can be addressed on theoretical grounds, but empirical observations and experimentation are required to test the theories, and importantly, to answer specific real-world questions. Experience has shown that in the general fields of ecology and environmental science, conventional research designs often are inadequate and the data involved usually have nonstandard features due to spatial and temporal variations and to social, economic, and political constraints. We cannot change natural selection and evolution, the behavior of organisms, and the complexities of population-community-ecosystem dynamics. Nor do we have perfect measuring devices for the many variables involved. Statistics becomes the key to dealing with the probabilistic nature of natural processes and their responses to man-made perturbations and stress.

Statistical theory and techniques already have contributed greatly to advances in the natural sciences and, together with technological advances in computing and other fields such as remote sensing, to environmental and ecological resource management. But serious problems remain. Definition and resolution of these problems require a multidisciplinary approach involving collaboration of biologists, ecologists, environmental scientists, social scientists and economists, and mathematical and applied statisticians. This collaboration also needs a coherent intellectual focus.

Several of the principal faculty participants at Penn State have been involved with multidisciplinary approaches to assessing and solving several real-world environmental problems of both qualitative and quantitative in nature. At Penn State, we have both the formal and informal mix of graduate programs in quantitative ecology and in environmental statistics which have largely resulted from the informal groups we have had on the campus for some years interested in statistics, ecology, and the environment.

Consider, for example, the research and training effort of the Center for Statistical Ecology and Environmental Statistics. Over the past twenty-five year period, we have had a variety of involvements with statistical methods and issues pertaining to environmental monitoring and assessment of a variety of ecological resources, such as: Forest Service, USDA: Spatial Statistics and Statistical Distributions, 1971; Stochastic Models and Sampling Designs for Forest Insect Populations, 1974; National Marine Fisheries Service, NOAA: Distributions and Geographic Distributions of Fish and Shell-Fish in Georges Bank, 1977-79; Chesapeake Bay Stock Assessment Committee: Monitoring, Sampling, and Assessment in the Chesapeake, 1985-1992; EPA Office of Policy, Planning and Evaluation: Mathematical Statistics and Stochastics for Data Interpretation and Improvement for Environmental Protection Research and Management, 1989-1993; Research and Outreach on Observational Economy, Environmental Sampling, and Statistical Decision Making in Statistical Ecology and Environmental Statistics, 1993- ; EPA Office of Research and Development: Research and Outreach on Sampling Design and Status Estimation for Environmental Monitoring and Assessment Program, 1993- ; NSF/EPA Research Program on Water and Watersheds: Statistical Approaches to Multiscale Assessment of Landscapes and Watersheds, 1995- ; EPA Office of Environmental Information, Advanced Raster Map Analysis for Sustainable Environment and Development Using Remote Sensing Data, 2000- . As per need, interest, and expertise, the faculty and research associates that have participated in these research and outreach activities include G. J. Babu, J. S. Fairweather, C. R. Goodall, S. D. Gore, A. Kaur, G. Lovison, W. L. Myers, J. K. Ord, Donna Peuquet, C. Taillie, S. K. Thompson, and several others. Several graduate students and interns have also benefitted from various disciplines, such as statistics, ecology, entomology, forestry, fish and wildlife, geoscience, plant pathology, and others.

Penn State is also uniquely positioned to proffer pioneering programs of graduate study in spatially-based statistical analysis of environment. The scope of such programs can extend well beyond the conventional concept of "spatial statistics".

The Office for Remote Sensing of Earth Resources (ORSER) (Co-Director: W. L. Myers) in the Environmental Resources Research Institute (ERRI, Director: Archie McDonnell) has formal roles in geographic informations systems (GIS) technology support for USDI, National Park Service (NPS) at the federal level and for the Department of Environmental Resources (DER) at the state level. Issues addressed in these support roles provide rich arrays of both real-world environmental challenges and extensive spatial databases on which to test new analytical approaches. ORSER has an interdisciplinary research focus which nicely complements the educational strengths of academic units.

The School of Forest Resources has an Analytical Landscape Statistics Laboratory (ALSL, Director: W. L. Myers) which serves both research and educational purposes. Like the facility at ORSER, this laboratory has multisystem GIS capability along with workstation Splus and GeoLink bridge between the statistical and spatial software domains. The School of Forest Resources is developing modularized programs of self-study in GIS whereby students with varying backgrounds can work progressively into GIS spatial technologies at an individually appropriate pace while minimizing course schedule conflicts.

The Department of Geography has an Advanced GIS Laboratory appropriate for group instruction in a classroom environment. This instructional facility supports a substantial complement of courses in spatial technologies offered by the Geography Department. The Geography lab has close liaison with the Earth Systems Science Center (ESSC, Director: Eric Barron) which is internationally known for its research in modeling of global hydrologic cycles using GCM approaches on super computers.

The Geography-ESSC and Forestry-ORSER spheres are linked across campus and beyond by the Penn State University computing backbone conduit to the Internet. The Land Analysis Laboratory in the Department of Agronomy is yet another facility with multi-system GIS capability. This laboratory has a special place in the campus distributed systems infrastructure by virtue of its global positioning systems (GPS) base station.

The College of Agriculture has long-standing emphasis on collaborative work in artificial intelligence and expert systems which complements the spatial and statistical capabilities interwoven through the campus fabric as outlined above.

The Penn State Department of Statistics is well known for its research and outreach in both disciplinary and cross-disciplinary statistics.

Penn State Graduate Program in Environmental Statistics

PhD in Statistics with Environmental Statistics Option. The requirements for the environmental statistics option are similar to those for the Ph.D. in statistics. They differ slightly in the course work. With the environmental statistics option, the student receives a Ph.D. in statistics, with a coherent and intensive specialization in environmental statistics. The difference in required courses is that 6 credits of mathematical analysis may be substituted for the 6 credits in advance probability.

Elective Courses. 1. Environmental statistics electives: 15 credits from 500-level courses on topics involving environmental statistics as approved by the program committee. Topics courses

may come from such areas as statistics, geography, remote sensing, quantitative ecology, resource modeling, environmental epidemiology, environmental policy, etc. Designated statistics electives include several statistics courses, such as Ecometrics, Quantitative Ecology, Statistical Ecology Spectrum, Environmental Statistics, and Statistical Distributions in Scientific Work; and 2. Mathematical statistics and probability: 9 credits from 500-level courses such as Stochastic Processes, Probability Theory, Categorical Data Analysis, Nonparametric Statistics, Multivariate

Analysis, Statistical Decision Theory, and Spatial Statistics.

Master's Degree in Statistics with Environmental Statistics Option. The goal in proposing a master's degree with environmental statistics option is to recognize degree candidates who have pursued a coherent academic program in environmental statistics. The requirements of the environmental statistics option are, in addition to those of the master's degree in statistics, at least 12 credits in topics involving environmental statistics.

The Advisory Committee on Environmental Statistics. An advisory committee on environmental statistics will consist of fifteen members at a time, six from statistics and nine from participating substantive departments. The role of the committee will be two-fold. First, to act as a focal point in graduate education in environmental statistics. This includes student advising, for students whose primary affiliation is statistics, but also for students whose primary affiliation is elsewhere with a strong interest in environmental statistics. Second, to take the initiative in obtaining additional support for environmental statistics, including training and internship grants, and to provide research assistantship funding to students in their last two years.

Penn State Graduate Quantitative Ecology Program

Ecology has entered a phase of development wherein much of the subject matter is approached by quantitative reasoning. This occurs through application of deterministic or stochastic models to theoretical questions in ecology, or it occurs through sophisticated use of applied statistics for experimental design and hypothesis testing. We experience greater fusion of these approaches as new generations of researchers become better educated in quantitative methods. Thus, there is a demand not only for the practitioners of quantitative ecology but also for field ecologists who are more accomplished in quantitative techniques.

The importance of quantitative ecology is apparent in recent developments of employment opportunities, both at academic and applied levels. A large fraction of job descriptions include mathematical modeling or statistical capability in addition to the ecological subject matter; this is a trend that will, undoubtedly, continue. The trend is visible in the increasing devotion of journals to the quantitative aspects of ecology; relatively new journals such as Theoretical Population Biology and Ecological Modeling join The American Naturalist as exclusively devoted journals, but also Ecology, Journal of Animal Ecology, Oikos, etc., are becoming increasingly quantitative in content.

Penn State has long been producing graduates in the field of quantitative ecology through the Biology, Ecology, and Statistics Graduate Programs. In order to centralize and formalize the student's education in this area, the Quantitative Ecology Option for the M.S. or Ph.D. in Ecology has been instituted. Graduation with this option will document the student's preparation and competence to contribute to this increasingly important area of science.

Program of Study. This intercollege program emphasizes the properties of ecosystems by focusing attention on interactions of single organisms, populations, and communities with their environment. It is designed to give students a basic understanding of ecological theory and research techniques and is complementary to other Penn State environmental programs that emphasize man's role in ecosystems. The program is administered through the Graduate School by an Interdisciplinary Committee on Ecology. The faculty comprises more than 45 members from twelve academic departments. Both Ph.D. and M.S. degrees are awarded. To ensure breadth of training in the fundamentals of ecology, students are required to take at least one course from each of three core areas: physiological ecology, population ecology, and community/ecosystem ecology. Beyond these basic courses, course requirements are determined by the student and the student's adviser and graduate committee. Completion of a thesis acceptable to the committee is required for both degrees.

The ecology program offers an option for the Ph.D. degree emphasizing the quantitative aspects of ecology, including mathematical and statistical modeling of ecological phenomena and applications of statistics to experimental design and data analysis. The option entails some extra course requirements plus a thesis in quantitative ecology directed by a member of the quantitative ecology faculty.

Research Facilities. The University Park campus is ideally situated for research in varied ecosystems ranging from undisturbed upland bogs, forests, streams, and lakes to areas severely affected by strip mining, deforestation, sewage effluents, and agriculture. Penn State has an excellent biological library. Collections include herbaria for mosses, fungi, and higher plants; excellent fish, herp, mammal, and bird collections; and the collection of the Frost Entomological Museum. Penn State has extensive animal-care facilities, experimental gardens, greenhouses, climate-controlled growth chambers, and well-equipped laboratories for air pollution monitoring, studies of plant and animal physiological ecology, and continuous culture of aquatic organisms. Remote-sensing equipment for photoanalysis and digital data analysis is available, as are completely equipped photogrammetry and photointerpretation laboratories. The University has a full range of computer facilities, including many terminals and many microcomputers. Consultants in statistics and computer science are available to graduate students. Students in the ecology program have access to the Stone Valley Experimental Forest, state gamelands, state forestlands, and the Spruce Creek Experimental Area.