This project links networks of undergraduate ecology courses to carry out a collaborative, hands-on research project focusing on a large-scale analysis of habitat requirements for amphibians. For many students, this project is their first experience working with large data sets, and the project's methods are designed to sharpen the highly transferable skills (i.e., data integration and analysis) necessary for doing science in the digital age. Students in each course match United States Geographical Service data on amphibian populations from their own region with habitat data from Google Earth. Their data are submitted to a shared database that is analyzed to determine amphibian-habitat associations.
Intellectual merit: The project provides useful information on the effects of land use on amphibian populations in the U.S. It also augments an important public database by adding habitat and landscape data for each amphibian survey location. Broader impacts: From an educational perspective, this project provides an authentic research experience for up to 400 students from a diverse group of schools including two-year colleges, public research universities, and historically minority-serving universities. Assessment of the project focuses on the effects of a collaborative research experience on students' interest in science, student attitudes about science, and student retention in science programs.
This project is being jointly funded by the Directorate for Biological Sciences, Division of Biological Infrastructure and the Directorate for Education and Human Resources, Division of Undergraduate Education as part of their efforts toward Vision and Change in Undergraduate Biology Education.
Toads, Road, and Nodes is a collaborative project between University of California, Santa Barbara’s National Center for Ecological Analysis and Synthesis (NCEAS) and Washington and Lee University. This project had the following major goals: 1) Create a data-based research project on the landscape ecology of amphibian populations for use in undergraduate ecology and conservation biology courses, enrolling undergraduate classes from different universities to compile and analyze data for their own region. 2) Host a meeting in which student and faculty representatives from participating classes come together to compile the data across classes and begin a pooled analysis of the data. 3) Use this project to provide a data-based research experience for roughly 200 undergraduate students and a more focused workshop experience for an additional 10-15 students. 4) Perform research of sufficient quality to produce novel scientific results that can be published as a stand-alone manuscript. 5) Use pre- and post-course surveys to measure what students gained from the project in terms of self-reported changes in confidence, knowledge, and experience with project learning objectives and to measure changes in attitudes towards scientific research, data analysis, and collaboration. The schools involved in this project were Eckerd College, University of South Carolina-Salkehatchie, Warren-Wilson College, Virginia Commonwealth University, Northern Virginia Community College, Hobart and William Smith Colleges, Clarkson University, University of Rhode Island, Anoka-Ramsey Community College, and Utah State University. Approximately 250 students from participating classes worked on the project during the winter/spring term, usually as a 4-6 week course unit. All participating faculty and 1-4 students from each class met at NCEAS to compile and analyze data. Additional tasks were assigned to both students and faculty following that meeting. Each class was assigned a data set, usually data from their home state. The project protocols and data were publicly shared on a project website. For most students and some faculty, this was their first significant exposure to Geographic Information Systems, analysis of large, existing datasets, and statistical analysis in the R computer programming language. Research focused on the following questions: a) over what spatial scales do landscape factors (road density, development, forest cover, agriculture, and wetland area) have the most influence on amphibian distributions, b) do the influences of roads on amphibian distributions depend on road type (e.g. primary versus secondary versus rural roads), and c) do different intensities of development have different kinds of effects on amphibians. Although data analysis is ongoing, the answers to these questions appear to be: 1) Factors varied in terms of the spatial scales over which they were most important. Roads and development had most effect at small scales, close to amphibian populations (e.g. <600m). Landcover (e.g. forest and agriculture) tended to have effects that were small but consistent over a range of scales up to several kilometers. There was also quite a bit of variation across regions, with the effects of agriculture particularly strong in the Midwest and the effects of roads particularly strong in the Northeast. 2) Across regions, negative correlations between road density and amphibian richness were stronger for primary/secondary roads than they were for rural roads. 3) When we controlled for the effects of road density and traffic, there were only weak effects of development on amphibian distributions across a range of development intensity. In addition to the scientific research, surveys were created to measure student learning. We received pre- and post-project surveys from the majority of the students. Collaborators at Purdue University are currently working on analyzing these survey data to determine what students gained from the project in terms of changes in confidence, knowledge, and experience with project learning objectives and to measure changes in attitudes towards scientific research, data analysis, and collaboration. NSF funding was provided to allow 10 students per year to attend a workshop and data analysis meeting at NCEAS. Participating campuses contributed additional funding to allow 18 more students to attend. These students received training in data management, compilation, and analysis. One manuscript has been accepted for publication in Biological Conservation and two more are planned when analysis is completed by Washington and Lee University. In addition to the outcomes listed above, several other unanticipated benefits arose from the project. Students at most participating campuses ended up doing oral or poster presentation based on the project at campus-wide or local conferences. A number of students ended up doing spin-off projects as independent research. Participating universities often supported additional students to attend a research conference in April. Several faculty members developed new lines of student-mentored research as spin-offs of the project. We hope that the success of this project will demonstrate the efficacy of a new model for education and research on large datasets.
