The intellectual merit and broader impacts of this grant combine to help ecological research maximally benefit from the large amounts of data that are increasingly available for ecological systems and from new general theories that predict large numbers of ecological patterns simultaneously.

Our research compares approaches to modeling ecological systems to find models that accurately describe patterns of diversity across different ecosystems and types of species. We do this by comparing the predictions of theoretical models to large compilations of ecological data from citizen science, government surveys, and academic research. We found that a new group models based on statistical physics showed potential for explaining a number of ecological patterns with only a small amount of information, but that these models exhibited several important deficiencies. We found that simple models based on birth, death, dispersal, and growth of organisms provided better descriptions of ecological patterns than the statistical physics approaches. We also developed a new set of methods for understanding how much information about ecological processes is present in different ecological patterns by describing the possible forms that these patterns can take. In combination our research is helping to change our approaches to testing ecological theory by leading the way in using data from across ecosystems, across the diversity of life, and across the array of predictions made by may theories.

Our broader impacts involved educational and research tool initiatives that are targeted across scientific and engineering disciplines with the intent of broadly improving the acquisition of data skills and development of data scientists. This makes the kind of data-intensive research we do possible for the next generation of scientists through a combination of training, software, and access to information. We trained thousands of scientists online and hundreds in classrooms and workshops on how to efficiently create, store, manage, and analyze data. We created software that is widely used to automate data cleaning and restructuring tasks so that scientists can focus on doing science. We created online resources to help scientists quickly find the data they need to answer a question and learn how to work with it. We wrote a series of papers on best practices for data management and scientific computing that have been widely read and are having a major impact on how scientists conduct research.

In addition to these broader community wide efforts, seven graduate students (including a veteran, two first-generation college students, an underrepresented minority, and a student with a chronic illness), two undergraduates, and two postdoctoral researchers, participated in this research and received in-depth training in advanced data and computational approaches to science. This prepared them for jobs in a diverse array of industries. Trainees from this grant are currently employed in the technology industry, government funded research organizations, research universities, and higher education.

Last Modified: 05/23/2016
Modified by: Ethan P White