Biocomplexity and the Art of Planetary Management

PARTICIPANTS
David Vieglais, University of Kansas Natural History Museum
David Stockwell, SDSC
Robert Waide, University of New Mexico

EOT-PACI
Government and New Communities

reader perusing a description of Kris Krishtalka's Earth Systems Science thrust area project on biodiversity might mistake it for an essay on Tibetan Buddhism. At its heart is the complex interconnection between all living things on Earth, and a goal of the project is to help individuals make informed decisions about how to maintain the delicate balance between creatures and environment. By creating a modeling system marrying biodiversity and ecology, Krishtalka envisions a tool to aid in what he calls "planetary management."

For years, biodiversity and ecology researchers have not worked closely together, a division that Krishtalka, director of the University of Kansas Natural History Museum, is glad to see disappearing. "Rita Colwell, the new director of the National Science Foundation, is very concerned about what she calls biocomplexity, which is essentially the merging of biodiversity and ecology studies with the study of the Earth's physical and chemical systems, and with human social systems," he said. "Research in each of these fields has an integral relationship to research in the other, and cumulatively they can accurately describe the interplay between species and their supporting environments, and help us understand how they co-evolve over time."

Ecological research has enjoyed great support in the United States, especially after the publication of Rachel Carson's Silent Spring in the early 1960s. The book investigated the relationship between environmental conditions and human-introduced contaminants such as pesticides, and spawned greater public awareness of ecological issues. Since then, such facilities as the NSF's Long Term Ecological Research (LTER) stations have been created and are devoted to monitoring, describing, and understanding how ecological systems across the nation function.

A recent biodiversity-scale modeling workshop held at SDSC, organized by Robert Waide of the University of New Mexico and John Helly of SDSC, brought together ecologists--including several from LTER sites--and biodiversity researchers to discuss where species data are still needed to complete both ecology and biodiversity models. "Learning where we fit together is an important part of our project," Krishtalka said. "There are questions to be answered like 'What role does biodiversity play in controlling ecosystem function and evolution?' or 'How much species extinction can an environment tolerate before it starts to deteriorate?'" As biodiversity and ecology studies become more integrated, accumulating species data becomes another major focus.

ACCESSING UNTAPPED RESOURCES

SHARING DATA

FOUNDATION FOR KDI EFFORTS

Figure 1. Biodiversity Species Workshop

The Species Analyst software developed by David Vieglais at the University of Kansas Natural History Museum interacts with the Biodiversity Species Workshop modeling software developed by David Stockwell at SDSC, allowing distributed analysis of biological data.

ACCESSING UNTAPPED RESOURCES

Scientists know there are between 20 and 50 million species on Earth, only five to 10 percent of which have been discovered. "There are plants and animals in many habitats, especially in the rain forests, that we haven't found or described yet," Krishtalka said. Most of the species that have yet to be discovered are arthropods and microbes that live in soils. The rest are freshwater, terrestrial, or marine invertebrates.

Beyond the sheer number of species, another challenge is introduced by a primary goal of biodiversity research--to understand the distribution of species through time and across geographic space. "Species and ecosystems evolve, and there is a push-pull relationship between the two," Krishtalka said. "In order to understand both ecological and evolutionary patterns and processes, we need to study how they have changed and affected each other over time. This is especially important for predicting how evolution will continue to occur based on current and future ecological conditions."

Luckily, biodiversity researchers have an enormous untapped resource available to them in the collections of natural history museums and herbaria around the world. "For nearly 300 years, scientists and collectors have meticulously sampled the Earth's various environments," Krishtalka said. "That provides us with three centuries of data across the geographic spectrum. If we add paleontological data to that, we have sample data going back 3.8 billion years." In addition, existing data can now be used in ways never before imagined, such as DNA sequencing. "We can construct a genetic-level history of a species using collection data in this way," Krishtalka said.

Accessing the data, however, can be difficult and tedious. For years, scientists have traveled to host museums or herbaria and combed through card files. Many museums, however, have now converted their collection records to databases, and Krishtalka's colleague Dave Vieglais, a research associate at the Kansas Natural History Museum, has developed software to link these databases around the globe, thereby creating an accessible community of disparate museum data. A video describing applications and research using such museum data is on SDSC's Biodiversity Web site.

Figure 2. Migration Model

Afrom an animation model of the migration of bird species Myiarchus swainsoni, or Swainson's Flycatcher. This collaborative project with Leo Joseph of the Philadelphia Academy of Natural Sciences is helping identify the migration patterns of long-distance migratory birds.

SHARING DATA

"A critical part of the Species Analyst--the software package we've created--is that it doesn't require the museum or herbarium to use any particular type of database software," Vieglais said. "Each institution can maintain their own database and they have complete control over their own data. A query through the Species Analyst, however, presents the user with data in a consistent visual format, regardless of the database format on the institution's side."

A user's query is simultaneously submitted to all museum databases employing the Z39.50 protocol on their servers. "Excel and ArcView are popular 'front doors' for viewing data," Vieglais said, "but it can be set up to use virtually anything." Information returned includes a description of the collection specimen, as well as the date and latitudinal and longitudinal information on the collection site.

The Species Analyst also interfaces with predictive distribution modeling software by David Stockwell of SDSC, and running on the center's Sun ES10000 server (Figure 1). "We need to account for gaps where species are likely to occur, based on environmental conditions, but where specimens haven't been collected," Stockwell said. "We also want to be able to layer the collections of multiple institutions to get a better prediction of the number of species found in a given region." The predictive modeling software allows a user to specify a species, acquire data on the species from the museum collections, and then map its distribution across a region (Figure 2). The user can also attach his own database to the modeling software over the Web and run models based on that data. "Eventually, these predictive models will be able to interact with ecosystem forecasting models, allowing a user to predict how environmental factors will impact the distribution of a species given statistics about the current conditions," Stockwell said.

Currently, 14 museums have implemented the Z39.50 database protocol, allowing researchers access to their collection data. "This is our next challenge," Vieglas said, "to get other institutions to employ the software and enable access." To spread the word about the Species Analyst, Vieglas and Krishtalka make the rounds of curator and museum association meetings, demonstrating the software and discussing the importance of museum-based data for biodiversity research.

FOUNDATION FOR KDI EFFORTS

Recently, Krishtalka's team--which includes researchers at the University of Kansas Natural History Museum, SDSC, the California Academy of Sciences, UC Berkeley, the Missouri Botanical Garden, the Museum of Southwestern Biology and LTER Network office at the University of New Mexico, the National Center for Ecological Synthesis and Analysis, and the Biological Resource Division of the Department of the Interior--received an NSF Knowledge and Distributed Information (KDI) grant to continue their biodiversity modeling work and integration with other ecological and ecosystem models. "The grant is going to enable us to go even further with the work we've started with NPACI's support," Krishtalka said.

The team will initially focus on a particular phenomena that concerns ecologists and biodiversity researchers alike--the disappearance of several species of amphibians in protected California wetlands. "Since the areas are protected, there shouldn't be any land development factors driving the disappearance of these species," Krishtalka said. "So it has to be something else. It may be related to our alteration of the environment and the impact that is having on soil, runoff water, climate, spread of disease, or food supply."

Their study will combine ecosystem and ecology models of the region with predictive distribution models for species that should be found in the area.

"The idea is to determine the causal factors driving the disappearance of the species, and then determine how those could be mediated," Krishtalka said. "This is excellent practice for developing other models that can actually inform public policy regarding land use and its impact. Ultimately, this should help us protect our valuable natural resources and practice better planetary management." --AF