The analysis and web output parts of the Avian Knowledge Network are still under development, and as such impacts are only starting to be apparent. Currently, the most obvious impact is that we have started altering "best practices" for the use of existing data resources by introducing a suite of analysis techniques unfamiliar to most ecologists. These novel (to ecologists) techniques have the advantage of being more likely to identify important influences on birds, and hence identifying appropriate targets for management action.
impact statement issue
In order to effectively manage North America's wildlife in the face of ever-increasing human impact on our environment, managers and conservation biologists need to be able to: (1) know where each species occurs and (2) understand the environmental features that cause each species to be present in some areas but not others. While all of the needed information is known for a few, intensively studied species, little is known about many other species, which hampers management efforts.
impact statement response
Our work provides wildlife managers an initial screening of hundreds of environmental features that potentially impact species, thus providing the initial knowledge with which to guide efficiently focused research and management efforts. We are designing processes for automated analysis of all major sources of bird monitoring data in North America, creating a web site that allows interested parties access to the raw data and analyses, and methods for exploring the results from analyses in order to identify the patterns of greatest interest to anyone interested.
impact statement summary
This project is the third facet of the creation of the Avian Knowledge Network (www.avianknowledge.net), an initiative to: (1) bring together all available bird monitoring data into a single, easily accessible repository; (2) produce exploratory analyses of these data in order to identify environmental features that have a high likelihood of affecting birds' distribution and abundance; and (3) create mechanisms for extracting from these analyses the patterns and relationships most relevant to an individual wishing to use analysis output. In order to predict bird distributions and understand the environmental factors that control these distributions, we are introducing new techniques for data analysis to ecologists. These techniques are both existing (machine-learning) methods from the computer sciences, as well as creating new techniques that merge the strengths of data analysis techniques from the computer and statistical sciences.