A century ago H.G. Wells commented that "statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write." Each day of our lives we are exposed to a wide assortment of statistics pertaining to such phenomena as stock market activity, unemployment rates, medical research findings, opinion poll results, weather forecasts, and sports data. Frequently, such information has a profound impact on our lives.1
Statistics
also have an important role in science and in the application of new information
to the pressing environmental issues of today. The use of statistics in
natural resource management can improve understanding of complex problems
and allow for more informed decisions. By "making sense of
data," statistical techniques enable researchers and managers to
describe habitat requirements, test ideas and concepts, evaluate management
alternatives and decisions, and predict changes.
The ready availability of high-speed personal computers has revolutionized the field of statistical analysis for scientific investigations, allowing for a variety of new estimation and inference procedures. U. S. Geological Survey statistician Brian Cade and programmer Jon Richards recently contributed to this growing field by releasing a Windows version of the Blossom software, a statistical software package used for a variety of natural resource applications. The statistical procedures used in Blossom are unique from those offered in commercially available software and include the multi-response permutation procedures, recently featured in a statistics textbook by Mielke and Berry 2001, and regression quantile estimates and permutation tests for linear models. The availability of these procedures allows scientists to better address natural resource issues important to managers.
For example, Blossom software was used to help determine the influence of periodic emergency haying on the composition and structure of grassland vegetation on Conservation Reserve Program lands. This study provided information on the long-term effects of emergency haying on wildlife habitat that could then be used to refine management guidelines. Blossom was also used in a study designed to quantify the relationship of grassland bird abundance to landscape characteristics in the urban open space around Boulder, Colorado. A final example can be found in the application of Blossom to estimating self-thinning relations in plants. Here, resulting information provided a better understanding of how competition for resources among plants affects plant performance. In all of these applications, Blossom provided additional information, or information that might have been missed using conventional statistical methods, by allowing scientists to use measures of variation more relevant to the ecological questions being asked.
1 Excerpted from Berenson, M.L. and D. M. Levine, 1990. Statistics for business and economics. Prentice-Hall, Inc., Englewood Cliffs, NJ. p. 1.
U.S. Department of the Interior | U.S. Geological Survey
URL: http://www.fort.usgs.gov/resources/spotlight/blossom/Default.asp
Page Contact Information: AskFORT@usgs.gov
Page Last Modified: 12:29:09 PM
