Stafford, S. G.; Klopsch, M. W.; Waddell, K. L.; Slagle, R. L.; Alaback, P. B. 1986. Optimizing the computational environment for ecological research. In: Michener, William K., ed. Research data management in the ecological sciences. The Belle W. Baruch Library in Marine Science No. 16. Columbia, SC: University of South Carolina Press: 73-91.
Computational tasks for ecological research database management, statistical and graphical analysis, word processing,communications, and accounting are diverse. Hardware and software options should be considered as a unit to create the most useful computational environment. Hardware includes mainframes, minicomputers (minis), microcomputers (micros), and supermicrocomputers (supermicros), and such attendant peripherals as printers, plotters, and tape drives. Ease of use, low cost, fast response time, and potential graphics capability make micros a popular choice for interactive computing (statistical analysis and wordprocessing). Software, which must be reliable and easy to use, includes statistical analysis, graphics, data management, word processing, modeling, accounting, and communications. Graphics software can be a powerful medium for presenting research results. Restricting equipment use to specific tasks and appropriately scheduling machines, providing manuals and short courses, and carrying maintenance contracts are important to maximizing output. Resources should be shared through networking to make the best use of existing equipment. Hardware acquired should be adequate for the job intended, and choice of vendor and support level must be carefully evaluated. Developing in-house softwar eshould be avoided if commercial packages are already available; a table of statistical software for micros, minis, and mainframes is provided. Although computing systems have hidden costs, the potential gains remain high for all but the most modest research projects.