A decade after Romesburg admonished wildlife biologists to establish and test hypotheses to gain more "reliable knowledge," we have added an incentive to bring rigor to our science. Wildlife biologists are finding themselves defending their science against often savage criticism. At least 2 factors are central to producing solid, defendable science: (1) the rigorous application of scientific methods and (2) the development of clear operational definitions for terminology. The hypothetico-deductive (H-D) process, in the form of statistical tests of hypotheses based on experimental data, is hailed as the superior means of acquiring strong inference and reliable knowledge. Results from experimental studies, however, are seldom available, and most management decisions are made on the basis of incomplete information. We argue that even in the absence of experimental information, the H-D process can and should be used. All management plans and conservation strategies have properties that can be stated as falsifiable hypotheses and can be subjected to testing with empirical information and with predictions from ecological theory and population simulation models. The development of explicit operational definitions for key concepts used in wildlife science-particularly terms that recur in legislation, standards, and guidelines-is a necessary accompaniment. Conservation management and planning schemes based on the H-D process and framed with unequivocal terminology will allow us to produce wildlife science that is credible, defendable, and reliable.
