|
Wildlife and Fisheries Sciences and the Scientific Method |
Ratti and Garton (1994) briefly described the scientific method as “a circular process in which previous information is synthesized into a theory, the theory is stated explicitly in the form of hypotheses, predictions are deduced from these hypotheses, the predictions are tested through experimentation or observations, the theory is modified or expanded on the basis of results of these tests, and the process starts again.” Hypothesis testing is central to the scientific method and requires formal analysis (i.e., use of statistics).
A hypothesis is a proposition or supposition regarding a theory that is tentatively accepted to explain known phenomena or to provide a basis for further investigation. A working hypothesis generally represents what is initially considered to be the most likely explanation of a phenomenon. Usually several alternative research hypotheses are formulated; these are based on a variety of possible explanations or reasons for facts observed. From general research hypotheses, statistical hypotheses are formulated. General research hypotheses represent theories, while statistical hypotheses represent predictions from theories (James and McCulloch 1985). Hypotheses are rejected or fail to be rejected depending on study results. When data support a hypothesis, it cannot be concluded that the hypothesis is true, only that it has not been rejected. In addition, a hypothesis is only rejected or not rejected at some statistical level. The scientific method cannot “prove” that a hypothesis or theory is correct, only that alternative hypotheses or theories are rejected. The potential correctness of a hypothesis or theory increases as alternate hypotheses are rejected.
It is common to hear statements about theories such as “there is no scientific proof,” “it has been proven scientifically,” or “the hypothesis was accepted.” Such statements indicate a lack of understanding about the scientific method. A hypothesis can only be rejected or not rejected; that is all.
The major drawback to the scientific method is that it can never “prove” that theories such as evolution, gravity, relativity, or global warming are correct. However, we can continue to test hypotheses associated with these theories and from the results, strengthen or weaken the theories. The primary strength of the scientific method is that when studies are correctly designed, implemented, and interpreted, human biases and beliefs are eliminated from the equation. There is a great difference between human belief and human knowledge; the scientific method provides knowledge-based information. Theories or hypotheses that cannot be scientifically tested are not really theories or hypotheses – they are beliefs.
Many facets of the study of wildlife and fisheries are not and probably never will be like the “pure” sciences. In “pure” sciences, such as chemistry and physics, there is a greater ability to control and modify variables associated with an experiment. This ability to control and modify variables results in experiment repeatability under the same conditions to confirm previous results. In wildlife and fisheries, or any biological science involving field work, it is more difficult to control and modify variables. For example, the population dynamics of Cactus Wrens in a certain locale may be determined. These determinations could be correctly made and then not be repeatable (either at a different location or even at the same location) because of differences in items such as mortality rates, growth characteristics, or hatching success resulting from variations in weather conditions, habitat modifications, or any number of other changeable factors. Some variables associated with the study of wildlife and fisheries are controllable in laboratory situations, and valuable information can be obtained from laboratory studies, but such information is not necessarily applicable to non-laboratory settings.
Much of our knowledge of wildlife and fisheries is based on observations and conclusions derived from associations, rather than from experimental tests of hypotheses (Ratti and Garton 1994). Such knowledge is correlative rather than indicative of cause and effect; thus, interpretation may be inaccurate and misleading. Romesburg (1981, 1991) and Johnson (1999, 2002) have been critical of this aspect of wildlife and fisheries as sciences. An example of knowledge obtained through association (often called a descriptive-type study) would be sampling a fish species from a number of locations, examining their stomach contents, and reporting what was eaten. The study is descriptive because it does not ask the “why” question, i.e., why are fish eating what was found, and a manipulative study was not designed to test a prediction. While descriptive-type studies are still conducted in wildlife and fisheries, there has been a strong trend in the last couple decades to test specific hypotheses and produce research results that are not solely based on associations. This is not to say that information based on an association has no value. Many hypotheses (generally called retroductive hypotheses) have been developed based on information obtained from descriptive studies and many untested hypotheses will evolve in the future from such information. However, hypothesis-testing research (i.e., using the hypothetico-deductive method [Romesburg 1981]) usually provides more reliable information and typically is the next step after completing a descriptive study.
The primary factors that have allowed wildlife and fishery biologists to progress from association-type studies to hypothesis-testing studies are the advent of improved statistical tests (i.e., Information-Theoretical Approach, Burnham and Anderson 1998), the almost universal use of computers, and agencies willing to invest the time and monies needed to conduct such studies. Increased complexity and power of statistical tests and constantly improving technology associated with computer hardware and software have combined to allow for greater research rigor in wildlife and fisheries.
The state of our current knowledge of wildlife and fisheries is incomplete. Because of this, biologists are often faced with the dilemma of making decisions based on insufficient information. This dilemma leads to questions such as – “Is the controlled burning of this piece of grassland at this time really the best manner of managing this resource?” or “Are you sure that this is the most effective way to increase this fish (or wildlife) population?” The answers to such questions are not “yes” or “no,” the answer should be “based on our current knowledge this is the best practice that can be implemented.” As our knowledge base increases we often learn that the best practice at some previous time was really not the most appropriate solution. Although this modifying of solutions has been called “waffling” by some, and the term is used to diminish credibility, the approach is unavoidable in science, as well as in other practices, where the objective is to find truth. Under the circumstances, all a biologist can do is use the best information currently available and then strive to obtain better information using the scientific method.
Literature Cited:
Burnham K. P., and D. R. Anderson. 1998. Model selection and inference; a practical information-theoretic approach. Springer-Verlag, New York.
James, F. C., and C. E. McCulloch. 1985. Data analysis and the design of experiments in ornithology. Pages 1-63 in R. F. Johnson, editor. Current ornithology. Volume 2. Plenum Press, New York, NY.
Johnson, D. H. 1999. The insignificance of statistical significance testing. Journal of Wildlife Management 63:763-772.
Johnson, D. H. 2002. The role of hypothesis testing in wildlife science. Journal of Wildlife Management 66:272-276.
Ratti, J. T., and E. O. Garton. 1994. Research and experimental design. Pages 1-23 in T. A. Bookout, editor. Research and management techniques for wildlife and habitat. 5th Edition. The Wildlife Society, Bethesda, MD.
Romesburg, H. C. 1981. Wildlife science: gaining reliable knowledge. Journal of Wildlife Management 45: 293-313.
Romesburg, H. C. 1991. On improving the natural resources and environmental sciences. Journal of Wildlife Management 55:744-756.