American Society of Naturalists

A membership society whose goal is to advance and to diffuse knowledge of organic evolution and other broad biological principles so as to enhance the conceptual unification of the biological sciences.

E. O. Wilson Award Address: “Where is natural history in ecological, evolutionary, and behavioral science?”

Posted on

Joseph Travis (July 2020)

Read the Article (Just Accepted)

There is an inherent tension between natural history, the careful observation of nature, and experimental science, the search for general principles of cause and effect. This tension is especially taut in the sciences that study nature’s variety: ecology, evolutionary biology, and behavioral biology. This paper examines the role that natural history plays in the scientific process, asking a series of questions about that role. Can scientists rely on observations in natural history to test scientific predictions? If experiments are necessary, how can natural history affect the relevance of experiments? In particular, the paper examines the pros and cons of two different approaches to science in these disciplines. Should scientists begin with an important hypothesis and find the best organism with which to test it? This is the foundation of what is called “question-driven science.” Alternatively, should scientists work from the opposite direction, that is, be careful natural historians so that they can ask which important hypothesis is most suited for testing with a particular organism? The paper offers a synthesis of these approaches, suggesting that strong science is built by having a command of natural history so that the scientist allows nature to tell her or him or them which question to ask instead of demanding that nature answer the question she, he, or they find most interesting.

<i>Hyla andersoni</i>, the Pine Barrens Treefrog, adult male in North Carolina.<br />(Credit: Joseph Travis)
Hyla andersoni, the Pine Barrens Treefrog, adult male in North Carolina.
(Credit: Joseph Travis)

The Pine Barrens Treefrog, Hyla andersoni, is found in acidic ponds and seeps in three separate regions of the eastern US: southern New Jersey, the eastern border between the Carolinas, and an area of west Florida and southwest Alabama. In the latter two areas, it is found almost entirely in small, shallow seepages. If we want to conclude that the species is a specialist in such habitats, what must we do? Is it enough to say that we find them only in such habitats? Ought we to survey, quantitatively, all kinds of habitats in which frogs occur so that we can be sure where H. andersoni is not? If we wish to understand why H. andersoni is found only in acidic seeps, should we transplant them to different habitats, even at the risk of causing high mortality in those habitats? Does our answer change if we remember that this species is endangered? This is an even more contrived example than the ones in the text of the paper, but it illustrates the tension between relying on natural history and using natural history as our foundation for scientific investigation.


Natural history is the careful observation of nature, wherever nature is. Ultimately, it is what ecological, evolutionary, and behavioral science are supposed to explain. It is difficult to use natural history alone to test hypotheses in these fields because of the complex paths between process and pattern. Few patterns are predicted by one and only one hypothesis, so experiments are almost always necessary. However, the robustness of experimental results depends on how well experimental conditions reflect the integration of natural history. Natural history also plays a vital role in how well we can apply Krogh’s principle to our work. Krogh’s principle is that scientists begin with an important hypothesis and find a system (organism, habitat, species interaction) with which to test it. However, natural history is essential for knowing if the question applies to the system or if we are forcing the question on the system. There is value in beginning one’s research not by identifying an interesting question and searching for the right system but with identifying an interesting system in which to ask the right question. This approach carries the danger of parochialism, which can only be avoided by having a command of theory as well as natural history. A command of both areas allows nature to tell us which question to ask instead of demanding that nature answer the question we find most interesting.