Observation and experiment in Ecology

The role of the scientific method may be an under appreciated aspect in any ecological investigation. Like astronomy, ecological studies are typically observational. Is the any need for the whole discipline of hypothesis, prediction and test here? I recall discussing the point with my colleague Alec Panchen over tea in the zoology department one day. At the time my experience was largely with observational ecology (A. J. McLachlan, 1974a, 1974b), and I maintained that in the context of pure observation, the full scientific method was not helpful. Alec was of a different opinion (Panchen, 1992). Now, many years later, I see that we were both partly right. The purpose of the first observations in a novel situation is to create hypotheses for testing (see also (Gould, 2014)). Thereafter such hypotheses can be tested, either by further observation or, depending on the nature of the hypotheses, by experiment. For example, research on the newly created Lake Kariba in tropic Africa, apart for knowledge of general ecological principles, required no specific hypotheses. Newly flooded soil would predictably be colonised by aquatic animals and plants.  Samples of mud led to discovery that larvae of a well know invader, the chironomid midge Chironomus transvaalensis, appeared in large numbers on freshly flooded soil. This observation leads to the testable hypothesis that flooding caused condition favouring C. transvaalensis. I tested the association between flooding and C. transvaalensis by repeating the same observations every year during the flood season. Such a string of observations is all rather obvious and does not require the careful thought involved in designing an experiment. It is the type of procedure that is carried out by scientists all the time without reference to a formal method.

 But to give an example of a simple experiment in the same situation; we wished to test the hypothesis that the presence of the dung dropped by grazing herbivores, mainly elephants (Fig.1.), promoted the release of nutrients and detritus during flooding (A. J. McLachlan, 1974a; A. J. McLachlan and Ladle, 2009).

Fig.1. An old photograph of the advancing shoreline during flooding on Lake Kariba with grazing elephants in 1965.

The test devised by Sandra McLachlan involved immersing dung or grass in water outside our lab. Implicit in her experimental design was a Control - grass only and a Treatment Control of water only. Under these carefully contrived conditions she predicted that nutrient would be released faster and in greater quantities from dung than from a grass alone or from water alone (Fig.2.). The result would be the creation a nutrient rich habitat for invaders in the wild.

Fig. 2. An experiment to determine the relative role of grass and elephant dung in the release of nutrients during flooding on Lake Kariba in 1965. Dung or grass have been immersed in lake water and allowed to soak for a week. The resulting colour is roughly proportional to the quantity of major inorganic and organic nutrients released. The experiment is duplicated with dung in three plastic bags on the left and grass in the three on the right in both duplicates. From (S. M. McLachlan, 1971).

Establishing a link between herbivores and insect larvae in this way is of interest because it illustrates an ecological principle - the interaction between terrestrial and aquatic ecosystems during development of a newly created lake.

A note added in April 2019

I have been rereading this essay and realise that I was really we could have done better. There is a better way. The framework of adaptive reasoning (A. J. McLachlan and R. Ladle, 2011), provides testable predictions to focus research effort. The regime of annual flooding on Lake Kariba mimics the environment experienced by organisms in the Zambezi Valley for million of years. Organisms are thus predictably pre-adapted to flood conditions. Therefore, invasions of animals to take advantage of flood conditions were predictable and observations could have been tailored to selectively record these effects. By adopting evolutionary thinking like this, it would have been possible to focus attention within a more rigorous and intellectually rewarding framework.

references

Gould, S. J. (2014). Stephen, J. Gould on the scientific method., http://www.atholmclachlan.blogspot.com/.

McLachlan, A. J. (1974a). Development of Some Lake Ecosystems in Tropical Africa, with Special Reference to the Invertebrates. Biological Reviews, 49, 365-397.

McLachlan, A. J. (1974b). Recovery of the Mud Substrate and its Associated Fauna Following a Dry Phase in a Tropical Lake. . Limnology and Oceanography, 10, 74 - 83.

McLachlan, A. J., and Ladle, R. (2009). The evolutionary ecology of detritus feeding in the larvae of freshwater Diptera. Biological Reviews, 84, 133-141.
McLachlan, A. J. and Ladle, R. (2011).  Barriers to Adaptive Reasoning in Community  Ecology. Biological Reviews. 86, 543-548. 

McLachlan, S. M. (1971). The rate of nutrient release from grass and dung following immersion in lake water. Hydrobiologia, 37, 521-530.

Panchen, A. L. (1992). Classification Evolution and the Nature of Biology. Cambridge UK: Cambridge University Press.