Evolutionary Biology Blogι**=7/3ψ
by Lars Witting

The pitfall of evolutionary interpretation

Brood size manipulation shows that individuals with the natural number of offspring (solid dots) are more fit than individuals with a manipulated number (open dots); but it provides no evidence for the traditional interpretation that it is the trade-off between current reproduction and future survival that selects the optimal number of offspring. Great tit data from Tinbergen and Daan (1990), and kestrel data from Daan et al. (1990).

Observations of fitness differences in populations in connections with evolutionary experiments are essential for a detailed understanding of selection in natural populations; yet, these observations have misguided evolutionary biologists for decades.

The best example is maybe brood size manipulation that was first introduced by Lack in 1947. By enhancing the number of eggs in some bird nests, and reducing them in others, these experiments show that it is typically the natural number of eggs that maximises the number of offspring that survives to the following generation. A smaller number of eggs increases the chance that each offspring will survive; but the total number of survivors will decline. And a larger number results in fewer survivors because the survival of each offspring, and/or the survival of the adult to the next breeding season, is declining too fast with more offspring to feed.

This is the perfect experiment that shows that the life history and its trade-offs are optimised by natural selection to a physiological optimum that maximises the fitness of the organism. And this is where the interpretation of the results should stop. But they are often interpreted further to mean that the optimal number of offspring is selected by the trade-off between current reproduction and future survival (see e.g., Stearns, 1992). But it may just as well be the trade-off and offspring number that is adjusted by selection for an optimal number of surviving offspring; and it is in fact straight forward to show theoretically that this is much more likely to be true (Witting, 1997, 2008).

The point to be made here, however, is not an argument for my alternative interpretation. It is instead that an observed fitness-related covariance between two (or more) traits does not imply a specific selection causality. It may be one trait that is selected by the fitness differences that is imposed by the other, or the other way around, or both traits that are adjusted by a joint selection that is imposed by a third unobserved component.

This raises one of the most essential problems in evolutionary biology. Basically all of our evolutionary theory is based on contingent models that use some of the traits in an observed covariance to generate a selection that “explains” the other co-varying traits. But how can we then be sure that our interpretations of evolutionary causality are correct? The point is that we cannot; and there is a high risk that basically all of the classical studies of evolutionary causality has fallen in the pitfall of flawed evolutionary interpretation.


  • Daan, S., C.Dikstra and J.M. Tinbergen 1990. Family planning in the kestrel ( Falco tinnunculus): The ultimate control of covariation of laying data and clutch size. Behavior 114:83--116.
  • Lack, D. 1947. The significance of clutch size. Ibis 89:302--352.
  • Stearns, S.C. 1992. The evolution of life histories. Oxford University Press, Oxford.
  • Tinbergen, J.M., and S.Daan 1990. Family planning in the great tit ( Parus major): Optimal clutch size as integration of parent and offspring fitness. Behavior 114:161--190.
  • Witting, L. 1997. A general theory of evolution. By means of selection by density dependent competitive interactions. Peregrine Publisher, Århus, 330 pp, URL http://mrLife.org.
  • Witting, L. 2008. Inevitable evolution: back to The Origin and beyond the 20th Century paradigm of contingent evolution by historical natural selection. Biological Reviews 83:259--294.