News & Views | Published:


Nets versus nature

Nature volume 450, pages 179180 (08 November 2007) | Download Citation

The life-histories of pike adjust quickly to shifts in the opposing forces of fishing and natural selection. Such rapid changes suggest that evolutionary dynamics must be incorporated into fisheries management.

People like to catch big fish, sometimes so much so that fish sizes overall become greatly diminished. According to one view, the continual removal of large fish from a population sets the stage for rapid, undesirable evolutionary changes, including slower growth, earlier adult maturation and permanently smaller size1,2. This occurs because removing the largest fish directly opposes natural selection, which tends to favour large size.

What happens when these two forces simultaneously oppose one another? Can evolution respond quickly enough to track changes in fishing selection, or does natural selection counteract it? Writing in Proceedings of the National Academy of Sciences3, Eric Edeline and colleagues illustrate the outcome of this dynamic tug-of-war between the forces of natural selection and fishing selection.

Until now, the theory underlying the management of fisheries has been based on ecological models that predict how the productivity of an exploited population changes in relation to its density, and the age and size at which fish are caught. The goal is to ensure a maximal but sustainable catch in perpetuity. Current management approaches do not take into account the potential for evolutionary change in response to fishing.

Why is evolution important to fisheries management? It could be argued that fishing merely adds an additional predator to the ecosystem. But from the fish's point of view, humans turn the rules of engagement completely upside down. Most natural predators attack smaller fish more frequently than larger fish. The bigger a fish gets, the lower its mortality (Fig. 1). Hence, growing fast early in life is a good strategy. Moreover, because big fish produce many more offspring than small fish, delaying maturation to larger size also increases fitness — that is, the likelihood that one's genes will be passed on to future generations. By causing greatly increased mortality at large sizes, fishing selects for fish that grow slowly and mature at small sizes. Numerous other physiological, behavioural and reproductive traits likewise evolve that can lower fitness4. Taken to its extreme, many generations of intense size-selective fishing could in theory cause the evolution of a population of runts.

Figure 1: The darwinian struggle between natural selection and fishing selection.
Figure 1

The graph depicts the contrast between mortality rates as a function of fish size in the absence and presence of mortality due to fishing. Natural rates of mortality decline dramatically with increasing size early in life, until reaching a low level for the remainder of life (purple). Fishing greatly increases the mortality of large fish (green). Arrows represent the direction of selection on body size in the absence (purple arrow) and presence of fishing (green arrow).

The introduction of darwinian principles into fisheries science has been controversial5,6. Some have argued that adequate proof of evolutionary changes caused by fishing has not been demonstrated. That would require changes in traits such as growth rate to be shown to have a genetic basis. This is extremely difficult to do in the wild because environmental and genetic influences are confounded, although new statistical methods have enabled the evolution of certain traits (such as size at maturity) to be revealed7. Lab experiments, in which environmental conditions are standardized, can demonstrate genetic change8 but have been criticized for not representing real fisheries in the wild5.

Edeline et al.3 now enter the fray. They took advantage of a unique 50-year time series of data on growth rates of pike (Esox lucius) in Lake Windermere in northwest England. This lake was fished for centuries until 1921, when the net fisheries were closed. Net fishing did not reopen until 1944. Each year from 1944 onwards, biologists tracked the age and growth of individual pike landed in the fishery by measuring the annual rings that form in certain bones, much like reading the rings in a tree trunk. They also tagged and recaptured pike, providing estimates of population size and mortality.

These highly detailed data enabled the authors to show in an earlier paper9 that fishing did indeed remove the larger, faster-growing fish whereas natural sources of mortality did the opposite. Hence, they hypothesized that the sudden resurgence of fishing in 1944 should cause an evolutionary decline in growth rate followed later by an increase as the fishery waned over the 50 years. After using statistical models to account for the effect of a suite of confounding environmental factors, the temporal trend in growth rate closely tracked the predicted pattern. Several twists and turns in growth trajectory seemed to coincide with episodes of excessively high fishing and with the large-scale death of perch (a prime food source for pike) in 1976. In addition, changes occurred in the level of reproductive investment by young females that were also as predicted from evolutionary theory. The authors conclude that evolutionary responses to the opposing forces of fishing and natural selection must be accounted for in managing fisheries.

Critics will contend that consistency with the predictions of evolution is not proof that the changes observed were in fact genetic. The responses are probably far too rapid to be entirely evolutionary as opposed to ecological in origin. With only one population under study, any interpretation of this sequence of growth changes contains an element of story-telling. Perhaps the changes in growth rate fit the predictions of evolution purely by coincidence.

Yet this is one of the most data-rich and comprehensive analyses of fishery-induced evolution ever published. Together with strong evidence also emerging from a variety of other harvested species7,10,11,12, the likelihood that all such studies are erroneous is becoming vanishingly small. Moreover, Edeline and colleagues' approach provides fresh incentive and the methodology to test for evolutionary change in the many other long-term data sets of age and growth that exist for heavily fished species.


  1. 1.

    ICES J. Mar. Sci. 57, 659–668 (2000).

  2. 2.

    , & The Exploitation of Evolving Resources (Springer, Berlin, 1993).

  3. 3.

    et al. Proc. Natl Acad. Sci. USA 104, 15799–15804 (2007).

  4. 4.

    , , & Ecol. Lett. 9, 142–148 (2006).

  5. 5.

    Fisheries 31, 554–555 (2006).

  6. 6.

    & Fisheries 32, 90–91 (2007).

  7. 7.

    et al. Nature 428, 932–935 (2004).

  8. 8.

    & Science 297, 94–96 (2002).

  9. 9.

    et al. Ecol. Lett. 10, 512–521 (2007).

  10. 10.

    & Genetica 112–113, 475–491 (2001).

  11. 11.

    , & Proc. R. Soc. Lond. B 274, 1015–1022 (2007).

  12. 12.

    et al. Nature 426, 655–658 (2003).

Download references

Author information


  1. David O. Conover is in the School of Marine and Atmospheric Sciences, Stony Brook University, Stony Brook, New York 11794-5000, USA.

    • David O. Conover


  1. Search for David O. Conover in:

About this article

Publication history



Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Newsletter Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing