Fishing of natural populations increases the variability of fish abundance. A unique data set from the southern California Current has allowed an evaluation of three hypotheses for why that should be so.
Understanding variation in the abundances of plants and animals has long been a central topic in population ecology1. It is one of particular significance when it comes to exploited populations, such as those of many fish stocks, and is revisited by Anderson et al. on page 835 of this issue2.
Extensive fluctuations of harvested fish stocks are clearly undesirable economically — too much uncertainty in expected income will adversely affect fishing communities. Such fluctuations are also harmful from a conservation perspective, as high variability may increase the (local) probability of extinctions. It has long been suggested3 by fisheries ecologists that fishing might itself increase the temporal variability of exploited populations. But long-term data on unexploited populations are needed for comparative (control) purposes, and the lack of such data has made it difficult to separate the effects of fishing from the effects of variations in environmental conditions.
In 2006, such a long-term data set — the California Cooperative Oceanic Fisheries Investigations (CalCOFI) record of larval fish abundance — was subject to comparative analysis by Hsieh et al.4, from which they concluded that, as their title put it, 'Fishing elevates variability in the abundance of exploited species'. But this study, pioneering as it was, did not look into the nature of the underlying mechanisms. Anderson et al.2, a group that includes many of the same authors, now report an extended analysis of the 50-year CalCOFI data set under the title 'Why fishing magnifies fluctuations in fish abundance'. In doing so they provide valuable, empirically based insights into the fluctuations of exploited populations. Their analysis convincingly shows that the observed increased variation of harvested fish stocks is caused by the selective removal of the larger (and older) individuals, leading to a decreasing average size and age of the fish that destabilizes the population dynamics.
Anderson et al.2 looked at three hypotheses. They found no support for the first one, that the observed variability of exploited fish stocks is a direct effect of variable fishing intensity. Then there is the selective removal of larger and so older individuals by fishing, known as the age-truncation effect. This 'juvenescence' of the population can affect the dynamics in two ways, leading to the second and third hypotheses.
The second hypothesis is that younger, smaller individuals may be more susceptible to environmental change than older, larger individuals that, as the authors put it, can survive hard times better. The third is that the changed age structure may affect demographic parameters, such as the age of maturation, and intensify the nonlinear nature of the processes involved in population dynamics (for example, through an increased population growth rate, or by increasing the nonlinear coupling of demographic parameters to environmental changes). Anderson and co-workers find some support for the former effect and strong support for the latter — that is, the observed increase in population variability seems to be caused by an increased nonlinearity of the underlying population dynamics of exploited fish stocks as compared with unexploited populations.
The higher mortality experienced by older and bigger fish, directly caused by size-selective harvesting, can induce earlier maturation of fish within the stock, and can do so in two ways that are not mutually exclusive. One is 'phenotypic plasticity', the ability of the fish to change its characteristics, or phenotype, in response to changes in its environment. This is a reversible response that is primarily an ecological effect5. The other is a potentially irreversible evolutionary response due to harvesting5,6,7,8. Many recent studies have provided evidence for this second effect, and show that the ecological–evolutionary consequences of harvesting can occur at a much faster rate than previously thought. Through the evolutionary effect, for which there is growing support7,9, age truncation can make the abundance of exploited fish stocks permanently more variable (Fig. 1).
Taken together with Anderson and colleagues' findings2, the implication is that fisheries management needs to give priority to precautionary measures. Juvenescence may be irreversible9. When the ecological effects of fishing a particular population are observed, the evolutionary consequences may have already set in, and may be irreversible, or at least only slowly reversible, depending on whether sufficient genetic variability remains in the stocks10. In other words, it will often be quicker to create a demographic change in a fish stock than to reverse that change and the increased variability in population abundance that stems from it.
The combined ecological and evolutionary juvenescence of exploited populations prompts various thoughts. Ecologists need to give more consideration to the ecological effects of life-history changes such as earlier maturation, and evolutionary biologists need to take more account of the ecological effects of evolutionary changes. Indeed, we would like to see many more combined ecological and evolutionary studies on the effect of exploitation. Such a research agenda is not easy to implement, given the scarcity of long-term data on unexploited fish stocks, and we require more data sets like the CalCOFI record. Otherwise, investigation of the heritability of the demographic parameters that Anderson et al. conclude amplify the nonlinear population behaviour would help us understand the evolutionary effect of the observed changes — and also help in assessing to what degree such an effect could be reversed through management policies.
May, R. Theoretical Ecology: Principles and Applications (Blackwell, Oxford, 1976).
Anderson, C. N. K. et al. Nature 452, 835–839 (2008).
Beddington, J. R. & May, R. M. Science 197, 463–465 (1977).
Hsieh, C. et al. Nature 443, 859–862 (2006).
Ernande, B., Dieckmann, U. & Heino, M. Proc. R. Soc. Lond. B 271, 415–423 (2003).
Olsen, E. M. et al. Nature 428, 932–935 (2004).
Jørgensen, C. et al. Science 378, 1247–1248 (2007).
Hilborn, R. Fisheries 31, 554–555 (2006).
Law, R. & Grey, D. R. Evol. Ecol. 3, 343–359 (1989).
Futuyma, D. J. Evolutionary Biology 3rd edn (Sinauer, Sunderland, MA, 1998).
About this article
Geographic Variation in the Reproductive Ecology of the Panamic Grunt in the Southeastern Gulf of California
Marine and Coastal Fisheries (2018)
Forecasting population dynamics of the black Amur bream (Megalobrama terminalis) in a large subtropical river using a univariate approach
Annales de Limnologie - International Journal of Limnology (2017)
Trends in Ecology & Evolution (2016)
Ecology and Evolution (2016)
Data-Limited Population-Status Evaluation of Two Coastal Fishes in Southern Angola Using Recreational Catch Length-Frequency Data
PLOS ONE (2016)