SEARCH

SEARCH BY CITATION

Keywords:

  • conservation biology;
  • conservation genetics;
  • fish;
  • fisheries management;
  • harvest, population size

Abstract

  1. Top of page
  2. Abstract
  3. Population genetics of marine fishes
  4. Significance of these observations
  5. Should we change fisheries management based on these observations?
  6. References

The world faces a global fishing crisis. Wild marine fisheries comprise nearly 15% of all animal protein in the human diet, but, according to the U.N. Food and Agriculture Organization, nearly 60% of all commercially important marine fish stocks are overexploited, recovering, or depleted (FAO 2012; Fig. 1). Some authors have suggested that the large population sizes of harvested marine fish make even collapsed populations resistant to the loss of genetic variation by genetic drift (e.g. Beverton 1990). In contrast, others have argued that the loss of alleles because of overfishing may actually be more dramatic in large populations than in small ones (Ryman et al. 1995). In this issue, Pinsky & Palumbi (2014) report that overfished populations have approximately 2% lower heterozygosity and 12% lower allelic richness than populations that are not overfished. They also performed simulations which suggest that their estimates likely underestimate the actual loss of rare alleles by a factor of three or four. This important paper shows that the harvesting of marine fish can have genetic effects that threaten the long-term sustainability of this valuable resource.

There is increasing concern over the potential genetic effects of the harvesting of marine fish (Laikre & Ryman 1996; Law 2007; Allendorf et al. 2008; Palkovacs 2011). Such genetic changes can either result from genetic drift caused by a reduction in population size or from natural selection. Harvest need not be selective to cause genetic change; for example, uniformly increasing mortality independent of phenotype will select for earlier maturation (Law 2007). Genetic changes caused by exploitation can increase extinction risks and reduce recovery rates of over-harvested populations (Walsh et al. 2006).

A few studies have shown that harvest can reduce effective population size (Ne) and cause the loss of genetic variation even in large populations of marine fishes by using a time series of samples (see citations Pinsky & Palumbi 2014). However, it is not known how general these effects might be. The important results of Pinsky & Palumbi (2014) demonstrate that loss of genetic variation has occurred in many of the world's most abundant marine fishes.

As expected, the effect of overfishing is more dramatic on the number of alleles than on heterozygosity (Ryman et al. 1995). Heterozygosity is expected to be reduced by a fraction 1/(2Ne) in each generation. This effect will be minor except in very small populations. In contrast, allelic diversity is much more sensitive to reduction in population size (Allendorf 1986). Moreover, the loss of allelic diversity will be much greater at loci with many alleles, while the expected reduction in heterozygosity is independent of the number of alleles and their frequencies.

Population genetics of marine fishes

  1. Top of page
  2. Abstract
  3. Population genetics of marine fishes
  4. Significance of these observations
  5. Should we change fisheries management based on these observations?
  6. References

Harvested marine fish populations are typically very large and highly connected (Waples 1998). A population's capacity for maintenance of average heterozygosity at selectively neutral loci is completely determined by the effective size (Ne). In contrast, the capacity for retention of neutral alleles depends on both the census (NC) and the effective (Ne) sizes, and there is no simple relationship between the number of alleles and either Ne or NC. In a population at drift-mutation equilibrium, the expected number of alleles occurring at a selectively neutral locus can be obtained for the infinite allele model by an expression containing the census population size (NC), the effective population size (Ne), and the mutation rate summed over all possible allele frequencies (Crow & Kimura 1970, equation 9.6.13).

image

Figure 1. About 400 tons of Inca scad (Trachurus murphyi) caught by a purse seiner off the coast of Peru in 1997. This is one of the overfished populations included in Pinsky & Palumbi (2014). Photograph by C.O. Rojas, National Oceanic and Atmospheric Administration, U.S. Department of Commerce.

Download figure to PowerPoint

Figure 2 depicts the number of neutral alleles expected when census size is NC = Ne, 10(Ne), and 100(Ne) with a mutation rate of 10−5. For a given Ne, increasing NC results in a larger number of alleles that can be retained, and the effect is most pronounced in large populations. Reducing a large population will result in a greater absolute and proportional loss of alleles than will reducing a smaller one (see also Ryman et al. 1995). At NC = 100Ne, for instance, a 99% reduction of NC from 10 000 000 to 100 000 is expected to result in a 98% loss of alleles (from 59.51 to 1.49 alleles), whereas a similar reduction of NC from 1 000 000 to 10 000 would yield a loss of only 84% (from 6.57 to 1.04 alleles).

image

Figure 2. Average number of alleles per locus expected at mutation-drift equilibrium for different effective (Ne) and census (NC) populations sizes under the infinite alleles model with a mutation rate of 10−5.

Download figure to PowerPoint

The simulations of Pinsky & Palumbi (2014) suggest that populations with effective sizes of around 3,000 or less are at greatest risk of erosion of variation. This is troubling because many studies have reported lower than expected values of Ne for marine fishes (Hare et al. 2011). Nevertheless, it is hard to compare these empirical estimates with the values used in Pinsky & Palumbi's simulations because of difficulties in estimating Ne in populations with overlapping generations experiencing gene flow. For example, estimates of Ne using the temporal method with gene flow are expected to systematically underestimate both local and global effective size when populations are connected by gene flow, and the bias is sometimes dramatic (Ryman et al. 2013). The problem is particularly likely to occur in marine fish populations where high levels of gene flow obscure identification of subpopulation boundaries. Similarly, the linkage disequilibrium (LD) method (Waples & Do 2010) will underestimate Ne in continuously distributed populations because isolation by distance will generate LD that will be confused with LD caused by small effective size (Neel et al. 2013). The overlapping generations life history of fish can also cause a dramatic underestimate of Ne (Waples et al. 2013). These authors found that the effective number of breeders per year (Nb) will generally be greater than Ne for long-lived species with high variability in reproductive success. Clearly, an estimated Ne of 3000 cannot be applied as an empirical threshold in fisheries management.

Significance of these observations

  1. Top of page
  2. Abstract
  3. Population genetics of marine fishes
  4. Significance of these observations
  5. Should we change fisheries management based on these observations?
  6. References

Are such small losses of genetic diversity, as reported by Pinsky & Palumbi (2014), likely to be important? An average 2% reduction of heterozygosity does not sound like much. Nevertheless, this is the amount of heterozygosity expected to be lost in a single generation if the effective population size is 25 (−1/2Ne per generation). Alternatively, this amount of loss of heterozygosity is expected with an Ne of approximately 250, if the loss is spread out over 10 generations. Such bottlenecks could have a major effect on a population's viability, but intense bottlenecks are more likely to have harmful genetic effects than the same amount of loss of heterozygosity spread out over many generations (England et al. 2003). Thus, it is important to know the time frame of loss in the harvested populations surveyed by Pinsky & Palumbi (2014) in order to understand their likely effects.

The observed 12% loss of allelic diversity could have an important effect, especially for loci having many alleles. Loci associated with adaptation often have more alleles than neutral loci because of balancing selection (e.g. the major histocompatibility complex). As expected, greater loss of allelic diversity has been observed at ‘adaptive’ loci that have many alleles than neutral loci (Sutton et al. 2011). Such loss might be more significant because of the need for future adaptation in view of climate change affecting ocean conditions (e.g. temperature, acidity, etc.).

Should we change fisheries management based on these observations?

  1. Top of page
  2. Abstract
  3. Population genetics of marine fishes
  4. Significance of these observations
  5. Should we change fisheries management based on these observations?
  6. References

The key implication of these results for management of marine fisheries is that the genetic effects of harvest cannot be avoided. Harvest inevitably reduces population size and therefore will increase the effects of genetic drift, as detected by Pinsky & Palumbi (2014). This effect is in addition to selective effects, which appropriate management can minimize, but cannot be avoided (Law 2007). We have two recommendations.

First, evolutionary impact assessments (EvoIA) should be used as a tool for the management of valuable marine fisheries (Jørgensen et al. 2007). An EvoIA involves predicting how harvest is likely to change the genetic composition of a population, and considering how such changes are likely to affect the stock's productivity (Eikeset et al. 2013).

Second, genetic monitoring (GeM) should be used to detect genetic changes induced by harvest that might affect long-term sustainability (Schwartz et al. 2007). GeM involves genetic analysis of a temporal series of samples to detect genetic changes over time. Genetic changes cannot be avoided, but we should detect them, measure their magnitude, and respond by modifying management. Genetic processes are rarely given prominence in fisheries management (Ovenden et al. 2013). Future GeM programs could increase the significance of genetic effects of harvest in the eyes of fisheries managers by screening functional genetic variation. Recent efforts to detect change in functional diversity of harvested species have focused on a candidate loci approach (Hemmer-Hansen et al. 2011; Jakóbsdottir et al. 2011). Genome-wide screening approaches (e.g. exome sequencing) hold promise to significantly increase the power to detect changes in functional genetic variation in non-model organisms (e.g. Lamichhaney et al. 2012).

Pinsky & Palumbi (2014) have detected an important effect of harvest on fish stocks. The challenge is now for fisheries managers to know when and how to respond to this effect, and for geneticists to engage with managers in order to provide appropriate guidance.

References

  1. Top of page
  2. Abstract
  3. Population genetics of marine fishes
  4. Significance of these observations
  5. Should we change fisheries management based on these observations?
  6. References

F.W.A. conceived this paper, wrote the initial version, and lead the writing process. O.B. and N.R. wrote various parts of the manuscript. O.B. made sure that we included considerations for managers in our recommendations, and N.R. contributed Fig. 2.