Fishing is one of the most ubiquitous forms of exploitation. It can have many indirect effects beyond reducing the abundance of target species (Jennings, Kaiser & Reynolds 2001). For example, it can deplete populations of species that are incidentally caught in fishing gear (e.g. Lewison et al. 2004). It can trigger trophic cascades – alternating increases and decreases in the abundance of sequentially lower trophic levels, which can sometimes reverberate all the way to primary producers and benthic composition (e.g. Dulvy, Freckleton & Polunin 2004; Frank et al. 2005). It can also exert a strong selective pressure favouring shifts in life history and behavioural traits that ultimately result in organisms growing faster (or slower), maturing earlier (or later), and becoming less active and shier than in unharvested populations (Conover & Munch 2002; Biro & Post 2008; Sharpe & Hendry 2009). A key question is whether such fisheries-induced shifts are temporary phenotypic responses or genetic changes that might take a long time to reverse if fishing is curtailed (Conover, Munch & Arnott 2009).
The indirect impacts of fishing can become particularly complicated when organisms have unusual life histories. This is the case of pandalid shrimps, which are the target of trawl and trap fisheries in temperate and subarctic regions of the world and the subject of the study by Chiba et al. (2013) in this issue. Pandalid shrimps, like many species of annelids, echinoderms, molluscs and fishes, are sequential hermaphrodites (Polikansky 1982). They start life as males, and once they reach a certain size or under favourable social conditions, they reorganise their internal plumbing and turn into females. As a result, in natural populations, most of the small individuals are male, and the larger ones are female. Using a combination of time series of population size and sex ratio data from a fished population and simple captive-rearing experiments, Chiba et al. show unambiguously that the timing of sex change in the Hokkai shrimp Pandalus latirostris varies plastically in response to population age and sex structure, but the authors go one novel step further. They demonstrate that the selective nature of fishing turns normally adaptive sex ratio adjustments into a maladaptive response.
The size or age at which a hermaphrodite changes sex is often determined by features of its social environment, such as the sex ratio of the social group, the size of the individual relative to others in the group and local density (reviewed in Munday, Buston & Warner 2006). For example, it can be advantageous for an individual of a certain size to change sex if there appears to be good breeding opportunities for individuals of that size, but of the opposite gender (e.g. Warner & Swearer 1991; Buston 2003). Chiba et al. confirmed the importance of social cues for sex-change decisions in Hokkai shrimp by showing that, in the laboratory, small shrimp rarely change sex when kept with large shrimp, but they readily do so when large conspecifics were absent. A 13-year-long time series of information on sex ratio from a natural population in a Hokkaido lagoon revealed that wild Hokkai shrimp do indeed fine-tune their sex ratio in the late spring in relation the relative abundance of mature males and females. Thus, some small males might wait an extra year to change sex if there are already too many mature females around, or large ones might change sooner than usual if mature females are rare. This is remarkable because Hokkai shrimp do not breed in the small mating groups typical of species that exhibit socially mediated sex change (Munday, Buston & Warner 2006). How the shrimp assess local operational sex ratio remains a mystery. At any rate and all else being equal, this facultative adjustment, which usually yields a slightly male-biased population sex ratio, would be adaptive by the time breeding starts in early autumn, but fishing makes things unequal.
In fact, fishing inflicts a double blow to Hokkai shrimp. First, the shrimp are harvested with traps that selectively retain large individuals. As a result, fishing severely biases the population sex ratio, from 53% males before the trapping season starts to 78% males by the end of it. In itself, this is not surprising. Selective exploitation, both on land and in the ocean, can skew sex ratios (e.g. ungulates, Ginsberg & Milner-Gulland 1994; ursids: McLoughlin, Taylor & Messier 2005; fishes: Hamilton et al. 2007). However, the second blow – the timing of fishing – is more unusual. Fishing occurs in the summer, after ‘decisions’ about sex change have been made but before breeding has occurred. The sex ratio that was adjusted according to the population's composition in early summer is therefore no longer appropriate for the heavily male-biased population that remains after fishing.
Chiba et al. offer a practical solution to the potential problem: shift the fishery to the spring, before shrimp are committed to remain male or change into females. This way, shrimp can change sex on the basis of the abundance of males and females that will actually be present and active on the mating ground. However, it is not completely clear that there is a problem to be solved, at least not yet. Skewed sex ratios can affect population dynamics (e.g. Mysterud, Coulson & Stenseth 2002; Rowe & Hutchings 2003), but there is yet no clear trend in the relatively short time series of catch-per-unit-effort for this fishery. Elucidating the link between sex ratio skew and population productivity is an obvious next step. This could answer the puzzling question of how the few females that escape fishing manage to sustain this population.