© Andrew Fox
I love it when a good collaboration bears fruit, and our latest paper is a good demonstration of that principle.
It all started a few years ago with an ARC Linkage Project grant we received to examine how the whaler shark fishing industry in Australia might manage its stocks better.
As I’m sure many are aware, sharks around the world aren’t doing terribly well (surprise, surprise — yet another taxon suffering at the hands of humankind). And while some populations (‘stocks’, in the dissociative parlance of the fishing industry) are doing better than others, and some countries have a better track record in managing these stocks than others, the overall outlook is grim.
One of the main reasons sharks tend to fair worse than bony fishes (teleosts) for the same fishing effort is their ‘slow’ life histories. It doesn’t take an advanced quantitative ecology degree to understand that growing slowly, breeding late, and producing few offspring is a good indication that a species can’t handle too much killing before populations start to dwindle. As is the case for most large shark species, I tend to think of them in a life-history sense as similar to large terrestrial mammals.
Now, you’d figure that a taxon with intrinsic susceptibility to fishing would have heaps of good data with which managers could monitor catches and quotas so that declines could be avoided. However, the reality is generally the inverse, with many populations having poor information regarding vital rates (e.g., survival, fertility), age structure, density feedback characteristics, and even simple estimates of abundance. Without such key information, management tends to be ad hoc and often not very effective.
© Andrew Fox
In Australia anyway, most shark species tend to be managed better than in other areas of the world, but this still doesn’t mean we have a lot better data with which to do it. So, without stock assessments (i.e., estimates of abundance), and patchy data derived mainly from the fishing industry itself, how do you manage the quotas for such sensitive species?
It’s not easy, but our latest paper spells out how we can get a little closer to setting safe(r) catch limits using modelling approaches.
Just out in ICES Journal of Marine Science, our (me, Tom Prowse, Mick Drew, Bronwyn Gillanders, Steve Donnellan, and Charlie Huveneers) paper entitled Predicting sustainable shark harvests when stock assessments are lacking examines the case of bronze whaler sharks (Carcharinus brachyurus) caught in the South Australian Marine Scalefish fishery as good example of the challenges facing shark-fisheries managers.
The bronze whaler is a temperate coastal species landed commercially around the world, but the South Australian fishery is considered small by comparison (about 60 tonnes annually). Nonetheless, there are no estimates of population size, and the fishery is managed exclusively by gear restrictions and catch-per-unit-effort ‘trigger’ points.
To help get a better handle on how fishing is (and could be) affecting this species, we developed a stochastic age-structured model based on the catch profile from the industry itself, as well as theoretical expectations of life-history parameters from allometric equations. Because we had no idea how many of these sharks are in the population, we examined relative changes in the rate of population change (r) to see if we could identify when scenarios of increasing offtake started to elicit large declines.
© Andrew Fox
I won’t go into all the details here, but we also uploaded to Github all the data and R code needed to reproduce the model.
We’re pretty happy with the outcomes, which essentially showed that increasing harvests up to about 1.25 times what they are now are likely to elicit unacceptable declines. We were also able to test the effects of various controls on minimum and maximum sizes of caught individuals, which in most cases were minimal.
Finally, we produced a global sensitivity analysis of the main parameters based on a Latin-hypercube sampling protocol we developed a few years ago. This showed the clear dominance of the age-specific survival rates driving variation in long-term (3 generations) r. In other words, without a good selection of individuals across the entire breadth of sizes/ages, predictions should be taken with a grain of salt.
Overall though, this represents an important tool in the growing toolbox that shark fishery managers have at their disposal, and we hope that more of them around the world start adopting these sorts of modelling approaches to manage sharks when (typically) relevant data are poor or unavailable.
CJA Bradshaw
