The dangers of data deficiency

Below is a guest post by Chris Mull, a PhD student in the Dulvy Lab at Simon Fraser University.  Chris studies shark biology and evolutionary neuroecology.  You can read more about his research here.

Over this past week science headlines have been flooded with the news that one-quarter of all chondrichthyans (sharks, skates, rays, and chimaeras) are threatened with extinction. This finding, according to a massive study by the International Union for the Conservation of Nature (IUCN) Shark Specialist Group, is alarming, and is helping to galvanize the already growing concern over the status and conservation of the world’s sharks and rays. However, lost amid the media frenzy over the headline statement is the equally disturbing finding that almost half of the world’s known chondrichthyan species are categorized as Data Deficient. This means we know so little about half of the world’s sharks and rays basic life history information, such as how quickly they grow or how often they reproduce, that it is extremely difficult to infer how their populations are faring, or how resilient they might be to the myriad of threats they face. This considerable gap in our basic understanding of sharks and rays may be the biggest unaddressed threat they face.

There are many methods to address this missing fundamental data, but unfortunately we are missing key opportunities to do so. The best life history information is collected in the field and often involves lethal sampling or the dedication of clever and intrepid researchers in fish markets and fishing fleets around the world. Fisheries and bycatch data collection is another untapped opportunity [for example, see Carr et al. 2013 – Ed]. It has been estimated that 100 million sharks are killed annually, and with no systems in place, little to no data is ever collected from these individuals.

Dulvy et al 2014 Fig 6

Fig. 1. Evolutionary uniqueness and taxonomic conservation priorities.  Threat among marine chondrichthyan families varies with life history sensitivity (maximum length) and exposure to fisheries (depth distribution). (A) Proportion of threatened data sufficient species and the richness of each taxonomic family. Colored bands indicate the significance levels of a one-tailed binomial test at p=0.05, 0.01, and 0.001. Those families with significantly greater (or lower) than expected threat levels at p<0.05 against a null expectation that extinction risk is equal across families (35.6%). (B) The most and least threatened taxonomic families. DOI: http://dx.doi.org/10.7554/eLife.00590.012

The necessity to collect vital missing data makes the shark cull recently undertaken in Western Australia particularly upsetting. Aside from the misguided policy of culling sharks to protect bathers (more can be read about the ineffectiveness of this strategy from numerous sources), the animals being killed are simply dumped overboard, despite the government’s assurances that samples would be provided to local research groups. Because these sharks are large and in some cases legally protected, the data that could be provided would be extremely difficult to collect any other way, making the cull a disastrous waste of both resources and potential knowledge.

While opportunistic sampling will take a long time to completely address the paucity of knowledge, predictive models of life history traits can be developed using what we know of life history in other vertebrate groups. In mammals and reptiles, life history traits (e.g. growth, age at maturity, offspring production) scale predictably with body size and environmental temperature, and sharks and rays may exhibit similar patterns. In fact, the novel models used by Dulvy et al. to predict the status of Data Deficient species are based on the idea that life history traits scale with body mass, and can serve as a proxy for a species’ risk of extinction. The one-quarter estimate of threatened chondrichthyans includes 68 species for which threat is estimated by body size, depth, and geographic range. With the development and testing of predictive models we can make inferences about the biological traits or extinction risk of sharks and rays that are either difficult to study, simply too sparse to collect adequate data, or immediately threatened with extinction, though this does not alter the need for direct measures to compare.

Compared with other vertebrates, sharks and rays are not only among the most threatened, but also the most poorly understood. For example, information is available for over 66% of amphibians, 85% of mammals, and 99% of birds. Our gaps in basic shark and ray knowledge are widely distributed both taxonomically and geographically. The focus of most research has been on coastal sharks and particularly the most charismatic species, notably the white shark. While the reason for this is partly logistical, as it is much easier to study a species at or near the surface, it is also due to the disproportionate amount of funding drawn by “sexy” research. For example, studying charismatic species that can be photographed, videoed or blogged about or focusing on the ecological roles, movement or behavior of a small number of species.

While our detailed understanding of threatened sharks and rays is certainly important, we have inadvertently perpetuated the idea that all sharks are apex predators that are slow growing, late to mature and have small litters. This generalization glosses over the true diversity of life history strategies, and how those vary across species and the environment. Sharks and rays differ tremendously in body size, growth rate, and reproductive output, and until we understand how these traits vary like in mammals and teleosts, we will struggle to conserve those species at risk.

While it is readily apparent that human activity is threatening many sharks, skates, rays, and chimaeras, ignoring data deficiencies could mean the number of species under threat could dramatically rise from 25% without us even noticing. It is critical that we collect the life history data to fill these fundamental gaps in regions with high biodiversity, endemism, and with high proportions of Data Deficient species. The impetus now should be on addressing the gaps in our basic understanding of shark and ray life history, while also developing predictive tools. As the old adage goes, “What you don’t know could kill you”…in this case, what we don’t know could kill what remains of our threatened shark, skate, and ray populations.

Leave a Reply