Risk assessment of cartilaginous fish populations

ICES Journal of Marine Science ICES Journal of Marine Science (2015), 72(3), 1057– 1068. doi:10.1093/icesjms/fsu157 Contribution to the Themed Section: ‘Risk Assessment’ Original Article Risk assessment of cartilaginous fish populations 1 Southeast Fisheries Science Center, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, 3500 Delwood Beach Road, Panama City, FL 32408, USA 2 Northeast Fisheries Science Center, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, 166 Water Street, Woods Hole, MA 02543, USA 3 Southeast Fisheries Science Center, National Oceanic and Atmospheric Administration, National Marine Fisheries Service, 101 Pivers Island Road, Beaufort, NC 28516, USA *Corresponding author: tel: +1 850 2346541; fax: +1 850 2353559; e-mail: Cortés, E., Brooks, E. N., and Shertzer, K. W. Risk assessment of cartilaginous fish populations. – ICES Journal of Marine Science, 72: 1057 –1068. Received 16 April 2014; revised 12 August 2014; accepted 14 August 2014; advance access publication 8 September 2014. We review three broad categories of risk assessment methodology used for cartilaginous fish: productivity-susceptibility analysis (PSA), demographic methods, and quantitative stock assessments. PSA is generally a semi-quantitative approach useful as an exploratory or triage tool that can be used to prioritize research, group species with similar vulnerability or risk, and provide qualitative management advice. Demographic methods are typically used in the conservation arena and provide quantitative population metrics that are used to quantify extinction risk and identify vulnerable life stages. Stock assessments provide quantitative estimates of population status and the associated risk of exceeding biological reference points, such as maximum sustainable yield. We then describe six types of uncertainty (process, observation, model, estimation, implementation, and institutional) that affect the risk assessment process, identify which of the three risk assessment methods can accommodate each type of uncertainty, and provide examples mostly for sharks drawn from our experience in the United States. We also review the spectrum of stock assessment methods used mainly for sharks in the United States, and present a case study where multiple methods were applied to the same species (dusky shark, Carcharinus obscurus) to illustrate differing degrees of model complexity and type of uncertainty considered. Finally, we address the common and problematic case of data-poor bycatch species. Our main recommendation for future work is to use Management Strategy Evaluation or similar simulation approaches to explore the effect of different sources of uncertainty, identify the most critical data to satisfy predetermined management objectives, and develop harvest control rules for cartilaginous fish. We also propose to assess the performance of data-poor and -rich methods through stepwise model construction. Keywords: chondrichthyans, demography, risk assessment, stock assessment, uncertainty. Introduction The field of risk assessment of chondrichthyan (sharks, skates, rays, and chimaeras) populations has lagged behind that of other vertebrate groups. This is due in large part to their comparatively low economic value, and as a consequence, their lack of basic life-history and fishery information. However, there is growing interest in this group, particularly sharks, sparked by the recent realization that many species have undergone substantial population declines (Stevens et al., 2000; Baum et al., 2003; Burgess et al., 2005; Myers et al., 2007; Dulvy et al., 2008; Dulvy and Forrest, 2010; Cortés et al., 2012). As a result, risk assessment of chondrichthyan populations, and the research to support it, is now drawing increased attention and resources. The approaches used to assess the risk of various stressors, notably fishing, on chondrichthyan populations have been heavily influenced by both the quantity and quality of available data. This process takes different forms depending on the discipline and the questions being asked. In a conservation context, the objective is typically the avoidance of large population declines or extinction, whereas in fisheries the goal is to maintain a healthy population while allowing for its sustainable, long-term exploitation. In both cases, a common objective is estimating current status and projecting Published by Oxford University Press on behalf of International Council for the Exploration of the Sea 2014. This work is written by (a) US Government employee(s) and is in the public domain in the US. Downloaded from http://icesjms.oxfordjournals.org/ at Adams State University on September 12, 2016 Enric Cortés1 *, Elizabeth N. Brooks 2, and Kyle W. Shertzer 3 1058 Figure 1. Continuum of risk assessment methods and the types of management products they generate. Although the figure presents the methods as a linear continuum, we recognize that there is overlap between the risk analysis categories. exploring the effect of different sources of uncertainty, and make recommendations for future work. Risk assessment methods Productivity and susceptibility analysis Data-poor situations are generally the norm when assessing risk of chondrichthyan populations. This group of fish is often taken as bycatch in many fisheries around the world and their biology is poorly understood. This situation gave rise to the use of productivity and susceptibility analysis (PSA, also known as ecological risk assessment or ERA), an approach initially designed to provide management advice when faced with cursory exploitation and biological information for a suite of species caught as bycatch (e.g. Stobutzki et al., 2001). This approach ranges from purely qualitative to quantitative, and is designed to provide management advice by assessing the vulnerability to fishing of a species or population. Vulnerability is expressed as a function of productivity, or capacity of the stock to recover after it has been depleted, and of susceptibility, or propensity to be captured by fishing practices and not survive the interaction. In its most widely used application, PSA is a semi-quantitative approach wherein the productivity and susceptibility components are defined by several attributes that are scored based on a predetermined numerical scale. The attribute scores are then averaged for each component and displayed graphically on an x–y (PSA) plot (Figure 2). Although not generally done, the range or a measure of variability of the attribute scores from different experts can also be displayed to convey “inter-expert” uncertainty. From this, vulnerability can be computed, for example, as the Euclidean distance from the origin to the coordinates of the productivity and susceptibility scores on the PSA plot. Examination of these plots provides a quick, practical tool to assess the potential or risk of a stock to b (...truncated)