Catching the Uncatchable: Use of Respondent-Driven Sampling for Surveying Specialised Recreational Fisheries

Recreational fishing is a popular sport and social activity undertaken by an estimated 11.5% of the global population, who catch an estimated 47 billion fish annually. Increasing availability, affordability and improvements of boats, searching technologies (e.g. GPS, sonar) and fishing tackle (e.g. electric reels), have resulted in increased efficiency and diversification of the recreational fishing sector. Specialised recreational fisheries have developed worldwide for many species that have traditionally only been caught commercially, such as billfish, tunas and deep-water demersal species. In some cases, this has led to conflict between commercial and recreational sectors. This is primarily due to recreational fishers claiming a greater share of shared stocks based on their perception that their activities generate superior economic and social benefits to the community than commercial fisheries. As a result, there is a need for fisheries scientists to obtain robust estimates of the recreational catch for inclusion in stock assessments to ensure resources are sustainable and shared equitably among stakeholders.

However, obtaining representative catch and effort data is problematic for specialised recreational fisheries that typically lack a complete list frame of participants (e.g. a fishing licence frame that contains no exemptions). Traditional probability-based sampling methods (e.g. creel or telephone surveys) are inadequate far too expensive for obtaining representative data from small ‘hard-to-reach’ components within recreational fisheries (e.g. gamefish anglers) that probably account for the majority of the total recreational catch for some species such as billfish, tunas and sharks.

Researchers in epidemiology and social sciences, routinely survey rare, 'hidden' or ‘hard-to-reach’ populations within the general community (e.g. HIV carriers, sex workers, illicit drug users) by penetration of social networks rather than by interception of individuals. This paper introduces the idea of using Respondent-Driven Sampling (RDS), a form of chain-referral (or “snowball”) sampling, as a cost-effective means of obtaining representative catch rate estimates from elusive specialised recreational fisheries that lack a complete sampling frame using retrospective (e.g. recall) or prospective (e.g. angler diary) surveys. Unfortunately, the total recreational catch cannot be reliably estimated using a single RDS survey since the total population size of fishers is unknown, which is required to expand catch rate estimates. However, by undertaking a capture-recapture survey within multiple RDS surveys, we demonstrate how the population size may be estimated using heterogeneous mark-recapture models, and thus, allowing the total harvest of particular species to be estimated for specialised recreational fisheries.