Fisheries science is a field whose very foundation (“counting the fish in the ocean”) creates doubt in many anglers' minds. Using smartphones to have recreational anglers upload their fishing information creates doubts in just about everyone’s minds—fisheries scientists included. However, that has not stopped a few groups from steaming forward under the belief that something created by and for anglers will cause them to report honestly and faithfully. The most extensive program to date is the Snook and Gamefish Foundation’s (SGF) iAngler app, the flagship app under its Angler Action Program1. Originally started as a way to provide state scientists with more data on snook fishing in Florida, it has expanded to include fresh- and saltwater fish across the country, inevitably turning some heads around the fisheries community.
Stock assessment scientists at the Florida Fish and Wildlife Conservation Commission’s Fish and Wildlife Research Institute (FWRI) were interested in getting as much data to help with snook assessments, but were also concerned about the reliability of this information. Likewise, on the federal level, fisheries scientists with the National Oceanic and Atmospheric Administration (NOAA) were skeptical about the validity of data that is self-reported in a non-random manner2. To provide the best chance of getting information that is representative of the whole angling population, there should be a fully randomized sample of anglers. This is not what an app like iAngler does; rather, it is utilized by whoever is interested in downloading it. What if only the most talented anglers use it (the anglers most fisheries and social scientists would expect to use such an app)? Then, the experts are left thinking all the fishers out there have such success when they drop their lines in the water. For reasons like this, an analysis of these volunteer fishing apps is necessary to begin solidifying or revising our assumptions.
NOAA’s Marine Recreational Information Program (MRIP) survey is a randomized, rigorously designed sampling initiative that has interviewers intercepting anglers at boat ramps and beaches for catch-per-unit-effort (CPUE) information and calling them on the phone for effort data. Because, as Professor John Shepherd once said, “Managing fisheries is hard: it’s like managing a forest, in which the trees are invisible and keep moving around,” we have no way of knowing the real values of the variety of fisheries metrics. However, something like the MRIP provides data about as close to the “truth” any any other program. So when we sought to gauge the validity of data from the iAngler app, we decided the best path would be to compare its information to that of the MRIP.
For specific comparisons, we chose the “Three Wise Men” of fisheries metrics (or “Three Stooges,” depending on your perception of fisheries): effort, catch, and catch-per-unit-effort (or catch rate). The results that followed were in some ways expected, but surprising in other ways. First, the only place that had a reasonable number of trips reported under iAngler was south Florida, the Atlantic side especially (where the app was created). Because of this, a lot of the fishing that goes on in other parts of Florida is not being captured by the app, so using this on a statewide scale would be risky. Also, the scale between the two programs was not comparable; the number of MRIP boat-ramp interviews dwarfed the number of iAngler reported trips. This app only began in 2012 and has been spread only by word-of-mouth, so that likely explains its relative size compared to NOAA’s 35-year-old nationwide sampling program. Also, the focus of the anglers using iAngler was directed toward Florida’s popular inshore species: common snook, spotted seatrout, and red drum. Even though Floridians as a whole also like to fish offshore for snappers, groupers, billfish, etc., the app is adequately capturing only these three species.
While this seems to be two strikes against the citizen-driven app, there was one big question left: how do the catch rates compare? The spatial bias of southeast Florida might not persist if anglers in other areas start using the app. And even though it only has sufficient information for a handful of species, scientists assess stocks individually anyway. We moved forward by looking at catch rates for the three inshore fish, but narrowed our focus to iAngler’s “hotspots,” in other words, south Florida specifically. This allowed the comparison to the MRIP’s catch rates to be more representative than a statewide comparison. When we added this specification, the iAngler catch rates were very similar to those of the MRIP—for each of the three fish we considered. This is surprising from a statistical standpoint, given the fact that these anglers were not randomly chosen to participate—it was voluntary, and thus, non-random.
To sum it up, SGF’s iAngler app provides recreational fisheries information that is spatially biased toward south Florida and contains mostly information on snook, seatrout, and red drum. However, when appropriate comparisons are made, the catch rates given by anglers are very similar to those estimated by the MRIP survey. If the participation were to increase and become more balanced throughout the state, a program like iAngler could provide valuable data to fisheries scientists, especially for relatively rare and perhaps poorly sampled fisheries like snook. It even has some advantages over traditional survey methods like the one utilized by the MRIP. Because boat-ramp interviews take place after a trip is completed, they miss a lot of detailed information about the fish that were thrown back—which is a lot of fish in Florida’s fisheries. Users of an app like iAngler can submit size, weight, and other information about every fish they caught, and not just the ones they brought back to land. Self-reporting programs will always carry an undeniable statistical risk, but being aware of and accounting for potential biases could give programs like iAngler a place in future recreational fisheries management.
1 1. Information about the Angler Action Program:
2 2. The link below provides a good summary of the risks of using non-random, self-reported data for fisheries science: