Mind the (Detection) Gap: Tracking Fish Across Coral Reef Habitats

November 14, 2017

The location of fishes can be tracked using underwater listening stations. But is the detection of electronically tagged fish by these stations equal across habitats? In their new publication in Canadian Journal of Fisheries and Aquatic Sciences, Nick Farmer and Jerry Ault develop a statistical method to account for gaps in tag detection. Read on to learn about their multi-year study in one of Florida’s marine protected areas. 

By Nick Farmer and Jerry Ault

Located about 70 miles west of Key West, Florida, the Dry Tortugas region is comprised of seven tiny islands surrounded by some of the most spectacular and luxuriant coral reefs in the United States. Discovered by Ponce de Leon in 1513, the islands were dubbed the Dry Tortugas due to their lack of freshwater ("dry") and the large populations of sea turtles (or "tortugas" in Spanish) nesting on their beaches. 

A beautiful but isolated and unforgiving environment, the Dry Tortugas are perhaps best known as the location of Fort Jefferson, a 16-million-brick military fortress originally built to guard the entrance to the Gulf of Mexico and Mississippi River following the war of 1812. The Fort later became a Union prison that housed the infamous Dr. Samuel Mudd. Dr. Mudd treated the conspirator John Wilkes Booth following his assassination of US President Abraham Lincoln, as in "your name is mud".

Dry Tortugas National Park. Fort Jefferson, on Garden Key, was a Union prison during the US Civil War. It now houses US National Park Service research and enforcement staff (National Park Service)

Research and recovery

The Dry Tortugas are a challenging place to conduct research—the intense ocean currents can leave you hanging on the anchor line with your fins cavitating behind you. Because of their isolation, you must bring all your own supplies including ship, scientists, dive tanks, food, and fuel. Despite these logistical challenges, the region contains some of the world’s most well-studied coral reefs. Satellite drifter tracks, sea surface temperature, and bathymetric maps have revealed that the Dry Tortugas are an important source of larval reef fish that support a multi-billion dollar fishing and diving industry in southern Florida and the Florida Keys. 

Since 1978, a multi-agency fishery-independent Reef-fish Visual Census (RVC) program has monitored the more than 300 reef-fish populations of the Florida Keys and Dry Tortugas. The RVC uses relatively high-resolution benthic habitat maps generated from a combination of multi-beam and side-scan sonar, LIDAR, and aerial photography to efficiently allocate SCUBA-based assessment surveys amongst patchy and rugose hardbottom habitats. Divers use a stationary visual census method and record the number and size distribution of each reef-fish species at thousands of randomly selected sites each year, allowing us to assess the status of their populations through time. Because we record every fish species seen and their size-structured abundance, along with information on the structure, rugosity, and composition of the coral reef ecosystem, we burn a lot of bottom time on Dry Tortugas RVC dives!

Based on survey data from 1978 to 1996, we used length-based assessment methods and found that 23 of the 35 (66%) key reef-fish species were overfished in the Florida Keys. In 2001, the National Oceanographic and Atmospheric Administration (NOAA) Florida Keys National Marine Sanctuary established the 518 km2 no-take Tortugas Ecological Reserves (TER) to rebuild the depleted reef fishery resources. Between 2002 and 2016, we participated in biennial cruises to the Dry Tortugas, joining teams of up to 77 divers from 17 agencies, each diver making four RVC dives a day for 20 days. Thanks to the new formal protections and their isolation, Dry Tortugas reefs and reef fishes are some of the healthiest and most beautiful in the United States!

Reef-fish Visual Census (RVC). Dr. Jerry Ault conducting a reef fish survey in the Dry Tortugas. (Dr. Jiangang Luo)

Tracking in the Tortugas

In 2007, the National Park Service implemented a 158 km2 Research Natural Area (RNA) that closed fishing of the western half of Dry Tortugas National Park (DRTO). The DRTO RNA was designed as a shallow water complement to the existing TERs, and at the time of its implementation was the largest no-take marine reserve network in the United States. To improve our understanding of how this network might function to promote sustainable fisheries, we deployed, calibrated, and maintained a large array of VEMCO VR2 passive acoustic receivers within the reserves. The quality of fishing more than made up for the long arduous days we spent range testing the acoustic equipment. In the Dry Tortugas, you're not "fishing"—you're catching! Over the course of several multiday trips, we implanted VEMCO V-16 acoustic transmitters into the abdominal cavities of a variety of snappers, groupers, jacks, and sharks. Following release, the movements of these fishes were recorded by nearby receivers for periods ranging from months to years, and the size of their home ranges was determined.  
Tiger Shark. A juvenile female tiger shark (Galeocerdo cuvier) was opportunistically caught and tagged when a longline gangion it was trailing from a previous encounter with commercial fishing gear became entangled in our tackle. (Joe Contillo)

The wide variety of reef habitats in the Dry Tortugas region contains abundant and diverse reef-fish populations. These habitats range from massive coral pinnacles adorned with sea fans, sponges, and stony corals to spectacular reef-terraces of large diameter platy-corals on the western bank (e.g., Sherwood Forest). There are broad stretches of hardbottom covered with forests of sea fans, soft corals, and gorgonians and fields of reef rubble comprised of dead staghorn and elkhorn corals. You can also find patch reefs isolated in sand. Each of these habitats has its own unique personality and fish communities. Within each habitat the probability of detecting a tagged reef fish is a unique function of the amount of time the fish spends in the habitat and the distance of the receivers from the habitat. 

Sherwood Forest. Massive swathes of plated coral make excellent reef fish habitat. (Dr. Jiangang Luo)

Minding the gap

Our earlier calibration work, which incorporated long underwater diver swims through the acoustic array while carrying a tag and dragging a surface GPS buoy, conclusively demonstrated that tag detection probability was not uniform across habitats in the array. Simple comparison of the number of detections in each habitat relative to the proportional availability of that habitat within the animal’s home range*  would lead to erroneous conclusions.  

In our new article, Accounting for detection gaps when evaluating reef fish habitat use in an acoustic array, we account for non-uniform tag detection within our acoustic array and extend our previous work by identifying specific habitat preferences of fish. 

Having a father and several friends working for the US Department of Defense prompted Dr. Farmer to consider detection probability as a target acquisition problem. Since distance between the tag and receiver is the primary factor affecting probability of a tag being detected, accounting for the non-linear decline in tag detections with distance (relative to all receivers in the array) was a novel way to significantly improve statistical models and to confidently identify habitat preferences for several species of reef fishes. Our approach is general and could be easily extended to other locations and passive monitoring technologies. It could also be made time-dynamic by incorporating daily detection frequencies from "sentinel tags" deployed at fixed distances from receivers throughout a study. 
The observed diversity of movement patterns, space requirements, and habitat preferences in the Dry Tortugas indicated reserves covering large tracts of contiguous reef or a mosaic of other habitat types aligned with natural boundaries to movement (e.g., deep sand channels) can confer substantial protection to exploited reef fishes. Fortunately, these were important considerations in the design of the Dry Tortugas marine reserve network. As a result, the Dry Tortugas reserve network has already begun to pay large dividends with increases in the abundance, spawning biomass, and spawning aggregations of many exploited reef fishes. It should continue to greatly support the Florida Keys coral reef fisheries ecosystem while remaining a world-class fishing and diving destination for generations to come.
Living Reef. School of permit (Trachinotus falcatus) swimming past a reef in the Dry Tortugas. (Dr. Jiangang Luo)

The article entitled "Accounting for detection gaps when evaluating reef fish habitat use in an acoustic array" is available on the NRC Research Press website.

*If you detected a fish once in habitat A and once in habitat B but there was only one cell of habitat A and 1000 cells of habitat B, you'd probably conclude that habitat A is more important to the fish. However, if habitat A is also right next to a receiver and all of the 1000 cells of habitat B are far from receivers, the probability of detecting a fish may not be equal and making conclusions about the importance of a particular habitat is not a straightforward computation. 

Header: A red grouper (Epinephelus morio) swims through a field of gorgonians. (Dr. Jiangang Luo)

Filed Under: Science News CJFAS

Post a Comment


Science Borealis: Canada's science blogging network