Counting fishes

Underwater video cameras (UVC) provide a non-lethal technique to sample fish in dense submersed aquatic vegetation. Fish often inhabit densely vegetated areas, but deficiencies of most sampling gears bias relative abundance estimates that inform fisheries management. 

Cameras have been used to document fish behavior -- including eating and breeding but never to count fish in e.g. underwater plant habitats. Australian researchers studying fish ecology have used cameras to count fish in the relatively clear waters at the Great Barrier Reef, but no research has peered through a lens to detect fish in thick vegetation.

However, it would be great if one could accurately count freshwater fish, even in the thickest of underwater vegetation. This is very important as it is commonly assumed that dense and invasive plants, e.g. hydrilla (Hydrilla verticillata), can drastically change fish habitat quality, primarily through changes in dissolved oxygen levels, water chemistry and habitat structure. Whether these changes are good or bad for fish has previously remained uncertain due to sampling problems in dense plant habitats. 

This is a big problem, especially with hydrilla, a plant that has invaded lakes throughout Florida, much of the U.S., Central America, South Africa and Australia. Florida likely spent up to $14 million per year throughout the 2000s to manage hydrilla, while the U.S. spent about $100 million per year in the 2000s for aquatic plant management. Another invasive aquatic plant with similar effects is Eurasian Watermilfoil (Myriophyllum spicatum).

Researchers from the University of Florida Institute of Food and Agricultural Sciences used underwater cameras to show that fish can be properly counted with this technology. Furthermore, they were able to show that fish actually do us habitats that were previously thought to be too stressful for fish habitat.

They lowered a camera into the water from a boat in three experimental ponds in Gainesville. The colleagues were able to count fish captured on video even those hidden in the nooks and crannies of hydrilla and other vegetation. Fish were counted during 13 weeks in the summers of 2011 and 2012, and then the ponds were drained to obtain actual fish densities.

In practical terms, researchers and conservation managers could use this technolgy to better understand how fish use invasive aquatic plants in general. Such approaches can be quite valuable in advising conservation plans and can help resolve stakeholder issues associated with these invasive plants.

This ability to use video cameras to estimate fish abundance is a tremendous asset to fisheries management, allowing us to evaluate fish habitat use in areas where previously no sampling method was effective,