Will Fish Become Significant Predators
of Zebra Mussels in Lake Champlain?

Mary Watzin, Jamie Rowder, Beth Lancaster, Bob Acabbo, and Leman Bronson
School of Natural Resources
The University of Vermont


When divers from the Lake Champlain Maritime Museum first examined a historic “railroad drawboat” in about 25 feet of water in Bulwagga Bay, near Port Henry, New York in 1999, they noticed large numbers of crushed zebra mussel shells on its deck. Thinking these shells could be the work of predators, we took an underwater remotely operated vehicle and video camera system to the site during the summer of 2000 and witnessed predation by sheepshead and yellow perch on zebra mussels on the wreck. Based on thes observations, we began a two-year study with the following objectives:
1. to determine which fish species were consistently eating zebra mussels on the railroad drawboat in Bulwagga Bay;
2. to explore areas outside of Bulwagga Bay in order to determine how widespread predation on zebra mussels might be in Lake Champlain;
3. to quantify the percent of the diet of the major fish predators that is comprised of zebra mussels
4. to explore whether zebra mussels in Lake Champlain have unique characteristics that facilitate their use as prey by bottom-feeding fishes

To address objectives 1 and 2, we used an underwater remotely operated vehicle (ROV) with an underwater video camera system, the Benthos Minimark II. Observations were made at the Bulwagga Bay railroad drawboat site and other locations in the main section of Lake Champlain between May 2001 and September 2002. To determine how much of each fish predator’s diet might be comprised of zebra mussels, fish were collected using gill or fyke nets, or by hook and line.

Each fish collected was measured and weighed, and then opened to remove the stomach and intestines. The contents were identified to major taxon and then the mass of each prey group was determined. Percent of the diet composed of each prey type was calculated based on these masses.

In order to quantify shell thickness, we collected mussels with a benthic sled from south Burlington Bay each year in mid-to late August, beginning in 1999. For comparison, mussels were also collected from Lake Ontario in 2002. Mussels about 14-18 mm in length were selected and then dissected, separating the soft tissues from the shell. After weighing each part, a shell:body mass ratio was determined. This ratio gives an indication of shell thickness, normalized for expected differences in weight with body size. As shell thickness declines, this ratio also declines.

Figure 1. Diet composition (by dry weight) for pumpkinseeds, sheepshead, yellow perch, and rock bass collected in 2002.

Figure 2. Shell:body mass ratios for zebra mussels collected in 1999-2002 from Lake Champlain (LC) and in 2002 from Lake Ontario. The shells from Lake Champlain are significantly thinner, making them easer to crush by fish predators.


The ROV was an excellent tool for observing fish behavior without disturbance. With this tool, we documented sheepshead, yellow perch, and pumpkinseed feeding on zebra mussels. Sheepshead swam and fed as solitary individuals. A feeding fish would swim down to the bottom, tear off a clump of mussels, chew the mussels, and then spit some of the broken shells back out of its mouth. Yellow perch also fed as individuals, but each fish would remove just a single small mussel at a time, gulping it down. Because yellow perch do not have pharyngeal teeth, they cannot crush the shells.

In contrast, pumpkinseed fish seemed to feed mostly in groups. They often appeared on the site in very large schools that would hang over the bottom without a lot of movement. Groups of fish would then move down in waves to feed on the mussels, each fish seeming to follow the leader other down to feed. Pumpkinseed also chewed the mussels after ingestion and we frequently observed the fish spitting broken shells back into the water.

We discovered that the fish moved into the Bulwagga Bay site in large, generally mixed species schools. The presence of fish on the site seemed to be related, in part, to temperature. Generally, feeding behavior did not occur unless the temperature was greater than 17.7 oC. Sheepshead did not feed until the temperature was above 21 oC. Similar patterns were noted at other sites.

Our fish collections targeted those species that we observed feeding on mussels in our underwater observations. Of those fish actively foraging, we found zebra mussels in significant numbers in the guts of pumpkinseeds, sheepshead, yellow perch and rock bass. Pumpkinseeds and sheepshead generally had the highest percentages of zebra mussels in their diet (Figure 1). The percentages for these fish were especially impressive because much of the shell material was spit out during ingestion and chewing.

The mean shell:body mass ratio for zebra mussels collected in south Burlington Bay, Lake Champlain has declined significantly (p<0.0001), since 1999 (Figure 2). The shells of adult mussels in Lake Champlain in 2002 (mean shell:body mass ratio = 17.6) were quite thin, much thinner than the shells of the mussels living in Lake Ontario (mean shell:body mass ratio = 27.3) (Figure 2). Even in 1999, when zebra mussel populations in Lake Champlain were at their lowest measured abundances and shell thickness was at its greatest, the mean ratio was very different from that measured in Lake Ontario in 2002 (p<0.0001).


Fish predation on zebra mussels is regularly occurring at a number of sites throughout Lake Champlain. Sheepshead, pumpkinseed, yellow perch, and rock bass have all learned to recognize zebra mussels as food. More than 50% of the pumpkinseeds and about 40% of the sheepshead collected had zebra mussels in their guts, suggesting that these fish are commonly consuming zebra mussels as prey. About 15% of the yellow perch collected contained zebra mussels in their guts. The large numbers of pumpkinseed, sheepshead, and yellow perch in the lake suggest that these fish have the potential to significantly reduce the abundance of zebra mussels over time. We also suspect that we have not yet documented all the fish species that are preying on zebra mussels in Lake Champlain.

Because the concentration of calcium in the water of Lake Champlain is very low, the zebra mussels in Lake Champlain have very thin shells. The thin shells of the zebra mussels in Lake Champlain may make them especially vulnerable to predation, and this may explain why we are seeing more predation on zebra mussels in this lake than in other North American locations.

The relationship we found between temperature and fish presence and foraging activity clearly suggest seasonality in the use of the zebra mussel beds. If we can begin to characterize the combination of water temperature, bottom habitat structure, water currents, and other factors that influence fish distribution, this may help us to estimate the overall impacts of fish predation on zebra mussels in Lake Champlain.

Sheepshead may be responsible for the large numbers of crushed zebra mussels shells documented at the railroad drawboat.

The Benthos Minimark II

Yellow Perch eating zebra mussels at Bulwagga Bay


Funding for this project was provided by the Argosy Foundation. We would also like to thank Mr. Fred Doane and Paradigm Analytical Laboratories for providing the remotely operated vehicle used for our underwater observations. Richard Furbush, Captain of the R/V Melosira, Brian Emerson, Susan Cobb, Mark Beekey, and Ryan Salem, assisted in various aspects of the field and laboratory experiments. We also thank Art Cohn, Pierre LaRoque, and Adam Kane of the Lake Champlain Maritime Museum for their help with the project.

Back to the main page