Note. Occurrence of Zebra Mussels in Near-shore Areas of Western Lake Erie

Transcription

1 J. Great Lakes Res. 23(1): Internat. Assoc. Great Lakes Res., 1997 Note Occurrence of Zebra Mussels in Near-shore Areas of Western Lake Erie Christine M. Custer 1 and Thomas W. Custer 2 U. S. Geological Survey Northern Prairie Science Center 1 Patuxent Environmental Science Center 2 P.O. Box 818 La Crosse, Wisconsin Abstract. We measured biomass, percent coverage, and length-frequency of zebra mussels in near-shore areas of western Lake Erie between 16 September and 10 November 1993 as part of a larger study on the ecological relationship between diving ducks and zebra mussels. Wet weight biomass of zebra mussels, determined by SCUBA diving, ranged from 0 to 3,611 g/m 2 and averaged (± 1 SE) 1,270 ± 380 g/m 2 (n = 11). Percent coverage of lake bottom by zebra mussels ranged from 0 to 70% and averaged 17 ± 4.0% (n = 27). Percent coverage of zebra mussels was relatively high in the northern portion (28 70% coverage) and in the southwestern portion (18-40%), but relatively low (< 5%) in the southeastern portion of the study area. Percent coverage by zebra mussels, determined from underwater videography, was highly correlated (r 2 = 0.96) with zebra mussel biomass. Analysis of length-frequency data indicated that there was prominent recruitment of juvenile zebra mussels at only three of eight sites. Average shell length ranged from 11 mm to 15 mm at the other five sites. The non-uniform distribution of zebra mussels, as determined from biomass and videography, may have important ramifications when assessing zebra mussel impacts on waterfowl. These data may also be used when assessing impact of zebra mussels on other aquatic organisms in the near-shore areas of western Lake Erie. INDEX WORDS: Biomass, distribution, Dreissena polymorpha, Lake Erie, underwater videography, zebra mussels. Introduction As part of an evaluation of the relationship between diving ducks and zebra mussels in the Great Lakes, we obtained information on zebra mussel biomass, percent coverage, and length-frequency distributions to assess impacts of zebra mussels on waterfowl (Custer and Custer 1996) in Lake Erie. We found that these data can be expensive and difficult to obtain. This may be why actual zebra mussel biomass has rarely been quantified in western Lake Erie (Hebert et al. 1991, Hunter and Bailey 1992, Leach 1993, Hamilton et al. 1994). Leach (1993) quantified zebra mussels on shoals in Lake Erie where the dominate substrate was bedrock, which makes up < 5% of the bottom of western Lake Erie (Bolsenga and Herdendorf 1993:79). Hamilton et al. (1994) quantified zebra mussel biomass at two sites east of Point Pelee, Ontario, Canada along the northern shore of western Lake Erie. Again, substrates at these two sites were composed of rock beds, so data from these two studies may not necessarily be indicative of other areas in western Lake Erie.

2 The objectives of this study were to obtain biomass, percent coverage, and length-frequency distributions of zebra mussels found on substrates other than bedrock in near-shore areas of western Lake Erie. In additional, we describe a method to assess percent coverage of zebra mussels using underwater videography in conjunction with actual biomass determinations. This may be an easy and cost-effective method to quantify zebra mussel biomass. Study Area Our study area in western Lake Erie extended from 20 km east of Toledo, Ohio, through the mouth of the Maumee River and north to near Detroit, Michigan (Fig. 1). The area was chosen to complement ongoing work on the ecological relationships between zebra mussels and migrating and wintering diving ducks (Custer and Custer 1996). The 11 sampling sites in the northern portion of the study area and the 16 in the southern portion were mainly in nearshore areas. Bottom deposits of the Ohio portion of western Lake Erie are composed mainly of mud (58%), bedrock (3%), and the rest (39%) a mixture of mud, sand, and gravel (Bolsenga and Herdendorf 1993). Methods Twenty-seven sites were sampled in western Lake Erie for the occurrence of zebra mussels (Fig. 1). Eleven of these sites were sampled for biomass by SCUBA divers on 17, 18, and 22 September Percent coverage of the lake bottom by zebra mussels was quantified at all

3 27 sites on 17, 18, and 22 September, 23 and 25 October, and 10 November We determined availability and size of attachment substrates at six of the biomass sampling sites. Zebra mussels were collected for biomass determination by SCUBA divers using one-third m 2 sampling frames and collecting all zebra mussels, including attachment substrates, by touch within the sampling frame. Zebra mussels in six sampling frames were collected at each site. Sampling frames were placed at 2-m intervals in a straight line beginning 3 m from the boat anchor. Zebra mussels from each sampling frame were brought to the surface and processed on shore the same or the next day by gently scraping or pulling mussels from substrates and rinsing them free of sediments. After draining, zebra mussels were weighed (total live weight [shell + soft tissue], Hebert et al. 1991, Hunter and Bailey 1992), frozen, and transported to the laboratory for length measurements. We calculated average zebra mussel biomass from six samples at all but two sites. At site #6 we had only three samples because the large quantity of zebra mussels required more storage capacity than we had available. Mechanical difficulties necessitated stopping after the fourth sample at Site #8. Percent coverage of the lake bottom by zebra mussels was obtained by two videography methods. With the first method, a SCUBA diver took underwater color video of the lake bottom just prior to beginning zebra mussel biomass collections at seven of eleven sites (Fig. 1). No video was taken at two sites due to poor underwater visibility and at two of the three sites with no zebra mussels. Approximately 20 m of the lake bottom were filmed. The second method utilized a drop video camera and was used at 20 other locations in western Lake Erie (Fig. 1). These color video pictures were taken from a boat with the video camera suspended in an aluminum frame. We set the boat anchor and rhythmically raised and lowered the video camera on and off the lake bottom as we slowly let out 30 m of rope. This resulted in pictures of the bottom per site. A plume of silt raised each time the video camera touched bottom, so lack of silt in the picture indicated that a different piece of bottom was being videoed. Video pictures were transcribed into percent coverage by replaying the video tape on a 8 x 10 cm video monitor and freeze-framing the video every 2 seconds for the video that was taken by SCUBA divers. For the drop video, we freeze-framed the video as the camera was being lowered and zebra mussels clearly became visible. This was generally cm above the bottom, but distance varied because of water clarity. Minimum picture size was approximately 150 cm 2. We then delineated each clump of zebra mussels on clear acetate. Once data were extracted from the video, percent coverage was calculated by laying the acetate over a 300-square grid and counting the number of squares that contained zebra mussels. These counts were converted to percent coverage. Type and size of substrates were quantified in the laboratory for six sites after zebra mussels were removed. We counted numbers of pebbles and pelecypods, other than zebra mussels, used as attachment sites (= foci) by zebra mussels. Type and quantity of attachment sites may influence biomass of zebra mussels because of their requirement for a hard attachment foci. Notes on other attachment substrates were made. Pebbles, which were used as attachment substrates, from sites #2, #6, #8, and #11 were measured along their longest axis. In the laboratory, we partially thawed zebra mussels and selected one or two clumps, depending on the clump size, per sample for length measurements. We measured all zebra mussels in the clump along their longest axis (Marsden 1992) with a digital caliper. Between 200 and 400 zebra mussels were measured per site. Data analysis of zebra mussel biomass and percent coverage was by ordinary least squares regression analysis. We regressed percent coverage against zebra mussel biomass from the six sites where underwater video was obtained. We also regressed number of pebbles/m 2 and average shell length of zebra mussels against zebra mussel biomass. Length of zebra mussels was analyzed with one-way analysis of variance (ANOVA) to test for differences in shell length of zebra mussels among sites (Sokal and Rohlf 1981); means were separated using Bonferroni tests.

4 Results Zebra mussels were present at eight of eleven sites sampled for biomass in western Lake Erie (Fig. 1). Maximum wet weight biomass was 3,610 g/m 2 and averaged (± 1 SE) 1,270 ± 380 g/m 2 (Table 1). In general, zebra mussels were most abundant in the northern portion of the study area near Pointe Mouillee State Game Area and at the mouth of the Maumee River. Percent coverage of the lake bottom by zebra mussels also varied by location in western Lake Erie (Fig. 1). The seven sites at and immediately south of Pointe Mouillee State Game Area and four of seven sites at the mouth of the Maumee River had high coverage of zebra mussels (18-70% coverage). Other areas generally had less than 5% coverage. Few zebra mussels occurred along the south shore of Lake Erie near Cedar Point (0.2% to 4.7% coverage, n = 4). Percent coverage east of Cedar Point increased with distance from shore (n = 3). Average coverage by zebra mussels in the west end of Lake Erie was 17 ± 4.0% (n = 27 sites). Percent coverage of zebra mussels closely predicted zebra mussel biomass (r 2 = 0.96, P = , n = 6) (Fig. 2). Percent coverage at these sites ranged from 2% to 70%. Zebra mussel biomass ranged from 229 g/m 2 to over 3,600 g/m 2.

5 Average length of zebra mussels varied by location, but did not conform to an observable geographical pattern (Table 2, Fig. 3). The average length of zebra mussels at Sites #3, #8, and #11 was significantly smaller (P < 0.001) than at other sites (Table 2, Fig 3). Sites #2 and #6 had the largest zebra mussels. There was no relationship between zebra mussel biomass at a site and average shell length of zebra mussels (P = 0.31).

6 Zebra mussels existed predominantly as clumps, which sat on, but were not attached to, the lake bottom. The foci for attachment were small pebbles and shell fragments. Fifty percent of the pebbles were < 22 mm long and 90% were < 41 mm long. Number of pebbles/m 2 that served as attachment sites for zebra mussels varied by location (Table 3). Sites #8, #9, and #11, at the mouth of the Maumee River, had few pebbles compared with sites #2, #3, and #6 farther north. Many zebra mussels at the mouth of the Maumee River were attached to other zebra mussels and not to pebbles. Densities of intact bivalve shells, other than zebra mussels, that served as attachment sites for zebra mussels were < 1 shell/m 2. Number of pebbles/m 2 accounted for 63.7% (r 2 ) of the variability in zebra mussel bio-mass (P = 0.057, n = 6, у = x). Discussion Zebra mussel biomass in soft-sediment areas of western Lake Erie was variable and averaged 1,270 g wet weight/m 2. This was higher than previously reported from western Lake Erie (Dermott et al. 1993, Leach 1993, Hamilton et al. 1994). Maximum zebra mussel wet weight biomass in this study was 3,610 g/m 2 near site #5. In other studies, maximum wet weight zebra mussel biomass was 60 g/m 2 near Point Pelee, along the northwestern shore of Lake Erie, in September 1991 (Hamilton et al. 1994) and 1,692 g/m 2 in October in northeastern Lake Erie in 1990 (converted from dry weight, Der-mott et al. 1993). In Lake St. Clair, maximum zebra mussel biomass was 1,409 g/m 2 in 1990 (Hunter and Bailey 1992). The variability in percent coverage and biomass has not previously been reported over such a large area in Lake Erie. Areas where zebra mussel biomass and percent coverage were highest, immediately south of Pointe Mouillee State Game Area and at the mouth of the Maumee River, were the two areas where waterfowl have been most numerous (С.М. Custer, unpubl. data), and where waterfowl had been collected during a previous study (Custer and Custer 1996). Tens of thousands of diving ducks migrate through and winter in these near-shore shallow waters and forage on zebra mussels (Custer and Custer 1996). For example, more than 95% of the diet of lesser scaup (Aythya affinis) in western Lake Erie was zebra mussels during the fall, winter, and spring of (Custer and Custer 1996). The large biomass of zebra mussels in these areas coupled with the relatively shallow water made these areas the most desirable for migrating and wintering diving ducks. Recruitment patterns of zebra mussels varied in the west end of Lake Erie in There was a predominance of small (< 5 mm long) zebra mussels at only three of eight sites indicating that large-scale recruitment of juveniles in 1993 was not uniform. Zebra mussels less than approximately 5 mm long were probably 1993 recruits (Griffiths et al. 1991,

7 Hunter and Bailey 1992, Dermott et al. 1993, Neumann et al. 1993), although age structure cannot always be determined simply from shell length because of differential growth rate due to differences in nutrients, currents, etc. (Griffiths et al. 1991). Peak veliger settlement in western Lake Erie is usually in late August (Garton and Haag 1993, Fraleigh et al. 1993). A poisson distribution with 90% of individuals < 6 mm indicates heavy recruitment from settlement in August of the current year (Hebert et al. 1991, Mackie 1993, Leach 1993). Dermott et al. (1993) found that newly settled mussels in northeastern Lake Erie had grown to only 3 mm by late September. It seems that there was little recruitment at site #5 (modal shell length was 15 mm and none < 4 mm), and only a small amount of recruitment at the other four sites in Length-frequency distributions have a bi-modal or more uniform distribution of shell length if there is zebra mussel recruitment each year (Griffiths et al. 1991, Hebert et al. 1991, Mackie 1991). A normal distribution around a larger mean shell length indicates a lack of recruitment in the current year (Dorgelo 1993). Variability in the length-frequency distributions of zebra mussels among sites reported in this study was similar to the variability reported from western Lake Erie by Griffiths et al. (1991). The presence of small (< 5 mm long) zebra mussels varied by site depending on whether there was recruitment in Dermott et al. (1993) found predominately small zebra mussels at their northeastern Lake Erie study site in the fall of They attributed the smaller average mussel size to heavy mortality each winter from ice scour and heavy surf. The size of zebra mussels present in western Lake Erie mirrored the size of zebra mussels consumed by diving ducks in the same area (Custer and Custer 1996). Average size of zebra mussels, at sites without juvenile recruitment (11-15 mm), was the same average size eaten by greater scaup (Aythya marila), common goldeneye (Bucephala clangula), and lesser scaup (Aythya affinis) in Lake Erie (Custer and Custer 1996). The only other diving duck, reported in that study, which ate substantial quantities of zebra mussels, was the bufflehead (Bucephala albeola). This is a much smaller duck than either the common goldeneye or lesser scaup [total length = 38 cm vs. 49 cm and 43 cm and 0.51 kg vs. 1.1 kg and 0.83 kg for the three species, respectively (Bellrose 1976)], and this may have accounted for their selection of smaller zebra mussels. A clump was the commonest form of zebra mussels because the majority of the bottom in western Lake Erie is soft without large, rocky substrates (Bolsenga and Herdendorf 1993). Attachment substrates for zebra mussels varied by location. At the mouth of the Maumee River, bottom substrates were very soft with few pebbles and rocks for attachment. Most zebra mussel clumps had zebra mussels as the attachment foci. Pebbles were more prevalent as attachment sites farther north. Few pelecypod shells, intact or fragmented, other than zebra mussels were used as attachment sites. This contrasts to southern Lake St. Clair where there were 2-8 unionid shells/m 2 available as attachment sites, but few pebbles or rocks (Hunter and Bailey 1992). Sites with more pebbles tended to have more zebra mussel biomass, but the number of pebbles/m 2 only accounted for 64% of the variation in biomass. Site #6, which had the second highest biomass, also had the most pebbles and rocks, however, sites could have substantial zebra mussel biomass (e.g., sites #8 and #9 [1,500 g/m2]) but have few pebbles (2-5 pebbles/m 2 ). Underwater videography was a useful technique to quantify percent coverage of zebra mussels in western Lake Erie without the cost associated with SCUBA diving. Because the relationship between biomass and percent coverage of lake bottom by zebra mussels was high (r 2 = 0.96), percent coverage can be converted to biomass. As a cautionary note, the regression equation between biomass and percent coverage may differ between water bodies or when substrate types differ. Also, underwater videography may not work where the thickness of the zebra mussel layers is variable, i.e., a single layer in some places and multiple layers in others. The underwater video technique has advantages over other techniques to quantify zebra mussel distribution. It eliminates the artifact associated with samplers (e.g., glass slides, scrubbers, concrete blocks) set out in the environment (Mackie 1991, Martel 1993, Yankovich and Haffner 1993). It also improves upon the assessment of the number of zebra mussels per unit area of pelecypod or rock (Leach 1993, Dermott et al. 1993, Mackie 1993),

8 which does not quantify actual densities because the density or abundance of the rocks or pelecypod is usually not known and can vary widely among sites. These zebra mussel data are the first available for soft substrates in western Lake Erie and provide a broader scale assessment in more detail of biomass, percent coverage, and lengthfrequency distribution of the zebra mussels in western Lake Erie than has previously been available. These data will be useful in assessing the impact of zebra mussels on waterfowl and on other links in the aquatic food chain in western Lake Erie. Acknowledgments We would like to thank Brett A. Solomon, Christine M. DuRussel, and H. Al Sidell for the SCUBA diving, and Rex W. Ainslie, Pointe Mouillee State Game Area and Gildo M. Tori, Crane Creek Wildlife Experiment Station for use of their facilities. We also thank Carl E. Korschgen, William B. Richardson, Don W. Schloesser, Ronald E. Kirby, and two anonymous reviewers for comments on drafts of the manuscript.