Best management practices for the control of blister worm on oyster farms

The research portion of our project was encapsulated in milestones #3 and #6 under which we tested preventative treatments for reducing or eliminating blister worm infestations on oyster farms. In the Northeast, oysters are grown using a variety of methodologies, including several that “plant” oysters directly on the bottom substrate of a lease site or in some form of caging system held near the bottom. In contrast, other methodologies hold oysters in surface cage systems. Our research, on reducing blister worm impacts, has focused on farms employing the latter type of culture system. One type of surface culture system common among oyster farms throughout the Northeast uses plastic mesh bags supported at the surface by small floats (Fig. R1). These can be flipped periodically to dry the bag surface in order to reduce the amount of fouling from tunicates, algae, mussels, barnacles and other sessile organisms. Fouling organisms can reduce water flow through the bag and many of the sessile invertebrates that are part of the fouling community compete with the oysters for suspended food. Thus, periodic flipping of the bag is necessary to dry and remove fouling organisms and sustain oyster growth.

The oysters inside the bag, however, are not dried and there is no opportunity to kill any fouling organisms that settle directly onto the oysters. Given that the life cycle of blister worms includes a dispersive larval stage, we hypothesized that periodic drying of oysters and washing of the oyster shell surface would disrupt blister worm larval settlement and reduce or eliminate infestations. We employed 4-bag Oyster-Ranch (Go-Deep International) cages in our experimental design. These cages consist of a wire mesh frame into which four individual plastic mesh bags containing oysters can be held near the surface suspended beneath two large plastic floats or pontoons. The majority of the time the oysters are submerged, but the cage system can be flipped exposing the wire frame, individual bags, and most importantly, the oysters (see Fig. R2). In comparison to oysters cultured in a typical flip bag, oysters in an Oyster-Ranch (or similar Oyster-Gro^TM) system can be air-dried and washed during periods of exposure. We used these caging systems to address the research hypotheses given, above.

Initially, we proposed to conduct this research at two oyster farms in Maine during the late spring to early fall (growing season) of 2014. Due to a delay in funding from NESARE, we obtained outside resources to start this research on one farm on the Bagaduce River during the 2014 growing season. The research was subsequently expanded to three farms in 2015, one farm on each of the Bagaduce, Weskaeg, and Damariscotta Rivers. We also continued this research for two additional seasons in 2016 and 2017 at the farm on the Bagaduce River (see Table R1).

Bagaduce River Farm (2014 to 2017) - During the growing season of 2014, we compared the frequency of new blister worm burrows in the shells of oysters cultured in “flip-bags” to those cultured in bags held within Oyster-Ranch cage systems. Using oysters provided by the host farm, we placed ~250 oysters (35-40 mm shell height) in each of four replicate flip bags like those typically employed at the farm (Control treatment). We also placed oysters (~250; 35-40 mm SH) in individual bags that were then placed inside one of the four slots of an Oyster-Ranch cage. We set up four such Oyster-Ranch cages; one cage was flipped every two weeks to provide 24 hours of air-drying (AD24) and one cage was flipped every two weeks to provide 4 hours of air-drying (AD4). The bags in one cage system were removed every other week and washed to remove built up sediment and detritus, after replacing the bags in the Oyster-Ranch cage the bags were then dried for 24 hours (AD24W). The bags the fourth cage system were removed every other week, washed, and then dried for 4 hours (AD4W). Washing the oysters can reduce the buildup of sediments and dead organic material (detritus) on the shell surface that provides material for blister worm larvae to build an initial mud tube. From this mud tube the blister worms excavate a burrow into the oyster shell by chemical means. Although our initial proposal included pressure-washing of the oysters in the latter two treatments, we did not have access to a suitable portable power-washer that could be used on the Bagaduce farm. Instead of power-washing, we “sudsed” the each bag by lifting the bag out of the water and vigorously plunging the bag back into the water up to a dozen times, as is often done by the farm operator. In addition to removing sediment from the surface of the oysters, “sudsing” has the added benefit of dulling the sharp edges of the oyster shells (leading to better oyster shape), reduces fouling, and breaks up oysters that have attached to one another (doubles).

The bags and cage systems were deployed on May 27, 2014. Starting on July 10 and continuing until November 8, 2014, we haphazardly sampled five oysters from each replicate bag in each treatment approximately every two weeks. At each sampling, individual oysters were brought to the University of Maine in Orono shucked, the meats discarded and the shells were scored for the presence of new worm burrows that contained worms (see Fig. R3). Although regular sampling reduced the number of oysters in each bag, we further reduced the number of oysters in each bag as they grew following standard practice employed at the Bagaduce River farm so that the final density on November 8 was ~125 per bag. Thinned oysters were returned to the host farm.

We employed the same basic experimental design at the Bagaduce River farm during the 2015 growing season. Notable changes to our protocol included starting with larger oysters (mean shell height 70.9 ± 0.82 mm) and the oysters provided by the farm had been exposed to blister worm the previous year. Thus, when we set up the experiment on June 11, 2015, we sampled 90 oysters and estimated the size of existing worm burrows (from edge of the oyster shell to where the burrow looped back to the shell edge – see Fig. R3). From these measurements we developed a size frequency histogram for the existing “older burrows” (Fig. R4); given this distribution we set a threshold of < 3 mm for determining whether burrows observed in subsequently sampled oysters represented a “new” versus “older” burrow. We visited the Bagaduce River farm approximately biweekly from June 24 to November 1, 2015. At each visit we employed the same strategy as we did in 2014, sampling five oysters haphazardly from each replicate bag, shucking the oyster and scoring both valves of each oyster for the presence of new burrows.

During the first two years of our work at the Bagaduce River farm we specified the dates on which the oysters were to be dried, or washed and dried in the experimental treatments. This likely reduced the effectiveness of drying, particularly on days when the weather was not suitable. For example, during mid-September in 2015 there was a period of unseasonably cool and rainy weather that was followed by a sharp increase in the number of new burrows not only in the oysters in the control treatment, but also in all four of the experimental treatments. Thus, in 2016 we requested that the farmer dry the oysters on warm, sunny days with low humidity and good airflow. In addition, because we observed only a minor reduction in blister worm settlement in the treatments that included washing during 2014 and 2015, our experimental design in 2016 focused on air-drying alone. Thus, in 2016 our experiment included oysters in eight plastic mesh bags distributed between two Oyster-Ranch cages that were dried weekly and oysters in another eight bags within two Oyster-Ranch cages that we dried every other week. We eliminated the washing step, and more importantly the cages were exposed on days that were most conducive to thorough drying of the surface of the oysters’ shells.

On May 12, 2016 we placed ~1000 oysters (30.3 ± 1.4 mm SH) into four replicate surface flip bags and ~500 oysters into each of 16 plastic mesh bags. The latter were subsequently deployed on the Bagaduce River in four separate Oyster-Ranch cages. The flip bags were turned weekly to control fouling without emersion of the oysters while the bags in the Oyster-Ranch cages were exposed, as described above. We visited the Bagaduce River farm every two weeks and haphazardly sampled five oysters from each replicate bag in each treatment. The oysters were processed and presence of new blister worm burrows (< 3 mm) was scored as in 2015.

We continued monitoring the oysters from 2016 into the 2017 growing season at the Bagaduce River farm. Oysters from the 2016 experiment were overwintered in one of two ways. The oysters in one control flip-bag were submerged at the Bagaduce River for the duration of the winter and redeployed on the surface on May 10, 2017. Oysters from the remaining three control flip-bags and all 16 of the Experimental treatment replicates were held in cold, damp storage (~38°F) for 3-4 weeks in a refrigerated trailer. Previous NESARE-funded research has shown that such treatment is generally successful at reducing the load of adult blister worms in infested oysters. Both control and experimental oysters were redeployed in their original mesh bags in their original Oyster-Ranch cage on May 10^, 2017. We haphazardly sampled five oysters from each replicate bag on this date and monthly, thereafter, until mid-October. As in previous years, we estimated the number of new blister worm burrows less than 3 mm in length among the oysters in this sample. From May 10 to August 31 we also estimated the number of live worms present in older, established burrows.

Weskeag & Damariscotta River Farms (2015 Research) - We tested the efficacy of drying and washing treatments at lease sites on two other rivers during the 2015 growing season. The depth underneath the Weskeag River farm at low tide (~1 m) precluded the use of Oyster-Ranch or Oyster-Gro style cages. At this site, we relied on the use of 30 plastic mesh flip-bags into which approximately 250 oysters (72.8 ± 0.64 mm SH), supplied by the farm owner, were placed. Five of the 30 bags constituted our control treatment; these bags were flipped biweekly to manage fouling on the bag surface, but the oysters inside were not exposed at any time during the experiment. Four other sets of 5 bags constituted our experimental treatments. One of the four sets was hauled out of the water and onto a floating dock located on the oyster farm and dried for 4 h every other week while a second set was dried on the dock for 24 h every other week. A third set of bags was hauled out onto the dock, washed with a pressure washer and dried from 24 h every other week and the last set was pressure washed and dried for 4 h. Because the oysters used in this experiment had previously been exposed to blister worm, we measured the length of >500 blister worm burrows in the shells of a preliminary set of 90 oysters collected on May 24, 2015. From these measurements we developed a size-frequency histogram, as described above for the 2015 Bagaduce River experiment. Based on this histogram we set a maximum threshold of 3 mm in length for scoring new blister worm burrows (data not shown). We revisited the farm approximately every two weeks from June 17 to Nov 5, 2015. As at the Bagaduce River farm, at each visit we haphazardly sampled 5 oysters from each replicate bag, returned the sampled oysters to the University of Maine in Orono where they were shucked and the frequency of new blister worm burrows in each valve were promptly recorded.

The lease-holder at the Damariscotta River farm provided us with a set of four Oyster-Gro cages which were each capable of holding six plastic mesh bags and also provided six regular surface flip bags. These 36 bags were stocked with roughly 500 oysters (~40-50 mm SH) on May 11^th, 2015. The experimental protocol mirrored that employed at the Bagaduce River farm in 2015; the flip cages were turned biweekly to control external fouling without drying the oysters, the oysters in one Oyster-Gro were dried for 24 h, the oysters in a second Oyster-Gro cage were dried for 4 h, the oysters in a third were both washed by “sudsing” and dried for 24 h, while those in a fourth cage were washed by “sudsing” and dried for 4 h. Each treatment was applied every other week. We visited the farm every other week from May 26 to October 9, at which time five oysters were haphazardly sampled from each replicate bag. These oysters were taken to the University of Maine’s Darling Marine Center in Walpole, ME where they were shucked and both valves of each oyster were scored for the presence of new blister worm burrows less than 3 mm in length.

Oyster Growth as a Function of Blister Worm Treatment - We explored whether the drying and washing treatments applied to control blister worm settlement affected the growth of oysters at the Bagaduce River farm and the Weskeag River farm in 2015. At the Bagaduce River farm, we individually measured the shell height, shell width and shell inflation on 90 oysters prior to marking each with colored and numbered “bee dots”. These 90 oysters were then distributed among the replicate bags belonging to each of our five experimental treatments on June 24, 2015. Similarly, at the Weskeag River farm 150 individually measured and tagged 150 oysters were deployed among the replicate bags in the five experimental treatments on June 17, 2015. If these oysters were inadvertently sampled during our biweekly sampling trips they were immediately returned to their respective bags. We recovered 71 of the oysters deployed at the Bagaduce River farm on November 11, 2015 and 92 of the tagged oysters deployed at the Weskeag River farm on November 5, 2015. Tagged oysters were missing at the final sampling due to both mortality during the experiment (and we could not be sure when they died) or because tags were no longer readable. We calculated specific growth for each recovered oysters using the formula ((ln[final shell height]-ln[initial shell height])/time elapsed) and then used a one-way ANOVA to determine whether mean specific growth rates varied among treatments at each farm.

In addition, we estimated the change in the mean size of oysters during each sampling interval at the Bagaduce River farm in 2016. Among the five oysters we collected from each replicate at each sampling, we measured individual shell height, shell width and shell inflation. From these measurements we calculated the mean shell height for the oysters from each treatment (n=20 per time point in Controls and n=40 in Experimental oysters) at each sampling. We then conducted a one-way ANOVA at each time point to determine whether there were significant differences in mean shell height over time.

Identification of Blister Worm Larvae - As part of milestone #3, we also conducted plankton tows near the Bagaduce River farm in 2014 with the intention of developing a guide for growers on the identification and monitoring of blister worm larval abundance on their farms. Our expectation was that, once trained, farmers could better time the application of preventative treatments to reduce blister worm outbreaks. Based on the literature, we anticipated that the life cycle of blister worms includes a dispersive larval stage that may last up to 45 or more days after which point they settle onto shellfish hosts and excavate a burrow in the host shell. Each time we visited the Bagaduce River Farm we took two plankton tows with an 80 μm net, one on either side of the rows of longlines constituting the central part of the oyster farm. From each tow we typically recovered 50-100 ml of concentrated plankton. After thorough mixing, several replicates of approximately 10-15 ml of concentrated plankton was preserved in 35-40 ml of 95% ethanol and returned to the University of Maine. Processing of the samples included re-sieving to remove the ethanol, resuspending the sample in distilled water, and transferring polychaete larvae into a separate 15 ml petri dish with distilled water. Each of the isolated polychaete larvae were measured using a stage micrometer on a dissecting microscope, photographed, and cataloged. We relied on the key in Blake (1969) to discriminate P. websteri larvae from those of closely related species. In addition, we isolated DNA from individual larvae and amplified a portion of the mitochondrial cytochrome oxidase I gene (mtCO1) for each using conserved primers. The PCR products were subsequently sequenced or used in a restriction fragment length assay diagnostic for P. websteri.