Development and Implementation of Integrated Pest Management of Burrowing Shrimp on Washington State Commercial Oyster Beds

Small arena experiments

Experiments to investigate whether an injury threshold could be defined were initiated in May 2002.  Sixty pieces of oyster seed (small juveniles or spat on oyster shell cultch) were placed on each of 12 small (2m x 2m) plots established at a location near the Cedar River long term monitoring site in Willapa Bay (Figure 1).  Four plots were placed in an area with very high ghost shrimp density (> 90 burrows m2) and four in an area of moderate density (20 - 35 burrows m2).  Each plot was surrounded with short (20cm high) plastic fencing to prevent shell loss due to physical weather effects.  The number of shrimp burrows in each of four 0.25 m2 sub-plots was recorded, along with the number of oyster seed, eelgrass turions, and percent algal cover measured at periodic intervals over 4 months.

A second set of plots was established in summer 2003 on four different oyster beds with lower shrimp abundance (two at the Cedar River location, one nearby but across the channel and one near the south end of the bay near Nahcotta.  Sixty pieces of oyster seed were again enclosed within each of several small (2m x 2m) plots of differing shrimp burrow density (<10 to > 40 burrows per m2) using plastic fencing to prevent seed loss.  Results were compared with several unfenced areas which were established to determine fence effects. The number of shrimp burrows in each of four 0.25 m2 sub-plots was recorded, along with the number of oyster seed in each plot during several subsequent tidal series.

Within 28 days after oyster shell placement, over 95% of shells in plots of very high burrow density had sunk beneath the substrate surface (Figure 2A).  Oyster shell survival was higher over longer intervals in plots of moderate burrow density, but that trend was not uniform (Figure 2B).  An average of 54% of shells in plots of 40 – 50 burrows / m2 sank within 40 days.

Results of experiments conducted in areas of slightly lower shrimp density in 2003 were highly variable (Figure 3) with high densities of polychaetes confounding burrow counts.  Multiple regression analysis reflected the seemingly random affects of burrow density on oyster survival.  Although a general linear model was significant (df = 62; F = 54.894), the small positive coefficient in the resulting equation suggests that oyster survival (OS) did not decline with increasing burrow density (OS = 13.56 + 0.03 B - 0.1 D + 0.07 F, where B is burrow density, D is number days since oyster placement, and F is presence or absence of fence; adjusted R2  = 0.71, SE = 1.87).

Large Plot Experiment

Another experiment measured the effects of burrowing shrimp over time at a larger scale.  A 1.4 ac (135 × 45 m ) plot was laid out on a commercial oyster bed (Bay Center Mariculture) that was heavily infested with burrowing shrimp (Figure 4).  Half of the plot was aerially treated with Sevin 80S (8 lb a.i./ac) in June 2003 while the other half remained untreated.  In August, after the treated substrate had become suitably firm, the southern halves of both treated and untreated plots were planted with seed shell (cultch) while medium sized developing (or fattening) oysters were planted on the northern halves.  Shrimp burrows and oysters were counted within each of twenty or more 1 m2 plots aligned along two intersecting diagonal transects in each plot before and after shell placement.  Oyster density was measured as percent cover and absolute numbers.  Both fattening and seed oysters increased in size in the large plot experiments.   As noted above, seed can be counted as individual pieces of cultch (shell with multiple small individual oysters on each piece), or as individual oysters as they grow.  While absolute abundance of either cultch or oysters is an important measure of yield, it was closely correlated to percent shell cover (R2 = 0.87, SE = 1.40, N = 20), which provided a more consistent and standardized unit of measurement in these plots.  Changes in oyster density relative to shrimp burrow density were even more apparent when percent shell cover was normalized to maximum cover in each treatment.   As expected, both seed and fattening oysters survived longer in sprayed plots compared to unsprayed plots (Figure 5).  Seed density and shell cover declined to near zero within 8 months of planting on the unsprayed plots.  Burrow densities were higher in the unsprayed plots (mean = 9.6 13.9 burrows/m2 vs 0.8 1.2 burrows/m2 in the sprayed plots), especially during the warm summer months when shrimp were most active, but the average at the end of the observation period was very similar to that at the outset of the experiment.  Burrow density was low in the sprayed plots when oysters were planted (mean = 0.5 0.6 burrows/m2), and increased only slightly over the twenty month observation period, most likely due to immigration by adult shrimp.  Densities of both fattening and seed oysters dropped sharply during the first winter after planting, even at low burrow densities, demonstrating the importance of seasonal events unrelated to shrimp density.  A relationship between oyster yield, burrow density, and time was evident, but a threshold could not be discerned.

The Financial Costs of Burrowing Shrimp Management

Preliminary information on farm economics, especially as related to the burrowing shrimp control program, was gathered from interviews conducted in 2004 with growers owning tidelands that had been closely observed by WDFW scientists for the last few years. Questions were standardized and  mailed to the growers before the interview took place and responses recorded in person at the interview (Table 1).

The six growers we interviewed owned together approximately 12,300 acres and actively farmed about 7,6000 acres of tidelands in Willapa Bay and Grays Harbor.  The farms ranged in size from 190 – 7,110 acres.  Most land under discussion was used for ground culture.  They were either “seed beds” where seed was planted and after 2 –3 years moved to a “fattening bed” until harvest, or seed-harvest beds where seed was planted and allowed to mature to harvest all on the same bed.  Most growers reported good growth and attributed this to either specific conditions like tidal elevation or proximity to channels or to recent ocean conditions.  Eight of the twelve beds were harvested by dredge; the rest were picked by hand.  Most of the beds were also harrowed with a pasture harrow at some point during the growth cycle to bring oysters up to the surface and/or break up clusters.  All of the beds had been previously treated with carbaryl.  Most beds were treated regularly on a 3-4 year cycle although at least one bed had not been treated for a decade.  The growers reported that shrimp were a consistent problem with only one bed being reported as receiving moderate re-infestation.  Eelgrass was reported present on all of the beds, but clearly more common on some.  Most of the growers answered that they could not afford not to plant oysters on these beds, except one grower who felt he might be able to do so if market downturns forced him too.  Most growers reported that they did not pay close attention to the shrimp burrow count information provided as part of the pesticide application program, but were concerned that the legally defined treatment threshold of 10 burrows per m2 was not exceeded.

An additional oyster grower was interviewed in 2005 following the procedures and questions presented in last year’s report.  This grower farms a very small area (~1 ac) without using carbaryl, relying instead on dozens of closely spaced PVC pipes to which oyster seed adheres and develops fairly naturally.  The grower relies on a small “you pick” market which partially supports his income.  Other growers have found such practices to be uneconomical and not sustainable at a larger scale.

Decision Tree for Control Action

However, IPM theory and practice can still be applied to the problem.  We suggest that a decision tree be used as an empirical economic action threshold based on 1) duration that oyster crop will remain on the bed, 2) treatment history, 3) recent shrimp recruitment patterns, and 4) a revised and adjustable minimum threshold burrow count (Figure 6).  Cultural and methodological hurdles exist, but level 1 integration using this approach seems feasible regardless of the tools chosen for shrimp control.

Habitat Crushing and Disruption

Prior to the implementation of carbaryl as the primary control for burrowing shrimp, Wiegardt and Sons oyster farm studied the ability to manage shrimp by habitat crushing and disruption, which would theoretically both injure shrimp and expose them to greater predation.  Under the direction of John L. Wiegardt Jr., a U.S. army “weasel” was used to pull spring tooth harrows across oyster beds heavily infested with burrowing shrimp. The harrowed land was rolled after harrowing, using “snow cats” to pull weights of differing composition and weight.  These efforts were minimally effective and not cost effective after the registration of carbaryl in 1963.  In 1960 the cost of harrowing and rolling was $50 per acre whereas in 1963, the cost of managing burrowing shrimp using Sevin was $10 per acre (Clyde Sayce, Director, WDF Nahcotta Laboratory 1955 – 1980, comment, January 31, 1991).  Other crushing trials featuring the semi-amphibious vehicle, the Rolligon™ were conducted by Taylor Shellfish in the mid-1980s with somewhat mixed results, but the experimental design was not well replicated or recorded.  We thought it would be worthwhile to revisit these tactics given more modern equipment and a better experimental design.

Burrow Crushing Trials – K. Patten (2002 – 05)
In August 2002, trials featured both a large-wheeled semi-amphibious vehicle (Rolligon™) (Figure 7) and a smaller tracked-wheel vehicle (Argo™).  The Rolligon was tested at a single site at Nahchotta but at three different tidal elevations, so the substrate was dry, moist, or covered with shallow water at test time.   Substrate compaction was recorded in untreated parts of each soil type as the amount of pressure (psi) needed to push a 1" diameter probe 3" into the substrate and number of burrows were recorded inside ¼ × ¼ m2 grids.

Burrow density was significantly reduced at two days after treatment (DAT) compared to pre-treatment counts, and in areas treated with the Rolligon vs immediately adjacent untreated areas (Table 1).  Substrate was significantly compacted in treated vs untreated areas of moist and shallow water substrates, but not in the dry area (Table 2).  Plot size was not large enough to prevent immigration from shrimp over a seasonal time frame so subsequent assessments showed little difference between treated and untreated areas.  Furthermore, the Rolligon became stuck in the substrate frequently enough to halt the tests.  Trials of the Argo were of similar design, but the vehicle was able to travel through the mud more easily than the Rolligon.  At short post application intervals densities of shrimp burrows were reduced to similar levels as those in the Rolligon experiments.

Trials continued in 2004 using the WSDA’s Marsh Master II (Figure 8) more commonly used for crushing the invasive cordgrass, Spartina alterniflora.  Trials comprised large plots (2 – 3 ac for the 3 crush plots; ½ ac for the 6 crush plot) that were crushed sequentially in June, July and September, 2004.  Plots were placed in areas of low, medium, or high initial shrimp density.  Number of shrimp burrows were counted in 25 × 25 cm grids after each event.

Burrow density was reduced relative to pre-trial levels in October 2004, especially in the high density plots, but had rebounded substantially by July 2005 (Figure 9).

Crushing preferentially targeted male shrimp, as more males were observed on the substrate surface immediately after crushing by a ratio of 5.6/1.0.  Uninjured but exposed female shrimp also burrowed back into the substrate faster than males.  Fifty percent of the female population reburrowed at an average of 6.0 min while 50% of the males took 7.6 min.

Trial of “The Kansas Machine” to suppress burrowing shrimp – S. Booth, D. Penny (2003)

Observations in 2002 during trials against the invasive cordgrass, Spartina alternaflora, suggested that shallow rototilling may have some potential to suppress burrowing shrimp.  A modified airboat (Master’s Dredging, Inc., Lawrence, Kansas) consisted of an exceptionally large engine and a front-mounted shaft equipped with tines that could be hydraulically raised and lowered to a maximum rototilling depth of 4" (Figure 10).

Arrangements were made in 2003 to assess the potential of  the modified airboat (e.g., “the Kansas machine”) to suppress burrowing shrimp.  The trials were contingent on ancillary jobs against aquatic weeds in California and additional tests against cordgrass in Willapa Bay.  Consequently, a trial was not conducted until October 5, when conditions were no longer optimal.  Burrowing shrimp become less active and burrow deeper as temperatures cool in the fall.  The maximum low daylight tide on that date in North Willapa Bay occurred at 4:00 PM and was +2 ft in height.  Consequently, the study was sited on slightly higher ground than that of most commercial oyster beds (Figures 11A and 11B) and was of moderate to low burrow density (mean = 13.8 vs > 20 burrows per m2 on heavily infested ground).  The exotic eelgrass Zostera japonica was patchily distributed over the study area, while the native species, Z. marina, was rare.  Four somewhat parallel transects, each ~200 m long, were placed perpendicular to the waterline at 2:30 (Figure 11C).

Beginning at 3:30 PM, the Kansas Machine rototilled twice along the four transects.  Each rototill pass lasted 4 – 5 minutes and, with discussion between passes, all eight were completed by 4:30.  A fifth, single pass path was rototilled into shallow water (~20 cm at the deep end) as the Kansas Machine left the study site.  No shrimp were observed leaving burrows during the course of the trial.  Shrimp burrow density inside three of the rototilled paths were compared with densities on adjacent untreated ground on October 14 during a +2 ft low tide.  Burrows were counted within 20 m2 grids spaced 10 m apart within and beside each transect.  Percent cover of Z. japonica was also estimated in each grid to compare affects of rototilling on that species.  Z. marina was absent from all grids.

Burrow density was significantly lower on untreated ground than on immediately adjacent rototilled ground  (Figure 12A).  Rototilling most likely temporarily exposed burrows that were less apparent in undisturbed substrate.  Although percent cover of Z. japonica was significantly lower in the rototilled plots (Figure 12B), only the grass stems had been cut and much of the root remained.

Although preliminary results suggest the Kansas Machine has low potential to suppress burrowing shrimp, they are strongly qualified by the sub-optimal conditions under which they were acquired.  Suppression would most likely be better during summer months and at lower tidal elevations when and where shrimp are more active and shallow.  While a similar trial against cordgrass concluded that the Kansas Machine currently lacks sufficient power and speed to efficiently suppress the large acreage of that weed, these concerns were less of a concern here, given the smoother substrate and smaller acreage of commercial oyster ground.  Nevertheless, more power and deeper rototilling could improve the machine’s potential to suppress burrowing shrimp.

Habitat Crushing: Summary and Evaluation

None of the crushing tactics offer much potential to suppress burrowing shrimp to low enough levels and for sufficient time periods to warrant further consideration.  Burrowing shrimp are highly adapted to their fossorial habitat and can apparently withstand large amounts of compaction and remain mobile in highly dense substrates.  In related laboratory experiments, burrowing shrimp were able to construct burrows in substrates with levels of densification much higher those found in the field (Weaver 2006).  Shrimp could only be crushed at stress levels found at least 1.5 m below the ground surface in association with very high surface loads.

Alternative Chemistries and Their Delivery

As previously mentioned, carbaryl was selected as the primary pesticide for burrowing shrimp control in the early 1960s following the assessment of several compounds.  Since then, pesticides with novel chemistries have been developed, implemented, and are common throughout most of American agriculture.  These include “biorational” compounds with low vertebrate toxicity or compounds that selectively target pests with minimal impact to beneficial or benign organisms.  Pyrethrins and pyrethrums are plant extracts with low mammalian toxicity that are certified for organic use.   Some “biorational” chemistries are less selective in the marine environment where, for example, insecticides that disrupt the development of chitin in the mandibles of insect pests (e.g., diflubenzeron) would affect all crustaceans, not just burrowing shrimp.  Other insect growth regulators (IGRs) like tebufenozide, methoprene, and pyriproxyfen target specific developmental hormones in some insect orders (e.g., moths, beetles or flies) but are less likely to effect only shrimp in an estuary.  Compounds that disrupt mating behaviors in some insects by mimicking sex pheromones (e.g., moths) have yet to be isolated and synthesized for burrowing shrimp.  Microbial control agents, though gaining importance in terrestrial systems, are less well known in aquatic systems and would be difficult to contain to local beds.

Nevertheless, we have examined a plethora of chemistries, beginning with those that might be relatively easy to register (e.g., those on the EPA’s 25 b list including cinnamon, clove oil, and garlic oil), but extending to other organic materials (azadiractin, sulfurs, pyrethrums) as well as some conventional materials (pyrethroids) and the “biorational” compounds described above (mostly neonicotinoids).  We also tested several adjuvants, including some stickers and sinkers, as well as lignosulfate, which reportedly enhances the ability of some compounds to stick to sediment particles.  The organic penetrants, Li700 and acetic acid, were tested as a potential synergists of pyrethrins.

In addition to the class of pesticide, the method of application can affect efficacy, selectivity, ease of use, and cost.  The current application method, aerial application using helicopters, is by far the most cost effective and quickest, but is limited to specific application dates as determined by the tides, weather, and the season of salmon migration.  Subsurface injection of pesticides is becoming a common technique to directly target subterranean pests with little contamination or exposure to surface organisms (Khalilian et al. 2000, Mann et. al. 1999, Ressler et. al. 1997).  The amount and distribution of injected material can be varied according to depth and spacing of the injectors and the speed of the vehicle.  Burrowing shrimp that are not directly killed by pesticides may be entombed, stressed, or forced to reconstruct burrows and so contact more pesticide.  Subsurface injection may reduce pesticide exposure to surface dwelling non-target organisms and offsite impacts due to drift.  Alternative chemistries were applied either to the bed surface using a CO2 powered backpack sprayer or subsurface delivery from land based vehicles or from boats or barges.  Water volumes for backpack applications were 10, 20, 30 or 40 gal/ac to approximate the water volume used in the standard applications of Sevin by helicopter (10 gal/ac).  Subsurface injections featured much larger water volumes (100 – 400 gal/ac).

All compounds were applied under Washington State Experimental Use Permits issued by WSDA, which limited total acreage applied per year to less than 1.0 ac for some compounds and less than 0.1 ac for others.  Plot sizes were therefore frequently small (i.e., 10×10m, 10×30m, 30×30m) and minimally replicated (3 or 4 plots per application).

●  Topical (broadcast) applications – K. Patten (2003 – 06)

○  25b list Materials

Many of the compounds on the EPA’s 25b list of minimum risk pesticides (exemption from registration) were tested for efficacy against burrowing shrimp.  Frankly, the expectation for many of these compounds was low, so many compounds were tested in single replicate plots to expedite the program.  Most 25b list materials were tested in 2002, 2003.  The following plant extracts showed little to no efficacy: 1) cinnamon oil, 2) citric acid, 3) citronella oil, 4) cedar oil, 5) eugenol, 6) garlic oil, 7) geranium oil, 8) linseed oil, 9) malic acid, 10)  peppermint oil, 11) rosemary oil, 12) thyme oil, and 13)  white pepper.  A variety of salts applied at different rates and in different formulations were also ineffective (Figures 13, 14).

○  Other Easily Registered Materials

Some materials with no or relatively low mammalian toxicity, such as organic pesticides, are easier to register than some conventional materials.  These include plant extracts (pyrethrums (Ecozone® and Pyganic®), azadiractin (Neemix®), capsaicin, habanjero pepper, and yucca), the fungal derivative sabdilla (Veratran®), elemental sulfurs (Thiosol® and Kumulus®), copper (Cutrine®), cryolite (Kryocide®), lime, nitrates, and sulfates.  We also tested some materials with existing aquatic use: Seaklean®, and Spectrus® are antifouling biocides used to sanitize bilge tanks in ships and clean water discharge pipes of algal and fungal growths.

Materials were compared against the both carbaryl (Sevin®) applied at a reduced rate and an untreated check and were generally ineffective (Tables 3,4).  The aquatic herbicide endothol-k, applied at four different rates, did not significantly suppress burrowing shrimp compared to an untreated check (Table 5).

○  Pyrethroids, Pyrethrums, and IGRs

These compounds, applied by broadcast, were compared together in a single large plot in 2006.  Results were very similar at both 13 and 102 days after treatment (Figure 15).  Carbaryl (Sevin) was among the most effective, even at the relatively very low application rate of 2 lb a.i. per ac.  Aside from Deltaguard (Bayer Corp.), the pyrethroids were also quite effective, especially Mustang (FMC).  The pyrethrin compound, (Pyganic; MGK), was somewhat effective compared to the untreated check, especially at the higher rate, but number of burrows remained above 10 per ac.  Pyrethrum (Evergreen; MGK) showed greater efficacy at 102 compared to 13 DAT, but also failed to pass the 10 burrow/ac threshold.  The IGR, pyriproxyfen (Esteem; Valent) was ineffective.  An ancillary trial showed the pyrethrin Pyganic to be intermediate in efficacy compared to carbaryl at 60 days after treatment (Table 6).

○  Neonicotinoids

Trials in 2003 also included the neonicotinoid, acetamiprid (Assail®; Cerexagri, Inc) in comparison with the standard carbaryl (Sevin 80WP®) and an untreated check (Table 7).  Assail was generally ineffective in these trials.  However, the 2004 broadcast trials showed the neonicotinoid, clothiniadin (Poncho®; Bayer Corp.), to be more effective than carbaryl, although the latter was applied at less than the standard rate (8 lb a.i./ac) (Table 8).

Both Cerexagri and Bayer indicated that they were not willing to participate in a registration process for their products’ use on Washington state commercial shellfish grounds, but the patent for the neonicotinoid imidacloprid (Admire, Pravado; Bayer Corp.) was scheduled to expire in 2006, so trials were initiated for that material with generally acceptable results.  Imidacloprid was effective at a range of rates (Table 9).

○  Adjuvants

Several adjuvants were compared for the ability to increase the efficacy of carbaryl, but had little affect (Figure 16).  In 2004, lignosulfate was also tested as a sticker in association both carbaryl and acetamiprid with encouraging results (Table 10).

A trial conducted in 2006 showed that the adjuvant, Sinker®, slightly but not significantly enhanced efficacy of carbaryl (Table 11).

The penetrants Li700 and acetic acid, both certifiable organic materials and reported as synergistic for pyrethrum-based materials, not only failed to enhance the efficacy of Pyganic, but apparently suppressed efficacy slightly (Table 12.) Results of another trial conducted in 2006 were more encouraging.  Burrow densities were lower in plots treated with Pyganic in combination with Li700 than in plots treated with Pyganic alone, but the difference was not significant (t-test; P = 0.05).

●  Preliminary trials of subsurface delivery and post-application harrowing or rolling – J. Durfey, S. Booth (2003)

Trials in 2003 featured comparisons of the reduced risk material acetamiprid (Assail 70 WP) applied at either 14.4 or 7.2 oz a.i./ac, two low rates of carbaryl (Sevin 80W @ 3 or 1 lb a.i./ac) using two different methods of application: 1) on the surface using CO2 backpack at ~30 psi 2) sub-surface shank injection using an apparatus designed by Mr. Durfey (Figure 17).  Two post-application tactics to enhance the penetration of material through the bed surface were also compared: 1) rolling of a large 500 lb water-filled cylinder and 2) harrowing.  The resulting 13 treatment combinations were made inside 2×6m plots arranged and replicated within strips.  The strips were also replicated among three blocks (Figure 18.  A 3 m buffer was placed between each strip (not shown in Figure 14).  Treatments were applied on July 30 at low tide and were assessed by counting shrimp burrows within a m2 grid placed in each plot at 1 and 28 days after treatment (DAT).  Results were analyzed using analysis of variance featuring orthogonal contrasts and t-tests.

Only 2 of 6 relevant contrasts were significant at 30 DAT (Table 13).  Significantly fewer shrimp burrows were observed in plots treated with the high rate of Assail than in plots treated with the high rate of carbaryl (16.03 ± 0.84 vs 20.73 ± 1.15; mean ± SE).  The low rate of Assail was not significantly more effective than the low rate of carbaryl, but Assail was more effective when both rates were pooled.  Application by shank was not significantly more effective than application by backpack and harrowing was not more effective than rolling.

●  Land-based subsurface injection – K. Patten, J. Durfey, S. Booth, D. Cheney (2004 – 05)

The objectives of these studies, executed in 2004 and 2005 were to 1) compare the efficacy of materials delivered subsurface against burrowing shrimp, and 2) research and develop alternative platforms for delivery.  We used a semi-amphibious heavy vehicle, the Rolligon™, that was modified to apply materials subsurface at high volumes and pressures in combination with vibrating shanks (Figure 19).  The Rolligon was operated in first gear at 3.57 ft/sec for a theoretical field capacity of 1.8 ac/hr or an estimated (80%) actual field capacity of 1.4 ac/hr.  In 2005, we also used a spikewheel ™ technology that featured 8 spikewheels at 4" spacing pulled behind an ATV with GPS guidance that also precisely regulated pesticide delivery (Figure 20).  A calibrated volume of material was released from each injector-spike precisely when it reached the bottom of the wheel’s revolution.

Trials were conducted on property owned by Richard Wilson at Rosetta Beach on the west side the Bay Center Peninsula (Figure 21).  Two distinct types of substrates were present at the area: 1) a sandy /silty substrate close to the beach, and 2) a very silty substrate beyond a shallow drainage channel located about 600 m from shore.  The 2004 trials were conducted mostly in the first type of substrate whereas all trials in 2005 were conducted at the lower, siltier tidal elevation.  In 2004, treatments were replicated over both space (usually 4 replicated plots per treatment) and time (usually 3 treatments on 3 consecutive days).  In 2005, treatments were more often replicated among plots rather than tidal interval, given the large number of compounds tested.  In both years, treatment plots were 24×48 ft and separated by 6 ft buffers.  Applications were made over a 2 – 3 day interval every 2 weeks at low tide.  Number of shrimp burrows and clam holes in ¼ m2 grids and percent cover of eel grass and water in 1 m2 grids, as well as water depth, were noted both before and after treatment.  Counts obscured by deep water or high percent eel grass cover were eliminated from the analysis.

○  2004 Results

Blocks A and B, treated during mid August, showed conclusively that subsurface injection of low rates of carbaryl (Sevin 80SP™) (1.5 and 3.0 v s the standard rate of 8.0 a.i./ac) significantly suppressed density of shrimp burrows (Table 14).  Other Rolligon treatments also significantly suppressed burrowing shrimp compared to the untreated check.  Treatment effects on clam holes were less apparent at 2 weeks after treatment and not significant when compared to pre-treatment levels. Shrimp burrows and clam holes at Block C were counted during the second low tidal interval after treatment, rather than along with Blocks A and B during the first. so results were analyzed separately.  A recent storm had also obscured the difference between shrimp burrows and clam holes.  Consequently, results were not so often significant, more difficult to interpret, so not presented here.

A second set of trials in late August again featured the extreme low rate of Sevin (1.5 lb a.i./ac) along with a low rate of clothianidin (Poncho™)) (1.5 lb a.i./ac), and a moderate rate of flowable sulfur (THAT™) (3.0 lb a.i./ac).  Because block effects (Blocks D, E, and F) significantly interacted with treatment effects on shrimp burrow density, results from each block were analyzed separately.  Neither shrimp burrow nor clam hole density was significantly affected compared to the untreated check at Blocks D and E, regardless of how they were measured (Tables 15, 16).

Mean number of shrimp burrows was significantly lower in all plots treated with pesticides than in the untreated check at Block F (Table 17).  According to the difference between counts before and after treatment, THAT flowable sulfur was especially effective.  Mean number of clam holes were not significantly different among treatments.

As the summer progressed, the weather became increasingly inclement and the lowest tides shifted into the night, so another set of trials featuring the organic sulfur (Kumulus™) was not replicated over consecutive days like the previous trials.  Instead, 8 plots per treatment were replicated within a single large block (Block G).  Results showed that both a moderate rate of Kumulus as well as half that rate significantly suppressed both shrimp burrows and clam holes (Table 18).

In general, these trials demonstrated that both the shanking of shrimp infested oyster ground, as well as subsurface injection of pesticides can effectively suppress burrowing shrimp under certain circumstances.  Some results also indicated that subsurface injection of appropriate materials (e.g., Poncho or THAT sulfur) may be less detrimental to clam populations.  Inconsistencies in efficacy among the various blocks likely resulted from differences in substrate type, water depth, and both the distribution and density of shrimp.  Trials were initiated in the firmer sandy areas for safety and logistic reasons, but, as noted above, Blocks A – E were located in sandy areas where shrimp were very patchily distributed and tidal channels frequently intersected the blocks.  Blocks F and G were located further from the shore in silty areas of higher, more uniform shrimp densities.  These results reiterated that large plots and high replication are needed for good experimental protocol in the extremely patchy inter-tidal areas of Willapa Bay.  Future enhancements of these experiments will feature enhanced vibration technique and operation from an oyster barge.

○  2005 Results

In trials conducted during the first tidal interval, experimental materials were significantly more effective than the untreated check, but less effective than Sevin applied at a low rate (the standard rate for aerial applications is 8 lb a.i./ac) (Table 19).  Shrimp burrow densities in plots treated with experimental materials were reduced by nearly 50% in the intervals between pre- and both post-treatment assessments.  At the first post-treatment assessment, higher rates of the pyrethrum, Pyganic, were just as effective as the low rate at the first treatment assessment, but burrow densities in plots treated with the mid-range rate of elemental sulfur, 20 lb F/ac, did not differ significantly from untreated plots.  The sulfur treatment was more effective at the second assessment than at the first.  According to comparisons with Sevin and the check, efficacy of the rest of the experimental materials did not did not change much between the first and second assessments.  A comparison between pre- and post-treatment clam hole densities was not possible, but clams were apparently less impacted by the alternative materials than the Sevin treatments.  Differences in mean clam hole density were not always significant, particularly at the first assessment.

Comparisons of the several materials applied at the second tidal interval in mid-July showed a range of efficacy according to comparisons with the untreated check and pre-treatment burrow densities (Table 20).  The hot pepper (habanjero extract) was the most effective material at the first post-treatment assessment, although the interval before assessment was substantially shorter than it was for plots treated with pyrethrum (9 vs ~ 32 days).  Mean burrow densities in the habanjero and Pyganic-treated plots were not significantly different at the second post-treatment assessment.  The pyrethrum Ecozone was similarly effective at both assessments.  Thiosol, Ecotrol, Cinnacure, UN-32, and aqua ammonia were relatively less effective.  Clam hole density was reduced by ~ 40 – 50% in all treated plots at the first sample date, but appeared to have relatively less impact in the habanjero-treated plots at the second sample date.

A set of synthetic pyrethroids and organic pyrethrums was compared to a low rate of Sevin and untreated check during the third and fourth interval (August 5, August 17-18).  At the first assessment, mean burrow density was significantly lower in all treated compared to untreated plots, with values close to 10 / m2 for all treatments except Deltaguard and Esteem (Table 21).  Burrow densities were substantially higher in all but the Mustang-treated plots by the second assessment, ~30 days later.  In general, the pyrethrums had a relatively lower impact on clam hole density than the pyrethroids or carbaryl.

Results of the 2005 Rolligon and Spikewheel trials were further by analysis of variance blocked across general application date (Figure 22).

Although no single compound was applied at the same rate using the same application method, a comparison of two compounds applied at the same rate using both the Rolligon and the spikewheel showed that application by shank injection via the Rolligon was significantly more effective than application by Spikewheel in two of four comparisons.  However, the length of the interval before assessment also strongly affected efficacy (Table 22).  Accordingly, the relative abilities of the two application techniques to affect compound efficacy could not be compared with full confidence.

Operation of the Rolligon, however, was mechanically challenging.  Our particular model was aging and suffered from engine problems and leaky tires.  The Spikewheel unit was much easier to operate, although it was less able to travel through water deeper than ~20 cm.

●  Subsurface injection from a floating platform – K. Patten

In 2006, the spikewheel apparatus was carried on a 10×16'  pontoon raft (Figure 23).  A 16' aluminum boat equipped with a 40 hp outboard engine pushed rather than pulled the raft over the bed, to minimize effects of the engine’s propellor.  The spikewheels were mounted on a bracket that could be raised or lowered to a maximum depth of 12' using a hand powered winch (Figure 24).  The bracket initially carried 6 spikewheels, but after the first few trials, the number was reduced to 4, each spaced 19" apart, to treat a 6' wide strip.   The spikewheels were mounted and rotated independently from each other, but calibration tests showed that they remained well synchronized.  Another, non-spiked drive wheel placed in front the Spikewheels rotated at the same speed and regulated injection timing via a chain-drive belt connected to a roller pump, which pumped material through the injector-spikes precisely when they reached the bottom of each spikewheel revolution; speed did not influence the amount of material injected.  The pump delivered 48 gal/acre at a line pressure of 15 psi with the boat motor running 2900 rpm and land speed was 1.8 mph.  Material was delivered to the pump from one of two 50 gal plastic tanks via a manually operated “Y” valve.  Each spikewheel held 12 4" long spikes.  As in the land-based system, material was injected through a pore ⅛” in diameter located on the right side of the spike just above the tip.  Preliminary tests using dyes, as well as observation using scuba, showed that material was injected to a depth of ~7".

Most treatment plots in 2006 comprised one or more 6×150' strips (900 ft2 ≡ 1 gal material).  Strip borders were marked by PVC pipes as visual cues for alignment during applications at high water.  Injectors were checked prior to each trial and two calibration passes were made using water as material on areas next to the target strip.  Plots were not treated when wave height was more than 2' or wind gusted more than 15 mph.  Applications were made on both in-coming and out-going tides, but only in the direction of the tidal current; not cross-current.  The spikewheel assembly was lowered as the barge crossed the beginning of the strip, causing the drive wheel to contact the substrate and engage the pump.  The assembly and drive wheel was lifted as the barge crossed the end of the strip, disengaging the pump and halting injection.  Treatments were checked for uniformity on the next day’s low tide by examining the spikewheel paths in the sand.  Efficacy was assessed at one or more post application intervals by counting the number of shrimp burrows.  Results were analyzed using analysis of variance or t-tests when appropriate.  In some instances, results from trials involving separate application dates were compared by computing each treatment’s percentage (proportion) of the untreated check.  Percentages were transformed to the arcsin for statistical analysis.

○  Preliminary trials with carbaryl

Carbaryl (Sevin 80SP) was used in preliminary trials to both test the operation of of the floating barge / spikewheel system and to provide comparisons for other alternative materials.  Burrow density was significantly lower in all carbaryl treated plots than in untreated plots at both 6 and 49 days after treatment, but were either higher in the treated than in the untreated plots at 20 and 256 days after treatment (Figure 24).  The lack of significant differences among treatments at the longest interval was likely due to shrimp moving into the narrow (6' wide) plots as well as poor long-term efficacy.

○  Sulfur

Elemental sulfur (Kumulus 80DF, Microflo Co.) was tested in 3 separate trials.  Results of each trial were assessed at different post application intervals and compared to appropriate untreated check.  In most cases, burrow density was significantly lower in plots treated with sulfur than in untreated checks, but burrow density remained above 10 per m2 (Figure 26).

○  Plant Extracts

Efficacy of different rates of habanjero extract, injected with oil or water, were poor to moderate (Table 23), but plots were examined at only 4 days after treatment; efficacy might have been greater at a longer post application interval.  Extracts of lilly (Veratran®) and  powdered chrysanthemums (Ecozone®) also showed moderate to poor efficacy at 17 days after treatment (Table 24).

Efficacy of pyrethrins (Pyganic®) varied somewhat among trials.  In a trial where burrow densities were low, pyrethrins showed excellent efficacy (burrow density was reduced to nearly 10 per m2)  at 6 days after treatment (Table 25).  Overall burrow density was higher in another trial conducted 10 days later and, although burrow density was lower in plots treated with Pyganic compared to untreated plots, densities were ultimately substantially higher than 10 burrows per m2.

○  Methoprene

The insect juvenile hormone analogue (JHA) methoprene showed moderate efficacy at 13 days after treatment, but burrow counts at longer post application intervals were substantially higher (Table 26).

○  Pyrethroids

Three pyrethroids were applied in 3 separate trials: 1) Belay®, 2) Mustang® at 0.01 lb a.i./ac, and 3) Mustang at 0.005 lb a.i./ac and Capture® at 0.01 lb a.i./ac.  The separate applications complicated direct comparison, but most treatments were moderately effective, judging by comparisons between each treatment and its appropriate untreated control plot (Figure 27).   Surprisingly, the higher rate of Mustang was less effective than the lower rate.

○  Imidacloprid

The neonicotinoid pesticide, imidacloprid (Admire®; Bayer Corp.), was tested multiple times at various rates and locations (Tables 27, 28, 29).  Usually, efficacy of imidacloprid was greater (final burrow density was lower) at higher rates, but the response was not always linear.  At a test area near Nahcotta, where substrates were primarily sandy, burrow density were substantially, if not significantly, higher at rates less than 0.2 lb a.i./ac.  This was especially true at longer post application intervals (e.g., 42 or 50 days after treatment) (Table 27, Trials 1, 2).  Efficacy was not necessarily greater in plots treated with imidacloprid at rates greater than 0.2 lb a.i./ac (Table 27, Trial 2: 2nd and 3rd post application interval; Trial 5).  Burrow density was also significantly lower in plots treated with 2.0 lb a.i./ac imidacloprid than in plots treated with 3.0 lb a.i./ac carbaryl (Table 27, Trial 1).

Results of a trial conducted on sandy/silty substrates were confounded somewhat by heavy growths of eel grass (primarily invasive Zostera japonica, but also Z. marinera), which slowed tidal drainage, left standing water on the bed, and obscured burrow counts (Table 28).

Another trial, conducted at the Willapa Bay Fish and Wildlife Refuge, featured applications of imidacloprid (0.2 lb a.i./ac) on four different types of substrate.   Burrows were counted in four 1 m2 quadrants within and in a single 1 m2 plot adjacent to each treatment plot.  Shrimp burrow density was significantly lower in all treated compared to untreated plots ( ± SE, 52.2 ± 15.7 burrows/m2; LSD, P=0.05), but was significantly higher in a plot of silty hummocks than in plots of other substrate types (Table 29).

●  Underwater injector sleds / harrows – K. Patten, J. Durfey, S. Booth, D. Cheney

Theoretically, pesticides could be injected to greater depths and with more precision using a harrow than a spikewheel.  In 2005, McGregor Company (Pasco, WA) designed, with the help of Jim Durfey,  and constructed a large iron harrow with skids that could be adjusted to different lengths (Figure 28).  Pressured lines secured to the back of each of six curved tines (12" apart) delivered material to injectors.  Unfortunately, the unit was too so heavy that it was difficult to pull.  It also had difficulty remaining level while harrowing and became easily entangled in eel grass and other immovable objects (Figure 29).  The unit was modified in 2006 by removing the skids, allowing the tines to better settle into the substrate.  Still, the unit could not be adequately controlled.

Also in 2006, J. Durfey and McGregor Co. designed and constructed a much lighter flexi-coil spring harrow equipped with 18 straight tines arranged in a 4×4 ft square.  The leading row (6 tines) carried material through pressurized lines to injector ports located near the bottom of each tine (Figure 30).  The model proved too light and merely skimmed the surface.  The addition of weights (Figure 31) allowed suitable penetration.

Results of trials using the flexi-coil spring harrow were generally satisfactory (Figure 32), but the system was still difficult to operate and the amount of material delivered was difficult to control.

●  Potential of alternative chemistries and subsurface application: A summary

Aside from many of the materials on the 25b list and some of the pesticides based on plant or fungal extracts (Esteem, Veratran), most of the materials tested during the past 3 years suppressed burrowing shrimp to at least a moderate degree; over half the shrimp in treated plots were killed compared to untreated plots or pre-treatment assessments.   Far fewer suppressed shrimp to levels below or near the accepted damage threshold of 10 burrows per m2.  These were primarily the neonicotinoids and the pyrethroids.  Both the EPA and Washington State Department of Agriculture have indicated that the potential for registration of a pyrethroid for use in Washington State estuarine waters would be extremely small.  Despite their low mammalian toxicity, pyrethroids have a relatively low LC50 against fish and have been implicated as endocrine disrupters.  They have broad spectrum efficacy against arthropods and would likely impact most other estuarine invertebrates.

According to both published toxicity criteria and recent field studies not presented here, the neonicotinoids possess greater selectivity towards burrowing shrimp; the LC50 for rainbow trout and sheepshead minnow are relatively high and impacts to marine benthic invertebrates do not differ significantly from untreated plots.  As previously mentioned, the producers of most neonicotinoids (i.e., Bayer and Cerexagri) will not support the registration of their products for use against burrowing shrimp.  Imidacloprid, however, came off patent from Bayer in 2006 and several other companies have begun to produce it.

Some materials that showed moderate efficacy would theoretically have an easier path towards registration.  These include the pyrethrum Pyganic and the sulfur compounds.  Results for these materials were somewhat inconsistent, especially for the sulfurs.  This could be due, in part, to the composition of the benthic substrate.  Given the proper pH (e.g., slightly acidic: ~6.0 compared to normal pH of ~8.0 for saltwater) H2S will diffuse into Crangon crangon (a common estuarine surface-dwelling shrimp), causing toxicity (Visman 1996).  A specific bacterial community featuring sulfate-reducers like Desulfoyibrio and Desulfotomaculum can also enhance sulfide levels (Wang and Chapman 1999).  These factors may account for the variability in sulfur efficacy here.

The trials also indicated that subsurface application from a floating barge was nearly as effective as topical application.  Although the two techniques were never precisely compared in space and time, a larger proportion of shrimp burrows was suppressed in topically-treated plots compared to plots treated using the barge-mounted spikewheels in 3 of 8 approximate comparisons (Table 30).

The other potential advantages of subsurface delivery of materials for burrowing shrimp control from a floating platform include: 1) an extended seasonal window for application, 2) less off-bed movement leading to 3) decreased non-target impact.  These variables have yet to be tested.

These trials remained preliminary in other respects as well, primarily due to size limitations imposed by the Washington State Experimental Use Permit.  Test plots were quite small (< 0.01 ac) relative to the size of commercial beds (5 – 20 ac), allowing shrimp from surrounding unsprayed areas to recolonize and confound measurements of long term efficacy.

Most of the test sites used in the last two years also featured areas where shrimp burrow density was frequently greater than 60 burrows per m2.  Most commercial beds have been previously treated and burrow densities rarely exceed 30 burrows per m2.  Moderately effective materials might theoretically suppress populations to less than 10 burrows m2 while killing less than 80% of the shrimp.

Controlled Microcosm Trials – J. Colt, S. Booth

Field experiments with burrowing shrimp are difficult because it is very time consuming and expensive to determine the number of shrimp at the beginning and ends of experiments. It is much easier to determine the number of shrimp under laboratory conditions but holding protocols for burrowing shrimp are not currently available. The primary objective of this research was to develop collection, holding, and feeding protocols for ghost shrimp. The first year's work will be restricted to the holding of unfed ghost shrimp without sediment cover. Following work will repeat the experiments with sediment cover.

Ghost shrimp were collected from the lower intertidal zone approximately ½ mile west of the Mukilteo Ferry Dock in Snohomish County, WA during November and December, 2004. Shrimp were collected using a 4" diameter PVC slurper tube. Shrimp were stored in 5 gal buckets filled with sea water for the1.5 – 2.0 hours during collection and transportation to the laboratory (about 0.5 miles). Shrimp were placed into a large glass aquarium (35×47×10") with flowing filtered seawater (1 – 2gpm). Light levels were maintained at low levels by lowering the window shades and keeping the overhead light off. The overhead lights were turned on during the mortality checks. The shrimp were counted the next morning to determine the number of shrimp killed or damaged by the collection process and distributed to the experimental units. No sand was added to the aquariums. The shrimp were not fed during the experiments. Mortalities were observed every 1 – 3 days and removed after each observation. The number of shrimp that did not appear to be acting normally, such as lying on their backs or sides, was also recorded. Shrimp were observed for 30 – 40 days. Cumulative mortality over time was plotted and time to 50% mortality (LT50) were estimated by a linear fit of the data. Three experiments were conducted: 1) Two aquaria at 28 shrimp/m2, 2) One aquarium at 60 shrimp/m2,and 3) One aquarium at 46 shrimp/m2 with substrate. These experiments were largely exploratory, so some of the experiments were not replicated and statistical analyses were fairly simple.

In experiment 1, 20% of the total shrimp collected were killed or damaged directly as a result of the collection process, usually due to the slurper tube slicing through the shrimp. Mortality was low during the first 10 days, followed by 25 days of higher mortality, and a final period of 10 days with no mortality (Figure 33). The time to 50% mortality (LT50) was in the range of 31 -- 35 days. The number of shrimp not acting normal was relatively constant over the entire experiment. At the end of the experiment, 6 shrimp were missing from the Tank A and 5 shrimp from the Tank B. Missing shrimp likely escaped through the drain stand pipe. The drain will be screened in the future to prevent such escapes. In Experiment 2, collection of the shrimp resulted in the death or damage to 25% of the total number collected. Mortality was rapid with an LT50 of 9 – 10 days (Figure 34). In Experiment 3, two types of substrate were used: 1) various 4" ABS plastic tees and elbows, and 2) 6" sections of 3/4" and 1" PVC pipe. Without exception, no shrimp were found inside either of the two types of substrate.

In 2005, similar experiments conducted in association with an ancillary study of the potential of lugworms (Arenicola pacifica) to biologically control burrowing shrimp showed similar results: shrimp can be maintained for at least several weeks in tanks. Since then, several other short-term experiments featuring burrowing shrimp in tanks have been conducted, mostly by Brett Dumbauld, USDA/ARS.

WRITE, IMPLEMENT, AND DELIVER THE IPM PLAN – S. Booth

A preliminary framework for the development of “A comprehensive plan toward the development of an integrated program for burrowing shrimp management on commercial oyster beds” was submitted to Washington Department of Ecology in March 2003 and was updated to much greater detail in 2007. The plan included sections on 1) IPM Definitions & Concepts, 2) IPM Plan Goals, 3) Principle Authorities, 4) Principle Policies, 5) IPM Elements (Funding, Research & Development, Implementation, and Evaluation / Regulatory Compliance) and 6) References (Appendix A). The 2007 plan Research & Development component of the plan was enhanced substantially in 2004 through the influx of a special Washington State proviso of $200,000 per year. The 2007 Plan also includes a detailed summary of all research projects completed since 2003.

In addition to the written document, the IPM Plan was presented via workshops and presentations. Grower workshops have been held in early spring every year since 2002 in Long Beach, Washington. The workshops were well attended by growers, researchers, agency representatives, and public citizens. Workshops included progress reports on research, funding, and regulatory related activities, as well as a special discussion session where growers provided comments, suggestions, and general feedback. Research results and program status were also presented at the Annual Pacific Coast Shellfish Growers Association / National Shellfish Association Conferences held during the fall of each year. Fewer growers, but more scientists, including those from outside the Pacific Northwest, attended a special Burrowing Shrimp session. The session usually included wide ranging discussions of research progress and priorities.

The IPM Plan is referenced in the West Coast Shellfish Research and Education 2015 Goals and Priorities as Goal 5.1 (Pacific Shellfish Institute, 2006). The Plan is briefly described as part of the rationale to manage burrowing shrimp and the major elements of the plan are listed as high priority goals. The IPM Plan is also currently included in the upcoming revision of the Pacific Coast Shellfish Growers’ Association Environmental Codes of Practice.

Two grower surveys were conducted during the last three years. The first was conducted as part of a grower workshop in spring 2003. The survey comprised 20 questions that targeted each grower’s type and size of oyster culture, their site-specific need to control shrimp, and economic ability to pay for alternative management tactics. Unfortunately, only 2 growers completed the survey. As these were representatives from the largest and smallest oyster companies, their answers hardly represented the majority of oyster growers in Willapa Bay / Grays Harbor. A second survey was conducted at the 2004 spring workshop, with equally dismal returns. Individual oral interviews have been much more productive, but are more time consuming and difficult to quantify.

Newsletters were mailed to ~100 persons in March 2004 and again in November 2005. In addition to topics directly related to IPM, the 2005 newsletter reported on the progress of an NPDES permit-required study to describe the sediment impact zone related to a commercial carbaryl application.

Another tactic to disseminate the IPM plan and allow for discussion of related topics has been a Yahoo discussion group. The group description is as follows: “This is a discussion forum for Pacific Coast shellfish growers, especially oyster growers. Topics are devoted to farming and environmental issues, primarily those related to burrowing shrimp and Spartina. This forum is open to anybody but please keep subject matter on topic and respect the rights of all to express their opinions.”

Bird. E.M. 1982. Population dynamics of thalassinidean shrimps and community effects through sediment modification. Ph.D. Dissertation, University of Maryland, College Park.

Brenchly, G.A. 1981. Disturbance and community structure: an experimental study of bioturbation in marine soft-sediment environments. J. Mar. Res. 39:767-790.

Brooks, K.M. 1995. Long-term response of benthic invertebrate communities associated with the application of carbaryl (Sevin) to control burrowing shrimp, and an assessment of the habitat value of cultivated Pacific oyster (Crassostrea gigas) beds in Willapa Bay, Washington, to fulfill requirements of the EPA carbaryl data call In Final Report. Aquatic Environmental Services, Port Townsend, Washington.Burrowing Shrimp Control Committee (BSCC).
Burrowing Shrimp Control Committee [BSCC], 1992. Findings and recommendations and an integrated pest management plan for the control of burrowing shrimp on commercial oyster beds in Willapa Bay and Grays Harbor, Washington State. Burrowing Shrimp Committee Report to the Grays Harbor and Pacific County Commissioners. 140 p.

DeWitt, T.H., K.F. Wellman, T. Wildman, D.A. Armstrong, and L. Bennett. 1997. An evaluation of the feasibility of using integrated pest management to control burrowing shrimp in commercial oyster beds. 127 pp.

Dumbauld, B.R. 1994. Thalassinid shrimp ecology and the use of carbaryl to control populations on oyster ground in Washington coastal estuaries. Ph. D. Dissertation, University of Washington, Seattle.

Dumbauld, B.R., E.P. Visser, D.A. Armstrong, L. Cole-Warner, K.L. Feldman, and B.E. Kauffman. 2000. Use of oyster shell to create habitat for juvenile dungeness crab in Washington coastal estuaries: status and prospects. J. Shellfish. Res. 19(1): 379 - 386.
Environmental Protection Agency (EPA). 1993. EPA for Your Information. Prevention, Pesticides and Toxic Substances (H7506C). 2 pp.

Khalilian, A., Williamson, R., Sullivan, M., Mueller, J. & F. Wolak., 2000. Subsurface injection versus surface application of composted municipal solid waste in cotton production. ASAE, Agricultural Utilization Research Session.

Mann C. L., Mengel D.B. & S.E. Hawkins 1999 Injection is most efficient method of uan placement. The Fluid Journal Vol. 3 No. 3
Pacific Shellfish Institute. West Coast Shellfish Research 2015 Goals and Priorities. Pacific Shellfish Institute, Pacific Coast Shellfish Growers Association, National Shellfisheries Association / West Coast Branch. Olympia, Washington. 28 pp.

Peterson, C.H. 1977. Comparative organization of the soft-bottom macrobenthic communities of Southern California lagoons. Mar. Biol. 43:343-359.

Posey, M.H., B.R. Dumbauld and D.A. Armstrong. 1991. Effects of a burrowing mud shrimp, Upogebia pugettensis (Dana) on abundances of macro-infauna. J. Esp. Mar. Biol. Ecol. 148:283-294.

Ressler, D. E., Horton R., Kaspar T.C. & J. L. Baker. 1997. Localized soil management in fertilizer injection zone to reduce nitrate leaching. Soil & Water Quality Res. 2150 Pammel Dr. Ames, IA.

Simenstad, C.A., and K.L. Fresh. 1995. Influence of intertidal aquaculture on benthic communities in Pacific Northwest estuaries: scales of disturbance. Estuaries 18:43-70.

Tamaki, A. 1994. Extinction of the trochid gastropod, Umbonium (Suchium) moniliferum (Lamark), and associated species on an intertidal sandflat. Res. Popul. Ecol. 36:225-236.

Vismann, B. 1996. Sulfide species and total sulfide toxicit in the shrimp Crangon crangon J. Exp. Mar. Biol. Ecol. 204:141-154.

Wang, F. And P.M. Chapman. 1999. Biological imiplications of sulfide in sediment – A review focusing on sediment toxicity. Env. Tox. and Chem. 18:2526-2432.
WDF/WDOE 1992. Use of the insecticide carbaryl to control ghost and mud shrimp in oyster beds of Willapa Bay and Grays Harbor. Washington Department of Fisheries / Washington Department of Ecology. Olympia, WA. 147 pp.