Coastal and Estuarine Research Federation Oysters, Crabs, and Burrowing Shrimp: Review of an Environmental Conflict over Aquatic Resources and Pesticide Use in Washington State's (USA) Coastal Estuaries Author(s): Kristine L. Feldman, David A. Armstrong, Brett R. Dumbauld, Theodore H. DeWitt, Daniel C. Doty Source: Estuaries, Vol. 23, No. 2 (Apr., 2000), pp. 141-176 Published by: Coastal and Estuarine Research Federation Stable URL: http://www.jstor.org/stable/1352824 . Accessed: 15/04/2011 10:25 Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at . http://www.jstor.org/action/showPublisher?publisherCode=estuarine. . Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. Coastal and Estuarine Research Federation is collaborating with JSTOR to digitize, preserve and extend access to Estuaries. http://www.jstor.org Estuaries Vol.23, No. 2, p. 141-176 April2000 Oysters, Crabs, and BurrowingShrimp:Review of an EnvironmentalConflictOver AquaticResources and Pesticide Use in WashingtonState's (USA) Coastal Estuaries KRISTINE L. FELDMAN1 DAVID A. ARMSTRONG School of Fisheries Box 355020 Universityof Washington Seattle, Washington98195 BRETT R. DUMBAULD WashingtonDepartmentof Fish and Wildlife Willapa Bay Field Station R 0. Box 190 Ocean Park, Washington98640 THEODORE H. DEWITT EnvironmentalProtectionAgency 2111 SoutheastMarine ScienceDrive Newport, Oregon97365 DANIEL C. DOTY WashingtonDepartmentof Fish and Wildlife Mail Stop 43200 600 Capitol WayNorth Olympia, Washington98501 ABSTRACT: Washington State's coastal estuaries are productive shallow water environments that support commercial fisheries for Dungeness crabs (Cancer magister) and English sole (Parophrys vetulus) by providing 0+ (settlement to age 1) populations with critical refuge and foraging habitats until subadults migrate to the nearshore coast. Intertidal mudflats also constitute prime areas for commercial oyster (Crassostreagigas) culture, an important industry for the coastal communities of Willapa Bay and Grays Harbor that supply much of the nation's oysters. Conflicts over natural resources and estuarine utilization have arisen over the last 37 yr due to the use of carbaryl (an organocarbamate pesticide) by oyster growers on their grounds to control populations of burrowing thalassinidean shrimp (Neotrypaea californiensis and Upogebia pugettensis). Burrowing shrimp, which have an indirect negative effect on oyster survival and growth through bioturbation and sediment destabilization, are killed by carbaryl, as are 0+ and subadult Dungeness crabs, 0+ English sole, and other non-target species present on the tideflats at the time of application. The pesticide is delivered at 9 kg ha-1 directly to the mudflat as a wetable powder during low tides in July and August. Commercial crabbers and other groups who have economic, recreational, and environmental interests in the estuaries have generally opposed use of the chemical that oyster growers maintain is essential to sustain production levels. For years, government natural resource agencies that regulate the use of carbaryl lacked critical information needed to effectively manage the program. An Environmental Impact Statement (EIS) and Supplemental EIS have provided much of that data and helped shape management decisions with regard to establishing carbaryl concentration rates and total allowable spray area. Additional research is needed to develop more economically and environmentally sound policies for shrimp control based on burrowing shrimp-oyster interactions on an estuarine-wide scale. In this paper we review issues pertaining to oyster culture, the use of carbaryl to control burrowing shrimp populations, and effects on non-target species, drawing upon research from published articles as well as unpublished data collected by the authors. We also discuss what is known of burrowing shrimp life history and ecology and emphasize the importance of integrating information on shrimp, such as timing of recruitment, variability in year class strength, and patterns of habitat use, into carbaryl control policies or alternative strategies that 1 Corresponding author; tele: 206/685-3609; fax: 206/685-7471; e-mail: kfeldman@fish.washington.edu. ? 2000 EstuarineResearch Federation 141 142 K. L. Feldmanet al. may be developed in the future. We recommend controlled experimentation be done to examine the ecological effects of delaying carbaryl application to some ghost shrimp beds until October after peak recruitment of 0+ ghost shrimp has occurred, allowing the number of hectares treated each year to vary based on fluctuations in pest population densities, and modifying the substrate by applying a dense layer of oyster shell to the mudflat (shell pavement) to reduce recruitment of ghost shrimp. Introduction Washington State's coastal estuaries are productive shallow water environments that provide critical habitat and feeding grounds for numerous species of aquatic invertebrates, fishes, migratory shorebirds, and waterfowl, and help support commercial fisheries for Dungeness crabs (Cancer magister) and English sole (Parophrysvetulus) by providing nursery areas for juvenile stages (Stevens and Armstrong 1984; Armstrong and Gunderson 1985; Gunderson et al. 1990). Located north of the Columbia River along the southwestern coast of Washington, Willapa Bay and Grays Harbor also constitute prime areas for commercial oyster culture. The Pacific oyster, Crassostreagigas, brought from Japan in the 1920s, is the principal species cultivated. The native Olympia oyster, Ostrealurida, once the basis for the industry, still occurs in some areas but is not significant commercially. Second only to Louisiana (MacKenzie 1996), Washington produces approximately 25% of the nation's oysters (Conway 1991), with Willapa Bay and Grays Harbor accounting for over 60% of the state's production (Hoines 1996). Although much of the physical environment of Willapa Bay and Grays Harbor is suitable for aquaculture, production is partially limited by two species of burrowing shrimp (Crustacea, Decapoda, Thalassinidea), the ghost shrimp Neotrypaeacaliforniensis (formerly Callianassa californiensis;Manning and Felder 1991) and the mud shrimp Upogebia pugettensis (hereafter referred to primarily by common names), which are native to estuarine intertidal and shallow subtidal sediments along the Pacific coast of North America (Stevens 1928; MacGinitie 1930, 1934). Mud shrimp tend to be pale orange in color, although blue and gray variations exist, grow 4-5 mm carapace length (CL) yr-~ on average in Washington (Dumbauld et al. 1996), and attain a maximum adult total length of approximately 15 cm. Ghost shrimp tend to be orange, yellow, or pink in color, grow 2-3 mm CL yr-l (Dumbauld et al. 1996), and reach an adult total length of approximately 13 cm. Both species are sexually dimorphic and have a lifespan of approximately 4-5 yr (Bird 1982; Dumbauld et al. 1996). Unlike red rock crabs (Cancer productus), Japanese oyster drills (Ceratostomainornatum), seastars (Pisaster ochraceus, Evasterias troschelli), and flat- worms (Pseudostylochusostreophagus),which prey directly on oysters (Stevens 1928; Woelke 1957; Chew and Eisler 1958; Baker 1995), thalassinidean shrimp affect oyster survival and growth indirectly through sediment disturbance. As shrimp burrow through the mud constructing extensive subterranean galleries up to 90 cm in depth with multiple openings to the surface (Stevens 1928; MacGinitie 1930, 1934; Thompson 1972; Swinbanks and Luternauer 1987), sediment compaction is reduced to the point that oysters growing directly on the benthos sink into the unconsolidated mud. Settling larvae and spat (juvenile oysters) are particularly vulnerable to burial or suffocation by suspended sediments, and growth of suspension feeding bivalves also is reduced due to clogging of gills and ciliary tracts (Stevens 1929; Loosanoff and Tommers 1948; Washington Department of Fisheries [WDF] 1970; Peterson 1984; Murphy 1985; WDF and Washington Department of Ecology [WDOE] 1985, 1992). Attempts to control burrowing shrimp in areas of aquaculture have occurred since the early 1900s following reported expansions of shrimp populations where oysters were formerly dominant. Stevens (1929) suggested that changes in oystering practices might have contributed to the shrimps' success in colonizing new grounds. She observed that natural beds of Ostrea lurida formed a 15-20 cm thick layer of cultch (live oysters attached to non-living shell) that was relatively impenetrable to burrowing shrimp and provided elevated substrate for the attachment of newly settled spat. Eventually, however, all the cultch was removed from natural beds to harvest the attached oysters. The shell substrate was never returned to the intertidal, which possibly enabled the shrimp to invade bare sediment. The reef-like structure of natural oyster beds was therefore replaced with a single scattered layer of oyster cultch. This type of oyster farming, while reducing competition among individual oysters for food, was characterized by greater susceptibility to shrimp colonization and bioturbation. Historical measures to protect cultivated oyster beds from the effects of burrowing shrimp included graveling, shelling, and burying wooden boards, none of which were particularly effective (Stevens 1929). Other methods of control such as harrowing and use of weighted plastic were considered costly and impractical for large-scale application (WDF 1970). Oysters,Crabs,and BurrowingShrimp In the early 1960s, experimental application of a broad range of pesticides tested by Dr. Victor Loosanoff for use on shrimp-infested oyster grounds identified carbaryl (1-napthol n-methyl carbamate; sold under the trade name Sevin?) to be an effective, practical, and relatively inexpensive method to control burrowing shrimp (WDF 1970). Carbaryl is a non-persistent organocarbamate pesticide that is extremely toxic to arthropods (Mount and Oehme 1981). Carbaryl inhibits activity of acetylcholinesterase, an enzyme that hydrolyzes the neurotransmitter acetylcholine; death results from muscle and respiratory paralysis (Estes 1986; Fukuto 1990). Other than some experimental use to control oyster drills on the East coast in the late 1950s (WDF and WDOE 1992), the chemical traditionally has been used in agriculture and forestry management to control or suppress outbreaks of insect pests. Carbaryl was selected for burrowing shrimp control based on its efficacy, rapid hydrolysis, and low mammalian toxicity (Carpenter et al. 1961; Karinen et al. 1967; WDF 1970; Mount and Oehme 1981; Rajagopal et al. 1984; Larkin and Day 1985; Cranmer 1986) and has been used in Washington State's coastal estuaries since 1963. Since its inception, the carbaryl program has been a focal point of contention among diverse economic, recreational, and environmental interests in Willapa Bay and Grays Harbor. Although carbaryl is effective at controlling shrimp populations, its use in aquatic environments is controversial due to its toxicity to non-target species, including juvenile Dungeness crabs that use live oysters as refuge and benthic invertebrates that form an integral part of estuarine food webs (Buchanan et al. 1985). Crab fishers, concerned that carbaryl impacts their industry by killing 0+ (settlement to age 1), and subadult (1 + and 2 +) crabs, have generally opposed use of the pesticide. Some environmentalists have sought to ban the use of carbaryl due to short-term and potential long-term impacts to the estuarine ecosystem. Similar concerns in Oregon led to the termination of their carbaryl program in 1984 (Bakalian 1985; Buchanan et al. 1985). Without carbaryl, oyster growers maintain that production would decline considerably and many grounds would have to be abandoned. Government natural resource agencies, charged with both managing fisheries and protecting the health and quality of the estuaries, require data that allow them to weigh the ecological costs and benefits of oyster and burrowing shrimp beds, quantify the relationship between shrimp density and oyster losses, assess impacts of carbaryl on estuarine fauna and trophic dynamics, and determine the recovery time and resilience of estuarine ecosystems to chemical applications. 143 This article provides an overview of issues pertaining to oyster culture and use of carbaryl to control burrowing shrimp populations in Washington State, incorporating published research and unpublished experimental data collected by the authors. We review what is known about shrimp life histories, trophic linkages, and distribution patterns since neither oyster culture techniques nor shrimp control policies have integrated much of this information into management practices. Our review includes the following sections: 1) physical and ecological characteristics of Willapa Bay and Grays Harbor relevant to oyster culture and burrowing shrimp distributions; 2) historical development of oyster culture and various cultivation practices; 3) shrimp life history traits, population characteristics, trophic contribution, and interspecific and intraspecific interactions; 4) justification for shrimp control as it pertains to oyster aquaculture, state regulatory standards, chemical efficacy and persistence in sediments, and impacts on non-target species; 5) the dilemma of carbaryl use, contrasting the benefits provided by oyster culture against the ecological consequences of removing burrowing shrimp on a large spatial scale; and 6) recommendations for management alternatives. Given the need to remove burrowing shrimp to maintain oyster production levels, we examine how information on shrimp ecology such as timing of recruitment, variability in year class strength, and patterns of habitat use might be better integrated into chemical or alternative control measures. While this issue is restricted to a particular geographic region, it is intended to serve as an example of competing issues involving estuarine use and natural resource management. Apart from the program in Washington State, the only other current use of carbaryl to control burrowing shrimp occurs in Colombia, South America, where the callianassid Lepidopthalmus sinuensis infests penaeid shrimp aquaculture ponds at high densities, lowering productivity (Lemaitre and Rodrigues 1991; Nates and Felder 1998). While the use of carbaryl as a management tool is limited to these two regions, estuaries throughout the world face similar issues with respect to multiple-use conflicts that must be resolved through implementation of science-based regulatory policies and effective management practices, as well as communication and cooperation among the parties involved. Willapa Bay and Grays Harbor PHYSICALCHARACTERISTICS Willapa Bay is located between the Columbia River to the south and Grays Harbor to the north and encompasses 260 km2 (31,970 ha) at mean 144 K. L. Feldmanet al. Fig. 2. Grays Harbor estuary, Washington. Commercial aquaculture of Pacific oysters (Crassostreagigas) is limited to 3% of the intertidal flats, or 364 ha, and is located in the North Bay and South Bay regions of the estuary. Stippling indicates intertidal habitats at mean low water. I Fig. 1. Willapa Bay, Washington. Arrows point to long-term monitoring sites for ghost shrimp (Neotrypaeacaliforniensis)populations (Palix River, Goose Pt., Nahcotta stations 1 and 2) and mud shrimp (Upogebia pugettensis) populations (Cedar River). Stippling indicates intertidal habitats at mean low water. high water. The estuary is oriented north-south and is protected from the Pacific Ocean by Long Beach Peninsula (Fig. 1). Nine rivers and several sloughs empty into Willapa Bay, draining a watershed of approximately 1,865 km2 (Hedgpeth and Obrebski 1981). Tides are semidiurnal and range up to 4 m. Much of the estuary is shallow (less than 15% of the estuary is deeper than 7 m), and over half of the surface area is exposed at low tide (Hedgpeth and Obrebski 1981). Broad expansive tideflats are drained by intertidal creeks that connect to deeper subtidal channels. Sediments consist of mixed sand and mud with sandy sediments prevalent near the mouth and central region of the estuary. Sand-mud composites are prevalent across the entrance towards the east and in the mid section, and mud is prevalent in the southern portion of the estuary (Clifton 1983). Salinities range from 7%o to over 30%o depending on location and season, with lower salinities occurring toward river mouths and from October through March when precipitation and river discharge rates are highest. Water temperatures (3-21?C) vary with location and season, with higher temperatures towards the shallow headwaters and in the summer months. Grays Harbor is oriented east-west and bounded by Point Chehalis to the south and Point Brown to the north (Fig. 2). Over 60% of the total surface area of the estuary (235 km2 or 22,137 ha at mean higher high water) is intertidal and exposed by semidiurnal tides that range up to 4 m (Loehr and Collias 1981). Six rivers flow into the estuary with a total drainage area of 6,204 km2 (Simenstad et al. 1992). During summer, salinities range from 5%o near the head of the estuary to over 30%o towards the entrance, whereas in the winter and spring, salinities range from 0.5-33%o (Loehr and Collias 1981). During summer, water temperatures average 19?C in the inner harbor and 15?C in the central section and near the mouth, but drop to a range of 2-8?C in the winter (Loehr and Collias 1981). Storms accompanied by gale and occasionally hurricane force winds hit the Washington coast during the fall and winter affecting sediment transport and water properties (e.g., salinity, temperature, and turbidity) in both Grays Harbor and Willapa Bay. ECOLOGICAL CHARACTERISTICS Willapa Bay and Grays Harbor share ecological similarities with respect to habitats and estuarine biota that are relevant to issues presented in this review. Eelgrass beds (Zosteramarina and Zosterajaponica) are common on intertidal and shallow subtidal mudflats in both estuaries, occurring between 1 m below to 2 m above mean lower low water 145 Oysters,Crabs,and BurrowingShrimp (MLLW). Eelgrass (particularly Z. marina) provides refuge and foraging opportunities for numerous benthic invertebrates, fishes, and waterfowl (Phillips 1984; Wyllie-Echeverria et al. 1994), spawning habitat for Pacific herring (Clupea pallasii), and is an important component of detritus-based food webs (e.g., Harlan and Throne-Miller 1981; Simenstad 1994). Epibenthic shell is also an important structural component of these estuaries, supporting higher densities of amphipods (Eogammarus oclairi), harpacticoid copepods (Dactylopodiacrassipes), cumaceans (Cumella vulgaris), crabs (Cancer magisterand Hemigrapsus oregonensis),and gunnels (Pholis ornata) than bare mudflats dominated by burrowing shrimp (Doty et al. 1990; Armstrong et al. 1992; Williams 1994; Eggleston and Armstrong 1995). Shell habitats consist of commercial oyster culture in both estuaries as well as relic surface deposits (death assemblages) of eastern softshell clam Mya arenaria (Palacios et al. 1994) and intertidal oyster shell habitat mitigation for juvenile Dungeness crabs constructed by the U.S. Army Corps of Engineers (COE) in Grays Harbor (Armstrong et al. 1992; Fernandez et al. 1993; Eggleston and Armstrong 1995; Iribarne et al. 1995). Ghost and mud shrimp are often a dominant feature of intertidal mudflats as well, with burrow densities approaching 600 holes m-2 in some areas (Armstrong et al. 1992). These shrimp comprise an additional estuarine habitat by virtue of their numerical abundance and ability to greatly influence benthic species composition (Peterson 1977; Brenchley 1981; Bird 1982; Posey 1985, 1986a; Posey et al. 1991; Dumbauld 1994; Tamaki 1994). Oyster Culture HISTORY AND DEVELOPMENT Prior to European settlements, the Olympia oyster, Ostrea lurida, was harvested historically by native Americans living along Willapa Bay and Puget Sound, Washington, and was a common food item found in ancient shell middens (Baker 1995). Natural beds occurred subtidally and intertidally as reefs comprised of several age classes. The commercial industry in Willapa Bay began in the mid1800s by harvesting these reefs, with the first shipments of 0. lurida arriving in San Francisco in 1851 (Conway 1991). Throughout the latter half of the nineteenth century, oysters were in high demand in the United States and considered a luxury item (MacKenzie 1996). Production in Washington increased until the completion of the transcontinental railroad which enabled East coast companies to expand their markets to California. As demand for Atlantic state oysters declined in the early 1900s due to contamination outbreaks, Washington once % A E ^ - Willapa Bay 1000- Year Q. , ', r 1935 1945 , i 1955 , 1965 . ' 1975 , . 1985 Year Fig. 3. Pacific oyster (Crassostreagigas) production for Grays Harbor and Willapa Bay, Washington, from 1935-1993. (Source: data from Hoines [1996]). again became a major supplier of oysters (Conway 1991). By this time, however, natural beds were quickly being depleted due to overharvesting and failure to replenish mudflats with shucked oyster shells to provide substrate for larval settlement. Burrowing shrimp were becoming a concern to harvesters practicing bottom culture and to those using dikes to hold transplanted oysters higher in the intertidal. Shrimp burrows extended deeper than the cement or wooden barriers, thereby draining ponds and exposing the temperature-sensitive oysters to freezing winter conditions (Stevens 1928). As a result of declining natural stocks, the eastern oyster, Crassostreavirginica, was imported to Washington around the turn of the century, but efforts to culture the species were unsuccessful. Then in the 1920s, the Pacific oyster, Crassostrea gigas, was introduced from Japan and since that time has been the basis of the industry. Production in Willapa Bay increased through the 1930s and 1940s, peaking at 1,234,200 gal of oyster meat in 1946 (Fig. 3). Grays Harbor produced less than 20,000 gal until the late 1940s when production began to increase, ranging from approximately 50,000 to 100,000 gal (Fig. 3). Production declined in Willapa Bay from the 1940s to 1970s due in part to market fluctuations and was associated with anecdotal accounts of burrowing shrimp expansions (Shotwell 1977), but has since rebounded as overharvesting, pollution, and disease have injured East and Gulf coast fisheries. From 1984-1993, an average of 426,000 gal was produced in Willapa Bay while an average of 114,000 gal was produced in Grays Harbor. Although oysters are sold locally, the vast majority are shipped out of state to supply domestic and foreign markets. Recent estimates place Washington's oyster industry at $25-27 mil- 146 K. L. Feldmanet al. lion and responsible for the direct and indirect employment of over 1,700 workers in the region (Conway 1991; Carkner and Harbell 1992; K. Chew personal communication). Since most designated oyster ground in Willapa Bay and Grays Harbor is not cultivated, Conway (1991) estimated that there is potential to expand production in these estuaries through development of beds that have remained fallow due to market fluctuations, relatively poor growing conditions, or high densities of burrowing shrimp. CULTIVATION PRACTICES Of the 17,200 ha of intertidal land in Willapa Bay, 10,522 ha are privately owned or leased for commercial oyster culture. Only 3,642 ha, or 35% of classified oyster grounds, are intensively cultured and capable of remaining economically profitable while producing market quality oysters (Burrowing Shrimp Committee 1992). The State of Washington operates an oyster reserve program on another 607 ha of tidelands that is managed by the Washington Department of Fish and Wildlife (WDFW; formerly Washington Department of Fisheries [WDF]) to supply industry with seed stock and 2-3 yr old oysters for transplantation. Oyster beds are located throughout the estuary, although the most productive beds are concentrated near the entrance. Several factors influence the quality and productivity of an oyster bed, such as substrate type and compaction, tidal height, phytoplankton supply, salinity, temperature, turbidity, and the density of predators and pests (Shotwell 1977; Peterson and Black 1987, 1991; Hofmann et al. 1992; Dumbauld et al. 1997). Although approximately 21% of intertidal ground in Willapa Bay is commercially farmed and managed for oysters, only 364 ha or 3% of the intertidal is farmed in Grays Harbor. Pacific oysters are grown using a variety of methods, each tailored to the environmental conditions of the physical location as well as to the particular retail market (Conte et al. 1996). Methods used in Willapa Bay and Grays Harbor include stake, longline, rack and bag, and bottom culture. Stake, longline, and rack and bag techniques involve suspending oysters off the benthos and are used principally to generate oysters for the single oyster or half-shell markets. Although stake culture may extend subtidally, suspended culture methods are generally restricted from 0.0 m to 0.6 m above MLLW due to the manual labor required to plant, maintain, and harvest the oysters (see Nosho 1989 for a complete description of suspended culture methods). Although growers utilize suspended culture techniques, bottom or ground culture is the most extensive method used in coastal estuaries, accounting for over 95% of production in Willapa Bay (Dumbauld 1994). Oysters are planted and grown from 0.5 m below to 1.1 m above MLLW directly on the sediment in a single layer, then harvested primarily for the fresh shucked oyster market. Bottom culture is more economical than other methods of growing oysters (WDF and WDOE 1992), in part because much of the culture cycle can be mechanized thereby reducing labor costs and increasing the scale of operation. Seed (also referred to as cultch: non-living oyster shell covered with spat Uuvenile oysters -2-3 mm length]) is produced by hatcheries or collected naturally by placing oyster shells on the mudflats or in plastic mesh bags before spawning occurs in the summer. Growers may plant seed in the fall, but they typically wait until spring (March through June) after the winter storm season. There are three common bed types that determine rotational cycles in ground culture: seed beds, fattening beds, and seed-growout beds. Seed beds are generally higher in the intertidal or further within the estuary where growth rates are slower. These areas are better protected from tidal currents and storms and are characterized by lower turbidity, siltation, and predator abundance than more exposed sites such as fattening beds. In this scenario, cultch typically is planted in the spring of year one to take advantage of phytoplankton blooms and then transplanted to fattening beds in the summer or fall of year three for a final year of growth before harvest and processing (Fig. 4). Fattening beds occur closer to the mouth of the bay and lower in the intertidal where growth is highest. Turnover is high since 3-yr old oysters are transplanted and harvested approximately every two years (Fig. 4). Since fattening beds are a limited resource comprising only 16% to 19% of cultivated ground, seed is rarely planted and left to grow in these areas. In other parts of the estuary, seed is left on the same bed over the entire cycle (seedgrowout bed; Fig. 4). This method is less common because seed and fattening beds are a more efficient use of intertidal land: seed distributed over 3 ha of ground can be transplanted onto 1 ha of fattening bed, thus maximizing production per hectare. The length of the average rotational cycle ranges from two to four years from planting to harvest, but individual cycles vary depending on oyster growth and market conditions. Oysters are dredged mechanically or picked by hand with the majority of harvest occurring from October through March. Bottom culture produces higher yields compared to other methods, averaging 1,540 gal ha- (Dumbauld 1994). Oysters,Crabs,and BurrowingShrimp Year2 Year1 I Seed 4A 44 Bed -1 Seed - ? .; ? .:.;:.;:-:::-:.::.:.:.:: J I I JJASOND 4A A : I I . U .. ,:.;.:.:,I ,~???~?, Year4 FMAM JJ ASOND 'J FMAM JJ ASONDIJ AA - : 1i !:i: Fattening Bed -z =_ r :::::::::::::::::::::::::::::o - S::!: ::!-: Growout Carbaryl Treatment Year3 I J FMAM JJ ASOND'JFMAM I -w Mud Shrimp 147 I I I I I I Ghnost V I I Shrimp Plant Harvest Oysters Present Larval Recruitment Fig. 4. Typical 4-yr oyster culture cycle used in Pacific Northwest estuaries. Seed oysters generally are planted on a seed bed in the spring and left to grow for 2 yr (top bar) before transplantation to a fattening bed (middle bar) where they complete another year of growth prior to harvest in the fall or early spring. Seed may also be planted and left to grow for the entire 3-yr period on the same bed (seed-growout bed, bottom bar). Also shown are the typical carbaryl treatment periods (July-August) and recruitment periods for mud shrimp (Upogebiapugettensis;April-June)- and ghost shrimp (Neotrypaeacaliforniensis;August-October), drawn with vertical arrows to intersect annual periods of oyster culture. (Source: modified from Dumbauld et al. 1996). Burrowing Shrimp Ecology To effectively manage densities of a pest population, it is important to understand the life cycle and ecological traits of the target species and to integrate this information into control policies. Under the carbaryl program, mud shrimp and ghost shrimp are treated as a single entity although each species has distinct life-history characteristics and behaviors. Although both species are infaunal burrowers, they differ in a number of life-history characteristics which include but are not limited to timing of reproduction and recruitment of postlarvae to the estuary and feeding strategy (for a more complete review of mud and ghost shrimp life histories see Table 1; Bird [1982]; Dumbauld et al. [1996]). In this section we review these life history traits in particular because shrimp recruitment bears directly on carbaryl policy with respect to the seasonal application of the pesticide, while feeding strategy bears on realized impacts to oysters via sediment disturbance (bioturbation) or potential impacts via food competition. We also examine data on shrimp densities and mortality rates to illustrate how quickly oyster beds can be colonized by shrimp and meet the minimum threshold for spraying carbaryl (10 burrow holes m-2) and to provide data that eventually may be used to help predict population densities a year or two into the future or at least help understand interannual fluctuations in abundance. Trophic linkages and interspecific and intraspecific interactions are discussed to highlight the importance of shrimp as prey and the role shrimp play in influencing benthic community composition, as well as features of particular habitats that might reduce population densities. We focus specifically on shrimp ecology in this and subsequent burrowing shrimp subsections and return to the relevance of this information to carbaryl practices, oyster damage, and ecosystem functions later. LIFE HISTORYTRAITS Reproductionand Postlarval Recruitment Periods of ovigery and postlarval settlement differ between mud shrimp and ghost shrimp (Bird 1982; Dumbauld et al. 1996). Mud shrimp are ovigerous from October through May. Eggs brooded on the female's pleopods begin to hatch in February, and the zoeae are exported out of the estuary to the nearshore coastal ocean, typically during night ebb tides (Dumbauld and Feldman unpublished data). Zoeae spend two to three weeks in the plankton and progress through three zoeal stages (Hart 1937) before molting into postlarvae, which are capable of settling and assuming a benthic lifestyle. Recruitment of postlarvae to Washington estuaries generally occurs from late April through June (Dumbauld et al. 1996). Ghost shrimp are ovigerous from April through August. The eggs begin to hatch in June and, similar to mud shrimp, zoeae are released primarily during the night ebbs of neap tide series and exported to nearshore coastal waters (Johnson and Gonor 1982; Pimentel 1983). Ghost shrimp progress K. L. Feldmanet al. 148 c) 0 - S . 0C . 0 cn c0 . , l l cr. ct . Eo ~ ~iu 1 + ghost shrimp. Over the 10-yr period, densities of > 1 + ghost shrimp were variable but rose from 228 shrimp m-2 in 1988 to 333 shrimp m-2 in 1997 (a mean change in abundance of approximately 12 shrimp m-2 yr- 1; Fig. 5). Densities of > 0+ mud shrimp were lower than those of ghost shrimp during the sampling period but also increased: from 54 shrimp m-2 in 1989 to 114 shrimp m-2 in 1997 (a mean change in abundance of approximately 8 shrimp m-2 yr-1; Fig. 5). Recruitment of 0+ shrimp may explain some of the fluctuation in the entire population. Interannual recruitment densities are highly variable, fluctuating by as much as two orders of magnitude, and are likely influenced by oceanic processes as are other decapod crustaceans with complex life cycles, such as Dungeness crabs (McConnaughey et al. 1992). We used data obtained from mud shrimp density cores at the Cedar River site as described in the preceding section to estimate the density of 0+ mud shrimp from 1989 to 1997. Visual analyses of length frequency histograms and 150 K. L. Feldmanet al. -- - . Mud shrimp Ghost shrimp..... E Q E 100- en + E - 50- 0O ,- 89 91 90 92 93 94 95 97 96 200 B. Ghost shrimp 88 90 92 94 * 96 Year Fig. 5. Mean density (? 1 SE) of ghost shrimp (Neotrypaea californiensis) at the Palix River station and mud shrimp (Upogebia pugettensis) at Cedar River station in Willapa Bay from 1988-1997. Samples (n = 4-12 benthic cores, 0.13 m2 X 60 cm depth, 3.2-mm mesh sieve) were taken during a single month in late summer or early fall and therefore include 0+ mud shrimp, but exclude newly settled 0+ ghost shrimp. (Source: modified from Dumbauld et al. 1996). published data on growth rates (Dumbauld et al. 1996) were used to distinguish 0+ shrimp from older age classes. Over the 9-yr period, densities of 0+ mud shrimp ranged from a mean low of 2 shrimp m-2 in 1989 to a mean high of 98 shrimp m-2 in 1995 (Fig. 6a), and fluctuations in 0+ densities appeared to mirror interannual fluctuations in the entire population (Fig. 5). To quantify 0+ ghost shrimp abundance, we used a 26 cm diam by 15 cm depth core and sieved the excavated material on a 0.5-mm mesh screen. We found that 0+ ghost shrimp densities varied considerably over time as well. For example, at the Palix River site, 0+ mean densities ranged from 144 shrimp m-2 in 1993, 22 shrimp m-2 in 1994, 0 shrimp m-2 in 1995 and 1996, to 6 shrimp m-2 in 1997 (Fig. 6b). Low recruitment of 0+ ghost shrimp in fall of 1995 and 1996 likely contributed to the reduction in - 1+ densities observed in 1996 and 1997 (Fig. 5). Densities of 0+ ghost shrimp also appeared to vary with respect to location within the estuary, although densities at different locations exhibited similar interannual fluctuations (Fig. 6b). In years with poor recruitment, such as 1995 and 1996, 0+ densities were low at all locations sampled. Estimates of Natural Mortality There is no published information to our knowledge on estimates of natural mortality for mud cx ~~~T 150- 15~3Palix River E~~~~~~~rl NahcottaStn 1 ED_ . s 2 100- + - Goose Point Nahcotta Stn 2 /" / i 50 , , 92 93 94 95 96 97 Year Fig. 6. (A) Mean density (+ 1 SE) of 0+ mud shrimp (Upogebia pugettensis) at the Cedar River station from 1989-1997 (n = 4-10 cores). Densities were calculated using MIX (refer to text for description of program) and estimated growth rates to separate 0+ shrimp from >1 + shrimp from benthic core samples (0.13 m2 X 60 cm depth). (B) Mean density (+ 1 SE) of 0+ ghost shrimp (Neotrypaeacaliforniensis) at four locations in Willapa Bay (see Fig. 1 for sampling locations) from 1992-1997 (0.05 m2 X 15 cm depth core, 0.5-mm mesh sieve; n = 3-12 cores). The absence of a bar indicates that no samples were taken, whereas the presence of a "0" indicates samples were taken, but no 0+ ghost shrimp were found at those locations. shrimp or ghost shrimp, nor for any other thalassinidean species. We used data from a long-term (3 yr) experiment conducted by Dumbauld et al. (1997) along the Palix River channel in Willapa Bay (Fig. 1) to derive estimates of mortality for ghost shrimp. Intertidal plots (10 m on a side, 100 m2) were treated with 5.6 kg ha-' of carbaryl in summer of 1989 to remove shrimp and serve as a starting point for 0+ recolonization in 1989 and thereafter (since carbaryl degradation is rapid, subsequent recruitment of 0+ ghost shrimp would not have been affected by any residual toxicity). By eliminating all age classes from the plots, it was Oysters,Crabs,and BurrowingShrimp Females Males 12 10 8 6 4 2 0 -. 0 > C 8 - : :- 6- ( I 1990 8 6 4 2 0 5 10 15 1991 - 12 10 8 6 4 2 0 0 5 10 15 1 ~ 0 0 5 5 10 10 15 15 5 10 15 20 0 5 10 15 20 5 5 10 10 15 15 20 20 12 10 20 1992 ' 0 8 6 4 2 0 .. 2 2 i 20 4 - LL 12 10 1210- () 151 ~8 20 20 12 10 - 0 0 CarapaceLength(mm) Fig. 7. Length frequency histograms of male and female ghost shrimp (Neotrypaea californiensis) from 1990 through 1992 that settled or immigrated into experimental plots (n = 4) which were sprayed with carbaryl in 1989 to remove pre-existing shrimp. Length frequency histograms were analyzed using MIX to identify modes and corresponding age classes and to estimate mortality of the 1989 year class which settled into plots 1-3 mo after carbaryl treatment. easier to identify the 1989 settling year class and track its abundance through time. Every August (1 yr, 2 yr, and 3 yr post-spray), a core sample (40 cm diam X 60 cm depth) was taken in the center of each plot (n = 4). Sediment was rinsed through a 3.2-mm mesh screen and sorted for shrimp which were measured and sexed. We estimated mortality of the 1989 year class beginning in 1990 at age 1 since 0+ shrimp settling immediately back into treated plots were not adequately sampled using a 3.2-mm mesh screen. Separate estimates were derived for males and females given skewed adult sex ratios (Dumbauld et al. 1996) which may be an indication of differential mortalities. Age classes and cohorts were identified by fitting lognormal components to length frequency histograms (Fig. 7) using MIX (Release 2.3, Icthus Data Systems; MacDonald and Pitcher 1979). In estimating proportions of shrimp in each age class we assumed that badly damaged unmeasurable and unsexable shrimp were equally represented among ages and between sexes. Because the MIX program often converged on more than one possible set of components, information on shrimp growth rates and recruitment of multiple cohorts (Bird 1982; Dumbauld et al. 1996) was used to choose the best scenario. Once numbers of shrimp were determined by age class, an estimate of natural mortality was derived based on simple linear regression of natural log-transformed abundances of the 1989 year class from 1990 to 1992 (Table 2). Mortality was also calculated between years based on the exponential model dN/dt = -ZN. Given the movement of older age classes into treated plots (e.g., the presence of 3+ shrimp in 1991, see Tables 3 and 4), it is likely that similar movements of individuals belonging to the 1989 year class into and out of plots biased our estimates. The derived estimates reflect the net sum of natural mortality plus emigration minus immigration, but for convenience we refer to estimates in this exercise as mortality since we were unable to assess the individual importance of these processes and their impact on calculations. 152 K. L. Feldmanet al. TABLE 2. Summary of natural mortality estimates for male and female ghost shrimp (Neotrypaeacaliforniensis)based on data collected from a field experiment conducted by Dumbauld et al. (1997). Between years, Z was calculated based on the exponential model dN/ dt = -ZN. An overall estimate of Z from ages 1 to 3 was calculated based on simple linear regression of natural log-transformed abundances over the period 1990-1992. Sex Male Female Number of Shrimp Year 1990 1991 1992 Overall Z 1990 1991 1992 Overall Z % Annual Survival 100 31 6 Z 31 19 182 39 113 21 290 Estimates of Z for males belonging to the 1989 year class are 1.17 from 1990 to 1991, 1.64 from 1991 to 1992, and 1.41 for the 2-yr period from 1990 to 1992 (Table 2). Three modes (cohorts) of 1+ males (4.1, 5.4, and 7.5 mm CL) totaling 100 individuals were identified in 1990 using MIX (Table 3; Fig. 7). In 1991, two modes of 2+ shrimp (8.7 and 11.2 mm CL) totaling 31 individuals were apparent, and in 1992, one mode of 3+ shrimp (15.6 mm CL) totaling six individuals was identified. Correct assignment of age classes to mean carapace length components is critical in estimating mortality. If, for example, the 7.5 mm CL component in 1990 actually represented 2+ shrimp (not 1 + shrimp) then the mortality estimate would SE r2 1.17 1.64 1.41 0.14 99 1.54 -1.06 0.24 0.75 9 (Dumbauld et al. 1996), and the potential for faster growth as a result of reduced competition (due to removal of older age classes), we suggest that the 7.5 mm CL component was most likely comprised of 1 + shrimp. Estimates of Z for females belonging to the 1989 year class are 1.54 from 1990 to 1991, -1.06 from 1991 to 1992, and 0.24 for the 2-yr period from 1990 to 1992 (Table 2). The slope of the regression curve over 1990-1992 was lower for females than males, but the statistical fit was much poorer (Table 2). Two modes (cohorts) of 1+ females (5.1 and 6.8 mm CL) totaling 182 individuals were identified in 1990 using MIX (Table 4; Fig. 7). In 1991, one mode of 2+ females (9.3 mm CL) totaling 39 individuals was apparent, and in 1992, one mode of 3+ shrimp (10.8 mm CL) totaling 113 individuals was identified. It is unclear whether the low number of 2+ shrimp in 1991 was the result of sampling artifacts or whether the net effect of mortality, emigration, and immigration accounted for the progression of 182 shrimp in 1990 to 39 be much lower [51 (1+) shrimp -> 31 (2+) shrimp -- 6 (3+) shrimp instead of 100 -> 31 -- 6]. It is possible that some of the males in the 7.5 mm mean CL component were actually 2+ individuals, but the MIX program was unable to converge on another modal peak. Given the available data, information on sex-specific growth rates TABLE 3. Results of MIX analyses on male ghost shrimp (Neotrypaea californiensis). Experimental plots sprayed with carbaryl at a concentration of 5.6 kg ha-' in August 1989 were sampled again August 1990, 1991, and 1992 for abundance of the 1989 settling year class. The 1989 year class was age 1 + in 1990, 2+ in 1991, and 3+ in 1992. Age classes were determined by fitting lognormal components to length-frequency histograms. Mean carapace lengths (CL), proportions, and numbers of males are given for each age class. Total numbers of male shrimp (calculated as the sum total of males in 4 core samples) were multiplied by the proportion of shrimp in each age class to calculate the number of males in each age class. An estimate of mortality was derived based on changes in the numbers of males comprising the 1989 year class from 1990 to 1992. Note the presence of three 1+ cohorts (modes) in 1990 and two 2+ cohorts in 1991 and 1992. Note also the presence of older (3+) males in 1991 that most likely immigrated into plots. If, instead, they were shrimp that had survived carbaryl treatment in 1989, they likely would have been present in samples as 2+ shrimp in 1990. Data from Dumbauld et al. (1997). Year Variables 1990 Mean CL (mm) Proportion of males Number of males in Mean CL (mm) Proportion of males Number of males in Mean CL (mm) Proportion of males Number of males in 1991 1992 Total Number of Male Shrimp in age class age class 100 in age class age class 63 in age class age class 73 Age Classes 1+ 1+ 1+ 4.1 0.12 12 4.5 0.07 4 5.4 0.38 28 5.4 0.39 39 6.4 0.21 13 7.5 0.49 49 2+ 2+ 3+ 8.7 0.25 16 8.0 0.23 17 11.2 0.24 15 11.1 0.31 22 14.1 0.23 15 15.6 0.08 6 Oysters,Crabs,and BurrowingShrimp 153 TABLE 4. Results of MIX analyses on female ghost shrimp (Neotrypaeacaliforniensis). Experimental plots sprayed with carbaryl at a concentration of 5.6 kg ha-~ in August 1989 were sampled again August 1990, 1991, and 1992 for abundance of the 1989 settling year class. The 1989 year class was age 1+ in 1990, 2+ in 1991, and 3+ in 1992. Age classes were determined by fitting lognormal components to length-frequency histograms. Mean carapace lengths (CL), proportions, and numbers of females are given for each age class. Total numbers of female shrimp (calculated as the sum total of females in 4 core samples) were multiplied by the proportion of shrimp in each age class to calculate the number of females in each age class. An estimate of mortality was derived based on changes in the numbers of females comprising the 1989 year class from 1990 to 1992. Note the presence of two 1 + cohorts (modes) in 1990, and the presence of older (3+) females in 1991 that most likely immigrated into plots. Data from Dumbauld et al. (1997). Year 1990 1991 1992 Total Number of Female Shrimp Variables Mean CL (mm) Proportion of females Number of females in Mean CL (mm) Proportion of females Number of females in Mean CL (mm) Proportion of females Number of females in in age class age class in age class age class in age class age class 182 116 182 in 1991 to 113 in 1992. The presence of older (3+) females in 1991 (11.8 mm CL; Table 4; Fig. 7) is suggestive of movement of shrimp into plots. If these older individuals had been present in plots in 1989 and survived carbaryl treatment, then some of these females should have shown up in length frequency histograms as 2+ shrimp in 1990. Based on the Z values calculated above, the abundance of male ghost shrimp within a given year class might decline 75% yr-~ on average and females 22% yr-1 on average. These values should be considered a rough approximation given that they were derived from a single experiment with low sample size. In addition, processes that regulate shrimp population structure, such as growth rates and movement patterns, obscured trends in mortality. Growth rates differ between sexes and cohorts (Dumbauld et al. 1996) and vary with conspecific density and food supply (Bird 1982), which make age determinations difficult, and there is some evidence of movement by shrimp (Posey 1985; Feldman et al. 1997) either through active choice or bedload transport. Finally, mortality rates may differ for mud shrimp populations based on differences in life history and behavior between the two species. We focused on ghost shrimp in this exercise because they pose a greater overall threat to oyster production, and the data set was more conducive to analysis than that from a similar experiment conducted by Dumbauld et al. (1997) on mud shrimp. TROPHICCONTRIBUTION Mud shrimp and ghost shrimp are prey for a number of species and, as such, are an important link in estuarine trophic pathways. Although shrimp inhabit burrows that extend up to 90 cm into the sediment, they are vulnerable to predators Age Classes 1+ 1+ 5.1 0.30 55 5.5 0.16 19 4.6 0.09 16 6.8 0.70 127 6.0 0.16 29 2+ 3+ 9.3 0.34 39 8.6 0.13 24 11.8 0.50 58 10.8 0.62 113 because shrimp often are situated near the sediment-water interface. Posey (1985) observed that ghost shrimp spent over 25% of the time within 2 cm of the entrance to their burrows, often with part of a chela lying exposed on the sediment surface. Occasionally, shrimp also have been observed crawling on the surface at low tide as well as at high tide (Posey 1985), exposing them to predation by epibenthic feeders. Effects of infaunal predators on prey density and distribution have received less attention than those of epibenthic predators, yet infaunal predators have been shown to structure some soft-sediment communities (Ambrose 1991) and regulate benthic recruitment (Desroy et al. 1998). One of the most common predators of burrowing shrimp in the Pacific Northwest is the staghorn sculpin, Leptocottusarmatus. Staghorn sculpin are opportunistic, generalist predators common in estuaries along the West coast of North America (Hart 1974). Burrowing shrimp have been found in the stomachs of sculpin (Tasto 1975; Williams 1994) and comprise a substantial portion of their diet, particularly during summer months (Posey 1986b; Armstrong et al. 1995). Armstrong et al. (1995) found that ghost shrimp and mud shrimp comprised 70% of the diet (expressed as index of relative importance [IRI]) of 1 + and older staghorn sculpin in July and 54% in August. Staghorn sculpin have also been shown to restrict the lower distribution of ghost shrimp. Posey (1986b) found that sculpins were four times more abundant immediately seaward of a dense shrimp bed than over it, and when they were excluded from the lower intertidal, ghost shrimp successfully migrated down and survived. On a mudflat in Boundary Bay, British Columbia, Swinbanks and Murray (1981) similarly noted that ghost shrimp densities were 154 K. L. Feldmanet al. highest in the mid-intertidal region and declined towards the lower intertidal and subtidal, perhaps due to predation. Other predators of burrowing shrimp identified by Posey (1985, 1986b) include cutthroat trout, Salmo clarkii, Dungeness crabs, Cancer magister,and Western gulls, Larus occidentalis. Although cutthroat trout preyed on ghost shrimp, they were rarely captured in beach seines over the 2-yr survey period, and thus were not likely to contribute substantially to shrimp population regulation in his study. Adult and juvenile Dungeness crabs were observed foraging over ghost shrimp beds, and in bare aquaria, attacked and consumed shrimp. In laboratory experiments, Feldman et al. (1997) found that 0+ crabs preyed on 0+ ghost shrimp buried in an epibenthic shell-mud substrate. Shrimp remains have also been reported in the stomachs of adult and subadult Dungeness crabs (Stevens et al. 1982). Posey (1985) noted the presence of ghost shrimp in the fecal pellets of Western gulls, suggesting that they occasionally prey on shrimp as well. Although staghorn sculpin was the only predator of significance in Posey's (1986b) study, other predators, including those listed above, may play a greater role in shrimp population regulation in different habitats and estuaries. Starry flounder, Platichthysstellatus, are common in Coos Bay, Oregon, and they, as well other species of fish, may prey on shrimp subtidally (Posey 1985). Leopard sharks, Triakis semifasciata, common along the coasts of California and Oregon, have been reported to prey on ghost shrimp and mud shrimp (Russo 1975), and there are anecdotal accounts of green and white sturgeon (Acipenser medirostrisand Acipenser transmontanus) predation on burrowing shrimp in the Pacific Northwest. Larvae of burrowing shrimp may also be a seasonal prey item for water column feeding fishes. Ghost shrimp larvae were found in the stomachs of Pacific herring (Clupea pallasii), chinook salmon (Oncorhynchus tshawytscha), and chum salmon (0. keta) captured in Tillamook Bay, Oregon, in June and July, and mud shrimp larvae were found in the stomachs of chinook salmon (Oregon Department of Fish and Wildlife 1977). Even the gray whale, Eschrichtius robustus, has been observed to feed extensively on ghost shrimp. Gray whales migrate between calving lagoons in Mexico and Arctic feeding grounds (Rice and Wolman 1971), but some individuals enter Pacific Northwest estuaries to feed during the summer (Darling 1984). Weitcamp et al. (1992) documented between 2,700 and 3,300 feeding pits (depressions 3-4 m2 and 10-15 cm deep) along 19 km of intertidal mudflats in Saratoga Passage, Puget Sound. They also reported seeing feeding pits in Willapa Bay. In Puget Sound, Washington, whales removed an average of 3 kg shrimp pit-~, and ghost shrimp standing stock was approximately five times lower inside feeding pits than outside pits (Weitcamp et al. 1992). Although feeding opportunities were restricted to high tides, they estimated that whales in north Puget Sound could meet their daily energetic requirements in 16-170 min by feeding exclusively on ghost shrimp. AND INTRASPECIFIC INTERSPECIFIC INTERACTIONS Thalassinidean shrimp have been the focus of several functional group hypotheses that seek to understand how benthic communities are structured by grouping organisms according to how they modify the sedimentary environment (see review by Posey 1990). Among the most common hypotheses are those that classify organisms based on effects of adults on settling larvae (Woodin 1976), trophic mode (Rhoads and Young 1970), and relative mobility (Brenchley 1981, 1982). The concept of assigning species to discrete functional groups is an appealing approach to understanding community interactions; however, organisms do not respond consistently to varying environmental conditions. Behavior, ontogeny, flow regime, and sediment transport all interact to create a unique set of conditions which affect benthic composition (Jumars and Nowell 1984). While adult-larval and trophic group amensalism hypotheses do not appear to be broadly applicable in describing soft-sediment communities (Mauer 1983; Black and Peterson 1988; Commito and Boncavage 1989; Hines et al. 1989; Snelgrove and Butman 1994), Brenchley's (1981, 1982) relative mobility mode hypothesis has been somewhat instrumental in identifying potential mechanisms by which species interact, particularly with respect to thalassinidean shrimp. Brenchley (1981, 1982) suggested that interactions between organisms of different mobilities would result in discrete distributions of species that tend to either destabilize or stabilize the sediment. Mobile infauna disturb the sediment by burrowing, depositing sediment grains on the surface, and resuspending particles in the water column which bury sedentary species, disturb infaunal tube-dwellers, and disrupt the roots and rhizomes of seagrasses. In turn, the tubes and roots of sedentary species bind the substrate inhibiting the burrowing activities of mobile organisms. Several studies have examined the effects of thalassinidean shrimp bioturbation on sedentary and mobile infaunal species. Dumbauld (1994) found that species richness and diversity were significantly lower in a ghost shrimp colony than in a mud shrimp colony. The ghost shrimp bed consisted Oysters,Crabs,and BurrowingShrimp primarily of mobile species whereas the mud shrimp bed consisted primarily of tube-dwelling and sedentary species. Ghost shrimp have been shown experimentally to have a negative effect on the abundance of sedentary infauna and a neutral or positive effect on mobile infauna (Posey 1986a). Tamaki (1988) found that Callianassajaponica had a positive effect on colonization by other mobile taxa, possibly by irrigating and fertilizing the sediment that stimulated the growth of microalgae and bacteria or by loosening up the sediment that eased burrowing and penetration. Negative effects of callianassid bioturbation also extend to seagrasses. Suchanek (1983) found that productivity and percent cover of the turtle grass Thalassia testudinum were negatively correlated with the density of Callianassa spp. mounds. Significant deterioration of T testudinumtransplants occurred in areas with high densities of Callianassa compared to control areas as a result of high turbidity or burial under sediment deposition. Bioturbation resulting from the grazing activities of epibenthic species such as sea urchins and cownose rays has also been shown to have negative impacts on seagrass beds (Ogden et al. 1973; Orth 1975; Valentine and Heck 1991). These interspecific interactions are not simply one-sided but rather can be reciprocal in nature. Structurally complex root-rhizome mats associated with seagrass beds have been shown to reduce the mobility of several burrowing species (Brenchley 1982) and to limit the distribution of burrowers to areas outside these habitats (Ringold 1979; Harrison 1987; Swinbanks and Luternauer 1987). Similar findings of reduced mobility have been reported in dense beds of tube-building polychaetes (Brenchley 1982) and phoronids (Ronan 1975). In both habitat types, the extent to which roots and tubes are capable of excluding burrowing organisms is a function of root or tube density, size and body morphology of the burrower, and degree of mobility (Brenchley 1982). Brenchley (1982) found that mean burial time increased significantly for ghost shrimp in root-rhizome and animal tube mats compared to pre-burrowed bare sediments, and in the majority of laboratory trials shrimp were unable to establish a burrow at all. Field surveys have been consistent with Brenchley's (1982) findings, noting the abrupt decline and low densities of ghost shrimp burrows in Zostera marina beds compared to adjacent intertidal mudflats (Swinbanks and Murray 1981; Swinbanks and Luternauer 1987). Harrison (1987) reported that an expansion of Z. marina and Zosterajaponica habitat was accompanied by a corresponding reduction in ghost shrimp density. He suggested that in temperate geographic regions, cycles of eelgrass and shrimp activity are sufficiently out of phase to en- 155 able the rhizomes of eelgrass to expand in early spring before shrimp become too active. Other species of callianassids appear to be restricted by root mats as well: Coleman and Poore (1980) noted a reduction in population densities of Callianassa australiensis and Callianassa limosa in areas where Zostera was present. In contrast to ghost shrimp, mud shrimp are better able to penetrate through root and tube mats (Brenchley 1982) and therefore may not be restricted from seagrass habitats. Dworschak (1987) noted that Upogebiapusilla burrows were more abundant in water-filled pools and Zosterapatches than in elevated and unvegetated areas. In Brenchley's (1982) study, burrowing time increased disproportionately with body size however, suggesting that populations may be skewed towards smaller mud shrimp in dense root and tube mats. Although mobility-mode interactions can account for many species distributions, support for this hypothesis is based principally on studies in which large-bodied taxa have had negative effects on small organisms (see Posey 1990 and references therein). The strength of the interaction between functional groups is density-dependent as well as size-dependent. It is difficult to categorize the effects of species as either stabilizing or destabilizing: sediment transport depends on interactions among flow, micro-topography, and species activities (Jumars and Nowell 1984). Animal tubes themselves are destabilizing biogenic structures that may induce sediment erosion at any density (Eckman et al. 1981); but tubes also enhance colonization by microbes which tend to bind and stabilize the substrate (Eckman 1985). Even burrowing shrimp have contrasting effects on sediment stability: the mucous-lined burrow of mud shrimp stabilizes the substrate but the expulsion of sediment during burrow construction results in destabilization. Posey et al. (1991) discovered that mud shrimp, like ghost shrimp, negatively affected the abundance of several sedentary species, which was unexpected given the relative differences in burrow characteristics and mobilities between the two species of shrimp. Finally, aggressive intraspecific and interspecific interactions may influence the abundance and distribution of burrowing shrimp and are important in helping to understand how populations are structured and maintained on oyster beds. Posey (1985) observed aggressive behaviors between ghost shrimp when burrow systems connected, until eventually one of the shrimp sealed off the intersection with sediment. Newly settled 0+ ghost shrimp also have been observed to seal points of intersection between burrow galleries (Tamaki et al. 1996 [Callianassa japonica]; Feldman personal 156 K. L. Feldmanet al. observation). Aggressive behavior has been observed in mud shrimp as well. When given the opportunity to construct burrows in sediment-filled aquaria, many individuals fought instead, tearing the limbs and chelae off conspecifics (Feldman personal observation). But ghost shrimp typically burrowed immediately into the sediment and showed little aggressive behavior towards others. Tunberg (1986) found that Upogebiadeltaura kept in aquaria without sediment also showed aggression and used their chelipeds to injure and kill each other. Damaged chelae and skewed sex ratios may further indicate that fighting is common in thalassinideans (Felder and Lovett 1989; Dumbauld et al. 1996). While aggressive intraspecific interactions may affect shrimp density, interspecific interactions may be involved in maintaining discrete populations of ghost shrimp and mud shrimp. Mud shrimp and ghost shrimp often occur together in mixed beds or in transition zones between single species beds, but tideflats with the highest densities of shrimp typically are populated by one species or the other. Virtually no studies have examined interspecific interactions, although Griffis (1988) observed aggressive behavior among Neotrypaea californiensis, Neotrypaeagigas, and Upogebia macginitieorum. Carbaryl JUSTIFICATION Thalassinidean shrimp have detrimental effects on ground culture of oysters, primarily through burial but also through feeding interference and perhaps competition as it applies to suspended culture of oysters. Bottom cultured oysters require sufficiently firm or compact substrate to grow and therefore may sink or be buried in unconsolidated sediments inhabited by burrowing shrimp. Although oysters of any age may sink or be buried, seed is particularly vulnerable on ghost shrimp beds since the small size and immobility of spat renders them unable to counteract rapid burial and suffocation. Shrimp activity and bioturbation increase in the spring and summer as water temperatures, salinities, and food concentrations rise (Posey 1987; Dumbauld et al. 1996). Rates of bioturbation have been shown to be species-specific: ghost shrimp sediment expulsion is much greater than that of mud shrimp (Swinbanks and Luternauer 1987; Dumbauld et al. 1997). The disparity in sedimentation rates most likely contributes to differing impacts on seed survival. Dumbauld et al. (1997) found that seed loss was more rapid and significantly greater on ghost shrimp beds than on mud shrimp beds. We documented a similar pattern in field experiments in which seed was plant- ---- Ghostshrimpbed ..... . Mudshrimpbed Cm E - 00 0 0. -o a) 0 Apr Jun Aug Oct Dec Feb Apr Jun Fig. 8. Mean density (? 1 SE) of oyster (Crassostreagigas) seed planted on a dense ghost shrimp (Neotrypaeacaliforniensis) bed (310 ghost shrimp m-2) and on a dense mud shrimp (Upogebia pugettensis) bed (150 mud shrimp m-2). Seed was planted in April at densities of 10-12 seed m-2 and within 2 mo dropped below 2 seed m-2 on the ghost shrimp bed, while seed densities remained initially higher (9 seed m-2) on the mud shrimp bed. After 1 yr, seed densities were close to 0 and 3 seed m-2 on the ghost shrimp and mud shrimp bed, respectively. ed on untreated plots, although the effect of species-specific activities on seed survival could not be clearly differentiated from the effect of shrimp density because there were twice as many shrimp at the ghost shrimp site than at the mud shrimp site (Fig. 8). Dumbauld et al. (1997) examined the role of shrimp density as it affects oyster mortality and found a significant negative correlation between densities of ghost shrimp and survival of oyster seed. No seed survived beyond 90 d where burrows exceeded 40 holes m-2, suggesting the presence of a density threshold response function. Armstrong et al. (1992) also found a significant negative correlation between ghost shrimp burrow counts and surface shell retention in studies examining the use of oyster shell to create habitat for juvenile Dungeness crabs. There, too, most of the shell sank within 2 wk after construction where burrows exceeded 40 holes m-2. In contrast to their results from ghost shrimp beds, Dumbauld et al. (1997) found that oyster seed survival was unaffected by mud shrimp density throughout the range examined (approximately 0 to 120 burrow holes m-2) when measured approximately 10 mo after initial planting. Suspended culture is also vulnerable to thalassinidean shrimp. The burrowing activities of shrimp combined with the scour caused by tidal currents and waves often cause structures to topple over so that oysters are subjected to the same processes that negatively affect ground culture, namely burial and siltation. Growers also maintain that in areas Oysters, Crabs,and BurrowingShrimp with relatively high densities of shrimp, oyster growth is negatively affected by feeding interference (suspended sediments obstructing gills and ciliary tracts) and competition with filter feeding mud shrimp for phytoplankton. There is some evidence that shrimp bioturbation can reduce growth rates in some suspension feeding bivalves. Although low levels of particle resuspension may enhance bivalve feeding (Bricelj et al. 1984; Grant and Thorpe 1991), higher levels of turbidity can inhibit filtering rates and increase production of pseudofeces (Loosanoff and Tommers 1948; Barille et al. 1993). Murphy (1985) found that juvenile hardshell clams Mercenaria mercenariawere negatively affected by high levels of suspended particulates caused by ghost shrimp bioturbation, reducing survival and growth compared to areas where shrimp were absent. Growers also contend that mud shrimp slow oyster growth by competing for and limiting phytoplankton even though there have been no controlled experiments conducted to evaluate claims of interspecific food competition. Dumbauld (1994) reported no significant difference in shell size, meat weight, or condition index of oysters grown over a 3-yr period in either the presence or absence of mud shrimp based on a field experiment in which plots were either treated with carbaryl to remove shrimp or left untreated to retain shrimp prior to planting oyster seed. Yet, the lack of effect in his study may have been an artifact of spatial scale (i.e., 100 m2 treated plots embedded within a large-scale shrimp bed) or due to some other factor, and therefore the results are inconclusive with respect to potential competitive interactions. REGULATION In order to spray carbaryl on fallow beds and beds planted with seed oysters, growers must certify that shrimp burrows meet or exceed a minimum of 10 holes m-2 (regardless of species composition) and submit chemical pest control applications to WDOE by May of each year. The WDOE authorizes and regulates the use of carbaryl under Washington Administrative Code 220-20-10 (16) in compliance with Washington State Special Local Needs Pesticide Registration No. WA670021 under authority of section 24(c) of the Amended Federal Insecticide, Fungicide, and Rodenticide Act. Until 1993, WDFW administered the program and initiated and supported research on minimizing nontarget impacts. Once these issues were adequately addressed to their satisfaction through an Environmental Impact Statement (EIS) and a Supplemental EIS, administrative oversight was transferred to WDOE, although WDFW continues to issue comments. 157 Carbaryl is only authorized for use on exposed oyster beds during daylight spring tides in July and August. The seasonal timing of spray is limited to these months based on attempts to avoid juvenile salmonid outmigration in the spring as well as research showing that the rate of hydrolysis is accelerated by alkaline conditions, the presence of organics, higher temperatures, sunlight, oxygen availability, and microbial activity (Carpenter et al. 1961; Karinen et al. 1967; Rajagopal et al. 1984; Larkin and Day 1985; Cranmer 1986). Beds are inspected prior to spray to record general physical characteristics of the area, use of the bed, presence or absence of oysters, burrow densities, and abundance of juvenile crabs. On the day of spray, the oyster grower or an industry representative must be present on site to witness the application and halt the procedure if wind velocities exceed 10 mph. The pesticide is delivered directly to the mudflat in a wetable powder form. Most tracts are sprayed by helicopter while smaller areas and bed perimeters are treated using back-pack hand sprayers. Per state and federal label requirements, carbaryl is prohibited within 200 ft of sloughs, channels, and harvestable oysters if applied by helicopter and within 50 ft if applied by hand sprayer. Application must be completed no more than 30 min after low tide. Shrimp are likely exposed to carbaryl primarily through ventilation of burrow water. Within minutes of delivery mud shrimp begin to exit their burrows and die on the surface, whereas ghost shrimp typically die in their burrows (DeWitt et al. 1997; authors' personal observations). Carbaryl concentrations and total allowable spray area have varied throughout the years based on results of efficacy experiments (see following subsection), non-target impacts (principally to Dungeness crabs), and extent of shrimp infestation. Concentrations have declined from 11.2 kg carbaryl ha-' to 9 kg carbaryl ha-1 while total annual spray area has doubled from 162 ha to 324 ha, with the majority of hectares concentrated in Willapa Bay. EFFICACYAND PERSISTENCE IN SEDIMENTS From 1963 to 1984, carbaryl was applied at 11.2 kg ha-1 which was shown to kill 90-95% of targeted shrimp on treated beds. Prompted by increased public concern over use of the chemical and requests by growers to spray a greater number of hectares, WDFW reduced the permitted concentration to 8.4 kg ha-~ in 1984. Subsequent studies determined that shrimp beds could be treated at concentrations lower than 11.2 kg ha-1 without compromising efficacy. Creekman and Hurlburt (1987) found no significant differences in the proportion of shrimp killed at 11.2 kg ha-1, 8.4 kg 158 K. L. Feldmanet al. californiensis Neotrypaea 1- 10.8- 0.8- 0.6: 0.6- 0.4: 0.41 -0 0.2 I D Expt3 0 c 0 1 *? Expt2 4 2 6 @ 0- A A /_\ . I . . 0 1: 0.8- 0.80.4- ELI@ ~ * 5.6 kg ha' o 1.7 kg ha'1 A 0.5 kg ha'~ ... ... 1 . 2 . . . w v l , 3 4 . I 5 Upogebiapugettensis _ 0.6 Al C1 0.2- 8 0 0. * * * 0- 0 oD * * 5.6 kg ha't o 1.7 kg ha'1 A 0.5 kg ha' 0.4: *Expt2 m C 4po 0.6- , ,// 0.2 * o Expt3 6 2 4 8 ApplicationRate (kg ha'1) 0.2 0 EA AA .:.._........ 0 1 2 3 4 5 Exposure Time (hr) Fig. 9. Dose response models for the proportion of ghost shrimp (Neotrypaeacaliforniensis)and mud shrimp (Upogebiapugettensis) killed 24 h after carbaryl application. Lines represent best-fitting binomial models. Application rate was significant for both species (upper left and lower left graphs), although lower concentrations were often as effective as higher concentrations for ghost shrimp when exposure time exceeded 1.5 h (upper right graph). (Source: Dumbauld et al. 1997). ha-1, and 5.6 kg ha-1. They also determined that a lower concentration (5.6 kg ha-1) applied in the summer (July-August) was as effective as a higher concentration (11.2 kg ha-l) applied in the spring (May-June), suggesting that efficacy, like hydrolysis, increased with warmer temperatures. Dumbauld et al. (1997) examined the effects of the duration of carbaryl exposure on efficacy and species susceptibility. Their data indicated that the optimal time to assess efficacy as measured by burrow counts was 1 mo after treatment in order to allow sufficient time for burrows to collapse. They also discovered that the duration of exposure was just as important as concentration to achieve a 90% kill for ghost shrimp, whereas concentration was more important for mud shrimp. Ghost shrimp grounds could be treated effectively at lower concentrations (2-5 kg ha-1) than mud shrimp grounds (7-9 kg ha-l) but required longer exposure times (2-5 h versus 1.5-3 h; Fig. 9). When carbaryl is applied to tideflats, much of the pesticide is adsorbed onto sediment grains. Persistence of carbaryl in estuarine sediments has not been investigated nearly as well as in terrestrial environments (Mount and Oehme 1981; Rajagopal et al. 1984), but studies suggest that residual concentrations, and hence toxicity, in sediments vary with season, time, and sediment type. Karinen et al. (1967) found that carbaryl applied at 11.4 kg ha-' in February or March persisted in sediments at levels over 1 ppm for 8 d and over 0.1 ppm for 42 d-much longer than degradation rates measured in the summer (WDF and WDOE 1992; Dumbauld et al. 1997). WDFW studied persistence as a function of time on a commercial oyster bed near the Palix River in Willapa Bay (WDF and WDOE 1992). Carbaryl concentrations on plots treated at 5.6 kg ha-1 in July averaged 62 ppm in the top 3 mm of sediment immediately after treatment and 41 ppm on plots treated at 8.4 kg ha-'. One day later concentrations declined to 0.78 ppm and 2.91 ppm respectively, and after 16 d levels were similar (0.023 ppm and 0.021 ppm). Carbaryl was still detectable albeit at very low levels (0.008 and 0.010 ppm) after 28 d, but not after 11 mo (< 0.001 ppm). Dumbauld et al. (1997) compared residual carbaryl concentrations at two sites in Willapa Bay, one dominated by mud shrimp and the other by ghost shrimp. Although sediments at both sites were similar (phi sizes of 2-3), a greater mean Oysters,Crabs,and BurrowingShrimp proportion of very fine sands and silt dominated the mud shrimp site (42% versus 10%). Plots sprayed at 5.6 kg ha-1 averaged 0.14 ppm on the ghost shrimp bed and 1.06 ppm on the mud shrimp bed 1 d after treatment. After 26 d, concentrations declined to 0.002 ppm and 0.03 ppm on the ghost shrimp and mud shrimp bed, respectively. Based on first order decay rate models, levels would have been detectable (> 0.001 ppm) for 43 d on the mud shrimp bed and 28 d on the ghost shrimp bed. Using a conservative 24-h EC50 (effective concentration that produces death or irreversible paralysis in 50% of test animals) of 0.01 ppm (Stewart et al. 1967), sediments could have remained toxic to 0+ burrowing shrimp for 28 d at the mud shrimp site and 12 d at the ghost shrimp site. Ghost shrimp recruitment began in August approximately 1 mo after carbaryl application and no negative response was observed; there were no significant differences in densities of 0+ shrimp among control plots and plots treated with 0.5, 1.7, and 5.7 kg ha- either in August or 2 mo later in October. Mud shrimp failed to recruit into treated plots the following May, but their absence was most likely due to changes in habitat or the absence of adult conspecifics rather than any residual carbaryl toxicity. NON-TARGETIMPACTS There have been and continue to be concerns about the short-term and long-term impacts of carbaryl on non-target species. These concerns led to the termination of the program in Tillamook Bay, Oregon, in 1984 where 40 ha of oyster grounds had been treated annually from 1964 to 1981 (Bakalian 1985; Buchanan et al. 1985). The Oregon Court of Appeals determined that the risk to resources and ecosystem functioning was too great given the lack of data on potential adverse effects on estuarine organisms. In 1984, WDF and WDOE called for an EIS to review and evaluate the carbaryl program in Washington State and address the concerns of the public, crab fishers, and natural resource agencies. The final EIS was completed in 1985, but it was apparent that these critical issues could not be adequately resolved without further study. Experiments were conducted to address these concerns and some of the results discussed below were included in a Supplemental EIS (WDF and WDOE 1992). Flora and Vertebrate Fauna Eelgrass, macroalgae, and benthic microalgae may be present on commercial oyster beds and thus directly exposed to carbaryl although no studies have examined the effect of the pesticide on species indigenous to Willapa Bay and Grays Har- 159 bor. The few studies that have been conducted on freshwater and marine algae indicate that shortterm (2-24 hr) disruption of "4Cuptake may occur at 10 ppm (Kentzer-Baczewska et al. 1984) and 3.7 ppm (Peterson et al. 1994) while photosynthesis was unaffected at 0.05 ppm (Ramachandran et al. 1984). The sum of available evidence suggests that carbaryl has no irreversible physiological effects on aquatic flora and given its rapid degradation is unlikely to harm estuarine plants (WDF and WDOE 1992). Many species of fish are adversely affected by carbaryl but show less sensitivity to the pesticide than invertebrates. The EC50values for carbaryl are generally an order magnitude higher for fishes than for many invertebrates (particularly crustaceans), but fish are more sensitive to 1-napthol, carbaryl's immediate breakdown product, than invertebrates (Table 5; Stewart et al. 1967). Acute toxicity bioassays have been criticized as being inapplicable to field conditions in which the dynamic nature and spatial scale of the environment work to rapidly dilute and degrade the chemical. Yet, many small fish may be retained on the mudflat at low water, trapped in shallow tide pools and exposed to carbaryl for hours. Surveys of treated oyster beds reveal that staghorn sculpin (Leptocottus armatus), saddleback gunnels (Pholis ornata), English and sand sole (Parophrys vetulus and Psettichthysmelanostictus), shiner perch (Cymatogasteraggregata),starry flounder (Platichthysstellatus), bay gobies (Lepidogobiuslepidus), and three-spine sticklebacks (Gasterosteusaculeatus) are killed following carbaryl application. Mortality varies depending largely on the amount of marine fish habitat on the oyster bed, defined as water > 5 cm depth, as well as speciesspecific spatial and temporal patterns of abundance. Hueckel et al. (1988) estimated that an average of 1,208 fish ha-~ were killed on three treated oyster beds, ranging from 489 gunnels ha-~ to 10 starry flounders ha- (Table 6); marine fish habitat comprised 20-98% of the area sampled on two of the tracts. Tufts (1989, 1990) conducted a single 3.05 m X 30.5 m transect survey on each treated oyster bed in 1986 and 1987 and found interannual differences in the amount of marine fish habitat, number of fish killed, and species composition. Marine fish habitat accounted for 34% of the treated area in 1986, but only 15% in 1987; the average number of fish killed was 93 fish ha-~ in 1986 and 46 fish ha-1 in 1987. In 1986, staghorn sculpin (63%) and saddleback gunnels (34%) dominated the fish kill, whereas in 1987 shiner perch (27%), saddleback gunnels (25%), sculpins (19%), and sole (27%) accounted for most of the mortality. Fish in off-tract habitats are unlikely to be affected by carbaryl based on studies showing 160 K. L. Feldmanet al. TABLE 5. Carbaryl and 1-naphthol EC50values for selected crustaceans, molluscs, polychaetes, and fishes. EC50values are effective concentrations that produced death or irreversible paralysis in 50% of the test animals (after Stewart et al. 1967). Exposure Time (h) EC50 (ppm) Carbaryl Amphipod (Gammarus lacustris) 24 0.04 Mud shrimp (Upogebiapugettensis) Larvae 24 0.03-0.16 6.2-13.7 Stewart et al. 1967 Ghost shrimp (Neotrypaeacaliforniensis) Larvae Adult 24 24 0.17-5.60 0.13 15.0-22.1 6.6 Stewart et al. 1967 Stewart et al. 1967 Dungeness crab (Cancer magister) Larvae Juvenile J2 Juvenile J9 Adult 24 24 24 24 0.08 0.076 0.35 0.49 - Shore crab (Hemigrapsusoregonensis) Adult 24 0.06-1.05 69.5-83.4 Stewart et al. 1967 Species EC,50(ppm) 1-naphthol Source Crustaceans Molluscs Bay mussel (Mytilus edulis) Larvae - - Mount and Oehme 1981 Buchanan Buchanan Buchanan Buchanan et et et et al. al. al. al. 1970 1970 1970 1970 48 1.4-2.9 0.8-2.2 Stewart et al. 1967 Pacific oyster (Crassostreagigas) Larvae 48 1.5-2.7 0.6-1.1 Stewart et al. 1967 Cockle (Clinocardiumnuttalii) Adult 24 7.3 5.1-7.8 Stewart et al. 1967 48 7.2 - Polychaetes Lugworm (Arenicolamarina) Fishes Shiner perch (Cymatogasteraggregata) Juvenile 24 English sole (Parophrysvetulus) Juvenile 24 Three-spine stickleback (Gasterosteusaculeatus) Brown trout (Salmo trutta) Rainbow trout (Salmo gairdnerii) 24 48 96 that little to no mortality of Dungeness crabs occurs in subtidal channels (Doty et al. 1990; WDF and WDOE 1992). Fish also are unlikely to die as a result of consuming contaminated taxa. Carbaryl concentrations in prey items are much lower than that needed to have an adverse impact on fish, although temporary shifts in consumption patterns may occur as prey availability changes (WDF and WDOE 1992). Other behavioral and physiological changes that have been observed in freshwater species exposed to non-lethal concentrations of carbaryl for 3 d or less include an increase in swimming activity (Arunachalam et al. 1980) and susceptibility to predation (Little 1990; Carlson et al. 1998), while longer-term exposure to the pesticide (27-60 d in studies cited) has been shown to impair reproductive potential (Bhattacharya 1993; Kaur and Dhawan 1996) and reduce growth rates (Arunachalam et al. 1980). Given the rapid deg- 3.8-4.0 Conti 1987 1.3-1.8 Stewart et al. 1967 3.2-5.0 2.4-2.7 Stewart et al. 1967 5.5-7.7 1.5 4.38 2.8-3.5 Stewart et al. 1967 Mount and Oehme 1981 Mount and Oehme 1981 - - radation and dilution of the pesticide upon flood tide, the results of the latter experiments are not applicable to estuarine field conditions unless changes in prey abundance and/or species composition temporarily depresses food intake and growth rates of fishes. Western gulls (Larus occidentalis) and Glaucouswinged gulls (Larus glaucescens) have been observed on oyster beds consuming contaminated burrowing shrimp after treatment. Shorebirds occasionally have been noted on treated mudflats as well, although carbaryl application occurs between peak migration periods (WDF and WDOE 1992). Based on oral toxicity studies, birds are insensitive to carbaryl at concentrations characteristic of field conditions (Mount and Oehme 1981). WDF and WDOE (1992) calculated that a seagull weighing 1 kg would have to ingest approximately 40 kg of burrowing shrimp before the incoming tide to Oysters, Crabs,and BurrowingShrimp TABLE 6. Estimated mean numbers of fishes killed on commercial oyster bed tracts in Willapa Bay treated with carbaryl. Data from Hueckel et al. (1988). Taxa Saddleback gunnel (Pholis ornata) Staghorn sculpin (Leptocottusarmatus) Bay goby (Lepidogobiuslepidus) Three-spined stickleback (Gasterosteusaculeatus) Starry flounder (Platichthysstellatus) Number ha-' 95% CI 489 361 309 40 10 620 791 630 141 40 even approach a lethal concentration. In North Dakota where carbaryl is distributed in a dry bait formulation for range grasshopper control, Fair et al. (1995) found that while killdeer (Charadriusvociferus) were exposed to carbaryl based on pesticide residue concentrations of whole-body and gizzard contents; no difference in brain acetylcholinesterase activity was detected between treated and untreated areas. George et al. (1992) found no inhibition of acetylcholinesterase activity in other birds and small mammals in the same system. While direct toxic effects to birds are unlikely, prey availability and habitat utilization may be affected by the application of carbaryl. In estuaries, foraging patterns of shorebirds may be affected like those of fish because both taxa prey on many of the same epibenthic and infaunal organisms such as tanaids, gammarid amphipods, harpacticoid copepods, ostracods, cumaceans, and polychaetes. While benthic species composition may be altered temporarily, total prey abundance may not change. Simenstad and Cordell (1989) found that densities of many taxa increased rather than decreased on treated plots relative to controls in a short-term (2 wk) study, suggesting that resources should not be limiting in areas sprayed with carbaryl if birds are capable of switching among alternative prey items. In northern Maine where carbaryl is aerially applied to control spruce budworm Choristoneurafumiferana,American black duck (Anas rubripes)and mallard (Anas platyrhynchos) ducklings exhibited decreased growth rates following treatment (Hunter et al. 1984) and several genera of Parulinae warblers emigrated from affected areas to forage in nearby untreated canopies (Hunter and Witham 1985). Dungeness Crab Seasonal Timing of CarbarylApplicationwith Respect to CrabSettlementand SubadultActivity.Acute toxicity studies indicate that Dungeness crabs are extremely susceptible to carbaryl, as are most other crustaceans, and that sensitivity varies with ontogeny (Table 5; Buchanan et al. 1970). Young-of-the-year (0+) crabs, which settle primarily in May and June 161 to intertidal habitats (Doty et al. 1990; Dumbauld et al. 1993; Eggleston and Armstrong 1995), come into direct contact with carbaryl as it is sprayed over them at low tide. Doty et al. (1990) determined that 100% of 0+ crabs present on oyster beds at the time of treatment were killed. However, survival of crabs placed on a treated mudflat in mesh bags 24 h after spray for 14 d was relatively high (x = 70%) and was not significantly different than that of crabs placed on a control plot (x = 90%). These data suggest that 0+ crabs should be capable of settling or recolonizing oyster grounds shortly after treatment. Indeed, crabs were found in samples collected at 2 wk, 4 wk, and 6 wk postspray although recovery to pre-spray densities differed by site. A few newly settled crabs were occasionally collected but the majority of crabs were older juveniles (25-35 mm carapace width, CW), indicating that immigration from adjacent untreated areas was primarily responsible for recolonization. Subadult 1 + and 2+ crabs that migrate into and make extensive use of estuaries during spring and summer months (Gunderson et al. 1990) move up from subtidal channels at high tide and are killed by consuming contaminated shrimp and other taxa lying on the surface as well as by exposure to toxic environmental conditions. Doty et al. (1990) conducted crab feeding experiments and investigated whether there was a dose-response relationship with respect to shrimp exposed to 3.4, 5.6, and 8.4 kg carbaryl ha-1 and crab mortality. In their first experiment, crabs fed contaminated shrimp treated at 8.4 kg ha-1 had greater mortality (38%) than crabs fed uncontaminated shrimp (2%). In their second experiment, no differences in mortality were detected among crabs fed shrimp killed at different carbaryl concentrations compared to a control, suggesting that lowering carbaryl concentrations would not improve crab survival. The risk to foraging crabs appears to be greatest in the first 24 h after spray. Creekman and Hurlburt (1987) found that the concentration of carbaryl in tissues of burrowing shrimp treated at 8.4 kg ha-1 declined by 90% within 1 d of treatment from about 4.5 ppm to 0.5 ppm. Thus, 1+ and 2+ crab mortality by ingestion is most likely to occur within the first 24 h, corroborating reports that freshly killed crabs are not generally observed on treated beds after 24 h. Effectsof Off-siteTransportof Carbarylon Crabs.Carbaryl sprayed during low tide can be carried substantial distances off treated tracts by the leading edge of an incoming flood tide and is detectable in water samples both on-site and off-site at concentrations lethal to Dungeness crabs. Doty et al. (1990) found that concentrations of carbaryl in wa- K. L. Feldmanet al. 162 I Crabmortality Carbaryl(ppm) --.---. --- 100 75- o"' -- ? 100 - . 10 0 0o or 0 50- , - 1 a . - 0 a) CD 3 25- - b, 0 On site 25 m i ,i 50m 75m i 0.1 0.01 100m 200m 800m Distance from treatment site Fig. 10. Average mortality (+1 SD) of 0+ Dungeness crabs (Cancer magister) exposed to carbaryl as a function of distance from the actual treatment site in the path of the incoming flood tide (n = 5 bags of oyster shell station-' and 10 crabs bag-1). Carbaryl concentrations (ppm) were measured next to bagged crabs by taking water samples 2.5 cm above the substrate when total water depth reached 10 cm at each station. (Source: modified from Doty et al. 1990). ter samples taken on-site 2.5 cm above the benthos declined from 16.7 ppm as the tide initially flooded over the tract to 0.25 ppm 30 min later as the chemical was transported and diluted. Immediately adjacent and downstream of the treated plot, concentrations fluctuated over the 30 min sampling period and generally increased with time as water depth increased and then gradually decreased. For example, at 50 m off-site, concentrations increased from 8.0 ppm when the tide first arrived to 12.6 ppm after 4 min and then decreased to 0.25 ppm after 30 min. A similar trend was observed 200 m off-site, although concentrations did not exceed 1 ppm over the sample period. Carbaryl was even detected at low levels (' 0.063 ppm) on a control plot located 800 m from the treated bed and beyond the direct pathway of transport. Average mortality of 0+ crabs held in shell-filled bags for 24 h on-site and off-site at distances corresponding to water sampling locations was uniformly high up to 100 m off-site, ranging from 83-100% (Fig. 10). At these locations off-tract, crabs were exposed to relatively high concentrations of 4-12 ppm carbaryl for about 4 min, sufficient to be lethal (Fig. 10). In contrast, mortality was significantly lower at 200 m and 800 m away from the treated bed (9% and 4%, respectively) where carbaryl had been diluted to sublethal concentrations below 0.1 ppm. Although Doty et al. (1990) found that close to 100% of crabs were killed up to 100 m off-site, it is difficult to predict or quantify off-tract mortality because tidal direction, water velocity, turbulence, and specific site characteristics affect the two-dimensional pathway of the chemical and how rapidly it is diluted in the water column (Hurlburt 1986; Tufts 1989; Doty et al. 1990; WDF and WDOE 1992). Furthermore, the extent of off-site 0+ mortality is influenced by the habitat over which carbaryl is transported since different habitats support varying densities of crabs (Doty et al. 1990; Fernandez et al. 1993; Eggleston and Armstrong 1995; McMillan et al. 1995). While off-site mortality of intertidal 0+ crabs may be substantial, adverse impacts to 1+ and 2+ crabs residing in subtidal channels adjacent to treated tracts appear to be insignificant. Doty et al. (1990) conducted subtidal trawls prior to and 24 h after carbaryl treatment at two locations in Willapa Bay and found very few dead or moribund crabs in post-spray assessments, amounting to only about 1% of the catch. Virtually all 1 + crabs placed subtidally (- -1 m at MLLW) in cages along the perimeter of the two treatment sites and held for 24 h following carbaryl spray survived: one crab out of 71 died at one location while no crabs died at the other location. Mortality of 1+ crabs held in cages placed in intertidal drainage creeks adjacent to treated beds was greater and more variable compared to subtidal locations. Average mortality for both sites combined was 21%, with 40% and 45% of crabs dead at one location while no crabs died in two replicates at another location. Site-specific characteristics may have encouraged drainage of water retained on the bed during treatment at low tide into nearby tidal creeks or directed pesticide transport towards tidal creeks at flood tide. Crab Kill Assessmentand Establishmentof a Quota System. Reports of dead crabs on treated tracts prompted WDFW officials to carefully regulate the timing and location of treatments and to assess the magnitude of crabs killed by carbaryl. In 1976, the agency began to conduct visual systematic surveys of treated beds 24 h after treatment to quantify the number and size of crabs killed, and in 1984, established a maximum kill quota of 12,000 sublegal adult equivalents (SLAe's) for the entire treated acreage. A 2+ crab (> 100 mm CW) was defined as one SLAe (a ratio of 1:1). Based on conservative natural mortality rates, 1 + crabs (40-100 mm CW) were converted to a SLAe using a ratio of 3:1, and 0+ crabs (< 40 mm CW) were converted using a ratio of 10:1 in order to compute an overall SLAe kill. In 1984, the newly settled Dungeness year class was exceptionally large, and 0+ crabs comprised 84% of total crabs killed. Total 0+ mortality was likely in error since visual transect surveys may grossly underestimate the true abundance of 0+ crabs (Doty et al. 1990). In 1985 WDFW predicted Oysters,Crabs,and BurrowingShrimp that greater numbers of 1 + crabs would be killed compared to previous years and restricted individual growers to a maximum kill of 75 SLAe's ha-1. Quotas were therefore established for each grower based on the area of their tracts. This system enabled the agency to halt treatment of an individual grower's beds if he had surpassed his crab quota while allowing other growers to continue spraying so long as neither the overall quota nor the individual's quota had been exceeded. As expected, 1+ crabs dominated the population and accounted for 88% of all crabs killed (Creekman and Hurlburt 1987). Despite efforts to spray beds with historically low crab densities and bait nearby sloughs and channels to divert crabs from treated beds, 19,275 SLAe's (121 SLAe's ha-l) were killed in 1985 on a total of 159 ha (78% of authorized acreage). In 1986, WDFW anticipated that the 1984 year class, present as 2+ crabs, again would dominate the crab kill; however, 2+ crabs comprised only 37% of the kill which amounted to a total of 9,141 SLAe's or 57 SLAe's ha-1 (Tufts 1989). Crab kill is in large part unpredictable, and short of severely restricting or eliminating carbaryl treatments (which WDFW declined to do in 1985 because some growers had not had the opportunity to spray any of their tracts before the overall quota had been exceeded), little can be done to reduce impacts to a large year class of subadult crabs. Crab kill assessments were conducted until 1992 at which time responsibility for the carbaryl program was transferred to WDOE after completion of the supplemental EIS. OtherInfaunal and Epifaunal Invertebrates Carbaryl has an acute lethal effect on many benthic invertebrates although mortality has been shown to vary among taxa as well as among species within a taxon. Bioassay studies reveal that bivalves and polychaetes are less sensitive to carbaryl and have higher EC50values than crustaceans (Table 5; Stewart et al. 1967; Buchanan et al. 1970; Mount and Oehme 1981; Conti 1987). While results of field studies have corroborated this general pattern by noting the lack of significant differences in the densities of several species of bivalves and polychaetes between treated and control plots (Armstrong and Millemann 1974; Dumbauld 1994), species-specific responses to carbaryl exposure are evident. Both Brooks (1993) and Dumbauld (1994) found that Macoma balthica, a deposit feeding tellinid bivalve, was unaffected by carbaryl application whereas Cryptomyacalifornica, a suspension feeding commensal clam associated with ghost shrimp burrows, was negatively affected, although the impact was immediate in Brooks' (1993) study and delayed in Dumbauld's (1994) investigation. 163 Armstrong and Millemann (1974) reported higher mortalities for the gaper clam Tresuscapax than the bent-nosed clam Macoma nasuta for both lower carbaryl concentrations (5.6 kg ha-') and higher concentrations (11.2 kg ha- ). Crustaceans also exhibit contrasting and variable responses to the pesticide. Brooks (1993) found a significant 10% reduction in the tanaid Leptochelia savignyi 48 h after spray while densities of the amphipod Corophium acherusicum were reduced by 97% and the cumacean Cumellavulgaris by 89% on treated commercial oyster beds. On the other hand, Simenstad and Cordell (1989) reported a significant rise in densities of C. acherusicumand C. vulgaris 24 h after spray on a treated oyster bed compared to a control bed in their study, but reduced abundance 2 wk after spray on the treated bed. They suggested that heightened densities of those organisms reflected a sublethal behavioral response to carbaryl that made them more susceptible to capture. They could not conclusively reject, however, the hypothesis that short-term increases were due to bedload transport and concentration of dead and moribund individuals in the benthic boundary layer since it was difficult to distinguish between live and dead individuals in the samples. The results of both of these studies differ from those of Dumbauld (1994) who found that densities of L. savignyi, C. acherusicum, and C. vulgaris were generally lower on small (100 m2) carbaryl treated plots compared to control plots 24 h, 2 wk, and 1 mo after spray, yet none of these differences were statistically significant (Table 7). Differences in sampling methodologies (i.e., van Veen grabs, epibenthic pump samples, benthic cores), tidal stage sampled (high tide versus low tide), replication, spatial scale (oyster beds versus 100 m2 treated plots nested within a broad shrimp bed), and site characteristics, including the presence of predators, may have contributed to the variation in these species-specific responses. Dumbauld (1994) did document significant impacts to the burrowing amphipod Eohaustorius estuarius and a small cyclopoid copepod (Hemicyclopssp.) on treated plots in an area dominated by ghost shrimp (Table 7). Despite some evidence of severe and immediate effects on benthic invertebrates, few if any generalities can be drawn with respect to the pesticide's impact on the estuarine community or food web interactions. First, virtually no studies have been conducted on the potential for and importance of sublethal effects on estuarine organisms. Buchanan et al. (1970) found that carbaryl could delay molting in Dungeness crab larvae, and results of other studies from aquatic systems suggest that carbaryl may affect swimming behavior (Capaldo 1987) and feeding efficiency (Donkin 1997). Sec- K. L. Feldmanet al. 164 TABLE 7. Summary of effects of carbaryl treatment on densities of infaunal invertebrates at the Palix River station which was dominated by ghost shrimp (Neotrypaeacaliforniensis) and at the Cedar River station which was dominated by mud shrimp (Upogebia pugettensis). Carbaryl was applied at 5.6 kg ha-~ to experimental plots and species densities were monitored at 1 d, 2 wk, 1 mo, 3 mo, and 1 yr post-spray on carbaryl-treated plots and untreated control plots. + indicates that species densities were significantly higher (t-tests, p < 0.05) on treated plots than untreated plots; - indicates that densities were lower on treated plots compared to untreated plots; 0 indicates that densities on treated and untreated plots were not significantly different. Data from Dumbauld (1994). Time (post-spray) Taxa Polychaetes Site Palix R. Cedar R. Molluscs Palix R. Cedar R. Crustaceans Palix R. Cedar R. Species Mediomastuscaliforniensis Paronellaplatybranchia .Hemipodusborealis Mediomastuscaliforniensis Pseudopolydorakempi Capitellacapitata Macoma spp. Cryptomyacalifornica Macoma spp. Clinocardiumnutalli Eohaustoriusestuarius Hemicyclopssp. Hemileuconcomes Cumellavulgaris Corophiumacherusicum Leptocheliasavignyi ond, the few studies that have examined longerterm patterns of species density and diversity on treated grounds have found that these areas either are recolonized rapidly through recruitment and bedload transport or exhibit no consistent changes or trends attributable to carbaryl. Simenstad and Cordell (1989) found that 2 wk after spray, densities of most invertebrate taxa approached if not exceeded pre-treatment levels. Brooks (1993) found that although most taxa declined or were nearly eliminated when sampled 2 d after spray, many had rebounded dramatically after 51 d. Nevertheless, densities of some species remained depressed or were slow to recover in both studies. The longest studies investigating non-target impacts to benthic invertebrates were conducted by Brooks (1995) and Dumbauld (1994). Brooks (1995) measured the impact of carbaryl on species densities and diversity on two oyster beds from 1992-1994. Although there were some significant short-term impacts (Brooks 1993), within 1 yr, densities of all species were as high or higher on treated beds than on control beds, while species diversity was similar for one of the treated beds and the two control plots but lower on the other treated bed. Dumbauld (1994) followed invertebrate densities and species diversity on treated and untreated control plots on both a ghost shrimp bed and a mud shrimp bed for 1 yr. He found some shortterm but few long-term effects of carbaryl on species densities (Table 7). The small bivalve Cryptomya californica and the amphipod Eohaustorius estuarius, commonly associated with ghost shrimp beds, were negatively affected by carbaryl or sub- 1d 2 wk 1 mo 3 mo 1 yr 0 0 0 0 0 + 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0- 0 0 0 0 0 + + 0 0 0 0 0 0 0 0 0 0 0 0 strate modification, but most species were largely unaffected. Moreover, there was a greater difference in species diversity between the mud shrimp bed and the ghost shrimp bed than between sprayed and non-sprayed plots. Diversity (Margalef's D) ranged from approximately 2.5 to 3.7 on the mud shrimp bed and 1.4 to 2.0 on the ghost shrimp bed over the year, but very few differences were observed between treated and untreated plots within each bed. In summary, although some species were negatively impacted by carbaryl in the short-term, recolonization via bedload transport and rapid reproduction largely offset impacts resulting from carbaryl application. Ecological Significance of Carbaryl Use Long before American colonists sought ways to harvest and cultivate the abundance of fisheries resources, native oysters, crabs, and burrowing shrimp were integral components of Pacific Northwest estuarine ecosystems. However, observations by oyster growers and fisheries biologists suggest that dramatic expansions of burrowing shrimp populations since the 1950s (whether the result of natural population cycling or anthropogenic influence) have upset the balance of species and habitat diversity. While treatment of intertidal mudflats with carbaryl has enabled the industry to largely maintain levels of harvest and protect highly successful grounds from production losses due to burrowing shrimp, some growers and scientists also view the use of carbaryl as a management tool to help sustain species assemblages. Although Crassostrea gigas is an exotic and, as farmed, is different Oysters,Crabs,and BurrowingShrimp * TotalO+crabs on oysterbeds - 165 O+crablosses on treatedbeds IA CM *E -------. -1-.o--. 150 (/) cn a) c () 0) Heavy eelgrass t0 Open mud E co 0 c. \ 10 75- 50- CD a) 0) C 0 - 25- + o + 50- 1986 Jun Jul Aug Sep Fig. 11. Mean density (+ 1 SE) of 0+ Dungeness crabs (Cancer magister) as a function of habitat type: heavy shell (-> 50% oyster seed [shell] cover), heavy eelgrass (> 50% eelgrass cover and < 10% oyster seed cover), and bare mud (< 10% oyster seed cover and < 10% eelgrass cover) from June through September, 1987, at an experimental site near the Palix River, Willapa Bay. Samples (n = 5-12) were taken by excavating a 0.25 m2 quadrat to a depth of 3-5 cm and sieving the contents through a 3.2-mm mesh screen. (Source: modified from Doty et al. 1990). from native populations of Ostrea lurida, research has shown that oysters play an important role in estuaries by enhancing water quality and clarity (Officer et al. 1982; Newell 1988; Dame 1996; Gottleib and Schweighofer 1996), providing a hard surface area for the attachment of epibionts such as mussels, macroalgae, and sponges, and serving as refuge habitat for juvenile crabs, fishes, shrimps, and other invertebrates (Ambrose and Anderson 1990; Doty et al. 1990; Breitburg 1991; Dumbauld et al. 1993). The dependence of 0+ Dungeness crabs on epibenthic cover has been well documented and serves to illustrate one of the many ecological tradeoffs when carbaryl is sprayed on tideflats. Intertidal shell, such as that provided by oyster culture, and to a lesser extent eelgrass, provide refuge from predators, and densities of 0+ crabs have been shown to be significantly higher in these habitats compared to open unvegetated sediments that may be occupied by burrowing shrimp (Dumbauld et al. 1993; Fernandez et al. 1993; Eggleston and Armstrong 1995; McMillan et al. 1995). In general, shell is superior crab habitat compared to either eelgrass or open mud (Fig. 11). Doty et al. (1990) found that heavy shell (HS: > 50% shell cover) and light shell (LS: 10-49% shell cover) provided 1987 1987 1988 Fig. 12. Comparison of estimated total 0+ Dungeness crab (Cancer magister) abundance on 2,400 ha of cultivated oyster ground in Willapa Bay and estimates of 0+ crabs killed on oyster grounds treated with carbaryl during July 1986, 1987, and 1988. (Source: modified from Doty et al. 1990). by live oysters supported 2-5 times higher densities of 0+ crabs than either heavy eelgrass (HE: > 50% eelgrass cover and < 10% shell cover) or light eelgrass (LE: 10-49% eelgrass cover and < 10% shell cover), especially after initial settlement pulses. For example, the highest densities recorded in HE were 3-4 crabs m-2 compared to 8-12 crabs m-2 in LS and 12-16 crabs m-2 in HS during June. Without any epibenthic shelter, densities of 0+ crabs declined to nearly zero soon after settlement to the benthos. In July, prior to carbaryl spray, crab densities ranged from 0-3 crabs m-2 in HE, 1-4 crabs m-2 in LS, and 4-6 crabs m-2 in HS. Based on extensive data sets from several sites, Doty et al. (1990) determined whether the number of crabs killed by carbaryl treatment was offset by the benefits of increased shell habitat. They estimated that 971 ha of commercial oyster culture produced 40.9, 75.7, and 35.5 million 0+ crabs in July 1986, 1987, and 1988, respectively, while treatment of 144, 145, and 111 ha of tideflat with carbaryl in each of those years killed an estimated 1.7, 3.7, and 1.3 million crabs, or 4-5% of the population (Fig. 12), but they did not include an estimate for off-site crab mortality. The overall loss to the estuarine population probably was much lower since shell and eelgrass habitats on non-cultured intertidal regions and subtidal channels would have provided habitat for crabs as well (Dumbauld and Armstrong 1987; Doty et al. 1990). Based on the number of hectares cultivated and sprayed at the time of their study, Doty et al. (1990) concluded that the oyster industry mitigated for its own negative impacts with respect to crabs, generating significantly higher number of 0+ crabs than eel- 166 K. L. Feldmanet al. grass or open sediments could have produced in the absence of commercial oyster culture. In addition to increased production of 0+ Dungeness crabs in the scenario described above, there is some evidence to suggest that application of carbaryl to shrimp-dominated mudflats may indirectly promote eelgrass expansion and growth under certain circumstances. Eelgrass is sympatric with commercial oyster culture in many areas (but see Pregnall 1993 and Everett et al. 1995 for examples of negative impacts of oyster culture on eelgrass distributions) and with beds of mud shrimp, but does not generally overlap with ghost shrimp populations. Continual sediment turnover and water column turbidity caused by ghost shrimp reduces shoot growth of the introduced Japanese eelgrass Zosterajaponica compared to areas without shrimp and impedes Z. japonica expansion by reducing seedling survival in Willapa Bay (Dumbauld and Wyllie-Echeverria unpublished manuscript). After eradication of ghost shrimp, establishment and percent cover of Z. japonica was significantly greater in carbaryl-treated plots compared to untreated control plots. Because Z. japonica is an annual (although in some areas it persists year-round), ghost shrimp may be able to reinvade open areas before seedlings become established again. The removal of ghost shrimp could result in the proliferation of the native eelgrass Zosteramarina at lower tidal elevations as well, but because the upper distribution of Z. marina is limited by desiccation (R. Thom personal communication) and because seed dispersal may be restricted to within a few meters of an existing bed (Orth et al. 1994), the use of carbaryl may not result in Z. marina colonization at higher intertidal sites. No studies have yet examined whether expansion of the exotic Z. japonica is ecologically desirable or detrimental, but in Washington, Z. japonica is granted the same protective status as the native species under current seagrass policy (Pawlak and Olson 1995). While the debate over carbaryl generally has focused on impacts to non-target species, virtually no consideration has been given to the ecological impacts of removing burrowing shrimp on such a broad scale given that thalassinidean shrimp as well as their burrows have been shown to be important features of estuaries and tropical lagoons. The benefits of sediment reworking, burrow ventilation, and shrimp metabolism include greater oxygen penetration into the sediments adjacent to the burrow wall (Koike and Mukai 1983; Forster and Graf 1992, 1995) and nutrient flux of ammonium and phosphate to the sediment surface and water column (Koike and Mukai 1983; Corredor and Morell 1989; Murphy and Kremer 1992). While Murphy and Kremer (1992) concluded that callianassid-re- lated nutrient flux accounted for less than 5% of primary productivity in their study, shrimp densities at their site were only on the order of 4 shrimp m-2, much lower than densities typically found in the Pacific Northwest. The physical structure of mounds and funnels that sometimes characterize shrimp burrows also affect benthic boundary layer flow and can induce flushing of water and ions in addition to currents created directly within burrows by shrimp fanning their pleopods (Allanson et al. 1992; Ziebis et al. 1996; Stamhuis and Videler 1998). The funnel can also act as a trap whereby detritus and other food particles may be deposited due to reduced shear stress associated with depressions. Higher concentrations of bacteria and benthic microalgae are associated with these biogenic structures (Branch and Pringle 1987; Dobbs and Guckert 1988). Corresponding studies on meiofauna, however, have been mixed. Dittmann (1996) found higher densities of meiofauna in the burrows of Callianassa australiensis than in sediments where the shrimp were excluded through use of a fine mesh barrier, whereas Branch and Pringle (1987) and Dobbs and Guckert (1988) found lower densities of meiofauna in the burrows of Callianassa kraussi and Callianassa trilobata,respectively, compared to ambient sediment adjacent to shrimp burrows. Burrow galleries also provide three-dimensional habitat for small fishes and crustaceans, especially in the absence of epibenthic structure. For example, the arrow goby Clevlandia ios and the northern hooded shrimp Betaeus harrimani utilize shrimp burrows either on a temporary or more permanent basis (Bird 1982; authors' personal observations). These species, as well as the commensal bivalve Cryptomyacalifornica, may also benefit from the feeding current and food supply generated by the shrimps' pumping activities, while the shrimp themselves are important in trophic pathways as a prey item for several species of resident or migratory vertebrates and invertebrates. Although they have no importance as a food item for human consumption (there is a small commercial and recreational fishery for them as bait), burrowing shrimp play an important role in ecosystem processes and often are a dominant component of the benthic community in terms of abundance and invertebrate production. In areas of Willapa Bay where shrimp are so dense that attempts to culture oysters have been abandoned, shrimp biomass may approach 7-10 t ha-~, comparable to annual harvest production of Pacific oysters (6.7 t meat weight ha-1; Dumbauld 1994). Treatment of 324 ha of cultivatable oyster ground with carbaryl results in the removal of an estimated 66 million shrimp or 476 t of biomass per year Oysters,Crabs,and BurrowingShrimp based on calculations by Dumbauld (1994). Given what we know and what we have yet to learn about their estuarine role, we should at least question what effect wide-spread eradication of burrowing shrimp habitat might have on the benthic community relative to the oyster habitat that replaces it. Clearly, there are species-specific costs and benefits associated with each habitat, but the use of carbaryl and other oyster aquaculture practices should be evaluated in a broader, systems or process-oriented approach (Simenstad and Fresh 1995) to better understand how the ecosystem functions and help managers protect and preserve the estuary's natural resources while permitting multiple-use activities. Recommendations for Management Alternatives Although carbaryl is still the only practical and effective control measure available to oyster growers, integrated pest management (IPM) was adopted as the preferred alternative in the Supplemental EIS (WDF and WDOE 1992). One of the principal goals of IPM is to better integrate shrimp lifehistory characteristics and oyster culture cycles into carbaryl policy or the development of alternative strategies (WDF and WDOE 1992; DeWitt et al. 1997). Distinct differences in shrimp life cycles suggest a species-specific and site-specific management approach that integrates each species unique ecological and behavioral characteristics with the particular location and method of oyster culture. Although IPM has not been implemented due to the lack of data necessary to develop an appropriate framework (DeWitt et al. 1997), the principles and goals of IPM are sound. IPM provides a dynamic and flexible approach to pest control that may be adaptable in part, if not entirely, to managing burrowing shrimp populations in areas of oyster culture while addressing the needs of other user groups and minimizing environmental impacts. Below we discuss how timing and variability in shrimp recruitment limit the effectiveness of the carbaryl program and suggest some management alternatives that might be explored to address those issues. In addition, we discuss how the physical modification of habitat through shell pavementing might be incorporated into oyster culture techniques to reduce shrimp recruitment under certain conditions. Carbaryl is used to minimize oyster mortality resulting from burrowing shrimp damage, yet policy regulations do not promote optimal efficiency of the pesticide with respect to establishment of threshold densities of shrimp and timing of application as it affects both oyster seed survival and shrimp recruitment. First, permit approval requires that beds be infested with a minimum of 10 167 holes m-2. That requirement, however, was established based on historical records noting that growers did not typically apply to spray beds below 5 holes m-2, while beds with 5-10 holes m-2 were evaluated on a case by case basis. Given that chemical pest control applications are due in May, beds may be assessed and burrow counts taken as early as March or April when shrimp are still relatively inactive. The relationship between burrow holes and shrimp is poor at this time of year, which obscures the true condition of the bed and likely underestimates actual shrimp abundance (Dumbauld et al. 1996). Although early assessment provides a conservative measure of shrimp abundance for the purpose of control, we recommend that burrow counts be taken as close to the scheduled spray period as possible so that decisions regarding which beds to treat are based on the most accurate information. Data from Dumbauld et al. (1997) and Armstrong et al. (1992) indicate that virtually all oyster seed or shell may be lost when ghost shrimp densities exceed 40 holes m-2, but the actual relationship between shrimp density and oyster damage has yet to be quantified. Because ghost shrimp cause more damage to oysters than mud shrimp, species-specific threshold levels need to be established. We imagine that beds dominated by mud shrimp could have a higher threshold minimum for spray than beds dominated by ghost shrimp. Second, because carbaryl can only be applied in the summer and many growers plant cultch on seed or seed-growout beds in the spring, there is a 1-4 mo period during which beds may endure high rates of seed loss prior to shrimp removal (Fig. 4). Growers can wait until after carbaryl application to plant cultch but then sacrifice growth opportunities and increase the risk of losing seed during fall and winter storms. In addition to the potential for seed loss, another drawback of the timing of carbaryl treatment with respect to shrimp control is that it occurs 1-2 mo prior to ghost shrimp postlarval recruitment. Whereas mud shrimp postlarvae settle in Willapa Bay and Grays Harbor from late April through June and are small 0+ at time of spray, ghost shrimp settle from August through October (Dumbauld et al. 1996). Consequently, 0+ mud shrimp as well as juvenile and adult mud shrimp and ghost shrimp are killed by carbaryl application in July/August, but 0+ ghost shrimp can reinvade treated areas immediately (Fig. 4), often at densities equal to or exceeding those on untreated grounds. In a study of one fallow oyster bed in Grays Harbor, we found that the density of 0+ ghost shrimp 4 mo after carbaryl treatment (181.8 shrimp m-2, S.E. = 13.4) was not significantly different from that on an ad- K. L. Feldmanet al. 168 ---- Ghostshrimpbed .......* 800 - Mudshrimpbed irvg lI !JU v 0 'E 100 - 0 0 .0 E E 400 - Q. C: 600- 0 0 10 - cn 200 - 0~~~~~~ + I 1 1d I I 1 yr 2 yr Ah~~~ 0 3 yr Time (post carbarylspray) Fig. 13. Mean density of shrimp on experimental plots that were first sprayed with carbaryl to remove existing shrimp and then monitored 1 d, 1 yr, 2 yr, and 3 yr post-spray for recolonization. Note that the ghost shrimp bed was rapidly recolonized by ghost shrimp with densities measuring 468 shrimp m-2 1 yr after treatment and remaining high throughout the period. In contrast, following carbaryl treatment densities on the mud shrimp bed remained very low (below 10 shrimp m-2) throughout the entire 3-yr experiment. (Source: data from Dumbauld et al. 1997). jacent untreated fallow bed (145.2 shrimp m-2, S.E. = 14.8; t-test; p = 0.09). We also examined data from Dumbauld et al. (1997) in which experimental plots were treated with carbaryl to remove for subseexisting shrimp and then monitored Ghost shrimp density rose quent recolonization. from 6 shrimp m-2 1 d after plots were sprayed to 468 shrimp m-2 1 yr later (primarily due to high and remained high up to 3 yr postrecruitment) treatment (Fig. 13). In contrast, density of mud shrimp remained at or below 8 shrimp m-2 during the entire post-treatment period (Fig. 13). Results of a recruitment experiment by Dumbauld (1994) suggest that adult conspecifics or other attributes associated with the presence of mud shrimp may be important for successful larval settlement or post-settlement survival of 0+ mud shrimp, whereas removal of pre-existing adult ghost shrimp can actually enhance densities of 0+ ghost shrimp recruits, perhaps by reducing intraspecific competition. Dumbauld (1994) discovered a density-dependent relationship between the numbers of 1+ and adult (> 1+) ghost shrimp whereby higher densities of adult shrimp were associated with lower densities of 1+ juveniles (Fig. 14). These patterns imply that ghost shrimp densities can quickly exceed the minimum threshold of 10 burrows m-2 for spray within 1 yr of treatment while mud shrimp densities may remain below this threshold for several years following carbaryl treatment. Fur- I 0 ,I I 40 I I I I I 80 * I I . I 120 . I 160 l 200 >1 + Shrimp (number m2) Fig. 14. Density-dependent function between the number of 1 + juvenile ghost shrimp (Neotrypaea californiensis) and > 1+ shrimp (i.e., 2+ and older) in benthic core samples, 1 yr after carbaryl treatment at an experimental site near the Palix River, Willapa Bay, in 1989. (Source: Dumbauld 1994). thermore, chemical application may influence shrimp distribution patterns and promote the conversion of mud shrimp beds to ghost shrimp habitat, as observed on occasion (Dumbauld 1994). Thus, while carbaryl is an effective short-term solution to shrimp control, rapid recolonization of treated beds and altered species distributions favoring expansion of ghost shrimp may create a cycle of more frequent use. Given continued use of carbaryl in oyster culture practices, we recommend small-scale experiments be done to examine ecological effects of delaying treatment of ghost shrimp beds until early fall. By shifting the timing of carbaryl application to October on fallow oyster beds where ghost shrimp are dominant, the spray would eliminate much of the newly settled year class in addition to older year classes, thereby reducing initial loss of seed planted in the subsequent spring and perhaps lengthening the time interval between sprayings. While Buchanan et al. (1985) and some growers have advocated experimentally shifting the timing of spray to early spring to reduce the loss of seed planted in March and April, the concurrent outmigration of juvenile salmonids and impacts to their prey resources as well as low water and sediment temperatures (5.8?C and 7.5?C, respectively in Oregon during February/March; Karinen et al. 1967), have largely suppressed further exploration of that prospect. An experimental investigation of an October application specifically for ghost shrimp beds, however, has merit. In addition to a greater reduction in 0+ ghost shrimp densities and greater seed sur- Oysters,Crabs,andBurrowing Shrimp vival, fewer 0+ and subadult Dungeness crabs would likely be killed, as the majority of crabs would have moved into subtidal channels by fall or migrated out of the estuary to the nearshore coast. The principal concerns of a fall application would include fewer daylight low tides, increased risk of storm systems delaying application, the potential for longer-term chemical persistence and reduced efficacy due to lower water temperatures (14.7?C in October as opposed to 19.5?C in July in Willapa Bay; Tufts 1989), and interaction with fall migration patterns of shorebirds. All of these issues would need to be reviewed and addressed through controlled and monitoring experimentation through the oyster harvest cycle to determine the costs and benefits of a spray schedule tailored to the recruitment period of each species of shrimp. Oyster beds as a whole are treated with carbaryl on average once every six years although the frequency of application depends on rates of shrimp reinfestation, bed type, and rotation schedule (WDF and WDOE 1992; Simenstad and Fresh 1995). The rate of shrimp colonization may depend largely on interannual fluctuations in postlarval recruitment and subsequent survival of the year class through time. Year class strength of 0+ mud shrimp and ghost shrimp can vary several fold, from an average as low as 0 shrimp m-2 up to 144 shrimp m-2 over the last 9 yr at our longterm sampling stations (Fig. 6). If 0+ year class strength is coupled to survival of the year class in subsequent years, then one might be able to roughly predict the condition of an oyster bed a year or two into the future. The ability to forecast future shrimp densities could improve the grower's ability to manage their crop as well as the use of carbaryl. It is just as likely that biological and environmental factors decouple initial recruitment densities from densities of the year class through time. A study designed to explore that question is currently underway to better understand shrimp population dynamics and its implications for oyster culture (Dumbauld et al. unpublished abstract). Given variability in year class strength and population densities of burrowing shrimp in general, we propose that regulators examine a carbaryl control program in which total number hectares would be allowed to vary from year to year rather than be restricted to a constant maximum. Growers lobbied for years to increase the limit from 162 to 324 ha, which they argued was necessary to maintain oyster production at current levels given the increase and expansion of shrimp populations in recent years. Since 1993 when their request was granted, growers have applied to treat nearly all 324 ha. We expect there may be years when pest densities are relatively low and a 324-ha cap is then 169 excessive. Yet growers are unlikely to voluntarily reduce pesticide use in those years perhaps in fear that regulators will operate under a use it or lose it philosophy. If growers apply to treat substantially fewer hectares than the maximum allowed, regulators may question the veracity of their initial claims. A program based on a sliding scale of treatable hectares would require good data on shrimp population dynamics (recruitment, population growth, mortality) and the relationship between shrimp density and damage to oysters, an active monitoring plan for sampling shrimp on oyster beds, and incentives to encourage growers to reduce pesticide application. The distribution of carbaryl credits might provide such an incentive in which growers would be rewarded for reduced use of carbaryl in years of low shrimp densities with provision to spray a greater number of hectares in years with strong shrimp recruitment and/or high adult densities. For example, if monitoring revealed that beds were in relatively good condition, growers as a whole might only need to spray 160 ha in a given year, but are given credit to treat more hectares in years when shrimp densities are high, up to a predetermined maximum level that may even exceed 324 ha, if deemed environmentally acceptable. Although it would be more costly to implement this type of plan, the intent and benefits of the system would include improved flexibility and responsiveness of the carbaryl program and more judicious and effective use of the chemical based on fluctuations in pest population densities. A number of alternative strategies to control burrowing shrimp have been proposed and to some extent tested over the years including chemical, biological, and physical methods (for a complete description and summary of alternative shrimp control techniques see Table 6.1 in DeWitt et al. 1997). Although the efficacy, technological feasibility, costs, and environmental risks of many of these control methods are unknown, research on physical habitat modification appears somewhat promising for ghost shrimp control under certain conditions. Specifically, research studies have been conducted to determine if placement of excess oyster shell on tideflats (beyond that used to produce seed oysters) reduces settlement and survival of 0+ shrimp by altering the physical and biological attributes of the mud habitat. Results of a field experiment in Grays Harbor indicate that establishment of a thick (10-15 cm) layer of epibenthic shell pavement reduces recruitment of 0+ ghost shrimp but not 0+ mud shrimp relative to bare unmanipulated mud (Fig. 15; Dumbauld et al. In press), due in part to differences in settlement patterns, burrowing abilities, and exposure to 0+ K. L. Feldmanet al. 170 920 A. Ghost shrimp ' E 150- c T T 100- 0O 0) - 50 0 0- Mud Subsurface Epibenthic Myashell shell shell I,~1n B. Mudshrimp Cm - 40- E Q. E 30- C) E T 20- -I 10- 0 0 Mud A Subsurface Epibenthic Myashell shell shell Substrate Fig. 15. Number of 0+ shrimp recruiting to open mud and various categories of shell cover including empty oyster shell buried 5-10 cm below the sediment surface (subsurface shell), surface deposits of oyster shell forming a thick (10-15 cm) layer (epibenthic shell), and relic surface deposits ("death assemblages") of eastern softshell clam Mya arenaria (Mya shell). Newly settled 0+ ghost shrimp (Neotrypaeacaliforniensis)were most abundant in open mud and subsurface shell (A), while 0+ mud shrimp (Upogebiapugettensis)were most abundant in epibenthic oyster shell and surface deposits of M. arenaria (B). (Source: Dumbauld et al. in press). Dungeness crabs that prey on shrimp (Feldman et al. 1997; Feldman unpublished data). In other experiments, application of carbaryl to plots prior to shell deposition to remove existing shrimp increased surface time of shell habitat compared to untreated plots where shell sank rapidly into the sediment, particularly where ghost shrimp were present (Feldman and Dumbauld unpublished data). These experiments provide an example of how chemical (carbaryl application), physical (habitat modification through the use of epibenthic shell), and biological (concentration of 0+ crab predators within the shell matrix) techniques can be integrated into an overall control strategy that may be more effective in combination than any single technique alone and thereby reduce the frequency of and need for control measures. Although application of shell pavement under certain circumstances may effectively reduce densities of ghost shrimp, the approach remains untested and unproven on a commercial scale and over an appropriate time scale. Economic and environmental costs and impacts need to be investigated as well. In a 2-yr field experiment, densities of oyster seed planted on plots treated with carbaryl and shell pavement were similar to or greater than densities of seed on plots treated with carbaryl alone, suggesting that shell pavement could serve as a suitable alternative substrate upon which to grow oysters (Feldman and Dumbauld unpublished data). However, the availability and cost of shell is uncertain as well as its effect on harvest operations. Shell pavement might interfere with harvest dredging by reducing the efficiency of the catch basket: the teeth on the dredge basket may be unable to dislodge the oysters from the surrounding substrate effectively or, on the other hand, the basket may fill with excess empty shell in addition to oysters, depleting the shell pavement substrate and hindering processing operations. There are also environmental impacts to consider in addition to economic and cultural issues. Shell pavementing clearly alters the physical and biological nature of the substrate thereby influencing the composition of flora and fauna found in association with shell habitat, and in areas of the estuary with natural seed set, shell pavement is likely to be transformed into an oyster reef (Dumbauld and Feldman unpublished data). Furthermore, deployment of large-scale shell pavement may encourage the conversion of ghost shrimp beds into mud shrimp beds where shell remains intact above the sediment surface (Fig. 15). Although this technique shows some promise, more research is needed to examine the feasibility of integrating it into oyster culture practices and impacts to estuarine habitats and biota. Conflicts involving oyster culture and burrowing shrimp control are complex and not easily resolved. Even with additional research, concerns about oyster culture practices on ecosystem functions will likely persist and the issue will ultimately rest largely upon environmental and economic values and the ecological costs and benefits associated with oyster culture, burrowing shrimp, and control measures, which are difficult if not impossible to quantify. At present, carbaryl is the only effective and proven tool to control burrowing shrimp on a commercial scale. The experimental data and suggested research studies presented in this paper, in Oysters,Crabs,and BurrowingShrimp conjunction with an active monitoring plan, should help advance development of economically and environmentally sound policies for shrimp control. Cooperation among oyster growers, natural resource agencies, scientists, and local communities is essential in the effort to identify practical and sustainable solutions to minimizing impacts of burrowing shrimp on commercial oyster beds. ACKNOWLEDGMENTS This work was done prior to one of the authors (T. H. DeWitt) joining the U.S. Environmental Protection Agency. This material has been subject to the Agency's peer and administrative review, and it has been approved for publication as an EPA document. Mention of trade names or commercial products does not constitute endorsement or recommendation for use. Financial support was provided by Grant Nos. 93-38500-8588, 94-38500-0049, 95-38500-1458, 96-38500-2674, and 97-385004041 from the United States Department of Agriculture to the Western Regional Aquaculture Center, by the Washington Department of Fish and Wildlife by Washington Sea Grant (NA36RG007-01,R/ES-1), and by the Pacific Northwest Ecosystem Regional Study. We thank Michael Herrle and Bruce Kauffman for field assistance and laboratory sample processing. Special thanks to oyster growers Leonard Bennett, Dr. Richard Wilson of Bay Center Mariculture, Lee Weigardt of Jolly Roger Seafoods, and Tim Morris of Coast Oyster Company for their insights and the use of their facilities and oyster beds. We also thank Steve Barry of the Washington Department of Fish and Wildlife and four anonymous reviewers for constructive comments on the manuscript. LITERATURE CITED B. R., D. SKINNER, ANDJ. IMBERGER. 1992. Flow in ALLANSON, prawn burrows. Estuarine, Coastal and ShelfScience35:253-266. AMBROSE, JR., W. G. 1991. Are infaunal predators important in structuring marine soft-bottom communities? AmericanZoologist 31:849-860. R. F. ANDT. W. ANDERSON. 1990. Influence of an arAMBROSE, tificial reef on the surrounding infaunal community. Marine Biology107:41-52. ANDS. B. MATHEWS. 1995. ARMSTRONG, J. L., D. A. ARMSTRONG, Food habits of estuarine staghorn sculpin, Leptocottusarmatus, with focus on consumption of juvenile Dungeness crab, Cancer magister.FisheryBulletin 93:456-470. D. A. AND D. R. GUNDERSON. 1985. The role of ARMSTRONG, estuaries in Dungeness crab early life history: A case study in Grays Harbor, Washington, p. 145-170. In Proceedings of the Symposium on Dungeness Crab Biology and Management. Lowell Wakefield Fisheries Symposium Series, Alaska Sea Grant Report No. 85-3. Fairbanks, Alaska. D. A., 0. IRIBARNE, P. A. DINNEL,K. A. MCGRAW,J. ARMSTRONG, A. SHAFFER, R. PALACIOS, M. FERNANDEZ, K. FELDMAN, ANDG. WILLIAMS. 1992. Mitigation of Dungeness crab, Cancermagister, Losses Due to Dredging in Grays Harbor by Development of Intertidal Shell Habitat: Pilot Studies During 1991. FRI-UW9205. University of Washington, Seattle, Washington. D. A. ANDR. E. MILLEMANN. 1974. Effects of the ARMSTRONG, insecticide carbaryl on clams and some other intertidal mud flat animals. Journal of theFisheriesResearchBoard of Canada 31: 466-470. AND S. ABOOBUCKER. 1980. ARUNACHALAM, S., K. JEYALAKSHMI, Toxic and sublethal effects of carbaryl on a freshwater catfish, Mystus vittatus (Bloch). Archivesof EnvironmentalContamination and Toxicology9:307-316. 171 A. B. 1985. The use of Sevin on estuarine oyster beds BAKALIAN, in Tillamook Bay, Oregon. CoastalZoneManagementJournal13: 49-83. BAKER,P. 1995. Review of ecology and fishery of the Olympia oyster, Ostrea lurida with annotated bibliography. Journal of ShellfishResearch14:501-518. 1993. No BARILLE, L., J. PROU, M. HERAL,AND S. BOURGRIER. influence of food quality, but ration-dependent retention efficiencies in the Japanese oyster Crassostreagigas. Journal of ExperimentalMarine Biologyand Ecology171:91-106. S. 1993. Target and non-target effects of antiBHATTACHARYA, cholinesterase pesticides in fish, p. 859-866. In W. Sloof and H. de Kruijf (eds.), Proceedings of the Second European Conference on Ecotoxicology. Elsevier Science, Amsterdam. BIRD, E. M. 1982. Population dynamics of thalassinidean shrimps and community effects through sediment modification. Ph.D. Dissertation, University of Maryland, College Park, Maryland. 1988. Absence of preemption BLACK,R. ANDC. H. PETERSON. and interference competition for space between large suspension-feeding bivalves and smaller infaunal macroinvertebrates. Journal of ExperimentalMarine Biologyand Ecology120:183-198. BRANCH,G. M. ANDA. PRINGLE.1987. The impact of the sand prawn Callianassa kraussiStebbing on sediment turnover and on bacteria, meiofauna, and benthic microflora. Journal of ExperimentalMarine Biologyand Ecology107:219-235. D. L. 1991. Settlement patterns and presettlement BREITBURG, behavior of the naked goby, Gobiosomabosci,a temperate oyster reef fish. Marine Biology109:213-221. G. A. 1981. Disturbance and community structure: BRENCHLEY, An experimental study of bioturbation in marine soft-sediment environments. Journal of Marine Research39:767-790. BRENCHLEY,G. A. 1982. Mechanisms of spatial competition in marine soft-bottom communities. Journal of ExperimentalMarine Biologyand Ecology60:17-33. BRICELJ, V. M., R. E. MALOUF, AND C. DE QUILLFELDT. 1984. Growth of juvenile Mercenariamercenariaand the effect of resuspended bottom sediments. Marine Biology84:167-173. BROOKS,K. M. 1993. Changes in Arthropod and Mollusk Populations Associated with the Application of Sevin to Control Burrowing Shrimp in Willapa Bay, July to September, 1992. Report to the Pacific County Economic Development Council, Raymond, Washington. 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 (Crassostreagigas) Beds in Willapa Bay, Washington, to Fulfill Requirements of the EPA Carbaryl Data Call In. Final Report to the U.S. Environmental Protection Agency, Region 10. Seattle, Washington. D. V., D. L. BOTTOM,ANDD. A. ARMSTRONG. BUCHANAN, 1985. The Controversial Use of the Insecticide Sevin in Pacific Northwest Estuaries: Its Effects on Dungeness Crab, Pacific Oyster, and Other Species. Proceedings of the Symposium on Dungeness Crab Biology and Management. Alaska Sea Grant Report 85-3, University of Alaska, Fairbanks, Alaska. D. V., R. E. MILLEMANN, ANDN. E. STEWART. BUCHANAN, 1970. Effects of the insecticide Sevin on various stages of the Dungeness crab, Cancer magister Journal of the Fisheries Research Board of Canada 27:93-104. BURROWING SHRIMP COMMITTEE. 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. Report to Grays Harbor and Pacific County Commissioners. South Bend, Washington. CAPALDO, P. S. 1987. Effects of carbaryl (Sevin) on the stage 1 172 K. L. Feldmanet al. zoeae of the red-jointed fiddler crab, Uca minax (LeConte). Estuaries 10:132-135. R. AND S. HARBELL.1992. Economic Assessment of CARKNER, Bottom Versus Off Bottom Oyster Culture in Grays Harbor and Willapa Bay, Washington. Final Report. Washington State Cooperative Extension Agency, Washington State University, Puyallup, Washington. CARLSON, R. W., S. P. BRADBURY, R. A. DRUMMOND, AND D. E. HAMMERMEISTER. 1998. Neurological effects on startle response and escape from predation by medaka exposed to organic chemicals. Aquatic Toxicology43:51-68. C. P., C. S. WEIL,P. E. PALM,M. W. WOODSIDE,J.H. CARPENTER, NAIRIII, ANDH. F. SMYTH, JR. 1961. Mammalian toxicity of 1naphthyl-N-methylcarbamate (Sevin insecticide). Journal ofAgricultural and Food Chemistry9:30-39. CHEW,K. K. AND R. EISLER.1958. A preliminary study of the Japanese oyster drill Ocinebrajaponica. Journal of the Fisheries ResearchBoard of Canada 15:529-535. CLIFTON,H. E.. 1983. Discrimination between subtidal and intertidal facies in Pleistocene deposits, Willapa Bay, Washington. Journal of SedimentaryPetrology53:353-369. N. ANDG. C. POORE.1980. The distribution of CalliCOLEMAN, anassa species (Crustacea: Decapoda) in Western Port Victoria, Australia. Proceedingsof the Royal Societyof Victoria91:7378. COMMITO, J. A. ANDE. M. BONCAVAGE.1989. Suspension-feeders and coexisting infauna: An enhancement counterexample. Journal of ExperimentalMarine Biologyand Ecology125:33-42. ANDR. L. RALONDE.1996. Oyster CONTE, F. S., S. C. HARBELL, Culture: Fundamentals and Technology of the West Coast Industry. Western Regional Aquaculture Center Publication 94101, Seattle, Washington. CONTI, E. 1987. Acute toxicity of three detergents and two insecticides in the lugworm, Arenicolamarina (L.): A histological and a scanning electron microscopic study. Aquatic Toxicology 10:325-334. CONWAY, JR., R. S. 1991. The Economic Impact of the Oyster Industry. Report to the Willapa-Grays Harbor Oyster Growers Association, Washington State Department of Community Development, and Pacific County Economic Development Council Port of Peninsula, Washington. CORREDOR, J. E. ANDJ. M. MORELL.1989. Assessment of inorganic nitrogen fluxes across the sediment-water interface in a tropical lagoon. Estuarine, Coastal and Shelf Science28:339345. M. F. 1986. Carbaryl: A toxicological review and risk CRANMER, 7:247-332. analysis. Neurotoxicology L. L. ANDE. F. HURLBURT. 1987. Control of BurrowCREEKMAN, ing Shrimp on Oyster Beds in Willapa Bay and Grays Harbor 1985. Washington State Department of Fisheries, Special Shellfish Report 3. Olympia, Washington. DAME,R. 1996. Ecology of Marine Bivalves: An Ecosystem Approach. CRC Press, New York. DARLING, J. D. 1984. Gray whales off Vancouver Island, British Columbia, p. 267-287. In M. L. Jones, S. L. Swartz, and S. Leatherwood (eds.), The Gray Whale: Eschrichtiusrobustus.Academic Press, Orlando. ANDE. THIEBAUT.1998. Infaunal preDESROY,N., C. RETIERE, dation regulates benthic recruitment: An experimental study of the influence of the predator Nephtys hombergii(Savigny) on recruits of Nereis diversicolor(O.F. Muller). Journal of Experimental Marine Biologyand Ecology228:257-272. DEVINE, C. E. 1966. Ecology of Callianassafilholi Milne-Edwards 1878 (Crustacea, Thalassinidea). Transactionsof the Royal Society of New Zealand Zoology8:93-110. DEWITT, T. H., K. F. WELLMAN, T. WILDMAN, D. A. ARMSTRONG, ANDL. BENNETT.1997. An Evaluation of the Feasibility of Using Integrated Pest Management to Control Burrowing Shrimp in Commercial Oyster Beds. Prepared for the Wash- ington Department of Ecology by Battelle, Pacific Northwest Division, Richland, Washington. S. 1996. Effects of macrobenthic burrows on infauDITTMANN, nal communities in tropical tidal flats. Marine EcologyProgress Series134:119-130. 1988. Callianassa trilobata(CrusDOBBS, F. C. ANDJ. B. GUCKERT. tacea: Thalassinidea) influences abundance of meiofauna and biomass, composition, and physiologic state of microbial communities within its burrow. MarineEcologyProgressSeries45:6979. DONKIN,P. 1997. Effect of neurotoxic pesticides on the feeding rate of marine mussels (Mytilus edulis). PesticideScience49:196209. DOTY, D. C., D. A. ARMSTRONG, AND B. R. DUMBAULD. 1990. Comparison of Carbaryl Pesticide Impacts on Dungeness Crab (Cancer magister) Versus Benefits of Habitat Derived From Oyster Culture in Willapa Bay, Washington. Report to the Willapa Bay/Grays Harbor Oyster Growers Association. FRI-UW-9020. University of Washington, Seattle, Washington. B. R. 1994. Thalassinid shrimp ecology and the use DUMBAULD, of carbaryl to control populations on oyster ground in Washington coastal estuaries. Ph.D. Dissertation, University of Washington, Seattle, Washington. B. R. ANDD. A. ARMSTRONG. 1987. Potential MitiDUMBAULD, gation of Juvenile Dungeness Crab Loss During Dredging Through Enhancement of Intertidal Shell Habitat in Grays Harbor, Washington. Final Report to the Army Corps of Engineers, Seattle District Office. FRI-UW-8714. University of Washington, Seattle, Washington. B. R., D. A. ARMSTRONG, ANDK. L. FELDMAN. DUMBAULD, 1996. Life-history characteristics of two sympatric thalassinidean shrimps, Neotrypaeacaliforniensisand Upogebiapugettensiswith implications for oyster culture. Journal of CrustaceanBiology16: 689-708. B. R., D. A. ARMSTRONG, ANDT. L. MCDONALD. DUMBAULD, 1993. Use of oyster shell to enhance intertidal habitat and mitigate loss of Dungeness crab (Cancer magister)caused by dredging. Canadian Journal of Fisheriesand Aquatic Sciences50:381-390. B. R., D. A. ARMSTRONG,ANDJ. SKALSKI. 1997. EffiDUMBAULD, cacy of the pesticide carbaryl for thalassinid shrimp control in Washington State oyster (Crassostreagigas, Thunberg 1793) aquaculture. Journal of ShellfishResearch16:503-518. B. R., E. P. VISSER,D. A. ARMSTRONG, L. COLE-WARNDUMBAULD, ANDB. E. KAUFFMAN. In press. Use of oyster ER,K. L. FELDMAN, shell to create habitat for juvenile Dungeness crab in Washington coastal estuaries: Status and prospects. Journal of Shellfish Research. P. C. 1987. The biology of Upogebiapusilla (PetagDWORSCHAK, na) (Decapoda, Thalassinidea) II. Environments and zonation. Marine Ecology8:337-358. ECKMAN, J. E. 1985. Flow disruption by an animal-tube mimic affects sediment bacterial colonization. Journal of Marine Research43:419-435. ECKMAN, J. E., A. R. M. NOWELL,ANDP. A. JUMARS.1981. Sediment destabilization by animal tubes. Journal of Marine Research39:361-374. 1995. Larval supply, EGGLESTON, D. E. AND D. A. ARMSTRONG. active substrate selection, and post-settlement survival as determinants of estuarine Dungeness crab recruitment. Ecological Monographs65:193-216. ESTES,P. S. 1986. Cardiovascular and respiratory responses of the ghost shrimp, Callianassa californiensisDana, to the pesticide carbaryl and its hydrolytic product 1-napthol. M.S. Thesis, Oregon State University, Corvallis, Oregon. EVERETT, R., G. Ruiz, ANDJ. T. CARLTON. 1995. Effect of oyster mariculture on submerged aquatic vegetation: An experimental test in a Pacific Northwest estuary. Marine EcologyProgress Series125:205-217. ANDL. C. MCEWEN.1995. Effects of FAIR,J. M., P. L. KENNEDY, Oysters,Crabs,and BurrowingShrimp carbaryl grasshopper control on nesting killdeer in North Dakota. EnvironmentalToxicologyand Chemistry14:881-890. FELDER,D. L. AND D. L. LOVETT.1989. Relative growth and sexual maturation in the estuarine ghost shrimp Callianassa louisianensis Schmitt, 1935. Journal of CrustaceanBiology9:540-553. ANDB. R. D. E. EGGLESTON, K. L., D. A. ARMSTRONG, FELDMAN, 1997. Effects of substrate selection and post-setDUMBAULD. tlement survival on recruitment success of the thalassinidean shrimp, Neotrypaeacaliforniensis, to intertidal shell and mud habitats. Marine EcologyProgressSeries150:121-136. 1993. Habitat FERNANDEZ,M., 0. IRIBARNE,ANDD. ARMSTRONG. selection by young-of-the-year Dungeness crab Cancermagister and predation risk in intertidal habitats. MarineEcologyProgress Series92:171-177. S. ANDG. GRAF.1992. Continuously measured changes FORSTER, in redox potential influenced by oxygen penetrating from burrows of Callianassa subterranea.Hydrobiologia235:527-532. S. ANDG. GRAF. 1995. Impact of irrigation on oxygen FORSTER flux into the sediment: Intermittent pumping by Callianassa subterraneaand "piston-pumping" by Lanice conchilega.Marine Biology123:335-346. FUKUTO,T. R. 1990. Mechanism of action of organophosphorous and carbamate insecticides. EnvironmentalHealth Perspectives 87:245-54. T. L., L. C. MCEWEN,ANDA. FOWLER. 1992. Effects of GEORGE, a carbaryl bait treatment on nontarget wildlife. Environmental Entomology21:1239-1247. 1996. Oysters and the GOTTLEIB, S.J. ANDM. E. SCHWEIGHOFER. Chesapeake Bay ecosystem: A case for exotic species introduction to improve environmental quality? Estuaries 19:639650. GRANT,J. ANDB. THORPE.1991. Effects of suspended sediment on growth, respiration, and excretion of the soft shelled clam (Mya arenaria). CanadianJournal of Fisheriesand Aquatic Sciences 48:1285-1292. R. B. 1988. Behavioral interactions between burrowing GRIFFIS, shrimp: Effects on distribution of sympatric thalassinids. American Zoologist28:179A. 1988. Effects of sediment type GRIFFIS,R. B. ANDF. L. CHAVEZ. on Callianassa californiensisDana and C. gigas Dana. Journal of ExperimentalMarine Biologyand Ecology117:239-253. 1991. A model of burrow GRIFFIS,R. B. ANDT. H. SUCHANEK. architecture and trophic modes in thalassinidean shrimp (Decapoda: Thalassinidea). Marine EcologyProgressSeries79:171183. GUNDERSON, D. R., D. A. ARMSTRONG, Y. SHI, AND R. A. McCONNAUGHEY.1990. Patterns of estuarine use by juvenile En- glish sole (Parophrysvetulus) and Dungeness crab (Cancermagister). Estuaries13:59-71. 1961. The biology of CalHAILSTONE,T. S. ANDW. STEPHENSON. lianassa (Trypaea) australiensisDana 1852 (Crustacea, Thalassinidea). Universityof QueenslandPapersDepartmentof Zoology1: 257-286. 1981. Nutrient enrichHARLAN,M. M. ANDB. THRONE-MILLER. ment of seagrass beds in a Rhode Island coastal lagoon. Marine Biology65:221-229. HARRISON, P. G. 1987. Natural expansion and experimental manipulation of seagrass (Zostera spp.) abundance and the response of infaunal invertebrates. Estuarine, Coastal and Shelf Science24:799-812. HART,J. L. 1937. Larval and adult stages of British Columbia anomura. CanadianJournal of Research15:179-220. HART,J. L. 1974. Pacific fishes of Canada. FisheriesResearchBoard of Canada Bulletin 180:1-740. HEDGPETH, J. ANDS. OBREBSKI. 1981. Willapa Bay: A Historical Perspective and a Rationale for Research. Office of Biological Services, U.S. Fish and Wildlife Service, FWS/OBS-81/03, Washington, D.C. HINES,A. H., M. H. POSEY,ANDP. J. HADDON.1989. Effects of 173 adult suspension- and deposit-feeding bivalves on recruitment of estuarine infauna. Veliger32:109-119. ANDE. A. WILSON. E. E., E. N. POWELL, HOFMANN, J. M. KLINCK, 1992. Modeling oyster populations III. Critical feeding periods, growth and reproduction. Journal of ShellfishResearch11: 399-416. HOINES,L. 1996. Washington State Fisheries Statistical Report: 1993. Washington State Department of Fish and Wildlife, Olympia, Washington. A. REAMAN,K. LI, ANDR. G., B. BENSON,D. PENTILLA, HUECKEL, 1988. Pilot Studies on Sevin Impacts to Marine Fish BUCKLEY. in Willapa Bay and Grays Harbor in 1987-1988. Washington State Department of Fisheries, Marine Fish Program, Habitat Investigations Unit, Olympia, Washington. HUNTER, M. L. ANDJ. W. WITHAM.1985. Effects of a carbarylinduced depression of arthropod abundance on the behavior of Parulinae warblers. Canadian Journal of Zoology63:26122616. HUNTER, M. L., J. W. WITHAM, ANDH. Dow. 1984. Effects of a carbaryl-induced depression in invertebrate abundance on the growth and behavior of American black duck and mallard ducklings. Canadian Journal of Zoology62:452-456. E. F. 1986. Control of Burrowing Shrimp on Oyster HURLBURT, Beds in Willapa Bay and Grays Harbor 1984. Washington State Department of Fisheries, Special Shellfish Report 2. Olympia, Washington. IRIBARNE,O., D. ARMSTRONG,AND M. FERNANDEZ.1995. Environmental impact of intertidal juvenile dungeness crab habitat enhancement: Effects on bivalves and crab foraging rate. Journal of ExperimentalMarine Biologyand Ecology192:173-194. JOHNSON,G. E. ANDJ. J. GONOR. 1982. The tidal exchange of Callianassa californiensis (Crustacea, Decapoda) larvae between the ocean and Salmon River estuary, Oregon. Estuarine, Coastal and Shelf Science14:501-516. P. A. ANDA. R. M. NOWELL.1984. Effects of benthos on JUMARS, sediment transport: Difficulties with functional grouping. Continental Shelf Research3:115-130. KARINEN, J. F., J. G. LAMBERTON, N. E. STEWART, AND L. C. TERRIERE. 1967. Persistence of carbaryl in the marine estuarine environment. Chemical and biological stability in aquarium systems. Journal of Agriculturaland Food Chemistry15:148-156. KAUR,K. AND A. DHAWAN.1996. Effect of carbaryt on tissue composition, maturation, and breeding potential of Cirrhina mrigala (Ham.). Bulletin of Environmental Contamination and Toxicology57:480-486. KENTZER-BACZEWSKA, A., A. KOSAKOWSKA,AND L. FALKOWSKI. 1984. Effect of carbaryl and 1-napthol on selected phytoplanktonic algae. Limnologica15:585-589. KOIKE,I. ANDH. MUKAI.1983. Oxygen and inorganic nitrogen contents and fluxes in burrows of the shrimps Callianassajaponica and Upogebiamajor.MarineEcologyProgressSeries12:185190. LARKIN,M. J. AND M. J. DAY. 1985. The effect of pH on the selection of carbaryl-degrading bacteria from garden soil. Journal of Applied Bacteriology58:175-185. R. ANDS. DEA. RODRIGUES. 1991. LepidophthalmussinLEMAITRE, uensis: A new species of ghost shrimp (Decapoda: Thalassinidea: Callianassidae) of importance to the culture of penaeid shrimps on the Caribbean coast of Colombia, with observations on its ecology. FisheryBulletin 89:623-630. LITTLE,E. E. 1990. Behavioral indicators of sublethal toxicity in rainbow trout. Archivesof EnvironmentalContaminationand Toxicology19:380-385. 1981. A Review of Water CharLOEHR,L. C. ANDE. E. COLLIAS. acteristics of Grays Harbor 1938-1979 and an Evaluation of Possible Effects of the Proposed Channel Widening and Deepening Project Upon Present Water Characteristics. Final Report for Contract DACW67-80-C-0009. Grays Harbor and Chehalis River Improvements to Navigation Environmental 174 K.L. Feldmanet al. Studies. Seattle District of the U.S. Army Corps of Engineers, Seattle, Washington. 1948. Effect of suspended V. L. ANDF. D. TOMMERS. LOOSANOFF, silt and other substances on rate of feeding of oysters. Science 107:69-70. 1979. Age-groups from P. D. M. ANDT. J. PITCHER. MACDONALD, size-frequency data: a versatile and efficient method of analyzing distribution mixtures. Journal of the Fisheries Research Board of Canada 36:987-1001. G. E. 1930. The natural history of the mud shrimp MACGINITIE, Upogebiapugettensis(Dana). Annals and Magazine of Natural History 6:37-45. MACGINITIE, G. E. 1934. The natural history of Callianassa californiensis (Dana). AmericanMidland Naturalist 15:166-177. MACKENZIE,JR., C. L. 1996. History of oystering in the United States and Canada, featuring the eight greatest oyster estuaries. Marine FisheriesReview58:1-78. MANNING,R. B. ANDD. L. FELDER. 1991. Revision of the American Callianassidae (Crustacea: Decapoda: Thalassinidea). Proceedingsof the Biological Societyof Washington104:764-792. MAUER,D. 1983. The effect of an infaunal suspension feeding bivalve Mercenariamercenaria(L.) on benthic recruitment. Marine Ecology4:263-274. MCCONNAUGHEY,R. A., D. A. ARMSTRONG,B. M. HICKEY,AND D. R. GUNDERSON. 1992. Juvenile Dungeness crab (Cancer magister) recruitment variability and oceanic transport during the pelagic larval phase. Canadian Journal of Fisheriesand Aquatic Sciences49:2028-2044. L. T. 1972. The ghost shrimp, Callianassa californiensis MCCROW, Dana, 1854, in Yaquina Bay, Oregon. M.S. Thesis, Oregon State University, Corvallis, Oregon. ANDP. A. DINNEL.1995. R. O., D. A. ARMSTRONG, MCMILLAN, Comparison of intertidal habitat use and growth rates of two northern Puget Sound cohorts of 0+ age Dungeness crab, CancermagisterEstuaries18:390-398. M. F. 1984. Bioturbation of intertidal quartz-rich sands: MILLER, A modern example and its sedimentologic and paleoecologic implications. Journal of Geology92:201-216. MOUNT,M. E. AND F. W. OEHME.1981. Carbaryl: A literature review. ResidueReviews80:1-64. MURPHY,R. C. 1985. Factors affecting the distribution of the introduced bivalve, Mercenaria mercenaria,in a California lagoon-The importance of bioturbation. Journal of Marine Research43:673-692. 1992. Benthic community meR. C. ANDJ. N. KREMER. MURPHY, tabolism and the role of deposit-feeding callianassid shrimp. Journal of Marine Research50:321-340. 1998. Impacts of burrowing ghost NATES,S. F. ANDD. L. FELDER. shrimp, genus LepidophthalmusCrustacea: Decapoda: Thalassinidea, on penaeid shrimp culture. Journal of the WorldAquaculture Society29:188-210. NEWELL,R. I. E. 1988. Ecological changes in Chesapeake Bay: Are they the result of overharvesting the American oyster, Crassostreavirginica?, p. 536-546. In M. P. Lynch and E. C. Krome (eds.), Understanding the Estuary: Advances in Chesapeake Bay Research. Chesapeake Research Consortium, Solomons, Maryland. NICKELL,L. A. AND R. J. A. ATKINSON.1995. Functional morphology of burrows and trophic modes of three thalassinidean shrimp species, and a new approach to the classification of thalassinidean burrow morphology. Marine EcologyProgress Series128:181-197. NOSHO,T. 1989. Small-scale Oyster Farming for Pleasure and Profit. Washington Sea Grant, Aquaculture Series 89-1, Marine Advisory Services, Seattle, Washington. ANDR. MANN.1982. Benthic filter C. B., T. J. SMAYDA, OFFICER, feeding: A natural eutrophication control. Marine EcologyProgress Series9:203-210. 1973. Grazing by OGDEN,J. C., R. A. BROWN,AND N. SALESKY. the echinoid Diadema antillarum Philippi: Formation of halos around West Indian patch reefs. Science182:715-717. OF FISH ANDWILDLIFE. 1977. IdentificaOREGONDEPARTMENT tion, Distribution, and Notes on Food Habits of Fish and Shellfish in Tillamook Bay, Oregon. Federal Aid Progress Report. Newport, Oregon. ORTH, R. J. 1975. Destruction of eelgrass Zosteramarina by the cownose ray, Rhinopterabonasus, in the Chesapeake Bay. ChesapeakeScience16:205-208. ANDK. A. MOORE.1994. Seed disORTH,R. J., M. LUCKENBACH, persal in a marine macrophyte: Implications for colonization and restoration. Ecology75:1927-1939. ANDD. A. ARMSTRONG. 1994. SeaPALACIOS, R., J. M. ORENSANZ, sonal and life-long variation of Sr/Ca ratio in shells of Mya arenaria from Grays Harbor (Washington)-An ancillary criterion in demographic studies. Estuarine, Coastal and ShelfScience 39:313-327. B. T. ANDA. M. OLSON.1995. Policies and management PAWLAK, practices of Washington State agencies as they pertain to the seagrasses, Zosteramarina and Zosterajaponica: An ecological analysis, p. 516-522. In Puget Sound Research Proceedings. Puget Sound Water Quality Authority, Olympia, Washington. C. H. 1977. Competitive organization of the soft-botPETERSON, tom macrobenthic communities of Southern California lagoons. Marine Biology43:343-359. C. H. 1984. Does a rigorous criterion for environPETERSON, mental identity preclude the existence of multiple stable points? AmericanNaturalist 124:127-133. C. H. AND R. BLACK.1987. Resource depletion by PETERSON, active suspension feeders on tidal flats: Influence of local density and tidal elevation. Limnology and Oceanography32:143166. C. H. ANDR. BLACK. 1991. Preliminary evidence for PETERSON, progressive sestonic food depletion in incoming tide over a broad tidal sand flat. Estuarine, Coastaland ShelfScience32:405413. H. G., C. BOUTIN,P. A. MARTIN, K. E. FREEMARK, N. PETERSON, AND M. J. MOODY.1994. Aquatic phyto-toxicity of J. RUECKER, 23 pesticides applied at expected environmental concentrations. Aquatic Toxicology28:275-292. R. C. 1984. The Ecology of Eelgrass Meadows in the PHILLIPS, Pacific Northwest: A Community Profile. Office of Biological Services, U.S. Fish and Wildlife Service, FWS/OBS-84/24, Washington, D.C. PIMENTEL,G. E. 1983. Recruitment of larvae of the ghost shrimp, Callianassa californiensis(Crustacea, Decapoda) in the South Slough estuary, Oregon. M.S. Thesis, Oregon State University, Corvallis, Oregon. POSEY,M. H. 1985. The effects upon the macrofaunal community of a dominant burrowing deposit feeder, Callianassa californiensis, and the role of predation in determining its intertidal distribution. Ph.D. Dissertation, University of Oregon, Eugene, Oregon. POSEY,M. H. 1986a. Changes in a benthic community associated with dense beds of a burrowing deposit feeder, Callianassa californiensis.Marine EcologyProgressSeries31:15-22. POSEY, M. H. 1986b. Predation on a burrowing shrimp: Distribution and community consequences. Journal of Experimental Marine Biologyand Ecology103:143-161. POSEY, M. H. 1987. Effects of lowered salinity on activity of the ghost shrimp Callianassa californiensis.NorthwestScience61:9396. POSEY,M. H. 1990. Functional approaches to soft-substrate communities: How useful are they? Reviews in Aquatic Sciences2: 343-356. 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. Journal of Experimental Marine Biologyand Ecology148:283-294. Oysters,Crabs,and BurrowingShrimp R. R. 1974. The functional morphology of the fore-guts POWELL, of the thalassinid crustaceans, Callianassa californiensis and Upogebiapugettensis. Universityof California Publications in Zoology 102:1-41. M. M. 1993. Regrowth and recruitment of eelgrass PREGNALL, (Zosteramarina) and recovery of benthic community structure in areas disturbed by commercial oyster culture in the South Slough National Estuarine Research Reserve, Oregon. M.S. Thesis, Bard College, Annandale-on-Hudson, New York. V. D. SINGH, G. P. BRAHRAPRABASH, B. S., B. R. REDDY, RAJAGOPAL, 1984. Effect and persistence of selected ANDN. SETHUNATHAN. carbamate pesticides in soil. ResidueReviews93:1-207. ANDV. K. R. NANDAKUMAR, RAMACHANDRAN, S., N. RAJENDRAN, 1984. Effect of pesticides on photosynthesis VENUGOPALAN. and respiration of marine macrophytes. Aquatic Botany 19: 395-399. RHOADS,D. C. ANDD. K. YOUNG.1970. The influence of deposit feeding organisms on sediment stability and community trophic structure. Journal of Marine Research28:150-178. RICE,D. W. ANDA. A. WOLMAN.1971. The life history and ecology of the gray whale (Eschrichtiusrobustus). SpecialPublication AmericanSocietyof Mammalogists3:1-142. RINGOLD, P. 1979. Burrowing, root mat density, and the distribution of fiddler crabs in the eastern United States. Journal of ExperimentalMarine Biologyand Ecology36:11-21. RONAN,T. E. 1975. Structural and paleo-ecological aspects of a modern marine soft-sediment community: An experimental field study. Ph.D. Dissertation, University of California, Davis, California. Russo, R. A. 1975. Observations on the food habits of leopard sharks (Triakis semifasciata) and brown smooth-hounds (Mustelus henlei). CaliforniaFish and Game61:95-103. SHOTWELL, J. A. 1977. The Willapa Estuary. Background Studies for the Preparation of a Management Plan. Department of Public Works, Planning Division, Pacific County, Washington. SIMENSTAD, C. A. 1994. Faunal associations and ecological interactions in seagrass communities of the Pacific Northwest coast, p. 11-18. In S. Wyllie-Echeverria, A. M. Olson, and M. J. Hershman (eds.), Seagrass Science and Policy in the Pacific Northwest: Proceedings of a Seminar Series. University of Washington, School of Marine Affairs, SMA 94-1. EPA 910/R94-004. Seattle, Washington. C. A. ANDJ. R. CORDELL. 1989. Effects of Sevin ApSIMENSTAD, plication on Littoral Flat Meiofauna: Preliminary Sampling in Willapa Bay,June-July 1988. Technical Report to the Willapa Bay, Grays Harbor Oyster Growers Association. FRI-UW-8904. University of Washington, Seattle, Washington. C. A., J. R. CORDELL, AND W. G. HOOD,J. A. MILLER, SIMENSTAD, R. M. THOM. 1992. Ecological Status of a Created Slough in the Chehalis River Estuary: Report of Monitoring in Created and Natural Estuarine Sloughs, January-December 1991. FRIUW-9206. University of Washington, Seattle, Washington. C. A ANDK. L. FRESH.1995. Influence of intertidal SIMENSTAD, aquaculture on benthic communities in Pacific Northwest estuaries: Scales of disturbance. Estuaries18:43-70. P. V. R. ANDC. A. BUTMAN.1994. Animal-sediment SNELGROVE, relationships revisited: Cause versus effect. Oceanographyand Marine Biology:An Annual Review32:111-177. E. J. ANDJ. J. VIDELER.1998. Burrow ventilation in STAMHUIS, the tube-dwelling shrimp Callianassa subterranea(Decapoda: Thalassinidea). II. The flow in the vicinity of the shrimp and the energetic advantages of a laminar non-pulsating ventilation current. TheJournal of ExperimentalBiology201:2159-2170. ANDP. A. W. J. DE WILDE.1998. E. J., J. J. VIDELER, STAMHUIS, Optimal foraging in the thalassinidean shrimp Callianassa subterranea:Improving food quality by grain size selection. Journal of ExperimentalMarine Biologyand Ecology228:197-208. STEVENS, B. A. 1928. Callianassidae from the West coast of North America. PublicationsPuget Sound BiologicalStation 6:315-369. 175 B. A. 1929. Ecological observations on Callianassidae STEVENS, of Puget Sound. Ecology10:399-405. 1984. Distribution, abunB. G. ANDD. A. ARMSTRONG. STEVENS, dance, and growth of juvenile Dungeness crabs, Cancermagister,in Grays Harbor estuary, Washington. FisheryBulletin 82: 469-483. ANDR. CUSIMANO. B. G., D. A. ARMSTRONG, 1982. FeedSTEVENS, ing habits of the Dungeness crab Cancer magister as determined by the index of relative importance. Marine Biology72: 135-145. ANDW. P. BREESE. 1967. Acute N. E., R. E. MILLEMANN, STEWART, toxicity of the insecticide Sevin and its hydrolytic product 1napthol to some marine organisms. Transactionsof the American FisheriesSociety96:25-30. SUCHANEK, T. H. 1983. Control of seagrass communities and sediment distribution by Callianassa (Crustacea, Thalassinidea) bioturbation. Journal of Marine Research41:281-298. D. D. ANDJ. L. LUTERNAUER.1987. Burrow distriSWINBANKS, bution of thalassinidean shrimp on a Fraser Delta tidal flat, British Columbia. Journal of Paleontology61:315-332. 1981. Biosedimentological D. D. ANDJ. W. MURRAY. SWINBANKS, zonation of Boundary Bay tidal flats, Fraser River Delta, British Columbia. Sedimentology28:201-237. A. 1988. Effects of the bioturbating activity of the ghost TAMAKI, shrimp Callianassa japonica Ortmann on migration of a mobile polychaete. Journal of ExperimentalMarine Biologyand Ecology 120:81-95. A. 1994. Extinction of the trochid gastropod, Umbonium TAMAKI, (Suchium) moniliferum (Lamarck), and associated species on an intertidal sandflat. Researcheson Population Ecology36:225236. TAMAKI, A., H. TANOUE, J. ITOH, ANDY FUKUDA.1996. Brooding and larval developmental periods of the callianassid ghost shrimp, Callianassa japonica (Decapoda: Thalassinidea). Journal of the Marine BiologicalAssociationof the United Kingdom76: 675-689. TASTO,R. N. 1975. Aspects of the biology of Pacific staghorn sculpin, Leptocottusarmatus Girard, in Anaheim Bay. In E. D. Lane and C. W. Hill (eds.), The Marine Resources of Anaheim Bay, California. CaliforniaFish and GameFisheryBulletin 165:135-154. THOMPSON, R. K. 1972. Functional morphology of the hind-gut gland of Upogebiapugettensis(Crustacea, Thalassinidea) and its role in burrow construction. Ph.D. Dissertation, University of California, Berkeley, California. TUFTS,D. F. 1989. The Use of the Insecticide Carbaryl to Control Burrowing Shrimp in Willapa Bay and Grays Harbor, Washington, in 1986. Washington State Department of Fisheries, Special Shellfish Report 4. Olympia, Washington. TUFTS, D. F. 1990. The Use of the Insecticide Carbaryl to Control Burrowing Shrimp in Willapa Bay and Grays Harbor in 1987. Washington State Department of Fisheries, Special Shellfish Report 5. Olympia, Washington. B. 1986. Studies on the population ecology of Upogebia TUNBERG, deltaura (Leach) (Crustacea, Thalassinidea). Estuarine, Coastal and Shelf Science22:753-765. VALENTINE, J. F. ANDK. L. HECK,JR. 1991. The role of sea urchin grazing in regulating subtropical seagrass meadows: Evidence from field manipulations in the northern Gulf of Mexico. Journal of ExperimentalMarine Biologyand Ecology154:215-230. DEPARTMENT OF FISHERIES. WASHINGTON 1970. Ghost Shrimp Control Experiments with Sevin, 1960-1968. Washington Department of Fisheries Technical Report 1. Olympia, Washington. DEPARTMENT OF FISHERIES WASHINGTON DEANDWASHINGTON PARTMENT OF ECOLOGY. 1985. Use of the Insecticide Sevin to Control Ghost and Mud Shrimp in Oyster Beds of Willapa Bay and Grays Harbor. Final Environmental Impact Statement. Olympia, Washington. 176 K. L. Feldmanet al. DEANDWASHINGTON OF FISHERIES DEPARTMENT WASHINGTON OF ECOLOGY. 1992. Use of the Insecticide Carbaryl PARTMENT to Control Ghost and Mud Shrimp in Oyster Beds of Willapa Bay and Grays Harbor. Final Supplemental Environmental Impact Statement. Olympia, Washington. WEITCAMP, L. A., R. C. WISSMAR,C. A. SIMENSTAD, K. L. FRESH, ANDJ. G. ODELL.1992. Gray whale foraging on ghost shrimp (Callianassa californiensis)in littoral sand flats of Puget Sound, U.S.A. Canadian Journal of Zoology70:2275-2280. WILLIAMS,G. D. 1994. Effects of a large-scale estuarine intertidal habitat modification on distribution patterns and food habits of epibenthic fish species in Grays Harbor, Washington. M.S. Thesis, University of Washington, Seattle, Washington. 1989. Some aspects of WITBAARD,R. ANDG. C. A. DUINEVELD. the biology and ecology of the burrowing shrimp Callianassa subterranea (Montagu) (Thalassinidea) from the southern North Sea. Sarsia 74:209-219. WOELKE,C. E. 1957. The flatworm Pseudostylochusostreophagus Hyman, a predator of oysters. Proceedingsof the National ShellfisheriesAssociation47:62-67. WOODIN,S. A. 1976. Adult-larval interactions in dense infaunal assemblages: Patterns of abundance. Journal of Marine Research 34:25-41. 1994. S., A. M. OLSON,ANDM.J. HERSHMAN. WYLLIE-ECHEVERRIA, Seagrass Science and Policy in the Pacific Northwest: Proceedings of a Seminar Series. University of Washington, School of Marine Affairs, SMA 94-1. EPA 910/R-94-004. Seattle, Washington. ZIEBIS, W., S. FORSTER,M. HUETTEL, AND B. B. JORGENSEN. 1996. Complex burrows of the mud shrimp Callianassa truncataand their geochemical impact in the sea bed. Nature 382:619-622. SOURCES OF UNPUBLISHED MATERIALS CHEW,K. Personal Communication. University of Washington School of Fisheries, Seattle, Washington. Abstract. ANDB. F. MACDONALD. B. R., D. P. CHENEY, DUMBAULD, Developing a pest monitoring plan for burrowing shrimp: A step towards integrated pest management. Submitted to Journal of ShellfishResearch. B. R. ANDS. WYLLIE-ECHEVERRIA. DUMBAULD, Unpublished manuscript. Influence of burrowing shrimp control practices on the distribution of intertidal seagrasses in Willapa Bay, Washington. THOM, R. Personal Communication. Battelle Marine Sciences Laboratory, Sequim, Washington. Receivedfor consideration,April 9, 1999 Acceptedfor publication,December21, 1999