Norntates PUBLISHED by the AMERICAN MUSEUM of NATURAL HISTORY CENTRAL PARK WEST at 79TH STREET, NEW YORK, N.Y

AMERICAN MUSEUM Norntates PUBLISHED BY THE AMERICAN MUSEUM OF NATURAL HISTORY CENTRAL PARK WEST AT 79TH STREET, NEW YORK, N.Y. 10024 Number 3042, 17 pp., 12 figures May 7, 1992

Distribution of Ghost Shrimp; North Beach, St. Catherines Island, Georgia

GALE A. BISHOP1 AND ERIC C. BISHOP2

ABSTRACT The Carolinian Ghost Shrimp (Callichirus ma- coast. Burrows are distributed as a burrow strand jor Say 1818) inhabits the open sandy beaches of parallel to the shore, but are somewhat more dif- the Sea Islands of Georgia. Burrows of this fos- fuse than those of the Carolinian Ghost Shrimp sorial thalassinidean are distributed along the fore- and reach much higher maximum densities (up to shore fromjust above mid-tide level to the shallow 483 burrows/M2). In unstable substrates burrow subtidal, forming a narrow band of high-density density of the Carolinian Ghost Shrimp is highly burrow apertures called the burrow strand. Burrow variable and generally reduced. Relict marsh muds, distribution along North Beach was investigated appearing along the lower beach of the erosive by random quadrat sampling every 7.5 m from front ofSt. Catherines Island, prevent colonization high-tide line to low-tide line for 4 km along tran- of such areas by this animal. Burrow distribution sects spaced 100 m apart. The burrow strand has is controlled shoreward by length ofexposure, sea- a steep shoreward (and a gentler seaward) gradient ward by predation, and laterally by substrate grain and exhibits variable densities along its length, size and stability. Use of Ghost Shrimp burrows peaking at approximately 15 burrows/M2. The in interpretation of the geologic record should be Georgian Ghost Shrimp (Callianassa biformis Bif- done with great caution because a broad spectrum far 1970) inhabits sheltered, rippled, muddy-sand of burrow morphologies can be produced by a tidal flats in the lower foreshore, usually bordering single species and many species produce similar sounds between the Sea Islands on the Georgia burrows. INTRODUCTION Fossorial (burrowing) shrimp are among but are seldom seen because oftheir infaunal, the most abundant and common macroin- subterranean mode of life. Yet in the fossil vertebrates in some modem environments, record, callianassid claws and chelipeds are

' Professor of Geology, Georgia Southern University, Statesboro, GA 30460-8149; Research Associate, American Museum of Natural History. 2 Student, Statesboro High School, Statesboro, GA 30458.

WALBURGSHOAL

PICNIC BLUFF

SEASIDEINLEI

McQUEENS INLET

O .5 2ml 0 .5 1 2km

Fig. 1. Index map ofthe southeastern United States, Georgia coast, and St. Catherines Island, Georgia. (from Shadroui, 1990)

among the most common and abundant re- Carolinian Ghost Shrimp; it evolved into a mains of decapod crustaceans. Research on study of the relative sizes, distribution, and the Carolinian Ghost Shrimp (Callichirus ecology of Ghost Shrimp on three Georgia major Say 1818) was initiated in August of barrier islands, Tybee, St. Catherines, and St. 1987, in order to assess that species' appli- Simon's (fig. 1). cability as a model which could be used to The Carolinian Ghost Shrimp (fig. 2) was compare fossil Cretaceous and Tertiary cal- described and named Callianassa major by lianassid claws. This research was subse- Say in 1818. Stimpson (1866) established a quently enlarged to include Callianassa bi- new genus, Callichirus, with Callianassa ma- formis Biffar 1970, when it became apparent jor as its type species. Manning and Felder that the burrow distributions of these two (1986, 1991) reviewed the systematic usage species overlapped. The present research pro- of this species, indicating that it was consid- gram originated as an investigation of the ered by various authors to be a subgenus of range of claw morphology exhibited by the Callianassa, a separate new genus, or a syn- 1 992 BISHOP AND BISHOP: DISTRIBUTION OF GHOST SHRIMP 3

Fig. 2. General morphology ofthe Ghost Shrimp as depicted in dorsal and lateral views. (after Biffar, 1971) onym of Callianassa. In a major systematic iment type and that those of C. californiensis revision, Biffar (1971) reviewed its relation- reached maximum densities of350-450 bur- ships to other callianassids of the southeast- row openings/M2 at low intertidal levels. ern United States. Manning and Felder (1986) The Georgian Ghost Shrimp, Callianassa reassigned it to Callichirus. Aspects of the biformis, described by Biffar (1970) from the ecology ofthe Carolinian Ghost Shrimp have south end of Sapelo Island, McIntosh Coun- been described by Lutz (1937), Pearse et al. ty, Georgia, is known from South Carolina (1942), Pohl (1956), Pryor (1975), Rodrigues to Georgia but probably has a more extensive (1976), Rodrigues and Suguio (1984), Rodri- range. This diminutive ghost shrimp is com- gues and Shimizu (1986), Rodrigues et al. monly found from mid to low-tide levels in (1986a, 1986b), Williams (1984), and Frey et protected areas behind sand bars associated al. (1978). Carolinian Ghost Shrimp were re- with sound channels where it burrows shal- ported to reach densities of 1780 animals per acre on South Carolina beaches (Pohl, 1946: lowly into ripple-marked muddy sand flats. 75). They occur from Beaufort Inlet, North Populations have been observed on the south Carolina to Santa Catarina, Brazil (Rodri- end of Sapelo Island (type locality), on the gues, 1983). Swinbanks and Leuternauer south end ofSt. Simon's Island, on the north (1987) described the distribution of burrows end of St. Catherines Island (this study), on of two other thalassinidean shrimp, Calli- the northern part of the McQueen Inlet ebb anassa californiensis and Upogebia pugetten- delta on St. Catherines Island, and at the sis, on the delta of the Fraser River. They mouth ofBig Bay Creek, Edisto Island, South reported the burrows to be segregated by sed- Carolina. 4 AMERICAN MUSEUM NOVITATES NO. 3042

ACKNOWLEDGMENTS heavily mineralized chelipeds with well-de- This research has been supported by grants veloped claws and four pairs of locomotor from the Georgia Geological Survey, the Na- appendages, the last much reduced and not tional Geographic Society's Committee for fused to the other sternal somites. The ab- Research and Exploration, Georgia Southern domen consists of two very soft and flexible College's Faculty Research Committee, and segments with very reduced pleopods used the St. Catherines Island Research Center for reproduction, followed by three more funded through the Edward John Noble highly mineralized segments with well-de- Foundation at the American Museum ofNat- veloped diaphragm-like pleopods, in turn fol- ural History, the University System ofGeor- lowed by the hardened sixth segment and tel- gia's Chancellor's Special Funding Initiative son with attached uropods forming a tail fan. for Research and Development, and the The Carolinian Ghost Shrimp spends most Smithsonian Institution's Short Term Visi- of its life burrowed into the sand of beaches tors Program. Assistance in the field has been and sand flats (fig. 4) fronting the open ocean. received from Nelda Bishop, Dan Good, Pe- Its presernce on a beach is indicated by holes ter and Anna Hayes, Ginny Millar, Richard marking burrow mouth apertures (fig. 3c), Petkewich, Cathy Rigsby, and JoAnne Shad- which often exhibit water spouts or sand roui. Special thanks are extended to Royce floods (fig. 3b) temporarily accumulating as Hayes and the staff of St. Catherines Island volcano-like mounds (fig. 3c) often ringed by for their support and assistance in facilitating extruded rod-shaped black fecal pellets com- this research. The manuscript was read and posed of mud (fig. 3d). Collapsed burrows strengthened by suggestions and comments (fig. 3e) occasionally form as the unstable from Royce Hayes, Sergio Rodrigues, and H. beach sand caves into, or washes into, the B. Rollins. underlying burrow. Few observations have been made of the animal outside its burrow (Pohl, 1946: 78; E. C. Bishop, 1988). The NATURAL HISTORY animal apparently feeds on detritus and par- ticulates in water pumped through its bur- DESCRIPTION row. It periodically extrudes masses of rod- The Carolinian Ghost Shrimp is a medi- shaped, cylindrical, channeled fecal pellets um-sized (30-100 mm) decapod crustacean consisting largely of mud containing 3-10% beautifully adapted to its burrowing mode of organic carbon (Pryor, 1975: 1246). The life (fig. 2). The pale animal is elongate, dor- composition ofthe fecal pellets indicates that soventrally flattened, cylindrical, and lightly these animals are actively ingesting large mineralized (fig. 3a). The cuticle is thin and quantities ofmud from some source. Because flexible over most of its body which consists the open beach consists mostly ofwell-sorted ofa short cephalothorax, appendages, and an quartz sand, the animal must ingest these large elongate abdomen. The cephalothorax is quantities ofmud from the turbid coastal wa- lightly sclerotized and flexible except for the ter being pumped through their burrows. The anterior of the dorsal shield which forms a production offecal pellets is copious and may hardened dorsal disk. The locomotor ap- result in the accumulation ofstringers ofpel- pendages are attached to the sternum beneath lets in ripple or runnel troughs and eventually the animal's carapace and consist of the in layers of mud in environments otherwise

Fig. 3. Ghost Shrimp from St. Catherines Island, Georgia. a, Dorsal view of male Carolinian Ghost Shrimp (Callichirus major) released in beach runnel; b, Carolinian Ghost Shrimp burrow mouth on North Beach showing a flood of sand being pumped onto the beach by a ghost shrimp; c, four active burrows on North Beach showing open burrow mouths, sand mounds, and a ring of fecal pellets of Carolinian Ghost Shrimp; d, burrow mouth of Carolinian Ghost Shrimp showing sand mound capped by a ring of rod-shaped fecal pellets; e, collapsed mouth of Carolinian Ghost Shrimp burrow indicating diameter of major shaft; f, lateral view of Georgian Ghost Shrimp (Callianassa biformis) above a 1 992 BISHOP AND BISHOP: DISTRIBUTION OF GHOST SHRIMP 5

WIN

11 WU- * m I

Epv 9sff-'6S-<-g ~ h -*-,-eetwe_ 1 ~~~23 ;2 Carolinian Ghost Shrimp to illustrate the difference in size of adults; g, active burrows of Georgian Ghost Shrimp on Walburg Creek Shoal showing great density, tiny size, and abundant fecal pellets (hoop diameter = 36 cm); h, active burrows of Georgian Ghost Shrimp (Callianassa biformis) in fossoria in aquarium. 6 AMERICAN MUSEUM NOVITATES NO. 3042

Fig. 4. Photograph of Callichirus major burrow strand looking north at low tide along North Beach at Picnic Bluff, St. Catherines Island; 10 cm scale in foreground. characterized by sand-sized quartz (Howard into the populations. Thompson and Prit- and Reineck, 1972: 92; Pryor, 1975). chard (1969), Swinbanks and Murray (1981), Large numbers ofCarolinian Ghost Shrimp and Swinbanks and Luternauer (1987) stated were collected during single years on Tybee, that the upper limit of the distribution of C. St. Catherines, and St. Simon's islands, using californiensis is controlled by critical tide lev- a Yabby Bait Pump (Manning, 1975). These el which exposes the burrows for 3.2 to 7.8 samples were analyzed by students Mike Klug days leading to anoxia in the burrow waters. and Ginny Millar at Georgia Southern Col- Felder and Lovett (1989) described the hab- lege for completion of senior thesis require- itat and some ecological parameters of Cal- ments; specimens exhibited similar patterns lianassa louisianensis Schmitt, 1935, in the of population structure, sexual composition northern GulfofMexico and mentioned that (70% female, 30% male), and relative pro- that species is included in the diet of the portions of right and left handedness (52% Whooping Crane. left, 48% right). The lack of small juveniles Carolinian Ghost Shrimp reproduce sex- (fig. 5) seems to indicate that recruitment oc- ually. The mechanism of sperm transfer is curs at a size ofabout 20 mm. However sam- not known. Egg masses developing in the pling along beach transects has demonstrated ovaries in the cephalothorax and abdomen that specimens on the lower part ofthe beach are visible in the females throughout the fall consist primarily of smaller animals as well and winter. In the spring (about mid-March) as a few large males. This pattern may in- females begin to lay their eggs and carry the dicate that larval animals settle and live sub- egg masses throughout the summer attached tidally. Posey (1986) demonstrated that the to the pleopods oftheir abdomens, primarily seaward edge of the distribution of Calli- of the first two abdominal somites. The eggs anassa californiensis on the mid-tidal zone hatch and the larvae enter a planktonic ex- in Oregon was controlled by fish predation. istence as is indicated by their abundance in Predation by sting rays may control the dis- plankton samples. Larval settlement into the tribution of Carolinian Ghost Shrimp in substrate is an unknown process. Georgia, as well as recruitment ofjuveniles Samples collected on the Georgia coast have 1992 BISHOP AND BISHOP: DISTRIBUTION OF GHOST SHRIMP 7

proportion of substratal mud. It is generally CalUIctiirus major Say associated with abundant acorn worms, mud 20 St. Catherim lsland L snails, hermit crabs, and other invertebrates. Picnic Bluff, 1988-89 The burrows of the Georgian Ghost Shrimp n = 452 are smaller, apparently shallower, and found 15 I in great densities of up to 483 burrows/M2 U Smallest vigerous Female on St. Catherines Island (fig. 3g, h). 8 10 The ecological ranges of Carolinian and Georgian Ghost Shrimp overlap on such 5.o. I1A sheltered tidal flats, although the respective burrows appear to be of different depth, ef- fectively partitioning them from one another. 0. KI, .4 di 1 I0 20 40 60 80 100 120 Total Leth. mm BuRRows Fig. 5. Histogram of total length (rostrum to When removed from the substrate and re- telson) of452 specimens of Callichirus major col- leased on an open sand surface, Carolinian lected at Picnic Bluff during 1988-89. Ghost Shrimp immediately begin to burrow, entering the substrate laterally, and disap- pearing beneath the surface within a few min- usually included a commensal crab, Pinnixa utes. Placement of animals in aquaria (Pohl, cristata Rathbun, which is very small, trans- 1946; GAB, this study) and in ant-farm-like versely elongate, and obviously adapted to frames (fossilorium) (Pohl, 1946) allowed the cohabiting the burrows of Carolinian Ghost burrowing process to be studied in detail. Shrimp. When females are ovigerous, small Once beneath the sand, the Carolinian Ghost amphipods have been observed in the egg Shrimp constructs a widened area called a masses, presumably feeding on the eggs. Pohl turnaround which allows the animal to re- (1946: 78) cited the presence of a bryozoan, orient in order to more efficiently remove Acanthodesia tenuis (Desor), encrusting the sand from the burrow (fig. 6). The burrow is burrow wall and reiterated the observation then commonly constructed with several made by Pearse in 1939, and by herself in branches from the turnaround, some hori- 1944, of a small red commensal copepod, zontal and some inclined at a steep angle to Clausidium dissimile Wilson, suggesting it the horizontal. Carolinian Ghost Shrimp might be the "curious parasite" found by Say burrows in analogous Pleistocene ancient (1818). barrier island sediments (Howard and Scott, The burrowing mode oflife has made Car- 1983: 176) indicate the burrows extend 3-5 olinian Ghost Shrimp difficult to capture and m into the beach. Ancient burrows of anal- study. Previously they had to be captured by ogous, but not necessarily conspecific, ghost digging them out ofthe sand, cutting offtheir shrimp from more ancient sediments (fig. 7) escape by blocking their burrow with a long preserve similar morphologies and are as- spade, or tricking them by dropping sand or signable to the ichnogenus Ophiomorpha. shell into the upper part oftheir burrow (Pohl, Carolinian Ghost Shrimp burrows on the 1946:71). The availability ofa suction device beach consist of three major parts: (1) a ver- called a Yabby Pump has revolutionized the tical upper constricted burrow aperture, (2) a sampling offossorial infauna, allowing rapid nearly vertical main shaft, and (3) one or more capture of large samples (Manning, 1975: approximately horizontal burrow mazes. The 318). upper constricted burrow aperture is vertical, The Georgian Ghost Shrimp (Callianassa about 5 mm in diameter, and usually about biformis Biffar 1970) resembles the Carolin- 15-20 cm in length. It opens into the main ian Ghost Shrimp but is differentiated by spe- burrow shaft which is vertical to inclined, cific morphological details and by its much several meters in length, and about 1-2 cm smaller size (fig. 3f). Ecologically the Geor- in diameter with periodic enlarged turn- gian Ghost shrimp tends to inhabit lower in- arounds and numerous horizontal and/or intertidal sand flats in sheltered areas which are clined branches. The lower part ofthe burrow typically rippled, and contain a considerable is thought to consist of anastomosing, inter- 8 AMERICAN MUSEUM NOVITATES NO. 3042

Fig. 6. Female Carolinian Ghost Shrimp turn- ing around in turnaround chamber in aquarium. Rear ofcephalothorax and thoracopods are visible at top and tucked-under abdomen below. Anterior of animal is to the right. Scale bar 1 cm. connected, horizontal tunnels. Rodrigues and Suguio (1984) found close correlation between the size of the burrow and the size of the animal inhabiting it. Felder (personal commun.) has correlated the number of sur- ficial burrow apertures with the number of burrow shafts to establish a closer estimate ofthe true population density. Ghost Shrimp probably spend most oftheir time in the lower part of the burrow, occasionally rising in the shafts to eliminate sand or waste. In un- consolidated sand the animal lines the bur- Fig. 7. Fossilized ghost shrimp burrows, as- row with a mucal-mud binding agent, pack- signable to the ichnogenus Ophiomorpha, exposed ing mucus-laden pellets into the burrow walls, in a road cut through Eocene Tobacco Road For- building a considerable burrow wall thick- mation along Old River Road, Burke County, ness, and imprinting a knobby texture onto Georgia. Notice the vertical shaft parallel to the the exterior of the burrow. When lithified, scale (10 cm on left) giving way to a turnaround these burrows are and a more or less horizontal burrow branch at assigned to the ichnogenus the bottom. Horizontal branches ofother burrows Ophiomorpha. Burrow morphology can vary show as oval or circular cross sections. dramatically in different sediment types as well as interspecifically (Frey et al., 1978). We emphasize that many burrow morphologies are constructed by a single species and similar 1 992 BISHOP AND BISHOP: DISTRIBUTION OF GHOST SHRIMP 9 burrow morphologies are constructed by dif- physical conditions such as revetments, beach ferent species. Burrows occasionally collapse, slope, steep sound channel walls, or presence forming chevronlike sedimentary structures ofrelict marsh sediment under the beach sur- (Frey et al., 1978). When erosional condi- face. The density ofburrows varies along the tions scour a beach containing Carolinian beach, building into hill-like local density Ghost Shrimp burrows, the burrows may be maxima separated by saddlelike minima but eroded to a level where the beach surface burrows are always present except where ex- intersects the main shafts. When this occurs cluded by sedimentological factors cited the main shaft is plugged and abandoned or above. On disturbed beaches (renourished a small, inclined shunt is constructed to the beaches and ebb deltas) burrow density gen- beach surface (Frey et al., 1978). erally decreases dramatically: the distribu- The burrows ofdifferent ghost shrimp spe- tion pattern broadens and becomes more dis- cies exhibit morphologies ranging from sol- persed. On normal, open oceanic beaches the itary burrows to communal burrow systems burrow distribution occupies a continuous (Frey et al., 1978), necessitating great caution narrow zone about 50 m wide along the beach when interpreting Holocene and ancient eco- from below low-water level tojust above mid- logical data (e.g., the interpretation of the tide level. When contoured and plotted on a ecology ofCallichirus major, ifbased on what true scale map this pattern is stringlike, hence is known about the ecology of Callianassa it can be referred to as the burrow strand. californiensis, would lead to largely erroneous The subtidal occurrence of the Carolinian conclusions). Erickson and Sanders (1991) and Georgian Ghost Shrimp has not been described Pleistocene burrows of the ghost investigated in this project because turbid shrimp, Callichirus major. water conditions prohibit subtidal sampling. Reports from the literature (Dorjes, 1972; BuRRow DISTRIBUTION Frey et al., 1978) indicate that the Carolinian Ghost Shrimp occurs subtidally to a depth of The density and distribution ofburrows of a few meters on the forebeach and on shallow the Carolinian Ghost Shrimp have been de- shoals. scribed by Pohl (1946), Pryor (1975), and Monitoring burrow mouth density on St. Eric Bishop (1988). Pohl (1946: 74) found the Catherines and St. Simon's islands over a maximum burrow density to lie just above two-year period has demonstrated the sta- mid-tide level ". . . in the quarter ofthe beach bility ofthe density distribution patterns. Only between the half-way and three-fourths minor fluctuations were observed in burrow mark.. ." Pryor (1975) cited Ghost Shrimp mouth density patterns except for one small burrow mouth densities of up to 450/M2 in area on North Beach, St. Catherines Island. tidal pools and protected shores of the Mis- That area, apparently underlain by relict sissippi Sound [although these probably be- marsh and veneered by beach sand (Morris longed to a smaller ghost shrimp such as Cal- and Rollins, 1977), experienced a dramatic lianassa biformis]. Eric Bishop (1988) decrease in burrow density from 1988 to 1989 surveyed and contoured the distribution of over a 100 m stretch ofbeach. A general de- Carolinian Ghost Shrimp burrow mouths on crease in burrow density was observed over beaches at Tybee, St. Catherines, and St. Si- the same period on North Beach at St. Cath- mon's islands, concluding that maximum erines and Old Village Beach at St. Simon's burrow mouth development occurs on stable Island. beaches just above mid-tide level. His study Burrows of the Georgian Ghost Shrimp demonstrated variation in burrow distribu- (Callichirus biformis) are much smaller, typ- tion across and along the beach. The burrow ically ringed by muddy sand mixed with fecal density increases rapidly from 0 bpm2 to a pellets, and have smaller fecal pellets (fig. 4g). maximum just above the mid-tide level then Depth (based on success in extracting these decreases slowly to low densities along the animals from shallow depths with a Yabby low water level. The zone of highest density Pump and observation in an aquarium, fig. generally follows the beach trend just above 3h) appears to be much less than that of Cal- mid-tide level, but this position changes with lichirus major. Burrow density data from the 10 AMERICAN MUSEUM NOVITATES NO. 3042

north end of St. Catherines Island (fig. 11) along the strike direction of the beach. The indicate that the density increases from 0 16 tosses were made in a crisscross pattern bpm2just above the mid-tide line to 483 bpm2 to assure four counts from each quadrant of just above the low-tide line with a lateral dis- the area around a station. Although not ran- tribution ofabundant Georgian Ghost Shrimp dom in the sense of using a random number along the shore for 700 m in an elongate hill- table to determine direction and distance of like pattern. The pattern of distribution ap- tosses, experience and consistency of resul- pears to depend more on the presence ofmud tant data indicate that the number of tosses in the substrate than on other physical pa- and methodology more than adequately rameters. compensate for lack of randomness. Burrow density near each point was calculated by tab- METHODOLOGY ulating the total number of burrow mouths MAPPING BuRRow DISTRIBUTION encircled by, or lying beneath, the hoop for each of 16 tosses and dividing the total by 4 A survey was made (fig. 8) on North Beach, [16 tosses x 1/4 m2 toss x 1/4 = burrows/M2 St. Catherines Island, by using an iron stake (bpm2)]. The data for all points along all lines near Picnic Bluff as a beginning point (point were then contoured using a contour interval 0) from which 100 m intervals were surveyed appropriate to the burrow density. The re- to the north and to the south using a 100 m sultant map of burrow distribution is dis- fiberglass tape. Points on the beach at high cussed under Results. Because of the greater tide line were numbered 0, IN, 2N, . . . , 20N density of burrows of the Georgian Ghost and 0, IS, 2S, . . . , 27S. Starting at the high- Shrimp it was impractical to use a /4 m2 hoop tide line (HTL), beach transects (called Line (as it was virtually impossible to tabulate bur- 0, Line IN, etc.) were established at each of row counts in such a large area). Therefore, the surveyed points and laid out perpendic- 10 tosses of a 14-in. diameter embroidery ular to the shoreline across the beach from hoop having a surface area of 0.095 m2 were the high-tide line to the low-tide line. Stations used and the burrow density was calculated were established at 7.5 m intervals along each [10 tosses x 0.95 m2/toss x 1/0.095 = bur- line beginning at high-tide line (Station 1) and rows/M2]. extending to the low-water level (both high- and low-tide levels fluctuate from day to day RESULTS but are roughly comparable as all surveys were done during spring tides). BuRRow DENSITY AND DISTRIBUTION The highest (i.e., highest on the beach or Carolinian Ghost Shrimp burrows on St. closest to the high-tide line) Carolinian Ghost Catherines Island are distributed in a com- Shrimp burrow mouths were located on the plex pattern that is controlled by physical beach and their position relative to the line conditions ofexposure and substrate quality. of transect recorded. These burrows are nor- Three detailed areas are presented here to mally easily recognized by their diameter (5 help illustrate the distribution patterns ob- mm), by their periodic activity (water spout- served on the north end of the island along ing, extrusion of sand floods, or extrusion of North Beach (fig. 9), on Sea Side Ebb Delta characteristic rod-shaped fecal pellets), or by (fig. 10), and on the Sound (fig. 1 1). On the their mucus lining manifested by a coherent middle and north end of North Beach, the collar ofsand around the burrow mouth. Bur- pattern is characterized by a 2.5 km burrow row occupancy was validated by periodically strand from the Sound to the south end of sampling with a Yabby Pump. the erosional bluff (Picnic Bluff) developed A round hoop with a 1/4 m2 area was ran- on the Pleistocene core of the island. Over domly tossed 16 times at each station along this interval the burrow strand approximate- each transect from the first station below the ly follows the mid-tide line with a steep gra- highest burrow to the lowest station nearest dient on its shoreward side and gentler gra- the low-water line. Tosses were kept within dient on the seaward side (fig. 9). Burrow half-station intervals ofeach point along the density ranges from 0 to 13.75 bpm2 along transect and within about 10 m ofeach point the burrow strand with three nodes of max- 1992 BISHOP AND BISHOP: DISTRIBUTION OF GHOST SHRIMP I1I

SURVEYED SAMPLING GRID Une 1 N I...JI...... LII I I I I I 2 3 4 5 6 7 8 9 10 11 12

a Iron Stake I Une 0 X R2 2 3 4 6 7 8 9 10 11 12 .P [seeflgure below]

8 7.5 meters between stations of transect Line 1 S

I N

_ Li I. Line of transect

7.5 m

Fig. 8. Transect grid used in North Beach density surveys and hoop tossing pattern used to assure even sampling across sampling area. imum density. Burrow density decreases dra- the flat, rippled sandy tidal flats forming the matically at both ends of the distribution as ebb shoal along the north edge of the island. it approaches St. Catherines Sound to the They then appear again in some abundance north and the relict marsh sediment exposed along the edge ofthe shoal near the low-water in front of Picnic Bluff on the older part of line and continue around the north end of the island to the south. In the sound, Caro- the island where they cohabit the beach with linian Ghost Shrimp are found along the steep a large population ofGeorgian Ghost Shrimp channel wall, but are largely excluded from (Callianassa biformis). 12 AMERICAN MUSEUM NOVITATES NO. 3042

Burrow density on the southern part of North Beach decreases to near 0 bpm2 along Black Hammock Spit between the bluff and Sea Side Ebb Delta. The mid-tide level along this stretch consists of exposures of relict marsh sediments (Morris and Rollins, 1977: fig. 6) which form an unsuitable substrate for Carolinian Ghost Shrimp. A few burrows have been observed during exceptionally low tides seaward ofthis relict marsh on the foreshore and a few burrows occur in sand-filled ancient tidal creek channels. On one occasion ephemeral burrows were observed in a sand veneer on top ofthe relict marsh muds. Tem- poral instability of sands over these relict muds is extreme as is the periodic invasion of thick sand veneers by Carolinian Ghost Shrimp. The south end of North Beach is domi- nated by Sea Side Ebb Delta. Burrow density on the shifting sands ofthe ebb delta (fig. 10) is generally low and highly variable, about 2.5-5.0 bpm2, but ranging from 0 bpm2 to 14.25 bpm2. Part of the ebb delta along the seaward side of Sea Side Spit is relict marsh mud which excludes Carolinian Ghost Shrimp from part of this area most of the time and in fact apparently controls the position of Sea Side Inlet (trapping the mouth of the inlet in the grip of the resistant relict marsh mud). Burrows become restricted to the channel of Sea Side Inlet and are found for some yet undetermined distance into the marsh lying shoreward of Black Hammock Spit and Middle Beach. Burrow density along the north end of the island along the south shore ofSt. Catherines Sound is generally very low and restricted to a narrow burrow strand because of the steep gradient of the channel margin, except for broad sand tidal flats at each "shoulder" of the island where current shadows apparently develop on the tail ends of flood and ebb tides. These tidal flats are developed in the Fig. 9. Distribution of Carolinian Ghost lower foreshore and are characterized by large Shrimp (Callichirus major) burrows along Picnic expanses of rippled, muddy sands inhabited Bluff on North Beach from Line 1S to line 6N by a plethora of infaunal and semi-infaunal showing burrow strand at approximately mid-tide organisms (Morris and Rollins, 1977) in- level and variation in burrow density along the cluding large populations of the burrow strand. Between Lines 4N and 5N a low- diminutive tide shoal or bar originates and trends southeast Georgian Ghost Shrimp, especially on the with the burrow distribution paralleling the bar. shoal at the mouth of Walburg Creek on the Contour interval is 5 burrows/M2. Note that dia- northwest shoulder of the island (fig. 11). gram is exaggerated 3.3 x perpendicular to the Burrows of the Georgian Ghost Shrimp are beach. found in a broad burrow strand parallel to 1992 BISHOP AND BISHOP: DISTRIBUTION OF GHOST SHRIMP 13

0 :..-: ;.: 8: 8: 6: I I 14S I I I I : .: .: .: I 9 : I I I I I I .::::::: 0.3 1.0 2.0 13 0.3 1.0 1I

AndeM 0 Tkhd 4: ::: Crook 0.5

1.3 2.3 1.0 1.8 06S 2.5 1.5 c ._3 o .t a i I = 16S _ I.; 0.8 01.8 0.8 1.3 1.8 2.3 1.3 0.8 0.3 &X

17S I I I I F I I I I I I 0 0.0 3.3 4.3 6.3 1.5

I ¢U~ a rea.

8,- , .------, . , I , . , . - 18S I I I I I I I ---j- 5 1.0 3.0 2.0 3.8 3.5 4.3 3.0 4.3 5.3

do- I.I Seea Side Ebb 19S I a_a I I _ -I-_ I-0-._ * -I 94, 2.3 4.5 3 5.'5 10.0 13.8 87. 1 4.3 14.0 : I i1.O I : : a I %10 II I I s o o o . : I0

I I I 20S I I I I I I I. I I - I co 0 3.0 2.3 2.8 5.5 8.3 8o 1 .3 9.3 7.5

7.5 m a

a Scales Andont Se Sde Inlet I NL

21S III I t I I I I t I I I 0 0.8 0.5 1.5 1.8 3.0 3.0 3.5 2.8 3.5 4. 7.3 6.8

- a Latra Exaggeration 3.3x Fig. 10. Distribution of the Carolinian Ghost Shrimp (Callichirus major) along the south end of Black Hammock Spit and on part ofSea Side Inlet ebb delta. Relict marsh mud prevents the colonization of the mid-beach by Ghost Shrimp except while sand-filled ancient channels provide adequate substrate at the correct level. Burrow distribution on Sea Side ebb delta becomes diffuse and somewhat lowered, typical of distributions on unstable substrates. Contour interval is 5 burrows/M2. Note that the diagram is exaggerated 3.3 x perpendicular to the beach. the shore along the stretch ofthe Sound beach length of the north end of the island, two which has a muddy-sand substrate. Although major distributional peaks ofGeorgian Ghost present in small numbers along the entire Shrimp are present on sand tidal shoals at 14 AMERICAN MUSEUM NOVITATES NO. 3042

the northwest (Engineer's Point) and the northeast "shoulders" of St. Catherines Is- land. The distribution at Engineer's Point (fig. 11) is well developed for over 700 m and is up to 50 m wide, consisting of three nodes ofmaximum density. Burrow density rapidly increases from 0 burrows/M2 from just above mid-tide level to maximum densities of 483 burrows/M2 just above normal low-tide level and begins to decrease more slowly to unknown densities in the subtidal waters of the Sound. The distributions of Carolinian and Georgian Ghost Shrimp overlap on Engi- neer's Point where Carolinian Ghost Shrimp burrows reach their highest density on the upper part of the beach and both are present on the middle and lower part of the beach.

CONCLUSIONS The burrows of the Carolinian and Geor- gian Ghost Shrimp on the north end of St. Catherines Island are characteristically distributed parallel to the strand line. These distributions can be related to available pre- ferred substrate and to tidal exposure (fig. 12). The upper edge ofthe burrow strand lies just below the line of demarcation of the lowest ofthe monthly high-tide sequences which de- fine long intervals of exposure twice each month. The lower edge of the burrow strand lies just below the line of demarcation of the highest low tide and just above the line of demarcation of the lowest low tide (subtidal line). This distribution pattern strongly sug- gests that the position of the burrow strand is controlled primarily by substrate and sec- ondarily by length ofexposure along its upper edge and perhaps by predation along its lower edge. The following conclusions can be made re- garding the distribution of ghost shrimp on Fig. 11. Distribution of the burrows of the St. Catherines Island: Georgian Ghost Shrimp (Callianassa biformis) on 1. The Carolinian Ghost Shrimp (Calli- protected, rippled, muddy-sand substrate on Wal- chirus major) is the most abundant and near- burg Creek Shoal on the northwest shoulder ofSt. ly ubiquitous macroorganism on sandy, oce- Catherines Island. The distribution ofthe burrows anic Georgia beaches. of Georgian Ghost Shrimp is tied more closely to 2. Open ocean sand beaches are inhabited muddy sands which tend to accumulate in current the shadows on entrances to sounds. Contour interval by Carolinian Ghost Shrimp whose pres- is 100 burrows/M2. Note that the diagram is ex- ence is indicated by active burrows. aggerated 3.3 x perpendicular to the beach. 3. Burrows ofthe Carolinian Ghost Shrimp reach their maximum density as a narrow 1992 BISHOP AND BISHOP: DISTRIBUTION OF GHOST SHRIMP 15

I ] I

T C. - U ~~E U A. 1 3 17 . 15.. _0III T II1T +-H 9 - 8 7 6 \ D. -h 5 5 9 13 17 4 3 2 300'-400

I0 200- _ MlidNightI Noon 6 100o Jl...-- B. E. I m~~ ~ ~ ~ . Fig. 12. Summary of distribution of St. Catherines Ghost Shrimp relative to tidal parameters, par- ticularly to mean lowest high spring tides (sltl) and mean highest low spring tides (shtl), and typical beach profiles on open beaches and protected tidal sand flat shoals. (A) Monthly tidal graph for September, 1989; (B) daily tidal graph for September 19, 1989; (C) idealized beach profiles for normal ocean beaches (stipple pattern) and muddy sand tidal flats (dashed stipple pattern); (D) typical abundance profile across the beach for distributions of Carolinian Ghost Shrimp (C. major); with number of individuals plotted on vertical axis against stations plotted on horizontal axis; and (E) typical abundance profile across the tidal flats of Georgian Ghost Shrimp (C. biformis) with number of individuals plotted on vertical axis against stations plotted on horizontal axis. Tidal data produced using Nautasoft's "MacTides."

band, termed the "burrow strand," parallel 7. The distribution ofthe burrows of Car- to the mid-tide line on open ocean sand olinian Ghost Shrimp (Callichirus major) and beaches. the Georgian Ghost Shrimp (Callianassa bi- 4. Burrow patterns on unstable sand sub- formis) overlap. strates become diffuse and generally less dense. REFERENCES 5. The Carolinian Ghost Shrimp is not found in areas where the beach is underlain Biffar, T. A. by relict marsh muds, but Ghost Shrimp bur- 1970. New species of Callianassa (Decapoda, rows may mark sand-filled ancient tidal creek Thalassinidea) from the Western Atlan- channels. tic. Crustaceana 21(3): 225-236. 6. Protected muddy, rippled sand sub- 1971. The genus Callianassa (Crustacea: De- strates at entrances to capoda. Thalassinidea) in South Flori- sounds are inhabited da, with keys to the Western Atlantic by abundant Georgian Ghost Shrimp (Cal- species. Bull. Mar. Sci. 21(3): 637-715. lianassa biformis) whose presence is indicat- Bishop, E. C. ed by abundant, small, active burrows. 1988. Burrow density ofghost shrimp on three 16 AMERICAN MUSEUM NOVITATES NO. 3042

Georgia beaches. Georgia State Science Pohl, M. E. Fair. 1946. Ecological observations on Callianassa D6rjes, J. major Say at Beaufort, North Carolina. 1972. Georgia coastal region, Sapelo Island, Ecology 27(1): 71-80. U.S.A.: sedimentology and biology. VII. Posey, M. H. Distribution and zonation of macro- 1986. Changes in a benthic community asso- benthic animals. Senckenbergiana Ma- ciated with dense beds of a burrowing rit. 4: 183-216. deposit feeder, Callianassa californien- Erickson, B. R., and A. E. Sanders sis. Mar. Ecol. Prog. Ser. 31: 15-22. 1991. Bioturbation structures in Pleistocene Pryor, W. A. coastal plain sediments of South Caro- 1975. Biogenic sedimentation and alteration lina, North America. Sci. Publ. Sci. Mus. of argillaceous sediments in shallow Minnesota, n. ser., 7(2): 5-14. marine environments. Geol. Soc. Am. Felder, D. L., and D. L. Lovett Bull. 86: 1244-1254. 1989. Relative growth and sexual maturation Rodrigues, S. de A. in the esturine Ghost Shrimp Calli- 1976. Sobre a reprodu,cao, embriologia e des- anassa louisianensis Schmitt, 1935. J. envolvimento larval de Callichirus ma- Crustacean Biol. 9(4): 540-553. jor (Say, 1818) (Crustacea, Decapoda, Frey, R. W., J. D. Howard, and W. A. Pryor Thalassinidea). Bol. Zoologia, Univ. Sao 1978. Ophiomorpha: its morphologic, taxo- Paulo 1: 85-104. nomic, and environmental significance. Rodrigues, S. de A., and R. M. Shimizu Palaeogeogr., Palaeoclimatol., Palaeo- 1986. Estimativa da biomassa de Callichirus ecol. 23: 199-229. major (Say, 1818) na Baia de Santos, Howard, J. D., and H. E. Reineck (Crustacea, Decapoda, Thalassinidea). 1972. Georgia coastal region, Sapelo Island, Ciencia e Cultura-Resumos SBPC, Sao U.S.A.: sedimentology and biology. IV. Paulo 38(7): 688-689. Physical and biogenic sedimentary Rodrigues, S. de A., and K. Suguio structures ofthe nearshore shelf. Senck- 1984. Ecologia e peleoecologia de Callichirus enbergiana Marit. 4: 81-123. major Say (1818) (Crustacea, Decapo- Howard, J. D., and R. M. Scott da, Thalassinidea). An. Sem. Reg. Ecol. 1983. Comparison of Pleistocene and Holo- 4: 499-519. cene barrier island beach-to-offshore se- Rodrigues, S. de A., J. M. Ferreira, H. F. Gebara, quence, Georgia and Northeast Florida and R. M. Shimizu coasts. Sed. Geol. 34: 167-183. 1986a. Estrutura temporal da populacao de Lutz, G. R., Jr. Callichirus major (Say, 1818) na Baia 1937. Notes on Callianassa major Say. de Santos (Crustacea, Decapoda, Tha- Charleston Mus. Leaflet 10: 1-15. lassinidea). RESUMOS. XIII Congres- Manning, R. B. so Brasileiro de Zoologia. Cuiaba So- 1975. Two methods for collecting decapods in ciedade Brasileira de Zoologia, p. 56. shallow water. Crustaceana 29: 317-319. Rodrigues, S. de A., A. F. Midio, M. V. H. To- Manning, R. B., and D. L. Felder motake, and R. M. Shikizu 1986. The status ofthe callianassid genus Cal- 1986b. Determinacao de arsenico en Callichi- lichirus Stimpson, 1866. Proc. Biol. Soc. rus major (Say, 1818) da Baia de Santo, Washington 99(3): 437-443. SP (Crustacea, Decapoda, Thalassinid- 1991. Revision of the American Callianassi- ea). Ciencia e Cultura-Resumos SBPC, dae (Crustacea: Decapoda: Thalassinid- Sao Paulo 38(7): 737. ea). Proc. Biol. Soc. Washington 104(4): 764-792. Say, T. Morris, R. W., and H. B. Rollins 1818. An account ofthe Crustacea ofthe U.S. 1977. Observations on intertidal organism as- Phil. Acad. Nat. Sci. 1: 57-458. sociations of St. Catherines Island, Shadroui, J. M. Georgia. I. General description and pa- 1990. Aeolian transport in a small dune field leoecological implications. Bull. Am. on St. Catherines Island, Georgia. Geor- Mus. Nat. Hist. 159(3): 89-128. gia Geol. Surv. Bull. 122: 48-64. Pearse, A. S., H. J. Humm, and G. W. Wharton Stimpson, W. 1942. Ecology ofsand beaches at Beaufort, N. 1866. Descriptions of new genera and species C. Ecol. Monogr. 12(2): 35-190. ofmacrurous Crustacea from the coasts 1992 BISHOP AND BISHOP: DISTRIBUTION OF GHOST SHRIMP 17

ofNorth America. Proc. Chicago Acad. Thompson, L. C., and A. W. Pritchard Sci. 1: 46-48. 1969. Osmoregulatory capacities of Calli- Swinbanks, D. D., and J. L. Luternauer anassa and Upogebia (Crustacea: Tha- 1987. Burrow distribution of thalassinidean lassinidea). Biol. Bull. 136: 114-129. shrimp on a Fraser Delta tidal flat, Brit- Williams, A. B. ish Columbia. J. Paleontol. 61(2): 315- 1984. Shrimps, lobsters, and crabs of the At- 332. lantic Coast ofthe Eastern United States, Swinbanks, D. D., and J. W. Murray Maine to Florida. Washington, DC: 198 1. Biosedimentological zonation of Smithsonian Institution Press. Boundary Bay tidal flats, Fraser River Delta, British Columbia. Sedimentol- ogy 28: 201-237.

Recent issues of the Novitates may be purchased from the Museum. Lists of back issues of the Novitates, Bulletin, and Anthropological Papers published during the last five years are available free of charge. Address orders to: American Museum of Natural History Library, Department D, Central Park West at 79th St., New York, N.Y. 10024.

THIS PUBLICATION IS PRINTED ON ACID-FREE PAPER. |