ORGANISM-SEDIMENT RELATIONSHIPS ON A MODERN TIDAL FLAT, BODEGA HARBOR, CALIFORNIA

Thomas E. Ronan, Jr. Fritz Jack Farmer Department of Earth and Space Sciences Department of Geology Department of University of California Vanderbilt University of California California 90024 Nashville, Tennessee 37235 Davis, California 95616

ABSTRACT Physical Setting

Sandflats in the protected Bodega Harbor are in- Bodega Harbor is a shallow coastal embayment habited by a diverse infaunal assemblage (over located i n southwest County about 65 miles species) consisting predominantly of molluscs, north of San Francisco (Fig. 1 on p. 1). The nearly chaetes and other worm-like animals, and crustaceans. landlocked embayment is sheltered from prevailing Sediment generally is well sorted, fine to medium northwesterly winds by a rocky peninsula, Bodega grained sand with small amounts of silt and clay. Head, and a beach and dune-covered neck of land t o Although ripples are formed by wind- driven and tidal the north that attaches it t o the mainland (Fig. 1). currents, micro- cross- lamination and other physical sedimentary structures are later obliterated by in- tense bioturbation. On the upper tidal flat the thalassinid shrimp Callianassa californiensis is an active deposit feeder. It produces predominantly staghorn- shaped burrow networks which extend 40 cm o r more below the sediment- water- interface. The deposit- feeding bivalve secta inhab- its the middle sandflat and is also an important turbator. On the lower flat the phoronid Phoronopsis viridis is abundant, producing dense clusters of ver- tical tubes. In general suspension feeders live more deeply within the sediment than do deposit feed- ers. Deposit feeders are abundant throughout the flat; they tend to burrow more deeply i n sediment of the high flat than in that of the low flat.

Gradients i n species abundances from mean lower lower water to higher areas on the flat suggest that physical factors such as exposure time are important controls of the species' distributions. However, experimental field studies in Bodega Harbor have shown that ecological interactions between species may also affect distributions (Ronan, 1975). Pre- sence of phoronids i n great abundance prohibits at- tainment of normal burrowing depths by M. secta, and in areas of overlap between phoronids and Callianassa californiensis, there is a high incidence of broken and disturbed phoronid tubes. Thus biological acti- vity may affect, directly or indirectly, the distri- butions of infaunal animals. The same may also have been true in the past, and ichnocoenoses should be viewed as reflecting interactions between animals as well as between the animals and their physical envi- ronment.

INTRODUCTION

Purpose Geologic and location map of Bodega Har- The purpose of the Bodega Harbor portion of bor. Arrow points to sandflat locality. (from Koenig, this guidebook is fivefold: 1) to briefly describe 1963). the physical and geological setting of Bodega Harbor, The harbor is bounded on the south by Beach, 2) to discuss sediment characteristics of the a dune-covered sandspit that curves out from the flat we will be visiting at the northern part of the mainland toward Bodega Head. On its west end, near harbor, 3) to describe the traces made by t h e domi- Bodega Head, the sandspit is truncated by a narrow nant trace- makers and delineate their distributions, tidal inlet and a ship channel that is maintained by 4) on the to illustrate biological interactions dredging. Maximum depths for the harbor to 25 Bodega sand flat and their implications regarding feet) are found i n this channel. The harbor entrance the structuring of modern and ancient communities, is stabilized by two rock jetties that project into and 5) to discuss distributional controls of traces Bodega Bay, a broad bight of coastline that stretches and possible paleoecologic implications. south to Point peninsula. Tidal exchange

15 16 through the harbor channel exceeds 80%on the spring Submerged topography i n estuaries around Bodega tide and 50%on the neap tide (Boyd, 1970); condi- indicates a sea level rise during the last 11,000 tions within the harbor are fully marine with sali- years since the last glacial advance. The dynamic nities from 32-34%. interplay of vertical and lateral movement, changing sea- level, and nearshore sedimentation processes have The Bodega Bay area includes a variety of ex- undoubtedly been complex during the history of the posed, semi- protected, and protected habitats. In Bodega Bay area. Bodega Head and smaller crustal the protected environment of the harbor, soft- bottom blocks caught up within the fault zone probably have substrates are especially conspicuous. At low tide played an important role in the evolution of Bodega (0.0 f t , mean lower low water) some 500 acres of Harbor by creating obstructions to southward flowing tidal sand and mud flats are exposed, which accounts sediment- laden longshore currents. f o r more than 60% of the total harbor area. The seemingly monotonous expanses of sand support a di- ECOLOGICAL STUDIES OF SOFT BOTTOM COMMUNITIES verse marine community dominated by the bi- valve but consisting of more than 200 des- AND THEIR IMPORTANCE TO cribed infaunal species (Standing et al., 1975). The intertidal flats maybe subdivided into several major The utility of trace fossils in subhabitats including high, mid, and low tidal flat, tal interpretation has been demonstrated repeatedly tidal channel and creeks, intertidal pond, and salt (Crimes and Harper, 1970; 1975; Frey, 1975). It is marsh; each subhabitat is characterized by particular based upon the correlation of changes i n ichnocoenoses substrate characteristics and species associations. with changes i n depth and related tors such as wave and current activity and substrate Geologic Setting characteristics. The assumption behind ichnological research has been that most marine communities, especi- Bodega Harbor is traversed by the San Andreas ally shallow- marine communities are " physically fault zone which is locally expressed as a 2.5 km wide modated", controlled by physical environmental zone of intensely sheared bedrock covered by Recent parameters (Sanders, and are maintained a t low marine and aeolian sediments (Fig. 1). Northeast of successional stages by physical environmental instabi- the fault zone the Franciscan Formation of Lake lity (Johnson, 1972). Jurassic to Late Cretaceous age (Bailey, Irwin and Jones, 1964, p. 115-123) is exposed. Locally it con- As pointed out by Rhoads Osgood (1975) sists of thick and massively bedded sandstones inter- and Pickerill trace fossils also provide im- preted as grain- flow deposits (Chipping, 1971, p. portant information about the nature of soft- bottom whereas elsewhere it shows characteristics of communities. Estimates of the percentage of dite deposition. To the north the Franciscan is a bodied species in modem communities whose only record tectonic melange consisting of blocks of intensely would be as trace fossils in paleocommunities range deformed low-grade metagraywacke, shale, chert and from 7% t o 68% (Lawrence, 1968; Schopf, 1978). Clear- greenstone i n a sheared shaly matrix, with rare blocks ly, soft- bodied organisms are important i n the ecolo- of eclogite, glaucophane schist and other high pres- gical structure of modern and presumably ancient com- sure metamorphic rocks. The co- occurrence of high munities. Knowledge of t h e traces made by various and low grade metamorphic assemblages is difficult to soft- bodied animals and of the interactions between interpret in detail. In general the sequence appears these organisms may aid i n the interpretation of paleo- t o have formed by tectonic mixing of wedges communities. The use of x- radiography i n observing shed into a trench environment and high pressure meta- and documenting the interactions of infaunal species morphic rocks formed at much deeper levels within the with the substrate has been especially enlightening subduction zone. (see Howard, 1968; Howard and Elders, 1970; Howard and Dorjes, and this technique also has important The San Andreas fault zone separates Franciscan applications to the study of ancient sediment- organism rocks on the mainland from Late Cretaceous relationships exposed on Bodega Head. Structurally the Bodega Head grandiorite is an extension of the Point grani- The most important advances i n the understanding t i c block and correlates in age and composition with of the ecology of modern soft bottom communities have granitic rocks on Pt. Reyes Peninsula, Point, come through the application of experimental field the Islands, and Mountain t o the methods similar to those developed for rocky intertidal south of San Francisco (Koenig, 1963, p. 6). The substrates by Connell Paine and Dayton Bodega Head granodiorite is significant as the north- (1971). Studies i n soft- substrate shallow marine ernmost outcrop of granitic rock west of the San communities have revealed that biological interactions Andreas fault zone. play an important role in structuring the community and i n modifying the physical environment San- Marine terraces and recent earthquake activity ders, 1958; Rhoads and Young, 1970; Rhoads, 1973; indicate that vertical and lateral movement has con- 1974, 1976); Kitchell (1979) has suggested that tinued to the present. Late Pleistocene to Recent the same may be true in deep sea environments. Ancient uplift resulted in the elevation of marine terraces, communities may also have been structured in part by and higher terraces overlain by Upper Pliocene marine biological interactions which would have affected the sediments occur farther inland. The most recent move- distributions of trace- producing animals. Occurrences ment along the San Andreas Fault in the Bodega Bay of trace fossils outside their typical ichnofacies may area occurred during the San Francisco earthquake of be a result of ecological interactions. Perhaps more 1906. Maximum right lateral displacement of 21 feet information can be gained about the structure of paleo- during that event was recorded near in communities i f the trace- fossil assemglage, the only County; vertical movement at Bodega Head was 18 record of the soft bodied viewed as potentially inches, with the west side (Koenig, 1963, biologically accommodated rather than necessarily p. 7-9). physically controlled. 17

SEDIMENT CHARACTERISTICS OF BODEGA HARBOR TIDAL FLATS Physical Sedimentary Structures

Sediment Source and Texture The most common physical structures observed on the harbor sand flats are transverse ripples and drag The primary source of sediment is medium t o fine or prod marks produced by floating algae or wood grained sand from the Salmon Creek dunes which is Bed forms usually preserved were produced carried by wind southward into Bodega Harbor. Al- by ebb tidal currents, although patches of interference though i n the past, especially after overgrazing in ripples may be present where the ebb tide ripples have the mid 19th century, this has provided enough sedi- been superimposed over the flood tide forms without ment t o partially fill the harbor, stabilization of completely destroying them. As is typical for most the dunes by dune grass in the 1930's greatly reduced tidal flats, the ebb and flood tide ripple sets are the amount of sediment entering the harbor so that now usually developed a t a high angle to one another. Very there is very little deposition. Of secondary impor- small ripples, generated by wind induced bottom cur- . tance is poorly sorted coarse- grained sand derived rents are often visible on the surfaces of the large from the granite of Bodega Head and deposited near tidal current ripples. Usual orientations of these the harbor entrance. small ripple forms reflect the northwesterly winds which dominate this area. Primary sedimentary struc- There are several trends in grain size of sedi- tures are rarely observed i n subsurface cores of the ment within the harbor. The first is a general de- Bodega Harbor sand flats. Apparently, the intense crease in grain size away from the inlet. Coarse reworking of the sediments by the burrowing granite- derived sand is present locally near the completely obliterates the primary fabric of the sand- entrance, but passes laterally into moderately well flat sediments i n a very short time. to well sorted medium t o fine grained quartz- rich sand which covers most of the harbor; at the north- The presence of thin diatom mats which frequent- ernmost comer fine sand, silt, and mud deposited l y cover large areas of the upper and mid flat may during periods of quiet water at maximum flood tide serve to stabilize the sediment and inhibit formation is abundant. The second trend is a decrease in sand- of ripple marks. Ronan (1973) observed that ripples grain size from high to low elevations across the are usually absent where the diatom mats are present. tidal flat (Table 1). In part this trend is related t o the dominance of algal cover a t lower tidal Effects of Animals on the Sediment levels or within bodies of standing water in higher areas of the flat. The algal cover i n many places is animals affect the compaction and water dense enough t o decrease current velocities suffici- content of the sediment and the formation of physical ently to allow deposition of fine grained sediment. sedimentary structures. The algal cover itself also contributes fine organic material. The trend toward fining toward lower ele- Compaction may be measured with a simple pene- vation does not hold for tidal channels, which are trometer consisting of a bar on which bar- bells of floored with coarser sand and lag deposits of pebbles various weights are placed. Penetrometer data for and shells. approximately the middle, upper, and lower portions of the in the north part of the harbor are Grain size Callianassa Macoma secta Phoronopsis- given i n Table 2. The upper flat is covered with a parameter bed bed Macoma secta crust 2 to 3 cmin thickness which resists penetra- tion; this crust probably is related to long emergence time and dessication by the wind. Below this level coarse 13.72 13.0 6.12 the sediment is poorly compacted, as indicated by the deep penetration (33 cm) when sufficient weight was added t o break through this crust. medium (%) 70.24 62.76 South North Sandf l a t Sandf l a t

fine 12.46 13.87 30.61 Loadings Callianassa Macoma Phoronopsis Phoronopsis bed secta -Macoma- Macoma bed secta Nasuta silt + clay 1.75 2.89 .51

5 lbs. 0.5 1.0 cm 0.5 cm 1.0 cm surface sand 74 lbs. 1.0 cm 2.5 cm 1.0 cm 1.0 cm + mud 65.23 68.0 63.66 12% 2.0 cm 5.5 cm 2.5 cm 1.5 cm lbs. 33.0 cm 15.0 cm 7.5 cm 4.0 cm 6 cm depth; sand: mud 56.14 36.0 65.33 Table 2. Penetrometer data for localities on north and south sandflats (after Ronan, 1975, Table 3). sorting Callianassa californiensis is a thalassinid crustacean (graphic well sorted moderately well sorted which constructs burrow networks t o a depth of 40 cm. well sorted o r more. Its burrowing activity serves to loosen the sand and reduce compaction. Similarly, in the mid flat Macoma secta bed, the maximum penetration depth Table 1. Textural of north sand- correlates well with the common burrowing depth of flat sediment. The Callianassa bed is on the upper this mobile bivalve. Near MLLW, t h e maximum flat, the secta bed is on the penetration depth (7.5 cm) corresponds roughly with lower flat, and the secta is on the lower flat. the average depth of the tops of phoronid tubes. . In (From Ronan, 1975). 18

the south sandflat near e where active bur- At and of the eel grass (Zostera) are rowers are absent and only the sessile M. nasuta abundant and the common animals include one snail and Phoronopsis are sediment is more (Polinices) and several polychaete worms highly compacted (Table 2). Apparently the burrowing Eupolymnia, Nepthys, Pectinaria, Notomastus, Glycera, of Callianassa and Macoma secta decrease compaction and Pista). Slightly higher a phoronid worm (Phylum whereas the of the sessile Phoronopsis either Phoronida, Phoronopsis), an echiuroid worm (Phylum increase it or do not affect it. Echiurida, Urechis) , a thalassinid (w- bia), and the bivalves Macoma nasuta, Tresus, and In the Callianassa bed (upper sand flat) Water Saxidomus are common. The fauna of the con- content increases with depth (Table 3); and where sists of polychaetes Lumbrineris), a M. secta is present (middle and lower flat) it nermertean worm (Phylum Nermertinea, Paranemertes), remains roughly constant or decreases slightly with bivalves (Macoma secta, Transenella, Protothaca) and depth. This contrasts with the south sandflat where a crustacean (Leptochelia). The thalassinid Callianassa and Macoma secta are both absent, but shrimp Callianassa is abundant i n the high intertidal, where sessile forms are present (Table 3). and i n the marsh the shore crab Hemigrapsus constructs by forms probably loosens the sediment, burrow systems around the roots of the marsh plants allowing a greater interstitial volume for accumula- Salicornia and Distichlis. tion of water. An illustration such as Fig. 2 gives the impres- South sion that faunal changes are quite abrupt. Actually, North Sandflat Sandflat the sandflat is over km wide, so that changes i n Macoma Phoronop- Phoronop- elevation and substrate characteristics over the Core Callianassa flat are gradual. The fairly continuous environmental Increments bed bed secta bed nasuta gradients are reflected in broad ecotones where spe- cies intergrade. Top 10 cm 19.85% 19.53% 20.45% 32.03% Burrow Morphology and Feeding Type Middle (10-20 20.71% 18.61% 18.97% 19.15% Burrow morphology, maximum depth of burrowing, feeding type and habitat of the common trace producers Bottom on the sandflat at Bodega Harbor are given i n Table 4. (20-30 cm) 41.40% 18.42% 18.05% 11.91% It should be noted that many animals produce traces which do not f i t neatly into one category. For 3. has a U-shaped burrow the arms Table Averaged water content, i n percent from of cross- burrow, and the replicate cores on the north and south sandflats shrimp Callianassa produces a shaped 2). i n Bodega Harhor (after Ronan, 1975, Table burrow network. Surface traces and burrow openings of some of the trace producers as well as radiographs Ronan (1973, 1975) suggested that the tubes of of burrowed sediment are illustrated in Plates 1, 2. Phoronopsis viridis prevent the formation of ripple This table and the plates may serve as a guide i n id- marks as well as increasing the sheer strength of entifying biogenic structures on the sandflat. the subsurface sediments. This is i n accord with Featherstone and Risk (1977) who showed that dense An abbreviated form of this table summarizing stands of tube- building stabilize surface sediments some of t h e information about the traces made by t h e and inhibit the formation of ripple marks on tidal suspension and deposit feeders in high and low flats in the Bay of Fundy. tidal flat areas is given in Table 5. Many of the worm-like animals listed as deposit feeders are actu- THE OF BODEGA ally omnivores feeding on a variety of food resources within the substrate including encrusted grains, frag- Distribution of the Animals on the Sandflat ments of eelgrass, metazoans, and fecal pellets; the relative importance of food items differs from species Transect surveys across the north tidal flat of to species (Ronan, 1977). Other animals are trophic Bodega Harbor over a period of years have demonstra- generalists, utilizing both deposit and suspension ted a distinct pattern of zonation of infaunal spe- feeding, and the relative contributions of each are cies with respect to elevation above MLLW (mean low- difficult to assess. er low water). The vertical zonation for a few of the most abundant of the more than 200 infaunal Discussion species present on the flat is illustrated in Fig. 2. Salt Marsh Data given i n Tables 4 and 5 about the sediment relations of burrowers i n Bodega Harbor re- veal several interesting facts and trends. The first is that the sediment is highly reworked; rate of chur- ning by animals clearly exceeds rate of sediment accu- mulation or reworking by wave and current activity. Physical sedimentary structures such as ripple marks are produced but would not be preserved i n the sedi- mentary record.

A second notable fact is that deposit feeders are abundant throughout the sand flat, even i n the highest areas where the sediment is very clean aeolian sand. Figure 2. Distribution of major infaunal species This indicates that suspension feeders in vertical or relative to elevation above mean low tide, north U-shaped tubes are not the only or necessarily even sand flat, Bodega Harbor (after Ronan, 1973). the dominant infaunal elements i n clean intertidal Subsurface Trace Character - Habit at istics (Diameters are for Species structures produced by adult! LOWER Vertical cylindrical tubes 3 Phylum (rarely crooked) up t o cm in diameter.

Variable: Straight shafts Cone - shaped mounds of fecal pel - 22 cm; tube top is Deposit - feeding Produces mottled sediment by chaeta to J- shaped burrows 5 lets 1 cm high. Density = 15- flush withthe sur - omnivore (tentacle tentacular movement and late - formia Spira i n diameter; vertically 1 face. feeding at and ral movement of entire dorm. brancha oriented just below in eelgrass beds. Trace fossil: Skolithoa, Arenicolites, Eupolymnia Weakly agglutinated Cone - shaped, mucous - richmounds 13 cm Deposit - feeding burrow i n mound. Crescent is shapedU 3 in 1 cm high, 2 omnivore (tentacle Trace fossils: Arenicolites, diameter. density = feeding at surface Chomatichniss. - shaped hurrow 5 i n dia - 2 holes 1 cm in 18 cm Trace fossil: Arenicolites meter, with gallery connec - . density = raptorial, (irregular and modified). ting arms of surf ace omnivore. burrows make systemcomplex and Cone -shaped agglutinated of projects just igrates laterally 1 - 2 dwelling tuhe used as an aboveSWI. Produces cone shaped deposit - feeding produces mottled sedi - exoskeleton; up t o 1 cm Tapereddepressions end or mounds of finer omnivore ment. Trace fossil: diameter, sediment 1 cm high, 2 cm dia - meter. One den - Agglutinated (fine sand, sity extends up to 5 65 Deposit - feeders where eelgrass not shell vertical One density extends- many abundant. Trace fossil: cylindrical tentacles in diameter. sediment surf ace. Weakly consolidated, recum - Openings commonly occluded sedi - 15 cm; rare Active hurrower; can bent Y - shaped tubes 1 cm in 2 12 cm, most yore. Deposit by 3Q cm of sediment. diameter; may he inter - density i n5 top feeder; predator. caecoides prefers muddy connected . sand, N. californiensis pre - fers clean sand. Irregular burrows; in 25 cm;is deepest Deposit feeder, used i n diameter; many horizontally active burrowing locomotion and feeding, pro - oriented. Burrow systems do on duces localized expansions of not open t o surface. Den- flat. burrow wall, Trace fossil: = Planolites (burrow not back - filled). Not well known; ?Irregular May produce on surface cm15 Carnivore; preys at channel margins or Class pathways through upper 10 cm above shell. (East coast species on shallowdwell- in intertidal ponds. Produces ropoda of sediment! snail is produces ridged trails of ing bivalves, concave i n umbo of t o apex of shell, ing morphology on surface). prey, and also egg cases of Maximum Depth Surface Trace

s Irregular Feeding Type t Class " burrows " through of fine sand and fecal 10 suspen - i n muddy sands Yacoma nasuta which siphons extend to Lets 5 in diameter withcentral sion feeder, hut (south flat) near MLLW. surface. siphon hole3 mm in diameter. to deposit 'Bent nose clam " . only 1 siphon at surface. at surface. density (ind) Vertical roughly Key - hole shaped opening 2 cm in 95 Suspension nuttalli drical shafts, expanding dimension, Eeeder downward. Elliptical in Density cross - section. Phylum Arthropo - Y- shaped t o complexly Cone shaped mounds of sediment cm Suspension Prefers muddier sediment high, Callianassa. Same a Class Crus - burrows up t o 5 cm in and fecal pellets 5 - 10 cm feeder than Upogebia diameter. Walls are smooth 10- 15 i n diameter. commensals as Callianassa. ugettepsis and mud- lined. fossil: Phylum U- shaped with constricted Circular openings 2 cm in dia- 45 cmnear MLLW net to turn around i n bur - cm row and switch Echiurida arms. Arms 2Q t o 8Q cm meter, commonly with on middle suspension Urechis apart near surface. collar of mucous hound sediment flat. feeder excurrent openings. (Plate or low conical mound of fecal pel - sals include a scale worm, lets. Max,density on north sand - a pea and a fish. flat = Trace fossil: -MIDDLE FLAT- Subvertical to vertical Cone shaped in long, Trace fossil: Skolithos Phylum Annelida - Deposit feeder cylindrical 5 in diametex, 2 - 3 cmhigh with a 15cm to max Rubrocincta diameter consisting of mu- tinctiye " curl " of sediment at row depth. (Plate cous Found fine sand, mound top. Density = Phylum Nemertea Irregular Thin dendritic to anastomosing Unknown preys he 25 cm long. Paranemertes traces 1 wide or just on Trace fossil: (Plate sand surface, phoronids. Arthropo - Weakly cemented, multi - Circular holes 1 i n diameter, 12 concen - Extremely abundant i n mud - da Crus - occluded sediment. trated near sur - feeder dier areas; especially on tacea Density of holes may he of face, sandflat. Disturbs Leptochelia thousands upper 4 cm. Phylum Irregular: lies on left Stellate rills of light sand. Mound cm Most deposit Class valve and extends siphons 1 cm o r less i n diameter of sedimen CI primarily; peri - mid flat. Shows depth strat - and fecal material, sometimes with odically suspen - ification by size, with Plate siphon hole. Density sion feeds. longer deeper. Trace at cm MLLW, at fossil: Produces spiral traces on Suspension Prefers clean sands of s t aminea surface after heing exhumed feeder mid flat. (Plate tidal scour. -HIGH-FLAT Arthro Unlined multibranched gal Deposit feeder of high Class Crus- - Many Prefers clean sands - leries 1- 5cmin density (ind,) Many open - intertidal. Many connected to surface ings surrounded hy conical mounds (see section on natural Callianassa tical shafts, Bulbous of sediment and fecal pellets 5 history). Trace fossil: lings (turn arounds) at high and up to cm in diameter. Thalassinoides. 2C) hiforcations. Shallow U - shaped Basically oval openings 3 - 5 cm in Usually 2 or more individu - oregonesis tions with lateral diameter between Salicornia and (Scavenger) a l s per burrow. Forage on (hrachyuran) galleries. stalks. marsh at low tide. Trace fossil: Name named, similar t o burrows described by 4. structures produced and feeding habits of common animals i n the lower, middle Miller, and upper sandflat in Bodega Harhor.Data from (1973, 1977) and observations of Farmer, studentsat Bodega Marine Lab, and Miller. sands. Encrusted grains and plant fragments, and fe- NATURAL HISTORY AND ANIMAL SUBSTRATE RELATIONS OF cal pellets apparently provide sufficient food re- DOMINANT ANIMALS IN BODEGA HARBOR sources even though the percent of silt and clay, which correlates with organic carbon content, is low. Callianassa californiensis

Other points of interest involve the maximum One of the most abundant animals in the middle depths t o which suspension and deposit feeders bur- t o high intertidal areas in Bodega Harbor and other row. Suspension feeders live much more deeply than west coast lagoons is the burrowing thalassinid ghost deposit in both high and low tidal areas. Similarly, shrimp Callianassa californiensis (Plate (1967) found that suspension feeders on tidal shrimp are decapod crustaceans, of the Suborder flats live much more deeply cm) than do Reptantia which includes crawlers such as lobsters, tidal deposit feeders cm). The deeper burrow- hermit crabs, and crabs. Because one thoracic appen- ing of suspension feeders within the tidal flat may dage is reduced i n thalassinids, they are included i n be related to mobility. Suspension feeders tend to the Section Anomura (hermit crabs) by some taxono- be more sessile than deposit feeders (Schopf, 1978) mists; others put them i n the Section Macura (lob- and thus are less able to move t o avoid environmental sters). Thalassinids are widely distributed in mar- perturbations. (The suspension feeders also tend to ine and estuarine environments, where they are active be larger, which would also decrease mobility.) A burrowers. Burrows a r e highly variable, but most deep burrowing habit provides a buffer as protection consist of deep networks with both horizontal and changes affecting the sediment- water interface, vertical components marked by expansions serving as The fact that Urechis burrows more deeply on the mid turn- around areas. Trace fossils interpreted as flat, where exposure would be greater than on the low thalassinid in origin include Thalassinoides and flat, supports this hypothesis. Conversely, algal iomorpha; the latter has been extensively reviewed by and bacteria encrusted sand grains, metazoans, and Frey e t al, 1978. fecal pellets would all be most abundant i n the upper part of the sediment column; thus, it is not surpri- Along the central coast of California there are sing that most deposit feeders are relatively shal- 3 species of thalassinid shrimp: californiensis, low C. and Upogebia pugettensis. is rare. Upogebia pugettensis builds from U- or Y-shaped bur- A final observation is that on the average depo- rows i n mud o r muddy sand (Smith and Carlton, 1975). sit feeders burrow more deeply on the upper flat than MacGinitie (1934) generalized that californiensis lower flat (Table 5); the deep burrowing Callianassa preferred muddy sand substrates. However, i n both and Macoma secta are responsible for this trend. Mugu Lagoon (located 50 miles or Los and Possible reasons for this may be that the upper flat Bodega Harbor it is most abundant i n clean sand, oc- is exposed for longer periods of time during the ti- curring in densities of (Mugu Lagoon) and dal cycle and thus the animals must burrow deeper t o (Bodega Harbor) (Ronan, 1975; Miller, avoid dessication or that bird predation is more in- I n some areas the surface is covered with mounds, tense on the upper flat. craters and burrow openings which are the surface representations of the three dimensional burrow sys- t o Maximum tems of californiensis (Plate LOW FLAT Deposit Feeders Depth (in cm) ___Burrows- and burrowing: Ronan (1975) studied the morphology of burrows constructed by Callianassa U-shaped : californiensis in Bodega Harbor and determined that max depth: Eupolymnia, Glycera the burrows are basically shaped with hori- cm; Y or J shaped: 22, 15 zontal branches (Plate 2C). I n areas of Bodega Har- s.d. Cirriformia, bor where sediment is poorly oxygenated and black, range : 10- Complex branching: -C. californiensis burrows have rust- colored, oxygen- 25 cm Glycera Irregular: 10, 15, 25 ated halos, indicating that they are often flushed Pectinaria, Nepthys, Notomastus with oxygenated water. SUMMARY : Suspension Feeders The number of burrow openings per burrow system max depth: U-shaped : Urechis 45 is not known, nor is it readily determined by resin cm; Y or J- shaped: Upogebia 70 casting. (1967) found that under Vertical: Pista, Tresus 65, 120 conditions californiensis burrows have two or more range: Phoronopsis, Saxidomus 25, 95 openings. There are apparently many more burrow open- cm Irregular: Macoma nasuta 10 ings than shrimp i n Mugu Lagoon. Numbers of burrow Predators openings at the surface and number of shrimp present Irregular: Polinices 15 to a depth of 50 cm were determined for 48 cores from MID TO muddy and sandy sediments i n Mugu Lagoon (Miller, UPPER FLAT results are shown i n Fig. 3. SUMMARY : Deposit Feeders max depth: Y or J- shaped: Callianassa 60 C. californiensis unlined burrows collapse read- Complexly branching: i l y when excavated. This has lead several workers s.d. Leptochelia, Callianassa 12, 60 (Thompson and Pritchard, 1969; MacGinitie, 1934) t o range: 12- Vertical: Axiothella 15 suggest that the burrows are impermanent. However, 60 cm Irregular: Macoma secta 50 using radiography and laboratory observations Ronan found SUMMARY : Suspension Feeders only disturbed or abandoned parts of the burrow max depth: U-shaped: Urechis 60 system were collapsed. The facts that cm; Vertical: Phoronopsis 25 Table 5. Common producers of biogenic structures s. on the lower and middle t o upper sand flat range: and their burrow shapes and maximum depths. Data from 60 cm Table 4. 22

-sis continually removes fecal material from the bur- duces compaction, which, i n conjunction with void row, that is constructs turn- around spots and side space by the burrows themselves, results in branches, and that its burrows are inhabited by com- a very loosely packed, water- rich substratum. Bur- mensals Cryptomya californica) are evidence rowing also alters the vertical grain size distribu- that burrows are permanent, for these are attributes tion of the sediment. characteristic of other shrimp inhabiting permanent burrows. Ronan (1975) implanted laminae of fine- grained lead in the Callianassa beds i n Bodega Harbor and documented subsequent lead movement with a series of SANDY radiographs of cores; there was significant lead dis- MUDDY persal after 7 days. MacGinitie and MacGinitie (1949, 287) estimated that 20 t o 50 cc of sand are depos- ited around the entrance of a burrow of iensis per day. Using observed densities of shrimp and burrow openings per core i n sandy and muddy sediments i n Mugu Lagoon, and making the assumption that each individual deposits sediment around only one burrow opening i n a 24-hour period, it is esti- mated that in a 24-hour period californiensis de- 0 posits on the surface a layer of sediment 3.4 to 8.5 thick in the sandy areas, and 1.5 t o 3.8 mm thick i n t h e muddy areas. This implies that populations of californiensis in Mugu Lagoon overturn the top 50 cm of sediment, i n which most live, in a period of 117 t o 330 days. As Warme (1967) pointed out, the # HOLES rate of reworking greatly exceeds the sedimentation rate of approximately yrs. Figure 3. Number of burrow openings at surface vs. number of Callianassa recovered from .06 cores Mass properties of the sediment including water taken i n sandy and muddy sediments of Mugu Lagoon content and stability are strongly affected by (from Miller, Fig. 19). californiensis burrowing. As shown i n Table 2, Ronan found a two-fold increase with depth i n the The depth to which Callianassa californiensis water content of cored sediment from the califor- burrows is not certain. Although Thompson and niensis zone i n Bodega Harbor. This contrasts with chard (1969, 116) did not find Callianassa i n the cores from other parts of Bodega Harbor, i n which upper 45 cm during low tide, and Warme (1971) indi- water content decreases with depth. Penetrometer data, cated that its burrows extend 75 cm or more into the which reflects sediment stability, is given i n Table sediment, other evidence suggests that most of its 2 . I n Mugu Lagoon where the sediment texture is sim- activity occurs closer to the SWI. MacGinitie (1934) ilar to that of the north of Bodega Harbor found that the burrowing mostly is confined t o the but the density of Callianassa is higher, the pene- upper 45 cm. Fig. 4 shows the depth distribution of trometer sank 6 cmwith the maximum loading or 5 C. californiensis recovered from hand excavated cores This indicates that sediment stability is correlative i n sandy or muddy sediments i n Mugu Lagoon during low with amount of burrowing activity. tide. Most individuals occur from 15 t o 30 cm deep, although some may extend beneath the 50 cm core depth. Substrate preference feeding habit: I n muddy sediments californiensis rarely recovered Although MacGinitie (1934) reported that califor- from depths greater than 40 cm. niensis prefers muddy sediments, within Bodega Harbor and Mugu Lagoon it occurs in greatest densities in clean sands. In Bodega Harbor, however, it is present i n impacted clay muds of t h e high intertidal flat at MUDDY the southeast end of the harbor. Another thalassinid shrimp, Upogebia pugettensis, is more characteristic of muddy sediment than californiensis. Miller quantitatively compared densities of forniensis in sandy (mean 2.9% silt and clay; mean 2.05 phi sand fraction) muddy (mean 15.9% silt and clay; mean 1.99 phi sand fraction) sediments I23 within Mugu Lagoon. For 24 samples each .06 in diameter excavated t o a depth of 50 cm, there was a mean of 10.2 shrimp i n the sandy sediments compared 0-15 deep cm t o 4.7 i n t h e muddy sediment.

Feeding habit and substrate preference are closely related for many benthic marine invertebrates. Figure 4. Vertical distribution of Callianassa As documented by many authors Sanders, 1958; californiensis within sediment column i n sandy and Purdy, 1964) deposit feeders dominate i n fine grained muddy substrates, Mugu Lagoon (from Miller, sediments whereas suspension feeders are more abun- 17). dant i n sands deposited under higher energy condi- tions. In fine- grained sediments small particles of The burrowing of californiensis profoundly organic debris are included i n the sediment, and the affects the sediment. Burrows provide conduits for small sized grains provide a large surface area for well oxygenated water to penetrate to greater depths attachment of micro- organisms. In contrast, food that would otherwise be possible. Reworking of sedi- particles in higher energy environments tend t o be ment destroys primary sedimentary structures and re- kept i n suspension, leading to a predominance of 23 pension feeders. can result in as much as a foot of sediment being ad- ded to the flat. Apparently little mortality results Feeding habits of callianassid shrimp vary from from such mass deposition, for within 24 hours numer- species to species; there is also evidence that some ous mounds and crater- shaped burrow openings appear, species combine suspension and deposit feeding. indicating that shrimp have regained contact with the lianassa filholi primarily deposit feeds, sediment- water- interface. (1967) observed a specimen i n an aquarium flicking sediment into the water column and feeding Studies i n Mugu Lagoon have shown that &- on the material as it settled. C. major is considered forniensis responds differently to disturbance and to be a suspension feeder, but also been found t o sedimentation i n subtidal vs. intertidal areas. sift sand for microscopic particles of organic matter Peterson (1978, p. 345) effectively removed &- 1946, p. 77, 80). forniensis from a large area (6 x 20 of the sub- tidal channel by walking over the surface and There is little agreement about the diet of bing the top 5 t o 10 cm of sediment; 24 hours later lianassa californiensis or about the method of feed- no burrow openings had reappeared. Based on rate of ing. According t o (1934, p. 169) *- reappearance of burrow openings, intertidal ghost iforniensis is a deposit feeder, feeding on minute shrimp are more able to survive disturbance. Miller detrital particles in the sediment. He observed trampled a 25 square foot area in the inter- californiensis feeding on fine grained sediment and tidal zone densely populated with californiensis. also cited the presence of chitinous structures with- After a week it was impossible to distinguish between in the gut as evidence of a deposit- feeding habit. the trampled and undisturbed areas. Rate of appear- Ricketts and Calvin (1968) and (1971) concurred ance of burrow openings appeared to be uniform with this interpretation, and Peterson (1978) consi- throughout the disturbed area. Although ghost shrimp dered californiensis a deposit feeder in his study bodies were not seen on the sediment surface, many of the structure of the soft- bottom community i n Mugu Callianassa must have been killed by the vigorous Lagoon. However, Powell (1974) suggested that &- stomping. Apparent recovery i n the intertidal zone iforniensis may obtain a significant portion of its may occur by migration of Callianassa into disturbed nutrition from suspension feeding. He noted that g- areas rather than re- establishment of contact with lianassa californiensis is most abundant i n very clean the sediment-water interface by the original inhabi- sands i n Bodega Harbor, where the supply of organic tants. Average density of ghost shrimp i n sand inter- debris would be quickly depleted by a population of tidal sediment is (Miller, whereas i n deposit feeders. Although he interpreted structures the subtidal sand it is (Peterson, of the as adaptions for dealing with inorganic 1978). Space may be at a higher premium i n the inter- grains, he did not see their presence as necessitating tidal areas, resulting in more rapid reburrowing of a deposit feeding habit; they would also handle grains disturbed areas. I inadvertently taken into the digestive system during burrowing activity. Powell (1974, p. 29) found that A field method f o r determining the ability of the mouth parts of californiensis closely resemble californiensis to withstand rapid sedimentation was those of the suspension feeding Upogebia pugettensis. designed by Miller Roughly circular rings of thin fiberglass were sunk several centimeters into I n Bodega Harbor during much of t h e year the tides the sediment; the resulting enclosures were deposit large quantities of algae on the mately 1.5 m i n diameter and high. Before em- californiensis bed; the algae is subsequently buried placement, the number of burrow openings was deter- by drifting sand and sand brought t o the surface by mined. For each treatment a layer of sand of a pre- C. californiensis. Ronan (1975, p. 139) found sizable determined thickness was added. Burrow openings at pieces of in 29 of 40 samples excavated by hand the surface of the sand were subsequently counted at in the californiensis bed; some even had ghost intervals of hours t o days. shrimp clinging to them. The presence of plant mate- rial i n the gut contents of shrimp is further evidence This method has several inherent difficulties, that the detritus and its epiflora is an important of which the primary one is the lack of one-to-one food source. correspondence between californiensis and burrow openings. Other investigations of anastropic burial I n Mugu Lagoon there is not enough plant debris such as that by Kranz (1974) have been laboratory present in the clean sands t o support the large popu- studies on animals which do not produce burrow sys- lation size of californiensis. Miller tems. However, t h e advantages of adding sediment i n suggested that the main food source may be epigranular the field where the ghost shrimp are undisturbed out- flora. Rapid replenishment of bacterial films on weigh the difficulty of not knowing how many shrimp grains would be enhanced by the diurnal movement of are present and actively burrowing before and after nutrient and oxygen-rich waters through the highly addition of sediment. Other unavoidable problems in- porous sediment honeycombed with burrows. clude the facts that it is difficult to accurately count the large number of burrow openings, and that Conclusions about californiensis are difficult the areas surrounding the enclosures are disrupted by t o draw, although it is likely that deposit feeding the procedure. provides some o r all of its nutritional needs. Results of the burial experiments are shown i n Ability californiensis withstand sedimen- Fig. 5. Clearly Callianassa was capable of rapidly tation: There have been no systematic field studies in re- establishing contact with the after burial Bodega Harbor on the ability of californiensis to sediment, and its rate recovery tended t o be faster withstand rapid sedimentation. However, an extensive when thinner layers of sediment were added. The dis- bed of californiensis is located near drifting crepancies between the replicate experiments reflect blowing dunes so that clean sand is added t o the sand the large experimental error caused by the procedural flat by prevailing winds. Storm induced dune slumps difficulties discussed above. The difference between values for the 10 cm trials was partly due to the fact that no count of one of the burrow openings was 24 taken after 25 hours; the recovery paths of other Macoma experiments show that the recovery rate is usually greater during the first day, and levels off in the General: The cosmopolitan bivalve genus Macoma second and third days. The recovery path for the 14 is a highly successful component of many soft sedi- cm experiment was clearly in error; this can be at- ment marine communities of intertidal and shallow tributed to difficulty or error in counting holes, tidal nearshore environments (Thorson, 1957). Macoma either initially or, more likely, after 75 hours. is typically most abundant i n muddy substrates of silt or fine sand (Stanley, 1970). Species of this genus have long, flexible siphons that can be extended t o many times the shell length, and most are active bur- rowers. Depth of burrowing is i n the range of 4to 40 cm; most species show depth stratification by size. The usual life orientation is horizontal on the left w valve with the siphons projecting upward t o the sedi- ment-water interface (Plate w The feeding biology of the is poorly understood (see reviews by Yonge, 1949; Pohlo, 1969). Macoma balthica is a trophic generalist (Brafield and 1961). Tunnicliffe and Risk (1977) concluded that in the Bay of Fundy balthica obtains much of its nutritional requirements from bacteria attached to sand grains, but that suspension feeding is also 0 40 SO 80 important. HOURS AFTER BURIAL -Macoma -secta: Although 5 species of Macoma 5. have been reported from the sand flats at Bodega Har- Figure of Callianassa burrow bor only two species, secta and nasuta are abun- opening at surface after burial with various thick- dant a t mid to high tidal levels. These two species nesses of sediment (from Miller, Fig. 21). live sympatrically i n the harbor, but their relative abundances change both with regard to elevation above To test whether or not burial was fatal to MLLW (increased elevation results in increased expo- anassa, buckets containing individuals were filled sure time) and substrate character. Macoma secta is with 10 and 15 cm of sediment and buried so that the primarily a deposit feeder that burrows t o depths of top of the bucket was level with the sediment sur- 10 t o 30 cm, and is most common i n t h e better sorted face. None of t h e ghost shrimp had died after two looser sands at intermediate elevations. Maximum den- days, indicating that the weight of the sediment sities on the north sand flat are between 60 alone does not harm Callianassa. and 80 cm above MLLW (Fig. 2; Ronan, 1973). is uncommon a t elevations above 100 cm, probably due Although the experimental difficulties prohibit to the intense bioturbation of the sediment by Calli- drawing firm conclusions, Callianassa apparently is anassa californiensis at these elevations and the capable of surviving burial by sediment, and the rate significant change i n mass properties that result from of recovery is inversely proportional to the thickness its burrowing activities. of sedimentary overburden. The rate of reburrowing when thicker layers were added may well have been as When deposit feeding, Macoma secta extends its fast or faster than when thinner layers were added, incurrent siphon to the surface and vacuums detritus but less of the burrowing activity may have resulted from the upper 3-5 mm of surface sediment. It i n burrow openings a t the surface. The rapid rate of produces surface mounds of sediment and fecal recovery shown i n both burial and trampling experi- pellets up to 1 cm high. Often a siphon hole, 5 mm. ments probably reflects the pressure to use all avail- or less in diameter, is found a t the center of this able space caused by the high population density of mound. Most surface traces produced by Macoma secta Callianassa i n this high intertidal area of Mugu during feeding are stellate or arcuate rills of light- Lagoon. er- colored, coarser material that were left as a lag deposit by the incurrent siphon (Plate This Commensal relationships: Several species live with contrasts with the typical surface traces reported for C. californiensis, including a polychaete worm, two M. balthica in the Bay of Fundy which are shallow species of crabs, a clam, and a fish. The rela- pressions 5 cm. i n diameter (Tunnicliffe and Risk, tionships have been discussed in detail by 1977). However, densities of similar sized (1934, p. 171-173). The polychaete and the ca in t h e Bay of Fundy averaged many times crabs, both of which are small and laterally elonga- greater than i n Bodega Harbor. ted, subsist largely on organic debris uncovered by burrowing of californiensis; the crabs may also be --Macoma nasuta: Below an elevation of 40 cm, and facultive suspension feeders. Cryptomya californica along the margins of standing ponds at higher eleva- is a small suspension feeding clam with short siphons. tions, Macoma secta numbers decline and the density It accrues the benefits of a deep burrowing lifestyle of nasuta reaches a maximum. nasuta is a shal- by living adjacent t o the burrows of lower burrower than being found at depths sis which provide it with a supply of oxygen and of 10 to 12 cm. Maximum densities of are food- rich water. The fish feeds off debris with- reached below 40 cm. on the south where the i n the burrow 1934, p. 173); these fish substrate is muddier, firmer, and dryer (Ronan, 1975). commonly flop out on the sediment- water interface The surface traces produced by nasuta are not dis- during excavation of Callianassa. tinguishable from those of although stellate deposit feeding traces are produced infrequently. This species is primarily a suspension feeder on planktonic diatoms and flagellates. 25

Polychaetes fore taxonomically significant, is open t o question.

Polychaete worms are abundant i n the lower por- COMMUNITY STRUCTURE tions of the sand flat at Bodega Head. The burrow morphology and feeding habits of 5 common species Community Concepts and the Macoma Community have been studied extensively by Ronan 1977) and will only be briefly summarized here. The Macoma community of Bodega Harbor is recognized as one of Thorson's (1957, 1958) now clas- Radiographs of some polychaete burrows are sic parallel communities which he defined as recurring in Plate 2, and surface traces are illustrated in associations of characteristic genera or species that Plate H. Ronan (1977) found that one species are found i n similar environments throughout the world. could produce a variety of biogenic structures; even The wide recurrence and high fidelity of this associa- Cirriformia branchiata which lives in a vertical bur- tion leads to the question of whether the Macoma com- row could move laterally. Burrow morphology does munity is a biological entity that through not clearly reflect feeding habit; all the species has become integrated at a higher level, or studied were detrivores or omnivores, yet two produ- whether it is a group of species brought together be- ced U-shaped burrows. Gut contents of all species cause of similar physiological tolerances, but little included a variety of food items although some had evolutionary interaction. If the species merely live food preferences. Although some of the burrows were together because of congruent tolerances, the bounda- meandering and similar to traces attributed to the ries of the assemblage may represent environmental activity of deposit feeders, none were filled with discontinuities that exceed the tolerances of the com- fecal material. ponent species. On the other hand, i f the boundaries are interfaces of biological interaction, the group Phoronopsis viridis may represent an interacting, biologically accomodated community. These two possibilities represent the end- The Phylum Phoronida consists of a small group points in the classical debate concerning the nature of worm-like, vertical tube- dwelling lophophorate of communities (see 1962, for review). suspension feeders. They occur i n patches i n densi- ties up t o in the sand flats of Bodega The question of the importance of biological in- Harbor (Ronan, 1978). teractions in soft- bottom benthic communities has been of recent interest. Several studies have shown that The tubes of P. viridis are about 3 mm in dia- a range of biologic interactions probably are impor- meter and up to 25 cm long, and are composed of ag- tant. and Young (1970) found that glutinated sand grains in a chitinous matrix. Orien- ing activity may alter substrate characteristics at ted vertically, the top of the tube is about 8 cm be- the sediment- water interface so that the sediment is low the sediment- water interface (Plate the top easily resuspended; this effectively excludes suspen- of the tube is connected t o the SWI by a well formed sion feeders. Sanders et al. (1962) documented the burrow. During low tide the phoronids retract into deleterious effect of massive populations of a shallow their tubes, leaving a small hole in the substrate burrowing mud snail on a tube- building on (Plate G). tidal flats in Barnstable Harbor (Mass.) and Peterson (1978) provided evidence that competition for space Ronan (1975, 1978) studied the natural history is important in structuring the tidal channel of Phoronopsis viridis including its feeding habits community of Mugu Lagoon. Studies by (1974, and ability to withstand rapid sedimentation. During 1976) also using field experimental techniques demon- high tide P. viridis extends its lophophore and col- strated the importance of adult- larval interactions lects food items resuspended from the sediment- water i n determining boundaries of clusters of animals, and interface; gut analysis indicate that it prefers the effects of interactions between species i n deter- small 100 encrusted grains floculent material, mining the abundance of polychaetes within a and fecal pellets (Ronan, 1978, p. 483). In dense ty; she also reviewed the subject. patches lophopheres are stratified, allowing each to be fulling extended. Biological Interactions in Bodega Harbor

In laboratory and field experiments Ronan (1975) To determine the importance of biological inter- found viridis able to withstand anastrophic burial. actions in controlling the distributions of some dom- Under laboratory conditions, 100% had inant species in the intertidal sand flats of Bodega tablished contact with the sediment- water interface Harbor, Ronan (1975) conducted several manipulative within 45 minutes of burial by 2 cm of sand. field experiments. The experiments focused on six hours after burial by 16 cm of sand, 92% californiensis, Macoma secta, and Phoronopsis had burrowed t o the SWI. viridis; each is abundant in the high, middle or portions, respectively, of the north sand flat Ronan (1975) added sediment t o field enclosures i n Bodega Harbor. Details regarding the experimental 2cmat a time, and t h e number of lophophores present techniques and methods of data processing are given observed a t the next high tide. After 15 cm had been in Ronan (1975). added, 90%of the specimens were above the SWI. Careful dissection of the container indicated that Phoronopsis viridis - Macoma secta many of t h e phoronids had abandoned their tubes, and some were i n the process of tube rebuilding. I n Bodega Harbor, Macoma secta is rare or absent where dense stands of Phoronopsis viridis occur near Phoronopsis has been suggested as a modem pro- or on the southern sand flat MLLW where individuals ducer of the ichnogenus Skolithos. Certainly the of secta found i n phoronid beds on the north sand size, morphology and orientation of the two traces flat are recovered from uncharacteristic depths and are very similar. Skolithos also often occurs i n i n abnormal orientations. Ronan's (1975) experimental dense patches, as does However, whether field methods tested the hypothesis that tubes of or not this is a behavioral characteristic and Phoronopsis viridis prevent the lateral and vertical migration of Macoma secta, thereby largely exluding of any animal dependent upon constant contact with it from the lower sandflat. the sediment-water interface.

To test the effect of viridis on lateral Importance of Biological Interactions migration of secta, Ronan transplanted large clams In Determining Community Structure ( 3 cm) into plots densely (7 sparsely or unpopulated by phoronids. Careful The work of Ronan (1975) on ecological and excavation showed that the mean depth of burrowing substrate interactions in Bodega Harbor i n addition (11 cm) of was significantly less in the to other recent studies 1974, 1976) re- densely populated plot than i n the sparsely and un- veals that biological interactions play an important populated plots (15 cm, 17 cm, respectively); the role in structuring the communities of soft substrate burrowing depth correlated with the tops of the intertidal habitats. Some animals within these com- ronid tubes, which terminate at about 8 cm below the profoundly affect the physical environment SWI. Although i n the low density and phoronid- free through burrowing activity which can affect the mass treatments there was a strong correlation between properties, water content, grain size and distribu- depth of burrowing and size of secta, this rela- tion and organic content of the sediment. This effect tionship did not hold i n dense stands of viridis. on the sediment i n addition to spatial competition, commensal relationships (commensals are abundant on The south sandflat i n Bodega Harbor consists of the sand flats in Bodega Harbor; see discussion of less well sorted and more compacted sand than the Callianassa) and predator- prey interactions is an im- north sandflat, with a greater proportion of fine portant component in the integration and structure of grained sediment. secta burrowed t o shallower these communities. depths in phoronid- free plots on the south plot, pre- sumably because of the greater degree of compaction. Is t h e Macoma community integrated in the evolu- The burrow depths were shallower than sediments trans- tionary sense, or is it a group of species with no planted from the north to south sandflat or from the real interactive structure? Although the evidence is south to north sandflat, or in phoronid- free sediment preliminary, recent studies do suggest that biologi- in the north flat. Lateral migration of Macoma secta cal interactions are important i n producing ecotonal was not significantly altered by presence of phoronid boundaries between species groups that are as impor- tubes. tant as those associated with tidal height or sub- strate characteristics. Both physical and biological The results of the vertical burrowing experi- factors enter into the determination of structure in ments suggest that although dense stands t h e Macoma community i n Bodega Harbor. The fact that do prevent deep burrowing by Macoma secta, the this association (one of Thorson's parallel communi- mass properties of the sediment, particularly water ties) recurs in similar environments throughout the content and degree of greatly affects world hints at the importance that biological integra- mobility of this bivalve. The effects of the tion may have played during its evolution. nid tubes are two fold: they provide obstacles to burrowing and they also serve as sediment stabilizers GERMAN RANCHO AND BODEGA HARBOR which further inhibits burrowing. ICHNOCOENOSES: A COMPARISON Callianassa californiensis - Phoronopsis viridis It is difficult to compare individual traces in Although viridis tends to live in lower ele- Bodega Harbor with those of the German Rancho Forma- vations on the tidal flat than californiensis, tion. Ophiomorpha and Thalassinoides in the German -P.~ viridis extends higher onto the flat and into con- Rancho probably were produced by thalassinid crusta- tact with Callianassa on the margins of several drain- ceans related to Callianassa californiensis and age channels. Zones of overlap 1 to 5 m wide between pugettensis. Upogebia is a suspension feeder the species are sparsely populated; these areas pro- whose burrow is mucous lined and permanent, whereas vide opportunities for studying interactions between C. californiensis probably is a deposit feeder whose these two animals. burrow is unlined and impermanent. A lined burrow would be advantageous for a suspension feeder. Per- Ronan (1975) assessed the impact of haps Ophiomorpha was produced by a suspension feeder niensis on viridis by planting viridis in plots and Thalassinoides by a deposit feeder; however, the within Callianassa bed and in a caged plot within the gradations between the two ichnogenera at Stump Beach Callianassa bed from which the shrimp had been exclu- are not compatible with this interpretation. Results ded. Excavation and radiography of box cores from of experiments on the escape potential of Callianassa the plots revealed that when Callianassa is present, californiensis suggest that the producers of the phoronid tubes were commonly at unnatural depths morpha and Thalassinoides may have been able to with- or orientations (20% of tubes; Plate or actually stand deposition of thick sands. broken (about 10%). This contrasts with the plot from which Callianassa was excluded, where only about More generally, the rocks at Stump Beach were 1%of the phoronid tubes were broken and 2% were dis- deposited i n deep sea fan environments with abundant oriented. evidence of environmental perturbation (debris flow, turbidites). This environmental instability would Ronan (1975) found former occupants of the dis- have led t o the establishment of " physically control- turbed tubes free in the sediment, constructing new led community" (Sanders, 1969) i n which physical tubes. Ronan suggested that the problem of maintain- stress and unpredictability would inhibit various eco- ing contact with the sediment-water interface and logical processes and subsequent development of a di- repair of damaged tubes is energetically expensive verse community with complex interactions between spe- and that this may result in viridis being abundant cies. In contrast, Bodega Harbor sediments are highly only when Callianassa is rare. He generalized (1975, bioturbated and contain a diverse fauna. Here physi- p. 156) that extensive sediment disruption by cal environmental perturbations appear t o be much less anassa may prohibit development of large populations intense. This has led t o the apparent development of 27

a " biologically accomodated" community (Sanders, feed differently, although they are 1969) i n which diverse ecological interactions are closely related and morphologically similar. This well developed; the existence of these interactions behooves paleoecologists to be careful in assigning has been demonstrated by Ronan (1975) but more work feeding habits for analysis of trophic structure. is needed t o evaluate their importance. Finally, study of the infauna i n Bodega Harbor PALEOECOLOGIC IMPLICATIONS has indicated that the structure of soft- bottom mar- ine communities is dependent upon biological inter- The degree t o which information about modem actions as well as on physical environmental para- traces and their producers is applicable in the in- meters. Thus, the distributions of trace fossils in terpretation of ancient biogenic structures is not some cases reflect ecological as well as physical con- clear. As indicated by Seilacher (1964) and Frey trols. Although the biological interactions are dif- (1970, 1971) and by the paucity of trace fossil ana- ficult or impossible to deduce from the rock record logs for Bodega Harbor traces, there are severe lim- this does not negate either the importance of these itations to the uniformitarian approach t o interactions or the utility of trace fossils in logy. Of the 20 animals listed in Table 4, only environmental and paleoecological reconstructions. about half (11) produce traces that can be assigned Paleoecologists and sedimentologists should be aware confidently to an ichnogenus. There is no evidence of these interactions, should look for evidence of that back filled burrows and traces with spreite, them i n the rock record, and use knowledge of their both common i n t h e fossil record, are being produced importance i n modern communities t o develop more on the Bodega Harbor sandflat; as Ronan (1975, 1977) realistic models of the structure of ancient marine noted, the polychaetes discard fecal material outside communities and of the distributional controls of of their burrow. Lack of direct correlation between ancient traces. modern and ancient traces may well be due, i n part, to preservational enhancement of biogenic structures ACKNOWLEDGMENTS by diagenesis (Seilacher, 1964; Frey, 1971). We thank Mary Graziose and Peggy Wrenne f o r pre- Although ichnologists cannot rely on direct com- paring the photographs and Kelly Boyte f o r typing the parison with modern traces to interpret the ethology manuscript. G.A. critically read the paper and of most trace fossils, they can use knowledge of the made helpful suggestions. We appreciate the help and relations between anatomy, behavior and trace morph- support of Jana Ronan. Financial support was provided ology t o develop internally consistent interpreta- by the Vanderbilt University Research Council. tions of the behavior represented by ichnofossils. The large amount of information compiled about be- REFERENCES CITED havior of animals living on and in sediment and their effects on the sediment by many workers many Bailey, E.H., Irwin, W.P. and Jones, D.L., 1964, papers by Rudolf Richter, Schaefer, 1972; Frey and Franciscan and related rocks and their signifi- Howard, 1969; Frey, 1970; Howard and Dorjes, 1972; cance i n the geology of western California: Dorjes and Hertweck, 1975; Basan and Frey, 1977) Calif. Div. Mines and Geology Bull., v. 183, provides a solid data base for broadly uniformita- 177 p. rian interpretations of ancient lebensspuren. In Barnes, R.D., 1974, Invertebrate Zoology, 2nd editi- addition, distributional patterns of bioturbation on: Philadelphia, W.B. Saunders, Inc., 870 p. and biogenic structures within rock sequences have Basan, P.B. and Frey, R.W., 1977, Actual- paleontology been compared with those i n modem nearshore sedi- and of salt marshes near Sapelo ments; the results have aided i n the interpretation Island, Georgia, T.P. and Harper, of the ancient environments Howard, 1972). J.C., Trace Fossils 2: Geol. Jour. Spec. Issue 9, p. 41-70. Data about the natural history, distribution Boyd, M.J., 1971, Fouling community structure and de- and structures produced by Bodega Harbor trace pro- velopment i n Bodega Harbor, California: Unpub. ducers give new information regarding Dissert., of California, Davis, interpretations of individual trace fossils and 191 p. factors controlling the distributions of their pro- Brafield, A.E. and R.C., 1961, The behavior ducers. First, the relationship between burrow mor- of Macoma balthica: Jour. Marine Biol. Assoc. phology and feeding mode is not clear cut. Both U.K., v. 41, p. 81-87. deposit feeders and suspension feeders inhabit ver- Chipping, D.H., 1971, Franciscan Formation near tical (trace and U-shaped burrows Bodega Bay and Jenner, California, Lipps, J. (trace Inhabitants of these and Moores, E.M., Geologic guide t o the northern burrow types may tend to be less mobile than other Coast Ranges, Point region, California: burrowers, and suspension feeders may tend t o be Ann. Field Trip Guidebook, Geol. Sacramento, sessile (Schopf, but there are many deposit p. 47-57. feeders and some carnivores, especially polychaetes, Connell, J.H., 1961, The influence of which live in vertical burrows (Barnes, 1974). competition and other factors on the distribution Secondly, a consistent relationship between grain of the barnacle Chthamalus stellatus: Ecology,

size of the sediment and feeding mode of t h e infauna v. 42.I.D . 710-723. is lacking. The relatively clean sands of the Bodega Crimes, T.P. and Harper, J.C., 1970, Trace fossils: Harbor sandflat support a diverse deposit- feeding as Geol. Jour. Spec. Issue 3, 547 p. well as suspension- feeding infauna. Some of t h e most -- , 1977, Trace fossils 2: common suspension feeders (Upogebia, Macoma nasuta) Geol. Jour. Spec. Issue 9, 351 p. are most abundant i n the muddiest areas. Thirdly, Dayton, P.K., 1971, Competition, disturbance and com- the feeding mode of most animals is not well under- munity organization: the provision and subse- stood and many combine feeding methods. Macoma quent utilization of space i n a rocky intertidal and nasuta feed both by ingesting sediment community: Ecol. Monographs, v. 41, p. 351-389. and by sucking water and suspended materials through Devine, C.E., 1967, Ecology of Callianassa filholi their siphons. Callianassa californiensis and 1878 (Crustacea, Thalassinidea): 28

Roy. New Trans., v. 8, p. 93-110. Plate 1. Dorjes, and Hertweck, G., 1975, Recent ses and ichnocoenoses i n shallow- water marine A. Microtopography produced on sediment surface environments, Frey, R.W., The study of trace by the burrowing of Callianassa fossils: New York, Springer- Verlag, p l 459-491. on the upper sandflat. Featherstone, R.P. and Risk, M.J., 1977, Effect of tube- building on intertidal sedi- B. Drag marks and fine dendritic trace of ments of the Minas Basin, Bay of Fundy: Jour. nemertes peregrina on the sediment surface. of Sed. Petrology, v. 47, p. 446-450. Frey, R.W., 1970, Environmental significance of Rec- C. Ripple marks and biogenic structures on sedi- ent marine lebensspuren near Beaufort, North ment surface. 1. Burrow opening of Urechis Carolina: Jour. Paleontology, v. 44, p. note surrounding mound of sediment. 519. 2. Stellate siphon mark of Macoma secta. , 1971, Ichnology - the study of fossil 3. Opening of Phoronopsis viridis burrow. and Recent lebensspuren, fossils: a 4. Axiothella rubrocincta burrow opening field guide to selected localities in Pennsyl- surrounded by material ejected by worm. vanian, Permian, and Tertiary rocks of Texas and related papers: School of D. Specimen of Callianassa californiensis about science, LSU Miscellaneous Publ. 71-1, p. of natural size. 125. , 1975, The study of trace fossils: New E. Two specimens of the echiurid Urechis caupo; York, Springer- Verlag, 562 p. larger is approximately 8 inches long. and Howard, J.D., 1969, A profile of genic sedimentary structures in a Holocene bar- F. Specimen of the polychaete Glycera robusta. rier island- salt marsh complex, Georgia: Gulf Coast Geol. Trans. V. 19, p. G. Occurrence of Macoma secta, as indicated by 444. siphon mark t o t h e left of the penny i n bed , Howard, J.D. and W.A., 1978, of Phoronopsis viridis. Holes are openings Ophiomorpha: its morphologic, taxonomic, and at the surface of the phoronid burrows. environmental significance: Palaeogeogr., Paleoclim., Palaeoecol. v. 23, p. H. 1. Callianassa californiensis burrow open- ~ Howard, J.D., 1968, X-radiography for the examination 2. Tubes of Axiothella rubrocincta of burrowing i n sediments by marine organisms: and surrounding Sedimentology, v. 11, p. , 1972, Trace fossils as criteria for I. Radiograph of box core taken i n phoronid bed recognizing shorelines in the stratigraphic rec- after introduction of Macoma secta. Note ord, J.K. and Hamblin, W.K., Recogni- that Macoma secta is either vertically orien- tion of ancient sedimentary environments: ted (not its normal position) or horizontal Econ. Paleontologists and Mineralogists Spec. at a shallow depth above the tops of the Publ. 16, p. phoronid tubes. (Approximately of actual and Elders, C.A., 1970, Burrowing pat- size; from Ronan, 1975, Plate 8). terns of haustorid amphipods from Sapelo Island, Georgia, Crimes, T.P. and Harper, J.C., Trace Fecal pellets of californiensis; key at fossils: Geol. Jour. Spec. Issue 3, p. left for scale. and Dorges, J., 1972, Animal-sediment relationships in two related tidal flats, K. Re-entry trace of Protothaca staminea. A elo Island, Georgia: Jour. Sed. Petrology, shallow burrowing bivalve, P. staminae is v. 42, p. easily exhumed, thus these are fairly Johnson, R.G., 1972, Conceptual models of common. marine i n Schopf, T.J.M., Models i n paleobiology: San Francisco, Freeman, Cooper and Co., p. 148-159. REFERENCES CITED, CONT. J.A., 1979, Deep-sea traces from the cent tral Arctic: an analysis of diversity: Deep- , Utility of trace fossils in sea Res., v. p. 1185-1198. paleoenvironmental interpretation: the distribu- Koenig, J.B., 1963, The geologic setting of Bodega tion of hiogenic structures in three modern and Head: Mineral Information Service (California), ancient nearshore environments: Unpuh. V. 16, p. 1-10. Dissert., UCLA, 373 p. Kranz, P.M., 1974, The anastrophic hurial of Osgood, R.G., 1975, The paleontological significance and its paleoecological significance: Jour. of trace fossils, Frey, R.W., The study of Geology, v. 82, p. 237-265. trace fossils: New York, Springer- Verlag, Lawrence, D.R., 1968, Taphonomy and information los- p. 87-108. ses in fossil communities: Geol. America R.T., 1966, Food web complexity and species Bull., V. 79, p. 1315-1330. diversity: Naturalist, v. 100, p. 65-75. MacGinitie, G.E., 1934, The natural history of Peterson, C.H., 1978, Competitive organization of the californiensis Dana: Am. Midland Natu- soft- bottom macrobenthic of southern ralist, v. 15, p. 166-177. California lagoons: Marine Biology, v. 43, and MacGinitie, N., 1949, Natural p. 343-360. history of marine animals: New York, Pickerill, R.K., 1976, Application of ichnology t o the Hill Book Co., Inc., 473 p. study of ancient marine benthic community Miller, M.F., Middle and Upper Ordovician ecology: a discussion and case example: Mari- genic structures and paleoenvironments, southern time Sed., v. 12, p. 49-52. Nevada: Jour. Sed. Petrology, v. 47, p. Pohl, M.E., 1946, Ecological observations on 1338.

30

B A

Plate 2

A. Radiograph of a field collected box core containing the biogenic structures produced by Notomastus (Polychaeta; Capitellidae). Arrows point to magnus proboscis strike mark. (From Ronan, 1977, Fig. 5).

B. Biogenic structures produced by the polychaetes crescentis (left), Pectinaria (right center) and Nepthys caecodes (arrow at extreme right). Sediment within the digestive tract of Pectinaria shows up as narrow darker zone. Fine grained sediment at top of radiograph appears lighter than underlying sediment; it was transported to the sediment surface by the feeding of naria californiensis. (Radiograph of field collected box core; from Ronan, 1977, Fig. 4).

C. Radiograph of box core collected from Callianassa bed into which phoronids were transplanted. Note broken phoronid tubes at Callianassa burrows such as and burrow galleries (c) have dis- rupted the phoronid tubes (from Ronan, 1975, Plate 12).

D. Radiograph of biogenic structures produced by the polychaete Glycera robusta. Note broadly U-shaped with cross- connecting galleries and abruptly terminating exploratory burrows (from Ronan, 1977, Fig, 3). 31

nassa major Say at Beaufort, North Carolina: 183 p. Ecology, V. 27, p. 71-80. Stanley, S.M., 1970, Relation of shell form t o life Pohlo, R., 1969, Confusion concerning deposit- feeding habits of the Geol. i n the Tellinacea: Proc. Malacological America Mem. 125, 296 p. London, v. 38, p. 361-364. Thompson, R.K. and Pritchard, A.W., 1969, Respiratory Powell, R.R., 1974, The functional morphology of the adaptions of two burrowing crustaceans fore- guts of the thalassinid crustaceans, nassa californiensis and Upogebia pugettensis: lianassa californiensis and Upogebia Biol. Bull. v. 136, p. sis: Univ. Calif. Publ. Zoology, v. 102, 41 p. Thorson, G., 1957, Bottom communities (sublittoral to Purdy, E.G., 1964, Sediments as substrates, Imbrie, shallow shelf), in Hedgpeth, J.W., Treatise on J. and Newell, N.D., Approaches t o marine ecology and paleoecology: Geol. gy: New York, John Wiley and Sons, p. 238-271. America Mem. 67, p. 461-534. D.C., 1967, Biogenic reworking of intertidal , 1958, Parallel level bottom communities, and sediments i n Barnstable Harbor their temperature adaptation, and their “balance” and Buzzards Bay, Massachusetts: Jour. Geol., between predators and food animals, V. 75, p. 461-476. Traverso, A.A., Perspectives i n marine biology: , 1973, The influence of deposit- feeding Berkeley, Univ. Calif. Press, 621 p. benthos on water turbidity and nutrient recyc- Tunnicliffe, V. and M.J. Risk, 1977, Relationships be- ling: Jour. Science, v. 273, p. 1-22. tween the bivalve balthica and bacteria , 1975, The paleoecological and environ- in intertidal sediments: Basin, Bay of mental significance of trace fossils, Frey, Fundy: Marine Res., v. 35, p. 499-507. R.W., The study of trace fossils: New York, J.E., 1967, Graded bedding i n the Recent sedi- Springer- Verlag, p. 147-160. ments of Mugu Lagoon, California: Jour. Sed. and Young, D.K., 1970, The influence of Petrology, v. 48, p. 419-425. deposit feeding benthos on bottom sediment , 1971, Paleoecological aspects of a modern stability and trophic structure: coastal lagoon: Univ. Calif. Publ. Geol. Sci. Jour. Marine Res., v. 28, p. 150-178. v. 87, 110 p. Ricketts, E.F. and Calvin, J., 1968, Between Pacific Whittaker, R.H., 1962, Classification of natural tides, 4th ed. (revised by J.W. Hedgpeth): munities: Botan. Rev., v. 28, p. 1-239. Stanford, Stanford Univ. Press, 614 p. S.A., 1974, Polychaete abundance patterns: T.E., Jr., 1973, An investigation of biologic- Ecol. Monogr., v. 44, p. 171-187. ally produced sedimentary structures within , 1976, Adult- larval interactions in dense the Bodega Marine Laboratory sandflat: unpub. assemblages: patterns of abundance: M.S. thesis, Univ. of Calif., Davis, 82 p. Jour. Mar. Res., v. 34, p. 25-41. , 1975, Structural and Yonge, C.M., 1949, On the structure and adaptations of cal of a modern marine soft- sediment the Tellinacea: deposit- feeding community: an experimental field study: Unpub. Phil. Trans. Roy. London 234, Dissert., Univ. of Calif., Davis, 220 p. p. 29-76. , 1977, Formation and paleoecologic recognition of structures caused by marine an- nelids: Paleobiology, v. 3, p. 389-403. , 1978, Food resources and the in- fluence of spatial pattern on feeding in the phoronid, Phoronopsis Biol. Bull., 154. 472-484. Sanders, H.L., 1958, Benthic studies in Buzzard’s Bay I. Animal-sediment relationships: Limnol. Oceanogr., v. 3, p. 245-258. , 1969, Benthic marine diversity and the stability- time hypothesis, Woodwell, G.M. and Smith, H.H., University and stability in ecological systems: Brookhaven Symposia i n Biology, 22, p. 71-81. , Goudsmit, E.M., Mills, E.L. and son, G.R., 1962, A study of the intertidal fauna of Barnstable Harbor, Massachusetts : Limnol. Oceanogr., v. 7, p. 63-79. Schaefer, W., 1972, Ecology and palaeoecology of mar- ine environments, Chicago, Univ. of Chicago Press, 568 p. Schopf, T.J.M., 1978, Fossilization potential of an intertidal fauna: Friday Harbor, Washington: Paleobiology, v. 3, p. 261-271. Seilacher, Adolf, 1964, Biogenic sedimentary struc- tures, Imbrie, and Newell, N.D., Approa- ches t o paleoecology: New York, John Wiley and Sons , p. 296-316. Smith, R.I. and Carlton, J.T., 1975, Light’s manual: intertidal invertebrates of the central Cali- fornia coast, 3rd ed: Berkeley, Univ. of Calif. Press, 716 p. Standing, J., Browning, B., and Speth, J.W., 1975, The natural resources of Bodega Harbor: Calif. Dept. Fish and Game Coastal Wetlands Series