Ecosystem Modeling in Cobscook Bay, Maine: A Boreal, Macrotidal Estuary 2004 Northeastern Naturalist 11(Special Issue 2):325–354 Late 20th-Century Qualitative Intertidal Faunal Changes in Cobscook Bay, Maine THOMAS J. TROTT * Abstract - Late 20th-Century changes in the intertidal distributions of macroinvertebrates within five sample sites in the Cobscook Bay, ME, region were evaluated by comparisons with qualitative baselines, some as old as 35 years. These baselines were generated by the Maine State Planning Office Critical Areas Program (1970–1987), which recognized the unique distributions of macroinvertebrates and high diversity of intertidal communities in Cobscook Bay that had attracted many zoologists dating back to the early 1800s. The sample sites were critical invertebrate areas registered by the Critical Areas Program between 1968 and 1976. None of the sample sites had been re-examined for at least 20 years, and all but one had been evaluated at least twice previous to this study. Many species, including those whose presence was used to designate habitats as critical, were common or abundant in original site descriptions, but rare or absent in 2002. The dramatic change in community composition away from species typical of hard bottoms to established mussel beds suggests a faunal shift has occurred. The principal driving force that produced this change is proposed to be disturbance from increased sedimentation that altered intertidal habitats. Potential sources of this disturbance and possible cascades that followed are discussed. Introduction Cobscook Bay, ME, is exceptional because of the extreme tidal range that characterizes this macrotidal estuary and creates an ecosystem with biodiversity unsurpassed at lower latitudes (Trott and Larsen 2003). Postglacial changes in climate and ocean circulation superimposed on the large tidal amplitude contributed to the assemblage of unique communities found there (Bousfield and Thomas 1975). However, compared to estuaries similar in size and commercial importance, general awareness of this ecosystem is poor in spite of its distinctiveness. Cobscook Bay, a rock-framed macrotidal estuary, is both the furthest east and the only boreal estuary on the eastern coast of the United States. Though Cobscook Bay serves as the basin for two rivers, the amount of freshwater entering the Bay is negligible (Campbell 2004) and differentiates it from many macrotidal estuaries which often receive substantial river input (Gleizon et al. 2003). All the water entering the innermost region *R.S. Friedman Field Station, Edmunds, ME 04628; Biology Department, Suffolk University, 41 Temple Street, Boston, MA 02114; codfish2@earthlink.net. 326 Northeastern Naturalist Vol. 11, Special Issue 2 of the Bay, formed by confluence of Whiting and Dennys Bays, must enter through a single narrow opening restricted by an island. All processes occurring in the Inner Bay dependent on mixing with offshore waters are reliant on this single input. Cobscook Bay has a historical record matched by few other marine locations in the United States since Eastport, ME, was where the United States Fish Commision established its second research station after recognizing the decline of fisheries on the eastern US coast in the mid-1800s. Before this, the intertidal communities of Cobscook Bay had drawn zoologists to this remote region for decades. The highly diverse macroinvertebrate fauna of Cobscook Bay is documented by a rich historical chronology that spans 162 years of collection records because of this attention (Trott 2004). Information of this kind is rare and useful in establishing species distributions and general knowledge of geographic ranges. It provides, however, no hint of relative abundance of the benthic community that inhabits intertidal rocky shores and mudflats. With the exception of Larsen and Gilfillan (2004), no quantitative information from systematic sampling in Cobscook Bay has been published, and even their study sampled exclusively subtidal commuities. One focus of historical ecology is the examination of long term changes in communities. These studies are difficult since pre-existing databases are required to make time-series comparisons. Surveys of marine benthic communities are few, especially in North America. Long-term changes have been best studied in northern Europe where ecological investigations of marine communities have more commonly surveyed community composition (e.g., Buchanan 1963, Ford 1923, Kühne and Rachnor 1996). Subsequent studies have revealed interesting changes in community composition and biogeography (Bamber 1993, Shillabeer and Tapp 1989, Tyler and Shackley 1980). The rarity of benthic community surveys in North America makes them attractive for investigating community change and studies for comparison with the eastern Atlantic. Five specific locations in the Cobscook Bay region were the focus of qualitative ecological assessments of intertidal macrobenthic invertebrates in the 1970s and 80s, as were other noteworthy invertebrate communities along the coast of Maine (Maine State Archives). These evaluations were conducted as part of the Critical Areas Program (CAP) of the Maine State Planning Office that operated officially from 1970 to 1987 and aimed to identify rare and unique features throughout the State. Many publications resulting from the CAP documented ecologically significant intertidal areas within Maine and the fauna found within these special locations (Doggett et al. 1978; Gilbert 1977a,b; Speel 1978; Weiss 1980). Because of their qualitative nature, measuring changes using the databases produced by these studies is difficult, 2004 T.J. Trott 327 though highly attractive, since such location-specific historical records are rarely available. Detailed site sketch maps, land owner information and tax maps, site descriptions, evaluations based on CAP criteria, species lists, and observations are available for most critical invertebrate areas. For some species, relative measures of their abundances based on collection effort for particular localities also exist. During the past six years of field studies in Cobscook Bay, many intertidal species once easily found, according to previous investigators and local knowledge, were rarely or never seen by the author and his colleagues who had sampled Cobscook Bay years before he began ecological studies there. The purpose of this investigation was to measure change, if any, in intertidal communities with assemblages of invertebrates that originally attracted attention to each critical invertebrate area. These were a group of species selected by their inclusion in previous CAP records, which included, but was not limited to, the circumboreal prosobranch gastropod, Margarites helicinus, the Arctic bivalve, Mya truncata, and the northern Atlantic brachiopod, Terebratulina septentrionalis (see Table 1). Ecological significance of critical invertebrate areas differed depending on whether they were located inside or outside the Cobscook Bay complex. Within Cobscook Bay, ecological signifi- cance of the Critical Invertebrate Areas was based on the change in vertical zonation of many cold water species from subtidal into the shallow intertidal zone. This expansion of vertical range is possible because of Table 1. Target species assigned ecological significance in historical reference information for the five critical invertebrate area sample sites in the Cobscook Bay region. The current distribution of each species is indicated by presence (+) or absence (–) for each sample site. Critical invertebrate areas were registered based on the presence of the species that abbreviations appear next to. West Outer Quoddy Gleason Birch Wilbur Crow Species Head Point Island Neck Neck Margarites helicinus (OBI, WN) + - + + + Colus stimpsoni (CN) - - - - - Neptunea lyrata decemcostata (CN) - - + 1 - - Lacuna vincta (CN) + - - + + Astarte complex (CN) - + - - + Mya truncata (GP) - - + [+] - + [+] Musculus discors (CN) + + + - + Musculus niger (CN) + - - - - Priapulus caudatus (CN) - - - - - Terebratulina septentrionalis (CN) [+] - + 1 - + 1 Solaster endeca (CN) - - - - - Crossaster papposus (CN) - - - - - Abbreviations: Gleason Point (GP); Wilbur Neck (WN); Outer Birch Island (OBI); Crow Neck (CN); shell [+]. 1Observed during post quadrat sampling reconnaissance. 328 Northeastern Naturalist Vol. 11, Special Issue 2 the physical oceanography and meteorological conditions of this boreal, macrotidal estuary that create cooler intertidal environments. Outside of Cobscook Bay, the high species diversity and the unusual southern distribution of Arctic species made these critical invertebrate areas ecologically significant. The qualitative baselines generated for each Critical Invertebrate Area in Cobscook Bay and vicinity served as reference points to measure changes in invertebrate distributions, community structure, and diversity. This study marks another point in the timeline for each critical invertebrate area in the Cobscook Bay region. Methods All five locations of this study were sampled in July to coincide within one month of the original historical reference data marked by dates that ranged from June through August on field evaluation forms obtained from the Maine State Archives. The duration of the quadrat sampling interval at each site ranged from 40 to 50 min. Each transect, later described, was sampled synoptically within a single low tide period by a team of three. Authorities for all species identified can be found in elsewhere in this issue in Trott (2004). Additional data recorded included tidal amplitude, air and sea water temperatures, weather conditions, and general observations of the sample site for evidence of disturbance and features identified in previous reports. Faunal sampling methods Location of study areas. The five study areas selected for evaluation were all in the Cobscook Bay region (Fig. 1). Three areas were situated within the Inner Bay complex: Crow Neck and Wilbur Neck, both located in Dennys Bay, and Outer Birch Island located in Whiting Bay. The two remaining areas, Gleason Point and West Quoddy Head, are located north and south, respectively, of the entrances to Cobscook Bay. Gleason Point borders Passamaquoddy Bay, and West Quoddy Head projects into Grand Manan Channel. Sample site selection. Reconnaissance surveys of each documented location were conducted as part of a pre-selection process for positioning sample transects. The primary goal was to locate one or more target species with distributions particular to each sample site (Table 1). Target species were not indicator species. Target species is a term created by this study. These species were considered ecologically significant among the set of those referenced from each site in CAP field evaluation forms for any of three reasons: (1) a historical record more extensive than a listing in critical area field evaluations only, (2) a recognition of uniqueness in geographic distribution by the CAP that was used in the decision on registering an area, and (3) available estimates of the relative density of these species made at the time of site evalu2004 T.J. Trott 329 ation. A secondary goal was to construct a qualitative species list of all species encountered in the process of searching for target species. Each mission began an hour before low water and its focus progressed from coarse large-scale sweeps to fine small-scale probing while moving towards the low water mark and following the falling tide. Maps contained in the original reports (Maine State Archives) were used for orientation to the specific designated regions within each specific location. These included navigational charts and both topographic and hand-drawn maps. In all instances, hand-drawn maps were used as the final reference to locate target species. They indicated landmarks, and other physiographic features, illustrated sufficiently to find the exact locations reported. Other supporting materials contained within each Figure 1. Map of Cobscook Bay, ME, showing locations of critical invertebrate areas. Inset illustrates general location of Cobscook Bay in relation to the Gulf of Maine. 330 Northeastern Naturalist Vol. 11, Special Issue 2 file, i.e., descriptions, correspondence, and evaluation forms, also aided in this process. Upon completion of a reconnaissance survey, two outcomes were possible based on presence/absence of target species: either (1) the target species was found and its distribution marked for assigning the boundaries of the sample transect, or (2) the target species was not discovered. In outcome (1), a sample transect was located within the boundaries of the critical invertebrate area indicated by its large-scale delineation on topographic maps and navigational charts. In outcome (2), the boundaries of the sample transect were selected according to the distribution of nontarget species found during reconnaissance that often corresponded to significant features of the physical environment. The rationale for following this selection method was based on the prediction that locations qualitatively characterized to have high species diversity would generate the best ecological inventory to compose a baseline update for an area. For example, at Gleason Point, the gaper clam, Mya truncata, was not found within the boundaries of the archived hand drawn map from the Gleason Cove critical area file (Maine State Archives). A transect was therefore located within Gleason Cove in an area adjacent to Gleason Point that formed a natural boundary where the greatest number of different species were found during the reconnaissance survey. Faunal transects. Quadrat sample transects were oriented as a 50-m2 band (50 x 1 m) parallel to the low waterline with a width overlapping the low intertidal-high subtidal fringe exposed at low water. This stratum had the highest probability of coinciding with the greatest densities of target species. Approximately 30 min before the time of low water, the endpoints of the 50-m long transect were anchored with markers, and the distance, divided into thirds, marked in turn. Approximately 20 min before the time of low water, a 1-m distance perpendicular to the transect and extending in the direction of the falling tide was marked at each endpoint and each intra-transect marker. At Crow Neck, Wilbur Neck, and Gleason Point, sampling effort and tide permitted a second sample transect (50 x 1 m), that extended the initial one to give a total sample transect area of 100 m2. It was positioned slightly higher (< 1 m) than the original stratum given the rise in tide. Faunal transect quadrat samples. Thirty replicate 0.1-m2 quadrats, 10 per each third of the 50-m2 transect area, were sampled without replacement. Quadrat selection was not truly random. A quadrat sampler was blindly tossed forward from an end of a subdivision with no intent towards distance or lateral direction provided it fell within the transect boundaries. This action was repeated until the opposite end of the subdivision was reached. Sampling continued if fewer than 10 quadrat samples were taken before reaching the end of the subdivision by fac2004 T.J. Trott 331 ing towards the area just sampled and repeating this procedure. Overlap between quadrats was avoided by marking the position of each sampled quadrat with golf tees. A subsequent selected quadrat that overlapped to any degree with a previously sampled quadrat was discarded and reselected. When a second sample transect was added, 60 total replicate quadrats were sampled. The presence of all macrobenthic epifaunal invertebrates was recorded before disturbing the quadrat by turning over rocks and cobble in search of additional species. Macroalgae were intensively searched for target species. Only target species were quantified by counting the number of individuals within a quadrat using adaptive cluster sampling. Faunal transect adaptive cluster samples. Many sample locations had been evaluated as ecologically significant and registered on the basis of the presence of species like the smooth top shell, Margarites helicinus, and the brachiopod, Terebratulina septentrionalis, that have patchy spatial distributions. An adaptive cluster sampling design (Thompson 1990) was implemented to estimate mean densities for these epifaunal species, since populations that are highly clumped are best sampled by this method. Briefly, it dictates that when a randomly selected quadrat contains a species of interest, additional quadrats are nonrandomly added to the borders of the original quadrat. These “neighborhood quadrats” initially have one side in common with the original quadrat and are examined for the targeted species. Neighborhood quadrats are added until “edge quadrats” (i.e., quadrats that are empty) result. When only edge quadrats are obtained, the sampling for that cluster stops. Adaptive cluster sampling is not effective for infaunal species and is inappropriate for dispersed species. Faunal transect benthic substrate samples. Ten substrate samples were taken from the quadrats along the initial 50-m transect. These quadrats were selected prior to sampling by randomly choosing quadrat numbers (1–30) from a table of random values (Zar 1996). All substrate was removed with a scoop within the selected 0.1-m2 quadrat until anoxic sediments were reached (≈ 2–6 cm) or a volume of 2.3 x 103 cm3 was collected, whichever came first. Substrate samples were processed in entirety on the same day of collection. They were sieved through 2.0 and 1.0 mm screens sequentially in the laboratory. Material retained on the screens was sorted with the aide of 50x dissecting microscopes. All organisms were identified to the lowest taxon practical, most often the species level. Invertebrates were transferred into fingerbowls after separating any predators (i.e., crabs), relaxed with a 7% filtered seawater solution of MgCl 2 for 24 h in refrigeration, then fixed with 10% formalin. Fixed samples were labeled with location, sample and quadrat number, and date before archiving. Archived samples were transferred into 70% ethanol for permanent storage after 6–8 weeks. 332 Northeastern Naturalist Vol. 11, Special Issue 2 Faunal transect sample data analysis. Species richness was estimated using the nonparametric jackknife estimate at each location for a simple measure of diversity (Heltshe and Forrester 1983). This estimate is based on the observed frequency of rare species in the community. Data from quadrats and substrate samples were used to calculate separate species richness estimates. As already described, mean densities for target species with clumped distributions were estimated using adaptive cluster sampling statistics (Thompson 1990). Whenever possible, current baseline information was compared with qualitative historical reference data collected during previous ecological assessments. Habitat characterization Habitat mapping transects. Sample sites were characterized to habitat type using semiquantitative methods described by Brown (1993) for a classification system of marine and estuarine habitats in Maine as applied to benthic habitats. Since examples of each habitat type defined by this classification system included some of the current sample sites, the utility and accuracy of this system could be tested. It also created a new baseline for future reference. Habitat mapping included more intertidal area than what was sampled for fauna since the entire vertical slope of the intertidal zone was surveyed along a line transect positioned perpendicular, not parallel, to the shore. The width of the intertidal zones exposed at low tide ranged from 35 to 80 m with tidal variation that ranged from -0.46 to 2.4 m around mean low water. All five locations were mapped using a procedure modified from Bailey et al. (1993). A habitat mapping transect was established starting at a georeferenced permanent marker at Spring High Water (SHW) mark that extended to the low water mark bisecting perpendicularly the horizontal faunal sample transect. The compass heading for the transect was recorded from the georeferenced marker to be used for roughly reestablishing the transect line in future mapping studies. The slope of the intertidal zone along the transect line was determined by measuring the change in elevation at 10-m intervals. Habitat mapping transect quadrat samples. A 1-m2 quadrat was sampled down slope at each 10-m interval by dividing the quadrat and collecting data from 0.25-m2 subsamples. The percentage cover of every macroalgal species that occupied at least 5% of a subquadrat was estimated by visual assessment. The substrate was exposed and a visual scan used to estimate the percentage area comprised of each sediment type, i.e., mud, sand, gravel, cobble, boulder, and rock, as defined by Brown (1993). Macroinvertebrates within the 1-m2 quadrat were noted. At West Quoddy Head, Wilbur Neck, and Crow Neck, each 0.25-m2 subquadrat was digitally photographed twice: (1) after estimating percent macroalgal cover and (2) after estimating percent2004 T.J. Trott 333 age of each substrate type. The purpose of photo documentation was to provide a quantitative reference for each mapped 1-m2 quadrat for future comparisons. Results Habitat characterization Habitat classification agreed in part with descriptions given in the intertidal subsystem of Brown (1993) that listed some of the Cobscook Bay critical invertebrate areas as typical for defined habitats (Table 2). West Quoddy Head substrate was classified as rock and differed from the published predominant substrate type of boulder, but energy level, i.e., partially exposed, was the same. This result indicates that the characterization of substrate type is highly dependent on the positioning of a habitat mapping line transect along the SHW horizon. Gleason Point (mixed-coarse: semi-protected) and Wilbur Neck (mixed-coarse and fine: protected) agreed with published descriptions for these locations. Outer Birch Island and Crow Neck were not directly comparable to published classifications since neither was listed as example sites for any of the intertidal habitat classifications in Brown (1993). However, the substrate types of both sample sites were within the possibilities for their categorized level of exposure. This agreement confirms the utility of the habitat classification system since only certain substrate types can occur for each level of exposure (Brown 1993). The level of accumulated fine mud sedimentation that occluded normal interstices among boulders and large cobble at the Inner Bay complex critical invertebrate area sample sites was substantial. Extensive patches of macroalgae, specifically Laminaria beds, described in correspondence and illustrated on maps in archived files were no longer present at these locations. The surfaces of rock, shell, and algae, were coated with a veneer of fine mud that strongly adhered and resisted removal by washing. In some sample locations, this condition was extreme, particularly Wilbur Neck, producing anoxic conditions just millimeters below the surface. Table 2. Habitat classification of the five critical invertebrate area sample sites in the Cobscook Bay region based on the Brown (1993) habitat classification system of benthic intertidal habitats in marine ecosystems. WQH = West Quoddy head, WN = Wilbur Neck, CN = Crow Neck, OBI = Outer Birch Island, and GP = Gleason Point. Sample Substrate (%) site Rock Boulder Cobble Gravel Sand Mud Brown classification WQH 61 16 11 3 10 0.4 Rock: partially-exposed WN 0.25 28 20 35 2 15 Mixed-coarse + fine: protected CN 1 4 4 51 0 40 Mixed-coarse + fine: semi-protected OBI 43 17 21 15 0.31 4 Mixed-coarse: semi-protected GP 0 35 23 24 18 0 Mixed-coarse: semi-protected 334 Northeastern Naturalist Vol. 11, Special Issue 2 Extensive beds of mussels were found at both Crow Neck and Wilbur Neck. One large mussel bed at Crow Neck extended along the sides to the point of the North by Northeast rock spit bordering a tide pool. This land form was once exposed, bare, current-swept rock and boulder according to Weiss (1980). Mud was not included in the geological description of the rock spit at Crow Neck. Layers of mussel shells underlie extensive mussel beds at Wilbur Neck, representing previous surface levels. At both locations, mussel beds were coated with a veneer of fine mud. Mussels were not reported to occur at Wilbur Neck by the original critical area evaluation. They are listed only in the revised species list for Crow Neck (Weiss 1980) and not the original lists from the field evaluation forms completed in 1968, 1970, and 1975–1977 (Maine State Archives 2001). The appearance of mussel beds could be relatively recent since the extensive detailed description of Crow Neck in the critical area report includes a specific focus on the North by Northeast rock spit fauna and never mentions mussel beds. Mussel beds were unlikely to be routinely overlooked. This idea is supported by the inclusion of Mytilus edulis in the description of the West Quoddy Head critical invertebrate area (Doggett et al. 1978, Maine State Archives 2001). Species composition and community structure Bryozoa, Porifera, and Ascidiacea, taxa that contain species that are epifaunal suspension feeders, consistently formed a minor proportion of representative phyla across all sample sites or were sometimes absent (Fig. 2). Cnidaria, epifaunal sit-and-wait predators, were also a minor segment of the community at all locations. Molluscs dominated in number of species present (37 to 51%) at all sample sites, except Outer Birch Island where samples contained nearly equal proportions of molluscs, polychaetes, and echinoderms comprising 24, 28, and 24% of the macroinvertebrate fauna, respectively (Fig. 2). Species area curves suggest that further sampling would not raise considerably the number of species at Crow Neck, Gleason Point, and Wilbur Neck, while at Outer Birch Island and West Quoddy Head, 30 replicate quadrat samples were insufficient to collect all species (Fig. 3). Differences in jackknife estimates of species richness (Fig. 4a,b) paralleled closely the number of species found in quadrat and substrate samples at each site (Fig. 3), although species historically present were under-represented in samples from all sites (Figs. 5–7). For example, Crow Neck had the highest species richness, but notably only 33% (13 out of 31) of species previously identified as ecologically significant to this sample site were found (Fig. 7). Gleason Point, West Quoddy Head, Outer Birch Island, and Wilbur Neck formed a group with nearly equal estimates of species richness. 2004 T.J. Trott 335 Mussels were clearly dominant at Crow and Wilbur Necks and occurred in 60 and 95% of quadrats, respectively; they were present in 23–47% of quadrats sampled at the remaining sites. Mussels and the gastropod Littorina littorea were the only species ubiquitous in occurrence in all transects at all sample sites. Distribution of target species The sea stars Solaster endeca and Crossaster papposus were not found at any sample site, though they were historically present intertid- Figure 2. Community composition by percentage contribution of major phyla represented at each critical invertebrate area sample site in the Cobscook Bay region. 336 Northeastern Naturalist Vol. 11, Special Issue 2 ally at Crow Neck, as were all targeted species (Fig. 7). In total, just 6 of the 12 target species reported to occur at Crow Neck were observed. Only a single Neptune whelk, Neptunea lyrata decemcostata, was discovered during reconnaissance at Outer Birch Island, the only one of the three target species reported for this site found in this study. No gaper clams, Mya truncata, were found at Gleason Point, registered as a critical invertebrate area based on the intertidal distribution of this species (Table 3, Fig. 5a). Gleason Point reconnaissance found Mya arenaria that had mis-shaped, deformed shells formed during growth in the coarse cobble/gravel/sand substrate (Table 2). Their initial appearance was similar enough to be confused with M. truncata and casts doubt on the report that this species occurred at Gleason Point (see Discussion). A gaper clam was found live under rocks at Outer Birch Island and in only one substrate sample taken from Crow Neck (Table 1). When shells of this species were counted in substrate samples, the same distribution among sample sites was obtained. Only one sample from Outer Birch Island and three from Crow Neck contained shells of M. truncata. This distribution was consistent with historical records (Figs. 6a, 7). Brachiopods, Terebratulina septentrionalis, were found live infrequently at two of the five sample sites during post quadrat sampling Figure 3. Species-area curves for the five critical invertebrate area sample sites in the Cobscook Bay, ME, region. 2004 T.J. Trott 337 Figure 4. Species richness of (a) quadrat samples and (b) substrate samples from the critical invertebrate area sample sites in the Cobscook Bay, ME, region. Vertical bars indicate 95% confidence intervals. 338 Northeastern Naturalist Vol. 11, Special Issue 2 Figure 5. Persistence measured in years of target (*) and nontarget species identified in previous evaluations and the present study of critical invertebrate areas located outside Cobscook Bay. (a) Gleason Point and (b) West Quoddy Head. a. b. 2004 T.J. Trott 339 Figure 6. Persistence measured in years of target (*) and nontarget species identified in previous evaluations and the present study of critical invertebrate areas located within Cobscook Bay. (a) Outer Birch Island and (b) Wilbur Neck. a. b. 340 Northeastern Naturalist Vol. 11, Special Issue 2 reconnaissance surveys only: Crow Neck and Outer Birch Island (Table 1). A shell of this species was found in one sediment sample taken from West Quoddy Head. The distribution of T. septentrionalis at West Quoddy Head and Outer Birch Island was not documented previously (Figs. 5b and 6a, respectively). Quantitative population densities were not estimated since this species did not occur within quadrats at any sample location. They were discovered at Crow Neck and Outer Birch Island only with intense, intentional directed searching when time permitted, i.e., at the end of the field season when quadrat sampling at all sites had been completed. Terebratulina septentrionalis was found on the undersides of only two out of eight large boulders at Crow Neck; one boulder had 11 lampshells attached to its underside and another 16. Figure 7. Persistence measured in years of target (*) and nontarget species identified in previous evaluations and the present study of the Crow Neck critical invertebrate area, located within Cobscook Bay, ME. 2004 T.J. Trott 341 This patchy distribution was similar to historical documentation on the occurrence of this species in the Crow Neck Critical Invertebrate Area (Table 3). On Outer Birch Island, only a single T. septentrionalis was found attached to the underside of a boulder. It was photographed and marked for future reference. Five target species were found within sample transects at various locations, but their abundance was not quantified using adaptive cluster sampling (Table 2). Three were infaunal bivalves found in sediment samples, i.e., the suspension feeding Musculus discors, M. niger, and Astarte complex, and the fourth was the gastropod Lacuna vincta, a grazer that did not have a clumped distribution. To summarize, eight of the 12 target species selected for study were found among the five Critical Invertebrate Area sample sites. However, only one of these, Margarites helicinus, was found in quadrat samples with distributions appropriate for estimating population density using adaptive cluster sampling. Margarites helicinus Abundance of Margarites helicinus differed markedly from documented qualitative estimates by Gilbert (1977a) that scored smooth top shells as “abundant” at Wilbur Neck (Table 3). Differences in densities among these sites were most apparent when the total number of snails counted using adaptive cluster sampling was compared. Within 50-m2 sample transects, totals ranged from 2 at Wilbur Neck to 320 at West Quoddy Head. Mean density among sites broke into two groups according to broad geographic location: (1) greatest outside Cobscook Bay—West Quoddy Head; and (2) lowest inner Cobscook Bay—Outer Birch Island, Wilbur Neck, and Crow Neck (Table 4). The occurrence of this species within sample sites Table 3. Locations and the relative abundance of macroinvertebrate species at the five Critical Invertebrate Area sample sites in the Cobscook Bay region that registration of each area was based on (Maine State Archives 2001). OBI = Outer Birch Island, WN = Wilbur Neck, GP = Gleason Point, WQH = West Quoddy Head, CN = Crow Neck. Sample Faunal area Species Common name affinity Relative abundance OBI1 Margarites helicinus Smooth top shell Epifauna Common (10–25 snails found at low tide) WN1 Margarites helicinus Smooth top shell Epifauna Abundant (> 25 snails found at low tide) GP2 Mya truncata ? Gaper clam Infauna Large, healthy population, but questionable ID WQH3 High species Species complex Epi- and No data diversity area infauna CN4 Terebratulina Lampshell Epifauna Rare to moderately septentrionalis common, patchy* 1Gilbert 1977a, 2Gilbert 1977b, 3Doggett et al. 1978, 4Weiss 1980. *One isolated patch with high density of 100/m2 (Weiss 1980). 342 Northeastern Naturalist Vol. 11, Special Issue 2 followed patterns documented in historical reference data with the exception of Gleason Point where it was absent, and W. Quoddy Head where it was found (Figs. 5–7). Among sample sites, M. helicinus was present in all transects except at Gleason Point (Table 1). Smooth top shells were found associated most frequently with the macroalgal fucoid, Fucus evanescens, a species common to quiet harbors though occasionally found on more exposed points, near or below low tide (Taylor 1976). At West Quoddy Head, M. helicinus was associated with a more diverse assortment of algae that included the red, Porphyra miniata. Discussion Faunal diversity of the critical invertebrate areas in the Cobscook Bay region, specifically those in the Inner Bay complex, was remarkably poor compared to the rich, diverse communities expected from historical baselines generated by the CAP. This was most apparent at Crow Neck, which had the best historical record (1968–1980) of all study locations for comparison with quadrat samples. Fewer than half of the documented ecologically notable species were found in quadrat and substrate samples from Crow Neck in the present study. Change over this time scale is notable and has not been observed in neighboring marine communities. No change was seen among three subtidal stations sampled in 1974 and 2000 (Wildish and Pohle in press). These St. Croix estuary stations, located in neighboring Passamaquoddy Bay, were isolated from anthropogenic influences. The absence of target species selected for study at sample locations where these species were distributed historically is evident also from qualitative comparisons of densities. Changes in abundance of at least one target species is reflected by the diminished presence of the smooth top shell, Margarites helicinus, now a minor part of the intertidal communities of the Inner Bay complex critical invertebrate area sample sites. The first reports describing the density of this species were based Table 4. Mean densities (mean ± 95% confidence interval) per quadrat (0.1 m2) of smooth top shells, Margarites helicinus, determined by adaptive cluster sampling in 50 m2 transects at the five critical invertebrate area sample sites in the Cobscook Bay region. N = total number of snails counted within each transect. Location N Mean density/0.1 m2 Outer Bay complex West Quoddy Head 320 0.61 ± 2.77 x 10-1 Gleason Point 0 0.0 Inner Bay complex Outer Birch Island 17 0.08 ± 7.99 x 10-4 Crow Neck 6 0.02 ± 2.34 x 10-2 Wilbur Neck 2 0.01 ± 2.03 x 10-3 2004 T.J. Trott 343 on a qualitative measurement relating effort to abundance, i.e., the number of individuals found at low tide on a walk through the intertidal zone (Gilbert 1977a). Based on this relative measure of abundance, Wilbur Neck had an “abundant” population of snails (> 25 snails found at low tide) that were also “common” (10–25 snails found at low tide) on Outer Birch Island (Table 3). Twenty-five years later, the Wilbur Neck populations have dwindled to rare and infrequent occurrences requiring intensive sampling to find any snails, e.g., the two snails found during the present study. Differences in abundance from the historical record are underscored by the nearly equal effort in search times for both field studies. In the present study, smooth top shells were most abundant at West Quoddy Head, which suggests seasonality in reproduction cannot explain the observed low population densities within Cobscook Bay. Insufficient sampling is a doubtful cause for low estimates: although species area curves indicate Outer Birch Island was undersampled, West Quoddy Head was also undersampled. A similar interpretation of abundance of the brachiopod Terebratulina septentrionalis, another target species, is problematic. Since they were absent in quadrat samples, population densities were not estimated with adaptive cluster sampling. Brachiopods were found during intensive searches directed at finding this species at Crow Neck post-faunal transect quadrat sampling. The patchy distribution of T. septentrionalis matches descriptions written as long as 26 years previous (Speel 1978, Weiss 1980) and resembled the clumped spatial pattern that characterize brachiopod distributions in general (Noble et al. 1976). Any reasonable population estimates of this species would require disruptive sampling that could damage this habitat extensively, a concern already raised by Weiss (1980). For this reason, adaptive cluster sampling was not executed after discovering brachiopods. The gaper clam, Mya truncata, was not found at Gleason Point, which Gilbert (1977b) registered as a critical invertebrate area because of the occurrence of this species. Our second survey completed one year after this present study and directed at the exact area marked in her hand-drawn, archived map did not find any M. truncata again. In a letter, Gilbert (Maine State Archives, File 371) raised the possibility that she could have mistaken M. arenaria for M. truncata during the field evaluation that the registration of Gleason Point rested on (Table 3). Her doubts of the initial identification of this species were expressed in this letter to Hank Tyler, Senior Planner, after finding “lots of truncated Mya arenaria” at Gleason Point in August 1977. Environmentally induced resemblance in shell form between these two species can result in similarities that require examination of the pallial sinus and hinge tooth as differentiating characters (Foster 1946). The results of our two surveys combined with Gilbert’s questionable idenification cast reasonable doubt on the presence of M. truncata at Gleason Point. 344 Northeastern Naturalist Vol. 11, Special Issue 2 Gaps in the historical record of faunal distributions within critical areas are less likely the result of inadequate sampling effort than absence of species from these sample sites. Sampling effort can be defined by the amount of time invested in search of species at a location. Most historical critical area evaluations were conducted during a single visit during one low tide. This is equitable to the initial sample site reconnaisance employed in this study. If the number of observers searching is considered in addition to reconnaisance time, then sampling effort was greater in the present investigation, i.e., three observers versus two at most in critical area evaluations. The added blocks of time (40–50 min) represented by quadrat sampling of transects during this re-evaluation accents its greater sampling effort. Species area curves convincingly illustrate adequate sampling at most locations, making sampling error an unlikely explanation for species absence (Fig. 3). Intentional searches outside of sample transects did produce positive records for only two species, Neptunea lyrata decemcostata and Terebratulina septentrionalis, pointing out that the chance effects of sample transect overlap with natural distributions of particular species can result in false scores of species absence. Reconnaissance prior to sampling for positioning sample transects intended to reduce this possibility. Species distributions can be influenced by natural fluctuations in abundance, both seasonally and inter-annually, and were given consideration a priori to sampling. Pronounced seasonal variation is known to occur in subtidal habitats in Maine, with the highest number of species found during the summer months (Ojeda and Dearborn 1989). The possibility that changes in distributions could result from seasonal reproduction and recruitment of target species was minimized by sampling protocols. To maintain congruency with reference to time, locations were sampled in the same season and within 30 days from the month when each critical area had been evaluated. The presence of target species at some locations and absence at others sampled within weeks of each other supports the idea that seasonal reproduction and recruitment was not a contributing factor to the observed changes in species distributions. To reduce the chance that species presence was the result of a random set of larvae, recently settled animals were excluded from quadrat observations by including only individuals easily seen, i.e., distinguishable to the naked eye. This increased the chances that only established populations were recorded and also reduced the chance that density estimates would be inflated. Inter-annual variations in occurrence could be argued as an explanation for the absence of target and other species, since this study was conducted over only one summer season. Some account for this can be made indirectly by considering longevity of some of the target species since only adult, established populations were surveyed. Gaps in the presence 2004 T.J. Trott 345 of species among previous critical area evaluations could address the likelihood that inter-annual variability accounts for their absence in this study. Among target species not found in quadrat samples, longevity is unknown, but might be estimated from congeners and related species. Musculus niger could have a life span similar to the 2–5 years for M. discors (Tyler-Walters 2001) and, likewise, Mya truncata to Mya arenaria with a life span of 10–20 years (Tyler-Walters 2003). Among the Buccinidae (C. stimpsoni, N. lyrata decimcostata), Neptunea antiqua is known to live up to 10 years and considerably older in colder waters (Pearce and Thorson 1967). Longevity for species of sun stars or the brachiopod T. septentrionalis could not be found, although Asterias rubens can live 7–8 years (Budd 2001) and the brachiopod Neocrania anomala has a lifespan of 5–10 years (Jackson 2000). Many of these target species were recorded during all previous critical area evaluations made at intervals of 2 to 10 years for up to 12 years. This persistence suggests that interannual variability during these years was not a major factor influencing the presence of target species. Most species historically present were not found during the present investigation 22 years after their last recorded appearance. Did they actually disappear or is this a result of inter-annual variability? To help answer this, found target species need to be considered, and the two rediscovered in the present study have longevities of 2–5 years for Musculus discors (Tyler- Walters 2001) and < 1 year for Lacuna vincta (Jackson 2003). Both species were recorded by evaluations 22 years earlier, and L. vincta twice by earlier evaluations made 4 years apart. While inter-annual variability can influence species occurrence, it does not seem to account for species absences in this study. If it were a major factor determining presence or absence of persistent fauna, then the rediscovery of these two species with such short life spans would be less likely. Other components of the sampling design also reduced possible confounding influences on species density estimates from movement of individuals within sample transects. Many of the species selected for study are motile which could potentially result in inaccurate estimates of species densities if data collected on different days were combined. This error was reduced by sampling transects synoptically within one low tide interval, which also matched the time frame used for relative abundance estimates in historical evaluations (Gilbert 1977a). Without doubt, explanations for species absence based in natural variation of seasonal abundance and interannual variation cannot be eliminated totally in lieu of annual, long-term seasonal monitoring lasting for years (Davidson 1934, Lacalli 1981). In spite of this limitation, surveys similar to this investigation can be valuable in assessing changes in species distributions (Cohen et al. 1998). What has happened in the last 20 years that caused the decline in diversity of species historically present intertidally and the faunal shift 346 Northeastern Naturalist Vol. 11, Special Issue 2 observed in this study? The possible effects of disturbance resulting from siltation of suspended sediments which could have produced the coatings of fine mud found covering surfaces of algae and rock are worth exploring. Changes in substrate characteristics resulting from siltation can have a direct effect on community composition through recruitment. Invertebrate larvae that actively select habitats for settlement may not prefer surfaces modified by sediment deposition (Butman and Grassle 1992, Crisp 1976, Pawlik 1992, Woodin et al. 1998) and will not recruit, while larvae of other species that do prefer this habitat will settle and replace the former. Once established, new colonizing species can enhance sedimentation through bioresuspension and biodeposition (Graf and Rosenberg 1997, Rhoads and Young 1970) and further inhibit colonization of previously established fauna. Increased suspended sediments can also decrease the survivorship of sessile filter-feeding fauna including ascidians, bryozoans, and sponges (Bakus 1968, Naranjo et al. 1996, Turk and Risk 1981, Zühlke and Reise 1994). These phyla were minor components of the intertidal communities surveyed by this study. Patterns of sedimentation in Cobscook Bay must have been altered over time if siltation is a reasonable explanation for the observed change in faunal composition. Present day community composition is characterized by the absence of species historically distributed intertidally and the predominance of mussels. Mussel beds cause significant siltation and modification of substrates in two ways. They increase sedimentation of suspended solids through hydrodynamic effects that result from turbulence created in the benthic boundary layer by the physically rough form of these aggregates (Butman et al. 1994, Frénchette et al. 1989). In addition, sediment composition can change from biodeposition of feces and pseudofeces by mussels and modify diversity through benthic macrofaunal succession (Dittmann 1990, Mattsson and Lindén 1983). Mussels were not reported at Wilbur Neck and for Crow Neck earlier than 1980, but were found by this survey to be ubiquitous and often the dominant species among all sample sites in Cobscook Bay. Greater abundance of mussels could have enhanced siltation, and recent work on the species composition of mussel beds in Cobscook Bay might help to explain population changes. Mussel beds in Cobscook Bay are of mixed species composition of Mytilus edulis and M. trossulus and their hybrids (Rawson et al. 2001, 2003). Frequencies of M. trossulus in mussel beds can vary from approximately 97% at Gove Point to 46% at West Quoddy Head (Rawson et al. 2001). In Newfoundland, Toro et al. (2002) found this species becomes mature at smaller sizes, has a higher reproductive output, and spawns over a much longer period of time (late spring to early autumn) than M. edulis (2–3 weeks in July). If reproductive capacity of M. trossulus is also high in Cobscook Bay, the combined effects of increased recruitment events facilitated by removal 2004 T.J. Trott 347 of grazers known to decrease mussel recruitment, e.g., Littorina littorea (Bertness et al. 1999, Petraitis 1987), might explain the current patterns of greater distribution and abundance of mussel beds. Massive recruitment of Mytilus edulis has occurred across a 125-km-long coastline in the southwest Gulf of Maine in 1995 (Witman et al. 2003). Recruitment on this scale has not been documented in the northern Gulf of Maine, but since the 1995 event, increased mussel recruitment has been reported at two offshore sites in the Gulf of Maine (Harris et al. 1998) and during the summer of 2000 in Long Island Sound, Narragansett Bay, and Isles of Shoals, Gulf of Maine (see Witman et al. 2003). Mussels beds might be considered unexpectedly persistent in spite of possible heavy predation by the invasive European green crab, Carcinus maenas. Green crabs were found by the present study to be very abundant in Cobscook Bay, and mussels are commonly eaten by this species (Bertness et al. 1999). Mussel clumps can also increase habitat availability for recruiting green crabs, with known impacts on community trophic structure and faunal diversity (Thiel and Dernedde 1994). A positive feedback between recruitment of mussels and green crabs would eventually limit mussel populations through increased predation, quite opposite of what was observed by the present investigation. However, the lower energy yield of M. trossulus that contain less meat (Freeman et al. 1994) could make them less preferable prey for green crabs that maximize energy in their diet by selective predation (Elner and Hughes 1978). If green crabs can discriminate between the two species of mussels based on sensory cues, the current species composition of mussel beds (Rawson et al. 2001) could also be the result of selective predation on more energetically profitable M. edulis and not solely a result of higher recruitment of M. trossulus. In addition, since predation by green crabs on periwinkles is well known (Lubchenco and Menge 1978, Trussell 1996), an increase in this predator’s abundance would indirectly facilitate mussel recruitment by removing these grazers from intertidal communities (Petraitis 1987). Sources of suspended solids in Cobscook Bay are worthy of examination if siltation facilitated by mussel beds has contributed towards changing intertidal communities. Potential sources of input are rivers and streams, erosion due to rising sea level, and sources outside of Cobscook Bay, i.e., Passamaquoddy Bay, Upper Bay of Fundy. The most parsimonious explanation for sources of suspended sediments would not require the examination of inputs of particulates from various sources outside Cobscook Bay, but would rely instead on resuspension from within the Bay. Expansive plumes of sediments stirred up by scallop-dragging vessels are evident in aerial photographs (Fig. 8, see also Brooks et al. 1997). The intensity of dragging for scallops has historically been high because Cobscook Bay was one of the most productive scalloping areas on the Maine 348 Northeastern Naturalist Vol. 11, Special Issue 2 coast. The suspension of large amounts of sediment from urchin dragging in autumn attenuates light enough to make Cobscook Bay light limited in respect to phytoplankton production (Phinney et al. 2004). Dredges used to harvest sea urchins undoubtedly create similar plumes. The degree of resuspension of sediments from dragging for scallops and sea urchins would be cumulative and vary according to fishing intensity, although the duration and overlap of open season for each fishery would contribute additional combined affects. Licensing regulations control the number of boats dragging for each species and are more stringent on urchin than scallop draggers. New licenses for urchin dragging have been issued by lottery since 1998, but the licensing of scallopers is unrestricted. Unlimited licensing has resulted in startling numbers of scallop boats appearing on opening day, with more than 200 scallopers working the Bay for the first two weeks of the season in 1989, and 178 boats appearing on opening day in 2001 (Department of Marine Resources, pers. comm.). Fewer boats were seen since 2001 due to changes in regulations that have made dragging less profitable for large boats working the Bay from outside the region. Currently, nearly 100 people in towns neighboring Cobscook Bay are licensed to drag or dive for scallops. While fewer boats harvest sea urchins than scallops because of regulated licensing, the duration of open season for urchin dragging was much longer prior to 2004 and overlapped the shorter sixmonth scallop season. When this fishery first started in earnest during 1987, urchin dragging had no closed season, i.e., harvesting was year round. The first season restriction was imposed in 1993, closing the fish- Figure 8. Opening day of scallop season 1 November 1995 in South Bay of Cobscook Bay. Plumes of sediment suspended from dragging activity are evident. Photograph courtesy of Bruce Kidman, The Nature Conservancy, Maine Chapter. 2004 T.J. Trott 349 ery for nearly three months and then approximately five months in 2003. Currently, the open season for 2004–05 is only 45 days. Fishing effort for urchins is difficult to reconstruct because of how catch data were collected, but during the 2002–03 season, of the 48 draggers licensed, 37 fished the Bay at least 10 times. Increased amounts of suspended sediments with redistribution dependent on circulation patterns in the Bay (Brooks 2004, Kelley and Kelley 2004) could result in the silt deposition observed at all of the Inner Bay complex critical invertebrate area sample sites. The decrease in predator populations, rather than the increase observed by Witman et al. (2003) following massive recruitment of mussels, could be the difference between sedimentation patterns within a bay complex versus an exposed open coast where transport away from the shore can take place. Sediments are more likely to remain within the Bay because of long residence times (Brooks 2004) and be trapped by mussel beds. In addition, the adhering veneer of fine sediment found coating all surfaces at the Inner Bay locations could be created by biostabilization of the resuspended silt by microphytobenthos (Friend et al. 2003, Stal 2003). Elevated standing stocks of microphytobenthos found within the Inner Bay (Phinney et al. 2004), where residence time increases (Brooks 2004), would enhance this process there. Removal of herbivorous snails by predation and the commercial fishery for Littorina littorea would eliminate the growth-limiting effects of grazing on microphytobenthos and associated biofilms. Over time, veneers of fine sediment like those found during this study would coat surfaces normally cleared by grazing snails. The influence of sedimentation of solids suspended through commercial dragging activity on community diversity in Cobscook Bay deserves further study. Inhibition of recruitment of historically important species on substrates altered by sedimentation in the Cobscook Bay ecosystem must be a concern if the diversity of species historically present is to be conserved. The paradigm suggested as an explanation for the observed faunal shift is based on the cumulative effects of hydrodynamically facilitated sedimentation by M. trossulus/M. edulis mussel beds of “recent” appearance and trapping of suspended solids on biofilms. Predation by green crabs may have indirectly accelerated these processes in two ways: (1) by reducing the abundance of blue mussels M. edulis and (2) reducing the abundance of grazing gastropods. Commercial harvest of periwinkles, L. littorea, would contibute towards the latter. Considerable changes in seascapes would be the overall result with loss of habitat for species found in the previously unique intertidal communities of Cobscook Bay and potentially for some currently harvested commercial species like scallops. The destabilizing effects of siltation and altered species recruitment could explain at least some of 350 Northeastern Naturalist Vol. 11, Special Issue 2 the significant changes in community structure observed over the past 20 years in Cobscook Bay. Acknowledgments This study was supported through a contract with The Nature Conservancy, Maine Chapter. I would like to thank Barbara Vickery, Director of Conservation Programs, for her support and discussions leading to the refinement of this project and manuscript. Suggestions on earlier drafts by John Sowles, Department of Marine Resources, and two anonymous reviewers improved this manuscript. The assistance of the Department of Marine Resources in providing sea urchin and scallop fishery information is gratefully acknowledged. Bruce Kidman, The Nature Conservancy, Maine Chapter, kindly provided the photograph of scallop fishing vessels. Appreciation and gratitude to Dr. Paul Langer for his voluntary participation on all levels is recognized. Andrey Avanesov assisted with quadrat sampling and substrate sample processing. Literature Cited Bailey, A., K. Ward, and T. Manning. 1993. A field guide for characterizing habitats using “A Marine and Estuarine Habitat Classification System for Washington State.” Washington Department of Natural Resources, Division of Aquatic Lands, Olympia, WA. 11 pp. Bakus, G.J. 1968. Sedimentation and benthic invertebrates of Fanning Island, Central Pacific. Marine Geology 6:45–51. Bamber, R.N. 1993. Changes in the infauna of a sandy beach. Journal of Marine Biology and Ecology 172:93–107. Bertness, M.D., G.H. Leonard, J.M. Levine, P.R. Schmidt, and A.O. Ingraham. 1999. Testing the relative contribution of positive and negative interactions in rocky intertidal communities. Ecology 80:2711–2726. Bousfield, E.L., and M.L.H. Thomas. 1975. Postglacial changes in distribution of littoral marine invertebrates in the Canadian Atlantic region. Proceedings of the Nova Scotia Institute of Science 27(Supplement 3):47–60. Brooks, D.A. 2004. Modeling tidal circulation and exchange in Cobscook Bay, Maine. Northeastern Naturalist 11(Special Issue 2):23–50. Brooks, D.A., M.W. Baca, and Y.-T. Lo. 1997. Tidal mixing, dispersion, and flushing in Cobscook Bay, Maine: A hydrodynamic model study. Texas A&M University, College Station, TX. Technical Report 97-2-T (December), 77 pp. Brown, B. 1993. A classification system of marine and estuarine habitats in Maine: An ecosystem approach to habitats. Part 1: Benthic habitats. Maine Natural Areas Program, Department of Economic and Community Development, Augusta, ME. 51 pp. + 1 appendix. Buchanan, J.B. 1963. The bottom fauna communities and their sediment relationships off the coast of Northumberland. Oikos 14:154–175. Budd, G.C. 2001. Asterias rubens. Common starfish. In Marine Life Information Network: Biology and Sensitivity Key Information Sub-programme [on-line]. Plymouth: Marine Biological Association of the United Kingdom. [accessed 16 September 2004]. Available at: http://www.marlin.ac.uk/species/ Asteriasrubens.htm. 2004 T.J. Trott 351 Butman, C.A., and J.P. Grassle. 1992. Active habitat selection by Capitella sp. 1 larvae. I. Two choice experiments in still water and flume flows. Journal of Marine Research 50:669–715. Butman, C.A., M. Frénchette, and W. R. Geyer. 1994. Flume experiments on the food supply to the blue mussel Mytilus edulis L. as a function of boundarylayer flow. Limnology and Oceanography 39:1735–1768. Campbell, D.E. 2004. Evaluation and emergy analysis of the Cobscook Bay ecosystem. Northeastern Naturalist 11(Special Issue 2):355–424. Cohen, A., C. Mills, H. Berry, M. Wonham, B. Bingham, B. Bookheim, J. Carlton, J. Chapman, J. Cordell, L. Harris, T. Klinger, A. Kohn, C. Lambert, G. Lambert, K. Li, D. Secord, and J. Toft. 1998. Puget Sound expedition: A rapid assessment survey of non-indigenous species in shallow waters of Puget Sound. Washington State Department of Natural Resources, Olympia, WA. 37 pp. Crisp, D.J. 1976. Settlement responses in marine organisms. Pp. 83–123, In Newell, R.C. (Ed.). Adaptation to Environment: Essays on the Physiology of Marine Animals. Butterworths, London, UK. Davidson, V.M. 1934. Fluctuations in the abundance of planktonic diatoms in the Passamaquoddy region, New Brunswick from 1924 to 1931. Contributions to Canadian Biology and Fisheries (new series) 8:359–407. Dittmann, S. 1990. Mussel beds: Amensalism or amelioration for intertidal fauna? Helgoländer Meeresuntersuchungen 44:335–352. Doggett, L.F., P.F. Larsen, and S.C. Sykes. 1978. Intertidal bedrock areas of high species diversity in Maine, and their relevance to the Critical Areas Program. Critical Areas Planning Report 55:1–106 + 1 appendix. Elner, R.W., and R.N. Hughes 1978. Energy maximization in the diet of the shore crab, Carcinus maenas. Journal of Animal Ecology 47:103–116. Ford, E. 1923. Animal communities of the level sea-bottom in the water adjacent to Plymouth. Journal of the Marine Biological Association of the United Kingdom 13:164–224. Foster, R.W. 1946. The genus Mya in the western Atlantic. Johnsonia 2:29–35. Freeman, K.R., K.L. Perry, T.G. DiBacco, and D.J. Scarratt. 1994. Observations on two Mytilid species from a Nova Scotian mussel farm. Canadian Technical Report of Fisheries and Aquatic Sciences, No. 1969. Department Fisheries and Oceans, Halifax, NS, Canada. Frénchette, M., C.A. Butman, and R.W. Geyer. 1989. The importance of boundary- layer flows in supplying phytoplankton to the benthic suspension feeder, Mytilus edulis. Limnology and Oceanography 34:19–36. Friend, P.L., M.B. Collins, and P.M. Holligan. 2003. Day–night variation of intertidal flat sediment properties in relation to sediment stability. Estuarine, Coastal, and Shelf Science 58:663–675. Gilbert, M.A. 1977a. The smooth top shell (Margarites helicinus [sic]) in Maine and its relevance to the Critical Areas Program of the State Planning Office. Maine Critical Areas Program Planning Report 51:1–15. Gilbert, M.A. 1977b. The gaper clam (Mya truncata) in Maine and its relevance to the Critical Areas Program of the State Planning Office. Maine Critical Areas Program Planning Report 29:1–16. 352 Northeastern Naturalist Vol. 11, Special Issue 2 Gleizon, P., A.G. Punt, and M.G. Lyons. 2003. Modeling hydrodynamics and sediment flux within a macrotidal estuary: Problems and solutions. The Science of the Total Environment 314–316:589–597. Graf, G., and R. Rosenberg. 1997. Bioresuspension and biodeposition: A review. Journal of Marine Systems 11:269–278. Harris, L.G., M. Tyrell, and C.M. Chester. 1998. Changing patterns for two sea stars in the Gulf of Maine 1976–1996. Pp. 234–248, In R. Moori and M. Telford (Eds.). Echinoderms: San Fransisco. A.A. Balkema, Rotterdam, The Netherlands. Heltshe, J.F., and N.E. Forrester. 1983. Estimating diversity using quadrat sampling. Biometrics 39:1073–1076. Jackson, A. 2000. Neocrania anamola. A brachiopod. Marine Life Information Network: Biology and Sensitivity Key Information Sub-programme [online]. Plymouth: Marine Biological Association of the United Kingdom. Accessed 16 September 2004. Available at: http://www.marlin.ac.uk/species/ Neocraniaanomala.htm. Jackson, A. 2003. Lacuna vincta. Banded chink shell. Marine Life Information Network: Biology and Sensitivity Key Information Sub-programme [online]. Plymouth: Marine Biological Association of the United Kingdom. Accessed 16 September 2004. Available at: http://www.marlin.ac.uk/species/ Lacunavincta.htm. Kelley, J.T., and A.R. Kelley. 2004. Controls on surficial materials distribution in a rock-framed, glaciated, tidally dominated estuary: Cobscook Bay, Maine. Northeastern Naturalist 11(Special Issue 2):51–74. Kühne, S., and E. Rachnor. 1996. The macrofauna of a stony sand area in German Bight (North Sea). Helgoländer Meersuntersuchengen 50:433–452. Lacalli, T. 1981. Annual spawning cycles and planktonic larvae of benthic invertebrates from Passamaquoddy Bay, New Brunswick. Canadian Journal of Zoology 59:433–440. Larsen, P.F., and E.S. Gilfillan. 2004. Preliminary survey of the subtidal macrobenthic invertebrates of Cobscook Bay, Maine. Northeastern Naturalist 11(Special Issue 2):243–260. Lubchenco, J., and B.A. Menge. 1978. Community development and persistence in a low rocky intertidal zone. Ecological Monographs 48:67–94. Maine State Archives. 2001. Critical Areas Program of the Maine State Planning Office. Maine State Planning Office, Augusta, ME. Accessed 2001. Mattsson, J., and O. Lindén. 1983. Benthic macrofauna succession under mussels, Mytilus edulis L. (Bivalvia), cultured on hanging long-lines. Sarsia 68:97–102. Naranjo, S.A., J.L. Carballo, and J.C. Garcia-Gómez. 1996. Effects of environmental stress on ascidian populations in Algeciras Bay (southern Spain). Possible marine bioindicators? Marine Ecology Progress Series 144:119–131. Noble, J.A.P., A. Logan, and G.R. Webb. 1976. The recent Terebratulina community in the rocky subtidal zone of the Bay of Fundy, Canada. Lethaia 9:1–17. Ojeda, F.P., and J.H. Dearborn. 1989. Community structure of macroinvertebrates inhabiting the rocky subtidal zone in the Gulf of Maine: Seasonal and bathymetric distribution. Marine Ecology Progress Series 57:147–161. Pawlik, J.R. 1992. Chemical ecology of settlement of benthic marine invertebrates. Oceanography and Marine Biology Annual Review 30:273–335. 2004 T.J. Trott 353 Pearce, J.B., and G. Thorson. 1967. The feeding and reproductive biology of the red whelk, Neptunea antiqua (L). (Gastropoda, Prosobranchia) Ophelia 4:277–314. Petraitis, P.S. 1987. Factors organizing rocky intertidal communities of New England: Herbivory and predation in sheltered bays. Journal of Experimental Marine Biology and Ecology 109:117–136. Phinney, D.A., C.S. Yentsch, and D.I. Phinney. 2004. Primary productivity of phytoplankton and subtidal microphytobenthos in Cobscook Bay, Maine. Northeastern Naturalist 11(Special Issue 2):101–122. Rawson, P.D., S. Hayhurst, and B. Vanscoyoc. 2001. Species composition of blue mussel populations in the Northeastern Gulf of Maine. Journal of Shellfish Research 20:31–38. Rawson, P.D., C. Slaughter, and P.O. Yund. 2003. Patterns of gamete incompatibility between blue mussels Mytilus edulis and M. trossulus. Marine Biology 143:317–325. Rhoads, D.C., and D.K. Young. 1970. The influence of deposit-feeding organisms on sediment stability and community trophic structure. Journal of Marine Research 28:150–178. Shillabeer, N., and J.F. Tapp. 1989. Improvements in the benthic fauna of the Tees estuary after a period of reduced pollution loadings. Marine Pollution Bulletin 20:119–123. Speel, J. 1978. The brachiopod, Terebratulina septentrionalis in Maine and its relevance to the Critical Areas Program of the State Planning Office. Maine Critical Areas Program Planning Report 71:1–15. Stal, L.J. 2003. Microphyobenthos, their extracellular polymetric substances, and the morphogenisis of intertidal sediments. Geomicrobiology Journal 20:463–479. Taylor, W.R. 1976. Marine Algae of the Northeastern Coast of North America. 6th Edition. University of Michigan Press, Ann Arbor, MI. 509 pp. Thiel, M., and T. Dernedde. 1994. Recruitment of shore crabs Carcinus maenas on tidal flats: Mussel clumps as an important refuge for juveniles. Helgoländer Meeresuntersuchungen 48:321–332. Thompson, S.K. 1990. Adaptive cluster sampling. Journal of the American Statistical Association 85:1050–1059. Toro, J.E., R.J. Thompson, and D.J. Innes. 2002. Reproductive isolation and reproductive output in two sympatric mussel species (Mytilus edulis, M. trossulus) and their hybrids from Newfoundland. Marine Biology 141:897–909. Trott, T.J. 2004. Cobscook Bay inventory: A historical checklist of marine invertebrates spanning 162 years. Northeastern Naturalist 11(Special Issue 2):261–324. Trott, T.J., and P.F. Larsen. 2003. Cobscook Bay, Maine: The crown jewel of regional biodiversity in the Gulf of Maine. In Proceedings of the 32nd Annual Benthic Ecology Meeting, Groton, CT. Trussell, G. 1996. Phenotypic plasticity in an intertidal snail: The role of a common crab predator. Evolution 50:448–454. Turk, T.R., and M.J. Risk. 1981. Effect of sedimentation on infaunal invertebrate populations of Cobequid Bay, Bay of Fundy. Canadian Journal of Fisheries and Aquatic Sciences 38:642– 648. 354 Northeastern Naturalist Vol. 11, Special Issue 2 Tyler, P.A., and S.E. Shackley. 1980. The benthic ecology of linear sandbanks: A modified Spisula subcommunity. Pp. 539–559, In M.B. Collins (Ed.). Industrial Embayments and their Environmental Problems: A Case Study of Swansea Bay. Pergamon Press, Oxford, UK. Tyler-Walters, H. 2001. Musculus discors. Green crenella. Marine Life Information Network: Biology and Sensitivity Key Information Sub-programme [on-line]. Plymouth: Marine Biological Association of the United Kingdom. Accessed 16 September 2004. Available at: http://www.marlin.ac.uk/species/ Musculusdiscors.htm. Tyler-Walters, H. 2003. Mya arenaria. Sand gaper. Marine Life Information Network: Biology and Sensitivity Key Information Sub-programme [online]. Plymouth: Marine Biological Association of the United Kingdom. Accessed 16 September 2004. Available at: http://www.marlin.ac.uk/species/ Myaarenaria.htm. Weiss, S. 1980. Report on the conservation of the marine invertebrate area at Crow Neck and vicinity, Trescott, Maine, including an updated inventory and species list. Maine Critical Areas Program of the State Planning Office Miscellaneous Report 45:1–20 + 1 appendix. Wildish, D.J., and G. Pohle. In press. Benthic macrofaunal changes in the Letang Inlet: A Quarter-century view of an industrialized area. In Hargrave, B.T. (Ed.). Environmental Effects of Marine Finfish Aquaculture. The Handbook of Environmental Chemistry. Vol. 5. Water Pollution. Springer Verlag, Heidelberg, Germany. Witman, J.D., S.J. Genovese, J.F. Bruno, J.W. McLaughlin, and B.I. Pavlin. 2003. Massive prey recruitment and the control of rocky subtidal communities on large spatial scales. Ecological Monographs 73:441–462. Woodin, S.A., R.L. Marinelli, and D.E. Lincoln. 1998. Process-specific cues for recruitment in sedimentary environments: Geochemical signals? Journal of Marine Research 56:535–558. Zar, J.H. 1996. Biostatistical Analysis. Prentice Hall, Upper Saddle River, NJ. Zühlke, R., and K. Reise. 1994. Response of macrofauna to drifting tidal sediments. Helgoländer Meeresuntersuchungen 48:277–289.