Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 45 2016 NORTHEASTERN NATURALIST 23(1):45–66 Relationship Between Epibenthic Invertebrate Species Assemblages and Environmental Variables in Boston Harbor’s Intertidal Habitat Elizabeth N. Eddy1,2,* and Charles T. Roman3 Abstract - The Boston Harbor Islands National Recreation Area has an extensive intertidal zone, with 47% of the area composed of mixed-coarse substrate. Given anticipated climatechange impacts such as sea-level rise and ocean warming, and other stressors associated with the urban environment, the critical ecosystem functions (i.e., species habitat, food-web support) provided by this dominant mixed-coarse habitat of Boston Harbor, and elsewhere throughout the Northeast, have been and will likely be further altered. To evaluate the present-day epibenthic invertebrate communities and to determine what environmental factors of the mixed-coarse substrate affect community structure, we used a stratified random design to sample epibenthic macroinvertebrates along with various physical and environmental variables from the intertidal zone. Epibenthic macroinvertebrate species assemblages and diversity differed significantly between wave-exposed and wave-protected sites, with higher diversity present at protected sites. We also found that environmental variables collectively explained up to 67% of the variation in species assemblages, with elevation, organic content, water content, and sediment type individually explaining up to 56%, 30%, 42%, and 33% of the variation, respectively. This study provides a baseline for long-term monitoring aimed at understanding the response of cobble and mixed-coarse intertidal communities to multiple disturbances, and a foundation to support habitat restoration or other management actions. Introduction Studies characterizing macroinvertebrate and other species assemblages of rocky intertidal habitats in the New England region and elsewhere are extensive, including several reviews and classic papers (e.g., Bertness 1999, Connell 1972, Cruz-Motta et al. 2010, Dayton 1971, Lubchenco 1980, Menge 1976, Murray et al. 2006). In addition, the assessment of macroinvertebrate assemblages in intertidal habitats of boulder, cobble, gravel, and mixed-coarse substrates have received attention in some regions such as Puget Sound, WA (Dethier and Schoch 2005, Schoch and Dethier 1996), and the Gulf of Maine, Cobscook Bay, ME (Trott 2004), yet an extensive literature is lacking. These unconsolidated habitat types are major components of many estuarine shores throughout the coastal northeastern US (e.g., Gulf of Maine [Foulis and Tiner 1994, Kelley and Kelley 2004], Buzzards Bay, MA [Hough 1940], and Narragansett Bay, RI [Boothroyd and August 2008, Schwartz 2009]). Within the Boston Harbor Islands National Recreation Area, the focus of 1Graduate School of Oceanography, University of Rhode Island, Narragansett, RI 02882. 2Current address - ORISE Research Participation Program, US Environmental Protection Agency, Washington, DC 20460. 3National Park Service, University of Rhode Island, Narragansett, RI 02882. *Corresponding author - eeddy@my.uri.edu. Manuscript Editor: Thomas Trott Northeastern Naturalist 46 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 this study, the dominant intertidal shoreline type is a mixed-coarse substrate, defined as cobbles, gravel, shell, and sand, each not exceeding 75% cover, and rock or boulder, each not exceeding 50% cover (Bell et al. 2005). However, characterization of this habitat from a faunal perspective is limited aside from qualitative species inventories (Bell et al. 2005, Matassa 2009). It is important to understand and document the structure of the biotic communities in the mixed-coarse intertidal zone because they are changing in response to sea-level rise and other climate-change factors and are threatened by shoreline modification, contaminant spills, and other anthropogenic factors. Factors driving species diversity or other characteristics of community structure are commonly studied in terms of climate change, temporal or spatial heterogeneity, or biological interactions such as competition and predation (Bertness and Leonard 1997, Menge 1976, Silva et al. 2006, Stachowicz 2001), though the role of physical or environmental factors in influencing marine intertidal biological community structure has also been emphasized (Covich et al. 2004, Dethier and Schoch 2005, Dethier et al. 2010, McQuaid and Branch 1984, Oak 1984, Schoch and Dethier 1996). The purpose of this study was to provide a quantitative inventory of epifaunal macroinvertebrate assemblages of Boston Harbor’s mixed-coarse intertidal habitats and understand the environmental variables that best explain the variation in the epifaunal assemblages, species richness, diversity, and density on wave-exposed and wave-protected shores. Field-Site Description The Boston Harbor Islands National Recreation Area consists of 34 islands and peninsulas nested within Boston Harbor at approximately 42°19'6.96"N, 70°50'44.88"W (Fig. 1). This area is unique as the only drowned drumlin field in the United States (Himmelstoss et al. 2006), formed by a series of glacial till deposits. Additionally, several of the outer islands on the edge of the Boston Basin are composed primarily of exposed bedrock (FitzGerald et al. 2011, Rosen and Leach 1987). Total land area of the park ranges between 6 km2 to 12.4 km2 from high tide to low tide, respectively, with semi-diurnal tides and a mean tidal range of approximately 2.9 m (Bell et al. 2005). The park area includes 56 km of shoreline, 30% of which are lined with coastal structures such as seawalls, and 40% with glacial bluffs (FitzGerald et al. 2011). The intertidal zone is composed of a range of habitats including rocky, gravel, and sand beaches, mud flats, and salt marshes, with unconsolidated mixed-coarse habitat dominating (Bell et al. 2005). Methods Site selection We initially selected for this study 8 islands (Bumpkin, Georges, Grape, Little Brewster, Lovells, Peddocks, Spectacle, and Thompson), of the 34 in the Boston Harbor Islands National Recreation Area, based on public-ferry availability to facilitate access. Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 47 On each of the 8 islands, we identified all mixed-coarse habitat from a GIS-based inventory of intertidal habitats (Bell et al. 2004). We randomly selected 18 sites for establishment of transects (Fig. 1). We categorized the mixed-coarse shoreline of each island as wave-exposed or wave-protected based on a maximum wave-energy model (i.e., storm events with winds from the northeast and northwest; FitzGerald Figure 1. Map of Boston Harbor with sampling sites identified by wave exposure. Northeastern Naturalist 48 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 et al. 2011, Hughes et al. 2007). Nine of the randomly located sites were waveexposed and 9 were wave-protected. Due to this random selection, only 6 islands (Bumpkin, Grape, Lovells, Peddocks, Spectacle, and Thompson) were represented in our inventory. When field sampling at each selected site, we ran a transect tape perpendicular to the shoreline from approximately mean lower low water (MLLW) to the upland boundary at mean higher high water (MHHW), and placed five 0.25-m2 quadrats along each transect using generalized random tessellation stratified (GRTS) sampling, which randomly selects quadrat locations while ensuring that each quadrat is an independent sample and that the entire elevation gradient along the transect is included in sampling (Steven and Olsen 2004). This method was performed by estimating the horizontal distance between the upper and lower boundaries of the mixed-coarse intertidal zone according to the data obtained from Bell et al. (2004), dividing that distance into 5 equal segments, and randomly selecting 1 quadrat location from each segment. Some sites had fewer than 5 quadrats because the inventory maps (Bell et al. 2004) overestimated the actual extent of the intertidal zone or sampling was conducted during neap tides when lower-elevation quadrats were submerged. Only 2 transects (both on exposed shorelines) were under-sampled, resulting in a total of 87 independent random quadrats (42 on exposed shores, 45 on protected shores). Macroinvertebrate sampling We collected all macroinvertebrate data during July and August 2012. We identified and quantified all epifaunal macroinvertebrate species, defined as species 1 mm in length or greater, by overturning all rocks within each quadrat and quickly collecting mobile species by hand. For colonizing sessile species such as tunicates and bryozoans, each colony was counted as 1 individual. Barnacle density was often too high to record individual counts in the field, so we took aerial, plan-view photos of each quadrat, i.e., photoplots, and estimated barnacle counts by counting individuals in each photoplot in the lab. Barnaclequantification methods likely resulted in an underestimate of total barnacle density since the photoplots are only a 2-dimensional representation of the quadrat and barnacles may have been present in contours unseen in the photos. When the surface layer of rock and cobble was cleared, we took a single shallow surface sample (10 cm diameter, 5 cm depth) from the center of each quadrat to collect smaller epifaunal organisms which were otherwise too difficult to detect or collect by hand. We rinsed the contents of the surface sample through a 1-mm– mesh sieve and extracted all amphipods and other epifaunal organisms. These species were preserved in 10% buffered formalin for storage, and transferred to 70% ethanol in the lab prior to identification. We omitted any infaunal species collected in the shallow surface sample from the analysis of species assemblages. Species identification was aided by various field guides and taxonomic keys (Bousfield 1973, Gosner 1978, Weiss 1995). We identified all individuals to the lowest taxon Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 49 possible, in most cases species, and scaled-up count data from the quadrats and cores to 1 m2 for data analysis. Environmental variables Latitude and longitude. We recorded latitude and longitude of each quadrat with a Garmin GPSMAP unit to assess how epifaunal assemblages varied over the geographical range of the harbor. Date. We included date of sampling as a variable in the environmental data analysis due to the likelihood that epibenthic assemblages could vary throughout the 2-month data collection period. Elevation, slope, and aspect. We recorded elevation as height in meters relative to MLLW for each quadrat using a Trimble Real Time Kinematic GPS on 18–20 September 2012. We determined the elevation for each corner of each quadrat and then averaged all values to obtain a quadrat elevation and used the elevations of the upper corners and lower corners of each quadrat to determine the slope of each quadrat. We recorded aspect, or the compass direction the shoreline slope faces, for each transect. Aspect is not directly related to wave exposure in Boston Harbor due to complex harbor morphology, as salients and sheltering effects of nearby islands influence shoreline exposure (FitzGerald et al. 2011). Thus, we treated aspect as an independent variable in our data collection. Rugosity. We determined rugosity, representing surface complexity of a substrate, prior to macrofaunal sampling using a method after Luckhurst and Luckhurst (1978). We laid a string over all the contours of both quadrat diagonals and recorded the average length (cm) divided by 71 cm, the length of a completely flat diagonal surface for a 0.25-m2 quadrat. A higher rugosity corresponds to an area composed primarily of boulder substrate, while a lower rugosity corresponds to quadrats containing mostly gravel. Algae cover. For the purposes of this study, algae cover was considered an environmental variable because the species assemblages are limited to epifaunal organisms, and there is evidence that algae may act as a predictor variable in the structure of faunal species assemblages (Bertness and Leonard 1997, Bolam et al. 2000, Hay 1981, Leonard 2000, Menge 1978, Stachowicz 2001). We used the 100-point–intercept method, adapted from Hutchinson and Williams (2003), to determine percent cover of upper-story, dominant red, green, and brown macro-algae using photoplots, or aerial, plan-view images of each quadrat, by projecting a grid containing 100 evenly distributed points onto each photoplot. Algae were identified beneath each point-intercept, where each point-intercept represents 1% cover. We estimated cover by algal group (red, green, and brown) and total algal coverage since identification to the species level was limited in many photoplots. We omitted encrusting and epiphytic algae from the analysis due to identification restraints associated with photoplots. Water and organic content. We took a sediment sample (approximately 100 cm3) from the top 5 cm of each quadrat. In the lab, we thoroughly mixed the collected sediment samples by hand with a spatula and removed 2 cm3. These samples were Northeastern Naturalist 50 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 weighed (initial weight), then placed in a muffle furnace at 100 °C for 24 hours to determine the dry weight (DW100) and again at 550 °C for 24 hours to determine organic content (i.e., loss-on-ignition method; DW550). We calculated percent of water and total organic content in each sample after Dean (1974). Sediment skewness. After determining water content and organic content of each sediment sample, we analyzed the dried samples for grain size using a Ro-Tap sieve shaker to sort the sample through a nested series of sieves (4750, 2000, 1700, 850, 425, 250, 160, 75, and 63 μm). We then used the GRADISTAT program (Blott and Pye 2001) to determine the distribution of grain size, in terms of skewness, for each sediment sample, where a negative skew represents sediment defined by finer grain sizes and positive skew represents coarser grain sizes. Data analysis Variability among epibenthic intertidal assemblages. We used routines available in the PRIMER 6 software program (Clarke and Warwick 2001) and the PERMANOVA+ add-on package (Anderson et al. 2008) to perform multivariate analyses of intertidal epibenthic assemblages and univariate analyses of density, species richness, and Shannon-Weiner diversity. For multivariate analysis of species assemblages, we first calculated the Bray- Curtis similarity index on fourth-root transformed species abundances, with a dummy variable of 0.001 added to the dataset where species abundances had a value of zero. The fourth-root transformation was used to de-emphasize large values (i.e., Balanus balanoides [Common Rock Barnacle], amphipod, Anurida maritima [Seashore Springtail]) in a highly skewed dataset. Based on the Bray-Curtis similarity measure, we tested differences in species assemblages between wave-exposed and wave-protected groups with a one-way permutational multivariate analysis of variance (PERMANOVA) and an analysis of the homogeneity of multivariate dispersions (PERMDISP). PERMANOVAs were performed with the wave-energy model as a fixed factor and partitioning was done with Type III sums of squares. We identified differences in species assemblages by calculating a distance-based pseudo-F statistic (ratio of between-group variance to within-group variance) with PERMANOVA and an F statistic with PERMDISP, using 9999 unrestricted permutations of the transformed data. Results are interpreted together to identify the source of dissimilarity between groups, i.e., location effects (variability between groups) and/or dispersion effects (variability within groups). We also constructed an nMDS plot to illustrate the differences in species assemblages between the 2 wave exposures. We then performed the similarity percentages (SIMPER) routine on the same transformed assemblage data to determine the contribution of each species to the average dissimilarity between wave-exposure groups. For univariate analysis, we calculated richness, density (individuals per m2), and diversity (H′) from the original, untransformed data set. We used the Euclidean similarity measure to calculate distances between samples, and performed PERMANOVA and PERMDISP analyses using the techniques described previously. Pielou’s evenness (J) was also calculated using the same methods to determine if differences in diversity were driven by differences in richness or evenness. Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 51 Relationships between epibenthic intertidal assemblages and environmental variables. We analyzed the relationships between environmental variables and epibenthic intertidal species assemblages, density, species richness, and diversity using distance-based linear models (DISTLM), after transforming the environmental variables with ln(x), ln(x+1), 1/x, or sqrt(x) in order to normalize the data, where applicable. A draftsman plot for environmental variables among all sites revealed that no 2 variables were strongly correlated (r ≥ 0.85), thus there was no redundancy of environmental variables. We performed DISTLM analyses on the same Bray-Curtis similarity matrix (for species assemblages) and Euclidean similarity matrix (for species richness, density, and diversity) used for the previously described PERMANOVA and PERMDISP analyses. We used marginal tests to assess the relationships between each environmental variable and the response variable independently, and sequential tests to determine which combinations of environmental variables best explain variability in the response variable. The sequential test was performed with a step-wise selection procedure using Akaike’s information criteria (AIC). We performed each DISTLM analysis twice, once with the total algae variable omitted and once with red, green, and brown algae variables omitted, to prevent redundancy in the models since red, green, and brown algae are components of total algae. The proportion of variation in assemblage characteristics explained by the other variables did not change when using different algae variables. Results Summary of biotic and environmental variables We detected a total of 25 epifaunal species (Table 1) and 11 algal species/genera (Table 2) in the quadrats in this study. A summary of environmental variables measured across all quadrats are included in Table 3. Note that we did not encounter any fringing salt marsh, dominated by Spartina alterniflora Loisel (Saltmarsh Cordgrass), along any of the sampled quadrats, although this habitat type does occur within the mixed-coarse substrate at Boston Harbor Islands (Bell et al. 2005). Variability in species assemblages We found a significant difference in epifaunal assemblages between wave-exposed and wave-protected groups (Table 4). This significant difference is due only to location effects and is not a result of differences in dispersion. An nMDS plot also illustrates the separation of sample sites (Fig. 2). Although there is considerable similarity between quadrats for the 2 wave exposures, the plot shows a small cluster of approximately 10 wave-protected quadrats that are clearly separated from wave-exposed sample sites, which may be contributing to the dissimilarity in epifaunal community structure between wave exposures. Similarity percentages (SIMPER) analysis indicated that 11 of the 25 observed species contribute up to 91% of the variability between exposed and protected shorelines (Table 5). Although all of these species contributed to the variability between exposure groups, only the herbivorous snails L. littorina (P = 0.02) and Northeastern Naturalist 52 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 Table 1. Epifaunal species present in the 87 sampling sites in the intertidal zone of Boston Harbor’s mixed-coarse habitat. The mean and maximum values represent abundance per m2. Frequency indicates the percent of sampling sites in which species were observed. Sites indicates whether the species were found on wave-exposed shores (E), wave-protected shores (P), or both (E/P). Phylum/Class Order Family Genus and Species Mean Max Freq. Sites Arthropoda Collembola Collembola Neanuridae Anurida maritima (Guérin-Méneville) (Seashore Springtail) 39.50 1400.56 5.75 E/P Insecta Unidentified insect larva 1.46 127.32 1.15 E Malacostraca Amphipoda Corophiidae Corophium volutator (Pallas) (Mud Shrimp) 1.46 127.32 1.15 P Gammaridae Gammarus oceanicus Segerstråle (Scud) 1.46 127.32 1.15 P Hyalidae Hyale plumulosa (Stimpson) (Amphipod) 52.69 763.94 13.79 E/P Microdeutopus Microdeutopus gryllotalpa Costa (Tube Builder) 1.46 127.32 1.15 P Unidentified amphipod Order (Latreille) 1.46 127.32 1.15 P Decapoda Paguridae Pagurus acadianus J.E. Benedict (Acadian Hermit Crab) 1.52 36.00 9.20 E/P Pagurus longicarpus Say (Long-Armed Hermit Crab) 1.01 40.00 4.60 E/P Portunidae Carcinus maenas (L.) (European Green Crab) 6.99 64.00 40.23 E/P Varunidae Hemigrapsus sanguineus (De Haan) (Asian Shore Crab) 6.16 104.00 35.63 E/P Maxillopoda Sessilia Archaeobalanidae Balanus balanoides (L.) (Common Rock Barnacle) 263.26 4136.00 48.28 E/P Bryozoa Gymnolaemata Cheilostomata Membraniporidae Membranipora membranacaea (L.) (Lacy Crust Bryozoan) 0.14 4 3.45 E/P Chordata Ascidiacea Pleurogona Botryllidae Botrylloides violaceus Oka (Orange Sheath Tunicate) 0.74 28.00 5.75 E/P Styelidae Styela clava Herdman (Clubbed Tunicate) 0.05 4.00 1.15 E Styela partita (Stimpson) (Rough Sea Squirt) 1.29 108.00 2.30 E Cnidaria Anthozoa Actinaria Diadumenidae Diadumene lineata (Verrill) (Orange-Striped Sea Anemone) 0.51 24.00 3.45 P Mollusca Bivalvia Mytiloida Mytilidae Mytilus edulis L. (Blue Mussel) 1.01 20.00 12.64 E/P Gastropoda Neogastropoda Nassariidae Nassarius obsoletus (Say) (Eastern Mudsnail) 0.09 8.00 1.15 P Nassarius trivittatus (Say) (Threeline Mudsnail) 0.05 4.00 1.15 E Neotaenioglossa Calyptraeidae Crepidula fornicata (L.) (Common Atlantic Slippershell) 1.70 72.00 6.90 E/P Crepidula plana Say (Eastern White Slippersnail) 5.19 108.00 12.64 E/P Littorinidae Littorina littorea (L.) (Common Periwinkle) 111.08 728.00 51.72 E/P Littorina obtusata (L.) (Flat Periwinkle) 0.32 24.00 2.30 P Littorina saxatilis (Olivi) (Rough Periwinkle) 0.78 16.00 11.49 E/P Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 53 Table 3. Summary of environmental variables characterizing the 87 sampled quadrats in the intertidal zone of Boston Harbor’s mixed-coarse habitat, where frequency indicates the percent of sampling sites in which algae were observed. Variable Mean Minimum Maximum Frequency (%) Slope (%) 5.85 -4.84 16.83 Rugosity 1.18 1.01 1.53 Soil water content (%) 13.72 2.09 35.14 Soil organic content (%) 1.48 0.34 4.14 Soil skewness 0.17 -0.29 0.70 Brown algae cover (%) 1.75 0.00 56.00 9.19 Green algae cover (%) 2.80 0.00 59.00 14.94 Red algae cover (%) 1.16 0.00 24.00 21.84 Total algae cover (%) 5.71 0.00 59.00 37.93 Table 2. Species list of algae detected in sampling sites. * denotes a crustose algae which was not included in percent cover data due to difficulty observing the a lgae in each photoplot. Phylum Class Order Family Genus and Species Chlorophyta Ulvophyceae Bryopsidales Codiaceae Codium fragile (Suringer) (Dead Man’s Fingers) Ulvales Ulvaceae Ulva sp. Ochrophyta Phaeophyceae Fucales Fucaceae Fucus distichus L. (Rockweed) Fucus spiralis L. (Spiral Wrack) Fucus vesiculosis L. (Bladder Wrack) Rhodophyta Bangiophyceae Bangiales Bangiaceae Porphyra sp. Florideophycaea Gigartinales Gigatinaceae Chondrus crispus Stackhouse (Chondrus Crispus) Phyllophoraceae Mastocarpus sp. Solieriaceae Agardhiella sp. Gracilariales Gracilariaceae Gracilaria sp. Hildenbrandiales Hildenbrandiaceae Hildenbrandia rubra* (Sommerfelt) Meneghini (Rusty Rock) Table 4. Summary of PERMANOVA (pseudo-F) and PERMDISP (F) analyses between waveexposed and wave-protected sites for epifaunal data, with * indicating significant results at P < 0.05. PERMANOVAs were performed as a one-way analysis. PERMANOVA PERMDISP (df1: 1, df2: 85) (df1: 1, df2: 85) Pseudo-F P-value F P-value Inference Species assemblages 4.42750* 0.0098* 0.67451 0.4277 Location effect only Density 0.19702 0.6773 1.90410 0.3587 No location or dispersion effects Species richness 1.36000 0.2507 1.91900 0.2054 No location or dispersion effects Diversity 6.62720* 0.0119* 0.29144 0.5893 Location effect only Northeastern Naturalist 54 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 L. saxatilis (P = 0.01) had significantly different abundances between exposure groups at P ≤ 0.05; however, note that several other epibenthic species were different at the more liberal P ≤ 0.10 (the barnacle B. balanoides [P = 0.10], the mussel M. edulis [P = 0.09], and the sea anemone D. lineata [P = 0.06]). Figure 2. nMDS plot of the 87 sampling sites, based on the 4th-root transformed Bray-Curtis similarities of species identities and abundances showing separation of wave-exposed and wave-protected sites. Table 5. Similarity percentages (SIMPER) analysis, identifying species contributing to the discrimination in epifaunal assemblages between wave-exposed and wave-protected shorelines. Average abundances presented in individuals per m2. Average dissimilarity between exposure groups was 82.66. Cumulative contribution of all species is 91.19%. * signifies species whose abundances were significant (P ≤ 0.05) between wave exposure groups . Exposed Protected Species average abundance average abundance Contribution % Littorina littorea* 69.2 150.1 22.0 Balanus balanoides 355.0 177.7 21.7 Carcinus maenas 5.7 8.2 9.9 Hemigrapsus sanguineus 4.7 7.6 9.9 Hyale plumulosa 51.5 53.8 8.8 Anurida maritima 30.3 48.1 6.9 Littorina saxatilis* 0.1 1.4 3.1 Crepidula fornicata 2.4 1.1 3.1 Mytilus edulis 1.5 0.5 2.2 Diadumene lineata 0.0 1.0 1.8 Pagarus acadianus 1.8 1.2 1.6 Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 55 Variability in density, species richness, and diversity We found no significant differences in density and species richness between exposed and protected shorelines (Table 4). However, diversity differed significantly between exposed and protected groups due to differences in location effects only (Table 4), where diversity was higher in protected sites (Fig. 3). To evaluate the source of the significant diversity result, we performed PERMANOVA on J, and determined that evenness is also significantly different between shore exposures (pseudo-F = 11.735, P = 0.0094), with average evenness higher on protected shores (J = 0.6224, n = 31) than on exposed shores (J = 0.45309, n = 20). Relationship between environmental data and species assemblages, density, species richness, and diversity Since PERMANOVA results indicated significant differences in species assemblages and diversity between wave exposure groups, we performed DISTLM analysis separately on these parameters for the exposed and protected categories. DISTLM was performed on the full dataset for density and species richness since no significant differences occurred between wave-exposure groups. Marginal tests indicated that latitude and longitude did not significantly explain variability in species assemblages or any parameters. All other variables contributed significantly to at least 1 parameter (Table 6). Sequential tests showed different combinations of variables in the best-fit models for each parameter and for each wave-exposure group among species assemblages and diversity. In some cases, certain variables improved the AIC selection criteria in the model development, but their contributions to the model were not statistically significant, therefore we chose to accept the models only including the significant variables (Table 7). Figure 3. Average diversity, species richness, and density of epifaunal species in waveexposed and wave-protected shores. Error bars are SE. Average values are per m2. Only differences in diversity between exposure groups were significant based on PERMANOVA analysis. Sample size: exposed n = 42, protected n = 45. Northeastern Naturalist 56 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 Discussion Summary of biotic and environmental variables The species list in this study is smaller than the 95 animalia species recorded in Bell et al. (2005). Of the species recorded by Bell et al. (2005), approximately 70 of those species are epifaunal; however, it cannot be determined on which islands and types of habitats each species was observed. Our study notably observed no echinoderms, hydroids, sponges, and other common intertidal organisms, possibly as a result of a less-exhaustive sampling design and smaller sampling area (Table 1). Bell et al. (2005) surveyed the entire intertidal shore of 21 islands, including 13 substrate types, not just the single mixed-coarse substrate. Species assemblages, density, species richness, diversity, and wave energy Wave exposure proved to be significantly correlated with epifaunal assemblages in the mixed-coarse intertidal zone of Boston Harbor (Table 4). A similar study on epifaunal species in Puget Sound, WA, including benthic invertebrates and algae in the overall assemblage structure, found that epifaunal assemblages varied by wave exposure as determined by wave fetch and mean maximum wind velocity, similar to our definition of wave exposure (Dethier and Schoch 2000). Diversity was also significantly different between the 2 wave-exposure groups, though species richness was not (Table 4). The higher diversity in wave-protected sites is perhaps best explained by evenness in the abundance of species. These findings are consistent with those of similar studies of benthic invertebrate communities on rocky intertidal shores, where evenness accounts for differences between trends in richness and diversity (Scrosati and Heaven 2007, Scrosati et al. 2011). The lower evenness Table 6. Summary of marginal tests, obtained from distance-based linear models (DISTLM), for epifaunal data. Values displayed indicate the proportion of variability explained by each variable. * indicates values significant at P < 0.05, and +/- indicate positive or negative correlations. Aln(x), Bln(x+1), C1/x transformed, Dsqrt(x) transformed. Protected = wave protected sites, exposed = waveesposed sites. Species assemblages Species Diversity Variables Protected Exposed Density richness Protected Exposed Latitude 0.0256 0.0441 0.0047 - 0.0006 - 0.0013 - 0.0084 - Longitude 0.0280 0.0113 0.0006 + 0.0050 - 0.0072 - 0.0212 - Date 0.0183 0.0531 0.0007 + 0.0334 + 0.0907* + 0.0346 + Slope 0.0431 0.0888* 0.0451* - 0.0438 - 0.0014 - 0.0297 - Aspect 0.0169 0.0882* 0.0735* + 0.0269 + 0.0145 - 0.0673 + Elevation 0.3070* 0.1830* 0.2490* - 0.4460* - 0.5640* - 0.2630* - Rugosity 0.0663* 0.1510* 0.0451C* - 0.0591C* - 0.0769 + 0.0769 + Water content 0.3070* 0.2860* 0.1620* + 0.4110* + 0.3080* + 0.4280* + Organic content 0.1830* 0.1700A* 0.0681D* + 0.262D* + 0.3010* + 0.2490A* + Skewness 0.0968* 0.2060B* 0.3310* + 0.2480* + 0.0870* + 0.0830B* + Brown algae 0.0162 0.0752* 0.0215 + 0.0676* + 0.0204* + 0.2350* + Green algae 0.0686* 0.0363 0.0108 - 0.0313 - 0.0746 - 0.0385 - Red algae 0.0272 0.1490* 0.1090* + 0.2270* + 0.0002 + 0.2250* + Total algae 0.0451 0.1420* 0.0101 + 0.0233 + 0.0263 - 0.1940* + Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 57 on wave-exposed mixed-coarse intertidal shores from this study is likely related to higher abiotic stress and dominance of more stress-tolerant species (e.g., barnacle). Barnacle and mussel abundance were statistically higher on wave-exposed sites. This observation is consistent with Bertness et al. (2006), who reported greater average barnacle density on exposed shores. These observations are also supported by studies demonstrating that filter feeders have a higher abundance on exposed shores since water movement enhances food supply and alleviates thermal stress (Bustamante and Branch 1996, Dethier and Schoch 2005, Hammond and Griffiths 2004, Harley and Helmuth 2003). Additionally, Bustamante and Branch (1996) determined that invertebrate predators are more abundant on exposed shores. The results of this study contradict those findings, with the crabs H. sanguineus and C. maenas Table 7. Summary of sequential tests, obtained from distance-based linear models (DISTLM), for epifaunal data. * indicates values significant at P < 0.05. +/-indicate additions to or subtractions from the model. Aln(x+1), B1/x transformed, Csqrt(x) transformed. Epifaunal data Variables Proportion Cumulative Species assemblages Protected +Elevation 0.30684* 0.30684 +Longitude 0.06763* 0.37447 +Rugosity 0.06822* 0.44269 +Water content 0.03093* 0.47362 +Date 0.02367 0.49729 +Green algae 0.02249 0.51978 Exposed +Water content 0.28557* 0.28557 +Rugosity 0.11889* 0.40446 +SkewnessA 0.04464* 0.44910 +Elevation 0.04412* 0.49322 Density +Skewness 0.33115* 0.33115 +Elevation 0.10547* 0.43662 +RugosityB 0.02514 0.46176 +Latitude 0.01948 0.48124 Species richness +Elevation 0.44567* 0.44567 +Water content 0.08573* 0.53140 +RugosityB 0.04031* 0.57171 +Red algae 0.04506* 0.61677 +Skewness 0.02277* 0.63954 +Organic contentC 0.01566 0.65520 -Water content 0.00099 0.65421 +Date 0.01371 0.66792 +Brown algae 0.00798 0.67590 Diversity Protected +Elevation 0.56446* 0.56446 +Rugosity 0.06500* 0.62946 +Latitude 0.04133* 0.67079 +Organic content 0.01632 0.68711 Exposed +Water content 0.42820* 0.42820 +Rugosity 0.04400 0.47220 +Total algae 0.03932 0.51152 Northeastern Naturalist 58 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 more abundant on protected shores, though the suite of predators in Bustamante and Branch (1996) differed from those in this study, and average abundances in this study were not statistically significant. The results of this study correspond more closely with Kitching and Ebling (1967), who found C. maenas abundances were greater on sheltered shores of a marine lough in Ireland. Grazers such as L. littorea and L. saxatilis were more abundant on the wave-protected shores than wave-exposed shores, and all amphipods besides H. plumulosa were found exclusively on wave-protected beaches, consistent with grazer distribution by wave energies observed in other regions (Bertness 1984, Bustamante and Branch 1996). L. littorea density was greater on wave-protected than wave-exposed shores, and barnacle density was lowest on protected shores (Table 5), consistent with Petraitis’ (1983) suggestion that high periwinkle abundance interferes with barnacle settlement. Wave action tends to extend biological zones of sessile species vertically upshore by increasing food supply and duration of immersion (Harley and Helmuth 2003, Murray et al. 2006, Ricciardi and Bourget 1999), while desiccation in protected sites limits vertical ranges (Bertness et al. 2006). Meanwhile, heightened wave energy tends to constrain distribution of mobile organisms (Hammond and Griffiths 2004, Menge and Olson 1990). These trends help to explain why the abundances of mobile species appear higher on wave-protected shores and the abundances of sessile species seem higher on wave-exposed shores, though these results are not all significant. Exposed shores have been found to have lower richness and diversity (Bustamante and Branch 1996, Scrosati and Heaven 2007), consistent with our results in which the wave-exposed shores had lower diversity, though richness did not significantly differ for the taxa detected (Fig. 3). It is likely that richness, and also density, were not significant since major differences between wave exposures were dependent on different species compositions. The wave-exposed shores were devoid of Littorina obstusata, Diadumene lineata, Microdeutopus gryllotalpa, Gammarus oceanicus, Corophium volutator, and an unidentified amphipod, while wave-protected shores were devoid of Styela clava, Styela partita, Nassarius trivittata, and an unidentified insect larva. Together, these results indicate that in mixed-coarse habitats, wave energy primarily affects species composition and species distributions, rather than the absolute number of species and individuals present. However, a study on coarse-sediment shores in Puget Sound found a clear trend of higher epibiotic species richness in areas of greater wave energy (Dethier and Schoch 2005). Also contrary to our results was a similar study that determined disturbance on bedrock shoreline by storm activity increased species richness (Zacharias and Roff 2001), whereas the model for our study showed no difference in species richness between protected and exposed sites. It is also difficult to compare results across different studies where relative exposure groups were used in analysis because the actual energy associated with storm winds may vary by geographic region. In Boston Harbor, significant wave height (average of the highest third of the waves) varies between 0 and 1.5 m, while Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 59 energy density reaches 3 J/m2 (FitzGerald et al. 2011), but without knowing wave height and energy density in other studies, caution must be taken when comparing results between studies. Although diversity was higher in the protected shores than the exposed shores, we cannot definitively say that diversity is negatively correlated with wave exposure because there were only 2 exposure groups defined in this study. Instead, the intermediate-disturbance hypothesis (Connell 1978; Sousa 1979, 1984) or environmental stress model (Menge and Sutherland 1987) could be relevant, in which diversity would be highest at an intermediate wave energy. However, these could not be tested with our data, since establishing a third exposure group was not possible given the methods used to define wave-exposed and wave-protected sites. Future improvements to this study may benefit by including additional waveexposure groups. Epifaunal assemblages and environmental variables Elevation was an important factor defining assemblage characteristics (i.e., explained up to 56% of the variation in assemblage characteristics; Table 6), consistent with similar studies that indicated elevation plays a primary role in regulating biomass, species richness, abundance, and diversity across a variety of benthic intertidal communities (Davidson 2005, Davidson et al. 2004, Scrosati and Heaven 2007, Wallenstein and Neto 2006). A trend of increased diversity, species richness, and abundance at lower intertidal elevations is evident (Figs. 4, 5). The percent of variation explained by elevation was about twice as high in waveprotected sites as in wave-exposed sites for species assemblages and diversity. This result is expected, as desiccation stress on protected shores limits distribution of particular species whereas higher elevations on exposed shores are less affected by this stress due to increased wetting from wave activity. Sediment skewness, organic content, and water content each explained up to 33%, 30%, and 43%, respectively, of the total variation in epifaunal assemblage characteristics (Table 6). Little information is available for the expected effect of sediment grain size on epifaunal communities since it is mostly studied in the context of infaunal communities (Dethier and Schoch 2000, Martin et al. 2005, Rodil et al. 2007, Ysebaert and Herman 2002, Ysebaert et al. 2002). However, the 3 variables are closely related because grain size affects permeability and the storage of water and organic matter (Bergamaschi et al. 1997, Masch and Denny 1966, Szarek-Gwiazda and Sadowska 2010), and both water and organic matter are essential in sustaining benthic intertidal communities (Levinton et al. 1984). Littorina littorea and L. saxatilis abundances (Fig 4) and diversity and richness (Fig. 5) all peaked at a water content of about 15–20%. Up to 23% of the variation in epifaunal assemblages can be explained by algae cover (Table 6), where red algae, brown algae, and total algae are significant in species assemblages and diversity on exposed shores. This finding is surprising as it was expected that algae would serve a larger role on protected shores, alleviating species assemblages from desiccation stress by providing shade and retaining moisture (Hay 1981, Leonard 2000, Menge 1978), though thermal buffering may Northeastern Naturalist 60 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 not be that significant in temperate climates (Bertness and Leonard 1997). Instead, algae may play a role on exposed shores by stabilizing the substrate and buffering mobile species from intense wave energy (Stachowicz 2001). Not surprisingly, red algae were positively associated with species richness and density and brown algae with species richness, indicating that perhaps algae do provide a buffer against stress in the overall community. Uniquely, green algae were significantly correlated with species assemblages on protected shores only. These results may indicate that the role of algae varies by algal group and by each measure of assemblage characteristics, though it is also possible that the algae and epibenthic community may be responding to similar environmental conditions. Future studies would benefit from analyzing algae as a response variable in addition to as a predictor variable, in order to determine how epibenthic assemblages may be indirectly affected by changes in algae throughout the harbor. Figure 4. Littorina littorea and Littorina saxatilis abundances related to elevation and water content. Elevation relative to MLLW. Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 61 Rugosity explained up to only 15% of total variation in species assemblages, density, and richness, and was not significant in explaining diversity. Surprisingly, rugosity had a negative relationship with richness and density. It was expected that a higher rugosity or surface complexity would be associated with more crevices and areas in or under which organisms could rest to escape thermal stress or predation and with a larger surface area for sessile organisms to colonize (Kostylev et al. 2005). However, among the Boston Harbor Island sites, the more complex substrates were at the more thermally stressful higher elevations. Sequential tests indicate that the environmental variables analyzed in this study collectively explain between 44% and 67% of the variation in assemblage characteristics when considering only the significant terms in the models (i.e., cumulative total of significant variables; Table 7). Other variables that may contribute additional variation are the porewater temperature, salinity, and pH, which could not be included in analysis due to insufficient data because porewater could not be collected in the dry surface sediments for many of the higher-elevation plots. It should be noted that it is a complex endeavor to fully understand the role of variations in porewater salinity and tidal water salinity that floods the intertidal, within mixed-coarse sediments of high hydraulic conductivity, in influencing benthic assemblages (Dethier et al. 2010). Additionally, biological effects such as predation or facilitation play important roles in structuring intertidal communities (e.g., Menge 1976, Stachowicz 2001), and the roles of these biological interactions may also contribute to assemblage characteristics. Figure 5. Diversity, species richness, and density related to their 2 most highly correlated environmental variables. Elevation relative to MLLW. Northeastern Naturalist 62 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 The models presented in this study are meant to evaluate the response of epifaunal assemblage characteristics along gradients of environmental variables. These variables may have a direct or indirect effect on the epibenthic intertidal macroinvertebrate communities or may act as surrogates for other physical, chemical, or biological (e.g., predation) factors and gradients. Management implications Cobble or mixed-coarse beaches, such as those found throughout Boston Harbor’s intertidal zones, are confronted with many natural and human-induced disturbances, such as climate change, storms, invasive species, shoreline-protection structures, contaminant spills, nutrient enrichment, and boat wakes. An initial step to protecting intertidal habitats and their essential food-web support functions for foraging shorebirds (e.g., Evans 1988, Hori and Noda 2001) and pelagic communities (e.g., Grossman 1986) is to quantify community structure. The findings of this study provide a baseline for long-term monitoring aimed at understanding the response of cobble and mixed-coarse intertidal assemblages to multiple disturbances, and lay a foundation to support habitat-restoration actions, if warranted. Further, this study highlights the importance of considering exposure to wave energy and other key environmental factors when designing long-term monitoring programs focused on boulder, cobble, or mixed-coarse intertidal shores. Acknowledgments Funding for this research was provided by the National Park Service, with funds administered through the North Atlantic Coast Cooperative Ecosystem Studies Unit at the University of Rhode Island. We thank Penelope Pooler Eisenbies for contributions to the study design; Sarah Waterworth, Annie Kreider, and the Green Ambassadors for their field assistance; Sheldon Pratt and Sebastian Kvist for taxonomic guidance; John King and Danielle Cares for sediment and grain-size analysis assistance; and Candace Oviatt, Carol Thornber, Mary-Jane James, and Marc Albert for their support throughout the study. This study was conducted as partial fulfillment of a Master’s degree program at the University of Rhode Island (Eddy 2013). Literature Cited Anderson, M.J., R.N. Gorley, and K.R. Clarke. 2008. PERMANOVA+ for PRIMER: Guide to Software and Statistical Methods. PRIMER-E Ltd., Plymouth, UK. 214 pp. Bell, R., M. Chandler, R. Buchsbaum, and C. Roman. 2004. Inventory of intertidal marine habitats: Boston Harbor Islands National Park Area. Technical Report NPS/NERBOST/ NRTR-2004/1. USDI National Park Service, Boston, MA. 146 pp. Bell, R., R. Buchsbaum, C. Roman, and M. Chandler. 2005. Inventory of intertidal marine habitats, Boston Harbor Islands National Park Area. Northeastern Naturalist 12 (Special Issue 3):169–200. Bergamaschi, B.A., E. Tsamakis, R.G. Keil, T.I. Eglinton, D.B. Montluçon, and J.I. Hedges. 1997. The effect of grain size and surface area on organic matter, lignin and carbohydrate concentration, and molecular compositions in Peru Margin sediments. Geochimica et Cosmochimica Acta 61:1247–1260. Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 63 Bertness, M.D. 1984. Habitat and community modification by an introduced herbivorous snail. Ecology 65:370–381. Bertness, M.D. 1999. The Ecology of Atlantic Shorelines. Sinauer Associates Inc., Sunderland, MA. 417 pp. Bertness, M.D., and G.H. Leonard. 1997. The role of positive interactions in communities: Lessons from intertidal habitats. Ecology 78:1976–1989. Bertness, M.D., C.M. Crain, B.R. Silliman, M.C. Bazterrica, M.V. Reyna, F. Hildago, and J.K. Farina. 2006. The community structure of western Atlantic Patagonian rocky shores. Ecological Monographs 76:439–460. Blott, S.J., and K. Pye. 2001. GRADISTAT: A grain-size distribution and statistics package for the analysis of unconsolidated sediments. Earth Surface Processes and Landforms 26:1237–1248. Bolam, S.G., T.F. Fernandes, P. Read, and D. Rafaelli. 2000. Effects of macroalgal mats on intertidal sandflats: An experimental study. Journal of Experimental Marine Biology and Ecology 249:123–137. Boothroyd, J.C., and P.V. August. 2008. Geologic and contemporary landscapes of the Narragansett Bay ecosystem. Pp. 1–34, In A. Desbonnet and B.A. Costa-Pierce (Eds.). Science for Ecosystem-Based Management: Narragansett Bay in the 21st Century. Springer Science and Business Media, LLC, New York, NY. 570 pp. Bousfield, E.L. 1973. Shallow-Water Gammaridean Amphipoda of New England. Cornell University Press, Ithaca, NY. 312 pp. Bustamante, R.H., and G.M. Branch. 1996. Large-scale patterns and trophic structure of southern African rocky shores: The roles of geographic variation and wave exposure. Journal of Biogeography 23:339–351. Clarke, K.R., and R.M. Warwick. 2001. Change in Marine Communities: An Approach to Statistical Analysis and Interpretation, 2nd Edition. PRIMER-E Ltd., Plymouth, UK. 250 pp. Connell, J.H. 1972. Community interactions on marine rocky intertidal shores. Annual Review of Ecology and Evolution 3:169–192. Connell, J.H. 1978. Diversity in tropical rain forests and coral reefs. Science 199:1302–1310. Covich, A.P., M.C. Austen, F. Barlocher, E. Chauvet, B.J. Cardinale, C.L. Biles, P. Inchausti, O. Dangles, M. Solan, M.O. Gessner, B. Statzner, and B. Moss. 2004. The role of biodiversity in the functioning of freshwater and marine benthic ecosystems. BioScience 54:767–775. Cruz-Motta, J.J., P. Miloslavich, G. Palomo, K. Iken, B. Konar, G. Pohle, T. Trott, L. Benedetti-Cecchi, C. Herrera, A. Hernandez, A. Sardi, A. Bueno, J. Castillo, E. Klein, E. Guerra-Castro, J. Gobin, D.I. Gomez, R. Riosmena-Rodriguez, A. Mean, G. Bigatti, A. Knowlton, and Y. Shirayama. 2010. Patterns and spatial variation of assemblages associated with intertidal rocky shores: A global perspective. PLoS One 5:e14354. Davidson, I.C. 2005. Structural gradients in an intertidal hard-bottom community: Examining vertical, horizontal, and taxonomic clines in zoobenthic biodiversity. Marine Biology 146:827–839. Davidson, I.C., A.C. Crook, and D.K.A. Barnes. 2004. Quantifying spatial patterns of intertidal biodiversity: Is movement important? Marine Ecology 25:15–34. Dayton, P.K. 1971. Competition, disturbance, and community organization: The provision and subsequent utilization of space in a rocky intertidal community. Ecological Monographs 41:351–388. Dean, W.E., Jr. 1974. Determination of carbonate and organic matter in calcareous sediments and sedimentary rocks by loss-on-ignition: Comparison with other methods. Journal of Sedimentary Petrology 44:242–248. Northeastern Naturalist 64 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 Dethier, M.N., and G.C. Schoch. 2000. The shoreline biota of Puget Sound: Extending spatial and temporal comparisons. Report for the Washington State Department of Natural Resources Nearshore Habitat Program. Washington State Department of Natural Resources, Olympia, WA. 30 pp. Dethier, M.N., and G.C. Schoch. 2005. The consequences of scale: Assessing the distribution of benthic populations in a complex estuarine fjord. Estuarine, Coastal, and Shelf Science 62:253–270. Dethier, M.N., J. Ruesink, H. Berry, A.G. Sprenger, and B. Reeves. 2010. Restricted ranges in physical factors may constitute subtle stressors for estuarine biota. Marine Environmental Research 69:240–247. Eddy, E.N. 2013. Characterizing the benthic invertebrate communities of the mixed-coarse intertidal habitat in Boston Harbor. Master’s Thesis. University of Rhode Island, Kingston, RI. 131 pp. Evans, P.R. 1988. Predation of intertidal fauna by shorebirds in relation to time of the day, tide, and year. Pp. 65–78, In G. Chelazzi and M. Vannini (Eds.). Behavioral Adaptation to Intertidal Life. Plenum Press, New York, NY. 524 pp. FitzGerald, D., Z. Hughes, and P.S. Rosen. 2011. Boat-wake impacts and their role in shoreerosion processes, Boston Harbor Islands National Recreation Area. Technical Report NPS/NERO/NRR-2011/403. USDI National Park Service, Fort Collins, CO. 140 pp. Foulis, D.B., and R.W. Tiner. 1994. Wetland trends for selected areas of the coast of Massachusetts, from Plum Island to Scituate (1977 to 1985–86). Ecological Services Report R5-94/2. US Fish and Wildlife Service, Hadley, MA. 14 pp. Gosner, K.L. 1978. A Field Guide to the Atlantic Seashore from the Bay of Fundy to Cape Hatteras. Houghton-Mifflin, Boston, MA. 329 pp. Grossman, G.D. 1986. Food-resource partitioning in a rocky intertidal fish assemblage. Journal of Zoology 1:317–355. Hammond, W., and C.L. Griffiths. 2004. Influence of wave exposure on South African mussel beds and their associated infaunal communities. Marine Biology 144:547–552. Harley, C.D.G., and B.S.T. Helmuth. 2003. Local- and regional-scale effects of wave exposure, thermal stress, and absolute versus effective shore level on patterns of intertidal zonation. Limnology and Oceanography 48:1498–1508. Hay, M. 1981. The functional morphology of turf-forming seaweeds: Persistence in stressful marine habitats. Ecology 62:739–750. Himmelstoss, E.A., D.M. FitzGerald, P.S. Rosen, and J.R. Allen. 2006. Bluff evolution along coastal drumlins: Boston Harbor Islands, Massachusetts. Journal of Coastal Research 22:1230–1240. Hori, M., and T. Noda. 2001. Spatio-temporal variation of avian foraging in the rocky intertidal food web. Journal of Animal Ecology 70:122–137. Hough, J.L. 1940. Sediments of Buzzards Bay, Massachusetts. Journal of Sedimentary Petrology 10:19–32. Hughes, Z.J., D.M. FitzGerald, N.C. Howes, and P.S. Rosen. 2007. The impact of natural waves and ferry wakes on bluff erosion and beach morphology in Boston Harbor, USA. Journal of Coastal Research (Special Issue 50):497–501. Hutchinson, N., and G.A. Williams. 2003. Disturbance and subsequent recovery of midshore assemblages on seasonal, tropical, rocky shores. Marine Ecology Progress Series 249:25–38. Kelley, J.T., and A.R. Kelley. 2004. Controls on surficial materials distribution in a rockframed, glaciated, tidally dominated estuary: Cobscook Bay, Maine. Northeastern Naturalist 11(Special Issue 2):51–74. Northeastern Naturalist Vol. 23, No. 1 E.N. Eddy and C.T. Roman 2016 65 Kitching, J.A., and F.J. Ebling. 1967. Ecological studies at Lough Ine. Advances in Ecological Research 4:197–291. Kostylev, V.E., J. Erlandsson, M.Y. Ming, and G.A. Williams. 2005. The relative importance of habitat complexity and surface area in assessing biodiversity: Fractal application on rocky shores. Ecological Complexity 2:272–286. Leonard, G.H. 2000. Latitudinal variation in species interactions: A test of the New England rocky intertidal zone. Ecology 81:1015–1030. Levinton, J.S., T.S. Bianchi, and S. Stewart. 1984. What is the role of particulate organic matter in benthic invertebrate nutrition? Bulletin of Marine Science 35:270–282. Lubchenco, J. 1980. Algal zonation in a New England rocky intertidal community: An experimental analysis. Ecology 61:333–344. Luckhurst, B.E., and K. Luckhurst. 1978. Analysis of the influence of substrate variables on coral reef fish communities. Marine Biology 49:317–323. Martin, D., F. Bertasi, M.A. Colangelo, M. de Vries, M. Frost, S.F. Hawkins, E. Macpherson, P.S. Moschella, M.P. Satta, R.C. Thompson, and V.U. Ceccherelli. 2005. Ecological impact of coastal defence structures on sediment and mobile fauna: Evaluating and forecasting consequences of unavoidable modifications of native habitats. Coastal Engineering 52:1027–1051. Masch, F.D., and K.J. Denny. 1966. Grain-size distribution and its effect on the permeability of unconsolidated sands. Water Resources Research 2:665–677. Matassa, C. 2009. Boston Harbor Islands intertidal bioblitz. Final Report. Marine Science Center, Northeastern University, Nahant, MA. 21 pp. McQuaid, C.D., and G.M. Branch. 1984. Influence of sea temperature, substratum, and wave exposure on rocky intertidal communities: An analysis of faunal and floral biomass. Marine Ecology Progress Series 19:145–151. Menge, B.A. 1976. Organization of the New England rocky intertidal community: Role of predation, competition, and environmental heterogeneity. Ecological Monographs 46:355–393. Menge, B.A. 1978. Predation intensity in a rocky intertidal community: Effect of an algal canopy, wave action, and desiccation on predator feeding rates. Oecologia 34:17–35. Menge, B.A., and A.M. Olson. 1990. Role of scale and environmental factors in regulation of community structure. Trends in Ecology and Evolution 5:52–57. Menge, B.A., and J.P. Sutherland. 1987. Community regulation: Variation in disturbance, competition, and predation in relation to environmental stress and recruitment. American Naturalist 130: 730–757. Murray, S.N., R.F. Ambrose, and M.N Dethier. 2006. Monitoring Rocky Shores. University of California Press, Berkeley, CA. 220 pp. Oak, H.L. 1984. The boulder beach: A fundamentally distinct sedimentary assemblage. Annals of the Association of American Geographers 74:71–82. Petraitis, P.S. 1983. Grazing patterns of the periwinkle and their effect on sessile intertidal organisms. Ecology 64:522–533. Ricciardi, A., and E. Bourget. 1999. Global patterns of macroinvertebrate biomass in marine intertidal communities. Marine Ecology Progress Series 185:21–35. Rodil, I.F., M. Lastra, and J. Lopez. 2007. Macroinfauna community structure and biochemical composition of sedimentary organic matter along a gradient of wave exposure in sandy beaches. Hydrobiologia 579:301–316. Rosen, P.S., and K. Leach. 1987. Sediment accumulation forms, Thompson Island, Boston Harbor, MA. Pp. 233–250, In D.M. FitzGerald and P.S. Rosen (Eds.). Glaciated Coasts. Academic Press, San Diego, CA. 364 pp. Northeastern Naturalist 66 E.N. Eddy and C.T. Roman 2016 Vol. 23, No. 1 Schoch, G.C., and M.N. Dethier. 1996. Scaling up: The statistical linkage between organismal abundance and geomorphology on rocky intertidal shorelines. Journal of Experimental Marine Biology and Ecology 201:37–72. Schwartz, M.L. 2009. Estuarine habitats of Narragansett Bay. Pp. 89–106, In K.B. Raposa and M.L. Schwartz (Eds.). An Ecological Profile of the Narragansett Bay National Estuarine Research Reserve. Rhode Island Sea Grant, Narragansett, RI. 176 pp. Scrosati, R., and C. Heaven. 2007. Spatial trends in community richness, diversity, and evenness across rocky intertidal environmental stress gradients in eastern Canada. Marine Ecology Progress Series 342:1–14. Scrosati, R., A.S. Knox, N. Valdivia, and M. Molis. 2011. Species richness and diversity across rocky intertidal elevation gradients in Helgoland: Testing predictions from an environmental-stress model. Helgoland Marine Research 65:91–102. Silva, G., J.L. Costa, P.R. de Almeida, and M.J. Costa. 2006. Structure and dynamics of a benthic invertebrate community in an intertidal area of the Tagus estuary, western Portugal: A six-year data series. Hydrobiologia 555:115–128. Sousa, W.P. 1979. Disturbance in marine intertidal boulder fields: The nonequilibrium maintenance of species diversity. Ecology 60:1225–1239. Sousa, W.P. 1984. Intertidal mosaics: Patch size, propagule availability, and spatially variable patterns of succession. Ecology 65:1918–1935. Stachowicz, J.J. 2001. Mutualism, facilitation, and the structure of ecological communities. BioScience 51:235–246. Steven, D.L., Jr., and A.R. Olsen. 2004. Spatially balanced sampling of natural resources. Journal of the American Statistical Association 99:262–278. Szarek-Gwiazda, E., and I. Sadowska. 2010. Distribution of grain size and organic matter content in sediments of submontane dam reservoir. Environmental Protection Engineering 36:113–124. Trott, T.J. 2004. Late 20th-century qualitative intertidal faunal changes in Cobscook Bay, Maine. Northeastern Naturalist 11(Special Issue 2):325–354. Wallenstein, F.F.M.M., and A.I. Neto. 2006. Intertidal rocky shore biotopes of the Azores: A quantitative approach. Helgoland Marine Research 60:196–206. Weiss, H.M. 1995. Marine animals of southern New England and New York: Identification keys to common nearshore and shallow-water macrofauna. Bulletin 115 of the State Geological and Natural History Survey of Connecticut. Connecticut Department of Environmental Protection, Hartford, CT. 344 pp. Ysebaert, T., and P.M.J. Herman. 2002. Spatial and temporal variation in benthic macrofauna and relationships with environmental variables in an estuarine, intertidal softsediment environment. Marine Ecology Progress Series 224:105–124. Ysebaert, T., P. Meire, P.M.J. Herman, and H. Verbeek H. 2002. Macrobenthic species response surfaces along estuarine gradients: Prediction by logistic regression. Marine Ecology Progress Series 225:79–95. Zacharias, M.A., and J.C. Roff. 2001. Explanations of patterns in intertidal diversity at regional scales. Journal of Biogeography 28:471–483.