Habitat Mapping in Gilbert Bay, Labrador: A Marine Protected Area Phase III Final Report

Submitted by:

Alison Copeland Dr. Evan Edinger Philippe Leblanc Department of Geography Department of Geography Dept.of Geography Memorial University Memorial University St. Memorial University St. John’s, NL A1B 3X9 John’s, NL A1B 3X9 St. John’s, NL A1B 3X9

Dr. Trevor Bell Dr. Rodolphe Devillers Dr. Joseph Wroblewski Department of Geography Department of Geography Ocean Sciences Centre Memorial University Memorial University Memorial University St. John’s, NL A1B 3X9 St. John’s, NL A1B 3X9 St. John’s, NL A1C 5S7

Standing Offer No. F6161-070001/001/XAQ Requisition No. F6161-07 0012 Draft final report, March 7, 2008

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report ii Cover image caption: Shoreline habitats in The Shinneys. Nearshore portions of The Shinneys were not mapped using multibeam sonar, due to shallow water and numerous hazardous shoals. The Shinneys constitutes the primary spawning and juvenile habitat for Gilbert Bay cod, yet little is known about the habitats Gilbert Bay cod use in the Shinneys.

Recommended citation: Copeland, A., Edinger, E., Leblanc, P., Bell, T., Devillers, R., Wroblewski, J., 2008. Marine Habitat Mapping in Gilbert Bay, Labrador – A Marine Protected Area. Phase III final report. Marine habitat mapping group report # 08-01, Memorial University of Newfoundland, St. John’s, 71 p.

Marine Habitat Mapping Group. Trevor Bell, Alison Copeland, Rodolphe Devillers, Philippe Leblanc Department of Geography, Memorial University of Newfoundland, St. John’s, NL A1B 3X9.

Evan Edinger Departments of Geography and Biology, Memorial University of Newfoundland, St. John’s, NL A1B 3X9.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report iii Overview.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report iv Table of Contents. Overview ………………………………………………………………… iii

Table of Contents ………………………………………………………… iv

List of Figures …………………………………………………………….. vi

List of Tables ……………………………………………………………… ix

Research Objectives and Rationale ………………………………………. 1

Research Approach ……………………………………………………….. 5

Field Methods …………………………………………………………….. 6

Laboratory Methods …………………………………………………… 12

Substrate classification …………………………………………… 12

Video substrate and faunal analysis ……………………………… 13

Statistical analysis ………………………………………………………... 14

Interpolation from sounding lines to bathymetric maps ………….. 14

Depth-substrate relationships in areas without multibeam coverage … 15

Faunal analysis ……………………………………………………. 15

Results ……………………………………………………………………. 17

Substrates observed ……………………………………………... 17

The Shinneys …………………………………………………… 26

River Out ………………………………………………………….. 33

Mogashu Tickle …………………………………………………… 37

Potential bedrock wall habitats …………………………………… 42

Habitat Classification …………………………………………….. 43

Defining mappable habitat types ………………………………….. 49

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report v Study Highlights and Future Directions ………………………………… 57

Summary ………………………………………………………………….. 59

References Cited ………………………………………………………….. 60

Appendix A: Bathymetric and substrate profiles in The Shinneys ………. 62

Appendix B: Faunal list …………………………………………………... 67

Appendix C: sediment grain size and organic content ………………….. 71

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report vi List of Figures. Figure 1: Locations of management zones associated with Gilbert Bay Marine Protected Area and place names used in the report.

Figure 2: Benthic substrate map for entire bay.

Figure 3. Substrate map of Zone 1B of the Gilbert Bay Marine Protected Area, including The Shinneys and River Out.

Figure 4. Alison Copeland operating the Garmin 178C GPS-sonar and Shark Marine Drop video systems.

Figure 5. Philippe Leblanc readying the Petit Ponar hand-deployed grab sampler.

Figure 6: Locations of all video and grab samples collected in Zone 1B in 2007 on multibeam backscatter intensity data.

Figure 7: Locations of drop video stations in Leg Island Basin to assess potential bedrock wall habitats.

Figure 8. Location and numbers of video transects collected in River Out, Mogashu Tickle, and The Shinneys.

Figure 9. Video frame grab of muddy gravel substrate, line 16R, Inner Shinneys.

Fig. 10. Sandy gravel substrate. Transect 17, Inner Shinneys.

Fig. 11A. Coralline algal encrusted gravel substrate, transect 16, Inner Shinneys.

Figure 11B . Example of coralline-algae-encrusted-gravel substrate class, as recovered in grab sampler.

Fig. 12A. Gravelly mud substrate.

Figure 12B. Gravelly mud substrate, Transect 25, Inner Shinneys.

Fig 13. Mud substrate. Transect 17, Inner Shinneys.

Fig. 14A. Nearshore gravel substrate, Transect 25, Inner Shinneys.

Fig 14B. Nearshore gravel substrate, Transect 25, Inner Shinneys.

Fig 15A. Sponges, rhodoliths, and sea stars on gravel, Mogashu Tickle Gravel.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report vii Fig 15B. Gravel with sun star and dead scallop shell, Mogashu Tickle gravel.

Fig 16A. Bedrock wall habitat, , site 418, Leg Island Basin.

Fig 16B. Bedrock wall, site 409, Leg Island Basin.

Figure 17. Distribution of substrates along bathymetric and video transect lines that were surveyed in the inner part of The Shinneys, overlad over multibeam backscatter.

Figure 18. Distribution of substrates along bathymetric and video transect lines surveyed in the Inner Shinneys, overlaid over substrate map generated from multibeam backscatter (Copeland et al., 2007a).

Figure 19A: Bathymetric profile along line 25, labeled by substrate type.

Figure 19B. Bathymetric and substrate profile, transect 28, Barr Tickle.

Figure 20. Boxplot showing depth distribution of video transect points in the Shinneys.

Figure 21. Linear and exponential regression lines of depth and substrate code in The Shinneys.

Figure 22: Fledermaus image of substrate classification draped over bathymetry, River Out (Copeland et al., 2007a).

Figure 23. Substrates in River Out, overlaid above multibeam backscatter.

Figure 24 Substrates in River Out, overlaid over substrates interpreted from multibeam sonar (Copeland et al., 2007a).

Figure 25A. Location of frame grab of muddy gravel occurring on steep slopes near mouth of Snook’s Arm, end of transect 7, River Out.

Figure 25B. Bathymetric and substrate profile of transect 7, near mouth of Snook’s Arm, River Out.

Fig. 26A, Soft coral, Gersemia rubiformis, on gravel, Mogashu Tickle gravel.

Fig 26B. Basket star, Gorgonocephalus, Mogashu Tickle.

Figure 27. Substrates in Mogashu Tickle, River Out, and the outer Shinneys, overlaid over multibeam backscatter data.

Figure 28. Substrates in Mogashy Tickle, River Out, and the outer Shinneys, overlaid over subsrate classification based on multibeam sonar (Copeland et al., 2007a).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report viii Figure 29A. Nearshore gravel substrate, shallow water portions of transect 9, Mogashu Tickle.

Figure 29B. Bathymetric and substrate profile, transect 9, running across Mogashu Tickle.

Figure 30. Boulder gravel, site 419, Leg island Basin.

Figure 31: 3-dimensional multidimensional scaling (MDS) plot of video sampled biota.

Figure 32: 3-dimensional multidimensional scaling (MDS) plot of biota from grab samples.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report ix List of Tables. Table 1: Wentworth Grain Size Definitions

Table 2. Numbers of video sequences of each substrate analyzed for faunal composition.

Table 3: The top 10 taxa from video sampled biota, ranked in order of decreasing contribution to faunal similarity, and their percent contribution to similarity within each substrate class.

Table 4: The top 10 taxa from grab sampled biota, ranked in order of decreasing contribution to faunal similarity, and their percent contribution to similarity within each substrate class.

Table 5. ANOSIM results comparing faunal composition of identical substrate types in areas covered by multibeam data and areas outside multibeam coverage. ANOSIM of video data revealed significant differences between all pairs. ANOSIM of grab sampled data revealsed significant differences for muddy gravel only.

Table 6: ANOSIM results table for 2007 video data. Numbers indicate probability of p- values, in percent, i.e. p=0.05 is represented as 5.0.

Table 7 : ANOSIM results table for 2007 grab sample data. Numbers indicate probability of p-values, in percent, i.e. p=0.05 is represented as 5.0.

Table 8. Contributions of individual taxa in video data to faunal dissimilarity between identical substrate types in all of Gilbert Bay (2006-7 data) and zone 1B (2007-8 data).

Table 9. Contributions of individual taxa in grab data to faunal dissimilarity between identical substrate types in all of Gilbert Bay (2006-7 data) and zone 1B (2007-8 data).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 1 Research Objectives and Rationale The primary goals of Phase III marine habitat mapping in Gilbert Bay Marine Protected Area

(MPA; Figure 1) are to collect detailed data on substrate and biotic composition in Zone 1B of

the MPA, especially to extend habitat classification to areas of The Shinneys outside multibeam

sonar coverage, to assess the nature of unclassified areas within River Out, and describe the

substrate and habitat of the Mogashu Tickle tidal rapids. A fourth research objective was to

target potential rock-wall habitats in Zone 1B and in Leg Island Basin (MPA Zone 3). Bedrock

wall habitats are prominent, and important, in many fjords, but our Phase II habitat maps

suggested that bedrock wall habitats are rare or absent in Gilbert Bay.

Research efforts were concentrated on Zone 1B for several reasons. First, large areas of Zone

1B, especially The Shinneys, were excluded from the habitat maps generated from multibeam

sonar data, because they were too shallow or too hazardous to be surveyed by the multibeam

sonar launch. Second, Zone 1B, especially the Shinneys, is of particular interest because of its

importance for cod spawning and juvenile habitat in Gilbert Bay. Third, River Out was

identified as an area of particularly high substrate and habitat diversity, and high species

diversity in Phase II of our mapping program. Furthermore, Phase II mapping showed that River

Out has high abundances of coralline-algae encrusted gravel, and true rhodoliths, or mobile balls

of branching coralline red algae (maerl, in European terminology). This habitat is highly bio-

diverse and possibly plays a role in the early life history of Gilbert Bay cod, therefore it was given special consideration in 2007. River Out also has a number of unclassified areas, where bathymetry, backscatter, and slope fell outside the statistically defined ranges of any of the six principal substrate types found within the bay. Finally, the area surrounding Mogashu Tickle,

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 2 the narrow strait connecting River Out and The Shinneys, was identified in our Phase II report as a key area of interest for its tidal rapids, and the consequent potential for a unique and highly diverse faunal assemblage.

Figure 2: Locations of management zones associated with Gilbert Bay Marine Protected Area and place names used in the report.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 3

Figure 2: Benthic substrate map for entire bay. Phase III mapping efforts focused on The Shinneys, River Out, and Leg Island Basin.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 4

Figure 3. Substrate map of Zone 1B of the Gilbert Bay Marine Protected Area, including The Shinneys and River Out. Note the extensive areas outside multibeam coverage, especially in The Shinneys. Mogashu Tickle is the narrow connection between the Shinneys and River Out.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 5

Research Approach Our benthic habitat classification approach relies on two important assumptions. First, we assume that substrate strongly influences the distribution of benthic biota; a relationship which has been established for nearshore and continental shelf environments elsewhere (Kostylev et al.

2001; Harney et al. 2006). Second, we assume a consistent relationship between substrate and multibeam sonar-derived values of depth, slope and backscatter (Todd et al. 1999; Kostylev et al. 2003; Ferrini and Flood 2006). The operational procedure for conducting the classification is described in our Phase II report (Copeland et al. in 2007a).

In Phase III, substrate classes identified throughout the bay were extended to the shallow areas of

The Shinneys using video transects coupled with georeferenced single-beam sonar bathymetry, but not backscatter. Additional substrate types were described as encountered. Faunal composition of substrates was determined using multivariate statistical analysis of video transects and grab samples. The operational procedure for conducting the classification is as follows:

1. Define major substrate classes from the measured grain-size distribution of grab-sampled

sediments and visual texture descriptions of video transects.

2. Create substate-classified lines from video using substrate classes defined for the bay as a

whole, and any new substrate types identified in video transects.

3. Create bathymetric profiles with substrate by combining sounder and video data.

4. Apply multivariate statistics to assess biotic composition of substrate types.

Analysis of substrates and faunal composition of potential bedrock wall habitats followed methods outlined in our Phase II report (Copeland et al., 2007a).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 6

Field Methods The goals of the Phase III field program were to conduct grab sampling and video data acquisition along transects in Zone 1B of the MPA, in which most transects included small areas covered by multibeam sonar data, and more extensive areas without multibeam coverage. The description of bottom type and dominant biota along these transects were used to to classify and map substrates and habitats for areas outside the multibeam coverage, using categories consistent with those developed for the areas with multibeam sonar data (fig 3). In River Out, video transects were conducted to cover previously unclassified areas near the edge of the multibeam coverage and drop camera stations were placed to fill gaps within the existing substrate map.

Transects were also planned to assess the nature of substrate and fauna in Mogashu Tickle.

Final, drop video sites within Leg Island Basin were sampled to assess the potential for bedrock wall habitats.

The fieldwork component of Phase II mapping was carried out over 12 days between

September 23 and October 04, 2007, and included 5.5 sampling days in Gilbert Bay, 4 travel days, and 2.5 days lost to weather or equipment failure. Sampling was conducted primarily from ashallow-draft speedboat, with the help of the boat operator, Mr. Jim Russell. The final sampling network appears in Figure 4, and the positional data for all stations are listed in

Copeland et al. (2007c). During video transects, time, position and depth were recorded using a

WAAS-enabled Garmin 178C GPS-depth sounder (figure 4). This, combined with the GPS overlay on the video, allowed for video classification to be translated directly into map format, and for construction of depth-substrate profiles along each transect.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 7

Figure 4. Alison Copeland operating the Garmin 178C GPS-sonar and Shark Marine Drop video systems.

Where possible, video transects were supplemented by grab sampling using a hand-operated

Petit-Ponar grab sampler (figure 5). The Petit Ponar, rather than the larger and heavier Van Veen

grab, was used. Conducting field work from the open deck speedboat precluded use of a heavy grab sampler requiring a winch or crab-pot hauler. Unfortunately, however, the Petit Ponar is

not heavy enough to effectively sample some of the harder gravely substrates, so many of the grab casts were unsuccessful and were returned empty.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 8

Figure 5. Philippe Leblanc readying the Petit Ponar hand-deployed grab sampler. After collecting subsamples for sediment texture and organic content analysis, samples were sieved through the floating mesh screen observed in the foreground.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 9 A total of 28 video transects were conducted in River Out, Barr Tickle, the outer Shinneys and

Mogashu Tickle, and the northern side of the Inner Shinneys, and recorded on approximately 13 hours of video tape. Transects in the Barr Tickle area of the Shinneys followed safe passage lines as navigated by the skipper. Transects were not sampled in the western edge of the Inner

Shinneys, because the skipper determined that this area was too hazardous.

A total of 28 grab samples were collected in 2007. Of these, 16 were collected opportunistically along the video transects where substrates appeared suitable for sampling with the small ponar grab (red dots on figure 6 below). The remaining 12 were collected at stations in River Out and the Shinneys to enhance the dataset from 2006 or sample locations that remained unclassified on the Phase II map (yellow dots on figure 6 below).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 10

Figure 6: Locations of all video and grab samples collected in Zone 1B in 2007 on multibeam backscatter intensity data. See more detailed maps below for grab and video drop site numbering. Video transect lines are represented schematically.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 11 Twenty-four drop video stations were sampled to assess potential bedrock wall habitats: in six stations in Zone 1B, indicated as black dots in figure 6, and 18 stations in Leg Island basin

(fig. 7). Potential bedrock targets were identified by high backscatter (>- 15dB) and high slope (>25°), following Copeland (2006).

Figure 7: Locations of drop video stations in Leg Island Basin to assess potential bedrock wall habitats. Red highlighted areas indicate potential bedrock wall locations with high backscatter and steep slope, while numbered black dots indicate camera drops.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 12 Laboratory Methods Laboratory research consisted of textural analysis of sediment samples collected in the field, identification of invertebrates collected in the field, analysis of underwater video footage, and statistical analysis of substrate and biological data.

Substrate Classification Substates were classified using visual classification of video footage and textural analysis of grab samples. Video classification of substrates followed the criteria established for visual substrate identification in Phase II of the habitat mapping program (Copeland et al., 2007a).

Textural Analysis: Sediment grab samples collected from Gilbert Bay were wet sieved to remove the mud fraction (silt and clay) from the sample. The weight of the mud that was washed out was calculated by weighing the dried sample before and after wet sieving. The remaining coarse material (sand and gravel) was dry sieved through a stack of 7 sieves with mesh sizes of −2,

−1, 0, 1, 2, 3 and 4ø. These sieves represent the boundaries of the grain size classes described by Wentworth (1922) which are shown in table 1. The grain size distribution of the whole sample could then be determined by combining the weight of material in each gravel and sand sieve with the known weight of the mud fraction. The results of the grain size analysis of the grab sampled sediments were combined with observations of the seabed from the video imagery to form a complete description of the substrate at each sampling site. These descriptions were then grouped into classes.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 13 Table 1: Wentworth Grain Size Definitions Phi unit Wentworth Grain Size Grain Size (mm) Fraction (ø) Description >−8 Boulder >256 gravel > −6 Cobble 63 to 256 gravel > −2 Pebble 4 to 63 gravel −1 Granule 2 to 4 gravel 0 very coarse sand 1 to 2 sand 1 coarse sand 0.5 to 1 sand 2 medium sand 0.25 to 0.5 sand 3 fine sand 0.125 to 0.25 sand 4 very fine sand 0.0625 to 0.125 sand <4 silt and clay < 0.0625 mud

Faunal identification: Fauna collected in grab samples were preserved in 70% ethanol in the field, then transported to Memorial University for identification under binocular microscope. Faunal identifications were based on Gosner (1971, 1979), Harvey Clark (1997) and additional references (---

ALISON!)

Video substrate and faunal analysis. Video transects were identified to substrate class visually, and by reference to the grab sample analysis. Following assignment to substrate class, a subsample of video segments of each substrate class were analyzed for fauna, at the presence-absence scale. The video footage of each transect line was reviewed to determine which of the 6 substrate classes from 2006 were present along it, or if any new substrates occurred. Each transect was then split into segments

> 10 seconds duration of a single substrate class (Table 2). Any substrate that was observed for less than 10 seconds (eg. a patch of gravel on a mud bottom) were not separated into a distinct segment. Once the transect had been broken into substrate class segments, the segments were reviewed for biotic composition. Fauna or algae present in each segment were identified and recorded for statistical analysis.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 14 Table 2. Numbers of video sequences of each substrate analyzed for faunal composition. Substrate Classification Number of Video Transect Segments Muddy Gravel 35 Sandy Gravel 10 Coralline Algae Encrusted Gravel 38 Mogashu Tickle Gravel 3 Nearshore Gravel 19 Gravelly Mud 36 Mud 21 1

Mapping video transect lines. Position data in UTM coordinates from the GPS-depth sounder were downloaded to mark the position and depth of all video transects and drop video targets. Segments of each transect were assigned to substrate classes based on visual assessment of the video data, and quantitative analysis of grab sample data where available. Transects were mapped according to substrate within ArcMap 9.2, and compared with multibeam backscatter and multibeam- derived substrate maps from Phase II. Bathymetric and substrate profiles for each transect were constructed by calculating the distance between each sounding point from its UTM coordinates, then plotting each transect as a cross-sectional bathymetric and substrate profile in Excel. Profiles for all transects are presented in Appendix A.

Statistical analysis.

Interpolation from sounding lines to bathymetric maps. An attempt was made to produce a continuous bathymetric map that would extend the initial bathymetric coverage derived from the multibeam soundings to the shallow water unmapped regions of The. Shinneys. Bathymetric soundings collected during Sept-Oct 2007 using the

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 15 Garmin 178C GPS-sounder were combined with the high-resolution multibeam data soundings in order to derive a dataset that would include all the available bathymetric data for the Shinneys and River Out sectors. The coastline from the NRCan 1:50,000 digital topographic maps was converted to points having a depth of zero meters, which has been added to the other bathymetric data. This allowed a control of the interpolation which stopped the interpolation process beyond the coastline. Interpolation procedure followed a nearest neighbour interpolation in ArcView 9.2.

Assessing depth-substrate relationships in areas without multibeam coverage.

In shallow water areas of The Shinneys, without multibeam coverage, Exploratory Data

Analysis (EDA) was used to visualize the depth distributions of the different substrate classes occurring in The Shinneys. Depth from the GPS-depth sounder and substrate class were compared statistically using EDA, ANOVA, and linear and exponential regression. 1-way

ANOVA was used to test the hypothesis that the different classes had distinct depth distributions. Finally, linear and exponential regression were used to test the ability of depth to predict substrate class in the shallow areas of the Shinneys outside the multibeam coverage.

Faunal analysis.

Species composition of the biota (both flora and fauna) was compared among sampled stations in the substrate types identified using the multivariate statistical techniques similarity percentage (SIMPER) analysis, non-metric multidimensional scaling (MDS), and Analysis of

Similarity (ANOSIM).. Analysis of all biotic data was conducted at the presence-absence level. MDS and ANOSIM analysis rely upon calculation of a Bray-Curtis similarity matrix, in

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 16 which the similarity of every sample in relation to every other sample is calculated. The

MDS ordination plot then shows the relative similarity of all samples; samples that plot close to one another are highly similar, while those plotting far from each other are highly distinct.

Similarity Percentage Analysis (SIMPER) Similarity percentage analysis (SIMPER) calculates the percent contribution of each taxon to the overall internal Bray-Curtis similarity of the biota associated with each substrate class, that is, the taxa that most clearly characterize the biota occupying each substrate. SIMPER also calculates the contribution of each taxon to the dissimilarity between all the substrate classes, but only the contribution to internal similarity is presented here.

In order to extrapolate substrates and habitats from the areas with multibeam sonar coverage to those without, we compared the faunal composition of each substrate class in the whole bay, based on the Oct 2006 data, with the faunal composition of that substrate class in the shallow water areas of Zone 1B, based on the Sept-Oct 2007 data. This comparison was conducted using ordination (MDS) analysis of similarity (ANOSIM), and the dissimilarity metric in the SIMPER, all conducted within PRIMER 5.0.. The extent to which the faunal composition of each substrate in the whole bay was similar to the composition of that substrate in the shallow areas indicates the reliability of extrapolating the substrate and habitat maps from the areas with multibeam data into the shallow areas outside the multibeam coverage.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 17 Results.

Substrates observed. In the areas of the Shinneys without multibeam coverage that were surveyed (figure 8), nearshore gravel, coralline-algae-encrusted-gravel, sandy gravel, muddy gravel, Fucus beds, gravelly mud, and mud were observed.

Figure 8. Location and numbers of video transects collected in River Out, Mogashu Tickle, and The Shinneys. In addition to the six substrate classes identified in Phase II, two additional substrate classes were recognized (see below). The characteristics of each of the eight substrate classes are described below.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 18 Muddy Gravel This class was characterised by pebbles, cobbles and occasionally boulders, in muddy sediment. The matrix (the sediment between the gravel) from grab samples in this class contained between 3 and 34% mud grains. Our analysis did not calculate the amount of silt and clay in the mud fraction of any of the substrate classes. Although the gravel was often draped by a thin layer of mud, this is a primarily gravel-bottom substrate. Boulder talus, where observed, consisted of coarse boulder gravel or exposed bedrock , sometimes covered in a thin layer of mud, generally on steep slopes. Potential bedrock and boulder gravel sites were mapped within the muddy gravel substrate class, because in most cases, boulder gravel was distinguished only by the coarse nature of the clasts.

Figure 9. Video frame grab of muddy gravel substrate, line 16R, Inner Shinneys.

Sandy Gravel The sandy gravel class was characterised by a matrix containing over 10% sand by weight and little mud. Most of the samples in this class contained sand which fell into the ‘very coarse sand’ (0ø) to ‘medium sand’ (2ø) range on the Wentworth scale (Wentworth

1922). The gravel component of this class was similar to the previous class, and included pebbles, cobbles and angular boulders.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 19

Fig. 10. Sandy gravel substrate. Transect 17, Inner Shinneys.

Coralline Algae Encrusted Gravel This substrate class contained pebbles, cobbles and boulder gravel that was at least 50% covered by branching coralline red algae. Coralline algae were observed in the other gravel classes, but samples where branching coralline algae, mostly Lithothamnion glaciale, formed a significant amount of the substrate structure were placed in this coralline-algae dominated class. The rhodoliths collected were densely branched and completely composed of coralline algae.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 20

Fig. 11A. Coralline algal encrusted gravel substrate, transect 16, Inner Shinneys.Image shown here captures true rhodoliths with associated sea stars.

Figure 11B . Example of coralline-algae-encrusted-gravel substrate class, as recovered in grab sampler. Branching coralline red algae, Lithothamnion glaciale, form biologically structured habitat on pebble-cobble gravel.

Gravelly Mud Samples in this class contained mud, occasionally with a very small amount of fine sand, and scattered gravel. Gravel of all sizes was observed, from small pebbles to boulders, scattered on a primarily mud bottom. Often a drape of mud covered the surface of the gravel.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 21

Fig. 12A. Gravelly mud substrate – note gravel clast with associated sun star. Scaling lasers 10 cm apart.

Figure 12B. Gravelly mud substrate, not scattered gravel clasts and mollusk shells. Transect 25, Inner Shinneys.

Mud Grab samples in the mud class were composed primarily of sediment grains smaller than 0.0625 mm (63μm) which fell into the silt and clay classes of the Wentworth scale

(Wentworth 1922). No gravel was observed.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 22

Fig 13. Mud substrate. Note burrow mouth. Transect 17, Inner Shinneys.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 23 Nearshore gravel. This class was characterized by gravelly sediment, sometimes with cobbles or boulder, and usually without either sand or mud, and found in very shallow water, especially in The Shinneys.

Fig. 14A. Nearshore gravel substrate, Transect 25, Inner Shinneys.

Fig 14B. Nearshore gravel substrate, Transect 25, Inner Shinneys.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 24 Mogashu Tickle Gravel. This class was characterized by gravelly sediment, without sand or mud, and was found only in the high-current areas surrounding Mogashu Tickle.

Fig 15A. Sponges, rhodoliths, and sea stars on gravel, Mogashu Tickle Gravel.

Fig 15B. Gravel with sun star and dead scallop shell, Mogashu Tickle gravel.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 25 Bedrock. This class was characterized by steeply sloping bedrock, as distinguished from boulder talus, which was included in the muddy gravel substrate. Only six sites of the 24 sampled contained true bedrock.

Fig 16A. Bedrock wall habitat, , site 418, Leg Island Basin.

Fig 16B. Bedrock wall, site 409, Leg Island Basin.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 26 Substrate distributions and bathymetry in the Inner Shinneys. The distribution of these substrates in the inner Shinneys is presented in figure 17, along with the backscatter map for the areas covered by the multibeam data. The depth profiles indicate a trough running WSW-ENE, approximately parallel to transect 25 filled with mud and gravelly mud, with bedrock ridges covered in coralline algae-encrusted gravel and muddy gravel to the W, E, and S, and the shore to the N. The bedrock ridge to the south of line 25 is also manifest as the two small islands separated by line 18. Similarly oriented bedrock ridges separating muddy troughs are likely to occur in the area SE of the multibeam coverage, but north of Barr Tickle. Contrary to expectations, Barr Tickle is not uniformly shallow nearshore gravel, but also contains considerable areas of mud and gravelly mud.

Continuity of substrate distributions. Mapping of substrate types from video transects in conjunction with the substrate classified multibeam data (figure 18) verified the substrate classifications in the multibeam data, and showed that the substrate distributions are generally continuous from the multibeam-covered areas to the areas outside multibeam coverage.

Following two pages: Figure 17. Distribution of substrates along bathymetric and video transect lines that were surveyed in the inner part of The Shinneys, overlad over multibeam backscatter. Figure 18. Distribution of substrates along bathymetric and video transect lines surveyed in the Inner Shinneys, overlaid over substrate map generated from multibeam backscatter (Copeland et al., 2007a).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 27

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 28

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 29 Attempted interpolation from survey lines to continuous bathymetry. The sounding lines collected in the Inner Shinneys during Sept.-Oct. 2007 were too sparse to allow a reliable mapping of the bathymetry, although several lines collected provided useful information on the bathymetric variability in the Shinneys. Some shallow water regions of the

Shinneys were not surveyed at all, as the skipper did not want to risk an accident in these zones. Indeed, the boat struck a rock and broke a propeller in Barr Tickle. The sounding lines in other parts of the Shinneys were too sparsely positioned offer reliable interpolations, as the seabed in the shallow water of the Shinneys is highly variable. A complete bathymetric map was generated but, as its reliability was very variable, it has not been used to derive a substrate or habitat map, nor released. To be able to use the information without taking a risk of over-interpreting the data, one would need to carefully understand the initial data used and the interpolation process. The production of a reliable complete bathymetric map of the shallow waters of Gilbert Bay would be possible, but would require considerable additional small boat time to collect a high density of bathymetric survey lines.

Substrate-Depth relationship in the Shinneys. The relationship between depth and substrate class in the Shinneys, including both areas covered by multibeam and not covered, was fairly consistent, and is well illustrated by line 25

(figure --). Nearshore gravel occurred in the shallowest waters, as did sandy gravel, in many but not all cases. Coralline-algal-encrusted-gravel and muddy gravel occurred in slightly deeper water, while gravelly mud and mud occurred in deeper troughs. Figure --- shows a similar profile for transect 28, in Barr Tickle. Profiles for all survey lines are included in appendix A.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 30

Video transect #25

1600 1400 1200 1000 800 600 400 200 0 0.0

-2.0

-4.0

-6.0

-8.0

Nearshore gravel -10.0 (m) Depth Coralline algae encrusted gravel Gravelly mud -12.0 Mud Muddy gravel -14.0 Sandy gravel -16.0 Distance (m)

Figure 19A: Bathymetric profile along line 25, labeled by substrate type. Note the general, but not absolute, relationship between substrate and depth.

Video transect #28

0 100 200 300 400 500 600 700 800 900 1000 0.0

-1.0 -2.0 -3.0 -4.0

-5.0

Depth (m) -6.0 -7.0 -8.0 Coralline algae encrusted gravel Gravelly mud -9.0 Muddy gravel -10.0 Distance (m)

Figure 19B. Bathymetric and substrate profile, transect 28, Barr Tickle. Note abundance of gravelly mud, and general but not uniform location of coralline algae and muddy gravel on ridges, but gravelly mud in troughs.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 31 Statistically, however, the relationship between depth and substrate was not strong. The water depths in which the different substrate classes in the Shinneys were found varied significantly (figure --- boxplot, ANOVA F(6,3968) = 489, p<0.001).

0

-10

-20 DEPTH -30

-40

-50

0 1 2 3 4 5 6 SUBS-CODE

Figure 20. Boxplot showing depth distribution of video transect points in the Shinneys. Substrate codes: 0: rockweed bed, 1: nearshore gravel, 2: coralline algae encrusted gravel, 3: sandy gravel, 4: muddy gravel, 5: gravelly mud, and 6: mud. Data derived from transect lines 11-25 and 28-30.

Nonetheless, depth was not a very strong predictor of substrate. Linear regression returned a r2 value of only 0. 261, while exponential regression gave an r2 value of 0.366. In both cases,

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 32 most variance in substrate class could not be explained by depth alone. Square-root transformation of depth increased the r2 value of the linear regression to 0.323, but did not alter the exponential regression of depth vs. substrate.

DEPTH_POS

Observed 50 Linear Exponential

40

30

20

10

0 0 1 2 3 4 5 6 SUBS-CODE

Figure 21. Linear and exponential regression lines of depth and substrate code in The Shinneys. Exponential regression explained more of the variance (r2=0.366) than did linear regression (r2=0.261), but neither accounted for more than 40% of the variance in substrate class using sonar-determined depth. Substrate codes: 0:Fucus bed, 1: nearshore gravel, 2: coralline algae encrusted gravel, 3: sandy gravel, 4: muddy gravel, 5: gravelly mud, and 6: mud. Data derived from transect lines 11-25 and 28-30.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 33 Detailed substrate and habitat mapping in River Out. Habitat mapping in River Out focused on further characterizing the coralline-algal-encrusted- gravel in this area, particularly with respect to development of rhodolith beds, determining the substrate and habitat in previously unclassified areas of River Out, and describing the substrate and habitat in Mogashu Tickle, the tidal rapids joining River Out and The Shinneys.

Coralline-algal-encrusted-gravel in River Out included extensive areas of rhodolith, in the form of round, mobile colonies of branching coralline red algae, especially Lithothamnion glaciale.

Figure 22: Fledermaus image of substrate classification draped over bathymetry, River Out (Copeland et al., 2007a). Coralline encrusted gravel in shallow water and muddy or sandy gravel in deeper water in RiverOut. Unclassified steep-walled low backscatter areas near mouth of Snooks Arm were determined to be muddy gravel.

Following two pages: Figure 23. Substrates in River Out, overlaid above multibeam backscatter.

Figure 24 Substrates in River Out, overlaid over substrates interpreted from multibeam sonar (Copeland et al., 2007a).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 34

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 35

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 36 Previously unclassified areas near the mouth of Snook’s Arm, in River Out, which had fallen outside the bathymetry, backscatter, and slope distributions of the six main substrates in

Gilbert Bay, were determined to be muddy gravel occurring on steep slopes (figure ---).

Figure 25A. Location of frame grab of muddy gravel occurring on steep slopes near mouth of Snook’s Arm, end of transect 7, River Out.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 37

Video transect #7

180 160 140 120 100 80 60 40 20 0 0.0

-5.0

-10.0

-15.0

-20.0 Depth (m) Depth Muddy gravel -25.0 Gravelly mud Mud -30.0 Sandy gravel Coralline algae encrusted gravel -35.0 Distance (m)

Figure 25B. Bathymetric and substrate profile of transect 7, near mouth of Snook’s Arm, River Out. L (180) = SW, R (0) = NE. Steep muddy gravel on NE end of transect represents previously unclassified region.

Mogashu Tickle. Copeland et al. 2007 predicted that Mogashu Tickle, the tidal rapids joining River Out and

The Shinneys, would be bedrock or boulder gravel, and that this area would have extremely high biodiversity. The substrate in Mogashu Tickle was determined to be not bedrock, but rather coarse cobble-boulder gravel, lacking in sand or mud. Steep current-swept gravel in

Mogashu Tickle was occupied by a diverse suite of suspension-feeding fauna, including sponges, the soft coral Gersemia rubiformis, and the basket star Gorgonocephalus sp.

Although the soft coral was also observed in scallop dredges in the upper part of Zone 2,

Gorgonocephalus was not observed elsewhere in the bay.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 38

Fig. 26A, Soft coral, Gersemia rubiformis, on gravel, Mogashu Tickle gravel.

Fig 26B. Basket star, Gorgonocephalus, Mogashu Tickle.

Next two pages: Figure 27. Substrates in Mogashu Tickle, River Out, and the outer Shinneys, overlaid over multibeam backscatter data.

Figure 28. Substrates in Mogashy Tickle, River Out, and the outer Shinneys, overlaid over subsrate classification based on multibeam sonar (Copeland et al., 2007a).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 39

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 40

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 41 Figure 29 shows the bottom type and bathymetric and substrate profile for transect 9, running perpendicular to the bidirectional water flow through the tidal rapids in Mogashu

Tickle. The shallowest water areas are nearshore gravel developed on talus derived from banded gneiss (figure 29A), while the deep-water areas are coralline algae encrusted gravel and mogashu tickle gravel (mtg; figure 29B).

Figure 29A. Nearshore gravel substrate, shallow water portions of transect 9, Mogashu Tickle.

Video transect #9

0 20 40 60 80 100 120 140 160 180 0

-5

-10

Depth (m) -15

Nearshore gravel -20 Coralline algae encrusted gravel Mogashu Tickle gravel -25 Distance (m)

Figure 29B. Bathymetric and substrate profile, transect 9, running across Mogashu Tickle.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 42 Potential Bedrock Wall habitats. Most potential bedrock wall habitats investigated in Leg Island Basin were determined to be dominantly boulder gravel derived from talus slopes, rather than true bedrock exposures.

Clast sizes appeared to be mostly large cobbles to large boulders. Many of these rocks had a thin layer of mud overlying them, causing much of the boulder gravel and bedrock sites to resemble the muddy gravel substrate mapped elsewhere in the bay. Of 18 potential bedrock wall targets in Leg island Basin, only six were actually bedrock, while the remainer were coarse muddy gravel derived from talus deposits. Because the majority of sites meeting the multibeam sonar criteria for potential bedrock wall environments were not actually bedrock, bedrock walls cannot be mapped within Gilbert Bay using ground-truthed multibeam sonar.

Figure 30. Boulder-gravel, site 419, Leg island Basin.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 43 Habitat Classification Habitat was mapped as the substrate-biotic units, as defined by samples collected in Oct 2006 and Sept-Oct. 2007. First, the biota associated with each substrate class were recorded based on the video transects and grab samples. Second, the degree of differentiation among the biota associated with each substrate class was assessed using multivariate statistics. Because species can be identified with a different degree of taxonomic precision in grab samples and video data, the two data sources are analyzed separately. The final habitat maps for each area show the substrate-biotic units that are statistically distinguishable in video. Characteristic fauna for each substrate type, as determined by SIMPER analysis, are indicated in Table 3 and 4, for video and grab sample data, respectively.

Muddy gravel The species contributing the most to similarity of grabs within the muddy gravel class are primarily epifauna present on sampled gravel. Video samples from the muddy gravel class were dominated by encrusting coralline algae, echinoderms such as the green sea urchin, sea stars and brittle stars, clams (mainly Macoma sp.), and hydroids, . (Tables -- and -

-) . Iceland scallops (Chlamys islandica) contributed 3 % to similarity in video data, indicating that muddy gravel may provide habitat for scallops. Characteristic fauna in grab samples included the bryozoan Schizoporella biaperta, the jingle shell Anomia squamula, and scallops.

Sandy gravel Like muddy gravel, the grabs from the sandy gravel class were characterised by epifauna present on the gravel component of the substrate. These included encrusting coralline algae, green sea urchins, sea stars, and clams.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 44 Coralline algae encrusted gravel SIMPER analysis of biota from video transects in the coralline algae encrusted gravel class indicated that the biotic assemblage of this substrate is characterised by branching coralline red algae and small epifauna associated with it. The highest contributor to group similarity among the grab sampled biota was the jingle shell

(Anomia squamula) which was particularly numerous among the branches of coralline algae.

Green sea urchins, sea stars, scallops, clams including Macoma, Astarte undata, the red chiton

(Tonicella rubra), Lepidonotus squamatus and Styela partita and whelks were also characteristic of the fauna in the densely branched coralline algae. Coralline algae encrusted gravel, including rhodolith, is likely to provide important habitat for scallops.

Gravelly mud. The biotic assemblage of the gravelly mud class was dominated by deposit feeding, infaunal species, such as mud stars, bivalves and brittle stars. The highest contributors to similarity in grab samples from this class were the Macoma calcarea, maldanid tube-dwelling worms, Crenella glandula, the jingle shell Anomia squamula, the cleft clam (Thyasira flexuosa) and Arctic boring clam (Hiatella arctica). The surface of the sediment observed in videos showed frequent obsevations of the green sea urchin, clams, and encrusting coralline reg algae. Large burrows, which were abundant at times, as well as siphon pits and other signs of bioturbation were common. Polychaete tubes were observed protruding from the sediment.

Mud In videos the surface of mud substrates showed evidence of biological activity such as trails on the surface, likely from mud stars, burrows, tubes, and bivalve siphon pits. Green sea urchins and clams, probably Macoma, were also characteristic in video. Mud bottoms were

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 45 primarily inhabited by infaunal bivalves, including the cleft clam Thyasira flexuosa, which alone contributed over half the biotic similarity of grabs in this class, as well as Macoma calcarea, Diastylis quadrispinosa, and tube-dwelling maldanid worms.

Nearshore gravel. Nearshore gravel substrates were characterized by their low biodiversity, and were characterized by only two dominant taxa: the green sea urchin Strongylocentrotus droebachiensis, and the blue mussel Mytilus edulis. The low abundance and diversity of fauna on this substrate is likely attributable to ice scour. Due to the hard compact nature of this substrate, it was not possible to collect grab samples.

Mogashu Tickle Gravel. Mogashu Tickle Gravel had a diverse and unique faunal assemblage. While many of the taxa that were characteristic of this substrate, such as the green sea urchin, the sea stars Henricia, Crossaster and Leptasterias, and scallops, were also characteristic of other substrates, this habitat was distinguished by suspension feeding fauna characteristic of high current settings. These included cnidarians, including the red soft coral

Gersemia rubiformis and hydrozoans, sponges such as Halichondria, and the basket star

Gorgonocephalus.

Bedrock. The fauna of the six bedrock sites was generally similar to that of muddy gravel substrates, as indicated by its location in the MDS ordination. Bedrock fauna was characterized by only three taxa: bivalve shell hash, branching bryozoans, and hydroids. The bivalve shell hash may be partly derived from dead scallop shells discarded from scallop draggers in Leg Island Basin. Because the six bedrock sites did not have unique acoustic

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 46 characteristics in multibeam data to distinguish them from muddy gravel, the bedrock class cannot be reliably mapped. Given its scarcity, and given the fact that only 6 of 24 potential bedrock targets were found to be true bedrock, mapping bedrock as a distinct habitat type would be inappropriate. Bedrock walls do not apparently constitute an important habitat type within Gilbert Bay.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 47

Table 3: The top 10 taxa from video sampled biota, ranked in order of decreasing contribution to faunal similarity, and their percent contribution to similarity within each substrate class. Coralline Algae Encrusted Nearshore Mogashu Muddy Gravel Sandy Gravel Gravel Gravelly Mud Mud gravel Tickle Gravel Bedrock Encrusting coralline Encrusting Branching coralline Strongylocentrotus Clams (30%) Strongylocentrotus Henricia (13%) Bivalve shell hash 1 algae (32%) coralline algae algae (24%) droebachiensis droebachiensis (59%) (37%) (36%) (85%) Strongylocentrotus Strongylocentrotus Strongylocentrotus Clam (36%) Strongylocentrotus Mytilus edulis Crossaster (13%) Branching 2 droebachiensis droebachiensis droebachiensis droebachiensis (7.6%) bryozoans (27%) (28%) (26%) (20%) (30%) Clams (15%) Leptasterias (13%) Leptasterias (16%) Encrusting Burrows (20%) Leptasterias (13%) Hydroids (7%) 3 coralline algae (9%) Leptasterias (7%) Clams (8%) Crossaster (11%) Tubes (5%) Tubes (13%) Strongylocentrotus droebachiensis 4 (13%)

Hydroids (3%) Crossaster (6%) Clams (10%) Burrows (4%) Hydroids (8%) 5

Crossaster (3%) Henricia (9%) Leptasterias (3%) Halichondria (8%) 6

Live scallops (3%) Live scallops (5%) Barnacle (8%) 7

Live scallop (8%) 8

Gersemia 9 rubiformis (4%)

Encrusting 10 coralline algae (4%)

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 48 Table 4: The top 10 taxa from grab sampled biota and their percent contribution to similarity within each substrate class. Grab samples were not collected from sandy gravel, neashore gravel, Mogashu Tickle gravel, or the boulder-bedrock sites.

Coralline Algae Muddy Gravel Encrusted Gravel Gravelly Mud Mud Schizoporella Anomia squamula Macoma calcarea Thyasira flexuosa 1 biaperta (46%) (31%) (27%) (45%) Anomia squamula Hiatella arctica (12%) Maldanidae (15%) Macoma calcarea 2 (39%) (12%) Live scallop (8%) Tonicella rubra (8%) Crenella glandula Diastylis 3 (15%) quadrispinosa (11%) Lepidonotus Anomia squamula Mya arenaria squamatus (7%) (13%) (10%) 4

Styela partita (7%) Hiatella arctica Maldanidae (7%) 5 (13%) Slime sponge (7%) Thyasira flexuosa Chaetozone setosa 6 (5%) (4%) Whelk (4%) Clinocardium Sipunculidae (4%) 7 ciliatum (4%) 8 Astarte undata (4%) Ophiura aculeata 9 (3%) 10 Lithothamnion (3%)

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 49 Defining Mappable Habitat Types: Mappable habitat types were defined by separating the substrate-biotic associations that were

statistically distinct, and combining the substrate-biotic associations that were statistically

identical. The statistical techniques used to make this determination were multidimensional

scaling (MDS) and analysis of similarity (ANOSIM). First, in order to extend map habitat units

defined for areas covered by multibeam sonar data to areas outside multibeam coverage, the

fauna of the same habitat in the two areas should be statistically identical. Second, the faunal

composition of unique substrate types identified in video transects was compared with that of the

previously defined substrates, to determine whether these could be mapped as distinct habitats.

Ordination using Multidimensional scaling (MDS) Ordination using multidimensional scaling illustrates the degree of difference in species

composition among samples. Figures 31 and 32 show the result of ordinations comparing faunal

composition between years where (habitat)_07 indicates the data for the entire bay, collected in

2006, and (habitat)_08 indicates the data for the areas outside multibeam coverage, collected in

2007, for muddy gravel, sandy gravel, coralline-algae-encrusted gravel, gravelly mud, and mud.

Ordination of the species composition data from video samples (Fig. 31) clearly shows a

separation between the biota of the hard substrates (muddy gravel, sandy gravel and coralline

encrusted gravel) and the soft substrate environments (gravely-sandy mud, gravely mud, and

mud), with a stress value of 0.14, indicating a highly reliable analysis (verified by ANOSIM, see

below). Comparison between years indicates relatively little differences between years for mud,

sandy gravel, and coralline-algae-encrusted gravel, and considerable differences between years

for muddy gravel and gravelly mud. Nearshore gravel compares most closely to muddy gravel,

while the Mogashu Tickle Gravel compares most closely to coralline-algae-encrusted-gravel.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 50

Figure 31: 3-dimensional multidimensional scaling (MDS) plot of video sampled biota. Mg: muddy gravel; sg: sandy gravel; cor: coralline-algal-encrusted-gravel; gm: gravelly mud; nsg: nearshore gravel; mtg: mogashu tickle gravel; bdrk: bedrock.

Ordination of the species composition data from grab samples (figure 32) shows a clear

separation between the fauna of the shallow hard substrate samples (sandy gravel and coralline-

algae encrusted gravel) and the soft substrates (gravely-sandy-mud, gravely mud, and mud), but

the muddy gravel substrate (black) was highly variable, and overlapped its composition with

samples from both hard and soft substrates. Similar to the inter-year comparisons for the video

data, ordination of the grab sampled faunal data indicated relatively little inter-year difference for

mud, and coralline-algae-encrusted-gravel, and considerable inter-year differences for muddy

gravel and gravelly mud. Sandy gravel was not represented in the grab samples from Phase 3,

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 51 and gravelly sandy mud was not represented in shallow areas of the Shinneys at all. The stress value of the analysis of the grab sample data (0.13) indicates a fairly reliable analysis, in which differences in biotic composition among substrate classes are slightly less clear than in the video data.

Figure 32: 3-dimensional multidimensional scaling (MDS) plot of biota from grab samples. Mg: muddy gravel; sg: sandy gravel; cor: coralline-algal-encrusted-gravel; gm: gravelly mud.

Analysis of Similarity (ANOSIM) Analysis of similarity found significant differences in species composition between areas on

identical substrate classes in the whole bay versus the shallow water areas of Zone 1B (table 5).

Considering video-sampled fauna, differences between years were significant for all substrates,

but in grab-sampled fauna, differences between years were significant for muddy gravel and

gravelly mud, but not for coralline-algae-encrusted-gravel or for mud.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 52 Table 5. ANOSIM results comparing faunal composition of identical substrate types in areas covered by multibeam data and areas outside multibeam coverage. ANOSIM of video data revealed significant differences between all pairs. ANOSIM of grab sampled data revealsed significant differences for muddy gravel only. Data type video Grab

Statistic R p R p

Muddy gravel 0.526 0.001 0.696 0.0001

Sandy gravel 0.35 0.013 n/a n/a

Coralline-algae- 0.423 0.001 0.048 0.276

encrusted-gravel

Gravely mud 0.424 0.001 0.483 0.004

Mud 0.657 0.001 -0.059 0.594

Analysis of similarity (ANOSIM) comparisions of faunal composition between substrates sampled in video transects found significant differences among all substrate types except muddy gravel vs. sandy gravel (table 6). ANOSIM comparisons of faunal composition between substrates grab sampled in 2007 found significant differences between muddy gravel and coralline-algae encrusted gravel, muddy gravel and mud, coralline-algae-encrusted gravel and mud, but lack of significant differences among the other classes (table 7).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 53 Table 6: ANOSIM results table for 2007 video data. Numbers indicate probability of p- values, in percent, i.e. p=0.05 is represented as 5.0. Muddy Sandy Coralline algae Gravelly Nearshore Mogashu gravel gravel encrusted gravel mud Mud gravel tickle gravel Muddy gravel − Sandy gravel 82.5 − Coralline algae encrusted gravel 0.1 0.1 − Gravelly mud 0.1 0.4 0.1 − Mud 0.1 0.1 0.1 0.1 − Nearshore gravel 0.1 0.2 0.1 0.1 0.1 - Mogashu Tickle 0.1 - Gravel 0.1 0.6 0.1 0.1 0.1 Bedrock 0.1 1.9 0.1 0.1 0.1 0.1 0.6

Table 7 : ANOSIM results table for 2007 grab sample data. Numbers indicate probability of p-values, in percent, i.e. p=0.05 is represented as 5.0.

Muddy Coralline algae Gravelly gravel encrusted gravel mud Mud Muddy gravel − Coralline algae encrusted gravel 3.7 - Gravelly mud 9.5 5.9 - Mud 0.2 0.2 7.1 −

Dissimilarity analysis using SIMPER. In order to determine the taxa responsible for inter-year differences in faunal composition within

substrate types, the SIMPER analysis was used to indicate the taxa responsible for greatest levels

of dissimilarity between years (tables 8, 9). This analysis showed that the faunal composition of

areas outside multibeam coverage differed from that of identical substrates in the whole bay

principally as a result of increased abundance of shallow-water taxa, especially encrusting

coralline algae, clams, and the green sea urchin Strongylocentrotus droebachiensis, and in

muddy substrates, unidentified tubes, and a reduced abundance of typically deeper-dwelling

species such as ophiuroids, bivalve shell hash, the sea star Asterias vulgaris, and the sponge

Halichondria panacea. Because the dissimilarity between samples of identical substrate classes

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 54 in the whole bay and in zone 1B was largely attributable to changes in the frequency of occurrence of typically shallow-water or deep-water species, these habitats are mapped as continuous habitat types, rather than as separate shallow-water and deep-water habitat types on the same substrate classes.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 55 Table 8. Contributions of individual taxa in video data to faunal dissimilarity between identical substrate types in all of Gilbert Bay (2006-7 data) and zone 1B (2007-8 data). Positive numbers indicate greater frequency of occurrence in zone 1B, while negative numbers indicate greater frequency of occurrence in the rest of Gilbert Bay (red). Taxa in bold consistently more frequent in zone 1B than in remainder of bay, while taxa in red text consistently less frequent in zone 1B than in remainder of the bay. Species Muddy Sandy gravel Coralline- Gravelly mud gravel algae- mud encrusted gravel Encrusting coralline 12.12 10.61 -6.02 7.41 algae Clams 8.24 6.53 9.5 7.04 9.83 Leptasterias polaris 6.56 -4.90 -3.61 -5.53 Live scallops 5.54 -7.28 -7.24 Hydrozoans 5.12 -5.16 -6.09 Orange sponge 4.54 -4.64 Strongylocentrotus 4.01 5.25 9.13 18.38 droebachiensis Branching bryzoans -5.02 4.35 Ophiuroids -5.71 -4.94 -5.23 -6.06 -10.56 Crossaster paposus -6.56 -6.01 7.71 5.04 Bivalve shell hash -7.74 -7.08 -6.61 -6.66 -5.56 Halichondria panacea -8.22 -6.50 -5.83 -5.60 Metridium senile -6.03 Asterias vulgaris -3.62 -7.57 -5.33 Branching coralline 5.55 -13.52 algae Henricia sp. 3.65 -4.17 9.05 Burrows -8.49 -9.00 Ctenodiscus crispatus 10.25 Tubes 6.45 13.50

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 56 Table 9. Contributions of individual taxa in grab data to faunal dissimilarity between identical substrate types in all of Gilbert Bay (2006-7 data) and zone 1B (2007-8 data). Positive numbers indicate greater frequency of occurrence in zone 1B, while negative numbers indicate greater frequency of occurrence in the rest of Gilbert Bay.

Species Muddy gravel Coralline-algae Gravely mud mud encrusted gravel Schizoporella biaperta 4.03 Anomia squamula 3.76 3.76 live scallops 2.48 Crenulla glandula 2.45 3.91 -5.17 Escharella immerse -2.28 granulate -2.3 -3.05 Balanus balanus -2.45 -2.71 Astarte undata -2.51 2.65 Thyasira flexuosa -2.81 -3.9 -4.36 -2.86 Stomachetosella sinuosa -2.96 Acmaea testudinalis -2.97 Spirorbis borealis -3.16 2.58 Ophiura robusta -3.24 Tubulipora -3.33 Nuculana tenuisulcata -4.32 -4.26 Branching Lithothamnion -3.25 Encrusting coralline algae -3.05 Puncturealla noachina -2.59 Hiatella arctica 2.75 3.81 slime sponge 2.96 Styela partita 2.99 Tonicella rubra 3.01 Lepidonotus squamatus 3.01 Ctenodiscus crispatus -3.49 -7.94 Pandora -7.71 Macoma calcarea 4.56 -5.08 Goniada maculate -3.06 -4.86 Nucula tenuis -3.43 -4.75 Paranidae -4.75 leathery mud tube -3.24 -4.26 Mya arenaria 4.79 Diastylis quadrispinosa 5.07 Maldanidae 3.57 5.93 unID polychaete -3.17 Clinocardium ciliatum 2.69

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 57 Study Highlights and Future Research Directions

Marine habitat mapping in the Shinneys and River out emphasizes a number of important results.

Inner Portions of The Shinneys

The Shinneys has been identified as important spawning grounds for Atlantic cod, and as

important habitat for juvenile cod, yet these features are not expected on mud bottoms that

constitute much of the central part of the Inner Shinneys. Mapping areas outside the multibeam

coverage verified the presence of bedrock ridges, covered with coralline algal encrusted gravel

and muddy gravel habitat, separating muddy basins. Substrate and bathymetry were clearly

linked in the shallow portions of the Shinneys outside multibeam coverage, but the correlation

between substrate and depth was not strong enough to allow direct translation of depth to substrate characterization. Furthermore, interpolation from the current suite of bathymetric survey lines to a continuous bathymetric surface was not reliable.

A reasonable benthic substrate and habitat map for The Shinneys could probably be constructed if two additional steps were undertaken. First, single-beam GPS-sonar data would need to be collected at higher spatial density throughout the Shinneys, from archival individual soundings

(e.g. data archives of Morris & Green), and from new bathymetric surveys. This would best be

done at high tide from a small shallow-draft boat. A dense collection of single-beam points could be interpolated into a bathymetric surface with greater reliability than what is possible given the current level of data resolution. Second, extrapolation from bathymetry to substrate

would rely on the underwater landforms, or the configuration of ridges and troughs, rather than

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 58 absolute bathymetric values. Such extrapolation could be executed using the Benthic Terrain

Modeler (BTM) extension on ArcMap (NOAA, http://www.csc.noaa.gov/products/btm/).

River Out – Mogashu Tickle. Among the most topographically and biologically complex areas in Gilbert Bay is the narrow passage connecting The Shinneys to the main part of Gilbert Bay,

River Out and Mogashu Tickle. Video transects in River Out verified the presence of abundant and large rhodoliths, and showed that unclassified areas consisted of muddy gravel on steep slopes, particularly close to Snook’s Arm. Rhodoliths, and coralline algae encrusted gravel habitat in general, have high biodiversity of invertebrates, and may constitute important recruitment area for Iceland scallops. Mogashu Tickle does not consist of bedrock ridges, but rather current-swept cobble and boulder gravel, with abundant coralline algae, sponges, soft corals, basket stars, and other hard-substrate fauna that thrive in areas of strong currents.

Substrate and faunal composition of Mogashu Tickle are distinct enough to justify its identification as a unique habitat type.

Steep rock walls in Leg Island Basin

Leg Island Basin is the only portion of Gilbert Bay where deep steep rock walls occur, a type of environment often found in fjords that typically forms a distinctive habitat. Video sampling of potential bedrock wall targets in Leg Island Basin showed that most of these targets were boulder gravel derived from talus slopes, rather than true bedrock exposures. Although fhe fauna of these boulder gravel and bedrock habitats was statistically distinct, it was generally most similar to that of muddy gravel and gravelly mud in the outer parts of the bay.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 59 Summary

Phase III of marine habitat mapping in Gilbert Bay, Labrador, found the following key results.

1. Habitats occurring in The Shinneys and River Out include the main habitat types

identified throughout the bay, plus faunally poor nearshore gravel and biodiverse

Mogashu Tickle gravel in the tidal rapids connecting the Shinneys to River Out.

2. Substrate and habitat distributions in areas of the Shinneys outside multibeam coverage

follow bathymetry, and are generally continuous with substrates and habitats observed on

multibeam.

3. River Out contains abundant and highly diverse rhodolith beds within the coralline algae

encrusted gravel habitat.

4. Previously unclassified areas of River Out include steeply inclined muddy gravel near the

mouth of Snook’s Arm. Most unclassified areas within River Out likely fall into this

category.

5. Bedrock wall habitats are extremely rare in Gilbert Bay, and are dominated by boulder

talus, rather than bedrock itself. These habitats have a faunal composition mostly like

that of muddy gravel. They do not constitute a unique mappable unit.

6. further mapping research in the Shinneys could interpolate bathymetry and habitats

between bathymetric and video transect lines if additional bathymetric data were

compiled and collected. Such mapping would use benthic terrain modeling.

7. Juvenile and adult fish distribution data from the Shinneys should be mapped in relation

to substrates observed in the habitat mapping program.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 60 References Anderson, J.T., R.S. Gregory and W.T. Collins. 2002. Acoustic classification of marine habitats in coastal Newfoundland. ICES J. Mar. Res. 59: 156-167.

Bell, T., Copeland A., Devillers, R, Edinger E, Hu, L., Wroblewski, J., 2006. Habitat mapping in Gilbert Bay Marine Protected Area, Phase II field report, Nov 2006.

Berglund B, ed, 1986 Handbook of palaeoecology and palaeohydrology. John Wiley and Sons, New York, 247-270.

Copeland, A., Bell, T., Edinger, E., Shaw, J. and Gregory, R. Benthic Habitat Mapping in Newman Sound – A Newfoundland Fjord. GeoHab – Marine Geological and Biological Habitat Mapping, Sixth International Symposium, Victoria, BC Canada. May 4th-7th 2005.

Copeland, A., Bell, T., Edinger, E., Hu, L., Wroblewski, J., 2006. Habitat mapping in Gilbert Bay, Labrador, a Marine Protected Area, Phase 1. Contract report to Department of Fisheries and Oceans., St. John’s, NL., 42 p.

Copeland, A., Edinger, E., Bell, T., Devillers, R., Wroblewski, J., Hu., L., 2007a. Habitat mapping in Gilbert Bay, Labrador, a Marine Protected Area: Phase II final report. Marine habitat mapping group Report 07-02, 108 p.

Copeland, A., Bell, T., Edinger, E., 2007b. Marine habitat mapping in Newman Sound, Terra Nova National Park, Newfoundland Marine habitat mapping group Report 07-01, 167 p.

Copeland, A., Leblanc, P., Edinger, E., Devillers, R., Bell, T., 2007. Marine habitat mapping in Gilbert Bay Marine protected Area, Phase III field report. Nov. 2007.

Cote, D., S. Moutlon, P.C.B. Frampton, D.A. Scruton and R.S. McKinley. 2004. Habitat use and early winter movement by juvenile Atlantic cod in a coastal area of Newfoundland. J. Fish Biol. 64: 665-679.

Gosner, K.L. 1971. Guide to Identification of Marine and Estuarine Invertebrates: Cape Hatteras to the Bay of Fundy. John Wiley and Sons, Inc.

Gosner, K.L. 1979. A Field Guide to the Atlantic Seashore: Invertebrates and Seaweeds of the Atlantic Coast from the Bay of Fundy to Cape Hatteras. Houghton Mifflin Company, Boston.

Green, J.M. and J.S. Wroblewski. 2000. Movement patterns of Atlantic cod in Gilbert Bay, Labrador: Evidence for bay residency and spawning site fidelity. Journal of the Marine Biological Association of the U.K., 80: 1077-1085.

Hall-Spencer, J.M. 1998. Conservation issues relating to maerl beds as habitats for molluscs. Journal of Conchology Special Publication No. 2: 271-286.

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 61

Hall-Spencer, J.M., Moore, P.G., 2000. Scallop dredging has profound, long-term impacts on maerl habitats. ICES J. Mar. Sci. 57: 1407-1415.

Hu, L. 2007. Modeling the resident fish production, the ecosystem carrying capacity and population dynamics of Atlantic cod (Gadus morhua) in Gilbert Bay, Labrador: A Marine Protected Area. MSc. Thesis Environmental Science Programme, Memorial University.

Kamenos, N.A., P.G. Moore, and J.M. Hall-Spencer, 2004. Small-scale distribution of juvenile gadoids in shallow inshore waters: what role does maerl play? ICES J. Mar. Sci. 61: 422-429.

Kostylev, V.E., B.J. Todd, G.B.J. Fader, R.C. Courtney, G.D.M. Cameron and R.A. Pickrill. 2001. Benthic habitat mapping on the Scotian Shelf based on multibeam bathymetry, surficial geology and sea floor photographs. Marine Ecology Progress Series 219: 121-137.

Kostylev, V. E, R.C Courtney, G. Robert and B.J Todd. 2003. Stock evaluation of giant scallop (Placopecten magellanicus) using high-resolution acoustics for seabed mapping. Fisheries Research 60: 479-492.

Morris, C.J. and J.M. Green. 2002. Biological characteristics of a resident population of Atlantic cod (Gadus morhua L.) in southern Labrador. ICES Journal of Marine Science, 59: 666-678.

Schneider, D.C., J-M. Gagon, and K.D. Gilkinson, 1987. Patchiness of epibenthic megafauna on the outer Grand Banks of Newfoundland. Marine Ecology Progress Series, 39: 1-13.

Wentworth, 1922.

Wroblewski, J.S., L.K. Kryger-Hann, D.A. Methven and R.L. Haedrich. 2007. The fish fauna of Gilbert Bay, Labrador: A marine protected area in the Canadian subarctic coastal zone. J. Mar. Biol. Ass. U.K. (in press).

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 62 Appendices.

Appendix A. Bathymetric and substrate profiles for all survey lines in Inner Shinneys.

Video transect #15

0 100 200 300 400 500 600 0.00

-5.00

-10.00

Depth (m) -15.00

-20.00 Gravelly mud Muddy gravel Coralline algae encrusted gravel -25.00 Distance (m)

Video transect #16

0 100 200 300 400 500 600 700 0.0

-5.0

-10.0

Depth (m) -15.0

Coralline algae encrusted gravel -20.0 Gravelly mud Muddy gravel -25.0 Distance (m)

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 63

Video transect #17

0 50 100 150 200 250 0

-2

-4

-6 Depth (m) -8 Gravelly mud Muddy gravel -10 Coralline algae encrusted gravel Mud -12 Distance (m)

Video transect #18

0 100 200 300 400 500 600 700 0

-5

-10

-15 Nearshore gravel Depth (m) -20 Mud Coralline algae encrusted gravel -25 Muddy gravel Gravelly mud -30 Distance (m)

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 64

Video transect #20

0 50 100 150 200 250 300 0.0

-2.0

-4.0 Nearshore gravel Coralline algae encrusted gravel -6.0 Gravelly mud

Depth (m) Depth Mud -8.0

-10.0

-12.0 Distance (m)

Video transect # 21

0 50 100 150 200 250 300 350 0.0

-2.0

-4.0

-6.0 Depth (m) -8.0 Nearshore gravel Coralline algae encrusted gravel -10.0 Mud Gravelly mud -12.0 Muddy gravel Distance (m)

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 65

Video transect # 23

0 50 100 150 200 250 300 350 0.0

-5.0

-10.0

Depth (m) -15.0

Coralline algae encrusted gravel -20.0 Gravelly mud Muddy gravel Mud -25.0 Distance (m)

Video transect #24

0 50 100 150 200 250 300 0.0

-2.0

-4.0

-6.0

-8.0

Depth (m) -10.0

-12.0 Fucus -14.0 Gravelly mud Mud -16.0 Distance (m)

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 66

Video transect #25

1600 1400 1200 1000 800 600 400 200 0 0.0

-2.0

-4.0

-6.0

-8.0

Nearshore gravel -10.0 (m) Depth Coralline algae encrusted gravel Gravelly mud -12.0 Mud Muddy gravel -14.0 Sandy gravel -16.0 Distance (m)

Video transect #28

0 100 200 300 400 500 600 700 800 900 1000 0.0

-1.0 -2.0

-3.0 -4.0

-5.0

Depth (m) -6.0

-7.0 -8.0 Coralline algae encrusted gravel Gravelly mud -9.0 Muddy gravel -10.0 Distance (m)

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 67 Appendix B: Species list of all species encountered in 2006 and 2007 field work.

Annelida: Polychaeta Pectinaria granulata ─ Trumpet worm Nothria (Onuphis) conchylega − Tube worm Harmothoe imbricata ─ Fifteen scaled worm Lepidonotus squamatus ─ Twelve scaled worm Pherusa plumosa Chaetozone setosa Nereis sp. ─ Clam worm Nephtys sp. − Red lined worms Goniada maculata – Chevron worm Diopatra cuprea ─ Plumed worm Spirorbis borealis ─ Hard tube worm Spirorbis spirillum ─ Hard tube worm Spirorbis granulatus ─ Hard tube worm Serpula sp. ─ Hard tube worm Phyllodoce sp. ─ Paddle worm Family Paraonidae Family Maldanidae ─ Bamboo worm Family Terebellidae

Aschelminthes: Priapulida Priapulus caudatus

Arthropoda: Pycnogonida Anoplodactylus lentus ─ Lentil sea spider

Arthropoda: Crustacea Balanus balanus − Rough barnacle Ampithoe rubicata − Tube building amphipod Leptocheirus pinguis − Tube building amphipod Gammarus sp. − Amphipod Casco bigelowi − Amphipod Hyas araneus − Toad crab Pagurus acadianus −Hermit crab Spirontocaris spinus − Caridean shrimp Lebbeus sp. − Caridean shrimp Sclerocrangon boreas – Red sand shrimp Diastylis quadrispinosa − Cumacean

Brachiopoda Hernithyris psittacea ─ Parrot beak lampshell

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 68

Bryozoa Membranipora sp. Stomachetosella sinuosa Escharella immersa Smittina sp. Eucratea loricata Flustra foliacea Porella sp. Schizoporella biaperta Tubulipora sp. Lichenopora sp. Scrupocellaria scabra

Chordata: Ascidiacea Styela partita − Rough sea squirt Didemnum albidum − White crust Ascidian

Cnidaria: Anthozoa Metridium senile − Frilled anemone Gersemia rubiformis – Red soft coral

Cnidaria: Hydrozoa Sertularia pumila Campanularia sp. Abietinaria sp.

Echinodermata: Asteroidea Crossaster papposus ─ Spiny sun star Solaster endeca ─ Smooth sun star (purple sun star) Henricia sp. ─ Blood star Hippasteria phyrygiana ─ Horse or cushion star Leptasterias polaris − Polar sea star Asterias vulgaris ─ Northern or common sea star Ctenodiscus crispatus ─ Mud star

Echinodermata: Crinoidea Heliometra glacialis

Echinodermata: Echinoidea Strongylocentrotus droebachiensis − Green sea urchin

Echinodermata: Holothuroidea Cucumaria frondosa − Orange-footed sea cucumber

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 69

Echinodermata: Ophiuroidea Ophiura sarsi ─ Boreal brittle star Ophiura robusta ─ Brittle star Ophiopholis aculeata − Daisy brittle star Gorgonocephalus arcticus – Northern basket star

Foraminiferans Calcareous foraminiferans Agglutinating foraminiferans

Mollusca: Bivalvia Clinocardium ciliatum ─ Iceland cockle Anomia squamula ─ Prickly jingle shell Astarte undata ─ Waved Astarte Cyclocardia borealis − Northern cardita Nuculana tenuisulcata ─ Elongate nut clam Hiatella arctica − Arctic rock borer or Arctic saxicave Yolida sapotilla − Short yolida Nucula tenuis ─ Thin nut shell Crenella glandula ─ Glandular bean mussel Mya truncate ─ Truncate soft-shelled clam Macoma balthica − Baltic macoma Macoma calcarea – Chalky macoma Thyasira flexuosa ─ Cleft clam Periploma papyratium − Paper spoon shell Chlamys islandica − Iceland scallop Cerastoderma pinnulatum ─ Little cockle Mytilus edulis − Blue mussel

Mollusca: Gastropoda Margarites helicinus ─ Smooth top shell Margarites costalis ─ Ridged top shell Turritellopsis acicula ─ Needle shell Puncturella noachina − Noah’s punctured shell Acmaea testudinalis ─ Tortoise shell limpet Velutina laevigata − Velvet shell Tachyrhynchus erosus

Buccinum undatum – Waved whelk

Mollusca: Polyplacophora Tonicella rubra ─ Red chiton Ischnochiton albus ─ White chiton

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 70 Porifera Halichondria panicea − Breadcrumb sponge Scypha ciliata − Vase sponge Finger sponge (possibly Haliclona oculata) Orange sponge (possibly Fig sponge - Suberites ficus) White sponge

Sipunculida: 1 unidentified species

Habitat Mapping in Gilbert Bay Marine Protected Area, Phase III final report 71 Appendix C: Sediment data for grab samples: grain size distribution and organic content.

Grab station Substrate Organic % 301 muddy gravel 6.10 304 mud 23.87 305 mud 23.25 306 mud 19.48 308 mud 20.76 309 mud 20.72 309dup mud 20.29 311 mud 15.24 312 mud 16.27 22A gravelly mud 6.33 23A gravelly mud 15.24 25A mud 13.02 25B mud 22.86 30A gravelly mud 18.69 30B gravelly mud 10.10 4B muddy gravel 16.21 5B gravelly mud 7.47