Deep Sea Coral Final Report Integrating Mapping and Fisheries Data for Deep-Sea Coral Habitats off South Carolina and Georgia by M. Scott Harris Department of Geology and Environmental Geosciences College of Charleston Charleston, SC 29401 843.953.0864 [email protected] George R. Sedberry Gray's Reef National Marine Sanctuary 10 Ocean Science Circle Savannah GA 31411 30 November 2011

EXECUTIVE SUMMARY Hard-bottom reefs in depths from 50-800 m off the coast of Georgia and South Carolina were mapped using multibeam sonar. Sonar images revealed rugged bottom topography along a nearly continuous reef at the shelf edge (centered at 50 m depth); at around 200 m on the iceberg-scoured upper slope off South Carolina; and on the Blake Plateau in depths from 400800 m. These complex reefs are important fish habitats, and many economically valuable fishes spawn in areas of steep slopes and rugged bottom topography that also support deep-sea corals and sponges.

INTRODUCTION Deepwater rocky reefs and associated sponges and corals provide important habitat (including spawning habitat) and feeding grounds for a diversity of fishes, many of which are economically valuable fishery species. In order to implement effective placed-based management, mapping of these essential fish habitats (EFH) is needed to protect sensitive life-history stages. Initial surveys of deep-sea coral habitats (defined as those supporting corals in depths >50 m) of the South Atlantic Bight (SAB, Cape Hatteras NC to Cape Canaveral FL) have provided useful imagery of complex bottom topography where deep-reef fishes occur and spawn (Figure 1).

Cape Romain SC

Figure 1. Multibeam sonar imagery obtained in 2006-2007 aboard the NOAA Ship Nancy Foster, plotted with reef-fish spawning locations obtained from the SCDNR-MARMAP program (Sedberry et al. 2006). The shelf-edge reef occurs along the 55-m depth contour and the upperslope reef is centered on the 200-m depth contour. Prominent features on the shelf-edge reef include spawning locations of several fishes. Slope-reef fishes spawn on the upper slope reef, which is made more complex by iceberg scours on the bottom (Hill et al. 2008). Also of note are numerous spawning locations on the shelf, shelf edge and upper slope that were not yet been mapped with sonar in 2007.

Upper-Slope Reef

2

Additional work is needed to evaluate substrate rugosity, potential and known sites for deep-sea coral growth, and associations of fishes with bottom complexity and corals. Databases and Geographic Information Systems (GIS) are needed to map these features together so that they can be incorporated with databases on biodiversity and environmental conditions to determine habitats that might need protection to conserve biodiversity or sustainable fisheries (Fautin et al. 2010; Sedberry et al. 2011). Additional analysis of existing mapping data, and incorporation of those data into accessible databases is needed to complete previous and current projects funded by the National Oceanic and Atmospheric Administration, and aimed at determining the factors that constitute deep-sea coral habitat and spawning grounds for associated reef fishes. Reef fish spawning locations and deep coral reefs are considered Habitat Areas of Particular Concern (HAPC) by NOAA Fisheries Service and the regional Fishery Management Councils, including the South Atlantic Fishery Management Council (SAFMC), which manages fisheries in federal waters of the SAB. Analysis of high-resolution multibeam sonar data taken from known spawning locations is needed to determine the characteristics of these sites that make them attractive as spawning grounds for many reef fishes. By characterizing the sites with sonar (groundtruthed with visual observations and studies of fish reproduction), we can then use rapid sonar surveys to map other lesser-known or previously unknown spawning locations, and discover additional EFH and HAPC. Because deep reef fishes often spawn in deep coral banks and other biogenically-altered bottoms, mapping location of those features will enable us to locate additional spawning grounds and other EFH. Mapping of deep coral banks, and investigating their diversity and ecology, are important missions of NOAA. The goal of this project was to further process and analyze existing multibeam sonar and fishery databases to map habitat details and to detect correlations among fish distribution, spawning and habitat features. This information will allow habitat and fishery managers to better plan placedbase management to include Marine Protected Areas (MPAs) such as HAPCs and National Marine Sanctuaries (NMS). To obtain this goal, the following objectives were addressed: 1) construct expanded detailed sea floor maps and habitat characterization maps with existing multibeam sonar data by additional analysis of backscatter and other sonar data; 2) use backscatter and other existing sonar data to locate areas of potential or existing deepwater coral habitat; and 3) create georeferenced bottom topography images for a GIS database and overlay existing fisheries survey data to map fish distribution and spawning in relation to bottom features. This project is aimed at mapping complex habitats, including low- to high-relief hard grounds and rocky reefs, smooth and bioturbated muds, and biologically-engineered habitats such as grouper excavations, tilefish burrows, worm reefs and coral reefs. We are particularly interested in mapping habitat of reef fishes. This mapping will complement previous and current NOAA-funded projects aimed at determining the factors that constitute spawning grounds for reef fishes, especially deep-reef species. EFH comprises HAPC, and sonar mapping and groundtruthing of sonar signatures of bottom features around known spawning locations are needed to determine the characteristics of these sites that make them attractive as spawning grounds for many reef fishes. By characterizing the sites with sonar (groundtruthed with visual observations and spawning data), we can then use rapid sonar surveys to map other lesser-known or previously unknown spawning locations, and discover additional EFH and HAPC (e.g. Figure 1). Because deep reef fishes often spawn in deep coral banks (Gilmore and Jones 1992) and

3

other biogenically-altered bottoms, mapping location of those features will enable us to locate additional spawning grounds and other EFH.

STUDY AREA AND DEEP-SEA CORAL ECOSYSTEMS OF INTEREST Deep-sea (>50 M) coral ecosystems (DSCE) in the southeastern U.S. exist along the shelf break or shelf edge at depths from 55-75 m, on the upper continental slope in depths around 200m and on the hard grounds of the Blake Plateau in depths from 400-1000 m. The shelf-edge reef off South Carolina is an important habitat for corals and reef fishes and supports recreational and commercial fisheries. Octocorals and sponges dominate the attached megafauna, and include the vase sponge Ircinia campana, the antipatharian whip coral Stichopathes spp., and the fan coral Muricea pendula (Fraser and Sedberry 2008). Many fish species spawn there, with some undertaking extensive migrations from shallower reefs on the shelf to spawn at specific deep-reef features (McGovern et al. 2005; Sedberry et al. 2006; Paz and Sedberry 2008). The upper slope reef at 200 m is often densely covered by globular sponges (Geodia sp.), soft corals such as Plumarella sp. and Trachymuricea sp., and the scleractinian coral Lophelia pertusa. The slope reef is an important habitat for deep-reef fishes such as snowy grouper (Hyporthodus niveatus), blueline tilefish (Caulolatilus microps), blackbelly rosefish (Helicolenus dactylopterus) and others. Off South Carolina, this hard reef is characterized by iceberg scours up to 400 m wide and 20 m deep (Figure 2), and associated plowed-up rock piles that add additional complexity to the bottom (Hill et al. 2008). Soft mud-clay bottom also exists at this depth off South Carolina and Georgia, where tilefish (Lopholatilus chamaeleonticeps) construct burrows in which they dwell. On the Blake Plateau off South Carolina and Georgia, the rugged bottom topography of the Charleston Bump supports deep-sea corals and several fishery species, including wreckfish, Polyprion americanus (Sedberry et al. 2001). The predominate coral on Charleston Bump reefs are the azooxanthellate, colonial scleractinian corals: Lophelia pertusa, Madrepora oculata, and Enallopsammia profunda; various species of hydrocorals of the family Stylasteridae, black corals of the order Antipatharia, and species of gorgonian octocorals including the bamboo coral of the family Isididae.

4

METHODS We used multibeam sonar aboard the NOAA Ship Nancy Foster to collect topography and backscatter data in June of 2006 and 2007 to address the specific objectives of mapping habitats associated with fish spawning. Additional surveys conducted over the years by us were combined with this dataset to obtain a more complete assessment of potential coral habitat on deep reefs of the SAB. Additional effort was directed at mapping the SAFMC Deepwater Snapper-Grouper MPAs along the shelf edge, which coincided with known spawning locations determined from the Marine Resources Monitoring, Assessment and Prediction (MARMAP) program of the South Carolina Department of Natural Resources (SCDNR). Details of MPA location and spawning determination can be found in Sedberry et al. (2006). Mapping was conducted at shelf-edge, upper slope and Charleston Bump hard grounds (Figure 3). Because various vessels were used, mapping technology varied. Generally, surveys were conducted in a similar fashion on all cruises, and along lines that paralleled the bottom feature being surveyed, with occasional cross-transects to further validate initial survey readings. Surveys consisted of high-resolution (at least 4 m/pixel) multibeam sonar recorded continuously using echosounders interfaced with GPS. Multibeam sonar data collected in 2006 and 2007 aboard the Nancy Foster were collected using a Multi-Beam 1002 Simrad 95 kHz System (30m1000m) interfaced with Scientific Computer System (SCS). The SCS included a HYPACK Data Acquisition and Processing System, which will also recorded continuous data from the ship’s Acoustic Doppler Current Profiler (ADCP). A Sea-Bird SEACAT SBE-19 CTD or equivalent was used to determine temperature and salinity profiles. Sound velocity profiles were used to calculate sound velocity corrections for depth measurements. A dynamic motion sensor or heave, roll, pitch and motion sensor collected heave, pitch and roll measurements to be applied to raw soundings data for correction during processing. Tidal time and ratio correctors were obtained from NOAA/HSD and applied during at-sea post-processing to 6-min tidal values based on the Charleston (or other local) tide gauge. Positioning information was collected using a Trimble DSM212L GPS Receiver or equivalent, with integrated DGPS VHF receiver. Differential corrections were received from the Ft. Macon NC, Charleston SC or Miami FL radio beacons as appropriate. Antenna positions were corrected for offset and layback and referenced to the position of sonar transducer(s) in use at the time. Accuracy requirements were met as specified in the NOAA Hydrographic Manual and Field Procedures Manual (FPM). The Horizontal Dilution of Precision and Estimated Position Error as specified in the FPM was monitored during on-line data collection. If the positioning degraded beyond the acceptable limits while on line, the data were rejected or smoothed, depending on the extent of the affected data. Coastal Oceanographics HYPACK software or equivalent was used for data acquisition. Processing of sounding data was accomplished using NOAA Hydrographic Processing System included in HYDROSOFT 9.4, Mapinfo software, and the HPS-MI MapBasic application (or equivalent/upgrade). Post processing was done using CARIS HIPS software that was used aboard the Foster and available in the lab at the College of Charleston. CARIS HIPS includes statistical-based data cleaning and processing, and data validation tools to translate raw sounding data into GIS-based maps and shapefiles.

5

6

Habitat data from all sonar surveys was incorporated into a geodatabase containing historical data on fish distribution, to aid in characterization of reef fish spawning sites and other habitats based on biological and physical data. After preliminary processing at sea, bathymetric data and backscatter data from all cruises were compiled for the region. Detailed seafloor maps and habitat characterizations were made using a modification of Battista’s methodology. MARMAP fishery data (Sedberry et al. 2006) were incorporated into the database. Multibeam bathymetric data were cleaned and processed, or existing data and backscatter data were extracted using CARIS and QPS Fledermaus software. Metrics of bottom complexity (depth, slope, rugosity, etc.) were calculated and compared to the original bathymetric data to estimate validity of boundaries of bottom types. Fishery survey data were compared to seafloor images. For the bathymetric analysis, the following metrics were extracted from the bathymetric data: Bathymetric Mean Bathymetric Standard Deviation Curvature Curvature Plan Curvature Profile Rugosity Slope Slope of Slope The metrics were then converted to 8-bit (scale of 1-256), masked to the area of data, and stacked. A Forward Principal Component Analysis (PCA) Rotation using a Correlation Matrix was calculated and the first three PCA bands merged, taking smaller segments and creating larger objects. Spatial, spectral, textural, color space, and band ratio attributes for each area were calculated. Spatial attributes included Area, Length, Compactness, Convexity, Solidity, Roundness, Form Factor, Elongation, Rectangular Fit, Main Direction of Major Axis Angle, Major Axis Length, Minor Axis Length, Number of Holes and Holes Ratio. Spectral attributes included Min pixel value Band_x, Max pixel value Band_x, Avg pixel value Band_x and Std Dev pixel value Band_x. Textural attributes included Texture Range, Texture Mean, Texture Variance and Texture Entropy. Color Space and Band Ratio attributes were Band Ratio, Hue, Saturation and Intensity. The analyses were then compared with the original bathymetric data to estimate validity of boundaries.

RESULTS AND DISCUSSION Three primary depth ranges were identified: (1) the 50-m ridge along the shelf edge, (2) 200-m depths adjacent to the shelf edge, and (3) deeper areas along the Charleston Bump (Figure 3). 50-m Ridge The 50-m ridge represents a discontinuous to continuous rocky outcrop with one to ten meters of relief along the shelf edge. This is the "shelf-edge" reef described by Struhsaker (1969) as 7

extending from 55-110 m and encompassing an important reef fish habitat. Sedberry et al. (2006) noted spawning at these depths among several species of reef fish and Schobernd and Sedberry (2009) observed courtship behavior on shelf-edge reefs off Florida and South Carolina. The importance of the high-relief areas of the shelf edge was confirmed in the present study by plotting spawning locations in relation to bottom features (Figure 4). Our analysis of the shelfedge sonar data indicated a high rugosity that provides substantial complex habitat, contributing to the high diversity of fishes found there (Sedberry and Van Dolah 1983; Schobernd and Sedberry 2009).

200-m Scour Zone on the Upper Slope Hard bottom on the upper slope off Georgetown SC supports a fishery for deepwater reef fishes, including snowy grouper, blueline tilefish and blackbelly rosefish. This is the "lower shelf" fish habitat described by Struhsaker (1969) as a smooth-to-rugged bottom. The area surveyed by us represents a sandy to silty section with abundant iceberg scours providing a broken surface for high-rugosity habitat. Smooth areas are typically silty to sandy with rock pavement in places. Depths ranged from approximately 160 m to over 230 m, exemplifying the bottom complexity in this area. Bathymetry data collected aboard the Nancy Foster in 2006 and 2007 revealed numerous kilometer-scale furrows in depths from 170 - 220 m (Figure 2, Figure 5). The grooved features are typically 10 - 100 m wide and 10 km. Some larger furrows are up to 400 m wide and 20 m deep. The furrows range from linear to arcuate, 8

with crosscutting tracks. Many of the furrow troughs are flanked by plowed-up lateral berms that are several meters high and that often terminate in semicircular pits ringed by several-meterhigh ridges (Hill et al. 2008). Video observations conducted along these berms (Schobernd and Sedberry 2009; personal ROV observations) show evidence of large upturned blocks and boulders. There are also numerous circular pits, 50–100 m in diameter and several meters deep, scattered across the area surveyed (Figure 5). The furrows were observed along an irregularlyshaped platform that is elevated above the surrounding slope by about 10 m (Figure 5). High backscatter across the platform suggests hardground material with very little sediment cover. Low backscatter was observed in both the deeper regions off the platform and the base of the furrows and pits. This interpretation was confirmed by bottom observations in 1983, 2002 and 2010, which showed an abundance of boulders interspersed with patchy sediment across the platform, whereas the deeper regions were sand covered (Schobernd and Sedberry 2009; personal observations). Snowy grouper were often observed in submersible and ROV dives (Figure 5).

9

Charleston Bump The Charleston Bump sonar surveys (Figure 6) took place on the Blake Plateau off South Carolina and Georgia, in depths from 400 – 800 m, and documented the high-relief areas scoured by the Gulf Stream (Sedberry et al. 2001). The area surveyed is an important spawning ground for wreckfish, barrelfish (Hyperoglyphe perciformis) and red bream (Beryx decadactylus). Fishermen target the high-relief scarps (Figure 7), and these three species have been caught in spawning condition on those scarps (Sedberry et al. 2006; Filer and Sedberry 2008; Friess and Sedberry 2011).

10

Figure 7. Wreckfish capture locations (red triangles) in relation to bottom topography on the Charleston Bump. Wreckfish occur near steep scarps, in depths from 400-650 m.

11

MANAGEMENT SIGNIFICANCE AND APPLICATION New information of significance to management comes from the 50-m and 200-m depth areas. In every case along the 50-m depth contour from South Carolina to Florida, high-relief rocky outcrops were identified suggesting a complete, although discontinuous, ridge of rocky formations exists along the shelf break. Relief is particularly spectacular off northern Florida and South Carolina, whereas the reef is less prominent off of Georgia (Fraser and Sedberry 2008; present study). Further identification of iceberg scour marks noted in surveys conducted since those of Hill et al. (2008) also indicates strong possibilities of additional areas of complex habitat preferred by snowy grouper and other deep-reef fishes along the blocky margins of the scours. These areas need to be further surveyed between the sites surveyed by Hill et al. (2008) off South Carolina and more recent Pathfinder surveys off of northeast Florida and Georgia, and the extent of this habitat delineated fully. Spawning locations for several reef fish species were located in areas of high relief along the 50m reef. Some of these locations are in existing MPAs where bottom fishing is prohibited, and some are located in areas open to fishing. Because there are spawning season closures of the fishery for many of these species, spawning fish are still protected, although they might still occur in bycatch outside of the MPAs, resulting in mortality of spawning fish. Shelf-edge reefs (40 - 60 m) in the middle of the SAB appeared to be particularly important spawning grounds. Some of these reefs coincide with the SAFMC MPAs that prohibit bottom fishing, including sites off South Carolina and Florida. The MPA sites off South Carolina appear to be particularly important as spawning grounds for several species. Vermilion snapper, for example were found spawning in almost all of the MPA sites. Although many reef fish spawn at numerous locations across the continental shelf (Sedberry et al. 2006), spawning in some of those species such as vermilion snapper (Rhomboplites aurorubens) is more concentrated at the 50-m shelf edge reefs. Red porgy (Pagrus pagrus) and bank sea bass (Centropristis ocyurus) also have broad distributions throughout the region, but spawning appeared to be more narrowly focused on deeper 50-m sites off South Carolina. Gag (Mycteroperca microlepis), scamp (M. phenax), red grouper (Epinephelus morio), knobbed porgy (Calamus nodosus) and gray triggerfish (Balistes capriscus) spawn mainly at shelf-edge reefs. Tagging of gag has indicated a spawning migration to shelf-edge reefs (McGovern et al. 2005). Tilefish, blackbelly rosefish, blueline tilefish, snowy grouper and yellowedge grouper (Hyporthodus flavolimbatus) are resident, at least as adults, on the upper slope, and all except for tilefish associate with hard and complex bottom such as that found in the iceberg scours off South Carolina. Spawning for these species is restricted to reef (or mud in the case of tilefish) habitats on the upper slope. Barrelfish, wreckfish and red bream live on the Charleston Bump, mainly in depths from 500 600 m (Sedberry et al. 2001; Filer and Sedberry 2008; Friess and Sedberry 2011). Spawning also occurs there, under the main axis of the Gulf Stream, and it is uncertain how these fishes are recruited back to the Charleston Bump.

12

Although many reef fishes important in commercial and recreational fisheries off the southeastern U.S. spawn across broad shelf areas, it is evident that some spawning is localized. Often, local spawning grounds are utilized by several species. In deciding among options for additional no-take MPA sites, consideration should be given to sites that are used as spawning grounds by several species. Although we do not fully understand why reef fish choose high-relief shelf-edge sites for spawning, the combination of habitat complexity, food availability and proximity of the Gulf Stream and associated gyres for larval dispersal are all probably important. Spawning maps show that spawning occurs in many other areas that have not yet been mapped with sonar. To fully understand the relative contribution of bottom habitat and oceanography to spawning success, all reef fish spawning sites should be mapped with high-resolution sonar.

LITERATURE CITED Fautin, D., P. Dalton, L.S. Incze, J-A. C. Leong, C. Pautzke, A. Rosenberg, P. Sandifer, G. Sedberry, J.W. Tunnell Jr., I. Abbott, R.E. Brainard, M. Brodeur, L.G. Eldredge, M. Feldman, F. Moretzsohn, P.S. Vroom, M. Wainstein and N. Wolff. 2010. An overview of marine biodiversity in U.S. waters. PLoS ONE 5(8): e11914. doi:10.1371/journal.pone.0011914. Filer, K.R. and G.R. Sedberry. 2008. Age, growth and reproduction of the barrelfish, Hyperoglyphe perciformis (Mitchill, 1818), in the western North Atlantic. J. Fish Biol. 72:861-882. Fraser, S.B. and G.R. Sedberry. 2008. Reef morphology and invertebrate distribution at continental shelf edge reefs in the South Atlantic Bight. Southeastern Naturalist 7:191206. Friess, C. and G.R. Sedberry. 2011. Age, growth, and spawning season of red bream (Beryx decadactylus) off the southeastern United States. Fish. Bull. 109:20–33. Gilmore, R. G. and R. S. Jones. 1992. Color variation and associated behavior in the epinepheline groupers, Mycteroperca microlepis (Goode and Bean) and M. phenax Jordan and Swain. Bull. Mar. Sci. 5: 83–103. Hill, J.C., P.T. Gayes, N.W. Driscoll, E.A. Johnstone and G.R. Sedberry. 2008. Iceberg scours along the southern U.S. Atlantic margin. Geology 36:447-450. McGovern, J.C., G.R. Sedberry, H.S. Meister, T.M. Westendorff, D.M. Wyanski, and P.J. Harris. 2005. A tag and recapture study of gag, Mycteroperca microlepis, off the southeastern U.S. Bull. Mar. Sci. 46:47-59. Paz, G. and G.R. Sedberry. 2008. Identifying black grouper (Mycteroperca bonaci) spawning aggregations off Belize: conservation and management. Proc. Gulf Carib. Fish. Inst. 60: 577-584. Schobernd, C.M. and G.R. Sedberry. 2009. Shelf-edge and upper-slope reef fish assemblages in the South Atlantic Bight: habitat characteristics, spatial variation and reproductive behavior. Bull Mar. Sci. 84:67-92. Sedberry, G.R., D.G. Fautin, M. Feldman, M.D. Fornwall, P. Goldstein, and R.P. Guralnick. 2011. OBIS-USA: A data-sharing legacy of the Census of Marine Life. Oceanography 24(2):166–173. 13

Sedberry, G.R., J.C. McGovern and O. Pashuk. 2001. The Charleston Bump: an island of essential fish habitat in the Gulf Stream. Pages 3-24, in G.R. Sedberry, editor. Island in the stream: oceanography and fisheries of the Charleston Bump. American Fisheries Society, Symposium 25, Bethesda, Maryland. Sedberry, G.R., O. Pashuk, D.M. Wyanski, J.A. Stephen and P. Weinbach. 2006. Spawning locations for Atlantic reef fishes off the southeastern U.S. Proc. Gulf Carib. Fish. Inst. 57:463-514. Sedberry, G.R. and R.F. Van Dolah. 1984. Demersal fish assemblages associated with hard bottom habitat in the South Atlantic Bight of the U.S.A. Env. Biol. Fish. 11(4): 241-258. Struhsaker, P. 1969. Demersal fish resources: Composition, distribution, and commercial potential of the continental shelf stocks off southeastern United States. U.S. Fish. Wi1dl. Serv., Fish. Ind. Res. 4:261-300.

14