The Benthic Ecosystem of Mountain Top Bank, a New Mesophotic Coral Reef in the Northern Gulf of Mexico

Abstract

The Gulf of Mexico, a geologically complex environment, supports mesophotic coral ecosystems, with reefs such as the Pinnacle Trend, Flower Garden Banks National Marine Sanctuary, the Florida Middle Ground reef system, and Pulley Ridge. Mountain Top Bank is a dome-shaped hardground feature located 60–150 m below the sea surface along the Mississippi–Alabama shelf. It appears to prolong the Pinnacle Trend towards the southeast, bridging the gap between mesophotic coral reefs east and west of the Mississippi Canyon. Shipborne high-resolution multibeam data (bathymetry, backscatter, and water-column) and an AUV photomosaic were collected over the site during several oceanographic expeditions. Data were analyzed and compiled into an ArcGIS geodatabase to produce the first benthic habitat map of Mountain Top Bank. The site is characterized by a network of outcrops and boulders interspersed within a predominately sandy environment. Different seabed features were correlated with the presence and abundance of a diverse array of biota across the phyla of Cnidaria, Porifera, Mollusca, Chordata, Echinodermata, and Rhodophyta. We found the benthic assemblage to be similar to those found at the Pinnacle Trend, supporting the hypothesis that Mountain Top Bank is part of the same reef system and acts as a topographic bridge between ecosystems on the east and west of the Mississippi Canyon.

1. Introduction

Multibeam echo-sounder (MBES) acoustic data and high-resolution visual data are often combined to provide powerful tools to characterize the seabed and produce robust thematic maps [1]. We combined them to produce a benthic habitat map (BHM), which is an important tool for both bolstering our ecological understanding of the seafloor by spatially representing its abiotic and biotic relationships and distributions as well as aiding in decision making for future survey efforts [2]. Fully characterizing the seafloor is a difficult task because of the inability to easily access it, the lack of real-time observations, and the cost of oceanographic expeditions. Hence, mesophotic reefs, despite new insights into their abundance, diversity, and role in reef connectivity, are among the ecosystems that still remain largely unexplored throughout the global oceans [3]. Interdisciplinary research efforts focused on the benthos provide a dynamic understanding of this equally dynamic environment. They also support robust mapping initiatives that can be expanded upon as deep-sea studies continue to proliferate with technological advances. The predictive power of BHMs could also promote the efficient use of time and money by pinpointing where ground truthing efforts should be focused for future studies. The potential predictive power and derived information from BHMs could specifically prove to be an essential tool for planning Marine Protected Areas (MPAs) and the assessment and monitoring of extra space required for marine conservation [4,5].

Mountain Top Bank (MTB) is an isolated dome-shaped hardground feature emerging from the seabed at a depth of about 150 m along the Mississippi–Alabama continental shelf in the Northern Gulf of Mexico (GOM, Figure 1). This feature emerges from a flat and featureless seafloor and is located between the Pinnacle Trend (offshore Alabama) and the Flower Garden Banks National Marine Sanctuary (FGBNMS, offshore Texas), two of the most prominent mesophotic coral reefs in the GOM. Separated by the Mississippi Canyon, these two reefs are over 1000 km apart. Structural, and therefore potential biological and ecological, connectivity between the two sites has been speculated through the existence of sites such as MTB. Although coral presence was previously reported at MTB [6], it has never been extensively investigated, and the benthic ecosystem has never been defined. In this paper, we present the first BHM, geodatabase, and species catalog of MTB. The data collected allowed us to display different components of MTB and better understand the interplay between factors such as, but not limited to, the general ecology, preferred substrate, and preferred assemblage of the benthic organisms found at MTB. The species catalog confirms the similarity of MTB with the Pinnacle Trend and the role it could play as a biological corridor for mesophotic ecosystem populations.

2. Materials and Methods

MTB, bound by the coordinates −88.45 W, 29.21 N, −88.42 W, 29.24 N, is located offshore of the MS-AL coast, approximately 90 nm from Gulfport, MS, and to the east of the MS River Delta. It is a dome-shaped structural height occupying 3 km²; emerging from a flat surrounding seabed, it extends from 56.62–145.07 m below the water’s surface. The dome appears to be capped by a 20 m thick hardground and shows a moat around its base. Its dome structure may be associated with salt diapirism, which affects large parts of the GOM subsurface geology. Alternatively, because of the proximity to the Pinnacle Trend, it may share the same geological origin as the relict topography of the last glacial cycles [8].

The Hydrographic Science Research Center at the University of Southern Mississippi began studying MTB in 2007. The first MBES survey was conducted using the EM2000 from the AUV Eagle Ray, a torpedo-shaped explorer class AUV typically employed to map seabed deeper than 300 m. It revealed an unexpected and complex morphology on the top of the mound which appeared to be primarily characterized by rough hardground. This rough hardground, depth, and proximity to other well-known mesophotic coral ecosystems (MCEs) within the northern GOM suggested that MTB could support an extensive MCE. In the summer of 2020, the Mola Mola AUV was deployed from the R/V Point Sur to collect new still photo images and additional data from the MBES mounted on the vessel.

Mola Mola is a SeaBED-class AUV, Woods Hole Oceanographic Institution (WHOI), Woods Hole, (MA) USA, with horizontal and vertical thrusters that allow it to maneuver at low speeds (0.15–0.25 m/s) 3 m above the seafloor (Figure 2). The vehicle is equipped with a downward-facing Allied Vision Technology industrial color video camera. The camera is centrally mounted on the lower housing of the AUV, with two LED arrays positioned fore and aft to illuminate the seafloor from different angles during image acquisition. To minimize the influence of ambient light, all surveys were conducted at night. The camera produces color-corrected seafloor imagery at a resolution of 2448 × 2050 pixels. The image footprint, when at a 3 m altitude, is about 2.6 m along-track and 3.2 m across-track [9].

The Kongsberg EM2040 multibeam, Kongsberg, Norway, was mounted on the side pole of the R/V Point Sur to collect bathymetry, seabed reflectivity, and water column backscatter data.

2.1. Data Analysis

2.1.1. MBES Acoustic Data

MBES bathymetry data were processed using QPS Qimera 2.7 following the standard processing workflow suggested by [10]. Topographic and slope maps were created at 3 m bin resolution. The seafloor backscatter data was processed utilizing Fledermaus Geocoder Toolkit (FMGT) v.7.9.7 following the workflow suggested by [11] and also gridded at a 3 m pixel resolution. Once gridded, backscatter values were further analyzed in ArcGIS Pro 3.4 to predict sediment type distribution and infer habitat class types.

2.1.2. Photo Mosaic

A total of 3345 still images were collated into a seafloor photo mosaic. High-densities of ground-truthing samples are required to increase the certainty in acoustic data interpretation and validation [12]. The mosaic was created through a set of Matlab version R2019b (MathWorks, Inc., Natick, MA, USA) tools that helped to determine which images should contribute to each pixel of the mosaic. Once those images were determined, color data was copied into the associated pixels of the mosaic grid. The colors were scaled with coefficients applied to the three-color bands, which provides a correction for differential absorption. The illumination was also balanced to prevent bright zones at the center and dark edges on the images. The resulting colorized mosaic was exported as a GeoTIFF and uploaded into ArcGIS. The images were also combined with vehicle navigation (altitude, pitch, roll, heave, yaw, heading, and position) to derive coefficients for an affine transform within a Cartesian coordinate system, allowing the mosaic to be overlaid onto the bathymetry and backscatter data layers within ArcGIS.

2.1.3. Organism Identification and Enumeration

The photos were analyzed frame by frame to identify, enumerate, and categorize sessile macrofauna and motile demersal organisms at the lowest taxonomic level possible from only visual observation and the substrate inhabited, allowing us to determine the soft and hard bottom communities. Different categories were used to describe shifts in habitat types and better understand relationships between organisms and habitat geology. The megabenthos analysis categories adapted from [13] are: sponges (Porifera), Echinodermata, bioturbators (indicate infauna), algae (Rhodophyta), corals (Cnidarians), Mollusca, and coral fragments/rubble. The geologic environmental descriptors as outlined in [14] are: sand, gravel/pebble, cobble/boulder, and substrate outcrops. The relationships observed between the two formed the basis for our establishment of possible surrogate relationships. Our photographic records were compared to other studies and reports [14,15,16,17,18,19,20] for both the purpose of narrowing down what organisms are present within the GOM and likely found at MTB and allowing for specific species identification where possible.

Species identification and enumeration from these photos allowed for further analyses including the calculation of the Shannon Diversity Index and richness levels where applicable. Richness refers to the number of groups or individuals within a determined category; species, class, and subclass richness were the three types used for these counts. The Shannon Diversity Index estimates the diversity of species within a community, considering the number of species living in a habitat and their relative abundance. The equation for this index is:

$$H=-\sum p_i\times \ln p_i$$

where H is the Shannon Diversity Index and p_i is the proportion of individuals of i-th species in a community. The greater the diversity of species, the higher the index value.

2.1.4. Geodatabase

Our geodatabase compiled our remotely sensed data to map and store the spatial distribution of benthic habitats at MTB. Specific layers that were incorporated include our photomosaic, bathymetry, and backscatter maps. Seabed classification with ArcGIS allowed us to “see” the seabed in “pseudo-colors” that provide information on the material and topography of the seabed, while also providing a visual means of evaluating the spatial distribution of different variables [21]. It is important to acknowledge and evaluate a variety of environmental factors because of the dynamic nature of the seafloor.

3. Results

3.1. Seafloor Morphology

MTB extends from 56.62 m to 145.07 m below the sea’s surface, an 88.45 m change in elevation. The average depth throughout the entire structure is 113.82 m, which is around the base of the structure (Figure 3). The topography is most complex at the top of the mound structure and decreases as you move downwards towards the surrounding seafloor. Two well-defined circular pockmarks are also present along the south foot of the mound (Figure 3 and Figure 4). MTB has a slope ranging from 0 to 59.53°, averaging at 4.09° (Figure 4). The greatest change in slope occurs along the edge of the transition from the top of the mound to the relatively evenly sloping sides.

3.2. Seafloor Backscatter

Figure 5 visualizes the seafloor backscatter derived from the multibeam data. Each gridded portion within the collected backscatter has a unique value, ranging from −18 to −62 dB. The light colors are associated with high intensities and dark colors associated with low intensities. The backscatter values of high intensity were correlated with rock outcrops with sand both beside and between outcrop features. When sand was the predominant feature within the grid, the backscatter was in the middle range of intensities, while the low intensities were associated with the muddy/sand areas.

3.3. Image Data

The 60 × 60 m photomosaic produced by the still images collected at the top of MTB is shown in Figure 6. This area correlates to the box highlighted in Figure 3, Figure 4 and Figure 5. We utilized the still images to catalog the benthic organisms and to ground-truth the substrate of their habitat.

3.4. Organism Identification and Enumeration

In the following subsections, we illustrate all of the species identified visually by the analysis of the still photos and comparison of similar studies. The standard scientific taxonomic citation for a species is adopted: genus + species in italics and scientist and year of first description. Those references are not reported in the manuscript’s Reference section but rather provid-ed together with the species still photos in the Supplementary Materials Figures S1–S67.

3.4.1. Chordata

Seventeen fish species were identified while twenty-four fish were present in the images but unidentifiable (

Table 1. Fish species observed throughout the MTB study transect.

Chordata (Demersal Fish)
Number of Individuals	Species Richness	Shannon Diversity Index
93 (identifiable) 24 (unidentifiable) 117 (total)	16	2.32
Species	Counts
Honeycomb cowfish (Acanthostracion polygonius)	5
Reef butterflyfish (Chaetodon sedentarius)	17
Yellowtail reef fish or royal gamma (Chromis enchrysura, Gramma loreto)	2
Spinycheek soldierfish (Corniger spinosus)	1
Spotted moray (Gymnothorax moringa)	2
Queen angelfish (Holacanthus ciliaris)	2
Rock beauty angelfish (Holacanthus tricolor)	1
Squirrelfish (Holocentridae sp.)	2
(Grey) snapper (Lutjanus sp., L. griseus)	15
Short bigeye (Pristigenys alta)	14
Red lionfish (Pterois volitans)	15
Greater amberjack (Seriola dumerili)	4
Grouper (Mycteroperca sp.)	8
Scorpionfish (Scorpaenidae sp.)	3
Tattler fish (Serranus phoebe)	1
Sea robin (Triglidae sp.)	1
Present but unidentifiable	24

). The species identified were: short bigeye fish (Pristigenys alta Gill, 1862), honeycomb cowfish (Acanthostracion polygonius Poey, 1876), red lionfish (Pterois volitans Linnaeus, 1758), spotted moray (Gymnothorax moringa Cuvier, 1829), reef butterflyfish (Chaetodon sedentarius Poey, 1860), greater amberjack (Seriola dumerili Risso, 1810), snapper and grey snapper (Lutjanus spp. Bloch, 1790, L. griseus Linnaeus, 1758), grouper (Mycteroperca sp. Gill, 1862), scorpionfish (Scorpaenidae sp. Risso, 1827), rock beauty and queen angelfish (Holacanthus tricolor Bloch, 1795, H. ciliaris Linnaeus, 1758), tattler fish (Serranus phoebe Poey, 1851), sea robin (Triglidae sp. Rafinesque, 1815), yellowtail reef fish/royal gamma (Chromis enchrysura Jordan & Gilbert, 1882, Gramma loreto Poey, 1868), squirrelfish (Holocentridae sp. Bonaparte, 1833), and spinycheek soldierfish (Corniger spinosus Agassiz, 1831). The identified groupers could be M. interstitialis but their positioning in the still images makes it difficult to see the defining features required to make this ID; this similarly applies to the snapper fish identified, though it is possible they are all grey snappers. Three of the unidentifiable fish could belong to Anthiadidae sp. Poey, 1861, but were left within the unidentifiable category. Altogether, there were 95 individual fish identified among those species, the greatest counts being for grey snapper, reef butterflyfish (which were usually observed in pairs of two or four), red lionfish, short bigeye fish, and groupers; because fish are mobile organisms, there is a possibility for count overlap. The fish were observed throughout the study site, primarily atop or alongside the rock outcrops.

3.4.2. Cnidaria

Fourteen corals and one sea pen were observed (

Table 2. List of corals observed throughout the MTB study transect.

Cnidaria
Species Richness	Subclass Richness		Order Richness
8	2		4
Species	Subclass	Order	Counts	Appearances
Antipathes spp.	Hexacorallia	Antipatharia		764
Aphanipathes sp.	Hexacorallia	Antipatharia	6
Muricea pendula	Octocorallia	Alcyonacea	11
Oculina varicosa	Hexacorallia	Scleractinia	47
Siderastrea sp.	Hexacorallia	Scleractinia	16
Stichopathes sp.	Hexacorallia	Antipatharia		2042
Swiftia exserta	Octocorallia	Alcyonacea		643
Tanacetipathes spp.	Hexacorallia	Antipatharia		406
Unknown mounding/encrusting coral Type 1 Type 2 Type 3 Type 4	-	-	67 16 84 28	-
Unknown soft corals	-	-	60
Unknown star coral	Hexacorallia	Scleractinia	9
Unknown sea pen	Octocorallia	Pennatulacea	4
Present but unidentifiable	-	-		2029

). Among the corals observed, the following were identified to the species level: Stichopathes sp., Swiftia exserta, Muricea pendula, Antipathes spp., Tanacetipathes spp. Opresko, 2001, Aphanipathes sp. Brook, 1889, Oculina varicosa Le Sueur, 1820, and Siderastrea sp. Blainville, 1830. The most observed species were Stichopathes sp., Antipathes spp., S. exserta, and Tanacetipathes spp. It should be noted that of the 2489 photos that had corals, 2029 photos had corals present that were too dark or blurry to identify. There were also white branching corals, orange branching corals, and star corals that could not be positively identified as well as mounding organisms that were either corals or sponges. Visual observation alone is a limiting factor in coral identification, and physical samples would be needed for verifiable identification of both the unknown and some of the identified species. Also, some coral counts were only documented as the number of appearances rather than number of organisms within that species observed because of dense assemblages, overlap between photos, and location along rock outcrops making exact counts difficult. Throughout our study site, we observed piles of coral fragments. They were documented in 210 images, indicative of 70 instances of coral rubble. Most of these fragments were in consolidated piles or spread out across sandy sediments. Although rare, there were some instances where they were found on top of and on the side of rock faces and crevices.

3.4.3. Porifera

Ten different sponges were observed (

Table 3. List of sponges observed at MTB.

Porifera
Number of Individuals	Species Richness	Shannon Diversity Index
254 (identifiable) 282 (unidentifiable) 637 (total)	5	1.17
Species	Class	Counts
Agelas clathrodes	Demospongiae	13
Halichondria sp.	Demospongiae	9
Ircinia sp.	Demospongiae	49
Geodia sp.	Demospongiae	34
Psuedoceratina/Aoilochroia crassa	Demospongiae	149
Unknown sponges Yellow Purple White Red and white Orange	-	36 1 13 96 237

). Of those, the following could be identified to the species level: Pseudoceratina/Aoilochroia crassa Hyatt, 1875, Geodia sp. Lamarck, 1815, Ircinia sp. Nardo, 1833, Agelas clathrodes Schmidt, 1870, and Halichondria sp. Fleming, 1828. There were also yellow, purple, white, red and white, and orange sponges that could not be positively identified. The most observed species were P. crassa, the unidentified red and white sponge, and the unidentified orange sponge. All of these sponges were found primarily in the encrusting and tubular form. Also, it is possible that some organisms identified as Geodia sp. were misidentified and are actually Siderastrea sp. and vice versa; identification was skewed by possible bleaching on the corals.

3.4.4. Rhodophyta

We observed that larger rock features such as boulders and outcrops were encrusted with algae in the crustose coralline and encrusting rhodophyte (Peyssonnelia sp. Decaisne, 1841) forms, the former being the dominant form of algae present, with observation in 2297 photos and the latter appearing in 349 photos (

Table 4. List of identified encrusted algae observed throughout the MTB study transect.

Rhodophyta
Species/Morphology	Appearances
Crustose coralline algae	2297
Hydrolithon sp.	579
Encrusting rhodophyte	349

). Another encrusting organism that was observed in 579 photos that may be algae was what was assumed to be Hydrolithon sp. Foslie, 1909.

3.4.5. Echinodermata

Crinoids, sea stars (Asteroidea spp. Blainville, 1830), and basket stars (Euryalida spp. Lamarck, 1816) were observed. Crinoids were the most frequently observed echinoderm, with presence in 1424 photos (

Table 5. List of Echinodermata observed throughout the MTB study transect.

Echinodermata
Species	Counts	Appearances
Sea star (Asteroidea spp.)	2
Basket stars (Euryalida spp.)	3
Crinoids (Crinoidea spp.)		1424
Unknown Echinodermata	4

). Similar to some corals, their dense assemblages, overlap between photos, and locations within crevices and on outcrop walls made accurate counts difficult. Counts within photos ranged from as few as one crinoid to as many as 8+, with the majority falling within the 2+ category; the + accounts for the possibility that there may be more, but lack of clarity makes it difficult to determine. The basket stars that were observed were attached to S. exserta. The sea stars that were observed were atop the sandy sediment portions of the study site.

3.4.6. Mollusca

Six whelks were observed throughout the study site (

Table 6. List of mollusks observed throughout the MTB study transect.

Mollusca
Species	Counts
Lightning whelk (Sinistrofulgur perversum)	6
Unknown mollusk	2

). Sedimentation and visibility factored into an inability to identify, but it is likely that they are lightning whelks (Sinistrofulgur perversum Linnaeus, 1758) based on relative size and shape. There were also two instances in which an organism appeared to be a mollusk, but both were unable to be identified further.

3.4.7. Infauna

Apart from the organisms we have listed, we also observed 371 instances of burrow holes within the study site, indicative of roughly 134 sets of burrow holes. The burrow holes always appeared with clusters of at least four holes within each set; some had over 10 holes. Lack of sediment grabs or cores made it impossible to accurately assign the organism(s) responsible for these holes. It is possible that they were squat lobsters (Eumunida spp. Smith, 1883). It should be noted that it is possible for overlapping counts to have occurred because most burrow holes were within the sandy sediment where there were no distinctive features by which to distinguish organisms.

3.4.8. Anthropogenic Debris

We also found evidence of anthropogenic pollution such as plastic cups and metal cans (Figure 7).

3.5. Organism Statistical Distribution

Cnidaria and Rhodophyta were the most frequently observed organisms; Mollusca and Chordata were the least frequently observed organisms. Although exact counts could not be made for each phylum, the number of appearances could be counted in order to estimate an organism percent composition at MTB as reported in the histogram in Figure 8. Of the 3345 photos taken, Cnidarians appeared in 2489 (74.41%), Rhodophyta appeared in 2313 (69.15%), Echinodermata appeared in 1449 (43.32%), Porifera appeared in 940 (28.10%), burrow holes (infauna) appeared in 370 (11.06%), Chordata appeared in 240 (7.17%), and Mollusca appeared in 25 (0.75%).

In Table 1, Table 2, Table 3, Table 4, Table 5 and Table 6, both identifiable and unidentifiable species are listed. Counts were included whenever possible; otherwise, only occurrences or appearances were recorded. Species richness calculations include only identifiable species, since unidentified organisms may belong to already recorded species and therefore cannot be considered distinct taxa, or because the number of unique species they represent is uncertain. Similarly, the Shannon Diversity Index was calculated using only identifiable species and should therefore be considered an under-approximation of the true diversity values.

3.6. Benthic Habitat Mapping

Table 7. Habitat class descriptions and benthic faunal assemblages and distributions derived from photo analysis.

Habitat Class		Seafloor Texture	Benthic Fauna	Distribution
	Sand	Only sand present (certain portions may have an exception where sandy mud/muddy sand is also present)	Infauna, echinoderms, mollusca, demersal fish	Widespread between, along, and beyond rock outcrops. Also present as a veneer atop certain portions of rock outcrops.
	Sand with boulders	Sand with boulders present	Infauna, echinoderms, mollusca, cnidarians, porifera demersal fish	Boulders are scattered throughout the sand portions or present at the interface between sand and rock outcrops
	Sand with coral fragments	Sand with coral fragments present	Infauna, cnidarian remnants, demersal fish	Coral fragments are found in piles in all places that sand is present (not specified on map)
	Sand with Gravel/ Cobbles	Sand with larger grain sizes present	Infauna, echinoderms, cnidarians, mollusca, demersal fish	Found around portions where rock outcrop and sand interfaces are as well as where boulders are also present (not specified on map)
	Rock Outcrop (Top)	Large rock outcrop where slope/rockface is not present to impact backscatter signal	All megabenthos categories except infauna	Found along the western portion of our transect at depths of about 63–64 m
	Rock Outcrop (Side)	Rockface of rock outcrop is predominant feature present, sand may be present at the interface between hard and soft terrain or weaving between rock outcrop features (as shown)	All megabenthos categories	Found along western portion of our transect

details the substrate classifications with their associated benthic fauna assemblages and distribution throughout the MTB transect. We recognized six different habitat classes: sand; sand with boulders; sand with coral fragments; sand with gravel/gobbles; rock outcrop (top); and rock outcrop (side), but only four consolidated substratum habitat classes were correlated with the backscatter data (Figure 9). These consolidated classes include sand, sand with larger grain sizes present (sand with boulders, gravel/cobbles, coral fragments), rock outcrop (side), and rock outcrop (top).

There is also the possibility that muddy sand or sandy mud was present in either sand class distributions. The rock outcrop features also had sand present in some cases such as a veneer on the top or between rock outcrop structures. The substratum classes were then combined with the benthic fauna distribution to derive the MTB BHM (Figure 10). Despite some discrepancies, we found a good agreement between the benthic classes and the acoustic predictor.

4. Discussion

The BHM derived for the 60 × 60 m photomosaic area was applied to predict benthic habitats across the entire dome by integrating acoustic classes with geological sample calibration points collected outside the photomosaic boundary. The strong agreement between substratum classes and control points suggests that the observed benthic faunal assemblage is representative of the broader dome structure.

MTB is characterized by a network of rock outcrops and boulders interspersed within a predominantly sandy matrix. The relief ranges from 1 to 4 m, with areas surrounding outcrops typically spanning 1–2 m. Surface rugosity varied from bare rock substrate—occasionally veneered with sand or coral rubble—to surfaces encrusted with algae, sponges, or corals. The transition zones between outcrops and sand are associated with coarser grain sizes, including gravels and cobbles, which also occur around scattered boulder features distributed across otherwise flat, sandy terrain. Although algae, corals, echinoderms, and porifera dominate the assemblage, demersal fishes, while constituting a minor proportion of observed organisms, were the most species-rich taxon, with 17 species identified.

Scattered boulders and outcrops provide essential attachment substrates for large soft corals and sea whips [7], and represent the preferred habitat for S. exserta, M. pendula, Stichopathes spp., Antipathes spp., and Tanacetipathes spp. While Stichopathes spp. and Antipathes spp. occurred in dense assemblages, the remaining larger soft corals were predominantly solitary. These assemblage patterns enhance structural complexity and refuge availability, and the vertical relief generated by these species over soft substrata contributes to elevated biodiversity in high-relief zones [22]. All identified coral species were recorded within their expected depth range, biogeographic setting, and community structure type. The occurrence of Porites spp. within rubble piles—a characteristically shallow reef species—may reflect historical shifts in community composition in response to sea-level change.

Low-relief reef areas subject to sand inundation are known to be dominated by sponges interspersed with coralline red algae [23], while structurally complex surfaces support dense assemblages of erect sponges and encrusting calcareous red algae [1]. The encrusting organisms and sponge communities observed on MTB outcrops are consistent with these established habitat associations. Demospongiae, the most diverse sponge class, encompasses a wide range of growth forms and serves as an important food source for associated fauna. The majority of algae identified at MTB was crustose coralline algae (CCA), a reef-building taxon predominantly found in deep, low-relief habitats [24]. The significantly higher algal cover observed at upper mesophotic sites relative to shallow reefs supports the relevance of MTB’s mid-mesophotic position, where CCA is not in direct competition with photosynthetic coral species [16].

Fish community structure at the nearby FGBNMS is positively correlated with both depth and habitat relief [16]. Our findings are consistent with these patterns and include the presence of the invasive red lionfish, P. volitans, recorded at low but potentially increasing densities in the GOM [16]. The proliferation of lionfish at MTB and adjacent reef structures underscores the importance of understanding regional reef connectivity. Although fish species richness typically correlates with live coral cover, the abundance of crevices observed throughout MTB likely supports a comparatively high diversity of cryptic species [3]. Equipment avoidance and attraction is also a limiting factor in using photographic data to characterize fish assemblages. It is therefore probable that species richness at this site exceeds observed values. Fish species were similarly recorded within their expected depth, biogeographic, and community structure context.

MTB exhibits the varied faunal assemblages characteristic of mesophotic reef ecosystems throughout the GOM and other ocean basins [1,4,14,16], supporting the broader observation that diverse reef communities can persist at depth. Based on taxonomic overlap with both the Pinnacle Trend and FGBNMS, and given MTB’s geographic proximity to the Pinnacle Trend, the faunal assemblage observed here suggests that MTB likely represents an extension of the Pinnacle Trend.

Taxa shared between the FGBNMS and MTB include the majority of identified fishes, several coral genera (Antipathes sp., A. atlantica, A. pedata, Siderastrea sp., Stichopathes sp.), all identified sponges (A. clathrodes, Halichondria sp., Ircinia sp., Geodia sp., P./A. crassa), Rhodophyta, and the echinoderm Crinoidea spp. Species shared with the Pinnacle Reef Trend [22] include additional coral and fish taxa. Conversely, species such as Tanacetipathes spp., Asteroidea spp., Echinoidea sp., Euryalina spp., and B. perversum, though absent from Pinnacle Trend and FGBNMS records, have been documented at other mesophotic reefs within the GOM (e.g., Sticky Grounds). The differential distribution of organisms across these sites may reflect endemism within the northern GOM or varying degrees of ecological connectivity among reef systems. It is important to note, however, that the complexity of the faunal assemblage and the large overlap among these systems make MTB not a simple ‘rest stop’ for larvae, as oil rigs [25] and/or shipwrecks [26] have been found to function, but rather a structural biological hotspot bridging the two regions.

Structural connectivity within the FGBNMS and surrounding deep-water areas provides the foundation for biological and ecological linkages among the more than one hundred reef structures along the TX and LA coastlines [15]. Depending on the extent of regional “habitat highways,” there exists potential for significant inter-site movement of species and individuals [15]. Should MTB lack direct connectivity with the FGBNMS, connectivity with other hardground banks in the NW GOM—many of which support extensive benthic invertebrate and reef fish communities [14,16]—remains plausible. Community composition data from MTB could be used to model assemblages across northern Gulf banks and, when integrated with genetic data, could inform assessments of regional connectivity.

The degree of connectivity between shallow and mesophotic coral reefs remains poorly understood. Genetic linkages between these zones could render mesophotic reefs a critical source of propagules for restoring shallow populations depleted by natural and anthropogenic disturbances [27]. However, localized larval retention driven by regional circulation patterns may promote high levels of endemism, and such patterns, combined with species longevity and diversity, may position mesophotic coral reefs as significant centers of speciation [28]. As potential extensions of adjacent shallow reefs, MCEs share some species with shallower counterparts while also harboring unique taxa, suggesting that conspecifics across these habitats may constitute components of larger metapopulations [29]. Limited larval dispersal over broad spatial scales, potentially linked to reproductive strategy [30], challenges the characterization of coral reefs as open systems. A comprehensive understanding of the multiple dimensions of connectivity is therefore essential for addressing existing knowledge gaps and informing effective conservation strategies.

5. Conclusions

The complex relationship between seafloor geology and biology underscores the need for interdisciplinary approaches in marine research. Identifying surrogate relationships is essential for advancing habitat mapping methodologies and improving the prediction and preservation of marine biodiversity. Continued investigation of the associations between physical properties and processes and biological communities will be necessary to further determine these linkages. This study at MTB contributes to that understanding by providing insight into these relationships within the northern GOM mesophotic zone and the potential for large-scale ecosystem connectivity across this region based on shared geomorphology and biodiversity patterns. At both the site-specific scale of MTB and the broader scale of the northern GOM, predictive habitat maps offer significant practical value—optimizing the allocation of ground-truthing efforts during sampling campaigns while reducing associated time and costs. Such tools could also prove to be essential for MPA planning and in supporting the assessment and monitoring frameworks required for effective marine conservation. Protection is particularly critical for reef communities harboring sensitive species that may be incapable of full recovery if exposure to degradation is not proactively prevented and managed.