Received: 26 March 2018 | Revised: 27 July 2018 | Accepted: 10 August 2018 DOI: 10.1111/ddi.12846 BIODIVERSITY REVIEW Mesophotic ecosystems: Distribution, impacts and conservation in the South Atlantic Marcelo de Oliveira Soares1,2 | Tallita Cruz Lopes Tavares1 | Pedro Bastos de Macêdo Carneiro3 1 Instituto de Ciências do Mar (Labomar), Universidade Federal do Ceará Abstract (UFC), Fortaleza, Brazil Aim: This study reviews recent research on the South Atlantic Mesophotic ecosys- 2 Institut de Ciència i Tecnologia Ambientals tems (MEs) and the pressures threatening them, and offers suggestions for their (ICTA), Universitat Autònoma de Barcelona (UAB), Barcelona, Spain management and conservation. 3 Universidade Federal do Piauí (UFPI), Location: The South Atlantic Ocean. Campus Parnaíba, Parnaíba, Brazil Methods: A comprehensive compilation of the scientific literature was performed to Correspondence examine the distribution, human impacts and conservation status of the South Marcelo de Oliveira Soares, Instituto de Atlantic MEs. Ciências do Mar (Labomar), Universidade Federal do Ceará (UFC), Fortaleza, Brazil. Results: Our review indicated that the South Atlantic Ocean (SAO) is one of the major Email: marcelosoares@ufc.br MEs areas in the world’s oceans. The South Atlantic MEs are composed of a mosaic Funding information of distinct seascapes, mainly rhodolith beds, mesophotic reefs (i.e., rocky and bio- Fundação Cearense de Apoio ao genic) and marine animal forests (e.g., sponge aggregations, octocoral and black coral Desenvolvimento Científico e Tecnológico, Grant/Award Number: PR2-0101- forests) that occur along the East South American and West African coasts, sea- 00008.01.00/15; Conselho Nacional de mounts and oceanic islands. Throughout the SAO, the distinct seascapes of MEs are Desenvolvimento Científico e Tecnológico, Grant/Award Number: 233808/2014-0 and usually formed on the middle and outer continental shelves, shelf-­edge, seamounts, 307061/2017 submarine canyons, incised valleys and paleochannels, reef structures and insular Editor: Maria Beger shelves. We highlighted sea temperature anomalies, ocean acidification, extreme floods and droughts, fisheries, invasive species, marine debris, mining, and oil and gas exploitation as major threats to these ecosystems. Main conclusions: Given the threats to the South Atlantic MEs, growing human pres- sures may degrade these ecosystems in the next years and undermine their unique biodiversity as well as their potential to provide connectivity between regions and depths. Our review revealed the existence of some extensive and unprotected for- mations, which urgently demand in-­depth investigations and conservation action. KEYWORDS climate change, coral reef, deep-sea refugia, marine biogeography, marine conservation, rhodolith bed 1 | I NTRO D U C TI O N deeper oceanic zones, providing a more detailed picture of the hid- den biodiversity in mesophotic ecosystems (MEs; Loya, Eyal, Treibitz, Developments in the marine sciences, especially in terms of method- Lesser, & Appeldoorn, 2016). These ecosystems are characterized by ology and equipment, have allowed the exploration of progressively the presence of light-­dependent corals and associated species (e.g., Diversity and Distributions. 2019;25:255–268. wileyonlinelibrary.com/journal/ddi © 2018 John Wiley & Sons Ltd | 255 256 | SOARES et al. algae, sponges and fishes), constituting complex communities at of physiological and reproductive modifications in deep-­water pop- mesophotic depths (i.e., between 30 and 150 m; Hinderstein et al., ulations that could prevent a recolonization of shallow habitats 2010). Studies on MEs have revealed that these deep-­water hab- (Shlesinger, Grinblat, Rapuano, Amit, & Loya, 2018; Smith et al., itats, when compared to their shallow-­water counterparts, usually 2016). Additionally, Semmler et al. (2017) observed different refu- display several characteristics in terms of health status (Tornabene, gium potential between upper (where ~30%–45% of total species van Tassell, Robertson, & Baldwin, 2016), coral coverage (Hoeksema, were shared with shallow habitats) and mid-­lower (~15%–25%) meso- Bongaerts, & Baldwin, 2017) and fish endemism (Lindfield, Harvey, photic ecosystems. Therefore, new studies are needed to assess the Halford, & Mcllwain, 2016). Often, for example, there seems to be a limitations and applicability of this and other ecological hypotheses turnover of species and a reduction in biodiversity with depth, but (i.e., beta diversity, turnover and endemism) about MEs (Bongaerts some taxa seem to peak at the mesophotic zone (Semmler, Hoot, & et al., 2017; Rocha et al., 2018). Moreover, particular attention should Reaka, 2017). On account of these characteristics, most researchers be given to regions where these ecosystems are poorly studied agree that MEs could act as marine biodiversity reservoirs, and thus (Turner et al., 2017), as in the case of the South Atlantic Ocean (SAO). should be more thoroughly studied (Kahng, Copus, & Wagner, 2017; In the present paper, a comprehensive compilation of the sci- Loya et al., 2016). entific literature (Supporting Information Appendix S1) was carried The good health status of MEs is often linked to their depth and out to examine the distribution, human impacts and conservation distance from the coast, which seem to reduce the impact of direct status of the South Atlantic MEs. For this review, we have con- anthropogenic pressures (Glynn, 1996; Kahng et al., 2017). However, sidered benthic communities located at mesophotic depths on a growing body of evidence suggests that even isolated formations the South Atlantic Ocean (SAO), between ∼0.00°S and 23.00°S may have been affected by pollution, sedimentation, oil spills, (Figure 1). This area contains many different seascapes, including bottom-­contact fishing and bio-­invasion (Andradi-­Brown et al., rhodolith beds, rocky reefs, biogenic reefs, sponge bottoms, and 2017; Baker, Puglise, & Harris, 2016; Rocha et al., 2018). Moreover, octocoral or black coral forests, to which we have attributed the Rocha et al. (2018) show that mesophotic reefs are ecologically dis- general term “mesophotic ecosystems.” The present study reviews tinct, threatened and in as much need of protection as shallow reefs. recent advances in knowledge about the geographical distribu- Therefore, there is scientific interest in understanding if and how tion of MEs in the SAO basin and discusses major anthropogenic MEs may act as refuge areas and ecological corridors, where marine pressures on them to provide useful and synthetic information for species would be kept away from local and global stressors, such as science, management and conservation. Finally, we suggest some thermal anomalies and pollution, which have been affecting shallow-­ urgent conservation measures to protect the tropical biodiversity water ecosystems (Turner, Babcock, & Kendrick, 2017). and ecosystem services of South Atlantic MEs in consideration of MEs (mainly the upper zone, 30–60 m depth) are frequently re- their role in a changing ocean. garded as extensions of shallow-­water reefs since both ecosystems share some species (Kahng et al., 2017). This suggests that meso- photic assemblages may act as deep-­sea ecological corridors provid- 2 | S O U TH ATL A NTI C M E S O PH OTI C ing large-­scale connectivity (i.e., stepping stones) between species ECOS YS TE M S (Rocha, 2003) and that they might be able to reseed or replenish populations from shallow endangered habitats. The latter possi- The SAO shelters distinctive mesophotic ecosystems. This ocean bility is referred to as the deep-­sea refugia hypothesis (Bongaerts, basin does share some species with neighbouring areas, such as Ridgway, Sampayo, & Hoegh Guldberg, 2010; Bongaerts et al., 2017; the Caribbean Sea, but this flux seems to be limited by biogeo- Glynn, 1996; Riegl & Piller, 2003; Smith et al., 2016), and it highlights graphical barriers, isolation by distance and differences in specia- the potential importance of these habitats for the future of marine tion and extinction rates (Leão et al., 2016; Pinheiro et al., 2018; conservation and management. Soares, Lotufo, et al., 2017). For example, whereas the Caribbean Despite their potential ecological importance, basic knowl- sustains dozens of scleractinian coral species, in the SW Atlantic edge about many aspects of MEs, such as their unique biodiversity there are only 23 of them, many of which are endemic (Leão et al., (Pinheiro et al., 2017; Rocha et al., 2018), susceptibility to human 2016), and so far, only 15 coral species have been recorded on the impacts and conservation status (Turner et al., 2017), is still lim- West African coast (Laborel, 1974; Moses, Helmle, Swart, Dodge, ited. Most studies have been conducted in the Red Sea (Shoham & Merino, 2003). Additionally, on both east and west sides of the & Benayahu, 2017), the Caribbean Sea and the Indo-­Pacific region SAO, the construction of biogenic reefs is at its most discontinu- (Hinderstein et al., 2010; Loya et al., 2016). Thus, large parts of the ous (Hopley, 2011). Accordingly, the contribution of scleractinian ocean have not yet been studied. Even in frequently surveyed areas, corals to MEs seems lower in the SAO compared to the Caribbean the limited number of studies makes it difficult to create hypotheses Sea and Indo-­Pacific. On the other hand, sponge bottoms, rhodo- to test the structure and functioning of MEs (Kahng et al., 2017). For lith beds and other hard-­bottom environments, such as rocky and example, although the deep-­sea refugia hypothesis has been sup- biogenic reefs (built mainly by coralline algae, scleractinian corals ported by statistical models and studies on genetic and community and bryozoans), which are structurally similar to coral-­dominated composition (Holstein, Paris, Vaz, & Smith, 2016), there is evidence ecosystems, are common in this ocean basin. SOARES et al. | 257 F I G U R E 1 Distribution of currently known mesophotic ecosystems, within biogeographical provinces and ecoregions (sensu Spalding et al., 2007), in the South Atlantic Ocean. The names on the map show relevant ecoregions. Legends: NE Brazil (Northeastern Brazil), E Brazil (Eastern Brazil) and VTC (Vitória-­Trindade Chain) [Colour figure can be viewed at wileyonlinelibrary.com] Many of these benthic ecosystems show three-­dimensional com- Soares, Lotufo, et al., 2017). Additionally, extensive coverage of plexity and sustain high marine biodiversity while providing import- sponges can be found atop submerged rocky outcrops (Soares, ant ecosystem services and performing functions, such as carbon Rossi, Martins, & Carneiro, 2017). sinks, nursery grounds and refuge areas. Rossi (2013), expanding the Besides the animal-­d ominated ecosystems, rhodolith beds, concept of reef, designated the assemblages dominated by sponges, while apparently absent from the African coast (Amado-­F ilho, cnidarians (hydrocorals, black corals, octocorals, scleractinian cor- Bahia, Pereira-­F ilho, & Longo, 2017), are a ubiquitous feature als), bryozoans and ascidians, organisms which are considered to along the Southwestern Atlantic Ocean (Horta et al., 2016). Where be “ecosystem engineers” based on their ability to modify sea-­floor they occur, these extensive algal build-­ups constitute an important habitats, as marine animal forests. These benthic species are primar- component of the MEs, with the coalescence of rhodoliths having ily responsible for the structure of animal-­dominated ecosystems, been hypothesized as a possible mechanism for the formation of especially in the lower mesophotic zone, where there is less light and carbonate reefs (Amado-­F ilho et al., 2016). In fact, many Brazilian sessile suspension feeders are abundant. biogenic reefs are primarily formed by the growth of calcareous In this regard, sponge bottoms (also known as sponge aggrega- red algae, bryozoans, milleporids and scleractinian corals (Bastos tions) are among the most important animal forests worldwide, oc- et al., 2018; Leão et al., 2016). curring mainly under suboptimal conditions for coral growth, such as Throughout the SAO, the distinct seascapes of MEs (rocky and turbid waters with sediment resuspension (Maldonado et al., 2017; biogenic reefs, sponge bottoms, rhodolith beds, octocoral and black Soares, Lotufo, et al., 2017). Among the SAO mesophotic ecosys- coral forests) are usually formed on the middle and outer continen- tems, there are many of these aggregations, which are composed tal shelves, shelf-­edge, seamounts, submarine canyons, incised val- of massive demosponges in the forms of balls, tubes, fans, vases leys and paleochannels, reef structures (such as reef walls and reef or branches, such as Agelas, Aplysina, Xestospongia, Callyspongia, banks) and insular shelves. Within these geomorphological settings, Clathria, Monanchora, Oceanapia, and Geodia (Moura et al., 2016; MEs often seem to be associated with drowned reefs or beachrock 258 | SOARES et al. lines, which were apparently formed in past periods of marine re- Thirty-­eight coral species (octocorals, scleractinians, hydro- gression, especially during the Last Glacial Maximum, when the sea corals and black corals) were later reported in this area (Cordeiro, level was approximately 130 m below the present, essentially ex- Neves, Rosa-­F ilho, & Pérez, 2015), along with a complex mosaic posing the entire continental shelf (Clapperton, 1993). As the ocean of carbonate structures and rhodolith beds (Moura et al., 2016). began to rise again, these now drowned reefs could not cope with Moura et al. (2016), considering the geographical distribution, di- the ascending sea level (Camargo, Araujo, Ferreira, & Maida, 2015; vided the Amazonian ME into three sectors, based on the degree of Silva, Gomes, & Vital, 2018); however, they developed into topo- exposure to the Amazon River, namely northern (under permanent graphic structures that offered suitable habitats for the establish- estuarine plume influence), central (seasonal plume influence) and ment of ecosystem engineers (e.g., algae, sponges and cnidarians) in southern (intermittent riverine influence), the latter constituting the mesophotic zone. the boundary with the Tropical Southwestern Atlantic Province. Finally, we have noticed the occurrence, on both sides of the Francini-­F ilho et al. (2018) suggested that these MEs comprise a SAO, of reefs at even greater depths (>150 m depth), reaching significant diversity of environments, including rhodolith beds, below the mesophotic zone. These deeper formations have also sponge grounds and reef structures, such as biogenic walls and received little scientific attention (Soares, Lotufo, et al., 2017) platforms. and are probably distinct and ecologically apart from the shallow These Amazonian MEs (the Great Amazon Reef sensu Francini-­ and mesophotic ecosystems (Hovland, 2008), being formed by Filho et al., 2018) are distributed across a large area (56,000 km²) ecosystem engineers, such as corals Lophelia pertusa, Madrepora of the North Brazil Shelf, between the shelf-­e dge and the upper oculata and Solenosmilia variabilis (Cordeiro, Kitahara, & Amaral, slope (70–220 m depth; Francini-­F ilho et al., 2018). Despite the 2012; Kitahara, 2007). It is outside the scope of this study to variability of seascapes, this ecosystem has a clear bathymetric discuss these deeper marine animal forests; however, a review gradient: Rhodolith beds and algal frameworks dominate the of their distribution and conservation status is urgently needed, shallowest sector (70–180 m), but in the deepest sector (180– since they are also vulnerable to human pressures, such as those 220 m), marine animal forests (composed of black corals, barrel caused by the fishing industry (Kitahara, 2009; Soares, Lotufo, sponges and octocorals) and associated fishes (e.g., butterflyfish) et al., 2017). are more common. Moreover, Vale et al. (2018) characterized the structure and composition of rhodoliths between water depths of 23 and 120 m. Important mesophotic ecosystem engineers, 3 | R EC E NT A DVA N C E S I N K N OW LE D G E such as bryozoans, coralline algae and encrusting foraminifera A B O U T TH E S O U TH ATL A NTI C M E s built these rhodolith beds while under the influence of the Amazon river. The South Atlantic MEs are distributed in distinct marine ecore- Overall, the reef biota (algae, sponges, cnidarians and fishes) gions along the continental shelves of South America and Africa recorded in the Great Amazon Reef inhabit a wide range of depths and on seamounts and oceanic islands (Figure 1). They are found (Moura et al., 2016). Red algae were the predominant benthic plant at the Amazon River mouth, Northeastern Brazil, Abrolhos Bank, group, with 25 species. The sponge assemblage comprised 61 spe- Eastern Brazil, Gulf of Guinea, seamounts (such as the Vitória-­ cies and was dominated by massive forms (Francini-­F ilho et al., Trindade Chain), oceanic islands (Trindade and Martin Vaz, Saint 2018; Moura et al., 2016). Two black coral species, Antipathes fur‐ Helena and Ascension islands, Fernando de Noronha Archipelago, cata and Tanacetipathes tanacetum, typical of mesophotic reefs, Saint Peter and Saint Paul Archipelago) and on the only atoll in were detected in the deeper zones. The most common sclerac- the South Atlantic (Rocas Atoll; Supporting Information Appendix tinian corals included small-­sized colonies, massive species and S2), and they comprise four main biogeographical provinces (sensu branching forms, such as Meandrina braziliensis, Montastraea cav‐ Spalding et al., 2007), namely the North Brazil Shelf, Tropical ernosa, Madracis decactis, Agaricia spp., Scolymia wellsii, Millepora Southwestern Atlantic, Saint Helena and Ascension islands and cf. alcicornis and Favia gravida. Overall, scleractinians comprised the Gulf of Guinea (Figure 1). only 12 species, with Octocorallia being the most diverse cnidarian group with 26 species. Octocoral animal forests (sensu Rossi, 2013) provide a unique seascape and probably form a “canopy” in this 3.1 | The North Brazil Shelf mesophotic zone. Finally, Moura et al. (2016) recorded 73 reef fish The North Brazil Shelf (sensu Spalding et al., 2007) is a poorly under- species in this ME on the North Brazil Shelf, most of which were stood biogeographical province in terms of MEs. The first evidence carnivores (86%). Also, aggregations of threatened and commer- for the occurrence of an extensive mesophotic ecosystem in this cially important fishes, such as Lutjanus purpureus and Hyporthodus province was provided by Collette and Rutzler (1977), who reported niveatus, have been detected using video-­surveys of the region the existence, next to the mouth of the Amazon River, of diverse (Francini-­F ilho et al., 2018). Considering the importance of these fish assemblages associated with sponge aggregations that possibly MEs, Francini-­F ilho et al. (2018) suggested that this ecosystem acted as a deep-­water ecological corridor between the Caribbean should be urgently included in a network of marine protected areas and Brazil (Rocha, 2003). (MPAs). SOARES et al. | 259 important component of the reefs in the northeastern Brazil. Two 3.2 | Tropical Southwestern Atlantic Province species (Stomozoa gigantea and Eudistoma saldanhai) stand out in the seascape because of their size and abundance. 3.2.1 | The Northeastern Brazil ecoregion The coral and fish compositions in the Northeastern equatorial In Northeastern Brazil, the outer continental shelves are relatively Brazil MEs between 35 and 37 m depths were similar to that reported shallower (60–80 m deep) than elsewhere along the South American for nearby shallow reefs (Freitas and Lotufo, 2015; Soares, Davis, Paiva, coast and are covered by warm and oligotrophic waters (Silva et al., et al., 2018). This finding suggests possible limited vertical connectivity 2018). Additionally, there are quite a few records of deep beachrock between the shallow and upper mesophotic zones (Morais & Santos, lines in this area, which are commonly seen as relics of ancient 2018) in these reef formations. Additionally, these equatorial MEs are shorelines (Camargo et al., 2015). These oceanographic and geologi- positioned approximately halfway between the Eastern Atlantic and cal characteristics provide a suitable habitat for the occurrence of the Amazonian reefs, and they could offer a more extensive east–west MEs, mainly sponge bottoms, rhodolith beds, and rocky and biogenic connection than the one provided by the existing shallow-­water coral-­ reefs associated with the continental shelf (Camargo et al., 2015; poor formations. However, the absence of genetic studies in this region Eduardo et al., 2018; Olavo, Costa, Martins, & Ferreira, 2011; Silva prevents precise conclusions about these hypotheses. et al., 2018; Soares, Davis, Paiva, & Carneiro, 2018). However, there The eastern sector of the Northeastern Brazil ecoregion also currently seem to be many large gaps in the occurrence of MEs along sustains several MEs. Sponge bottoms in mesophotic depths off the continental margin of Northeastern Brazil, which we attribute the states of Rio Grande do Norte, Paraíba, Pernambuco and to the lack of benthic habitat mapping in this area. The currently Alagoas represent an important habitat for species, including en- available data, for example, came mostly from the upper mesophotic demic sponge-­dwelling fishes. Moreover, although not necessarily zone (30–70 m; Camargo et al., 2015; Eduardo et al., 2018; Feitoza, in direct association with the sponges, angelfish, surgeonfish, but- Rosa, & Rocha, 2005; Morais & Santos, 2018; Rocha, Rosa, & Feitoza, terflyfish, parrotfish and triggerfish were found inhabiting these 2000). We hypothesized that the Amazonian ME is not restricted to formations (Rocha et al., 2000). Feitoza et al. (2005) analysed reef the North Brazil Shelf. This ecosystem is probably semicontinuous fish off this coast in the upper mesophotic depths (35–70 m depth). on the Northeastern Brazil ecoregion (especially on the equatorial A total of 158 fish species belonging to 49 families was recorded, margin) due to occurrence of similar seascapes and biodiversity, the and the most abundant families, in order of importance, were absence of biogeographical barriers and the connectivity provided Carangidae, Gobiidae, Lutjanidae, Labridae, Serranidae, Haemulidae by the North Brazil current. Therefore, we hypothesized the exist- and Scaridae. These fishes occurred in shelf-­edge reefs composed ence of a large tropical mesophotic ecosystem in this ecoregion, es- of sandstone outcrops dominated by algae and massive sponges pecially on the outer shelf, shelf-­edge and in incised valleys. that were also characterized by the occurrence of living corals This ecoregion (sensu Spalding, Ravilious, & Green, 2001) com- (such as Montastraea cavernosa, Porites branneri, Siderastrea stellata, prises the tropical coast of Northeastern Brazil (Ceará, Rio Grande Meandrina braziliensis). The main finding was that these MEs might do Norte, Paraíba, Pernambuco, Sergipe and Alagoas states) and may indeed function as a corridor for fish populations between Brazil and be divided into two distinct sectors based on the orientation of the the Caribbean because of the presence of several species previously continental margin: I) the equatorial sector (comprising Ceará and considered to have disjunct or anti-­equatorial distributions (Feitoza Rio Grande do Norte states), and II) the eastern sector (encompass- et al., 2005). Moreover, Eduardo et al. (2018) described elevated fish ing Paraíba, Pernambuco, Sergipe and Alagoas Brazilian states). diversity in specific seascapes along the outer shelf, but particularly The occurrence of MEs in the Brazilian equatorial margin, in in the upper mesophotic zone (40–60 m). This region, characterized an area not under Amazon River influence, has been recently de- by macroalgae, sponge bottoms and coralline algae formations, con- scribed (Almeida, Vital, & Gomes, 2015; Silva et al., 2018; Soares, centrates fishing resources and benthic biodiversity. Davis, Paiva, et al., 2018). Freitas and Lotufo (2015) described reef Morais and Santos (2018) described coral communities along a fishes in a mesophotic reef (36 m), and Soares, Davis, Paiva, et al. bathymetric gradient (3–61 m) in reefs off the Paraíba state coast. (2018) described reef fishes and scleractinian corals at depths of The authors reported that coral assemblages presented twice as 37 m. Both articles reported the presence of a mosaic of benthic more gamma diversity in shallow than in MEs (13 vs. seven species), habitats, composed of carbonate sediment deposits and reef out- and only three out of 17 species occurred along the entire gradient. crops located in incised shelf valleys off the Ceará state coast. Two On the other hand, the alpha diversity was similar between shallow resilient scleractinian coral species were mainly found on these reef and deep reefs. The authors concluded a limited potential of MEs to outcrops, namely Siderastrea stellata and Montastraea cavernosa. serve as refuges, probably only for the two dominant coral species Silva et al. (2018) provided important geological information on (Siderastrea stellata and Montastreaea cavernosa). The mesophotic shelf-­edge reefs located in the Açu incised valley off the Rio Grande reefs off Northeastern Brazil also sustain endemic species. On the do Norte equatorial coast. The described mesophotic reefs occurred Pernambuco state coast, Pereira, Santos, Lippi, and Silva (2016) in a depth range of 30–55 m and were covered by red algae, scler- studied patterns of parrotfish (endemic Scarus zelindae) ontogenetic actinian corals (e.g., Montastraea cavernosa) and sponges. In ad- foraging activity and feeding selectivity in mesophotic reefs (30– dition, Soares, Lotufo, et al. (2017) indicated that ascidians are an 35 m depth) due to the presence of different fish life-­phases. The 260 | SOARES et al. preservation of topographic features off Pernambuco state coast, ecosystems in the region are significantly threatened by human ac- such as wave-­cut erosion steps, shelf valleys and canyons results tivities. For the MEs that are not yet legally protected, particularly in greater biodiversity and therefore a greater potential for MEs those that have been adequately described (such as the deeper reefs (30–50 m depth), especially on drowned reefs and beachrock lines and rhodolith beds in Eastern Brazil), it is recommended that they (Camargo et al., 2015). should be given protection as soon as possible by establishing new conservation areas or expanding existing ones. For example, on the Abrolhos Bank, the marine protected areas (MPAs) do not cover 3.2.2 | The Eastern Brazil ecoregion important and geomorphologically diverse mesophotic habitats One of the most extensive reef complexes in the SAO is located on (Moura et al., 2013). the Abrolhos Bank, in the Eastern Brazil ecoregion (sensu Spalding et al., 2007). The occurrence of reefs and rhodolith beds has been 3.2.3 | Oceanic islands and seamounts reported at depths from 30 to 93 m (Bastos et al., 2013; Moura et al., 2013) and 40 to 200 m (Olavo et al., 2011). Geomorphologically, Several islands and seamounts occur off the continental shelf and these reefs are described as banks, paleochannels and submerged constitute important biodiversity hotspots within the Tropical SW pinnacles as well as coalescent structures with sinkhole-­like depres- Atlantic. They include the archipelagos of Fernando de Noronha and sions (cup-­shaped), which are known as “buracas.” These sinkhole-­ Saint Peter and Saint Paul (SPSPA), the Rocas Atoll and the Vitória-­ like features are novel and unusual (Bastos et al., 2013) and may Trindade Seamount Chain (VTC). These formations vary in size (the enhance biomass and productivity. These MEs are often described largest is Fernando de Noronha) and distance from the mainland (the as drowned reefs, with walls dominated by encrusting coralline algae most remote is SPSPA, located on the Mid-­Atlantic Ridge). These and low coral coverage. In the mesophotic reef pinnacles, M. caver‐ islands and seamounts may provide insights into the evolutionary nosa is the dominant species, and there are also rare occurrences of history, ecology of marine taxa, and seascape-­wide connectivity several species from different genera, such as Siderastrea, Agaricia, (Pinheiro et al., 2017). For example, Peluso et al. (2018) detected Porites, Madracis, Favia and Scolymia, as well as the black corals migration and connectivity of the reef-­building coral M. hispida be- Cirripathes and Antipathes (Moura et al., 2013). tween the Eastern Brazil and oceanic islands. The MEs may provide a Shelf-­edge reefs (40–200 m) have been recorded across the more continuous area of suitable habitat for this coral, favouring the Eastern Brazil, from Salvador city, across the Abrolhos Bank to the maintenance of gene flow between the Trindade Island, Rocas Atoll southernmost region (Espírito Santo and Rio de Janeiro coast). In and Fernando de Noronha Archipelago through the oceanic currents this region, fish species from the families Serranidae, Lutjanidae, (Peluso et al., 2018). Malacanthidae, Muraenidae, Sparidae, Balistidae, Carangidae, The VTC extends approximately 1,200 km offshore from the Haemulidae, Scorpaenidae and Priacanthidae were observed to be Brazilian continental shelf, from the Vitória seamount to the iso- associated with marine hard-­bottom environments at mesophotic lated oceanic islands of Trindade and Martin Vaz. The VTC consists depths (Olavo et al., 2011). of 11 seamounts with summits reaching euphotic and mesophotic The Abrolhos Bank encompasses the largest known continuous zones (~10–110 m depth) (Pereira-­F ilho et al., 2011). MEs found on rhodolith bed worldwide, which occupies an area of approximately the VTC are a mosaic of habitats dominated by macroalgae, includ- 20,900 km2. Rhodolith-­forming calcareous red algae are ubiquitous ing crustose coralline algae (CCA) at approximately 40 m deep, and along the eastern South American coast, but, like biogenic reefs, fleshy species on the rhodolith beds below 50 m deep (Meirelles they seem to thrive around the Abrolhos bank, apparently becom- et al., 2015). The scleractinian corals Siderastrea spp., M. cavernosa ing the foremost hard-­bottom feature in the mesophotic zone of and M. hispida, along with sponges, comprise the main benthic sus- the eastern Brazil ecoregion (Amado-­Filho et al., 2017; Horta et al., pension feeders in these MEs (Pinheiro et al., 2017). With a total 2016). Among the rhodoliths, some reef-­building coral species could of 211 species recorded on the seamounts (Pinheiro et al., 2015), be found (such as Mussismilia hispida, M. cavernosa and Siderastrea the structural reefs shelter richer and more abundant fauna than spp.), but they usually achieved low coverage (Amado-­Filho, Moura, the surrounding flatbeds. The fish fauna has a composition partly et al., 2012). Simon et al. (2016) reported 74 fish species, including similar to that of coastal reefs, but also shares certain character- new species and new records for the SAO. The authors also found istics, such as the abundance of planktivores, with Fernando de threatened species in these MEs, highlighting the importance of Noronha and Rocas Atoll (Pinheiro, Ferreira, Joyeux, Santos, & these habitats for conservation action. Moreover, due to its sheer Horta, 2011). size, this rhodolith bank may produce significant amounts of CaCO3, The insular complex formed by the Rocas Atoll and Fernando de suggesting that this area is an important area for calcium carbonate Noronha Archipelago represents a significant portion of the island deposition in the oceans (Amado-­Filho et al., 2017). surface in the SAO. The productive oceanic waters here are import- Concerning conservation of this large and rich coral complex, ant for the reproduction and feeding of many taxa (UNESCO, 2017), Francini-­Filho and Moura (2008), found that only 2% of the Abrolhos and they were designated a Natural World Heritage Site in 2001. Bank was designated as a “no-­t ake” area and that implementation They are also protected by MPAs (Biological Reserve in Rocas Atoll, was inadequate even for this small fraction. As a result, the fragile and Marine National Park and Environmental Protected Area in SOARES et al. | 261 Fernando de Noronha Archipelago). It is noteworthy that the Rocas, Irving (1989) described benthic communities of Ascension located at the top of a chain of seamounts, which are 267 km off the Island and reported the occurrence of the corals Astrangia solitaria, coast of Brazil and at 4,000 m depth, is the sole atoll in the South Madracis decactis and Siderastrea radians (Irving, 2013). In addition, Atlantic (Leão et al., 2016). This atoll is approximately 150 km west Zibrowius, Wirtz, Nunes, Hoeksema, & Benzoni (2017) detected the of the Fernando de Noronha Archipelago, with which it shares some presence of the coral Cladocora debilis in the lower mesophotic zone endemic species (Soares, Lotufo, et al., 2017). MEs in the Rocas Atoll (72 m depth). Moreover, fishes from shallow and mesophotic depths (Amado-­Filho et al., 2016) and Fernando de Noronha Archipelago (0–60 m) were also associated with the slopes of these seamounts have been recently described (Amado-­Filho, Pereira-­Filho, et al., and insular shelves (Wirtz et al., 2017). 2012; Pereira-­Filho et al., 2015; Santos et al., 2016). In these stud- MEs in these remote seamounts and oceanic islands may act ies, rhodolith beds were reported as the principal seascape. These as centres of endemism and “stepping stones,” providing oppor- rhodoliths appeared both as free and coalesced forms and produced tunities for certain species to expand their ranges between the patch reefs on both islands (Amado-­Filho et al., 2016; Pereira-­Filho Eastern and Western Atlantic (Irving, 1989, 2013; Wirtz et al., et al., 2015). 2017). Indeed, seamounts in this ocean basin (especially near Saint The most remote MEs in this area are located in the Saint Peter Helena Island) are the least known habitats for the occurrence of and Saint Paul Archipelago, on the mid-­Atlantic ridge. Spalding et al. MEs due to the lack of research conducted away from the coast (2007) defined this archipelago as a distinct marine ecoregion due to in Brazil and Africa. Overall, there is little information (i.e., map- its distance from the coast and its unique biodiversity (Supporting ping and quantitative studies) about the MEs around these remote Information Appendix S2). In this area, some shifts in the benthic South Atlantic islands, which clearly suggests the need for more and reef fish assemblages between shallow and mesophotic depths studies. have been reported (Amaral et al., 2000; Magalhães et al., 2015; Rosa et al., 2016; Soares, Lotufo, et al., 2017). At least one bryozoan, 3.4 | The Gulf of Guinea Province Margaretta buski, is considered to be an ecosystem engineers species and is abundant at depths of 10–45 m depth. This bryozoan forms This region comprises the marine ecoregion of the Gulf of Guinea erect colonies, providing an animal forest for the associated species islands on the west coast of Africa (Spalding et al., 2007). The MEs (such as crustaceans and echinoderms). Rosa et al. (2016) observed in the Eastern Atlantic Ocean are composed of rocky reefs with that scleractinian corals and macroalgae (turf and fleshy) were more biogenic patches that support shore fishes (Morais & Maia, 2017; abundant between 30 and 40 m depths. As the depth increased, Wirtz et al., 2007). This area has very distinct oceanographic fea- they were progressively replaced by more animal-­dominated assem- tures that prevent the development of extensive coral reefs, such blages (such as sponges, bryozoans and black corals). These authors as large inputs of freshwater from the mainland and cold water from also found that some reef fishes were clearly associated with the oceanic currents. As a result, corals are often limited by temperature black coral forests (Tanacetipathes spp.), indicating that these animal to the first 20 m (Spalding et al., 2001). Wirtz et al. (2007) report forests play a key nursery function in the lower depth limit of these fishes mainly in shallow waters (0–30 m) and in mesophotic depths MEs. (between 32 and 45 m depth) around São Tomé and Príncipe islands. Considering that shallow-­water scleractinian corals in this prov- ince are usually restricted to warm waters, the presence of MEs in 3.3 | Saint Helena and Ascension islands cool waters (São Tomé island), recently described by Morais and The Saint Helena and Ascension islands (Central Atlantic) consti- Maia (2017), is surprising. Although the warm shallow-­water forma- tute one of the least studied biogeographical provinces in the SAO. tions (<30 m) were dominated by coralline algae, with few species of Ascension Island (7°57′S 14°22′W) is one of the most remote vol- scleractinian corals (M. cavernosa and Siderastrea sp.), which are also canic islands in the South Atlantic and is located ~1,200 km from common in the South America (Leão et al., 2016) and other African St Helena (the nearest island) and ~2,300 km from Recife (Brazilian islands (Laborel, 1974), the upper mesophotic zone (35–50 m) coastline). Seamounts, canyons and insular shelves are common fea- showed a benthic composition that changed due to the thermocline. tures around the Ascension and Saint Helena islands, which provide At this depth, the seascape became dominated by a marine animal habitat for the occurrence of MEs without the presence of coral forest of black corals (mainly Antipathes gracilis and Tanacetipathes reefs (Irving, 2013). For example, the shallowest part of the Grattan spinescens) and associated reef fishes, such as Paranthias furcifer, seamount near Ascension Island is 72 m deep, with the occurrence of Clepticus africanus, and Lutjanus fulgens (Morais & Maia, 2017). In algae (dominated by coralline algae), marine animal forests (sponge this case, it would be possible to test whether temperature-­driven bottoms, scleractinian corals and octocoral animal forests) and asso- differences in the assemblages would prevent the areas acting as ciated reef fishes. Rocky formations around Ascension Island consist effective deep-­sea refugia. Moreover, the shallow and mesophotic of bedrock reefs, vertical cliffs and steep boulder slopes, as well as biodiversity of this African archipelago is of scientific interest due to a variety of caves, canyons and lava tubes providing significant geo- the occurrence of easterly flowing Equatorial currents that link the diversity for the little-­studied MEs (Irving, 1989, 2013; Wirtz et al., western and the eastern Atlantic at this tropical latitude (Laborel, 2017). 1974; Wirtz et al., 2007). 262 | SOARES et al. 4 | H U M A N PR E S S U R E S O N TH E S O U TH The lack of monitoring of fishing activity in the South Atlantic ATL A NTI C M E S MEs (Magalhães et al., 2015; Morais & Maia, 2017; Moura et al., 2016; Soares, Lotufo, et al., 2017) means that it is now urgent to MEs are commonly viewed as being less affected by disturbance than quantify fishing impacts to test the refugia hypothesis (Lindfield shallow-­water ecosystems (Kahng et al., 2017). Local and regional et al., 2016). In addition, it is important to monitor the effects of human impacts (e.g., overfishing, sedimentation, dredging and pol- fishing and other human pressures on the shallow and mesophotic lution) and natural disturbances (e.g., storms, extreme heat waves, reef fishes. Pinheiro et al. (2018) provided a recent database on hurricanes, severe floods and droughts) tend to decrease with in- reef fishes (0–150 m depth), which detected 733 fish species, of creasing depths and distance from the coast (Baker et al., 2016; Loya which 405 were Southwestern Atlantic (SWA) resident reef fishes et al., 2016). However, MEs may also be less resilient than shallow or strictly reef species, 111 were endemic to the SWA and 78 were reefs (Pyle et al., 2016), and assessments of their health status and considered to be threatened with extinction. possible threats are surely needed to ensure their conservation. MEs Bottom trawling, bottom-­set long lines, bottom-­set gillnets and are gradually becoming part of the international list of vulnerable abandoned fishing gear are particularly important threats to meso- marine ecosystems (VMEs; Aguilar, Perry, & López, 2017), and in the photic ecosystems (Soares, Lotufo, et al., 2017), and they have also South Atlantic Ocean, protective measures need to be developed to been reported in the South Atlantic MEs (Supporting Information ensure their conservation. Appendix S2). Ecosystem engineers (e.g., sponges and black corals) The South Atlantic MEs are receiving pressure from fishing, car- in mesophotic depths have low growth and recovery rates, which bonate deposit surveys, submarine cables, biotechnological product suggest the vulnerability of these taxa to fishing activities. Bottom development, mining, oil and gas exploration, dispersion of invasive trawling is a fishing activity that has been impacting shallow marine species and marine debris (Supporting Information Appendix S2). ecosystems along the Brazilian coast and has extended into offshore These increasing pressures have been observed along the tropical areas in the last decades. Indeed, bottom trawling has a negative and subtropical coasts of both South America and Africa (all four impact on MEs worldwide (Baker et al., 2016; Cánovas-­Molina et al., biogeographical provinces considered in the present review) and 2016). Another regional pressure that occurs in MEs on the South thus are threatening the MEs in these areas, including reef species Atlantic is the presence of marine debris (e.g., macroplastics, micro- (Supporting Information Appendix S2). plastics, pellets and abandoned fishing gear; Supporting Information At regional levels, these anthropogenic pressures can be man- Appendix S2). The marine debris in the mesophotic zone can intro- aged (e.g., fisheries regulations, marine spatial planning and marine duce invasive species as well as threaten the life of marine organ- protected areas) within the regions covered in the present review. isms by entrapment, ingestion of solid wastes or contamination by On the other hand, global stressors, such as "global warming and microplastics. Moreover, plastic debris stresses reef-­building species sea-surface temperature anomalies (SST) anomalies, acidification, through anoxia, light deprivation and toxin release, increasing the and extreme floods and droughts, are unmanageable at local and risk associated with pathogens (Lamb et al., 2018). regional scales. However, they should still be included in global con- On the Brazilian and West African coasts, mining activities that servation strategies (e.g., reduction in carbon emissions). target oil, gas, metals and carbonates have been receiving strong public support and private investments, especially in deeper waters. In the VTC, mining activities are expected to increase in the next 4.1 | Local and regional stressors decades by targeting metals on slopes and carbonates on seamount Fishing is one of the main anthropogenic pressures on MEs, and all summits (Pinheiro et al., 2015). These mining activities could threaten biogeographical provinces included in this study sustain important local MEs. Recent discoveries of oil and gas fields near mesophotic fishing grounds (Supporting Information Appendix S2). Artisanal ecosystems may also increase the pressure on MEs. International and commercial boats operate mainly with lines, traps and bot- companies have acquired exploratory drilling blocks that will soon tom trawling. The MEs, characterized by high population densities be producing oil, if approved by the national agencies, near the ex- and species richness, concentrate fishing grounds and sustain an tensive MEs off the North and Northeastern coast of Brazil, increas- important multispecific reef fishery (Eduardo et al., 2018; Pinheiro ing the risk of bio-­invasion and oil spills in a little-­studied ecosystem. et al., 2015; Rosa et al., 2016). These activities affect even the It is therefore crucial to carefully evaluate the exploitation of these remote Saint Peter and Saint Paul Archipelago where overfishing marine resources and assess possible environmental risks associated may have caused local extinction of the shark species Carcharhinus with mining activities, considering their direct and indirect impacts galapagensis (Luiz & Edwards, 2011). Significant fisheries for the on MEs (Creed et al., 2017; Silva, Etnoyer, & MacDonald, 2016). It Southern red snapper, Lutjanus purpureus (2,900 metric tons/ is also important to note that these activities are already subject to year), and spiny lobsters, Palinurus spp. (1,360 metric tons/year), environmental licensing in Brazil (Santos et al., 2016; Soares, Lotufo, were recorded for the MEs along the Amazon coast. Reef fisher- et al., 2017) and therefore require a detailed description of the MEs ies have been carried out by small-­ to medium-­sized boats (8 to and the effects of the proposed activities. 20 m lengths) operating with traps (for lobsters) and hand lines or Finally, biological invasions usually have negative consequences long lines (for reef fishes) on the outer shelf (Moura et al., 2016). on MEs (Andradi-­Brown et al., 2017) by affecting ecosystem goods SOARES et al. | 263 and services and leading to financial losses and biodiversity changes the South Atlantic MEs with some degree of resilience to thermal (Baker et al., 2016). Recent investigations have identified the intro- stresses. The susceptibility to temperature anomalies must take into duction and proliferation of alien species worldwide (e.g., Carijoa consideration not only the low-­bleaching threshold of South Atlantic sp., Avrainvillea sp., Pterois spp.), which may alter the mesophotic MEs but also the exposure to elevated temperatures. Greater depths community structure at the expense of native biodiversity (Baker may also buffer these deeper ecosystems from some pressures (e.g., et al., 2016). Brazilian and African coasts have experienced signifi- irradiance and local impacts). For example, at Trindade Island within cant transoceanic ship traffic since the 16th century, but the con- the VTC, tissue necrosis and bleaching affected >90% of the coral sequences of this maritime transport on bio-­invasions have only colonies in shallow waters, whereas >90% of the coral colonies in recently received academic and governmental attention (Castro, MEs were healthy (Meirelles et al., 2015). However, this resilience Fileman, & Hall-­Spencer, 2017). Indeed, recent studies have re- may be compromised in the face of the increased frequency and ported the presence of invasive species in the South Atlantic MEs duration of thermal stress predicted as a result of climate change (Supporting Information Appendix S2), for example, ophiuroids and other stressors (eutrophication, acidification and sea-­level rise; (Ophiothela mirabilis) on the Amazon shelf (Moura et al., 2016). Hoegh-­Guldberg et al., 2017). Other potential alien invaders are lionfish (Pterois volitans) and A commonly overlooked global driver of pressures on MEs is orange cup corals (Tubastraea coccinea and T. tagusensis), which have the occurrence of extreme events, such as droughts and floods. been associated with impacts on marine ecosystems in the Western Lack of information on the impact of such events is partly due to Atlantic (Andradi-­Brown et al., 2017; Creed et al., 2017). Lionfish has the distance between MEs and the coastline, which diminishes the been reported in shallow waters on the east South American coast influence of estuarine plumes, freshwater inputs and precipitation (Ferreira et al., 2015) but have also been observed at mesophotic regimes. However, MEs in North Brazil are seasonally under the in- depths in the Caribbean Sea (Baker et al., 2016). Orange cup corals fluence of large rivers and are directly influenced by the input of were the first invasive scleractinian species introduced into the SAO, fine sediments, nutrients and turbid water. The Amazon coast has with known occurrences near the Northeastern Brazil MEs (Soares, already been affected by droughts and floods due to climatic anom- Davis, & Carneiro, 2018) and the Abrolhos Bank (Costa et al., 2014). alies (Marengo & Espinoza, 2016). Moura et al. (2016) have indicated Considering the threats imposed by invasive species to biodiversity, that such extreme events are on the rise under climate change sce- it is essential that international and national environmental agencies narios and will probably influence the functioning of MEs off the operating in the SAO adopt guidelines to restrain invasive species Amazon River. expansion. 5 | FU T U R E D I R EC TI O N S : R E S E A RC H A N D 4.2 | Global stressors M A N AG E M E NT Regarding global environmental stressors, little is known about their effects on MEs, mainly due to the lack of long-­term monitoring. For The present review integrates studies on the geographical distribu- example, ocean acidification, which is expected to accelerate in the tion, human impact and conservation status of the MEs in the SAO 21st century (Hoegh-­Guldberg et al., 2017), compromises carbonate and aims to be a first step in the elaboration of a more detailed sce- production by reef-­building species (e.g., calcareous algae and cor- nario regarding these deep-­water habitats within the tropical zone als) and is the least studied global change stressor in the SAO (Kerr of this ocean basin. It was outside the scope of our review to inte- et al., 2016). Nevertheless, coralline algae are among the most im- grate information on subtropical, temperate and polar ecosystems portant framework builders in the South Atlantic MEs. Thus, ocean in the mesophotic zone of the SAO. Nevertheless, integrative stud- acidification is a real threat to these MEs, which may experience a ies on these southern environments are much needed, due to their considerable shift in their structure and function in the near future ecological importance. For example, it is known that the distribution (Amado-­Filho, Moura, et al., 2012; Horta et al., 2016). of rhodoliths and rocky reefs in Brazil extends beyond the tropical Coral-­bleaching events, mainly in shallow-­water reefs, are related zone, crossing through the subtropical zone to the temperate coast to both local and global stressors, such as anomalies in SST, including of Santa Catarina state (Horta et al., 2016). several weeks of above-­average temperatures and positive hotspot Future research should focus on four main topics: (a) baseline as- values (Leão et al., 2016; Soares, Lotufo, et al., 2017). Despite increas- sessment (biodiversity, ecology, physiology, geodiversity and chem- ing rates of bleaching and coral diseases in the shallow-­water reefs ical and physical oceanography), (b) evaluation of the ME goods and of the SAO, mass mortality of reef-­building species has not yet been services, (c) impacts and (d) conservation strategies. Work on the detected in coastal reefs. This may be due to the higher resilience first and second topics is essential to provide the basis on which the of corals in the SAO, which are often adapted to suboptimal condi- other two topics can be studied and to investigate the potential role tions (Leão et al., 2016). Thus, we hypothesize that the presence of of MEs as refuges in a changing ocean. resilient reef-­building corals (e.g., Siderastrea spp., M. cavernosa and In terms of biodiversity, it is necessary to conduct studies (i.e., Mussismillia spp.) in the mesophotic zone and the abundance of non-­ beta diversity, nestedness and turnover) comparing shallow (<30 m), coral organisms (e.g., sponges and calcareous algae) may provide upper (30–60 m) and lower (>60 m) mesophotic zones to verify the 264 | SOARES et al. similarity of their assemblages using meta-­analysis and large-­scale the recent and future human pressures. Therefore, consider- studies with a standardized methodology. Studies in the Caribbean ing the economic activities with environmental impacts on MEs Sea and Indo-­Pacific Ocean have shown that shallow and upper me- (Supporting Information Appendix S2), we suggest five main man- sophotic zones are frequently similar (Kahng et al., 2017; Slattery, agement and conservation actions: (a) adoption of ocean zoning Lesser, Brazeau, Stokes, & Leichter, 2011). However, the lower com- (Yates, Schoeman, & Klein, 2015) and large marine spatial plan- munity is a distinct assemblage with a high degree of endemism ning (Dunstan et al., 2016); (b) creation and expansion of marine (Kahng, Copus, & Wagner, 2014). Semmler et al. (2017), studying a protected areas (MPAs) using systematic conservation planning; dataset of over 9,000 species found throughout the Gulf of Mexico, (c) rigorous fishing regulations and environmental licensing for observed that the taxonomic overlap between shallow and pro- human activities, based on sound scientific knowledge; (d) reduc- gressively deeper zones declined steadily with depth. The authors tion in carbon emissions to the atmosphere, considering the vul- concluded that the mesophotic zone is home to three ecological nerability of MEs to ocean acidification and SST anomalies; and communities: the upper, which is confluent with shallow reefs; a dis- (e) extending environmental monitoring programs to include the tinct mesophotic assemblage spanning 60–120 m; and a third that South Atlantic MEs. extends onto the outer continental shelf, with limited potential to The major challenges in the conservation of MEs in the SAO are serve as a refuge. related to surveillance in offshore areas (especially in seamounts), Another important aspect is the limited connectivity between the financial cost of environmental monitoring and the lack of de- the shallow and mesophotic ecosystems, observed only in a small tailed seabed mapping to identify required management actions. number of species in the Caribbean Sea and the South Atlantic The latter should be the subject of scientific research efforts to (Bongaerts et al., 2017; Morais & Santos, 2018). Within each zone, obtain detailed information on the distribution of MEs. Regarding it is necessary to verify the patterns of bathymetric distribution environmental monitoring and surveillance, the use of technolo- and genetic connectivity of marine populations, particularly con- gies should be encouraged, such as the use of remote sensing by sidering the peculiarities of large rhodolith beds, sponge bottoms, satellites, the Internet of the oceans and the use of automatic iden- black coral forests, octocoral animal forests and reefs in turbid tification systems for surveillance of fishing activities and shipping waters of the SAO. Finally, horizontal connectivity (i.e., between lines in MEs. Kroodsma et al. (2018), who detected the global foot- MEs) remains poorly understood worldwide (Kahng et al., 2014) print of industrial fisheries and its results, found a lack of data from and it should also be assessed in the SAO, especially in relation to the fisheries in the Southwestern Atlantic, which harbours unique previous hypotheses of large-­s cale connectivity between Brazilian MEs. and Caribbean species (Francini-­F ilho et al., 2018; Rocha, 2003; The management plans of the MPAs that cover mesophotic hab- Rocha et al., 2000). itats in the SAO generally do not directly contemplate actions to It is necessary to understand the cumulative impact of human conserve these ecosystems. Therefore, these documents need to be stressors that affect ecosystem goods and services provided by the updated to include specific actions for deeper ecosystems, such as MEs in order to discuss the challenges faced and the role of science no-­take zones, environmental monitoring and surveillance. Moreover, in improving South Atlantic Ocean governance in this twilight zone Soares and Lucas (2018) discussed aspects of the recent and large (Baker et al., 2016). The mapping and assessment of these goods and MPAs in Vitória-­Trindade Chain and St. Paul and St. Peter Archipelago, services would be a highly valuable source of information and would as well as the need for a participatory management plan. The authors contribute to an understanding of their current and potential bene- also noted that most of the area (~87%) covered by these specific fits to society. MPAs was under a regime of sustainable use, which permitted a range The present review has compiled information (Supporting of human activities, including fishing and mining. Giglio et al. (2018) Information Appendix S2) on anthropogenic pressures on MEs. The and Magris and Pressey (2018) revealed the absence of large no-­take quantitative impacts of these pressures are still poorly understood zones in vulnerable habitats (including MEs), uninformed opportunism worldwide (Turner et al., 2017), and this is another limitation of this by the Brazilian government in the creation of these MPAs, and poor review. A better comprehension requires consideration of scales at adherence to best practices in MPA planning (e.g., lack of connectivity which these pressures are acting, an assessment of the degree of between habitats). change in key mesophotic ecological indicators, and an analysis of The increase in MPAs in the world’s oceans as well as their man- the resilience of these ecosystems. It also should be considered that agement should aim the protection of multiple components of meso- the way these pressures act is cumulative and synergistic (Mumby, & photic biodiversity (taxonomic, phylogenetic and functional), and the Anthony, 2015). Thus, appropriate metrics and indicators should be ecosystem goods and services. However, a clear spatial mismatch be- used (Cánovas-­Molina et al., 2016). tween the existing degree of protection and all conservation priorities The current scarcity of information on the structure and worldwide has been detected (Lindegren, Holt, Mackenzie, & Rahbeck, functioning of the SAO MEs should not limit conservation ac- 2018). The MEs provide an opportunity to avoid past mistakes in the tion and the management of these ecosystems. The “precau- design of MPAs planning and to provide effective action to protect tionary principle,” known in maritime and environmental laws these vulnerable habitats (e.g., no-­take zones). Specific studies for the (Santos et al., 2016), should be adopted in this case and consider creation and expansion of MPAs should provide information about SOARES et al. | 265 their size, spacing, connectivity, management and shape in order to Aguirre (Eds.), Rhodolith/Maërl Beds: A global perspective. Coastal achieve ecological effectiveness for conservation action. research library. (vol. 15, pp. 299–317). New York, NY: Springer. https://doi.org/10.1007/978-3-319-29315-8 Amado-Filho, G. M., Moura, R. L., Bastos, A. C., Francini-Filho, R. B., Pereira Filho, G. H., Bahia, R. G., … Motta, F. S. (2016). Mesophotic 6 | CO N C LU S I O N S ecosystems of the unique South Atlantic atoll are composed by rh- odolith beds and scattered consolidated reefs. Marine Biodiversity, 46(4), 933–936. https://doi.org/10.1007/s12526-015-0441-6 In this study, we reviewed recent research on South Atlantic MEs Amado-Filho, G. M., Moura, R. L., Bastos, A. C., Salgado, L. T., Sumida, P. and the pressures threatening these ecosystems. We also pro- Y., Guth, A. Z., … Thompson, F. L. (2012). Rhodolith beds are major vided options for their conservation. Conservation policies have CaCO3 bio-­factories in the tropical south West Atlantic. PLoS ONE, generally focused on mesophotic biogenic reefs. However, our 7(4), e35171. https://doi.org/10.1371/journal.pone.0035171 Amado-Filho, G. M., Pereira-Filho, G. H., Bahia, R. G., Abrantes, D. review indicated that there is a need for a change of approach to P., Veras, P. C., & Matheus, Z. (2012). Occurrence and distribu- conservation and that measures should be taken to protect the tion of rhodolith beds on the Fernando de Noronha Archipelago distinct seascapes identified in the mesophotic zone (e.g., sponge of Brazil. Aquatic Botany, 101, 41–45. https://doi.org/10.1016/j. aggregations, biogenic and rocky reefs, octocoral forests, black aquabot.2012.03.016 Amaral, F. M. D., Hudson, M. M., Silveira, F. L., Migotto, A. E., Pinto, S. M., & coral forests and rhodolith beds). Longo, L. L. (2000). Cnidarians from St. Peter and St. Paul Archipelago, Recent descriptions of MEs in the South Atlantic demonstrate Northeast Brazil. Proc. 9th Intern Coral. Reef. Symp., 1, 567–572. how little we know about this ocean and how much of its biodiversity Andradi-Brown, D. A., Vermeij, M. J. A., Slattery, M., Lesser, M., Bejarano, may still be hidden. A solid scientific knowledge base for the South I., Appeldoorn, R., … Exton, D. A. (2017). Large scale invasion of Atlantic MEs is essential to build a clear unifying picture to share Western Atlantic mesophotic reefs by invasive lionfish potentially undermine culling-­based management. Biological Invasions, 19(3), with stakeholders and to gain public support for MEs conservation. 939–954. https://doi.org/10.1007/s10530-016-1358-0 Baker, E., Puglise, K., & Harris, P.T.(2016). Mesophotic coral ecosystems: a lifeboat for coral reefs? The United Nations Environment Programme AC K N OW L E D G E M E N T S (UNEP) and GRID-Arendal, Nairobi and Arendal. 98p. Retrieved from http://www.unep.org/ourplanet/may-2016/unep-publications/ Financial support was provided by the Conselho Nacional de mesophotic-coral-ecosystems Desenvolvimento Científico e Tecnológico—CNPq (Grants Bastos, A. C., Moura, R. L., Amado-Filho, G. M., D’Agostini, D. P., Secchin, 233808/2014-­0 and 307061/2017 to MOS), PRONEX FUNCAP/ N. A., Francini-Filho, R. B., … Thompson, F. L. (2013). Buracas: CNPq (Grant PR2-­0101-­0 0008.01.00/15) and INCT AmbTropic Novel and unusual sinkhole-­like features in the Abrolhos Bank. Continental Shelf Research, 70, 118–125. https://doi.org/10.1016/j. (National Institute of Science and Technology for the Tropical csr.2013.04.035 Marine Environment). Bastos, A. C., Moura, R. L., Moraes, F. C., Vieira, L. S., Braga, J. C., Ramalho, L. V., … Webster, J. M. (2018). Bryozoans are major modern builders of South Atlantic oddly shaped reefs. Scientific Reports, 8, DATA ACC E S S I B I L I T Y 9638. https://doi.org/10.1038/s41598-018-27961-6 Bongaerts, P., Ridgway, T., Sampayo, E. M., & Hoegh Guldberg, O. All bathymetric and environmental GIS layers, and results generated (2010). Assessing the “deep reef refugia” hypothesis: Focus on for this study are available as raster grids from the Pangaea database. Caribbean reefs. Coral Reefs, 29, 309–327. https://doi.org/10.1007/ s00338-009-0581-x Bongaerts, P., Riginos, C., Brunner, R., Englebert, N., Smith, S. R., & ORCID Hoegh-Guldberg, O. (2017). Deep reefs are not universal refuges: Reseeding potential varies among coral species. Science Advances, Marcelo de Oliveira Soares http://orcid. 3(2), e1602373. https://doi.org/10.1126/sciadv.1602373 org/0000-0002-4696-3166 Camargo, J. M. R., Araujo, T. C. M., Ferreira, B. P., & Maida, M. (2015). Topographic features related to recent sea level history in a sediment-­ starved tropical shelf: Linking the past, present and future. Regional Studies in Marine Science, 2, 203–211. https://doi.org/10.1016/j. REFERENCES rsma.2015.10.009 Aguilar, R., Perry, A. L., & López, J. (2017). Conservation and manage- Cánovas-Molina, A., Montefalcone, M., Bavestrello, G., Cau, A., Bianchi, ment of vulnerable marine benthic ecosystems. In S. Rossi (Ed.), C. K., Morri, C., … Bo, M. (2016). A new ecological index for the sta- Marine animal forests: The ecology of benthic biodiversity hotspots tus of mesophotic assemblages in the mediterranean based on ROV of the world. Major Reference Book (pp. 1–43). New York, NY: fotography and video footage. Continental Shelf research, 121, 13–20. Springer. https://doi.org/10.1016/j.csr.2016.01.008 Almeida, N. M., Vital, H., & Gomes, M. P. (2015). Morphology of sub- Castro, M. C. T., Fileman, T. W., & Hall-Spencer, J. M. (2017). Invasive marine canyons along the continental margin of the Potiguar Basin, species in the Northeastern and Southwestern Atlantic Ocean: A re- NE Brazil. Marine and Petroleum Geology, 68, 307–324. https://doi. view. Marine Pollution Bulletin, 116, 41–47. https://doi.org/10.1016/j. org/10.1016/j.marpetgeo.2015.08.035 marpolbul.2016.12.048 Amado-Filho, G. M., Bahia, R. G., Pereira-Filho, G. H., & Longo, L. L. Clapperton, C. M. (1993). Nature of environmental changes in (2017). South Atlantic rhodolith beds: Latitudinal distribution, spe- South America at the Last Glacial Maximum. Palaeogeography, cies composition, structure and ecosystem functions, threats and Palaeoclimatology, Palaeoecology, 101(3), 189–208. https://doi. conservation status. In R. Riosmena-Rodríguez, W. Nelson, & J. org/10.1016/0031-0182(93)90012-8 266 | SOARES et al. Collette, B. B., & Rutzler, K. (1977). Reef fishes over sponge bottoms off Hoeksema, B. W., Bongaerts, P., & Baldwin, C. C. (2017). High coral cover the mouth of the Amazon river. Proceedings of the 3rd International at lower mesophotic depths: A dense Agaricia community at the Coral Reef Symposium, Miami, Florida. leeward side of Curaçao, Dutch Caribbean. Marine Biodiversity, 47, Cordeiro, R. T., Kitahara, M. V., & Amaral, F. M. D. (2012). New records 67–70. https://doi.org/10.1007/s12526-015-0431-8 and range extensions of azooxanthellate scleractinians (Cnidaria: Holstein, D. M., Paris, C. B., Vaz, A. C., & Smith, T. B. (2016). Modeling Anthozoa) from Brazil. Marine Biodiversity Records, 5, 1–6. vertical connectivity and mesophotic refugia. Coral Reefs, 35(1), 23– Cordeiro, R. T. S., Neves, B. M., Rosa-Filho, J. S., & Pérez, C. D. (2015). 37. https://doi.org/10.1007/s00338-015-1339-2 Mesophotic coral ecosystems occur offshore and north of the Hopley, D. (2011). Encyclopedia of modern Coral Reefs: Structure, form Amazon River. Bulletin of Marine Science, 91, 491–510. https://doi. and process. Encyclopedia of Earth Sciences Series. New York, NY: org/10.5343/bms.2015.1025 Springer. https://doi.org/10.1007/978-90-481-2639-2 Costa, T. J. F., Pinheiro, H. T., Teixeira, J. B., Mazzei, E. F., Bueno, L., Hora, Horta, P. A., Riul, P., Amado-Filho, G. M., Gurgel, C. F. D., Berchez, F., M. S. C., … Rocha, L. A. (2014). Expansion of an invasive coral species Nunes, J. M. C., … Figueiredo, M. (2016). Rhodoliths in Brazil: Current over Abrolhos Bank, Southwestern Atlantic. Marine Pollution Bulletin, knowledge and potential impacts of climate change. Brazilian 85(1), 252–253. https://doi.org/10.1016/j.marpolbul.2014.06.002 Journal of Oceanography, 64(2), 117–136. https://doi.org/10.1590/ Creed, J. C., Fenner, D., Sammarco, P., Cairns, S., Capel, K., Junqueira, S1679-875920160870064sp2 A. O. R., … Oigman-Pszczol, S. (2017). The invasion of the azooxan- Hovland, M. (2008). Deep‐water coral reefs: Unique biodiversity hot‐spots, thellate coral Tubastraea (Scleractinia: Dendrophyllidae) throughout 1st ed. (pp. 278). Amsterdam, the Netherlands: Springer. the world: History, pathways and vectors. Biological Invasions, 19(1), Irving, R. A. (1989). A preliminary investigation of the sublittoral habi- 283–305. https://doi.org/10.1007/s10530-016-1279-y tats and communities of Ascension Island, South Atlantic. Progress in Dunstan, P. K., Bax, N. J., Dambacher, J. M., Hayes, K. R., Hedge, P. T., Underwater Science, 13, 65–78. Smith, D. C., … Smith, A. D. M. (2016). Using ecologically and biolog- Irving, R. A. (2013). Ascension Island’s hermatypic but non-reef building ically significant marine areas (EBSAs) to implement marine spatial corals. In C. R. C. Sheppard (Ed.), Coral reefs of the United Kingdom over‐ planning. Ocean and Coastal Management, 121, 116–127. https://doi. seas territories, coral reefs of the world 4 (pp. 213–221). Dordrecht, The org/10.1016/j.ocecoaman.2015.11.021 Netherlands: Springer. https://doi.org/10.1007/978-94-007-5965-7 Eduardo, L. N., Frédou, T., Lira, A. S., Ferreira, B. P., Bertrand, A., Ménard, Kahng, S. E., Copus, J. M., & Wagner, D. (2014). Recent advances in the F., … Frédou, F. L. (2018). Identifying key habitat and spatial pat- ecology of mesophotic coral ecosystems (MCEs). Current Opinion terns of fish biodiversity in the tropical Brazilian continental shelf. in Environmental Sustainability, 7, 72–81. https://doi.org/10.1016/j. Continental Shelf Research, 166, 108–118. https://doi.org/10.1016/j. cosust.2013.11.019 csr.2018.07.002 Kahng, S., Copus, J. M., & Wagner, D. (2017). Mesophotic coral eco- Feitoza, B. M., Rosa, R. S., & Rocha, L. A. (2005). Ecology and zoogeog- systems. In S. Rossi (Ed.), Marine animal forests (pp. 185–206). raphy of Deep-­Reef Fishes in northeastern Brazil. Bulletin of Marine New York, NY: Springer International Publishing. https://doi. Science, 76, 725–742. org/10.1007/978-3-319-21012-4 Ferreira, C. E. L., Luiz, O. J., Floeter, S. R., Lucena, M. B., Barbosa, M. C., Kerr, R., Cunha, L. C., Kikuchi, R. K. P., Horta, P., Ito, R. G., Muller, M. Rocha, C. R., … Rocha, L. A. (2015). First record of invasive Lionfish N., … Copertino, M. S. (2016). The Western South Atlantic Ocean (Pterois volitans) for the Brazilian coast. PLoS ONE, 10(4), e0123002. in a High-­CO2 World: Current measurement capabilities and per- https://doi.org/10.1371/journal.pone.0123002 spectives. Environmental Management, 57, 740–752. https://doi. Francini-Filho, R., Asp, N. E., Siegle, E., Hocevar, J., Lowyck, K., D’Avila, org/10.1007/s00267-015-0630-x N., … Thompson, F. L. (2018). Perspectives on the Great Amazon Kitahara, M. V. (2007). Species richness and distribution of azooxan- Reef: Extension, biodiversity, and threats. Frontiers in Marine Science, thellate Scleractinia in Brazil. Bulletin of Marine Science, 81(497–518), 5, 142. https://doi.org/10.3389/fmars.2018.00142 2007. Francini-Filho, R., & Moura, R. L. (2008). Dynamics of fish assemblages Kitahara, M. V. (2009). The deep-­sea demersal fisheries and the azooxan- on coral reefs subjected to different management regimes in the thellate corals from southern Brazil. Biota Neotropica, 9, 1–10. Abrolhos Bank, eastern Brazil. Aquatic Conservation: Marine and Kroodsma, D. A., Mayorga, J., Hochberg, T., Miller, N. A., Boerder, K., Freshwater Ecosystems, 18(7), 1166–1179. https://doi.org/10.1002/ Ferretti, F., … Worm, B. (2018). Tracking the global footprint of aqc.966 fisheries. Science, 359, 904–908. https://doi.org/10.1126/science. Freitas, J. E. P., & Lotufo, T. M. C. (2015). Reef fish assemblage and zoogeo- aao5646 graphic affinities of a scarcely known region of the western equatorial Laborel, J. (1974). West African reef corals, an hypothesis on their ori- Atlantic. Journal of Marine Biological Association of the United Kingdom, gin. Proceedings of the Second International Coral Reef Symposium, 95(3), 623–633. https://doi.org/10.1017/S0025315414001404 Brisbane, Australia, 1, 425–443. Giglio, V. J., Pinheiro, H. T., Bender, M. G., Bonaldo, R. M., Costa‐Lotufo, Lamb, J. B., Willis, B. L., Fiorenza, E. A., Couch, C. S., Howard, R., Rader, D. L. V., Ferreira, C. E. L., … Francini‐Filho, R. B. (2018). Large and marine N., … Harvell, C. D. (2018). Plastic waste associated with disease on protected areas are flawed and raise concerns: Comments on Soares coral reefs. Science, 359(6374), 460–462. https://doi.org/10.1126/ and Lucas (2018). Marine Policy, 96, 13–17. https://doi.org/10.1016/j. science.aar3320 marpol.2018.07.017 Leão, Z. M. A. N., Kikuchi, R. K. P., Ferreira, B. P., Neves, E. G., Sovieroski, Glynn, P. W. (1996). Coral reef bleaching: Facts, hypotheses, and H. H., Oliveira, M. D., … Johnsson, R. (2016). Brazilian coral reefs in a implications. Global Change Biology, 2, 495–509. https://doi. period of global change: A synthesis. Brazilian Journal of Oceanography, org/10.1111/j.1365-2486.1996.tb00063.x 64, 97–116. https://doi.org/10.1590/S1679-875920160916064sp2 Hinderstein, L. M., Marr, J. C. A., Martinez, F. A., Dowgiallo, M. J., Lindegren, M., Holt, B. G., Mackenzie, B. R., & Rahbeck, C. (2018). A Puglise, K. A., Pyle, R. L., … Appeldoorn, R. (2010). Theme section global mismatch in the protection of multiple marine biodiversity on ‘mesophotic coral ecosystems’: Characterization, ecology and components and ecosystem services. Scientific Reports, 8, 4099. management’. Coral Reefs, 29, 247–251. https://doi.org/10.1007/ https://doi.org/10.1038/s41598-018-22419-1 s00338-010-0614-5 Lindfield, S. J., Harvey, E. S., Halford, A. R., & Mcllwain, J. L. (2016). Hoegh-Guldberg, O., Mumby, P. J., Hooten, A. J., Steneck, R. S., Greenfield, Mesophotic depths as refuge areas for fishery-­t argeted species P., Gomez, E., … Scheffer, M. (2017). Coral reefs in the Anthropocene. on coral reefs. Coral Reefs, 35, 125–137. https://doi.org/10.1007/ Nature, 546, 82–90. https://doi.org/10.1038/nature22901 s00338-015-1386-8 SOARES et al. | 267 Loya, Y., Eyal, G., Treibitz, T., Lesser, M. P., & Appeldoorn, R. (2016). Pereira-Filho, G. H., Amado-Filho, G. M., Guimarães, S. M. P. B., Theme section on mesophotic coral ecosystems: Advances in Moura, R. L., Sumida, P. Y. G., Abrantes, D. P., … Francini-Filho, knowledge and future perspectives. Coral Reefs, 35, 1–9. https://doi. R. B. (2011). Reef fish and benthic assemblages of the Trindade org/10.1007/s00338-016-1410-7 and Martin Vaz Island Group, southwestern Atlantic. Brazilian Luiz, O., & Edwards, A. L. (2011). Extinction of a shark population in the Journal of Oceanography, 59(3), 201–212. https://doi.org/10.1590/ Archipelago of Saint′s rocks (equatorial Atlantic) inferred from the S1679-87592011000300001 historical record. Biological Conservation, 144, 2873–2881. https:// Pereira-Filho, G. H., Francini-Filho, R. B., Pierozzi-Jr, I., Pinheiro, H. T., doi.org/10.1016/j.biocon.2011.08.004 Bastos, A. C., Moura, R. L., … Amado-Filho, G. M. (2015). Sponges Magalhães, G. M., Amado-Filho, G. M., Rosa, M. R., Moura, R. L., Brasileiro, and fish facilitate succession from rhodolith beds to reefs. Bulletin of P., Moraes, F. C., … Pereira-Filho, G. H. (2015). Changes in benthic Marine Science, 91, 45–46. https://doi.org/10.5343/bms.2014.1067 communities along a 0-­60 m depth gradient in the remote St. Peter Pinheiro, H. T., Bernardi, G., Simon, T., Joyeux, J. C., Macieira, R. M., and St. Paul Archipelago (Mid-­Atlantic Ridge, Brazil). Bulletin of Marine Gasparini, J. L., … Rocha, L. A. (2017). Island biogeography of ma- Science, 91, 377–396. https://doi.org/10.5343/bms.2014.1044 rine organisms. Nature, 549, 82–85. https://doi.org/10.1038/ Magris, R. A., & Pressey, R. L. (2018). Marine protected areas: Just for nature23680 show? Science, 360(6390), 723–724. https://doi.org/10.1126/sci- Pinheiro, H. T., Ferreira, C. E. L., Joyeux, J. C., Santos, R. G., & Horta, ence.aat6215 P. A. (2011). Reef fish structure and distribution in a south-­western Maldonado, M., Aguilar, R., Bannister, R. J., Bell, J. J., Conway, K. W., Atlantic Ocean tropical island. Journal of Fish Biology, 79, 1984–2006. Dayton, P. K., … Young, C. M. (2017). Sponge grounds as Key Marine https://doi.org/10.1111/j.1095-8649.2011.03138.x habitats: a synthetic review of types, structure, functional roles, and Pinheiro, H. T., Mazzei, E., Moura, R. L., Amado-Filho, G. M., Carvalho- conservation concerns. In S. Rossi (Ed.), Marine animal forests: the ecol‐ Filho, A., Braga, A. C., … Joyeux, J. C. (2015). Fish biodiversity of the ogy of benthic biodiversity hotspots of the world (pp. 145–183). Basel, Vitória-­Trindade Seamount Chain, Southwestern Atlantic: An up- Switzerland: Springer. https://doi.org/10.1007/978-3-319-21012-4 dated database. PLoS ONE, 10, e0118180. https://doi.org/10.1371/ Marengo, J. A., & Espinoza, J. C. (2016). Extreme seasonal droughts and journal.pone.0118180 floods in Amazonia: Causes, trends and impacts. International Journal Pinheiro, H. T., Rocha, L. A., Macieira, R. M., Carvalho-Filho, A., Anderson, of Climatology, 36(3), 1033–1050. https://doi.org/10.1002/joc.4420 A. B., Bender, M. G., … Floeter, S. R. (2018). South-­western Atlantic Meirelles, P. M., Amado-Filho, G. M., Pereira-Filho, G. H., Pinheiro, H. T., reef fishes: Zoogeographical patterns and ecological drivers reveal de Moura, R. L., & Joyeux, J. C. (2015). Baseline assessment of meso- a secondary biodiversity centre in the Atlantic Ocean. Diversity and photic reefs of the Vitória-­Trindade Seamount Chain based on water Distributions, 24(7), 951–965. https://doi.org/10.1111/ddi.12729 quality, microbial diversity, benthic cover and fish biomass data. PLoS Pyle, R. L., Boland, R., Bolick, H., Bowen, B. W., Bradley, C. J., Kane, C., ONE, 10, e0130084. https://doi.org/10.1371/journal.pone.0130084 … Spalding, H. L. (2016). A comprehensive investigation of meso- Morais, R. A., & Maia, H. A. (2017). Lush underwater forests in meso- photic coral ecosystems in the Hawaiian Archipelago. PeerJ, 4, e2475. photic reefs of the Gulf of Guinea. Coral Reefs, 36(1), 95. https://doi. https://doi.org/10.7717/peerj.2475 org//10.1007/s00338-016-1523-z Riegl, B., & Piller, W. E. (2003). Possible refugia for reefs in times of envi- Morais, J., & Santos, B. A. (2018). Limited potential of deep reefs to serve ronmental stress. International Journal of Earth Sciences, 92, 520–531. as refuges for tropical Southwestern Atlantic corals. Ecosphere, 9(7), https://doi.org/10.1007/s00531-003-0328-9 e02281. https://doi.org/10.1002/ecs2.2281 Rocha, L. A. (2003). Patterns of distribution and processes of speciation Moses, C. S., Helmle, K. P., Swart, P. K., Dodge, R. E., & Merino, S. (2003). in Brazilian reef fishes. Journal of Biogeography, 30(8), 1161–1171. Pavements of Siderastrea radians on Cape Verde reefs. Coral Reefs, https://doi.org/10.1046/j.1365-2699.2003.00900.x 22, 506. https://doi.org/10.1007/s00338-003-0346-x Rocha, L. A., Pinheiro, H. T., Shepherd, B., Papastamatiou, Y. P., Luiz, Moura, R. L., Amado-Filho, G. M., Moraes, F. C., Brasileiro, P. S., Salomon, O. J., Pyle, R. L., & Bongaerts, P. (2018). Mesophotic coral ecosys- P. S., Mahiques, M. M., … Yager, P. L. (2016). An extensive reef system tems are threatened and ecologically distinct from shallow water at the Amazon River mouth. Science Advances, 2, e1501252. https:// reefs. Science, 361(6399), 281–284. https://doi.org/10.1126/science. doi.org/10.1126/sciadv.1501252 aaq1614 Moura, R. L., Secchin, N. A., Amado-Filho, G. M., Francini Filho, R. B., Rocha, L. A., Rosa, I. L., & Feitoza, B. (2000). Sponge-­dwelling fishes of Freitas, M. O., Minte-Vera, C. V., … Bastos, A. C. (2013). Spatial Northeastern Brazil. Environmental Biology of Fishes, 59(4), 453–458. patterns of benthic megahabitats and conservation planning in the https://doi.org/10.1023/A:1026584708092 Abrolhos Bank. Continental Shelf Research, 70, 109–117. https://doi. Rosa, M. R., Alves, A. C., Medeiros, D. V., Coni, E. O. C., Ferreira, C. M., org/10.1016/j.csr.2013.04.036 Ferreira, B. P., … Francini-Filho, R. B. (2016). Mesophotic reef fish Mumby, P. J., & Anthony, K. R. N. (2015). Resilience metrics to inform assemblages of the remote St. Peter and St. Paul’s Archipelago, ecosystem management under global change with application to Mid-­Atlantic Ridge, Brazil. Coral Reefs, 35(1), 113–123. https://doi. coral reefs. Methods in Ecology and Evolution, 6, 1088–1096. https:// org/10.1007/s00338-015-1368-x doi.org/10.1111/2041-210X.12380 Rossi, S. (2013). The destruction of the “animal forests” in the oceans: Olavo, G., Costa, P. A. S., Martins, A. S., & Ferreira, B. P. (2011). Shelf-­ Towards an over-­simplification of the benthic ecosystems. Ocean edge reefs as priority areas for conservation of reef fish diversity and Coastal Management, 84, 77–85. https://doi.org/10.1016/j. in the tropical Atlantic. Aquatic Conservation: Marine and Freshwater ocecoaman.2013.07.004 Ecosystems, 21(2), 199–209. https://doi.org/10.1002/aqc.1174 Santos, C. S. M., Lino, J. B., Veras, P. C., Amado-Filho, G. M., Francini-Filho, Peluso, L., Tascheri, V., Nunes, F. L. D., Castro, C. B., Pires, D. O., R. B., Motta, F. S., … Pereira-Filho, G. H. (2016). Environmental licens- & Zilberberg, C. (2018). Contemporary and historical oceano- ing on rhodolith beds: Insights from a worm. Natureza & Conservação, graphic processes explain genetic connectivity in a Southwestern 14(2), 137–141. https://doi.org/10.1016/j.ncon.2016.06.002 Atlantic coral. Scientific Reports, 8, 2684. https://doi.org/10.1038/ Semmler, R. F., Hoot, W. C., & Reaka, M. L. (2017). Are mesophotic coral s41598-018-21010-y ecosytems distinct communities and can they serve as refugia for Pereira, P. H. C., Santos, M., Lippi, D. L., & Silva, P. (2016). Ontogenetic shallow reefs? Coral Reefs, 36(2), 433–444. https://doi.org/10.1007/ foraging activity and feeding selectivity of the Brazilian endemic s00338-016-1530-0 parrotfish Scarus zelindae. PeerJ, 4, e2536. https://doi.org/10.7717/ Shlesinger, T., Grinblat, M., Rapuano, H., Amit, T., & Loya, Y. (2018). peerj.2536 Can mesophotic reefs replenish shallow reefs? Reduced coral 268 | SOARES et al. reproductive performance casts a doubt. Ecology, 99(2), 421–437. Marine Sciences, 74(9), 2309–2320. https://doi.org/10.1093/icesjms/ https://doi.org/10.1002/ecy.2098 fsx085 Shoham, E., & Benayahu, Y. (2017). Higher species richness of octo- UNESCO (2017). Brazilian Atlantic Islands: Fernando de Noronha and corals in the upper mesophotic zone in Eilat (Gulf of Aqaba) com- Atol das Rocas reserve. Retrieved from http://whc.unesco.org/en/ pared to shallower reef zones. Coral Reefs, 36(1), 71–81. https://doi. list/1000 org/10.1007/s00338-016-1528-7 Vale, N. F., Amado-Filho, G. M., Braga, J. C., Brasileiro, P. S., Karez, C. Silva, M., Etnoyer, P. J., & MacDonald, I. R. (2016). Coral injuries observed S., Moraes, F. C., … Moura, R. L. (2018). Structure and composition at Mesophotic reefs after the Deepwater Horizon oil discharge. Deep of rhodoliths from the Amazon River mouth, Brazil. Journal of South Sea Research Part II: Topical Studies in Oceanography, 129, 96–107. American Earth Sciences, 84, 149–159. https://doi.org/10.1016/j. https://doi.org/10.1016/j.dsr.2.2015.05.013 jsames.2018.03.014 Silva, L. L. N., Gomes, M. P., & Vital, H. (2018). The Açu reef morphology, Wirtz, P., Bingeman, J., Bingeman, J., Fricke, R., Hook, T. J., & Young, distribution, and inter reef sedimentation on the outer shelf of the J. (2017). The fishes of Ascension Island, central Atlantic Ocean – NE Brazil equatorial margin. Continental Shelf Research, 160, 10–22. New records and an annotated check-­list. Journal of the Marine https://doi.org/10.1016/j.csr.2018.03.011 Biological Association of the United Kingdom, 97, 783–798. https://doi. Simon, T., Pinheiro, H. T., Moura, R. L., Carvalho-Filho, A., Rocha, L. A., org/10.1017/S0025315414001301 Martins, A. S., … Joyeux, J. C. (2016). Mesophotic fishes of the Abrolhos Wirtz, P., Ferreira, C. E. L., Floeter, S. R., Fricke, R., Gasparini, J. L., Shelf, the largest reef ecosystem in the South Atlantic. Journal of Fish Iwamoto, T., … Schliewen, U. K. (2007). Coastal fishes of São Tomé Biology, 89(1), 990–1001. https://doi.org/10.1111/jfb.12967 and Príncipe islands, Gulf of Guinea (Eastern Atlantic Ocean) – An Slattery, M., Lesser, M. P., Brazeau, D., Stokes, M. D., & Leichter, J. J. update. Zootaxa, 1523, 1–48. (2011). Connectivity and stability of mesophotic coral reefs. Journal Yates, K. L., Schoeman, D. S., & Klein, C. J. (2015). Ocean zoning for conser- of Experimental Marine Biology and Ecology, 408(1–2), 32–41. https:// vation, fisheries and marine renewable energy: Assessing trade-­offs doi.org/10.1016/j.jembe.2011.07.024 and co-­location opportunities. Journal of Environmental Management, Smith, T. B., Gyory, J., Brandt, M. E., Miller, W. J., Jossart, J., & Nemeth, 152(1), 201–209. https://doi.org/10.1016/j.jenvman.2015.01.045 R. S. (2016). Caribbean mesophotic coral ecosystems are unlikely Zibrowius, H., Wirtz, P., Nunes, F. L. D., Hoeksema, B. W., & Benzoni, climate change refugia. Global Change Biology, 22(8), 2756–2765. F. (2017). Shallow-­water scleractinian corals of Ascension Island, https://doi.org/10.1111/gcb.13175 Central South Atlantic. Journal of Marine Biological Association Soares, M. O., Davis, M., & Carneiro, P. B. M. (2018). Northward range of United Kingdom, 97(4), 713–725. https://doi.org/10.1017/ expansion of the invasive coral (Tubastraea tagusensis) in the south- S0025315414001465 western Atlantic. Marine Biodiversity, 48(3), 1651–1654. https://doi. org/10.1007/s12526-016-0623-x Soares, M. O., Davis, M., Paiva, C. C., & Carneiro, P. B. M. (2018). B I O S K E TC H Mesophotic ecosystems: Coral and fish assemblages in a tropical marginal reef (northeastern Brazil). Marine Biodiversity, 48(3), 1631– Marcelo de Oliveira Soares obtained her PhD at the Federal 1636. https://doi.org/10.1007/s12526-016-0615-x University of Rio Grande do Sul (Brazil) in 2010 and completed Soares, M. O., Lotufo, T. M. C., Vieira, L. M., Salani, S., Hajdu, E., Matthews- postdoctoral studies at the Autonomous University of Barcelona in Cascon, H., … Kikuchi, R. K. P. (2017). Brazilian marine animal forests: A New world to discover in the Southwestern Atlantic. In S. Rossi (Ed.), 2016. Currently, he is a Professor in the Institute of Marine Sciences Marine animal forests: The ecology of benthic biodiversity hotspots of the (Labomar), Federal University of Ceará (Brazil). Her research fo- world. Major Reference Book (pp. 73–110). New York, NY: Springer. cuses on the dynamics of tropical ecosystems (coral reefs and man- Soares, M. O., & Lucas, C. C. (2018). Towards large and remote pro- groves) and their responses to global environmental change. tected areas in the South Atlantic Ocean: St. Peter and St. Paul′s Archipelago and the Vitória-­Trindade Seamount Chain. Marine Policy, Author contributions: M.O.S conceived the original idea. Data anal- 93, 101–103. https://doi.org/10.1016/j.marpol.2018.04.004 ysis and paper writing was done by M.O.S., T.C.L.T and P.B.M.C. Soares, M. O., Rossi, S., Martins, F. A. S., & Carneiro, P. B. M. (2017). The forgotten reefs: Benthic assemblage coverage on a sandstone reef (Tropical South-­western Atlantic). Journal of the Marine Biological S U P P O R T I N G I N FO R M AT I O N Association of the UK, 97, 1585–1592. https://doi.org/10.1017/ S0025315416000965 Additional supporting information may be found online in the Spalding, M. D., Fox, H. E., Allen, G. R., Davidson, N., Ferdaña, Z. A., Supporting Information section at the end of the article.   Finlayson, M., … Robertson, J. (2007). Marine ecoregions of the world: A bioregionalization of coastal and shelf areas. BioScience, 57, 573–583. https://doi.org/10.1641/B570707 Spalding, M. D., Ravilious, C., & Green, E. P. (2001). World atlas of coral How to cite this article: Soares MDO, Tavares TCL, Carneiro reefs. Berkeley, CA: University of California Press. PBDM. Mesophotic ecosystems: Distribution, impacts and Tornabene, L., van Tassell, J. L., Robertson, D. R., & Baldwin, C. C. (2016). conservation in the South Atlantic. Divers Distrib. 2019;25: Repeated invasions into the twilight zone: Evolutionary origins of 255–268. https://doi.org/10.1111/ddi.12846 a novel assemblage of fishes from deep Caribbean reefs. Molecular Ecology, 25, 3662–3682. https://doi.org/10.1111/mec.13704 Turner, J. A., Babcock, R. C., & Kendrick, G. A. (2017). Deep thinking: A systematic review of mesophotic coral ecosystems. ICES Journal of