Keywords

Introduction

The deep continental margin , extending from 140 to 3500 m depth and covering approximately the 11% of the seafloor, exhibits particular biotic and abiotic characteristics as well as a remarkable habitat heterogeneity and biodiversity (Levin and Sibuet 2012; Menot et al. 2010).

Different papers published along the last thirty years, mainly in the framework of the Census of Marine Life (COMARGE Project), have highlighted the main deep-sea characteristics (Gage and Tyler 1991; Herring 2002; Wefer et al. 2003; Carney 2005; Brandt et al. 2007a, b; Levin and Dayton 2009; Levin et al. 2010; Menot et al. 2010; Ramírez-Llodra et al. 2010), changing the old perceptions on the continental slope as a monotonous and stable habitat . So, the time-stability hypothesis (Sanders 1969) was challenged by new findings promoted by the implementation of many international programs and huge advances in sampling technology and now the deep-seas are regarded as a dynamic part of the global biosphere (Rex and Etter 2010) that harbors many different ecosystems driven by their geomorphological , geochemical and hydrographic features (Levin and Sibuet 2012).

But, although deep-water research began in the late 19th century and despite the increasing threats to these ecosystems, most of deep-sea habitats and species, including those of continental margins, remain still unknown (Ramirez-Llodra et al. 2010; Levin and Sibuet 2012). This is especially evident in Northwest Africa , where the benthic ecosystems, mainly in waters deeper than 30 m, are reported among the most unknown worldwide (Decker et al. 2004). In fact, Northwest Africa was not included within the analysed regions in the last reviews of global knowledge on marine biodiversity accomplished within the Census of Marine Life (Costello et al. 2010; Levin and Sibuet 2012).

Since the nineteenth century many early and emblematic expeditions, such as the Challenger , Travailleur and Talisman , Chazalie, Hollande VII, Prince Albert I of Monaco, Michael Sars, Atlantide and the Calypso, have visited Northwest Africa including Mauritanian waters, and important collections of fishes and invertebrates were obtained in this area. The scientific results were included in the reports of the expeditions but also in others papers, many of them with regional scope, published over a century and a half, which makes an exhaustive review almost impossible. An extensive bibliography was compiled by Le Loeuff and Von Cosel (1998) in the only review paper that analyzes the biodiversity patterns of benthic fauna on the tropical Atlantic coast of Africa. The zoobenthos of upwelling areas off NW Africa was also reviewed by Thiel (1982).

Despite the first survey being accomplished in 1905 (Gruvel 1905), the current knowledge on Mauritanian benthos remains scarce and fragmentary. In most cases the previous works were only focused on some biological or ecological aspects of particular commercial species from fishing grounds on coastal areas , as Galgos Bay and Cape Blanc region (Maigret 1980; Diop 1988a, b; Mint 1987; Ly 2009; Ould Baba 2010).

The knowledge on Mauritanian deep-sea benthos is mostly related with the surveys carried out in deep-shelf and slope off West African coast . The fifth cruise of President Theodore Tissier was conducted in 1936, from Villa Cisneros to Sierra Leone, and the cruises of the Belgian school ship Mercator worked from Cape Boujador to Conakry from 1934 to 1936. The area stretching from the Canary Islands to Tamxat (northern Senegal River) , including the Banc d’Arguin and surrounding waters, was explored by the Thalassa in 1962, 1968 and 1971, conducting trawling and dredging operations to 1200 m depth. The results of these surveys provided the description of the main benthic biocenosis and also the first distribution maps from Mauritanian soft-bottom s (Maurin 1968; Bonnet et al. 1971). The CANCAP project was a Dutch long-term program (1974–1989) focused on taxonomical and biogeographical studies of the benthos from coastal to deep-waters (0–4000 m) in the Macaronesian region (den Hartog 1984; van der Land 1987) and during the CANCAP-III survey the northern coasts of Mauritania were extensively sampled (van der Land 1987).

The most comprehensive knowledge of the benthos of Mauritania comes from the Tyro Mauritania- II expedition developed in 1988 on the continental shelf and slope (16–1900 m depth) of the northern area; this study was devoted to clarify the functioning of the Banc d’Arguin ecosystem and its interactions with the open ocean systems (van der Land 1988). The main results have been compiled in a special volume of Hydrobiologia (Wolff et al. 1993) and those obtained by Duineveld et al. (1993) about composition and biomass distribution of trawl megafauna , and by van Soest (1993) on the sponges should be emphasized. Taxonomic and biogeographical results of the CANCAP and Tyro Mauritania- II expeditions resulted on the publication of more than one hundred papers, and the study of these collections is still continuing. Some of these references are listed in the faunistic chapters in this book.

The EUMELI project, developed between 1989 and 1992, was focused on the study of particle flow in three ocean sites at Cape Blanc latitude, from a high productivity coastal upwelling system to a low productivity oceanic area (Auffret et al. 1992; Morel et al. 1996); the benthic communities were also sampled during the EUMELI 2 survey (Sibuet et al. 1993; Cosson et al. 1997; Galéron et al. 2000).

However, none of the above cited reports provide an integrated approach to the composition , structure and distribution of benthic communities in Mauritanian waters.

During the four Maurit surveys , accomplished by the Spanish Institute of Oceanography (IEO) and the Mauritanian Institute of Oceanographic and Fisheries Research (IMROP) between 2007 and 2010 (Maurit-1107, Maurit-0811, Maurit-0911 and Maurit-1011), 291 trawling stations were carried out in deep-shelf and slope (Hernández-González et al. 2008; Ramos et al. 2010). The identification of the main benthic invertebrate groups and the analysis of data obtained during the surveys allowed us to perform a first overview on the biodiversity and distribution of the megabenthos inhabiting deep-waters off Mauritania.

Materials and Methods

Benthos Sampling on-board

During November-December from 2007 to 2010 four research cruises focused on the prospection of demersal resources and the study of benthic communities of sedimentary bottom s were carried out on-board the Spanish R/V Vizconde de Eza at the continental margin off Mauritania. A total of 291 stations were sampled with an otter trawl (Lofoten type) on the deep-shelf and continental slope at depths between 79 and 1867 m, following a random stratified sampling procedure (see Chaps. 1 and 4 for a detailed description of methodology , stations data and map).

At each station, megabenthos was sorted into high-range taxa . Subsequently, a preliminary identification to species or morphospecies level was accomplished and data on abundance and biomass (wet weigh) were also obtained. Numerical abundance (individuals or colonies) and wet weight of all benthic taxa were obtained from the total catch or from a subsample, when the large volume of the capture made exhaustive sorting impossible. In this case, total abundance and biomass were estimated by a weighting coefficient . At each station, photographs of the total catch, as well as all benthic invertebrate species and some of their anatomical details were taken. Finally, a representative collection was fixed and preserved in 70% ethanol or 4% formaldehyde solutions for further taxonomic study .

Geographical position and depth were obtained for every station; near-bottom temperature and conductivity were recorded in 189 hauls with a net sensor SB37, and sediment parameters (grain size , organic matter and carbonates content) were analysed in 60 stations (see Chap. 2 for detailed methodology).

Abundance , biomass and specific richness were spatially represented by geo-statistical techniques using ArcGIS krigging and a spherical semivariogram, with 12 neighbours, and the output cell size of 1000 m.

Statistical Analyses

Matrices of numerical abundance and weight by (morpho) species and station were standardized to swept area and expressed by 0.1 km2, area which was close to the real surface swept in each individual trawling (see detailed methodology in Chaps. 1 and 4). Pelagic cephalopods were removed from the matrices.

Univariate diversity descriptors were estimated on untransformed data using DIVERSE routine of PRIMER 6 (Clarke and Warwick 2001). For each station, numerical abundances (N), total number of species (S), ln-transformed Shannon-Wiener diversity index (H′) and Pielou’s evenness index (J′) were estimated.

In order to establish possible large-scale biodiversity patterns , averages of main diversity variables were analysed by one degree latitude and 200 m depth strata ; species richness , Shannon-Wiener and Pielou indexes , abundance and biomass trends across latitudinal and depth gradients were tested by regression fitting and variance analyses.

A Bray-Curtis similarity matrix was obtained transforming the biomass by species (208 species) and station (283 stations) by fourth-root and subsequently used for the application of multivariate methods. Species with low biomass (<1 kg) were removed from the matrix. Hierarchical clustering analysis , using the group average sorting algorithm, and a non-metric multidimensional scaling (MDS) were subsequently performed to analyse faunistic similarity among localities (Clarke and Warwick 1994). The SIMPER routine was used to identify species that could potentially discriminate station groups or assemblages . An analysis of similarities (ANOSIM) was performed in order to check the statistical robustness of the differences in the faunistic composition between areas or assemblages (β diversity) .

BIOENV procedure, following Clarke and Ainsworth (1993), was applied to a matrix of 70 stations for which the datasets of 9 environmental variables were available, with the aim to test their correlation with megabenthos distribution . The variables considered in this analysis were: latitude, longitude, water depth (m), near-bottom temperature (°C), gravel (% > 2 mm), sand (% 2–0.0625 mm), mud (% < 0.0625 mm), organic matter (OM %) and carbonate content (CaO3 %) of the sediments.

All statistical procedures were performed using the PRIMER software package v 6 (Plymouth Routines in Multivariate Ecological Research) (Clarke and Gorley 2006).

Results

Global Overview

A total of 551,281 specimens, included in 37 high-range taxa (phylum, class, order), were collected at 283 valid trawling stations in the deep-shelf and slope off Mauritania. The total corresponding wet mass was 2.2 tons.

Decapoda showed the highest occurrence , collected at all stations but one (99.6%), followed by benthic Cephalopoda (233 stations, 83.2%), and by Actiniaria , Hydrozoa, Echinoidea and Ophiuroidea , with rates over 50% (Fig. 7.1).

Fig. 7.1
figure 1

Occurrence (percentage of stations) of 37 main macrobenthic taxa in Mauritanian deep-waters (main exclusively suspension-feeders groups are represented in light colour)

Full size image

The best-represented group in terms of species richness was also Decapoda, represented in Mauritanian deep-waters by 118 species, followed by Polychaeta , Porifera and Hydrozoa , which oscillated from 74 to 63 species (Fig. 7.2). We want to stress that the taxonomic work is still in progress and that the final figures, once the identification will be accomplished, may display some differences. Among the currently identified taxa, some new species of crustaceans were described (de Matos-Pita and Ramil 2014, 2015a, b, c) and the description of other new species will be published shortly.

Fig. 7.2
figure 2

Specific richness (S, preliminary number of species) of the 37 main taxa composing the macrobenthos of Mauritanian deep waters (dark colour taxa already identified to species level)

Full size image

Regarded as a whole, the megabenthos of the deep-shelf and slope off Mauritania seems clearly dominated by detritivorous holothurids , which constituted the 83% of total biomass , sharing the numerical abundance with the decapods (34 and 39%, respectively) (Fig. 7.3). Benthothuria funebris and Paelopatides grisea , due to its large size, together with the small but highly abundant benthopelagic Enypniastes eximia, are responsible for these dominance patterns in biomass and abundance, respectively. Holothurians were followed by the polychaetes in terms of abundance (12%) and the decapods in terms of biomass (8%). The remaining taxa are much less represented (<5% both in abundance and biomass). Suspension-feeders were represented by approximately 280 species (38.4% of total of estimated γ diversity) , but these constituted only the 9.5% of abundances and did not even reach 1% of the total biomass . In the pictures of Figs. 7.4, 7.5, 7.6 and 7.7 are displayed the most representative species of Mauritanian megabenthos.

Fig. 7.3
figure 3

Global faunistic composition , numerical abundances and biomass standardized to 0.1 km2 (in %) of deep-waters Mauritanian megabenthos (in green main and exclusively filter- and suspension-feeders taxa)

Full size image
Fig. 7.4
figure 4

Some of the most representative megabenthic species in the deep-shelf assemblage off Mauritania: Octopus vulgaris (1), Sepia elegans (2–3), Munida speciosa (4), Macropipus rugosus (5), Parapenaeus longirostris (6), Nemertesia sp (7), Centrostephanus longispinus (8) and Parastichopus regalis (9). The main small-size components of suspension-feeders community over the tubes of Chaetopteridae polychaete (10–11), alcyonids (12), ascidians (13), sponges (14), bryozoans (15) and actinians (16) (© Ana Ramos)

Full size image
Fig. 7.5
figure 5

Slope assemblages : Decapods are certainly the most diverse group and the best numerically represented in deep assemblages of Mauritanian slope . In the pictures: the striped red shrimp, Aristeus varidens , a species of great commercial value (1), Glyphus marsupialis (2), Acanthephyra pelagica (3), Nematocarcinus africanus (4), Stereomastis talismani (5), Solenocera africana (6) and Pasiphaea semispinosa (7) (© Ana Ramos)

Full size image
Fig. 7.6
figure 6

Slope assemblages : Echinoderms, mainly holothurids , constituted the most representative taxa in number and above all in biomass in the slope assemblages. Huge amounts of the sea-cucumbers Enypniastes eximia (1, 2, 4) and Benthothuria funebris of up to 6 kg weight (2, 3, 5), together Paroriza pallens (6) were caught after 1000 m depth. Also sea-urchins belonging to Echinothuroidea order, particularly the species Phormosoma placenta (7, 8), besides Hygrosoma petersii (9) and Calveriosoma hystrix (10) are very abundant between 1000 and 2000 m depth. Other important components in deep assemblages: the sea-stars Psilaster cassiope (11), Hymenaster roseus (12), Pseudarchaster gracilis gracilis (13), Brisinga sp (14) and Zoroaster fulgens (15), and the brittle-stars Ophiomusium lymani (16), Ophiura flagellata (17) and Ophiernus alepidotus (18) (© Ana Ramos)

Full size image
Fig. 7.7
figure 7

Slope assemblages : Species belonging to others different taxa are also very characteristics of Mauritanian deep-waters. Among them are found some actinians belonging to Hormathiidae family (1, 2) and the barnacles of Scalpellidae family (3, 4)—still in identification process—the zoanthid Epizoanthus paguriphilus associated to the hermit-crab Parapagurus pilosimanus (5, 6), polychaetes (7) and the opisthobranch mollusc of Scaphander genus (8) (© Ana Ramos)

Full size image

Biodiversity Distribution Patterns

Global richness of megabenthos in the deep-shelf and slope off Mauritania (γ diversity) , includes more than 700 species with an average richness of 22 species by station. Local values of specific richness (α diversity) exhibit a wide range, oscillating from 59 to only two species (see distribution map in Fig. 7.8). However, the cumulative curves (original and predicted) obtained by using bootstrap index do not approach an asymptote despite the high sample numbers considered—almost 300, corresponding to a 30 km2 sampling area—(Fig. 7.9).

Fig. 7.8
figure 8

Geographical distribution of specific richness (S) (left), numerical abundances (N) (centre) and biomass (B, in kg) (right) (quantitative data by station standarized to a 0.1 km2 swept area)

Full size image
Fig. 7.9
figure 9

Accumulation curve of the observed species richness (in blue, Sob) and estimated cumulative species richness by bootstrap index (in red). Cumulative curves are plotted both in relation to the number of samples and total swept surface (in km2)

Full size image

Shannon-Wiener diversity index showed in general low global and local values (mean H′ = 1.494; minimum = 0.017; maximum = 3.180), as well as Pielou evenness index with most values around 0.502, a minimum value of 0.008 and a maximum of 0.962. Mean abundances and biomass by station were 2032 specimens and 69 kg, even though strong variations were observed between stations: from 27,700 specimens and 1370 kg, to only few individuals and less than 1 kg (Fig. 7.8).

Biomass showed highest values in a wide area south of Nouakchott between 1500 and 2000 m depth (Fig. 7.8) coinciding with the highest numerical abundances (Fig. 7.8); similar high densities, but not with such high biomass , were also observed at the same depth in a less extensive area north of Nouakchott, and areas with high density but low biomass were reported in the deep-shelf and upper slope off Nouakchott and close to the Senegalese border . The biomass exhibits the lowest values all along the deep-shelf, upper- and middle slope (to 1000 m depth approximately); densities showed the lowest figures at the same depths, but only at Nouackchott latitude, with some intermediate values at Cape Blanc and north of Cape Timiris. In deep-waters off Cape Blanc minimum values for both, densities and biomass, were found (Fig. 7.8).

Specific richness exhibits a different pattern (Fig. 7.8), with maximum values in the deep-shelf and upper slope at Cape Blanc, Banc d’Arguin and Tioulit canyon , but also in the deep-slope between 1500 and 2000 m depth, off Cape Timiris.

The analysis of mean values of abundance , biomass and diversity indexes showed interesting latitudinal and bathymetric patterns along the Mauritanian continental margin for the global megabenthos. The average values of diversity descriptors represented by one degree latitude intervals showed evident patterns for all variables (Fig. 7.10). Even though the average of specific richness remains around 20–25 species at all latitudes, Shannon and Pielou indexes exhibit a decreasing trend from the northern zone, at Cape Blanc—Banc d’Arguin latitude (20°N)—to the southern part of Mauritania (16°N). Abundances and biomass displayed a reversed trend, increasing southwards, three times (from 1065 to 3181 specimens) and almost 8 times (from 17 to 134 kg), respectively. Specific richness excluded, the high values of regression coefficient (R2 > 0.9), as well as the values of ρ < 0.05 obtained in the variance analyses by one degree latitude (Table 7.1) confirm that the latitudinal-related trends along the Mauritanian continental slope are statistically significant.

Fig. 7.10
figure 10

Average values and confidence intervals (±95%, α = 0.05) by latitude degree of main variables and diversity indexes with superimposed fitting lines and respective regression coefficients

Full size image
Table 7.1 Results of variance analysis (t-test) for the various diversity indexes between one latitude degree areas
Full size table

Bathymetrical variation of mean values of the same biotic variables plotted by 200 m depth intervals is represented in Fig. 7.11. Richness , Shannon and Pielou indexes show relatively small fluctuations until 1300 m depth, with an initial decreasing to 700 m, followed by an increasing trend. From 1300 m onwards Shannon and evenness showed a drastic fall, reaching their minimum at 1700 m while richness maintained the increasing trend to a maximum at 1900 m depth.

Fig. 7.11
figure 11

Average values by depth (200 m interval) of main variables and diversity indexes

Full size image

The distribution patterns of abundances and biomass are clearly different until the mid-slope: while abundances increased between 100 and 300 m and dropped gradually to 1300 m, biomass showed its minimum values to 900 m. Both variables increased in deep-waters, reaching their maxima at 1700 m depth (Fig. 7.11). All variables considered follow almost identical bathymetric patterns in the Northern, Central and Southern areas (Fig. 7.12). The biomass, the variable which integrates the community structure the best, goes from insignificant values, not surpassing 16 kg by 0.1 km2 between 80 and 900 m, to mean values of 373 kg (20 times higher!) at 1700 m. The variance analysis carried out for individual values of biomass by depth strata pointed out the significance of this relationship (p < 0.05, Table 7.2). This significant relationship between biomass and depth, was also checked by the fitting of Maurit-0811 biomass data by station to depth2 (Fig. 7.13) which shows a good regression coefficient (R2 = 0.59).

Fig. 7.12
figure 12

Mean values of different diversity indices by depth strata in Northern (blue), Central (red) and Southern areas (green)

Full size image
Table 7.2 Results of variance analysis (t-test) for individual values of biomass by depth strata
Full size table
Fig. 7.13
figure 13

Dispersion graphics of biomass by station during Maurit-0811 (96 stations) in relationship to depth, with superimposed lines and exponential regression coefficient

Full size image

Community Structure

The multivariate analysis of biomass datasets (species with standardized biomass >1 kg, 283 stations) identifies the main assemblages along the Mauritanian coast. The MDS ordination plot (Fig. 7.14), with a stress value of 0.1, discriminates two major groups: the first including deep-shelf and upper slope stations (80–440 m) and the second those located at middle and lower slope (460–1900 m). Stations are segregated by depth along the first MDS axis (Fig. 7.14). The two main assemblages are separated by a spatial gap in which only 8–10 stations located between 420 and 550 m are represented. The dendrogram (Fig. 7.15) confirms the MDS results, but also identifies four groups within the two main clusters . In the first cluster, group 1 includes the shallowest stations at the deep-shelf (80–180 m) and group 2, those located at the upper slope (180–440 m). Within the deep cluster, group 3 is composed by the stations found at the middle-slope (460–1200 m) and group 4 by the ones at lower-slope (1200–1900 m).

Fig. 7.14
figure 14

Two-dimensional multidimensional scaling (MDS) plot of the biomass data of megabenthos obtained in 281 trawl stations during Maurit surveys with superimposed values of depth, organic matter and biomass (grey deep-shelf and upper slope stations; black middle and deep-slope)

Full size image
Fig. 7.15
figure 15

Dendrogram resulting from multivariate analysis based on biomass matrix of megabenthos (biomass by station standardized to 0.1 km2 swept area , \(\surd\surd\) transformed data, Bray-Curtis similarity index)

Full size image

The results of the SIMPER analysis for all groups are shown in Table 7.3 (dissimilarity) and Table 7.4 (similarity) . Similarity within each group ranges from 25.60% (Deep-Shelf assemblage) to 34.36% (Middle-Slope assemblage) , while dissimilarity among groups oscillates from 81.6 to 99.9%. The four assemblages were significantly different (ANOSIM Global R = 0.8, ρ < 0.001, all pairwise ρ < 0.001) in terms of community structure (Table 7.5).

Table 7.3 Summarized results of the SIMPER analysis: Average of dissimilarity between the four main assemblages
Full size table
Table 7.4 Results of similarity percentages (SIMPER analysis) within the four megabenthic assemblages
Full size table
Table 7.5 ANOSIM test results: R values between the four assemblages identified by the cluster analysis, at a significance level of 0.1%
Full size table

In the deep-shelf assemblage cephalopods , represented by Octopus vulgaris (34.5%) and Sepia elegans (10.0%), are the main contributors to the similarity and together with the decapods Munida speciosa (12.0%), Macropipus rugosus (6.5%) and Parapenaeus longirostris (6.4%), the hydrozoan Sertularella gayi gayi (4.4%) and the sea-urchin Centrostephanus longispinus (4.1%) reach the 78.1% (Fig. 7.4).

The upper and middle slope assemblages are characterized by decapods, which account for the 79.0 and 63.6% of similarity respectively. Five species, Parapenaeus longirostris, Plesionika heterocarpus, Solenocera africana , Munida speciosa and Pasiphaea semispinosa , were the main contributors to the similarity in the upper slope . The middle slope assemblage is typified by seven different species, Aristeus varidens , Glyphus marsupialis , Acanthephyra pelagica, Hymenopenaeus chacei, Nematocarcinus africanus , Stereomastis talismani and Systellaspis debilis , accounting for the 70.9% of similarity with the contribution of two other non-decapod species, Phormosoma placenta (3.9%) and Benthoctopus sp (3.4%) (Fig. 7.5).

In the deep slope assemblage the accumulation of 13 species was necessary to reach the 70.2% of similarity . The most important contributors to this cluster were the holothurids Enypniastes eximia (13.6%) and Benthoturia funebris (7.9%), which together with Paelopatides grisea (5.6%) account for 27.5% of similarity (Fig. 7.6). Seven decapods, A. pelagica, Pasiphaea tarda , G. marsupialis, Neolithodes asperrimus , S. talismani, H. chacei and Benthesicymus bartletti , two actinians , Phelliactis sp and Actiniaria indet , and one sea-star, Pseudarchaster gracilis gracilis , are the remaining contributors (Figs. 7.5, 7.6 and 7.7).

Table 7.6 shows the mean values of environmental factors and biotic variables for each assemblage, including their increasing or decreasing trends in the right column (arrows). The abiotic factors , except for the carbonate contents, are depth-related, with positive correlation for mud and organic matter and negative for bottom temperature, salinity , gravel and sand . In Chap. 2, distribution maps of carbonates and organic matter contents , and a synthesis of the sediment size-grain composition found, are showed.

Table 7.6 Mean values of environmental factors and ecological indexes for the four megabenthos assemblages identified by the multivariate analysis
Full size table

The BIOENV results (Table 7.7) showed the highest correlation between environmental variables and faunal biomass for the bottom temperature (ρ = 0.857). Depth and bottom temperature (ρ = 0.825) would be the main factors structuring the megabenthos assemblages off Mauritania, since both offered the best correlation figures in all abiotic variables matches.

Table 7.7 Results of BIOENV analysis: Best matches of biotic and abiotic similarities matrices for each combination of variables (No)
Full size table

Discussion

About the Methodology

As Menot et al. (2010) pointed out, research expeditions on the continental margin are rarely specifically focused on biodiversity studies or taxonomy . The sampling of the benthos in deep waters is more difficult than in shallow waters because it requires specialized and powerful oceanographic vessels, large and heavy equipment and special skills; moreover, handling a sampling device placed at the end of several kilometers of wire is time-demanding and has high failure rates (Eleftheriou and Holme 1984; Gage and Bett 2005). The study of deep-sea megabenthos entails other particular issues linked to the lower densities and relative rarity of many species, which increase the need of more numerous and larger samples. Another gap is the scarcity of quantitative data for the megafauna because, as a result of these low densities, the area sampled and the number of specimens collected by the traditional devices, such as grabs and box-corer, are insufficient and/or inadequate for quantitative estimation (Gage and Tyler 1991). Estimations of abundance and/or biomass of fishes and invertebrate epifauna are very important in order to determine secondary production or consumption rates, as well as for ecosystem modelling (Reiss et al. 2006).

Large commercial otter trawls have been successfully used in deep-waters soft bottom s, at more than 1500 m depth, for demersal fish catch and usually a large amount of benthic megafauna is also collected as by-catch (Gage and Bett 2005; Reiss et al. 2006). Despite trawls being the most effective sampling device for deep-sea megabenthos , commercial trawls have been rarely used for this purpose because the operation of these heavy gears requires specialist expertise and twin warps, which research vessels are not normally equipped with (Gage and Bett 2005). In addition, trawls have not been traditionally considered as a quantitative method (Eleftheriou and Moore 2005) because the abundance of megafauna is underestimated. Thus, Billett (1991) estimated that trawl catches are only about one half to one-fifth of the megafauna , but Bett et al. (2001) showed, comparing trawl and photographic samplings , that the otter trawl has an average fishing efficiency of 68%.

Recently the use of acoustic telemetry systems for the tow monitoring allows knowing accurately the distance covered across the seabed by the gear as well as the swept area . Howell et al. (2002) considered as quantitative samplings the ones accomplished with an epibenthic sledge and as semi-quantitative those carried out with an otter trawl . But otter trawl data were considered useful for comparative purposes (Eleftheriou and Holme 1984; Ramos 1999) and have also been used by Howell et al. (2002) for abundance analyses at stations where epibenthic sledge data were lacking.

The Spanish R/V Vizconde de Eza , devoted to fishery research and equipped with two winches, large commercial trawls and a specialized crew, provides a suitable platform for sampling demersal fishes and the associated megafauna . At the beginning of the African fisheries research program developed by the Spanish Institute of Oceanography (IEO) between 2002 and 2010, a megabenthos sampling protocol was specifically designed and further implemented in all surveys carried out in different countries. Although the Maurit surveys were mainly focused in the prospection of demersal resources , the study of megabenthos was also considered as an important objective during the surveys’ planning (Hernández et al. 2008; Ramos et al. 2010). The benthos program was carried out on-board by the same scientific team, affiliated to IEO and to the University of Vigo and the subsequent taxonomic study of the invertebrate collections and data analysis were accomplished in the framework of the EcoAfrik joint project.

The sampling program carried out in Mauritanian continental margin is undoubtedly exceptional because, in addition to the high number of trawling stations , the use of a swept area method and a stratified sampling design, resulted in a dense coverage of the slope seabed. The use of net sensors measuring its vertical opening coupled with those mounted in the trawl-doors, allowed an accurate trawling monitoring and the standardization of quantitative data. Furthermore, the total surveying of the Mauritanian slope by a multibeam echosounder offers the possibility to further locate and characterize the main deep-sea habitats. In this way, a series of fishing trawling surveys have offered the opportunity to obtain a comprehensive overview of the composition and distribution of the megabenthos , perhaps never obtained before at any other continental margin worldwide. Moreover, this overview could provide the starting point for the monitoring of long-term biodiversity changes in an area where the displacement of demersal fishing fleets to increasingly deeper waters and the oil and gas exploitation are eminent threats to its ecosystems.

A similar methodology , taking advantage of the trawling commercial surveys to improve the knowledge of the megabenthos , has been previously used in African and Antarctic waters (Ramos 1999; Ramos et al. 1991, 2005, 2010).

General Overview

The exceptional richness of the Mauritanian benthos was already highlighted by Bonnet et al. (1971) who, based in the faunistic results obtained during the Thalassa’s surveys, considered this region as one of the most important in the world oceans. Likewise, Le Lœuff and von Cosel (1998), in their review on the biodiversity of Tropical West Africa , noted the Mauritanian-Senegalese coast as one of the richest in the region, as a consequence of its location in the northern zone under the alternating influence of important upwelling phenomena and the seasonal displacement of the Cape Blanc thermal front (see Chap. 3). The first faunistic results of four Spanish-Mauritanian’ surveys, based on the analysis of quantitative data collected in 291 trawl stations, showed that benthic communities of the deep-shelf and continental slope off Mauritania are composed by more than 700 species. This estimation is based only on samples collected on soft bottom s (90% of the seabed between 80 and 2000 m depth), basically constituted by muddy sediments.

The continental margins were considered as a system of high habitat heterogeneity and biodiversity , which is among the highest on the Planet (Levin and Dayton 2009; Menot et al. 2010; Levin et al. 2010). Moreover, the existence of hotspot ecosystems, associated to hard bottom s located in the muddy slopes, strongly enhances the biodiversity in the continental margins at regional scale (Menot et al. 2010). Vulnerable ecosystems of Mauritanian slope—the giant coral mounds barrier , the Arguin and Timiris canyon systems and the small Wolof’s Seamount—extensively described in the corresponding chapters of this volume (see Chaps. 13, 14 and 15), are not considered in this paper. Many benthic species , probably with different environmental requirements or living exclusively associated to hard-bottom habitats , were only recorded in these vulnerable hotspot ecosystems . If these species were included, the γ-diversity of Mauritanian slope would undoubtedly increase, confirming the widespread perception that the continental margins support high biodiversity levels worldwide. However, recently McClain and Schlacher (2015) highlighted that the apparent paradox of high diversity of the deep-sea benthos in a homogeneous, stable and food-scarce environment which should rather inhibit than enhance biodiversity , has to be revisited. But in any case, despite the high number of stations and the extensive sampling coverage, it was not possible to predict the total theoretical species number inhabiting the Mauritanian slope , because in the species accumulation plot the asymptotic values were not reached. Grassle and Maciolek (1992) had already pointed out that this phenomenon occurs in general in deep-sea macrobenthos due to the high percentage of species represented by a single specimen, which mean that a greater sampling effort is required to obtain a comprehensive overview of diversity . This concurs with our observations in the Mauritanian slope since the 40% of species were collected only once and the 56% solely at one or two stations, despite the sampled area in the 291 trawling stations that covered approximately 30 km2. But, we must take into account that this swept area only represents the 0.01% of the total surface of the studied area. Results of previous works developed in other deep-water areas are basically based on box-corer samplings and are not comparable with this study since those correspond to the infaunal macro- and meiobenthos (Hessler and Sanders 1967; Sanders 1968; Grassle and Macioleck 1992; Etter and Grassle 1992; Etter and Mullineaux 2001; Snelgrove and Smith 2002; Stuart et al. 2003). Moreover, Menot et al. (2010) remarked that the difficulty to access such a vast geographical area, together with the high species richness and the taxonomy impediment (decreasing taxonomy expertise and ageing of taxonomists), also make the comparisons of datasets across habitats and regions within the continental margins difficult. Even in our case, and despite the evident advantages of the wide sampling coverage in Mauritanian slope , unusual in deep-sea research , as well as the extensive taxonomic effort engaged to achieve species identifications of major megabenthic groups, comparisons still remain difficult.

The only recent trawling surveys in Mauritanian waters have been carried out with beam-trawl (Duineveld et al. 1993; Sibuet et al. 1993; Galéron et al. 2000); nevertheless their results are not comparable due, not only to methodological differences and the bathymetric range sampled, but also to the different taxonomic levels reached in these studies (see Chap. 8 for discussion).

Large commercial trawls provided with net sensors have been recently and successfully used on-board research vessels for the study of epibenthic and demersal communities in the Cantabrian and Mediterranean seas (Sánchez et al. 2008; Serrano et al. 2011; Cartes et al. 2009; Ramírez-Llodra et al. 2008; Papiol et al. 2012; Fanelli et al. 2013). Despite differences in the number of stations, depth range and taxa considered in these works, even the results of Serrano et al. (2011), which analyse a historical data series (>20 years and >300 trawling stations), only recorded 147 megabenthic species . These figures are clearly lower than those achieved in Mauritania. The results obtained during the Maurit surveys for the macrobenthos (see Chap. 8), as well as for some particular megabenthic taxa , such as decapods and cephalopods (see Chaps. 9 and 10), confirm that Mauritanian diversity is clearly higher than that recorded in deep-waters of other North Atlantic areas . This high diversity in Mauritania is related with the elevated productivity linked to the upwelling conditions in the area, with the high habitat heterogenity in the Mauritanian slope and also with its geographical location in a faunistic transitional region between the Lusitanian and the Tropical Eastern Atlantic Provinces. Its biota benefits from both temperate and tropical faunal components, which also enhances the regional biodiversity (Massin 1993; Van Soest 1993; Le Loeuff and von Cosel 1998; de Matos-Pita 2016).

Biodiversity Distribution Patterns

The relation between diversity and productivity is a scientific subject of central interest in ecology (Mittelbach et al. 2001; Stuart and Rex 2009). Nevertheless studies aimed at understanding how productivity is related to diversity patterns are rare, particularly in the deep-sea where the food limitation is well-known. Diverse hypotheses were proposed in order to explain the productivity -diversity relationship, often inferred from proxy variables such as depth and latitude, but they have been poorly tested up to date (Rex and Etter 2010; McClain and Schlacher 2015).

The two central paradigms about marine diversity distribution patterns , the increasing trend of species richness with depth to a maximum around 2000 m, and the latitudinal cline from poles to tropics, have been established over the last decades (Rex 1981; Rex et al. 1993; Gray 2001). Nevertheless, the conclusions of Census of Marine Life (COMARGE project) point out that the geographical distribution gradients are currently much less understood than depth-related patterns (Menot et al. 2010).

Rex et al. (1993, 2000) and more recently the synthesis work of Rex and Etter (2010) showed the existence of a latitudinal diversity pattern in North Atlantic deep-sea benthos (500–4000 m), which increases from poles to tropic, but it is not as obvious for the South Atlantic where strong interregional variation are observed. Nevertheless these authors pointed out that their results in the North Atlantic are based only on some taxa (bivalves, gastropods and isopods), on a limited samples number and a restricted spatial coverage. The results for the South Atlantic are even scarcer. Preliminary results from Fridjof Nansen surveys pointed out that, at regional level and for the whole megafauna , this cline does not seem consistent along the Northwest African coast (Ramos et al. 2015). Nevertheless, even though diversity does not follow the expected increasing cline southwards and despite that no latitudinal patterns were described for the macrobenthos (see Chap. 8), the results of the most intensive and evenly developed sampling program with the commercial trawl revealed some geographical patterns at regional scale .

Megabenthos diversity is two times higher north of Cape Blanc than in the southern area, reaching the highest values along the Western Saharan slope (Gonzalez-Porto et al. 2007; Ramos et al. 2012, 2015). Our results from the Mauritanian continental margin showed a clear latitudinal pattern for some diversity indexes. Although specific richness remains around 20–25 species per station all along the slope , mean values of Shannon diversity and evenness showed a very significant decreasing trend southwards. These results would support the hypotheses that productivity drives the megafauna diversity in the Northwest African upwelling system . Seasonal input of organic matter to the seabed linked to the seasonal pulses of productivity at high latitudes was invoked as a possible limiting factor for diversity in bathyal soft sediments, opposed to tropical areas where productivity remains constant along the year (Rex et al. 1993, 2000; Culver and Buzas 2000). At Sahara—Cape Blanc latitude, upwelling conditions and its associated high primary production are permanent during the whole year (Pastor et al. 2008; Chap. 3). In addition, the giant and highly productive filament of Cape Blanc which stretches itself 600 km offshore (Sangrá 2015) could be acting as an important exporting mechanism that enhances the food input to deep-sea (Thiel 1982; Duineveld et al. 1993; Ramos et al. 2015). In fact, during the EUMELI project eutrophic conditions have been reported at 1600–2100 m depth off Cape Blanc latitude (Sibuet et al. 1993; Cosson et al. 1997; Galéron et al. 2000). Therefore, the high and permanent productivity throughout the whole year could underlie the highest diversity found off Cap Blanc.

Direct relationships between abundance and biomass of benthic organisms and the input of food to the seabed have been commonly observed for many taxa in regions characterized by high oceanic surface productivity or elevated carbon flux to deep-sea benthos (Gooday 2002; Smith et al. 2008; McClain et al. 2012a, b). Nevertheless, the abundance and biomass values do not seem to reflect the high productivity at the Cape Blanc area. Thus both variables showed the lowest figures at Cape Blanc—Banc d’Arguin latitude (20°N) with a strong increasing trend southwards. Other environmental factors, like sediment slides (see Chap. 2) and the terrigenous inputs of Senegal River , coupled with a lower trawl-fishing pressure in the south, could explain this different pattern.

Despite of this remarkable latitudinal cline , all biotic variables maintain a similar depth-related pattern along the entire slope , recurrent in North, Centre and South areas . That would mean that the differences in the geomorphological features between areas (deep canyons, giant coral mounds barrier , and width shelf) do not seem to play a major role in the community structure of soft-bottom megabenthos. The bathymetric gradient is strongly marked mainly for biomass , very low at shallower depths than 900 m, but sharply increasing up to 1700 m depth. This minimum of biomass found at depths shallower than 900 m could be linked directly to the intensive fishing activity traditionally developed in Mauritania EEZ. In fact, foreign commercial fleets have intensively exploited demersal resources for more than 50 years (FAO 2012a, b), being the Banc d’Arguin zone, the widest shelf area, where historically the highest fishing efforts were concentrated. During the last decade a dangerous displacement of trawling fleets to deeper waters, exactly to 1000 m depth, has also been observed in Mauritania (FAO 2006, 2012a, b) threatening ecosystems whose diversity and functioning we are only starting to understand (Menot et al. 2010).

The extraordinary densities and biomass concentrations in the deepest and southern zone of Mauritanian slope could be related to its proximity to the mouth of Senegal River and the input of organic matter to the seabed moving northwards within the deep-slope Poleward Undercurrent (Peña-Izquierdo et al. 2012; Pelegrí and Peña-Izquierdo 2015). These high values are related to the dominance , in the four surveys, of holothurians , particularly the species Enypniastes eximia Théel, 1882, whose high abundance in the slope had also been recorded in other areas and associated to the instability, turbidity current and slides (Gage and Tyler 1991). In this case, the area coincides exactly with the location of the macro-landslide of Mauritanian slope , partially included in the old delta of Senegal River (see Chap. 2).

Faunistic Composition and Assemblages

The composition of soft-bottom epifauna of Mauritanian slope , contrarily to what one would expect in one of the most productive upwelling areas of the world ocean, shows that suspension-feeders are practically absent and that a general dominance of deposit-feeder, scavenger and carnivorous fauna is observed. Although important communities of sponges and cnidarians, characteristic suspension-feeders taxa in the deep sea (Ramírez-Llodra et al. 2010), were located in deep-waters of Western Sahara and Morocco (Ramil et al. 2005; González‐Porto et al. 2007; Ramos et al. 2015), these groups are virtually absent along the entire Mauritanian slope .

Nevertheless, Thiel (1982) pointed out that 40 years ago large-sized filter feeders, mainly Octocorallia , Crinoidea, Actiniaria and Porifera , constituted dense communities in the deep-shelf and slope along the entire Northwest African upwelling area, particularly off Cape Blanc. Currently, although suspension-feeders diversity seems high in Mauritanian slope (approximately 280 species, the 38.4% of estimated γ diversity) , its contribution to abundances (9.5%) and biomass (1%) in soft bottom s is scarce, similar to values obtained with the beam-trawl (see Chap. 8). The main suspension-feeder groups collected in Mauritanian soft bottoms were not important in weight (e.g. hydrozoans) , or were represented by small sized species (e.g. sponges) . The same results were obtained for the sponge’s fauna of Mauritanian continental shelf in the Banc d’Arguin by van Soest (1993) who remarked the high diversity (about 100 different species), but also the small size of all species collected during the Tyro Mauritania- II survey.

Highly diverse suspension-feeder communities were located in some areas of the upper slope , mainly in front of Nouakchott, but also off Cape Blanc and close to the Senegalese border (Fig. 7.16). In all cases, these concentrations were composed by bryozoans , small sponges , alcyonids and ascidians , growing on tubes of polychaetes belonging to the filter-feeder family Chaetopteridae . Moreover, in the northern area dense aggregations of the crinoid Leptometra celtica have been recorded (Chap. 12). But in no case, suspension-feeders characterize the Mauritanian assemblages and only the hydrozoan Sertularella gayi gayi was representative in the deep shelf.

Fig. 7.16
figure 16

Geographycal distribution of numerical abundances (N) of suspension-feeders by station (quantitative data by station standarized to a 0.1 km2 swept area)

Full size image

Probably, despite the high productivity that supplies food resources for suspension-feeders , the absence of suitable hard bottom s for larval settlement and the intensive fishing exploitation , prevent the development of stable long-lived suspension-feeder communities . These communities are currently restricted to special habitats, like canyons, coral mounds and the seamount , not accessible for trawling fleets and where the presence of hard bottom offers appropriate substrata to sessile epibenthos (see Chaps. 13, 14 and 15; Freiwald, Senckenberg Institute, unpublished data, 2015).

Decapods, characterized by its scavenger and carnivorous habits, constituted the most diverse and abundant group, while echinoderms, particularly deposit-feeder holothurians , represent the main biomass component in Mauritanian slope soft bottom s. The relevance of decapods in Northwest African waters was highlighted by Muñoz et al. (2012) and García-Isarch and Muñoz (2015) and its importance within the megabenthos is also evident, typifying the upper and middle slope megabenthic assemblages ; moreover decapods constitute an unusual association between 400 and 550 m (Chap. 9), at the depth range where the main faunistic discontinuity has been detected.

The holothurids also constitute an important component of benthic assemblages in extensive areas in the Northeast Atlantic deep-sea (Billett 1991; Miller and Pawson 1990), where they can account for 95% of the megafaunal biomass . Nevertheless, the species characterizing the Mauritanian deep-slope assemblage, Benthothuria funebris , Paelopatides grisea and E. eximia, are different from those of other North Atlantic areas and they were reported forming dense aggregations . Such aggregations have also been previously recorded in the deep slope off Morocco (Ramil et al. 2005; Ramos et al. 2006, 2008a, b, 2009), Namibia (Soto et al. 2006; Ramos et al. 2008a, b) and Gabon (Ramos, Spanish Institute of Oceanography , unpublished data, 2002), as well as in the abyssal plains of Cape Verde and Madeira (Billett et al. 1985). These aggregations were related with detritus-rich sediments (Gage and Tyler 1991), phytodetritus accumulations (Billet et al. 1983) and also with areas influenced by terrigenous inputs (Sibuet 1977). In the Mauritanian slope , we located the main holothurid concentrations south to Nouakchott, close to the Senegalese border down to 1500 m depth. The particular conditions of this area, with unstable bottom conditions, large sediment slides, turbidity currents (see Chap. 2) and under the influence of river inputs , could explain the holothurids abundance . A second aggregation of holothurids located off Banc d’Arguin  seems to be related with the canyon system, which enhances the transport from shallow to deep-waters and increases the sedimentation rate and the accumulation of detritus on the seafloor (Rowe et al. 1982; Harrold et al. 1998).

Noteworthy is the large size of P. grisea and B. funebris; their individual weights reached 5–6 kg. Other similar cases of gigantism in holothurids have been reported for some deposit-feeder species inhabiting the abyss (Ramírez-Llodra et al. 2010). Gigantism has usually been related to the scarce food availability in deep-waters because the energetic requirements are lower in big sized invertebrates (Rex and Etter 1998). Nevertheless, the existence of such dense aggregations of holothurians seems to be more probably coupled to areas with high rates of organic matter, like the delta of Senegal River and the canyon systems off Banc d’Arguin , than to areas with oligotrophic conditions . In addition, the most abundant holothurid in the same areas, E. eximia, is a relatively small species, which seems to support the idea that other selective issues different from food limitation may drive the evolution of body size in deep waters (Ramirez-Llodra et al. 2010). The swimming capacity displayed by E. eximia, B. funebris and P. grisea could confer them some ecological advantages for living in unstable seafloor habitats of the Mauritanian slope and to exploit their fluctuant feeding resources . Active foraging by swimming was already highlighted in some holothurians by Miller and Pawson (1990) and this behaviour can promote aggregations in suitable areas as well as enhance their reproductive efficiency.

The review of Carney (2005) confirmed some zonation patterns along the continental margins worldwide, but the final results may be heavily influenced by biogeography (local occurrence of fauna) and methodology (e.g. sampling gear). Nevertheless, Carney (2005) highlighted that the main boundaries for a wide range of taxa at local scale could be related with sharp changes in seabed morphology and water column properties. Our results for the megabenthos in Mauritanian continental margin are consistent with this interpretation. The multivariate analysis identified two main assemblages with a boundary located between 420 and 550 m depth, a bathymetric range typified by particular features in both the seabed and water column conditions. At this depth the giant coral mounds barrier interrupts the muddy slope along almost the entire Mauritanian coast (see Chap. 13) and also coincides with the lowest oxygen values measured in the water column (Chap. 3). At this depth zone, megafaunal diversity was strongly depleted and only decapods seem to profit from this particular habitat (Chap. 9). The consistence of this boundary in the Mauritanian slope was also corroborated by the distribution of macrobenthos (Chap. 8) that showed the same pattern despite differences in taxa (benthic invertebrates and demersal fishes) and sampling methodology (beam trawl).

Conclusions

Diversity of Mauritanian megabenthos , about 700 species, seems clearly higher than the one recorded in deep-waters of other North Atlantic areas , and the cumulative curve predicts even more elevated γ diversity .

Diversity showed a latitudinal pattern , with highest values in the northern area, followed by a southwards decreasing trend. Abundance and biomass displayed the opposite pattern with the lowest values in the north, increasing southwards.

The bathymetric distribution showed its highest figures in deep-waters for specific richness , abundance and biomass but for Shannon and Pielou indexes a decreasing trend with depth was observed.

Multivariate analysis of biomass datasets identified four megabenthic assemblages segregated by depth. The giant coral mounds barrier coupled with the lowest oxygen values obtained at the same depths (400–550 m) perform the main faunistic discontinuity in Mauritanian slope , separating the deep-shelf and upper slope assemblages from those located at middle and lower slope .