Taxonomic revision of West African cone snails (Gastropoda: Conidae) based upon mitogenomic studies: implications for conservation

In the last few years, a sharp increase in the number of descriptions of new species of West African cone snails, particularly from the Cabo Verde Archipelago, has taken place. In previous studies, we used mitogenome sequences for reconstructing robust phylogenies, which comprised in total 120 individuals representing the majority of species (69.7%) described from this biogeographical region (except Angolan endemics) and grouped into seven genera within the family Conidae. Here, we add another 12 individuals representing endemic species that were missing in the previous studies. We use the phylogenies to identify monophyletic groups and a genetic divergence threshold (0.2% uncorrected p distance) to determine the number of valid species. As a result, the number of valid West African cone species could be drastically reduced to at least 40%, indicating that some recent poor-quality descriptions loosely based on phenotypic characters prone to convergence such as the shape and color patterns of the shell have contributed substantially to taxonomic inflation. Several previously accepted species with a reduced geographical distribution now become phenotypic forms of the remaining valid species, which increase their distribution ranges. In contrast, several cryptic species are now uncovered and described. For instance, Africonus insulae sp. nov. and Kalloconus canariensis sp. nov. are hereby introduced as new species. A detailed systematic account with illustrations and relevant information is presented. Lectotypes are designated for Conus trochulus and Conus irregularis, and neotypes for Conus crotchii and Conus diminutus. According to our results, it is strongly recommended that any future introduction of new taxa names for cone snails from West Africa should be supported by molecular and/ or anatomical rather than exclusively shell morphological data. The taxonomic decisions here taken have direct implications for conservation and will eventually require re-evaluation of the Red List risk status of an important number of species. European Journal of Taxonomy 663: 1–89 (2020) 2


Introduction
The more than 900 described species of cone snails (Caenogastropoda, Conidae;WoRMS, accessed March 2020) are widely found in all tropical and subtropical seas from intertidal zones to deep waters. For a long time, studies on biodiversity of cone snails have mostly been focused on species from the Indo-West Pacific region (e.g., Röckel et al. 1995;Duda et al. 2001) to the detriment of others such as West Africa (e.g., Monteiro et al. 2004;Cunha et al. 2005;Duda & Rolan 2005). This region extends from Angola along the African coast to Morocco, including the Cabo Verde and the Canary Islands, among other archipelagos. The few endemic cone species found south of Angola, inhabiting the area comprised between Namibia and the Cabo Agulhas, are considered members of the South African malacological province . While several works partially treated the description of the cone snail fauna from independent geographical areas in West Africa (Senegal: Pin & Leung Tack 1995;Cabo Verde: Rolán 1980, 1990Angola: Rolán & Röckel 2000;Saint Helena: Tenorio et al. 2016), the first comprehensive review of the cone snail fauna from West Africa and the Mediterranean was published by Monteiro et al. (2004). That work covered and illustrated 99 species, subspecies and forms of cone snails from the Mediterranean Sea, the Atlantic coast of southern Spain and Portugal, the coast of Africa from Morocco to south Angola, and the archipelagos of Canaries, Cabo Verde and São Tomé e Príncipe, as well as the island of Saint Helena. Since 2004, the number of new species names introduced in the literature for West African cone snails has risen to the spectacular number of 178 as of December 2019 (Fig. 1). However, many of these new species names were introduced in non-peer reviewed amateur magazines (Cossignani 2014;Cossignani & Fiadeiro 2014a, 2014b, 2015a, 2015b, 2015c, 2017c, 2018a, 2018b, 2018c, 2019a, 2019b) based on rather poor-quality descriptions of shell characters only, neither with specific references to intraspecific variability nor with statistical analyses of shell morphometry. Only in a few cases, the new taxa described were supported by molecular analyses (Tenorio et al. , 2017. Studies focused on West African cones have revealed high levels of endemicity and peaks of species diversity concentrated in subtropical areas around Senegal and Angola on the continent, and most prominently in the Cabo Verde Archipelago, which was estimated to harbor about 10% of cone species diversity worldwide (Cunha et al. 2005. Several species endemic to the Cabo Verde Archipelago were described by Hwass, Reeve, Kiener and G.B. Sowerby II in the 18 th and 19 th centuries. However, in the original descriptions, the type localities were erroneous (i.e., "America"), too vague ("West Africa"), or simply not stated. Modern studies on the cones of the Cabo Verde Archipelago began with the description of the taxon Africonus anthonyi Petuch, 1975a, and with the works of Burnay & Monteiro (1977) and Röckel et al. (1980a). These pioneering works were followed by the descriptions of multiple new taxa by Trovão (1978Trovão ( , 1979, and most notably by Rolán and co-workers (see Rolán 1990 and references therein), who extensively sampled the archipelago in the 1980's. The first molecular work on cones from Cabo Verde dates back to the first decade of this century and reconstructed their phylogeny based on partial mitochondrial and nuclear gene sequences (Cunha et al. 2005;Duda & Rolán 2005).
In 2011, a workshop for the International Union for the Conservation of Nature (IUCN) Red List of Threatened Species held in Chicago assessed threats to over 630 species of cone snails (Peters et al. 2013(Peters et al. , 2016. The results of this assessment indicated that three-quarters of the species evaluated were not considered at risk of extinction. However, in the West African region, 41.8% of the 98 cone snail species evaluated from this area were classified as threatened or near threatened with extinction, mostly due to their restricted distributions often in habitats under high human pressure. In subsequent years, the very sharp increase in the number of descriptions of new species of West African cone snails, particularly from the Cabo Verde Archipelago, complicates monitoring programs as well as future evaluations and eventual updates of their risk status. Furthermore, the number of endemic gastropod mollusks in the Cabo Verde Archipelago (of which cone snails constitute the main component) has been one of the factors recently used to support the exclusion of Cabo Verde from the remaining Macaronesian archipelagos and the claim for its own biogeographic subprovince within the West African Transition province (Freitas et al. 2019). Therefore, if the number of valid endemic species of cone snails in the Cabo Verde Archipelago is modified significantly, the new figures would have an impact on the results of this study and its conclusions would accordingly require a re-evaluation in the light of more accurate data. It also may affect various downstream analyses such as, e.g., the estimation of speciation rates (Phuong et al. 2019).
Here, we reconstruct a new phylogeny based on 54 mitogenomes representing valid species as deduced from previous studies (Abalde et al. 2017a(Abalde et al. , 2017b as well as new mitogenomes of 10 species endemic to Senegal, Gambia, Canary Islands and Cabo Verde not previously sampled, and two species endemic to Angola, another important hotspot of endemism for cone snails along the West African coast. The total number of recorded species in West Africa (without Angola) is 145 and we include 101 species in the phylogenies of Abalde et al. (2017aAbalde et al. ( , 2017b plus this work (i.e., 69.7%). In this monograph, we discuss in detail the taxonomic implications of the reconstructed phylogenies for West African cones taking into account the 0.2% divergence threshold. The number of valid West African cone species is drastically reduced into almost one half. This implies a modification of the distribution ranges for many of the remaining valid endemic species, thus directly affecting in many cases their Red List risk status, which will eventually require re-evaluation for proper downgrade or upgrade of their respective risk classifications.
The systematic account herein presented covers in detail all valid cone species from West Africa excluding those endemic to Angola. The mitogenomes of cone specimens from this latter area are now being systematically sequenced, and the results will be reported elsewhere in due course. We have taken some necessary taxonomic actions affecting the status of certain taxa names and type material. The phylogenetic analysis here performed has also revealed two previously undescribed new taxa, namely Africonus insulae sp. nov. from Santa Luzia Island, Cabo Verde, and Kalloconus canariensis sp. nov. from the Canary Islands. These two new species are formally described herein.

Material and methods
Most of the material studied here was previously deposited in institutional repositories (MNCN and others; see Abbreviations below). Descriptions and measurements are based on shells oriented in the traditional way: spire up with the aperture facing the viewer. The taxonomy used in the present work follows Tucker & Tenorio (2013), who recognize up to 91 genera within the family Conidae. Instead, other authors have proposed only 4-6 genera within the family Conidae (Puillandre et al. 2015;Uribe et al., 2017), with the vast majority of species within the genus Conus (a total of 761 members according to WoRMS, accessed March 2020). The clades indicated in the phylogeny and their association to particular biogeographic units (Africonus to Cabo Verde; Varioconus to continental coast of Africa; Lautoconus to the Mediterranean Sea) lead us, however, to use generic names given the high level of sequence divergence observed. Furthermore, compared to other gastropod groups, WoRMS lists, e.g., 27 genera for 620 species in Nassaridae (23 spp. on average per genus), and 52 genera for 261 species in Cypraeidae (5 spp. on average per genus). Considering that Conidae sensu Puillandre et al. (2015) could have 111 genera according to Tucker & Tenorio (2013), the average number of species per genus would be 9.4, i.e., a number similar to those observed for other gastropod genera. Maps of sampling sites including the names of localities mentioned throughout the text are included as part of the electronic Supplementary Material File 1 for reference, as well as a map showing latitudinal variation in species richness of cone snails along the West African coast and Macaronesian islands reproduced from Cunha et al. (2014) by permission of Oxford University Press.

Morphological analyses
For the description of new species, we use the following procedure. We describe shell morphology using the terminology established in Röckel et al. (1995). For morphometric comparisons, adult shells selected among available specimens in the collections of the MNCN and other sources (private collections) were measured with a digital caliper, and the measurements rounded to 0.1 millimeter. All the measurements TENORIO M.J. et al., Taxonomic revision of West African cone snails 5 are in a spreadsheet, deposited as electronic supporting information (Supplementary Material File 2). For comparison of shell morphometry, we performed analysis of the covariance (ANCOVA) for different shell parameters, namely maximum diameter (MD), height of the maximum diameter (HMD) and spire height (SH), using species hypotheses as factor and shell length (S L ) as covariate. Additionally, we compared the mean values of S L statistically using t-and U-tests. Statistical tests were carried out using STATGRAPHICS XVII-X64, after all the measurement sets passed the normality tests. We use the terminology for radular morphology of Tucker & Tenorio (2009), and the abbreviations in Kohn et al. (1999) and Rolán & Raybaudi-Massilia (2002). The radular sac was dissected from the cone snail and soft parts were digested in concentrated aqueous potassium hydroxide for 24 hours. The resulting mixture was then placed in a Petri dish and examined with a binocular microscope. The radular teeth were removed with fine tweezers, rinsed with distilled water, then mounted on a slide using Aquatex (Merck) Mounting Medium, and examined under a compound microscope. Figure photos were obtained with a CCD camera attached to the microscope.

Molecular and phylogenetic analyses
Most of the samples, mitogenome (mt) sequences and phylogenetic analyses involved in this taxonomic revision have originally been reported in previous papers (Cunha et al. 2009;Abalde et al. 2017aAbalde et al. , 2017bAbalde et al. , 2019Uribe et al. 2017). Here, we sequenced additional mitogenomes from the species listed in Table 1, which also provides information on localities, vouchers, genome sizes, and GenBank accession numbers (see a full list of previous and current mitogenomes in the electronic Supplementary Material File 3). All laboratory procedures and sequence analyses involving the new mitogenomes were performed as in Abalde et al. (2017a) except if otherwise mentioned. For Africonus curralensis (Rolán, 1986), Kalloconus pseudonivifer (Monteiro, Tenorio & Poppe, 2004), Kalloconus canariensis sp. nov. and Varioconus echinophilus (Petuch, 1975b), long PCRs and Illumina sequencing were performed (Abalde et al. 2017a). However, the samples of Varioconus equiminaensis (Schönherr, 2018), V. fernandi (Petuch & Berschauer, 2018), V. gambiensis (Petuch & Berschauer, 2018), V. pineaui (Pin & Leung Tack, 1989b), V. rikae (Petuch & Berschauer, 2018), V. trovaoi (Rolán & Röckel, 2000), V. trencarti (Nolf & Verstraeten, 2008) and V. wolof (Petuch & Berschauer, 2018) rendered DNA of poor quality and long PCRs did not work. Therefore, we used universal cox1 LCO1490 + HCO2198 (Folmer et al. 1994) and rrnL 16SAR-L + 16SBRH primers (Palumbi et al. 1991), and designed up to 23 pairs of specific primers (see Supplementary Material File 4) to amplify through standard PCRs overlapping fragments of 500-1000 bp covering the entire mitogenome except the control region. Standard PCR conditions were as in Abalde et al. (2017a), and annealing temperatures for each pair of primers were 48ºC except for the pair ANG-ND1-F/R, which was 52ºC. Amplified fragments were sequenced using the Sanger method using the PCR primers at Macrogen (Seoul, Korea). Illumina sequences were assembled as in Abalde et al. (2017a) whereas Sanger sequences were assembled manually using Geneious ® 8.0.3. Once annotated, all mt protein-coding and rRNA gene nucleotide sequences were aligned and concatenated. The selected bestfit partitions according to the Bayesian Information Criterion (BIC) included the three codon positions in the case of protein-coding genes plus the two rRNA genes combined. PartitionFinder2 (Lanfear et al. 2017) was used to select best-fit model for each partition according to the BIC (GTR+I+G for each codon position; HKY85+I+G for the rRNA genes) and a maximum likelihood (ML) tree was reconstructed (see Supplementary Material File 5) using IQtree 1.6.10 (Nguyen et al. 2015;Chernomor et al. 2016) in the CIPRES gateway (Miller et al. 2010). Statistical support was measured with 10 000 ultrafast bootstrap pseudoreplications (-bb). The recovered topology was used as constraint for performing a Bayesian estimation of divergence times using a relaxed molecular clock calibrated as in (Abalde et al. 2017a). Two nodes with low support, A. longilineus, A. regonae, A. felitae (bb = 76) and V. trecarti, V. unifasciatus, V. guanche (bb = 66) were not constrained during topology fixation.

Species delimitation
The remarkable diversification of cone snails in West Africa has occurred through radiation events concentrated in geographically restricted hotspots. As a result, species delimitation within the group is challenging as morphological differences in many cases are subtle, and genetic divergences are low. Following Abalde et al. (2017aAbalde et al. ( , 2017b, we use the reconstructed phylogeny (that delimits monophyletic groups) and pairwise uncorrected p distances estimated among the mitogenome sequences with a threshold of 0.2% as primary criteria for species separation. This threshold was set in (Abalde et al. 2017a) based on divergences within two clades in the phylogeny: (1) Africonus verdensis and its sister species Africonus raulsilvai + Africonus gonsaloi; (2) Varioconus guanche and its sister species Varioconus unifasciatus. We assumed that A. verdensis living off Santiago Island in Cabo Verde only and V. guanche, from the Canary Islands, and far away from Senegal cones, are both valid species, and thus their sequence divergence to the corresponding sister group species could be used as minimum threshold. Actually, this was a rather conservative criterion when compared with literature. Our threshold is equivalent to a 0.5% Kimura 2-parameter (K2P) distance for the barcoding cox1 fragment in our dataset, and thus it is significantly lower than the reported mean intraspecific K2P distance of 0.8% for various gastropods (Borges et al. 2016). In those few cases in which the mitogenome was not available, we take a conservative approach. Those species having distinctive morphological and biogeographical features are given the consideration of provisionally valid. This status is assigned to cone snail species from westernmost islands of Cabo Verde (as those for which a mitogenome is available show important sequence divergence); to the cone endemics at Brava and São Nicolau islands; and to most cone snails endemic to Angola, which are currently under study. For recently described species showing only subtle differences in shell pattern or apparent differences in shell morphometry not statistically tested in the original descriptions, we propose to synonymize the name of the taxon under consideration with that of the morphologically closest species with known mitogenome (specific arguments for each case are provided and discussed in the systematic account below). A complete alphabetic list of names, synonyms (with sources for synonymy) and other relevant data is included in the Supplementary Material File 3.

Phylogeny and time tree of West African cones based on mitogenomes
The almost complete mitogenomes (except the control region) of Africonus curralensis, a new specimen of Kalloconus pseudonivifer, Kalloconus canariensis sp. nov., Varioconus echinophilus, V. equiminaensis, V. fernandi, V. gambiensis, V. pineaui, V. rikae, V. trovaoi, V. trencarti and V. wolof were sequenced. The following primer pairs did not work in few species: ANG-ND4-F/R in Varioconus gambiensis and Varioconus fernandi; ANG-ND1-F/R in Varioconus trovaoi; universal 16SAR-L/SBR-H in Varioconus rikae. A highly resolved (most nodes show maximal statistical support) phylogenetic tree of West African cones is reconstructed based on mitogenomes and using Monteiroconus tabidus (Reeve, 1844), Genuanoconus genuanus (Linnaeus, 1758) and Chelyconus ermineus (Born, 1778) as outgroup taxa (see Supplementary Material File 5). The recovered topology was used as constraint to reconstruct a chronogram (Fig. 2) under a relaxed molecular clock calibrated with fossils as in Abalde et al. (2017a). The first split in the ingroup separated specimens within genus Fig. 2. Phylogenetic relationships of West African cones and species delimitation. A time tree (relaxed molecular clock) was reconstructed following methods and calibrations in Abalde et al. (2017a). Dates are in million years. Bold names indicate new mitogenomes. A red line indicates the threshold (uncorrected p distance of 0.2%, based on the divergences between Africonus verdensis (Trovão, 1979) and its sister group and Varioconus guanche (Lauer, 1993) and its sister group) for species delimitation.
Kalloconus from the remaining analyzed cones. Within Kalloconus, two main clades were recovered. One corresponded to the continental Kalloconus pulcher ( [Lightfoot, 1786]) and one pulcher-like individual from Gran Canaria, Canary Islands, whereas the other included all species of Kalloconus endemic to Cabo Verde: K. marimaris  was sister to K. ateralbus (Kiener, 1850) and both were sister to a clade including K. venulatus (Hwass in Bruguiére, 1792) sister to K. pseudonivifer + K. trochulus (Reeve, 1844) (Fig. 2). Note that Kalloconus atlanticoselvagem  and most K. pseudonivifer (except those from Praia Canto in Boa Vista; a second specimen from this locality was sequenced to confirm previous results in Abalde et al. 2017a) did not show enough divergence to K. trochulus, and had to be synonymized with this latter species (see Abalde et al. 2017a). On the other hand, the pulcher-like specimen from the Canary Islands showed a large sequence divergence from mainland Kalloconus pulcher indicating that they were not conspecific (Fig. 2). Consequently, we have introduced the new taxon name Kalloconus canariensis sp. nov. for the pulcher-like individuals from the Canary Islands, which must be considered representatives of a full species sister to K. pulcher.
The remaining analyzed cones were recovered in three main clades (Fig. 2). A first clade corresponded to the genus Varioconus da Motta, 1991 (type species Varioconus variegatus (Kiener, 1848)) and included two monophyletic groups, one grouping cones endemic to Senegal and V. guanche (Lauer, 1993a) comb. nov. from the Canary Islands and the other including cones endemic to Angola (cox1 sequences indicate that the taxon V. jourdani (da Motta, 1984), endemic to Saint Helena Island, also appeared deeply nested within the clade of endemic cones from Angola; see Tenorio et al. 2016). Three main lineages were retrieved within Varioconus endemic to Senegal and Canary Islands. Most phylogenetic relationships within these three lineages were reported in Abalde et al. (2017b). Here, we only focus on those involving the newly sequenced mitogenomes. The species Varioconus pineaui and V. wolof have to be synonymized with V. guinaicus (Fig. 2). The species Varioconus fernandi, V. gambiensis and V. rikae have to be synonymized with V. mercator. The species Varioconus trencarti remains valid and is sister to V. unifasciatus (Fig. 2). The species Varioconus echinophilus is valid and sister to V. bruguieresi (Fig. 2). Note that the cones endemic to Senegal and the species from Canary Islands were previously assigned to the genus Lautoconus which resulted a polyphyletic group in the reconstructed tree (Tucker & Tenorio 2009). To resolve this problem, the clade could receive a new genus name but the low sequence divergences to the sister group did not justify such taxonomic action, and instead supported ascription of these species to the genus Varioconus as the most senior supraspecific taxon name (Fig. 2). Therefore, the genus Lautoconus was restricted to Lautoconus ventricosus (Gmelin, 1791) from the Mediterranean Sea and the adjacent Atlantic Ocean region. This species was recovered as sister to cones endemic to Cabo Verde belonging to the genus Africonus Petuch, 1975a (Fig. 2). Note that the two mitogenomes of Lautoconus ventricosus from Faro (Portugal) and Formentera Island (Spain) show an important sequence divergence that might suggest a potential case of a cryptic species complex and could merit further study with an extended taxon sampling.

Material examined
Holotype (not examined) Holotype was in collection H. Cuming, present whereabouts unknown; reference is figure 254 on plate 6 of Reeve (1849). Three specimens in the NHMUK type collection were formerly labeled as "probable syntypes", but these were not part of the type material. Given the relevance of this polymorphic taxon as an important reference for the Cabo Verde cone fauna, we hereby designate a neotype matching the figure in Reeve (1849). The originally stated type locality "Saldanha Bay, South Africa" was erroneous and it was corrected to Santa Mónica, Boa Vista Island, Cabo Verde Archipelago (Coomans et al. 1985a).

Geographical distribution
Multiple localities around Boa Vista Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Endangered. Status revision required.

Remarks
The name crotchii has traditionally been applied to specimens from southwest Boa Vista Island matching figure 254 in Reeve (1849). However, based upon molecular phylogenetic studies, this has now been found to be the most senior name applicable to a large clade of specimens from multiple localities around Boa Vista Island with a very large phenotypic variability in shell pattern. Differences in radular morphology among the members of this clade are, however, rather subtle in agreement with the low DNA sequence divergence. This clade includes representative individuals of no less than 14 taxon names, which are now considered to be synonym (color forms) of the nominal taxon. Thus, the taxon A. salreiensis, which was initially classified within the category 'critically endangered' in the IUCN Red List, now shows to be a mere color form of A. crotchii distributed in and around the Sal Rei area in northwest Boa Vista Island. Similarly, the taxon A. irregularis also now becomes a junior syonym of A. crotchii. In the mitogenome phylogeny, the analyzed individuals of A. irregularis from different populations appear distributed in two separate clades. Most of the individuals fell into the A. crotchii clade, whereas two were recovered within the A. maioensis clade. In order to attribute the senior name to each of the clades, we needed first to identify which of the analyzed individuals matched the available type material of A. irregularis. There is a series of three syntypes of this taxon at the NHMUK "from West Africa" as the type locality. We hereby designate the specimen 197871/1 (dimensions: 28.3 × 18.1 mm) in the series as lectotype for the taxon Conus irregularis G.B. Sowerby II, 1858 (Fig. 3H). Specimens 197871/2 and 197871/3 from the same lot became paralectotypes. We consider that the specimens identified as A. irregularis from Estancinha, near Ponta do Sol in northwest Boa Vista match the type series at the NHMUK, and hence we propose a change in the type locality to Estancinha, north of Sal Rei, Boa Vista Island, Cabo Verde Archipelago. The sequenced specimens of A. irregularis from Baía das Gatas, Morro de Areia and Estancinha were recovered within the A. crotchii clade, which becomes the senior name for this taxon. Being a polymorphic species, many of the recently introduced names could be assigned to the taxon A. crotchii. Thus, A. antonioi and A. josegeraldoi are morphologically related to A. crotchii form fiadeiroi, whereas the green color and small size of A. calhetinhensis most likely correspond to a juvenile specimen of the highly variable A. crotchii form irregularis. (Reeve, 1843) Fig. 3I

Remarks
It was initially reported in Röckel et al. (1980a) as Conus decoratus form D. In the phylogenetic tree, it belongs to the same clade as A. decoratus, but appears only distantly related. A distinctive color form living sympatrically in Santa Luzia Island was initially confused with this species in Abalde et al. (2017a). The sequencing of a specimen fully matching the holotype of A. curralensis has shown that the alleged color form is actually a separate, sister species that is introduced hereby with the name Africonus insulae sp. nov. (see below).

Geographical distribution
Multiple localities around Boa Vista Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List Least concern.
Remarks This species is widely distributed around Boa Vista Island, with a high variability in shell pattern. This polymorphism has led to the introduction of several taxa names associated with certain populations, which are all considered synonyms hereby. Thus, the holotypes of A. diegoi and A. roquensis exhibit all the shell features characteristic of A. damottai (just a more elongated shell shape in the case of A. diegoi, which does not justify separation at the species level). The holotype of A. purvisi most likely represents a juvenile specimen of A. damottai, whereas the holotype of A. pinedensis resembles certain variations of A. damottai with a well-developed color pattern (i.e., Monteiro et al. 2004: pl. 89). The taxon A. galeao from Maio Island was initially introduced as a subspecies of A. damottai. Despite the morphological similarities in shell pattern and radular tooth, the phylogenetic analyses indicate that A. galeao can be regarded as a separate, valid full species. (Röckel, Rolán & Monteiro, 1980) Fig. 3L, 5L

Geographical distribution
Calhau to Saragaça, on the southeast coast of São Vicente Island, and Curral area, on the southwest coast of Santa Luzia Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Vulnerable.

Remarks
This species belongs to a lineage that includes species endemic to the westernmost islands in the Cabo Verde Archipelago and is sister to species from Santiago and Maio islands. Despite obvious differences in shell morphology, the species sister to A. decoratus according to the phylogeny is A. grahami. The type locality from which A. decoratus was originally described is now part of a large and touristic urban area in the vicinity of the modern Port of Mindelo. This population (decoratus form A in Röckel et al. 1980a) is believed to be now extinct, but specimens from other population (decoratus form B in Röckel et al. 1980a) are occasionally found on the east coast of São Vicente Island from Calhau to Saragaça, and more commonly on the deserted island of Santa Luzia, where the specimens used in the present study were collected. This beautifully patterned species shows some degree of variation in its background color, which ranges from light brown to black. (Trovão, 1979

Geographical distribution
Northeast coast to Porto Ferreira on the East coast of Boa Vista Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.
Remarks This species was originally described from the population at Baía das Gatas in Boa Vista Island, and the shell color is characterized by a pattern of numerous small irregular dots on a light to dark brown background. It was treated in Röckel et al. (1980a) as Conus cuneolus form B. Other populations of cones around the northeast and east of Boa Vista Island, introduced recently as separate species, exhibit very distinct shell patterns but share the same elongated radular tooth morphology observed in A. delanoyae . Representative specimens of these taxa examined molecularly exhibit mitogenome sequences almost identical to that of A. delanoyae, and can be considered conspecific. Additionally, the type material of A. joserochoi is fully consistent with the concept of A. delanoyae. The author stated that the main features separating A. joserochoi from A. delanoyae were the relative height of the spire (always eroded) and the "less triangular and more pyriform shell shape" in the former. However, no morphometric data are given, nor were these differences statistically tested, so there is no sufficient evidence to consider A. joserochoi as a separate species from A. delanoyae. We can conclude that A. delanoyae is a polymorphic species, only distantly related to A. cuneolus, comprising several well-defined forms with characteristic shell patterns, sharing the same radular morphology and mitogenome sequence. The sister species A. vulcanus is separated by a genetic uncorrected p distance of 0.3%, slightly above the threshold used to delimit species. Given the different radular morphology (radular tooth broad and robust in vulcanus, but narrow and elongated in delanoyae) we consider them tentatively separate species in spite of the close genetic proximity. Afonso & Tenorio, 2011 Fig . 3N Africonus denizi Afonso & Tenorio, 2011: 127, pl. 1, figs 1-8

Geographical distribution
Northeast coast of São Vicente Island at Praia Grande, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Near threatened.

Remarks
This small-sized species is morphologically similar to A. miruchae from Sal Island, which is however only distantly related. In the phylogeny, it was recovered as sister to the recently described A. freitasi, a morphologically related allopatric species. So far, it has only been found at the type locality. (Trovão & Rolán, 1986) Fig. 3O

Material examined
Holotype (not examined) Holotype (type locality: Cabo Verde; Boa Vista Island, Baía de Sal Rei; intertidal to 1 m depth) stated to be at MNCN in Trovão & Rolán (1986), but apparently it was never deposited (as per a note in the collection). Instead, a lot of three paratypes was present, with registration number MNCN 15.05/1011. We have selected the specimen from which the mitogenome was sequenced as neotype for the taxon Conus diminutus Trovão & Rolán, 1986

Geographical distribution
West and northwest coast of Boa Vista Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Near threatened. Status revision required.

Remarks
Africonus diminutus is a small, highly variable species endemic to the west coast of Boa Vista Island. Whereas typical specimens of A. diminutus are found around the locality of Sal Rei, specimens from Morro de Areia introduced under the name A. morroensis exhibit almost identical mitogenome sequences despite the variations in shell morphology, and are considered conspecific with A. diminutus. This represents a significant expansion of the distribution range for the species. According to the reconstructed phylogeny, A. diminutus is sister to A. boavistensis, also very variable in shell pattern and widely distributed around Boa Vista Island. Cossignani & Fiadeiro, 2017 Fig. 3P

Geographical distribution
Only known from the type locality.

Conservation status in IUCN Red List
Not evaluated.

Remarks
This species has not been examined molecularly. It is morphologically similar to A. antoniaensis, but smaller sized. Radular examination (E. Rolán, pers. com. to MJT) indicated a fully-developed, mature tooth for such a small-sized shell, consistent with that of an adult individual. This observation rules out the possibility of this taxon being a juvenile of another species. We hereby consider it as provisionally valid, pending confirmation by molecular studies. (Rolán, 1990) Fig. 4A

Geographical distribution
Only known from the type locality at Sal Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Vulnerable.

Remarks
This small-sized species endemic to Sal Island is closely related to A. cuneolus. Its distinctive shell and radular morphology as well as the phylogenetic data are consistent with its status as a valid species with a very reduced distribution range. (Tenorio, Afonso & Rolán, 2008) Fig. 4B

Geographical distribution
South coast of Santo Antão Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Endangered.

Remarks
Prior to its introduction as a full species, it was referred to as Conus cuneolus form K in Röckel et al. (1980a). Molecular data clearly confirm that it is a valid species endemic to Santo Antão Island, where it lives in the vicinity of Porto Novo harbor in rocky areas. In the tree, it is recovered as sister to A. curralensis and A. insulae sp. nov. from Santa Luzia Island. Other material CABO VERDE • 1 spec., 14.5 mm; same collection data as for holotype; 12 Apr. 2016; GenBank mitochondrion, partial genome: MF491601; MNCN 15.05/78562.

Geographical distribution
From Calhau to Saragaça, southeast coast of São Vicente Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Not evaluated.

Remarks
This small-sized species is morphologically similar to A. miruchae from Sal Island, and to A. denizi. Phylogenetic analyses show that A. freitasi, A. denizi and A. miruchae are distinct species despite their partial similarity in shell morphology. The phenotypic resemblance of the shell of A. freitasi to A. miruchae is possibly the result of convergence, whereas the similarity to A. denizi may reflect their close phylogenetic relationship, sharing a relatively recent last common ancestor. (Rolán, 1990) Fig. 4D

Conservation status in IUCN Red List
Least concern.

Remarks
This taxon was initially introduced as a subspecies of A. verdensis. This species has not been examined molecularly. However, the geographic isolation of Brava Island within the Cabo Verde Archipelago together with the distinctive morphology of shell and radula suggest that this is most likely a valid species, well differentiated in genetic distance from A. verdensis. (Röckel, Rolán & Monteiro, 1980) Fig. 4E

Geographical distribution
Multiple localities around Boa Vista Island and north Maio Island (Baía do Navío Quebrado and Praia Real), Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.

Remarks
This taxon was initially introduced as a subspecies of A. damottai distributed in Maio Island. There are clear morphological similarities in shell pattern and radular tooth. However, the phylogenetic analysis indicates that A. galeao is actually a separate species, whose lineage diverged from that of A. damottai in the middle Pliocene. It is sister to A. calhetae, a parapatric valid species from Maio Island, which had been also initially introduced as a subspecies of A. navarroi. There is a number of morphs of A. galeao with distinctive shell patterns (i.e., reticulated, greenish, etc.) corresponding to well-defined populations from different bays in northern Maio Island, but all these have shown to be conspecific. Among these, A. claudiae from Praia Real is a color form of A. galeao. This must not be confused with another color variation of A. calhetae found in the west coast of Maio Island, which also resembles phenotypically the A. galeao form claudiae.  Fig. 4G

Geographical distribution
Only known from the type locality, Maio Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Not evaluated.

Remarks
This was the first endemic species described from the northeast coast of Maio Island. It is sister to A. raulsilvai, an allopatric, closely related species widely distributed along the east and north coasts of Maio Island. Both, A. raulsilvai and A. gonsaloi are members of a clade sister to A. verdensis, the only endemic species from the neighboring Santiago Island. (Röckel, Cosel & Burnay, 1980) Fig. 4H

Geographical distribution
Eastern side of São Vicente and southeastern extremity off the neighboring island of Santa Luzia at Praia Francisca, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.

Remarks
The elongated-ovate aspect of the shell, its greenish color and typical pattern, as well as the high spire and convex profile easily separate A. grahami from most other species from the Cabo Verde Archipelago, with the exception of A. longilineus from Sal Island, which was in fact described initially as a subspecies of A. grahami. However, A. longilineus and A. grahami are two separate species, only distantly related according to the phylogeny. Thus, the similarities in shell morphology are most likely due to convergence. The name A. grahami luziensis (initially introduced as Conus grahami pseudoventricosus Röckel, Rolán & Monteiro, 1980, and then renamed due to homonymy with Conus mediterraneus var. pseudoventricosa Sacco, 1893, a fossil) was applied to populations of A. grahami from the island of Santa Luzia. We have not been able to examine molecularly specimens from this population. However, given the lack of clear morphological features separating the populations from São Vicente and Santa Luzia, and the past connectivity between these two islands, we hereby consider A. grahami luziensis as a synonym (form) of A. grahami. The species sister to A. grahami according to the phylogeny is A. decoratus, which however exhibits very distinct shell morphology. (Rolán, 1990) Fig. 4I

Geographical distribution
Maio Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.
Remarks This is a species widely distributed around Maio Island, very especially along the east and north coasts. It is sister to A. roeckeli from Boa Vista Island, and only distantly related to other species with similarities in shell morphology such as A. antoniomonteiroi from Sal Island.  shell. Moderately small to small. Maximum length: 30.6 mm. Shell profile broadly and ventricosely conical to broadly conical, with convex sides adapically, and straight below. Spire of moderate height, of straight to concave outline. Protoconch not observed due to erosion, but assumed paucispiral. Teleoconch whorls flat or slightly concave, with 3-4 spiral cords becoming rather obsolete in late whorls. Shoulder angulate. Early teleoconch whorls white. Suture of spire whorls pigmented with dark brown. Late teleoconch whorls yellow to light brown with radial white irregular blotches. Ground color yellow to light brown. Last whorl overlaid in variable amount with oblique white lines forming zigzag pattern. Small white spots and axial flecks arranged in spiral band frequently present. Columella white to light purple. Aperture white and porcellanous inside, with pale purple to light brown diffuse area near the outer lip crossed by two narrow bands of a lighter color. Periostracum yellow-brown, thin and translucent. radular teeth. Radula studied from two specimens. 27-35 teeth in radular sac. Radular tooth of medium relative size (S L /T L = 50-65), with well-developed apical barb. Anterior section shorter than posterior section of tooth (T L /AP L = 2.4-2.6). Waist evident. Blade rounded, covering most of anterior section (100B L /AP L = 83-87%). Serration with 10-15 denticles, which do not reach apical portion, arranged in one row ending in small terminating cusp. Base rounded, with small spur (Fig. 4M).

Distribution and habitat
Southwest coast of Santa Luzia Island (Curral and Praia de Palmo Tostão), Cabo Verde Archipelago. On rocks and under stones, in 2 to 5 m depth.

Conservation status in IUCN Red List
Not evaluated.

Remarks
This species was previously confused with A. curralensis, which lives sympatrically on Santa Luzia Island. In fact, the specimen identified as A. curralensis in Abalde et al. (2017a) actually corresponds to the taxon A. insulae sp. nov. Molecular examination of a specimen (Fig. 5K) morphologically perfectly matching the holotype of A. curralensis (Fig. 3J) showed a genetic divergence consistent with the separation at the species level. Hence, A. curralensis and A. insulae sp. nov. are two closely related sister species living in sympatry. Both species essentially have the same shell shape and almost identical radular teeth morphologies (Fig. 4M). However, they can easily be separated by differences in shell pattern. The ground color in A. curralensis is composed of different shades of olive green with two narrow spiral bands of a lighter color, whereas in A. insulae sp. nov. the ground color is plain yellow to light brown. The pattern in A. curralensis consists of irregularly shaped white flecks and blotches, whereas in A. insulae sp. nov. it is better defined, composed of fine white zigzag lines and small arrow head-shaped markings. This pattern of zigzag lines on the shell also resembles that present in certain forms of A. decoratus. This species however has a much darker ground color, usually dark brown to black, a white-colored aperture, and is phylogenetically divergent. The shell of the recently described A. santaluziensis Cossignani & Fiadeiro, 2015 (hereby considered as provisionally valid) has a similar ground color, but it is very sparsely patterned and with a white-colored aperture. Besides, the shell of this species has rounded shoulders and a more ventricose appearance, instead of conical.  Fig. 4J

Geographical distribution
West and northwest of Maio Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.
Remarks Africonus isabelarum, endemic to Maio Island, was initially considered conspecific with A. decoratus from São Vicente and Santa Luzia, due to similarities in shell pattern (A. decoratus form C in Röckel et al. 1980a). In the phylogeny, it stands alone, sister to a large clade comprising many species distributed in Boa Vista and Maio islands, and only distantly related to A. decoratus.

Geographical distribution
Multiple localities around Boa Vista Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Near threatened. Status revision required.

Remarks
This species is very closely related to A. borgesi. With regards to the conservation status, A. josephinae used to be abundant around its type locality in Boa Vista Island. However, the expansion of the town of Sal Rei due to touristic and industrial development is seriously compromising the habitat of this population. The alleged conspecific population from Maio Island has been now established as a separate species, namely A. perrineae Cossignani & Fiadeiro, 2018, in agreement with molecular data (Abalde et al. 2017a;Tenorio et al. 2018). This significantly reduces the distribution range of A. josephinae. On the other hand, several populations distributed around Boa Vista Island, which have been recently introduced as new species, actually correspond to different color and pattern variations of A. josephinae. These names are here synonymized accordingly. Thus, A. guiandradoi is applied to pale-colored individuals of A. josephinae with a pattern of sparse fine brown spiral lines. This pattern is commonly observed in specimens of A. perrineae from Maio Island, but it is rare in Boa Vista. The name A. demisgeraldoi applies to typical brown specimens of A. josephinae that display one midbody band irregularly patterned with white blotches. The alleged differences in shell shape were not statistically tested, and the presence of white blotches on the shell does not warrant the separation from A. josephinae. Taking into consideration all these changes, a risk status revision for A. josephinae is strongly recommended. (Tenorio, Afonso & Rolán, 2008) Fig

Geographical distribution
Southwest coast of São Nicolau Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Near threatened.

Remarks
Prior to its introduction as a full species, it was referred to as Conus cuneolus form L in Röckel et al. (1980a). This species has not been examined molecularly. However, the geographic isolation of São Nicolau Island within the Cabo Verde Archipelago together with the distinct morphology of the shell and radula suggest that this is most likely a valid species, not closely related to A. cuneolus. (Röckel, Rolán & Monteiro, 1980) Fig. 4M

Geographical distribution
East coast, Sal Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.

Remarks
This taxon was initially introduced as a subspecies of A. grahami. Despite superficial resemblance in shell morphology, A. longilineus is a separate species endemic to Sal Island, only distantly related to A. grahami. Although the type locality for A. longilineus was not specifically mentioned in the original description, its distribution range was stated to be along the east coast of Sal Island, with another population in the north of the island. Several representative specimens of recently named taxa described from the east coast of Sal Island exhibit mitogenome sequences almost identical to that of A. longilineus, and are therefore considered synonyms (forms) of this species. The species A. nelsonandradoi was compared in the original description to the very variable A. cagarralensis, which is a synonym (form) of A. longilineus (Abalde et al. 2017a). We propose to consider that A. nelsonandradoi represents merely a northern population of the latter. On the other hand, A. minimus likely is a juvenile of A. longilineus. Besides, the name Africonus minimus Cossignani & Fiadeiro, 2015b (Fig. 4N).

Conservation status in IUCN Red List
Critically endangered.

Remarks
This species has not been examined molecularly. Specimens matching the type material of this species were found in the 1970's and 1980's in Matiota beach in São Vicente Island, but this is now part of a large and touristic urban area in the vicinity of the modern Port of Mindelo. This population is believed to be now extinct. We are not aware of live specimens of this species having been observed for over 20 years. Some empty shells of this species have been collected in recent years, and the search for live specimens in new locations along the north coast of São Vicente Island continues, without success for the moment being. It is feared that the species might have become extinct. The claim for specimens of A. cf. lugubris found at Calhau and Saragaça to the east of São Vicente Island most likely corresponds to individuals of A. freitasi. (Trovão, Rolán & Félix-Alves, 1990) Fig. 4O

Material examined
The holotype was in Museu Bocage in Lisbon and currently assumed to be lost due to a fire which destroyed the building of the hosting institution. A neotype was subsequently designated (Monteiro et al. 2014).

Geographical distribution
Maio and Boa Vista islands, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.
Remarks This is one of the most representative species of Maio Island, as reflected by its name. Off Maio Island it is distributed along the north coast, from Baía do Navio Quebrado to Baía do Galeao. Each of the populations has its own characteristic shell pattern on a green to black background color. This variability has led to the introduction of several new taxa names. However, an examination of the mitogenome sequences indicates that all these names correspond to plain color forms of one single polymorphic species, namely A. maioensis. Some individuals of the polyphyletic taxon A. irregularis also fell within the A. maioensis clade. One of them was from the population from Porto Cais, on Maio Island, and the other from Baia Grande de Derrubado, Boa Vista Island. Once that we have designated the lectotype for the taxon Conus irregularis G.B. Sowerby II, 1858 and its type locality as "Estancinha, north of Sal Rei, Boa Vista Island", and established that this is a synonym (form) of A. crotchii (see above), it becomes clear that the irregularis-like specimen from Boa Vista in the maioensis clade is not conspecific with A. irregularis (= A. crotchii), but with A. maioensis. This indicates that the distribution range of A. maioensis is not restricted to Maio Island, as it expands to Boa Vista Island. Furthermore, the recently described species A. zinhoi from south Boa Vista Island also fell within the A. maioensis clade, and it is therefore another synonym (form) of A. maioensis. The holotype of A. cossignanii is virtually identical to A. crioulus, a form of A. maioensis. The alleged differences in shell length and spire profile were not tested, and the minimal differences in shell pattern do not justify a separation at the species level. The recently described A. decolrobertoi is likely a juvenile of the population of A. maioensis from Praia Real, north Maio (which had previously been introduced as a separate species named A. marcocastellazzii). This population is characterized by a predominance of greenish shades on the shell pattern. Other than that, it is merely another morph of A. maioensis.

Geographical distribution
North of Sal Island, Cabo Verde Archipelago. Also present in Rabo de Junco, on the west coast of Sal Island.

Conservation status in IUCN Red List
Least concern.
Remarks According to the phylogeny, A. miruchae is sister to all the other endemic species from Sal Island. This small-sized species is adapted to the rough sea conditions present in the barren northern coasts of Sal Island. It is often found crawling on the rocky walls of large pools formed during low tide. A population of this species that is present at Rabo de Junco, at the north end of Murdeira Bay, is often confused with A. longilineus due to similarities in shell morphology (Monteiro et al. 2004). (Rolán, 1986) Fig. 6A

Geographical distribution
Southeast coast São Vicente Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Near threatened.

Remarks
This small, distinctive species with a very restricted distribution range belongs to a lineage that includes species endemic to the westernmost islands in the Cabo Verde Archipelago. It is sister to A. fernandesi from Santo Antão Island, and to the pair A. curralensis / A. insulae from Santa Luzia. It is only distantly related to A. calhetae from Maio Island, which was initially introduced as a subspecies of A. navarroi due to similarities in shell morphology, now considered convergent.  (Fig. 6B).

Geographical distribution
North of Maio Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Not evaluated.

Remarks
Mitogenomic analyses revealed that the populations of A. josephinae from Boa Vista and the alleged ones from the north of Maio Island actually are not conspecific (Abalde et al. 2017a). This led to the introduction of a new species, which was formally described as A. angeluquei based upon shell and radula morphological studies and examination of its mitogenome sequence ). However, the name A. perrineae was almost simultaneously introduced in a rather succinct description for exactly the same species, and was published one week earlier than A. angeluquei, becoming the senior name for this species (Cossignani & Fiadeiro 2018b). In spite of this, a wealth of useful information about this taxon can be found in the detailed original description of A. angeluquei . (Rolán, Monteiro & Fernandes, 1998) Fig. 6C

Geographical distribution
West and north coasts of Maio Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.
Remarks This species is sister to A. gonsaloi, allopatrically distributed on the northeast coast of Maio. Both A. raulsilvai and A. gonsaloi are members of a clade, which is sister to A. verdensis, the only endemic species from the neighboring Santiago Island.

Geographical distribution
West and north Sal Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Vulnerable.

Remarks
Prior to its introduction as a full species by Rolán (1990), it was referred to as Conus cuneolus form I in Röckel et al. (1980a). This species is very closely related to A. cuneolus, and lives in sympatry with its form fontonae. It is particularly abundant in rocky areas to the northwest of Sal Island. (Rolán, 1980) Fig. 6E

Geographical distribution
North and northeastern Boa Vista Island, Cabo Verde Archipelago.

Least concern.
Remarks This is a very variably patterned species from Boa Vista Island. It is sister to A. infinitus from Maio Island, and stands phylogenetically fairly distant from other species in Boa Vista Island. The taxon A. damioi is conspecific, and represents a mere color form of A. roeckeli. Rolán, 1986

Geographical distribution
Only known from the type locality.

Conservation status in IUCN Red List
Not evaluated.

Remarks
This relatively recently described species has not been examined in detail. It is morphologically (and most likely phylogenetically also) related to A. fuscoflavus. It is hereby considered as provisionally valid only, pending confirmation by future molecular analyses. Cossignani & Fiadeiro, 2015 Fig. 6G

Geographical distribution
Only known from the type locality.

Remarks
This recently described species has not been examined. We did not find individuals in our survey of Santa Luzia in 2016. Given the observed genetic divergence present in most of the taxa reported from the islands of São Vicente and Santa Luzia, we consider this taxon as provisionally valid pending future detailed molecular studies.  (Fig. 6H).

Geographical distribution
Northwest and northeast coasts of Maio Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Not evaluated.

Remarks
This species has not been molecularly examined. Its radular tooth morphology suggests a relationship to the species A. calhetae or A. galeao, but its distinctly elongated shell shape readily separates it from the latter two. Initially described from the area of Praia Santana, to the northwest of Maio, a population recently introduced as A. gonsalensis from Praia Gonçalo, to the northeast of the island, has an almost identical shell morphology and is therefore considered conspecific. The description of A. santanaensis was published in April 2014, and the taxon A. gonsalensis appeared in July 2014. Hence, A. gonsalensis becomes a junior synonym for A. santanaensis.

Geographical distribution
São Vicente and Santa Luzia islands, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Near threatened.

Remarks
Prior to its introduction as a full species by Rolán (1986), it was referred to as Conus cuneolus form E in Röckel et al. (1980a). This species has not been molecularly examined. Given the observed genetic divergence present in the taxa reported from the islands of São Vicente and Santa Luzia, we consider this taxon as provisionally valid pending future detailed molecular studies. (Trovão, 1979) Fig. 6J

Geographical distribution
Santiago Island, Cabo Verde Archipelago.

Conservation status in IUCN Red List
Least concern.

Remarks
The type locality for this species is most likely erroneous, as it seems to be endemic to Santiago Island, being particularly abundant around the Tarrafal area to the northwest. The species introduced under the name nelsontiagoi had been illustrated and referred to as Conus cuneolus form M in Röckel et al. (1980a). This form appears distributed mostly along the east coast of Santiago Island, with a transition area to the north where it intergrades with typical A. verdensis. Morphological studies of shell and radula had already suggested the close relationship of this morph with the taxon A. verdensis (Rolán 2005). The molecular phylogeny and mitochondrial genome sequence divergences support the conspecificity of nelsontiagoi with A. verdensis.   (Fig. 6L).

Geographical distribution
West and East Atlantic Ocean.

Least concern.
Remarks This is an amphiatlantic (i.e., present on both shores of the Atlantic Ocean) species and the only piscivorous cone in the Atlantic Ocean. Its venom has been subject of detailed proteomic (Rivera-Ortiz et al. 2011;Quinton et al. 2013;Echterbille et al. 2017) and transcriptomic ) studies. It is widely distributed along the West African coast, being locally abundant at certain locations, but nowadays it is much more scarce in the western Atlantic than in the past. For a long time it has been speculated that possibly the eastern and western Atlantic populations might represent two separate species. Until detailed molecular studies are performed including the entire mitogenome examination of western Atlantic individuals, we maintain all the populations from both shores of the Atlantic Ocean as members of the same species, namely Chelyconus ermineus.

Geographical distribution
West Africa, from Senegal to Angola including Cabo Verde Archipelago and São Tomé e Príncipe. There are occasional reports of live specimens found on the Canary Islands (Bettocchi 2018), where it is exceedingly rare.

Least concern.
Remarks This is a typical West African species, being the only representative of the genus Genuanoconus. The different names available in the literature have all been synonymised with G. genuanus. The name foscaclaudiae Assi, 2010 applied to a mere form of G. genuanus is unavailable according to the ICZN articles 45.5 and 45.6 (infrasubspecific; introduced as form after 1960). As discussed in Abalde et al. (2019), the previous placement of G. genuanus in the genus Kalloconus by Puillandre et al. (2014) was due to a misidentification of a sample in the original work (Cunha et al. 2005). A mitogenomic analysis clearly indicates that Genuanoconus is not a synonym of Kalloconus but a distantly related genus. Genuanoconus genuanus has an specialized vermivorous diet, preying on amphinomid polychaetes (fireworms) such as Hermodice carunculata (Pallas, 1766) (Rolán 2005). On the phylogenetic tree in Abalde et al. (2019), G. genuanus is recovered as sister to Rhombiconus imperialis Linnaeus, 1758, another specialised amphinomid polychaete worm eater from the Pacific Ocean. A novel bioactive guanine derivative named genuanine has been isolated from the venom duct of G. genuanus. This finding indicates that not only conopeptides, but also small molecules contribute to the activity of cone snail venoms (Neves et al. 2015).

Geographical distribution
Calheta Funda and Algodoeiro, southwest coast of Sal Island, Cabo Verde Archipelago.

Remarks
This beautiful species is endemic to Sal Island. Its pattern of white triangular or tessellate markings on a black background color seems fairly constant. Its classification as 'Endangered' resulted upon considering all the populations of Kalloconus on the island as members of this species. Several phylogenetic studies (Cunha et al. 2008;Abalde et al. 2017a;Tenorio et al. 2018) have shown that under the apparently same phenotype there are actually two cryptic sister species in Sal Island, namely K. ateralbus and K. marimaris. As a consequence, the distribution range of K. ateralbus has been drastically reduced to the southwest coast of Sal Island, an area that has underwent a large touristic development in the recent years, with subsequent pressure on marine habitats. Although the species seems to be fairly common within its distribution range, the limited geographical area and the human activities indicate that an upgrade in the risk status is most likely required. (Röding, 1798) Fig. 6O

Geographical distribution
Western Sahara and Mauritania.

Conservation status in IUCN Red List
Not evaluated.

Remarks
Specimens from the type locality have not been molecularly examined. However, the mitogenome of a specimen from Senegal with a shell morphology matching that of typical K. byssinus is almost identical to that of K. pulcher (Abalde et al. 2017a). There are no significant differences between the radular teeth of K. pulcher and K. byssinus. Although we consider K. byssinus as provisionally valid only, pending molecular examination of specimens from the type locality, it is however likely that this taxon represents just a phenotypic variation of the widespread K. pulcher, typical of the northern end of its distribution range along the West African coast.

Etymology
The name canariensis refers to its distribution range throughout the Canary Islands. The name 'canariensis' van Rossum, 1997 is an unavailable forma name (nomen nudum; no intention to introduce a new name, stated to be infrasubspecific after 1960) that has been occasionally and incorrectly been used to refer to specimens with pulcher-like shells from the Canary Islands. The name has usually been applied to subadult specimens with a characteristic sparse pattern of purplish brown dots and dashes aligned in a spiral on a creamy white background with two broad light brown spiral bands (Fig. 7I). We hereby rescue this name for the new taxon and provide a formal description compliant with ICZN standards.

Material examined
Holotype shell. Moderately large to very large. Maximum length: 167 mm. Shell profile conical to broadly conical, with convex sides adapically, straight or slightly concave below. Spire low, of straight to concave outline, some times slightly stepped. Protoconch white. Teleoconch whorls smooth, flat or slightly concave. Shoulder angulate or subangulate. Area below the shoulder often swollen, conferring a turnip-shaped appearance to certain specimens. Early teleoconch whorls white. Late teleoconch whorls white with radial brown irregular blotches. Ground color white. Last whorl overlaid in variable amount with alternating purplish brown and white spiral bars, as well as lines of dots and dashes in different shades of brown on light brown-orange and white broader spiral bands. Two sparsely patterned spiral bands located respectively above and below the midbody, where the white ground color predominates. Similar spiral band often present around basal region, just above the columella. Great variability in the color pattern, with shades of purple, brown and orange, which tend to fade with time to a more uniform brown color. Young specimens with characteristic sparse pattern of purplish brown dots and dashes aligned in spirals on creamy white background with two broad light brown to white spiral bands (Fig. 9I). Columella white to light purple. Aperture white. Periostracum yellow-brown, thin and translucent.

SPAIN -Canary Islands
aniMal. Sole of foot tan to light brown with brown streaks. Dorsum of foot of same color, overlaid with purplish brown irregular markings. Siphon yellowish white with dark purple tip. Mantle off-white, bordered with thin yellowish ribbon. Operculum small, teardrop shaped. radular tooth. Radula studied from two specimens, including paratype 1 (Fig. 7J). Radular tooth of medium relative size (S L /T L = 36-49), with short, but well-developed apical barb. Anterior section distinctly longer than posterior section of tooth (T L /AP L = 1.59-1.63). Waist rather indistinct, not very well marked. Blade moderately short and rounded, covering about one third to one half of anterior section (100B L /AP L = 39-48%). Serration with 50-55 denticles, arranged in major row flanked by numerous smaller serrations. Rounded terminating cusp present. A basal spur is present.

Distribution and habitat
Canary Islands. Its potential presence outside this archipelago (i.e., Madeira) needs confirmation.

Conservation status in IUCN Red List
Not evaluated.

Remarks
Mitogenomic analyses have revealed that the populations of pulcher-like cones from the Canary Islands actually correspond to a separate species hereby introduced as K. canariensis sp. nov., sister to K. pulcher (Figs 2, 8A) and most likely endemic to this archipelago. This species is present on all the islands, living between 1 and 50 m deep, most frequently between 5 and 20 m. The shell of K. canariensis sp. nov. usually is more brightly colored than that of K. pulcher, with shades of purple and orange in the pattern which are usually absent in the latter. In K. canariensis sp. nov. the area below the shoulder is often swollen, conferring a somewhat turnip-shaped appearance. However, these differences are not constant. Furthermore, there are no statistically significant differences in mean shell length or maximum diameter between K. pulcher and K. canariensis sp. nov.: a comparison of two sets constituted by 40 individuals of K. canariensis sp. nov. and 40 individuals of K. pulcher did not yield significant results in ANOVA tests using species as factor and S L (F = 2.96, p = 0.0892) or MD (F = 2.40, p = 0.1254) as variables. Likewise, and analysis of the covariance using species as factor, MD as variable and S L as covariate did not yield significant results either (F = 0.32, p = 0.5762). Despite the fact that individuals of K. pulcher (especially those coming from the Gulf of Guinea) may attain a greater maximum shell lenght than individuals of K. canariensis sp. nov., actually there is no statistically significant difference in mean shell lenght between the two species: mean S L = 96.7 mm for K. canariensis sp. nov., and mean S L = 110.2 mm for K. pulcher (t = -1.721, p = 0.089; U = 618.5, p = 0.082). There are no significant differences either in the radular teeth of both species (Fig. 7J-K), suggesting similar dietary habits. Hence, the separation of K. pulcher and K. canariensis sp. nov. based upon shell and radular morphologies cannot always be accurate, relying only on subtle differences in shell features and pattern. The name siamensis Hwass in Bruguiére, 1792 has been used in the literature to refer to these cones from the Canary Islands, either as a separate species (Monnier et al. 2018) or as a geographic subspecies (Röckel 1978(Röckel , 1982Monteiro et al. 2004;Filmer 2011). This name has been synonymized with pulcher ( [Lightfoot, 1786]) by several authors (Kohn 1992;van Rossum 1997). However, the type locality of siamensis is "Indian Ocean" (erroneous). The dimensions of the lectotype of Conus siamensis at MHNG (Fig. 9B) overlap completely with those of both K. pulcher and K. canariensis sp. nov. (Fig. 8). Hence, it is not possible the unambiguous assignation of the lectotype of C. siamensis as representative of the new species constituted by the individuals from the Canary Islands based upon its shell features only. Therefore, we consider the name siamensis as a junior synonym of pulcher in agreement with Kohn (1992), and the new name K. canariensis sp. nov. is hereby introduced for species from the Canary Islands.

Conservation status in IUCN Red List
Not evaluated.

Remarks
This recently described species was formerly identified as K. venulatus or K. ateralbus. Kalloconus venulatus is absent from Sal Island, and K. ateralbus has enough mitogenome sequence divergence to be considered a separate species, sister to K. marimaris (see above). The latter is split into two populations inhabiting areas to the east and west coasts of Sal Island. This species has some morphological similarities to K. ateralbus and hence the confusion. Kalloconus marimaris is very variable in background color, with yellow, orange, milky white or brown morphs, whereas in K. ateralbus the background color is always black. (Monteiro, Tenorio & Poppe, 2004) Fig. 9E

Geographical distribution
East and northeast coast of Boa Vista Island.

Conservation status in IUCN Red List
Least concern. Status revision required.

Remarks
This name was introduced in order to solve the existing confusion with the name Conus nivifer G.B. Sowerby I, 1833 (a synonym of Conus venulatus Hwass in Bruguière, 1792) which had been incorrectly applied to shells from Cabo Verde having a bluish white background with a distinctive pattern of spiral interrupting thick lines of deep purple or dark brown and a purple aperture (Monteiro et al. 2004). The specimens of this species are morphologically very close to K. trochulus (Reeve, 1844), and in fact the latter might be considered as a patternless version of K. pseudonivifer. Previous population genetic analyses European Journal of Taxonomy 663: 1-89 (2020) 56 showed that most specimens of individuals identified as K. pseudonivifer from Boa Vista and Maio islands clustered with K. trochulus, with the exception of one specimen from Praia Canto in northeast Boa Vista Island (Cunha et al. 2008). We examined the mitogenome of many individuals of alleged K. pseudonivifer from Boa Vista and Maio islands (Abalde et al. 2017a). As in Cunha et al. (2008), all individuals but the one from northeast Boa Vista Island clustered together with K. trochulus (Abalde et al. 2017a). Here, we sequenced the mitogenome of a second specimen from Praia Canto in northeast Boa Vista Island, which confirmed the segregation of these two individuals from the clade grouping K. trochulus with pseudonivifer-like individuals, at a genetic distance consistent with separate species status (Fig. 2). Given the fact that the type locality of the taxon Conus pseudonivifer is "Baía das Gatas", which is located to the northeast of Boa Vista Island, the name pseudonivifer is only applicable to specimens distributed along the east and northeast coasts of the island, whereas the specimens from other areas should be considered patterned forms of K. trochulus. ([Lightfoot], 1786) Fig. 9A-B

Geographical distribution
Boa Vista and Maio Islands, Cabo Verde Archipelago. Also present in Santiago Island, where it is very scarce.

Conservation status in IUCN Red List
Near threatened. Status revision required.

Remarks
Patternless-as well as patterned-shell forms of K. trochulus exist. The latter are widely distributed throughout the islands of Boa Vista, Maio, Santiago and offshore banks, and each of the existing populations has its own distinctive pattern. The name atlanticoselvagem was introduced for the offshore population inhabiting João Valente Bank in between the islands of Boa Vista and Maio (Fig. 9G). However, an analysis of the mitochondrial genome of patternless and patterned individuals from different populations recovered all of them within a compact clade, with the exception of two patterned individuals from Praia Canto, Boa Vista Island, which were recovered in a sister clade. These individuals are assigned to the taxon K. pseudonivifer (see above). Apart from this, the phylogenetic data suggest that all the other populations studied, both patterned and patternless, are members of one single, variable species that bears the name K. trochulus as the most senior. The hereby-designated lectotype of K. trochulus corresponds to the patternless form, particularly abundant in the northwest coast of Boa Vista Island. The recently described Africonus cazalisoi Cossignani & Fiadeiro, 2018 was incorrectly assigned to the genus Africonus. It is a representative of the population of sparsely patterned K. trochulus from Morro de Areia, southwest of Boa Vista Island, which was thoroughly studied in Cunha et al. (2008), and hence another synonym for this species. The same applies to K. stanchinensis, which is just the patterned form of K. trochulus from Estancinha to Ponta do Sol area, northwest of Boa Vista Island (Fig. 9H). Representative individuals from the Santiago Island populations are figured in Rolán (2005).

Geographical distribution
Boa Vista and Maio islands, Cabo Verde Archipelago. Also present in Santiago Island, where it is very scarce.

Least concern.
Remarks This is the largest-sized cone snail endemic to the Cabo Verde Archipelago, where it is widely distributed in Boa Vista and Maio islands. There are occasional reports of its presence off Santiago Island as well (Rolán 2005). Very variable in shell length and pattern, it has received a number of different names, which are all considered synonyms. Detailed populations genetic studies of the venulatus group (Cunha et al. 2008) and more recent mitogenomic phylogenetic analyses (Abalde et al. 2017a) indicate that despite the extreme phenotypic variability, all the populations scattered throughout the archipelago are members of one single species. This is also applicable to the recently introduced taxon K. josefiadeiroi, which is just a representative of the population of K. venulatus from Praia Canto, northeast coast of Boa Vista Island (Fig. 9J) and is included in our phylogenetic study.

Material examined
Holotype (not examined) The holotype was in the collection F.J. Stainforth, present whereabouts unknown; neotype in ZMA (Coomans et al. 1980). The original type locality was not known. For neotype: coast of Senegal, West Africa (Coomans et al. 1980).

Geographical distribution
Senegal (south of Dakar), Gambia and northern Guinea Bissau.

Conservation status in IUCN Red List
Least concern.
Remarks This species has not been molecularly examined, but the distinctive features of its shell and radula warrant its consideration as a valid species. The taxa names griseus and miser are considered synonyms of M. ambiguus, whereas gernanti represents a pale violet color form of this species without taxonomical value. (van Rossum, 1996) Fig. 10B

Geographical distribution
Off Guinea Conakry.

Conservation status in IUCN Red List
Data deficient.

Remarks
This is a rare species, first obtained from fishing nets of trawlers operating in waters off Guinea Conakry in the 1980s. Since the fishing activities ceased, no more specimens have been brought to surface. Hence, it has not been molecularly examined. Its radular morphology is also unknown. The shell of this species is very similar to that of M. ambiguus, but it is characterized by the presence of numerous irregular reddish axial lines. It might represent just a mere patterned form of M. ambiguus from a particular geographic area. We hereby consider it as a provisionally valid species, pending future studies that might eventually validate its status as such, or alternatively demonstrate its conspecificity with M. ambiguus. (Reeve, 1844) Fig. 10C

Material examined
Holotype (not examined) Holotype in ZMUC, GAS-000366. Since May 2012 the specimen is not present in the collection. The curator team was notified, but apparently the specimen has not been located hitherto. The type locality is not known.

Geographical distribution
West African coast from Senegal to Angola including Cabo Verde Archipelago and São Tomé e Príncipe.

Least concern.
Remarks This is a fairly common, widespread species that has been molecularly examined as a representative of genus Monteiroconus. Historically, this species has been mistaken for M. ambiguus. However, there are constant features in shell morphology that allow a distinction between these two species, mainly the larger size and angulated, broader shoulder in the latter. In any case, a molecular examination of M. ambiguus might clarify whether the separation from M. tabidus at the species level is justified. We provisionally consider both of them as separate valid species.
Genus Varioconus Da Motta, 1991Da Motta (1991 originally introduced the taxon Varioconus as a subgenus of Conasprella Thiele, 1929 using the contour of the shell as criterion. The subgenus included West African species from Angola, Senegal and elsewhere, but not from the Cabo Verde Archipelago. Later, Tucker & Tenorio (2009) elevated the rank of Varioconus to genus, and only maintained in it the endemic cones from Angola, whereas most of the remaining West African cones were placed in the genus Lautoconus. Now, we find that the reconstructed phylogeny and the observed genetic distances support the original proposal by Da Motta (1991), and the genus Varioconus actually includes species from Angola, Senegal, Gambia and other locations like Canary Islands or Saint Helena. The species of Varioconus endemic to Angola are not included in the present revision. These taxa are being sequenced systematically, and the results will be reported elsewhere in due course. (Pin & Leung Tack in Pin, 1989a) comb. nov. Fig. 10D Conus belairensis Pin & Leung Tack in Pin, 1989a: 33, pl. 1, figs 2-4, pl. 2A, figs 1-2.

Geographical distribution
Dakar area including Madeleine Island, and to the south along the Petite Côte, Senegal.

Conservation status in IUCN Red List
Endangered. Status revision required.

Remarks
This is a valid, phenotypically variable species, which is closer to V. reticulatus than to V. mercator, both phylogenetically and according to its radular morphology (Abalde et al. 2017b). The distinction between V. belairensis and V. reticulatus based upon conchological or even radular features might not be effective in all cases. Whereas the original population from Bel-Air in Dakar seems to be extinct due to pollution from the large harbor of the city, the species occurs in many other locations around Dakar and to the south. As it happens with most species of cone snails living on the peninsula of Dakar, the individuals seem to have adapted very well to anthropogenic conditions, and populations thrive even in densely populated areas as long as they are not affected by chemical contamination, i.e., fuel from port activities. (Kiener, 1846) comb. nov. Fig. 10E Conus bruguieresi Kiener, 1846: pl. 56, fig. 2. Conus bruguieri Kiener 1849: 221 (incorrect subsequent spelling).

Material examined
Holotype (not examined) The holotype was in the collection of Madame Dupont, its present whereabouts are unknown. The type locality was not stated, but was designated Dakar, Senegal by Coomans et al. (1982).

Geographical distribution
Dakar area, Senegal.

Conservation status in IUCN Red List
Endangered.

Remarks
This species is characterized by its elongated greenish shell. The name Conus bruguieresi was originally published on plate 56, fig. 2 by Kiener, ahead of the text, which appeared published in 1849. In the text, the name used for this taxon was Conus bruguieri, which was a misspelling leading to subsequent misinterpretations in the literature (e.g., Wagner & Abott 1978). Even though there was no description, the publication of the plate with a name (caption) on it is sufficient to make bruguieresi available as the earlier valid name for this taxon (Coomans et al. 1982;Faber 2011). Varioconus bruguieresi is very closely related to its sister species V. echinophilus, up to the point that juvenile individuals of the former can not be discriminated based upon conchological features only. (Walls, 1978) comb. nov. Fig. 10F Conus cloveri Walls, 1978: 2, pl. on p. 5. Conus soaresi Trovão, 1978 fig. 1-1a.

Geographical distribution
Dakar area and to the south along the Petite Côte, Senegal.

Conservation status in IUCN Red List
Endangered.

Remarks
The name soaresi for this taxon appeared published shortly after the description of Conus cloveri, which became the senior name for this species. This is a distinct species distributed from Pointe des Almadies to the south of Dakar along the Petite Côte. Two color forms exist, one with wavy brown narrow axial lines that overlap to produce fine tents on a creamy white background, and another one white, almost patternless with only faint traces of milky white axial lines. The latter seems to be more common at the southern end of the distribution range for the species. In the mitogenome phylogeny, V. cloveri is sister to V. mercator. (Monnier & Limpalaër, 2010) comb. nov.

Geographical distribution
Known only from the type locality.

Conservation status in IUCN Red List
Near threatened.

Remarks
The mitogenome is not available, as we have not examined this taxon. However, partial sequences of cox1, rrnL and rrnS genes were reported (Puillandre et al. 2014). In the corresponding phylogeny, V. dorotheae appeared closely related to V. pineaui (a synonym of V. guinaicus) and V. guanche (Puillandre et al. 2014). Radular morphology also confirms its placement in the V. guinaicus group. Given its geographically restricted habitat on the west coast of Madeleine Island in relatively deep water (between 10 and 15 m) (Monnier & Limpalaër 2010), we retain this taxon as provisionally valid in spite of its close phylogenetic relationship to V. guinaicus. Examination of its mitogenome in the future might confirm this assumption, or alternatively prove otherwise. (Petuch, 1975) comb. nov. Fig. 10G Conus echinophilus Petuch, 1975b: 180, figs 1-3, 7.

Material examined
Holotype (not examined) The holotype is in CAS with catalog number 55454 according to the original publication. However, Coomans et al. (1986) stated that one of the paratypes (CAS 61495, 10.7 × 6.0 mm) was deposited instead of the holotype, which was still missing in 1988 (Filmer 2011 (Fig. 10G).

Geographical distribution
Dakar area, Senegal.

Conservation status in IUCN Red List
Endangered.

Remarks
This is a species very closely related to V. bruguieresi. In fact, in a previous phylogenetic analysis (Abalde et al. 2017a), the mitogenome of one juvenile specimen initially identified as V. cf. echinophilus turned out to be identical in sequence to that of V. bruguieresi. However, the mitogenome of another specimen of V. echinophilus sequenced here was recovered as sister to that of V. bruguieresi, at a genetic distance consistent with a valid species (Fig. 2). Bruguière, 1792)

Geographical distribution
Dakar area, Senegal.

Conservation status in IUCN Red List
Endangered (evaluated as hybridus).

Remarks
The lectotype of the taxon Conus franciscanus Hwass in Bruguière, 1792 designated by Kohn (1968) is at the MNHG. It measures 56.6 mm (stated 55 mm in Kohn 1968). It has been a controversial species, and several authors have synonymized this name either with Conus ventricosus Gmelin, 1791(Kohn 1992 or Conus guinaicus Hwass in Bruguière, 1792(Röckel 1989Lauer 1993b;Filmer 2011). The homonym name franciscanus (variety mediterraneus Hwass) Philippi, 1836 applicable to a color form of ventricosus further contributed to increase the confusion. The type locality ("Africa") and the size and general appearance of the lectotype of franciscanus are more consistent with a guinaicuslike species than with the Mediterranean L. ventricosus. In this sense, we agree with the opinions of Röckel (1989) and Lauer (1993b), also followed by Filmer (2011) and Tucker & Tenorio (2013). Specimens matching the lectotype of V. franciscanus are relatively abundant around Gorée Island in Dakar. These uniform dark brown shells with a white central band had provisionally been assigned to the taxon V. guinaicus, despite being lighter and more elongated than the typical form of the latter species (Monteiro et al. 2004). We examined the mitogenome of one of such dark guinaicus-like specimens from Gorée Island. It showed almost no sequence divergence with a specimen of V. hybridus from NGor. This result was confirmed by sequencing of the universal cox1 gene fragment (Folmer et al. 1994) of additional guinaicus-like specimens from Gorée Island (Abalde et al. 2017b). Typical V. guinaicus from Popenguine are recovered as the sister clade. These results indicate that the guinaicus-like specimens from Gorée Island, representatives of the taxon V. franciscanus, are essentially identical to V. hybridus, and different from typical V. guinaicus. If we assume the conspecificity of the taxa franciscanus Hwass in Bruguière, 1792 and hybridus Kiener, 1847, then the latter becomes a junior synonym to represent a form of franciscanus, which stands as the senior name for the species.

Geographical distribution
Canary Islands (absent in the western islands), southern Morocco, West Sahara and Mauritania.

Conservation status in IUCN Red List
Least concern.
Remarks Until its introduction as a new species by Lauer (1993a), there was a large confusion regarding certain populations of cones present in the Canary Islands, referred to as ventricosus, guinaicus or desidiosus (Bandel & Wils 1977;Coomans et al. 1985b;Raybaudi 1992). The taxon Lautoconus desidiosus (A. Adams, 1853) is nowadays considered a form of L. ventricosus (Filmer 2011) or a separate species (Monteiro et al. 2004;WoRMS, accessed March 2020) depending on the author, but in any case is restricted to the Mediterranean Sea and neighboring Atlantic and is not a West African species. In the phylogeny, V. guanche appears deeply nested within the Varioconus clade together with Senegal endemics and is only distantly related to L. ventricosus. It is indeed sister to the species V. unifasciatus, and close to V. guinaicus (Abalde et al. 2017b). The non-planktonic development of V. guanche supposedly limits its capacity of dispersal, resulting in populations having characteristic shell colors and patterns. However, phylogenetic studies focused on V. guanche showed no differentiation of the species within the Canary Islands or between the archipelago and the western African coast indicating recurrent gene flow . These findings clearly indicate that the recently introduced taxon Lautoconus saharicus Petuch & Berschauer, 2016 represents the local form of V. guanche from Dahkla Bay, Western Sahara. It lacks taxonomical value and must be considered a junior synonym. The name nitens, initially introduced as a subspecies of V. guanche, is applied to high-spired shells typically from Lanzarote. However, typical V. guanche and high-spired individuals co-occur, hence the separation at subspecific level seems unjustified and nitens represents a synonym (form) of V. guanche.

Geographical distribution
Senegal, Gambia, Guinea-Bissau, and Guinea. Also reported from Ghana.

Conservation status in IUCN Red List
Vulnerable. Status revision required.

Remarks
This is the most representative taxon of a very compact clade, which includes a number of very closely related species, and certain populations with minimal differences in mitogenome sequences that are now considered synonyms (forms), such as taslei or pineaui (Fig. 10O). All these taxa share the same radular tooth morphology: narrow and elongated radular tooth with the anterior section longer than the posterior section, rounded blade covering less than half of the anterior section, and more than 40 denticles in the serration (Abalde et al. 2017b). The population from Gorée Island, characterized by its dark brown shell, does not correspond to the taxon guinaicus but to its sister species V. franciscanus. The distribution range of V. guinaicus is quite large, from Dakar area south to Guinea including Gambia, and also reported from Ghana. The taxon V. wolof (Fig. 10N), recently described from Gambia, has been sequenced. The sequence of its mitogenome is essentially identical to that of V. guinaicus (Fig. 2). Therefore, this name represents a mere local form of V. guinaicus and is hereby considered a synonym. (Linnaeus, 1758) . 11G).

Conservation status in IUCN Red List
Endangered. Status revision required.

Remarks
This is one of the most classical species from Senegal. Specimens from Gorée Island (Fig. 11B) match perfectly the lectotype of this species (Fig. 11A), which is widely distributed from Dakar area to the south. This is a polymorphic species. Several taxon names introduced in the literature as distinct species (e.g., cacao and senegalensis) correspond to the taxon V. mercator. Abalde et al. (2017b) showed that two different morphs identified as V. mercator, one from Goreé Island (Fig. 11B) and another from Pointe des Almadies (Fig. 11C), were actually separated by a genetic uncorrected p distance at exactly the threshold that delimits species. Based upon these grounds, the name V. stimpsonorum has been recently introduced for the population of mercator-like shells from Pointe des Almadies (Cossignani & Allary 2019). However, the subtle differences in the reticulated shell pattern (coarse in V. stimpsonorum but fine in V. mercator) do not justify the introduction of a new name. Furthermore, a transcriptomic study of the venom gland of two specimens of V. mercator, one from Almadies (i.e., form stimpsonorum) and another from Ndayane (form cacao), has shown that both individuals share a significant percentage of conotoxin precursors in their venom repertoires within the expected intraspecific variation (Abalde et al. 2020). Therefore, we consider that the taxon V. stimpsonorum falls within the genetic and morphological variability of V. mercator and propose to consider it a mere form of the latter. Specimens of the species recently described from Gambia, including fernandi (Fig. 11D), gambiensis (Fig. 11E) and rikae (Fig. 11F), have been sequenced. The sequences of the mitogenomes of these taxa were almost identical, and showed a minimal divergence to that of V. mercator (Fig. 2).
The type material of the taxon Conus orri Ninomiya & da Motta, 1982 (Fig. 11G) described from Gambia is consistent with the observed variability of V. mercator. We therefore also consider orri as a synonym of the latter.
Conversely, a mitogenome examination has shown that certain names traditionally synonymized with V. mercator (e.g., reticulatus) actually correspond to separate, valid species (see below). As it happens to most species living around the Dakar peninsula, V. mercator seems to be very resilient and has adapted very well to environments under heavy human influence as long as they are not affected by chemical or industrial contamination. (Born, 1778)

Conservation status in IUCN Red List
Not evaluated.

Remarks
The taxon Conus reticulatus Born, 1778 had been traditionally synonymized with Conus mercator Linnaeus, 1758(Monteiro et al. 2004Filmer 2011;Tucker & Tenorio 2013). The phylogenetic analyses suggest that V. reticulatus actually is a valid species, distinct from V. mercator and sister to V. belairensis (Abalde et al. 2017b). Varioconus mercator and V. reticulatus also differ in radular tooth morphology. A separation of V. reticulatus and V. mercator based upon the shell pattern only is not always accurate, and may eventually require confirmation by DNA or radular morphology analysis. Specimens collected along Petite Côte near Ndayane and initially identified as juveniles of V. mercator form cacao turned out to be V. reticulatus. (Boyer & Pelorce, 2009)

Geographical distribution
Dakar area, Senegal.

Conservation status in IUCN Red List
Vulnerable.

Remarks
This species has not been molecularly examined. In the original description it was compared to Conus desidiosus (which does not live in Senegal), bruguieresi and echinophilus. The species is morphologically related to the latter two, and might represent a variation of one of these. We consider this species as provisionally valid only, pending future molecular examination. (Nolf & Verstraeten, 2008) comb. nov. Fig. 11K Conus trencarti Nolf & Verstraeten, 2008: 1, pls 1-2.

Geographical distribution
Dakar area, Senegal.

Conservation status in IUCN Red List
Near threatened.

Remarks
This is another member of the V. guinaicus clade, with a close phenotypical resemblance to V. guinaicus form pineaui. The sequence divergence of the mitogenome of one individual from the type locality suggests that this is a valid species, sister to V. guanche and V. unifasciatus (Fig. 2). At variance with most other species of Varioconus from Senegal, V. trencarti has a relatively deep-water habitat, between 20 and 40 m, and a restricted distribution range around Pointe des Almadies, Dakar. (Kiener, 1850) comb. nov. Fig. 11L Conus unifasciatus Kiener, 1850: 61, pl. 110, fig. 4.

Geographical distribution
South of Dakar, Petite Côte area, Senegal.

Conservation status in IUCN Red List
Endangered.

Remarks
This species is sister to V. guanche, and is characterized by its uniform dark brown shell, with the occasional presence of lighter-colored spiral bands just as in the holotype. It has been synonymized with V. franciscanus in Petuch & Berschauer (2016). However, according to the phylogeny, the latter is a distinct, valid species sister to V. guinaicus. In any case, the guinaicus clade is very compact, and all its members are closely related to each other.

Discussion
In the present work, we have reviewed in detail the species of cone snails from West Africa, excluding the endemics from Angola. The analysis of almost complete mitogenome sequences for 131 individuals has proved to be of utmost importance in the reconstruction of robust phylogenetic trees, which were used as an evolutionary framework to revise the taxonomy of these gastropods (Abalde et al. 2017a(Abalde et al. , 2017bthis work). As of December 2019, 178 species names for West African cone snails have been introduced, taking into consideration the revision by Monteiro et al. (2004) as a starting point. After the present revision, 100 species are considered valid or provisionally valid. Additionally, two new species have been introduced herein, elevating the total number of species to 102 (Figs 1, 12). This number represents most likely an overestimation, as it still provisionally considers valid all the endemic species from Angola listed in Monteiro et al. (2004) plus all the species described from this area since then (a total of 35 species). The phylogeny of Angolan cone snails is currently under study, and results will be reported in due course. Hence, about 40% of the named species from the West African region are considered invalid, as they constitute junior synonyms describing particular color forms of other polymorphic species, indicating that shell colors and patterns are prone to convergence, and hence rather homoplasious (Fig. 12). In the case of the endemic cones from the Cabo Verde Archipelago, the number is reduced from 108 named species (100 Africonus + 8 Kalloconus) to a total of 48 valid or provisionally valid species (43 Africonus + 5 Kalloconus). This indicates that less than one half of the names introduced for Cabo Verde endemic species of cones correspond to at least provisionally valid species. The main reason for this huge taxonomic inflation comes most likely from the rapid publication of taxa names in amateur magazines addressed to shell collectors, without proper peer review and based exclusively on shell material (actually only a minimum part of these species turned out to be valid). However, even in the studies in which several independent traits have been taken into consideration (i.e., animal features, radular morphology, egg capsules, or even cox1 gene fragment sequencing), the very close phylogenetic proximity among the different taxa have made the delimitation of species very hard. The same applies for many cone species recently described from other areas of West Africa, such as Senegal or Gambia. In these areas, the shell pattern or distribution range represent traits that in most cases are not reliable enough as proxies for species separation, unless proven otherwise by discriminant molecular Fig. 12. Stacked bar charts of the accumulated number of species described in the period 1974-2020 before (A) and after (B) revision, showing the contribution to the total number of species of the Cabo Verde endemics (blue), Angola endemics (red) and the remaining species from West Africa (green). Note that Angolan cone endemics still await a thorough genetic revision.  (Reeve, 1849) as an example of an increase in range distribution, and thus an estimated decrease on its risk status. B. The case of Kalloconus ateralbus (Kiener, 1850) as an example of a decrease in range distribution, and thus an estimated increase on its risk status.
(i.e., mitogenomes and ideally sets of nuclear markers to detect potential instances of hybridization or incomplete lineage sorting) methods. Hence, we strongly suggest that new taxa names for cone snails from West Africa based upon shell characters only and not supported additionally by robust molecular studies in the original descriptions should be treated with utmost caution. In spite of meeting the ICZN criteria to be accepted as available names, their status as valid species should be put on hold by the scientific community until proper confirmation, in order to avoid unnecessary taxonomic inflation. It is interesting to note that the accumulated number of valid or provisionally valid species follows a lineal trend (R 2 = 0.9855) since the year 1974, with a slope value of 1.5. This means that during this period only 1-2 valid species have been discovered per year on average in the West African region. Over time, this rate will most likely be reduced, as the number of species is not expected to grow indefinitely.
The drastic reduction of the number of valid species results in important changes affecting the distribution ranges of many of these taxa. In most cases, distribution ranges are extended due to synonymy, but in several instances they are very much reduced due to the discovery of cryptic species. One of the former cases corresponds to A. crotchii, which was restricted to the southeast coast of Boa Vista Island but now is found to be widely distributed on the island (Fig. 13A). One of the latter cases corresponds to the taxon K. ateralbus, endemic to Sal Island in Cabo Verde. It was classified as 'endangered' in the IUCN Red List (Tenorio 2012), but now its distribution range (southwest of Sal Island) appears much more restricted than initially considered due to the assignment of some of its populations to K. marimaris (Fig. 13B). Therefore, a new assessment of the risk status of this species using the new information hereby provided will likely result in an upgrade of its Red List category (Peters et al. 2016). Likewise, studies on ecology and biogeography relying on the inventory of cone snail species from eastern Atlantic areas (i.e., Freitas et al. 2019) should take into consideration the results of the present work. This has been the case of the estimation of speciation rates for Cabo Verde cone snails that varied considerably when taking into consideration the number of valid species based on genetic criteria (Phuong et al. 2019). We are currently extending our mitogenome sequencing approach to address the phylogenetic relationships of cones from Angola as well as from the Mediterranean and nearby eastern Atlantic region. Results will be reported in due course, and the conclusions will undoubtedly have an effect on the figures of the cone snail fauna diversity in these regions.