After decades of ocean warming, an increasing number of works have postulated a tropicalization of species assemblages in temperate biogeographic transition zones such as southwestern Australia (Wernberg et al. 2016), some parts of the Mediterranean (Horta Costa et al. 2014), the NW of the Iberian Peninsula (Cuesta et al. 2016) and the Macaronesian archipelagos in the northeastern Atlantic (Brito et al. 2005, Wirtz et al. 2008, Afonso et al. 2013). This process has been associated with global warming in many cases (Perry et al. 2005, Occhipinti-Ambrogi 2007, Wernberg et al. 2016). In response to warmer conditions, marine species tend to shift their distributions to higher latitudes, and organisms are predicted to keep shifting under climate change scenarios (e.g. Perry et al. 2005). Some recent reviews have accounted for this topic in temperate marine ecosystems, emphasizing that climate-driven changes in biotic interactions can profoundly alter ecological communities and reconfigure ecosystems, and are expected to have important social and economic implications (Vergés et al. 2014, Wernberg et al. 2016).

Within the Macaronesian ecoregion (Azores-Madeira-Canaries) of the Lusitanian biogeographic province, the Canary Islands region occupies its southern frontier, limiting with the warmer West African Transition province (Spalding et al. 2007). Moreover, the Canary Islands are located in front of the northwestern coast of Africa (104 km from Morocco). The Canaries are under the influence of the subtropical gyre of the eastern-central Atlantic, which facilitates the transport of planktonic larvae and rafting organisms to the archipelago. Average seawater temperature around these islands is 18.5°C in February, rising abruptly to 24°C in August-September (Barton et al. 1998). A distribution of larval communities has been described in filaments of the upwelling system from the African coast reaching the archipelago (e.g. Landeira et al. 2009). Also, the Canaries are geographically located on a very important maritime route, and both ships (Brito et al. 2011, González et al. 2012) and oil platforms (Triay-Portella et al. 2015, Pajuelo et al. 2016) have been recognized as major vectors for the introduction of non-indigenous species.

Human-caused introduction and spread of non-native species are considered one of the main threats to marine biodiversity, and the degree of interest in introduced species has increased worldwide in recent years (Streftaris et al. 2005, Molnar et al. 2008, Galil et al. 2011). Biological invasions are a growing environmental problem because non-indigenous organisms may affect the structure and functioning of native communities (Molnar et al. 2008, Galil et al. 2011, Sant’Anna et al. 2012).

During the last 30 years, the occurrence and even establishment of non-native marine species around the Canary Islands have been reported, and the arrival of subtropical/tropical fishes (Brito et al. 2005, Triay-Portella et al. 2015, Pajuelo et al. 2016) and crustaceans (González et al. 2012, Moro et al. 2014, González 2016) has increased.

To date, according to González (2016), the portunoid family Portunidae Rafinesque, 1815 is represented in Canary Islands waters by four sublittoral species: Laleonectes vocans (A. Milne-Edwards, 1878), Portunus (Portunus) hastatus (Linnaeus, 1767), Portunus (Portunus) inaequalis (Miers, 1881) and Thalamita poissonii (Audouin, 1826).

This paper records for the first time the presence and spread of the pantropical crab Cronius ruber (Lamarck, 1818) (Portunidae) in the Canary Islands. Previously known from off Cape Verde Islands and Dakar, Senegal (Monod 1956, Manning and Holthuis 1981, Fransen 2015), this is the northernmost record of the species in the eastern Atlantic. Crabs have been identified by both morphological characters and DNA barcoding (16S). The reasons for occurrence of this species at this latitude are discussed herein.

An experienced fisherman working in the area provided us with photographic evidence (they were eating a crab new to the Canaries!) and then with four live specimens of a brachyuran species unknown to him collected at Castillo del Romeral, on the southeastern coast of the island of Gran Canaria (Canary Islands, northeastern Atlantic) in early June 2016. This first set of crab individuals were captured by hand from hard bottoms, in sublittoral caves at depths between 8 and 10 m (approximately at 27°47′31″N 15°28′22″W) (Fig. 1).

From late June to early July 2016, in the framework of an overall ecological study on brachyuran crabs from sand with Cymodocea nodosa seagrass meadows at Bahía de Santa Águeda, on the southwestern coast of Gran Canaria, some individuals of the same brachyuran species were observed and video-recorded, and then a second set (22 specimens) were captured on these soft substrates, at depths between 2 and 10 m (approximately at 27°45′13″N 15°40′46″W). These crabs were caught by means of a small, ad hoc collecting trap for sampling in shallow waters. Each trap measured 46×20 cm base length and 20 cm in height, was covered with 1×2 mm mesh, and had one 16×6 cm truncoconical opening (Triay-Portella et al. in prep.). Each trap was baited with pieces of frozen Spanish mackerel.

Each crab specimen was taxonomically identified in the laboratory by checking their morphological characters with suitable bibliography on Atlantic subtropical/tropical portunids (i.e. Monod 1956, Manning and Holthuis 1981, Spiridonov et al. 2014). Systematic classification follows Ng et al. (2008), but also takes into account the latest changes in Portunoidea (Schubart and Reuschel 2009, Spiridonov et al. 2014).

For the purposes of the present paper, standard carcinological measurements were taken using Vernier callipers (±0.1 mm), following Attrill et al. (1991) and the update by Biscoito et al. (2015): carapace width (CW), defined as the maximum transversal diameter at the midline of the carapace, excluding the ninth dorso-lateral spines; and carapace length (CL), defined as the longitudinal distance from the rostral teeth to the posterior carapace margin. CW and CL were taken on the upper side of the crab, in the left-right and anterior-posterior direction, respectively. Sex (based on abdomen morphology and presence/absence of gonopods) and ovigerous condition (based on the presence of external or remaining eggs on the pleopods) were also recorded. Number, sex, ovigerous condition and size of all vouchers by type of substrata are presented in Table 1.

The only ovigerous female collected was used for the morphological study of eggs and to preliminary estimate fecundity. Following Triay-Portella et al. (2014), pleopods with attached eggs were removed, and egg mass was then placed on a 100-mm mesh, washed and isolated from pleopods. After removal of the excess wash water, 40 eggs were measured in sea water.

Voucher specimens were labelled, curated, databased and deposited, available for verification (Turney et al. 2015), in the ICCM study collection at the University of Las Palmas de Gran Canaria.

Traffic information of oil platforms in the Port and Bay of Las Palmas between 2010 and 2015 was obtained from local newspapers, port logs and annual reports of activities of the port authority. Information gathered allowed Pajuelo et al. (2016) to build the main routes of oil rigs arriving in Gran Canaria. These routes and the chronology of voyages were used to check coincidences with the occurrence of the newly recorded crab.

DNA barcode sequences were obtained for a male specimen (76.13 mm CW, 49.05 mm CL) of Cronius ruber collected off the Canary Islands and deposited in the ICCM study collection under the accession code ICCM422. Equivalent sequences were also obtained for a female specimen of Cronius ruber from Guinea deposited in the IEO Crustacean Decapod Collection at Cadiz Oceanographic Centre under the accession code IEO-CD-CCLME11/719.

The molecular confirmation of the identification of the crabs was based on a partial sequence of the mitochondrial gene 16S mtDNA, but a partial sequence of the Cytochrome c oxidase subunit I (COI) was also obtained. Total genomic DNA was extracted from muscle tissue from one pereiopod, following a modified Chelex 10% protocol by Estoup et al. (1996).

Target mitochondrial DNA from the 16S rRNA and COI genes was amplified with polymerase chain reaction (PCR) using the following cycling conditions: 2 min at 95°C, 40 cycles of 20s at 95°C, 20s at 45°C (COI) or 48°C (16S), 45s (16S) or 47s (COI) at 72°C, and 5 min at 72°C. Primers 1472 (5′- AGA TAG AAA CCA ACC TGG-3′) (Crandall and Fitzpatrick 1996) and 16L2 (5′-TGC CTG TTT ATC AAA AAC AT-3′) (Schubart et al. 2002) were used to amplify 516 bp of 16S, while primers COH6 (5′- TAD ACT TCD GGR TGD CCA AAR AAY CA -3′) and COL6b (5′- ACA AAT CAT AAA GAT ATY GG -3′) (Schubart and Huber 2006) were used to amplify 634 bp of COI. The PCR products were sent to Stab-Vida Laboratories to be purified and then bidirectionally sequenced.

Sequences were edited using the Chromas software version 2.0. With the obtained final DNA sequences, a BLAST search was executed on the NCBI webpage to get the sequence that best matched. All DNA sequences obtained were deposited in GenBank under accession numbers KY111025 to KY111028.

An evolutionary distances analysis was carried out in MEGA6 (Tamura et al. 2013). Alignment of 16S sequences was built using sequences obtained from the specimens studied as well as from the other congeneric species downloaded from GenBank (http://www.ncbi.nlm.nih.gov, FJ152143 to FJ152146). The phylogeny reconstruction analyses were inferred from neighbour-joining using the p-distance method. The nodal confidence of the obtained topologies was assessed via 2000 bootstrap replicates (Fig. 2).

One brachyuran decapod species, Cronius ruber, has been newly recorded for the Canary Islands herein, and this represents the northernmost record (27º47’31”N) of the species in the eastern Atlantic Ocean (Fig. 1). This is also the first record for the genus Cronius Stimpson, 1860 from the Canary Islands, and C. ruber is the fifth portunid (Portunidae) recorded from the Canaries (cf. González 2016). The present newly accounted species increases the Canarian brachyuran carcinofauna to 133 marine species (cf. González 2016).

All portunids previously occurring in Canary Islands waters appear to exhibit similar distributional (shallow-water species) and warm-affinity biogeographic patterns, as well as nocturnal and more or less sciaphilic habits. All of them inhabit from a few metres to a maximum depth of 60 m (P. hastatus), with the exception of T. poissonii, which reaches a maximum depth of 110-120 m (cf. González 2016).

L. vocans is an amphi-Atlantic species of warm affinity; its occurrence in the Canaries is based on 6 individuals (4 from Tenerife, and 2 from El Hierro) found in sublittoral sciaphilic caves (5-10 m) between 1994 and 2014 (cf. González 1995 as Portunus sp., Moro et al. 2014).

P. hastatus is a Guinean species first recorded from the Canaries by Brullé (1839, as Lupa hastata) and, since that time, regularly reported as a common species with a moderate well-established population nowadays (cf. González 1995, 2016); adults are typically found on sand and sandy-mud bottoms (González 1995), but juveniles have pelagic stages (Türkay 1987, reported on five juveniles near La Palma).

P. inaequalis is a Guinean species; it has recently been first recorded from the Canaries by González (2016), based on the finding of one individual from Tenerife (cave, 8 m, in 2014) and another from Gran Canaria (at the base of a breakwater, 12 m, in 2015).

T. poissonii is a pantropical species first recorded from the Canaries by Brullé (1839, as T. admete) and then regularly reported as a common species with apparently a moderate well-established population nowadays (cf. González 1995, 2016, Moro et al. 2014).

However, according to communications by fishermen working in the area, individuals of C. ruber suddenly appeared on the southern coast of Gran Canaria five or six years ago, resulting in a rapid colonization of sublittoral caves in rocky shores. It is not possible to establish a chronology on the colonization of these hard bottoms and also soft substrates characterized by sand with seagrass meadows (adjacent to rocky bottoms). It probably colonized both habitats based on its biological and ecological traits (relatively large size, aggressiveness, swimming capacity) (e.g. Spiridonov et al. 2014), with the help of a set of local disorganized intertidal and sublittoral ecosystems with many vacant niches.

In the marine environment there are records of invasive species including crustaceans (Tavares and Mendonça 1996, Dauvin et al. 2009, González et al. 2012), and portunids of genera Charybdis De Haan, 1833 (Dineen et al. 2001, Tavares and Amouroux 2003, Sant’Anna et al. 2012), Callinectes Stimpson, 1860 (Molnar et al. 2008, Minchin et al. 2013, Manfrin et al. 2015) and Scylla De Haan, 1833 (Tavares and Mendonça 2011) are among the most widespread invasive groups.

The assessment of species-level ecological traits associated with the rafting behaviour has recently been proposed by Luiz et al. (2015) to better understand, and even predict, the species’ capacity for long-distance oceanic dispersal in isolated areas of benthic habitats. They used a dual classification of seafarer or castaway species. Rafter species are "seafarers” because they possess traits that facilitate long oceanic journeys, whereas species marooned by chance in the flotsam are "castaways”. Knowing their biological and ecological traits and their dispersal capacity, Canarian portunid species could be evaluated with the criteria of Luiz et al. (2015). Also, this information combined would be useful for discussing the establishment status of these species at the Canaries, and even for predicting future displacements to the north within the current scenario of tropicalization. In this regard, L. vocans and P. inaequalis, being markedly sciaphilic forms restricted to caves and similar habitats, seem to be castaway species. P. hastatus and T. poissonii, being capable of colonizing sandy and mixed (sandy gravels) bottoms, would be seafarer species. Finally, C. ruber, being capable of colonizing all types of substrata (from caves to sand vegetated bottoms), seems to possess the necessary ecological traits to be a typical invasive species. Therefore, an ambitious, in-depth biological and ecological study is now being implemented by the second author (RT-P). Its results will be published elsewhere.

Regarding the establishment status of the portunid species occurring in the Madeira archipelago (north of the Canary Islands, between 32 and 33°N), L. vocans, P. hastatus and P. inequalis were first recorded by Türkay (1976) based on a few specimens deposited in the Funchal Natural History Museum; they have no local common name and seem to be rare or uncommon species in that archipelago. T. poissonii and C. ruber have not yet reached the Madeiran shores (cf. Araújo and Wirtz 2015) but, at least for C. ruber, its arrival there using the Selvages islets (at 30ºN, between the Canary and Madeira Islands) as stepping stones could probably be predicted. In this prediction one must take into account that C. ruber has been reported as high as 32-33°N in both the western Atlantic and eastern Pacific Oceans.

Based on the evidence gathered herein, two hypotheses can be proposed in order to explain the natural arrival to Gran Canaria Island, or the introduction route of the present genetically proven African individuals of C. ruber.

The first theory explaining the occurrence of C. ruber in Canary Islands waters is based on a natural range extension (i.e. a natural population expansion) as a result of the high oceanographic connectivity (trade winds, marine currents and larval flow, see Introduction section) between this archipelago and the neighbour northwestern African coasts. Dispersal capabilities depend on the mobility of the decapods in question. Adults of brachyuran portunids are benthic and have proven to have moderate-high capabilities for dispersal. As the larvae are pelagic, it is assumed that dispersal takes place during the larval phase. However, C. ruber larvae should have overcome the biogeographical boundary represented by the Banc d’Arguin (Fransen 1991, Spalding et al. 2007). This area coincides with the meeting point of the southwards, cooling Canary Current and the northwards, warming Guinea Current, and these currents may play a role in determining the faunal distributions in the region; the matter still remains unresolved (Fransen 2002). However, Mittelstaedt (1983) postulated that the only natural means of arrival to the Canaries of larvae of tropical African fauna is the possible occasional displacement of African waters by certain conditions of the wind from the south when the ascending coastal current (a reverse current) flows copiously past Cape Blanc and reaches Cape Bojador, an occurrence that appears to take place during winter.

The second theory explaining the occurrence of C. ruber in the Canary Islands is a human-caused introduction of one or more phases of its life cycle. The first plausible possibility would be ship ballast mediation and dispersal during the larval phase (Wonham et al. 2000, Brito et al. 2011, Galil et al. 2011). Many small benthic marine fishes, chordate species, small-sized invertebrates and plankton (introduced as eggs, larvae or juveniles) are first recorded from regions with major commercial ports, and the method of transport associated is via the large amounts of ballast water carried by international shipping (Wonham et al. 2000, Lockett and Gomon 2001) or ship’s hull fouling (Cuesta et al. 2016).

A second possibility would be the transport of juvenile, subadult or/and adult crab specimens on oil platforms (Friedlander et al. 2014, Triay-Portella et al. 2015, Pajuelo et al. 2016). Oilrigs are large structures in the ocean where they act as artificial reefs and serve as an important habitat for a number of species (e.g. Atchison et al. 2008). These artificial habitats are colonized by diverse ecological communities, increasing the growth and survival of individuals and affording shelter for protection from predation and spawning substrate (e.g. Friedlander et al. 2014). There is a noteworthy coincidence between the establishment of oilrig maritime routes connecting the West African coast with Gran Canaria and the presumed occurrence of C. ruber on the coasts of this island. Recently, Pajuelo et al. (2016) reported the introduction of non-native marine fish species to the Canary Islands waters through oil platforms as vectors, mapping the main routes used by oilrigs arriving at Gran Canaria between 2010 and 2015. In the east Atlantic, a significant part of these voyages started from ports of Cape Verde Islands, the Ivory Coast, Cameroon, Nigeria, Congo-Brazzaville, Namibia and South Africa.

The present results from the DNA barcoding analysis clearly support an African origin for Cronius ruber from the Canary Islands. The low number of mutations in the 16S sequences among specimens of the east Atlantic, west Atlantic and east Pacific indicate that they all belong to the same species (Cronius ruber) and only minor differences, based on geographical distances, were detected. The greatest distance was found between sequences of specimens from Brazil and Panama, with only eight mutations (98%).

According to these data, pending new gene analysis, there is no support for distinguising Western African and eastern Pacific populations as Cronius milleri (A. Milne-Edwards, 1868) and C. edwardsii (Lockington, 1877), respectively.

Comparison of the 16S and COI sequences obtained herein for C. ruber from the Canaries revealed a very high proximity to C. ruber specimens from Guinea (10°27′N 14°56′W). Is this a coincidence or the result of crab introduction from the West African coast via ship ballast and/or oil platforms?